Journal of Molecular Liquids 382 (2023) 122005

Contents lists available at ScienceDirect

Journal of Molecular Liquids

journal homepage: www.elsevier.com/locate/molliq

A disparity in solvatochromism of C– – O and S–

– O vibrational probe: A
study of structurally similar acetone and dimethyl sulfoxide
Suranjana Chakrabarty, Anjan Barman *, Anup Ghosh *
Department of Condensed Matter Physics and Materials Sciences, S. N. Bose National Centre for Basic Sciences, Kolkata 700106, India

A R T I C L E I N F O A B S T R A C T

Keywords: Solvatochromic studies of small IR probes are beneficial to study the structural dynamics of biomolecules as well
Solvatochromism as measuring the strength of surrounding electric fields in condensed phases. Herein, we have investigated the
Hydrogen Bond Population C––O and S––O vibrational probes of structurally similar acetone and dimethyl sulfoxide respectively to compare
Alpha C-H/D interaction
the solvatochromism. Interestingly in water C– –O formed both one hydrogen bond and two hydrogen bond
Disparity of acetone and DMSO
complexes wherein S– –O formed only two hydrogen bond complexes. Not only this, but we have also investi­
gated how different solvents alter the IR spectra of the C– –O and S– –O vibrating modes during the interaction
with α C–H of acetone and DMSO respectively. The IR absorption spectra of C– –O of acetone altered after
α deuteration (acetoned6) in aprotic organic solvents as well as aqueous solvents. But, surprisingly for DMSO and
DMSOd6, the IR spectra of S– –O mode remain intact during α isotopic exchange (CH3 to CD3) in organic aprotic
solvents but in strong coordinating polar protic solvents (D2O) it is altered remarkably. The equivalent frequency
gap of C––O of acetone and acetoned6 in organic solvents and water represent the no interaction of α C–H/D
with solvent molecules. Parallelly unaltered IR spectral signature of S– –O of both DMSO and DMSOd6 in aprotic
organic solvent represents the stationary state of the ‘S’ atom and altered S– –O IR absorption in water signifies
the different α C–H and α C-D interaction with the surrounding water molecules. These results are the
cautionary of perturbation on IR probe during α C–H interaction with solvents as well as guest molecules.

1. Introduction molecules have been extensively investigated because of their electric

Understanding the secret of solute–solvent interactions has a large
range of fundamental importance as these interfaces initiate many nat­
ural processes. To gain a detailed understanding of solute–solvent
interaction, IR spectroscopy is significantly used. In the condensed
phase, the normal mode frequencies of a vibrational probe are signifi­
cantly affected by its intermolecular communications with molecules of
surrounding solvents. [1–3] It is well established that the intermolecular
solute–solvent interactions have an undeniable impact on the IR fre­
quencies and intensities of vibrational probes. [4–10] Vibrational fre­
quency shifting of the solute molecules towards the red or blue region in
attendance of different solvents strongly depends on polarity as well as
intermolecular hydrogen-bonding interaction with the surrounding
solvent molecules. [2] To treat such shifts in vibrational frequencies, lots
of models have already been proposed. [11–14] The normal vibrational
frequencies of a solute molecule can be interrupted by a solvent-induced
force. [11–15] In recent years lots of studies have been done to explore
solute–solvent interactions. [16–26] The C– – O probe-containing
field sensitivity and hydrogen bonding capability in different solvated
environments. [27–50] In organic molecules, as well as various bio­
molecules, the C– – O bonds are the most predominantly studied part.
[51–54] So to determine the different biomolecule structures, detailed
C–– O bond interaction identification become an important aspect in the
modern scientific era. Parallelly, having an extensive range of applica­
tions in the pharmaceutical industry and being a very good solvent/
cosolvent agent, recently similar S–– O bond containing DMSO molecule
has gained interest to explore. [55–58] Based on the IR peak assignments
of S–– O of DMSO, there are lots of vibrational studies that have been
conveyed. [59–65].
To check how C– – O and S– – O vibrational probes interact with
different aprotic and protic solvents, we have studied similar kinds of
structures, acetone and DMSO. The interesting aspect of DMSO is a
similar structure to acetone where the central ‘C’ atom is replaced by a
more polarizable ‘S’ atom. So, to explore the discrimination in solvation
as well as spectral properties of DMSO and acetone, they have been
investigated in different solvents. To recognize the detailed

* Corresponding authors at: Department of Condensed Matter Physics and Materials Sciences, S. N. Bose National Centre for Basic Sciences, Kolkata 700106, India.
E-mail addresses: anupg86@gmail.com, anup.ghosh@bose.res.in (A. Ghosh).

https://doi.org/10.1016/j.molliq.2023.122005
Received 25 November 2022; Received in revised form 25 April 2023; Accepted 1 May 2023
Available online 3 May 2023
0167-7322/© 2023 Elsevier B.V. All rights reserved.
S. Chakrabarty et al. Journal of Molecular Liquids 382 (2023) 122005

quantification of hydrogen bond interaction in water (D2O) and the
disparities in physicochemical properties of acetone/acetoned6 and
DMSO/DMSOd6 in various solvents such as DCM, ACN and D2O, we
have employed IR absorption, DFT calculations and MD simulations.
However, collective experimental, and theoretical studies offer us to
understand the hydrogen bond population disparity of acetone and
DMSO as well as the miscellaneous interaction of CH3 of acetone and
DMSO with different solvents. In this regard, our present report is ad­
ditive studies wherein we have explored inconsistent behavior of
DMSO/DMSOd6 and acetone/acetoned6 depending on the α H/D in­
teractions with different types of organic and protic solvents. Moreover,
the diversity in the interactions of DMSO and acetone with organic
solvents and water have been explored to understand how α C–H/D
interaction can affect the IR absorptions of S–
– O/C–– O stretching modes
differently. Interestingly we observed that IR spectra of S–– O of DMSO
and DMSOd6 remain unaltered in organic solvents but altered in pres­
ence of water wherein in the case of acetone and acetoned6 C– – O fre­
quency is altered in every solvent (DCM, ACN) including D2O. Overall,
this type of finding plays a prominent role to explain the details about
the implications of micro solvation environment on different vibrational
spectra, condensed phase structure, and dynamics of acetone and DMSO.
2. Materials and methods
Sample preparation. Dimethyl sulfoxide (DMSO 99.9 %), dimethyl
sulfoxide d6 (DMSOd6, 99.9 %), acetone (99.9 %), acetoned6 (99.9 %)
and all solvents (99.9 %) namely dichloromethane (DCM), acetonitrile
(ACN), and water (D2O) are bought from Sigma Aldrich and used
without further purifications. All the recorded linear IR absorption
spectra were obtained by using an FTIR spectrometer (Bruker, Tensor
70). 0.1 M solutions of DMSO/DMSOd6 and acetone/acetoned6 mole­
cules in different solvents were used during the study. Between the 2 mm
thick CaF2 windows separated by a Teflon spacer, all the mentioned
sample solutions were placed.
3. Theoretical Section
DFT Calculations. All the molecular geometries of acetone/ace­
toned6, as well as DMSO/DMSOd6 in organic solvents and their H-bond
complexes in water, were optimized by density functional theory with

Fig. 1. (a) IR absorbance of C–

–O vibrational mode of acetone in D2O, (b) Black-Experimental IR spectra of DMSO in D2O. Cyan-IR absorbance of S–
–O vibrational
mode, red- CH3 rocking mode, green, blue- two others mode and Orange-Cumulative spectra (c) Second-order derivative spectra of C– – O of acetone in D2O, (d)
Second-order derivative spectra of DMSO in D2O.

2
S. Chakrabarty et al.
the employment of the B3LYP (6-311G +(d,p)) level of theory in
Gaussian 09. [49] Herein we have used atomic Mulliken charges from
the optimized structure of acetone and DMSO in DFT calculation.
MD Simulation. MD simulations were carried out with the
CHARMM general forcefield using the GROMACS version 4.6.5 package
for acetone and DMSO in water. [66–68] All the forcefield parameters of
acetone and DMSO were taken from virtualchemistry.org, which has
been systematically tested earlier. [62–63] We have used the TIP4P
water model for the MD simulation of acetone and DMSO. [62–63] A 5
nm cubic box, filled with water molecules was used to run an MD
simulation of acetone and DMSO. The MD simulations were performed
for each minimization, using the steepest-descent algorithm followed by
the equilibrium in the NVT ensemble (1 ns) and equilibrium in NPT (1
ns) at 1 bar pressure at 300 K temperature. [62–63] Final coordinates
were recorded for 1 ns MD simulation run and used for the hydrogen
bond population of acetone and DMSO in neat water. Hydrogen bond
populations were computed using standard 3.5 A donor (D)-to-acceptor
(A) distance cutoff, and a 30◦ angle cutoff. The distance parameter was
selected to match the first minimum of the D − A radial distribution
function.
4. Results and discussion
To understand the solvation of acetone and DMSO in water, IR ab­
sorption spectra of 0.1 M of each have been collected. All the collected
IR absorption spectra of acetone and DMSO in D2O are depicted in
Fig. 1a and 1b respectively. Second derivative IR spectra have also been
calculated for a better understanding of the spectral signature (Fig. 1c,
1d). The gaussian line shape of C– – O IR absorption at 1696 cm− 1 dic­
tates either 1hb and 2hb complexes are equally populated with almost
equal extinction coefficients or they are very close to separate out.
Another possibility also may come either of 1hb or 2hb population is
negligibly small to separate. Correspondingly, for DMSO in Fig. 1b, a
gaussian shaped S– – O IR absorption spectrum at 1012 cm− 1, one CH3
rocking mode, and two other modes were obtained and the probable
preliminary explanation is similar to acetone. To solve the experimental
difficulty for hydrogen bond population calculations by IR spectra, MD
simulations have been performed in water to obtain a better represen­
tative picture of comparative H-bond populations of acetone and DMSO
and correspondingly probable explanation of IR line shapes. The MD
simulation predicts a discrepancy between C– – O and S– – O H-bond
populations of acetone and DMSO shown in Fig. 2. The C– – O mode of
acetone forms approximately a significant amount of 1hb (38 %) and
2hb (60 %) with a negligible amount of 0hb (~2%) in water. While in
Fig. 2. Bar diagram of H-bonded complexes. The H-bonded complex of acetone
through C– –O (Orange) and the H-bonded complex of DMSO through
S–
– O (Green).
3
Journal of Molecular Liquids 382 (2023) 122005
the case of DMSO, S– – O is mostly 2hb (~96 %) populated wherein the
negligible amount of 0hb (~0%) and 1hb (~3%) population is observed
(Fig. 2). So from the MD simulation, we have observed though acetone
and DMSO are similar in structure, the H-bond interaction patterns are
completely different. The mentioned H-bond population calculations
command close proximity of C– – O IR absorption of 1hb and 2hb com­
plexes which makes them very difficult to separate. But, certainly, for
DMSO, the S– – O IR absorption is mostly coming from 2hb populated
complexes.
To investigate the reason led behind this disparity in H-bond pop­
ulations, we have done multiple DFT calculations to compute the for­
mation energy of 1hb and 2hb complexes of acetone and DMSO
simultaneously. Due to the similar effect of water on acetone and DMSO,
we have ignored the effect of solvent dynamics in DFT calculations. The
calculated formation energy of the 2hb complex is − 22kj/mol and the
1hb complex is − 11.5 kJ/mol of acetone. Wherein for DMSO, they are
about − 42 kJ/mol and − 21 kJ/mol respectively. The graphical repre­
sentation in Fig. 3, clearly shows the corresponding formation energies
and the energy gaps between 1hb and 2hb for both acetone and DMSO.
Due to a small stabilization energy gap (Δ9.5 kJ/mol), 1hb and 2hb
conformation of acetone can be converted on either side very easily. On
contrary, due to the high stabilization energy of 2hb and high energy gap
(Δ21.5 kJ/mol) between 1hb and 2hb of DMSO, they can’t convert
either side of the equilibrium easily (Fig. 3). The equilibrium is mostly
shifted towards the 2hb complex side. Now to establish the reason of the
huge discrepancy in stabilization energies of 2hb and 1hb complexes, we
have looked on the atomic charge distribution of acetone and DMSO
(Fig. 4). The δ- charge on the ‘O’ atom (-0.871) of DMSO is much higher
than the ‘O’ atom (-0.517) of acetone. The disparity in charge distri­
bution over the ‘O’ atom is due to the higher polarizability of the ‘S’
atom than the ‘C’ atom of DMSO and acetone respectively. So the
chances of hydrogen bond formation with S– – O are higher for DMSO
than C– – O of acetone and the coulombic attraction probability of H
atoms of water is also higher towards S– – O than C–– O mode. As the
solvent is the same for both cases, the dynamical effect on acetone and
DMSO are presumably similar. But surprisingly we have different
hydrogen bond populations. So even though solvent dynamics may have
some effect on hydrogen bond population, the intrinsic properties of the
solute play a key role here.
We have also done a comparative study of α C–H interactions of
acetone and DMSO with water. For such investigation, we have
considered acetone and isotope substituted acetoned6 as well as DMSO
and DMSOd6. To resolve the detailed molecular interaction strategy of
acetone/acetoned6 and DMSO/DMSOd6 in different solvents a series of
infrared absorption spectra of 0.1 M acetone and acetoned6 in different
solvents such as DCM, ACN and D2O have been collected (Fig. 5).
Similarly, for DMSO and DMSOd6, we have also taken a series of IR
spectra in the same solvents (Fig. 6). We have deconvoluted the IR ab­
sorption spectra of DMSO and DMSOd6 after second order derivative
spectra calculations (See Figure S1 and S2). Based on the second-order
derivative spectra in the region from 950 to 1100 cm− 1 we have
clearly assigned the individual fitted peak (S– – O symmetric stretching
modes-red, other modes- green, blue- and CH3/CD3 rocking modes-
magenta) of DMSO/DMSOd6 (Section S1). All the Experimental fre­
quencies of acetone and acetoned6 as well as DMSO/DMSOd6 are sum­
marized in Table 1. The constant frequency gap (10 cm− 1) of C– – O IR
absorption spectra of acetone and acetoned6 are observed in each sol­
vent, shown in Fig. 5. Wherein, the S– – O IR absorption of DMSO and
DMSOd6 remain unaltered in each organic solvent (DCM, ACN,) but
altered in D2O, depicted in Fig. 6 and Table 1.
Parallelly, we have further evaluated the IR absorption frequency of
acetone/acetoned6 and DMSO/DMSOd6 in different organic solvents as
well as in D2O by DFT calculations. All the calculated frequencies and
the respective frequency gaps of C– – O of acetone/acetoned6 and S–– O of
DMSO/DMSOd6 in different solvents are shortened in Table 2. Where
the C–– O IR absorption frequency gap is almost constant for acetone and
S. Chakrabarty et al.
Fig. 3. Stabilisation Energy of water hydrogen-bonded complexes of acetone (blue) and DMSO (red) in water.
Fig. 4. Charge distribution of acetone and DMSO.
acetoned6 in organic as well as protic solvents (D2O). Again S– – O IR
absorption of DMSO and DMSOd6 are comparable in organic solvents
but significantly altered by 23 cm− 1 in D2O (Table 2). The calculated
results exactly matched with experimental observations which drive us
to find out the reason of the disparity in IR absorption of C–
– O and S–
–O
of acetone/acetoned6 and DMSO /DMSOd6 respectively. We have seen
from DFT calculations that weak vibrational coupling exists between the
modes in DMSO. However, as weak coupling cannot explain the large
frequency shift, we have not considered the coupling.
In the case of acetone and acetoned6, both the atomic mass of the ‘C’
and ‘O atoms are comparable. So, C and O mutually vibrate simulta­
neously during C– – O stretching vibration. As the relative mass of the
central ‘C’ atom (attached with the heavier CD3 group) is more in ace­
toned6 than acetone, C– – O IR absorption shifted towards the lower
frequency region. Whereas for DMSO and DMSOd6, the heavy ‘S’ atom
of S–– O mode remains in the stationary state and mostly the ‘O’ atom
vibrates significantly. So, during S– – O stretching vibration, the addi­
tional perturbation from the CH3/CD3 mass is almost negligible in
different organic solvents. But on the contrary, despite the structurally
and chemically similar DMSO and DMSOd6, S– – O IR absorption is
altered remarkably in D2O (Fig. 6). These could be the possibility for
different kinds of H-bond interaction of α C–H/D of DMSO and DMSOd6
with water. Herein, the high polarizability of the S–– O bond makes α H/
D acidic and suitable to form H-bond with D2O.
4
Journal of Molecular Liquids 382 (2023) 122005
It is well known that the amplitude of the C – H vibration is larger
than C − D, which leads to the strengthening of the H-bond bridge of C
− D—OD2 than C – H—OD2. The higher mass of the − CD3 group ro­
tates slower over the C–S bond in DMSOd6, wherein the − CH3 group
rotates faster over the C–S bond, which makes the C – H—OD2 H-bond
weaker than the C − D—OD2 H-bond. We have also done DFT calcula­
tions to study the H bond formation energy of α-H/D of DMSO/DMSOd6.
For DMSOd6 it is about − 7.6310 kJ/mol; for DMSO it is − 7.0798 kJ/mol
which is significantly smaller than DMSOd6. So definitely the α-D H-
bond is stronger in DMSOd6 than the α-H H-bond of DMSO. So the
delocalization of the bond over the S–– O mode of DMSOd6 is more than
DMSO. Consequently, bond order decreases more in S– – O of DMSOd6
than DMSO. Subsequently, we have more red-shifted S– – O IR absorption
peaks in DMSOd6 than in DMSO.
On the other hand, the constant frequency gap of IR absorption
spectra of the C–– O mode of acetone and acetoned6 in organic solvents
and water also suggests that α C–H/D negligibly interacts with water
(Table 1 and Table 2). Also, the carbonyl ‘C’ atom of acetone and ace­
toned6, is not polarizable, and the acidity of α H/D is almost negligible.
Hence, D2O rarely interacts with α C–H/D of acetone/acetoned6. The
calculated CH3 rocking mode of acetone suggests the non-interaction of
α H wherein alteration of the mode confirms the α H interaction of
DMSO with D2O (Section S1). These could be the feasible reason for
dissimilar vibrational C–– O and S– – O IR absorption peaks for both
S. Chakrabarty et al. Journal of Molecular Liquids 382 (2023) 122005

Fig. 5. Experimental IR absorption spectra (highest intensity point normalized) of C–

–O mode of acetone ((a) in DCM (b) ACN, (c) D2O) and acetoned6 ((d) in DCM
(e) ACN, (f) D2O). The red line is drawn to show the frequency gap between acetone and acetoned6.

Fig. 6. The black spectra denote the experimental IR absorption spectra (highest intensity point normalized) of DMSO ((a) DCM, (b) ACN and (c) D2O) and DMSOd6
((d) DCM, (e) ACN and (f) D2O). All the spectra are fitted with Voigt line shape functions. S–
– O stretching spectra (red), magenta peaks signify CH3 and CD3 rocking
modes and the green, blue represent other modes of vibrations. The cyan color denotes cumulative spectra.

Table 1 Table 2
Experimental IR absorption peak positions of acetone/acetoned6 and DMSO/ Calculated IR absorption frequencies of acetone/acetoned6 and DMSO/DMSOd6
DMSOd6 in different solvents. in different solvents by employing the MP2/cc-pVDZ level of theory.
Solvents νC¼O of acetone (cm¡1) νC¼O of acetoned6 (cm¡1) Δν (cm¡1) Solvents νC¼O of acetone νC¼O of acetoned6 ΔνC¼O
(cm¡1) (cm¡1) (cm¡1)
DCM 1713.3 1702.8 ~10
ACN 1712.5 1703.0 ~10 DCM 1753.8 1744.8 9.0
D2O 1692.0 1682.5 ~10 ACN 1747.0 1738.0 9.0
Solvents νS¼O of DMSO (cm¡1) νS¼O of DMSOd6 (cm¡1) Δν (cm¡1) D2O 1746.0 1736.8 9.2
DCM 1059.0 1058.0 ~1 Solvents νS¼O of DMSO νS¼O of DMSOd6 (cm¡1) ΔνS¼O
ACN 1061.0 1060.0 ~1 (cm¡1) (cm¡1)
D2O 1012.0 989.0 ~23 DCM 1059.5 1061.3 1.8
ACN 1055.1 1058.1 3.0
D2O 1021.0 1001.0 20.0

5
S. Chakrabarty et al.
acetone/acetoned6 and DMSO/DMSOd6 in a distinct organic solvent and
water.
5. Conclusions
In this work, we inspected the disparity of solvatochromism of
acetone and DMSO through IR absorption spectroscopy, DFT calcula­
tions and MD simulation. Even though DMSO and acetone have equiv­
alent chemical structures except for central sulphur and carbon atom,
they showed significant discrepancies in hydrogen bond populations
and IR absorptions. During hydrogen bond formation dissimilar per­
centages of the population have been obtained for acetone (1hb-40 %
and 2hb-60 %) and DMSO (1hb-3 % and 2hb-96 %). On the other aspect,
dissimilarity was also observed in the IR absorption spectra of S–
– O and
C– O in different organic and protic solvents. We showed that the iso­
–
topic exchange of α C–H to α C-D, the IR absorption of the S–– O mode of
DMSO and DMSOd6 remains unaltered in organic solvents but altered in
aqueous solvent. Wherein the C– – O frequency of acetone and acetoned6
significantly altered in every solvent including D2O. Herein we have
exposed that α C–H/D of DMSO/DMSOd6 and acetone/acetoned6
interact differently in presence of different organic aprotic and protic
solvents and their perturbation on S– – O and C–– O modes IR absorption
frequency is significantly different. The α C H/D of acetone and ace­
– [20] A. Idrissi, F. Sokolic, A. Perera, A molecular dynamics study of the urea/water
toned6 couldn’t form hydrogen bonds with D2O whereas in the case of
DMSO/DMSOd6, they formed hydrogen bonds which induce a signifi­
cant shift in the S–
– O IR absorption frequency. In conclusion, our results
delivered an outline of the contradiction between DMSO and acetone
which are structurally comparable compounds. We hope that in the
future, our findings will help in solvatochromic studies of biomolecules
such as in DMSO and acetone in the arena of IR spectroscopy.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgments
A.G thanks the DST- India for financial support and SNBNCBS, Kol­
kata for instrumental facilities.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.molliq.2023.122005.
References
[1] W. Liptay, Electrochromism and Solvatochromism, Angew. Chem. internat. Edit. 8
(3) (1969) 177–188.
[2] M. Cho, Vibrational solvatochromism and electrochromism: Coarse-grained models
and their relationships, J. Chem. Phys. 130 (9) (2009), 094505.
[3] B.A. Lindquist, K.E. Furse, S.A. Corcelli, Nitrile groups as vibrational probes of
biomolecular structure and dynamics: an overview, Phys. Chem. Chem. Phys. 11
(32) (2009) 8119–8132.
[4] K.S. Schwiezer, D. Chandler, Vibrational dephasing and frequency shifts of
polyatomic molecules in solution, J. Chem. Phys. 76 (5) (1982) 2296–2314.
[5] C.M. Huggins, G.C. Pimentel, Infrared Intensity of the C — D Stretch of
Chloroform-d in Various Solvents, J. Chem. Phys. 23 (5) (1955) 895–898.
[6] C.H. Veas, J.L. McHale, Intermolecular resonance coupling of solute and solvent
vibrational modes, J. Am. Chem. Soc. 111 (18) (1989) 7042–7046.
[7] R.C. Lord, B. Nolin, H.D. Stidham, Quantitative Study of the Bonding of
Chloroform-d in Various Solvents by Infrared Spectrometry, J. Am. Chem. Soc. 77
(5) (1955) 1365–1368.
6
Journal of Molecular Liquids 382 (2023) 122005
[8] D.L. Akins, Theory of Raman scattering by aggregated molecules, J. Phys. Chem. 90
(8) (1986) 1530–1534.
[9] J.L. McHale, Intermolecular vibrational resonance coupling: Intensity borrowing in
polarized Raman spectroscopy, Int. J. Quantum Chem. Symp. 40 (S25) (1991)
593–602.
[10] A. Mortensen, O.F. Nielsen, J. Yarwood, V. Shelley, Anomalies in the Isotropic
Raman Spectra of Liquid Mixtures of Isotopomers of Formamide. Intermolecular
Interactions in the Carbonyl Stretching Region, J. Phys. Chem. 99 (13) (1995)
4435–4440.
[11] A.M. Benson Jr., H.G. Drickamer, Stretching Vibrations in Condensed Systems:
Especially Bonds Containing Hydrogen, J. Chem. Phys. 27 (5) (1957) 1164–1174.
[12] A.D.E. Pullin, The variation of infra-red vibration frequencies with solvent,
Spectrochim. Acta 13 (1–2) (1958) 125–138.
[13] A.D. Buckingham, Solvent effects in vibrational spectroscopy, Trans. Faraday Soc.
56 (1960) 753–760.
[14] F.G. Dijkman, J.H. Van der Maas, Inhomogeneous broadening of Morse oscillators
in liquids, J. Chem. Phys. 66 (9) (1977) 3871–3878.
[15] H. Wang, L. Wang, S. Shen, W. Zhang, M. Li, L. Du, X. Zheng, D.L. Phillips, Effects
of hydrogen bond and solvent polarity on the C=O stretching of bis(2-thienyl)
ketone in solution, J. Chem. Phys. 136 (12) (2012), 124509.
[16] J.L. Finney, D.T. Bowron, A.K. Soper, The structure of aqueous solutions of tertiary
butanol, J. Phys. Condens. Matter. 12 (8A) (2000) A123.
[17] D.T. Bowron, J.L. Finney, A.K. Soper, Structural Investigation of Solute− Solute
Interactions in Aqueous Solutions of Tertiary Butanol, J. Phys. Chem. B 102 (18)
(1998) 3551–3563.
[18] J. Turner, A.K. Soper, J.L. Finney, Water structure in aqueous solutions of
tetramethylammonium chloride, Mol. Phys. 77 (3) (1992) 411–429.
[19] S.E. McLain, A.K. Soper, A. Luzar, Orientational correlations in liquid acetone and
dimethyl sulfoxide: A comparative study, J. Chem. Phys. 124 (7) (2006), 074502.
mixture, J. Chem. Phys. 112 (21) (2000) 9479–9488.
[21] A. Idrissi, S. Longelin, F. Sokolic, Study of Aqueous Acetone Solution at Various
Concentrations: Low-Frequency Raman and Molecular Dynamics Simulations,
J. Phys. Chem. B 105 (25) (2001) 6004–6009.
[22] J.J. Max, C. Chapados, Infrared spectroscopy of acetone–water liquid mixtures. II.
Molecular model, J. Chem. Phys. 120 (14) (2004) 6625–6641.
[23] Y. Takebayashi, S. Yoda, T. Sugeta, K. Otake, T. Sako, M. Nakahara, Acetone
hydration in supercritical water: 13 C-NMR spectroscopy and Monte Carlo
simulation, J. Chem. Phys. 120 (13) (2004) 6100–6110.
[24] D.S. Venables, C.A. Schmuttenmaer, Spectroscopy and dynamics of mixtures of
water with acetone, acetonitrile, and methanol, J. Chem. Phys. 113 (24) (2000)
11222–11236.
[25] P.E. Mason, G.W. Neilson, J.E. Enderby, M.L. Saboungi, E.D. Christopher, A.
D. MacKerell, J.W. Brady, The Structure of Aqueous Guanidinium Chloride
Solutions, J. Am. Chem. Soc. 126 (37) (2004) 11462–11470.
[26] A.K. Soper, A. Luzar, neutron diffraction study of dimethyl sulphoxide–water
mixtures, J. Chem. Phys. 97 (2) (1992) 1320–1331.
[27] W.K. Orland, Effect of Hydrogen Bonding Solvents on the Infrared Absorption Band
for the Fundamental Vibration of the Carbonyl Group in 1,1,3,3-Tetramethylurea,
Trans. Kans. Acad. Sci. 102 (1999) 53–56.
[28] L.C.J. Almeida, P.S. Santos, The anomalous solvent effect in the vibrational
spectrum of 2,3-diphenyl-cycloprop-2-enone: an experimental and theoretical
investigation, Spectrochim. Acta: A 58 (14) (2002) 3139–3148.
[29] R.A. Nyquist, H.A. Fouchea, G.A. Hoffman, D.L. Hasha, Infrared Study of β
-Propiolactone in Various Solvent Systems and Other Lactones, Appl. Spectrosc. 45
(1991) 860–867.
[30] R.A. Nyquist, C.L. Putzig, D.L. Hasha, Solvent Effect Correlations for Acetone: IR
versus NMR Data for the Carbonyl Group, Appl. Spectrosc. 43 (6) (1989)
1049–1053.
[31] R.A. Nyquist, C.L. Putzig, L. Yurga, Solvent and Concentration Effects on Carbonyl
Stretching Frequencies: Ketones, Appl. Spectrosc. 43 (6) (1989) 983–991.
[32] R.A. Nyquist, T.D. Clark, R. Streck, Infrared study of alkyl carboxylic acids in CCl4
and/or CHCl3 solutions, Vib. Spectrosc. 7 (3) (1994) 275.
[33] P. Bruni, E. Giorginy, E. Maurelli, G. Tosi, Fourier transform infrared spectrometry
investigation of solvent effect on NH and CO stretching modes in N-
acylaminopyridines, Vib. Spectrosc. 12 (2) (1996) 249–255.
[34] R.A. Nyquist, C.L. Putzig, T.D. Clark, Infrared study of 1,3-dimethyl-2-imidazoli­
dinone in various solvents, Vib. Spectrosc. 12 (1) (1996) 81–91.
[35] J.B.F.N. Engberts, G.R. Famini, A. Perjessy, L.Y.J. Wilson, Solvent effects on C=O
stretching frequencies of some 1-substituted 2-pyrrolidinones, Phys. Org. Chem. 11
(4) (1998) 261–272.
[36] A. Ghosh, B. Cohn, A.K. Prasad, L. Chuntonov, Quantifying conformations of ester
vibrational probes with hydrogen-bond-induced Fermi resonances, J. Chem. Phys.
149 (18) (2018) 184501–184511.
[37] A. Ghosh, Vibrational Coupling on Stepwise Hydrogen Bond Formation of Amide I,
J. Phys. Chem. B 123 (37) (2019) 7771–7776.
[38] K.I. Oh, K. Rajesh, J.F. Stanton, C.R. Baiz, Quantifying hydrogen-bond populations
in DMSO/water mixtures, Angew. Chem. Int. Ed. 56 (38) (2016) 11375–11379.
[39] K.I. Oh, C.R. Baiz, Crowding Stabilizes DMSO− Water Hydrogen-Bonding
Interactions, J. Phys. Chem. B 122 (22) (2018) 5984–5990.
[40] S.M. Kashid, G.Y. Jin, S. Chakrabarty, S. Bagchi, Y.S. Kim, Two-Dimensional
Infrared Spectroscopy Reveals Cosolvent Composition-Dependent Crossover in
Intermolecular Hydrogen Bond, J. Phys. Chem. Lett. 8 (7) (2017) 1604–1609.
[41] S.M. Kashid, G.Y. Jin, S. Bagchi, Y.S. Kim, Cosolvent Effects on Solute− Solvent
Hydrogen-Bond Dynamics: Ultrafast 2D IR Investigations, J. Phys. Chem. B 119
(49) (2015) 15334–15343.
S. Chakrabarty et al.
[42] I.A. Borin, M.S. Skaf, Molecular Association between Water and Dimethyl Sulfoxide
in Solution: The Librational Dynamics of Water, Chem. Phys. Lett. 296 (1–2)
(1998) 125–130.
[43] A. Luzar, D. Chandler, Structure and Hydrogen-Bond Dynamics of Water-Dimethyl
Sulfoxide Mixtures by ComputerSimulations, J. Chem. Phys. 98 (10) (1993)
8160–8173.
[44] S.M. Kashid, R.K. Singh, H. Kwon, Y.S. Kim, A. Mukherjee, S. Bagchi, Arresting an
Unusual Amide Tautomer Using Divalent Cations, J. Phys. Chem. B 123 (40)
(2019) 8419–8424.
[45] A. Wulf, R. Ludwig, Structure and Dynamics of Water Confined in Dimethyl
Sulfoxide, Chem Phys Chem. 7 (1) (2006) 266–272.
[46] R. Ludwig, T.C. Farrar, M.D. Zeidler, Temperature Dependence of the Deuteron and
Oxygen Quadrupole Coupling Constants of Water in the System Water/Dimethyl
Sulfoxide, J. Phys. Chem. D 98 (1994) 6684–6687.
[47] S.D. Fried, S. Bagchi, S.G. Boxer, Measuring Electrostatic Fields in Both Hydrogen-
Bonding and Non-Hydrogen-Bonding Environments Using Carbonyl Vibrational
Probes, J. Am. Chem. Soc. 135 (30) (2013) 11181–11192.
[48] S.G. Boxer, S. Bagchi, S.D. Fried, N. Levinson, M. Saggu, L. Xu, Vibrational Stark
Spectroscopy Directly Probes Electric Fields in Proteins, Biophys. J . 104 (2) (2013)
355A.
[49] M.S. Kashid, S. Bagchi, Experimental Determination of the Electrostatic Nature of
Carbonyl Hydrogen-Bonding Interactions Using IR-NMR Correlations, J. Phys.
Chem. Lett. 5 (18) (2014) 3211–3215.
[50] S. Chakrabarty, A. Ghosh, Inconsistent hydrogen bond-mediated vibrational
coupling of amide I, RSC Adv. 13 (2023) 1295.
[51] A. Greenberg, C.M. Breneman, J.F. Liebman, The Amide Linkage: Structural
Significance, Chemistry, Biochemistry and Material Science, Wiley, New York, NY,
USA, 2000.
[52] L. Brunton, B. Chabner, B. Knollman, Goodman and Gilman’s the Pharmacological
Basis of Therapeutics, MacGraw-Hill, New York, NY, USA, 2010.
[53] V.R. Pattabiraman, J.W. Bode, Rethinking amide bond synthesis, Nature 480
(2011) 471–479.
[54] A.B. Hughes, Amino Acids, Peptides and Proteins in Organic Chemistry, Wiley-
VCH, Weinheim, Germany, 2009.
7
Journal of Molecular Liquids 382 (2023) 122005
[55] N.C. Santos, J.F. Coelho, J.M. Silva, C. Saldanha, Multidisciplinary utilization of
dimethyl sulfoxide: pharmacological, cellular, and molecular aspects, Biochem
Pharmacol. 65 (7) (2003) 1035–1041.
[56] C.D. Wood, J. Wood, Pharmacologic and Biochemical considerations of Dimethyl
Sulfoxide, Ann N Y Acad Sci. 243 (1) (1975) 7–19.
[57] W.R. Fawcett, A.A. Kloss, Attenuated total reflection fourier-transform IR
spectroscopic study of dimethyl sulfoxide self-association in acetonitrile solutions,
J. Chem. Soc. Faraday Trans. 92 (18) (1996) 3333–3337.
[58] W.J. Stanley, H. Robert, Pharmacology of DMSO, Cryobiology 23 (1) (1986)
14–27.
[59] R.H. Figueroa, E. Roig, H.H. Szmant, Infrared study on the self-association of
dimethyl sulfoxide, Spectrochem. Acta. 22 (4) (1966) 587–592.
[60] T. Clark, J.S. Murray, P. Lane, P. Politzer, Why are dimethyl sulfoxide and dimethyl
sulfone such good solvents? J Mol Model. 14 (8) (2008) 689–697.
[61] S. Singh, S.K. Srivastava, D.K. Singh, Raman scattering and DFT calculations used
for analyzing the structural features of DMSO in water and methanol, RSC Adv. 3
(13) (2013) 4381–4390.
[62] M.I.S. Sastry, S.J. Singh, Self-association of dimethyl sulphoxide and its dipolar
interactions with water: Raman spectral studies, Raman Spectrosc. 15 (2) (1984)
80–85.
[63] S.A. Kirillov, M.I. Gorobets, M.M. Gafurov, K.S. Rabadanov, M.B. Ataev,
Temperature dependence of associative equilibria of DMSO according to raman
scattering spectra, Russ. J. Phys. Chem A 88 (1) (2013) 175–177.
[64] S. Chakrabarty, H.S. Deshmukh, A. Barman, S. Bagchi, A. Ghosh, On− Off Infrared
Absorption of the S=O Vibrational Probe of Dimethyl Sulfoxide, J. Phys. Chem. B
126 (24) (2022) 4501–4508.
[65] S. Chakrabarty, A. Barman, A. Ghosh, Anomalous Infrared Absorbance of S=O: A
Perturbation Study of α-C–H/D, J. Phys. Chem. B 126 (29) (2022) 5490–5496.
[66] P.J. Stephens, F.J. Devlin, C.F. Chabalowski, M.J. Frisch, Ab Initio Calculation of
Vibrational Absorption and Circular Dichroism Spectra Using Density Functional
Force Fields, J. Phys. Chem. A 98 (45) (1994) 11623–11627.
[67] A.D. Becke, Density-functional thermochemistry. III. The role of exact exchange,
J. Chem. Phys. 98 (7) (1993) 5648–5652.
[68] S. Chakrabarty, S. Maity, D. Yazhini, A. Ghosh, Surface-Directed Disparity in Self-
Assembled Structures of Small-Peptide L-Glutathione on Gold and Silver
Nanoparticles, Langmuir 36 (38) (2020) 11255–11261.
 RSC Advances
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

View Article Online

PAPER View Journal | View Issue

Inconsistent hydrogen bond-mediated vibrational

Cite this: RSC Adv., 2023, 13, 1295
coupling of amide I†
Open Access Article. Published on 05 January 2023. Downloaded on 1/6/2023 5:20:04 AM.

Suranjana Chakrabarty and Anup Ghosh *

Received 12th November 2022
Accepted 16th December 2022
DOI: 10.1039/d2ra07177k
rsc.li/rsc-advances
Using infrared spectroscopy and density functional theory (DFT) calculations, we scrutinized an amide
(dimethylformamide) as a “model” compound to interpret the interactions of amide 1 with different
phenol derivatives (para-chlorophenol (PCP) and para-cresol (CP)) as “model guest molecules”. We
established the involvement of amide I in vibrational coupling with symmetric and asymmetric C]C
modes of different phenolic derivatives and how their coupling was dependent upon different guest
aromatic phenolic compounds. Interestingly, substitution of phenol perturbed the pattern of vibrational
coupling with amide I. The symmetric and asymmetric C]C modes of PC were coupled significantly
with amide 1. For PCP, the symmetric C]C mode coupled significantly, but the asymmetric mode
coupled negligibly, with amide I. Here, we reveal the nature of vibrational coupling based on the
structure of a guest molecule hydrogen-bonded with amide I. Our conclusions could be valuable for
depiction of the unusual dynamics of coupled amide-I modes as well as the dependency of vibrational
coupling on altered factors.

bond is present in many organic molecules and

Introduction
Hydrogen bonds are pervasive in protein molecules, and are
involved in biological processes such as molecular association,
catalysis, and signal transmission.1–15 How biologically active,
small organic molecules interact with other molecules to elicit
different biological effects is an important research area.16–20
However, identifying the vibrational modes of biologically
active molecules (e.g., proteins) is difficult. To overcome such
difficulties, the structural and environmental properties of
biomolecules have been investigated using “vibrational
probes”.21–26
Many biomolecules contain “amide I”, “amide II”, “amide
III”, and “amide A” modes of vibration. However, the amide-I
mode is studied widely as a vibrational probe.27,28 The amide
a, Department of Condensed Matter of Physics and Materials Sciences, S. N. Bose
National Centre for Basic Sciences, JD Block, Sector-III, Salt Lake City, Kolkata –
700 106, India. E-mail: anupg86@gmail.com; anup.ghosh@bose.res.in
† Electronic supplementary information (ESI) available: We have provided videos
showing coupling of the: symmetric C]C and C]O modes of PC and DMF (AV1);
asymmetric C]C and C]O modes of PC and DMF (AV2); symmetric C]C and
C]O modes of PCP and DMF (AV3); asymmetric C]C and C]O modes of PCP
and DMF (AV4); symmetric C]C and C]O modes of para-ethylphenol and
DMF (AV5); asymmetric C]C and C]O modes of para-ethylphenol and DMF
(AV6); symmetric C]C and C]O modes of para-nitrophenol and DMF (AV7);
asymmetric C]C and C]O modes of para-nitrophenol and DMF (AV8);
symmetric C]C and C]O modes of meta-chlorophenol and DMF (AV9);
asymmetric C]C and C]O modes of meta-chlorophenol and DMF (AV10);
symmetric C]C and C]O modes of ortho-chlorophenol and DMF (AV11);
asymmetric C]C and C]O modes of ortho-chlorophenol and DMF (AV12). See
DOI: https://doi.org/10.1039/d2ra07177k
biomolecules.29–32 Most importantly, as an infrared (IR) probe,
amide I is employed extensively because of its sensitivity to the
native electric eld, solvation, and large molar extinction
coefficient.33–38 In particular, vibrational spectroscopic
measurements of the amide-I band are used to monitor shis in
the transition frequency, which is sensitive to the local electric
elds as well as interactions with specic “guest molecules”.24–27
Many studies have focused on the relationship between vibra-
tional couplings and conformational dynamics of proteins/
peptides.34–41 Various theoretical methodologies and multidi-
mensional IR-spectroscopy methods have been employed to
investigate the vibrational coupling and structural details of
biological systems.42–49 Vibrational coupling and the interac-
tions between different vibrational modes have been investi-
gated.50 The vibrational coupling between hydrogen bonds
associated with amide-A and amide-I/II modes within the same
amide component for several dipeptides has been studied using
two-dimensional IR spectroscopy.51 The hydrogen bonding
between amide I and phenol derivatives, dimethylformamide
(DMF), and dimethyl acetamide has been considered.52–54
Investigation of the amide-I vibrational mode is very complex
because it is delocalized in biomolecules. However, to study the
molecular perceptions and sensitivity of the amide-I mode in
the presence of intermolecular hydrogen bonds, we used DMF
as a “model” molecule. We measured the IR absorbance of the
C]C mode involved in vibrational coupling during intermo-
lecular hydrogen bonding with amide I. Correlations between
the hydrogen bond-induced vibrational coupling of C]C and
C]O transitions with different factors were investigated by

© 2023 The Author(s). Published by the Royal Society of Chemistry RSC Adv., 2023, 13, 1295–1300 | 1295

RSC Advances
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
Open Access Article. Published on 05 January 2023. Downloaded on 1/6/2023 5:20:04 AM.
Scheme 1 Chemical structure of the phenol derivatives para-cresol
(PC) and para-chlorophenol (PCP).
employing linear IR spectroscopy. We revealed how pervasive
formation of hydrogen bonds in the presence of phenolic
compounds (hydrogen-bond contributors) could disturb the
amide-I transition and symmetric/asymmetric C]C transition
of guest molecules. Hydrogen-bond formation as well as the
dependency of vibrational coupling upon different orientations
between coupled modes were also investigated in our work.
We employed linear IR spectroscopy and density functional
theory (DFT) calculations as theoretical approaches. The
frequency gap between symmetric and asymmetric C]C
stretching of phenol derivatives and the C]O vibrational mode
of DMF, as well as the enhancement factor in IR absorption
during vibrational coupling, were monitored in the presence of
different donor molecules. Vibrational coupling in biomole-
cules is important to understand the many biological interac-
tions and processes at the microscopic level, so the coupling of
amide I must be investigated. Overall, this structural evidence
of vibrational coupling can be used to elucidate many biological
and chemical effects.
Experimental section
Para-chlorophenol (PCP; 99.9% purity), para-cresol (PC; 99.9%
purity), and DMF (99.9% purity) were purchased from Milli-
poreSigma and used without additional purication. The
Fig. 1 Linear IR spectra of amide I (dimethylformamide) in the presence of different concentrations of (a) para-cresol and (b) para-chlorophenol.
The colour of the spectra is described in the inset with different concentrations of para-cresol and para-chlorophenol. The green (up) and red
arrow (down) represent enhancement and suppression of absorbance, respectively.
1296 | RSC Adv., 2023, 13, 1295–1300
View Article Online
Paper
chemical structure of PCP and PC are drawn in Scheme 1. A
solution of DMF (0.1 M) in carbon tetrachloride (CCl4) was used
for linear IR spectroscopy. The sample was placed in a home-
made Fourier transform infrared (FTIR) sample cell with CaF2
windows and a Teon™ spacer (60 mm). Linear IR absorption
spectroscopy was undertaken using an FTIR spectrometer
(JASCO-FTIR-6300). The background of the solvent (CCl4) was
measured and subtracted from all spectra of interactions
between DMF and phenol derivatives. The Beer–Lambert law
was validated by plotting the area of IR absorbance for the C]C
mode vs. concentration (Fig. S1, ESI†).
Theoretical section
We wished to gain detailed knowledge about the IR absorption
spectra of the C]O mode in DMF and C]C mode of different
phenol substitutions, so we undertook DFT calculations
employing Gaussian 09. In a preliminary manner, all the initial
geometries of DMF and different phenolic complexes were
optimized by the B3LYP/6-311G+ (D, P) level of theory. Then,
calculations to determine the frequency of IR absorption were
done for all DMF–phenol hydrogen-bonded complexes.
Results and discussion
A series of linear IR spectra of DMF solution (0.1 M) in CCl4 were
taken with increasing concentrations of PC and PCP from 0 M to
0.05 M (Fig. 1). Preliminarily, the IR absorption frequency of the
C]O mode was shown to be 1686 cm−1 (black single peak) in
the absence of phenolic compounds (0.00 M). With gradual
addition of PC or PCP, the IR absorbance of amide 1 decreased
progressively and the frequency shied towards a lower-
wavenumber region (Fig. 1a and b, respectively). Fig. 1 reveals
that increasing the concentration of PC and PCP led to
hydrogen-bond formation of C]O and gradual shiing of the
peak position of IR absorption. However, in PC (0.1 M) and PCP
(0.1 M), with a gradual increase in the DMF concentration, the
IR absorption spectra for symmetric and asymmetric C]C
modes showed anomalous behaviours (Fig. 2). For PC, the IR
© 2023 The Author(s). Published by the Royal Society of Chemistry
 View Article Online

Paper RSC Advances

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
Open Access Article. Published on 05 January 2023. Downloaded on 1/6/2023 5:20:04 AM.

Fig. 2 Linear IR absorption of the C]C mode of (a) para-cresol (0.1 M) and (b) para-chlorophenol (0.1 M) in the presence of dimethylformamide
at 0.00 M (black), 0.02 M (red), 0.04 M (blue), 0.06 M (green), 0.08 M (violet), and 0.10 M (dark-yellow). The green arrow represents enhancement
of absorbance. The red cross represents no enhancement of absorbance.

absorption peaks for symmetric and asymmetric C]C modes
were at 1597.5 cm−1 and 1618.5 cm−1 (Fig. 2a) whereas, for PCP
they were at 1594.0 cm−1 and 1606.5 cm−1, respectively (Fig. 2b).
For PC and PCP, with an increasing concentration of DMF
(0.000–0.10 M), a signicant difference in IR absorbance was
observed between symmetric and asymmetric C]C modes
(Fig. 2). For PC, with a gradual increase in the DMF concen-
tration from 0.000 M to 0.10 M, IR absorbance for symmetric
and asymmetric C]C modes was enhanced signicantly.
However, in contrast with PCP, though IR absorbance for
symmetric C]C stretching was enhanced, IR absorbance for
asymmetric C]C stretching was altered negligibly throughout
the experiment (Fig. 2). The IR-absorbance enhancement ratio
for the symmetric and asymmetric C]C modes of PC was 1.35
and 1.40 whereas, for PCP, it was 5.97 and 1.00, respectively
(Table 1), as calculated from Fig. 2. The transition dipole
moment of the symmetric and asymmetric modes of PC and
PCP changed accordingly (Table 1). These unusual phenomena
focused our attention on the intermolecular interactions
between amide (C]O) and phenolic compounds (PC and PCP).
The increase in IR absorbance of the C]C mode could have
been due to the altered hydrogen bonding with vibrational
modes or n–p* bonding of PC and the PCP ring with the C]O
mode of DMF (Fig. 2a and b). To elucidate the precise reason
underpinning the enhancement, we undertook IR spectroscopy
of a 1 : 1 DMF : anisole mixture. The unaltered enhancement of
IR absorbance for the C]C mode of the DMF : anisole (1 : 1)
mixture signied no interaction of the non-bonded electron of
the oxygen atom of DMF with the p electron cloud of anisole.
We did not know whether the enhancement was due to the
altered electron density of the phenolic ring (C]C mode) aer
hydrogen bonding with DMF, so we undertook IR spectroscopy
of a mixture of acetonitrile (ACN) and phenol at a ratio of 1 : 1
(Fig. 3b). The shiing to a higher IR absorbance frequency of CN
signied formation of a hydrogen bond between ACN with
phenolic OH (Fig. 3b). An absence of enrichment of IR absor-
bance of C]C explained the non-involvement of hydrogen
bonds. Hence, these results suggested that the enhancement in
IR absorbance of the C]C mode was not due to n–p* bonding
or hydrogen bonding. Therefore, the enhancement was prob-
ably due to vibrational coupling between the C]C and C]O
modes of amide (DMF) and phenol derivatives (PC and PCP).
To check that our hypotheses on vibrational coupling were
robust, we carried out DFT calculations (b3lyp, 6311 G (d, p)) for
PC and PCP hydrogen-bonded with DMF, and the videos are
provided in ESI† (AV1, AV2, AV3, AV4). Vibrational couplings
were visualized between the C]C of phenol derivatives and
C]O of DMF. Interestingly, our experimental observations
aligned with the videos for DFT calculations. In AV1 and AV2,
the symmetric and asymmetric C]C modes of PC were coupled
signicantly with amide 1 of DMF; for PCP, the symmetric C]C
mode was coupled signicantly (AV3) but the asymmetric mode
was not (AV4). PC and PCP are phenol derivatives and form
hydrogen bonds with amide 1 of DMF, but they showed
different vibrational coupling. Hence, phenolic substitution
altered the pattern of vibrational coupling.
Vibrational coupling is dependent upon the frequency gap of
IR absorption, the distance between two vibration modes, and

Table 1 Vibrational frequencies of IR absorption for the C]C mode and the ratio of IR absorption area for the hydrogen-bonded C]C mode
and free C]C mode of phenol derivatives. The enhancement factor and transition dipole moment ratio were calculated from the experimental
data shown in Fig. 2

IR enhancement Transition dipole

Phenol derivatives Stretching mode Frequency (cm−1) ratio (R) moment ratio

PC Symmetric 1597.5 1.35 1.16

Asymmetric 1618.5 1.44 1.20
PCP Symmetric 1694.0 5.97 2.44
Asymmetric 1606.5 1 1

© 2023 The Author(s). Published by the Royal Society of Chemistry RSC Adv., 2023, 13, 1295–1300 | 1297
 View Article Online

RSC Advances Paper

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
Open Access Article. Published on 05 January 2023. Downloaded on 1/6/2023 5:20:04 AM.

Fig. 3 (a) Linear IR absorbance spectra of the C]C mode of anisole (0.1 M) in the absence (black) and presence (red) of dimethylformamide (0.1
M). (b) IR spectra of the CN mode of ACN (0.1 M) in the absence (black) and presence (green) of phenol (0.1 M) in CCl4. (c) IR spectra of the C]C
mode of phenol (0.1 M) in the absence (black) and presence (red) of ACN (0.1 M).

orientation. The frequency gap between the asymmetric C]C
mode of PC and C]O stretching mode of DMF was smaller
(56 cm−1) than that of the symmetric C]C and C]O (76 cm−1)
modes of PC (Fig. 4a). However, the intensity of asymmetric and
symmetric C]C modes was enhanced simultaneously. The
frequency gap was smaller for the asymmetric C]C mode
(67 cm−1) than symmetric C]C mode (79.5 cm−1) of PCP with
amide 1 of DMF. Signicant enhancement in IR absorbance was
observed for the symmetric C]C mode, but negligible
enhancement was observed for the asymmetric C]C mode
(Fig. 4b). The distance between the C]C and C]O modes of
phenol derivatives and DMF according to DFT calculations are
shown in Fig. 4c and d. An identical distance (5.2 Å) between
C]C and C]O could not explain the disparity in vibrational
coupling. Hence, we assumed that a different transition dipole
angle between the C]C mode and amide-I mode was the cause

Fig. 4 Linear IR spectra of amide I (dimethylformamide) and the C]C mode of (a) para-cresol and (b) para-chlorophenol. The IR spectra of
amide I (violet), free amide I (red), hydrogen-bonded amide I (blue), cumulative spectra (green), and C]C (gold) are shown. (c) Distance between
C]O and C]C of dimethylformamide and para-cresol. (d) Distance between C]O and C]C of dimethylformamide and para-chlorophenol. A
distance of 5.2 Å was calculated (by DFT) for para-cresol and para-chlorophenol, respectively, with amide I of dimethylformamide.

1298 | RSC Adv., 2023, 13, 1295–1300 © 2023 The Author(s). Published by the Royal Society of Chemistry
 View Article Online

Paper RSC Advances

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
Open Access Article. Published on 05 January 2023. Downloaded on 1/6/2023 5:20:04 AM.

Fig. 5 Red and green denote the linear vibration modes of C]O and C]C, respectively. bij is the coupling constant, and mi, and mj are the
transition dipole moments of the C]C mode and C]O mode, respectively. Orientation of C]O and C]C modes- (a) parallel (b) antiparallel (c)
linearly parallel (d) oppositely parallel and (e) perpendicular.

of this difference in vibrational coupling. To ascertain the AV10). The symmetric C]C mode of ortho-chlorophenol was
reason for this anomalous coupling behaviour, we exposed coupled with amide I, but the asymmetric C]C mode was not
different orientations between vibration modes (C]C and C] (AV11, AV12). Hence, the C]C symmetric and asymmetric
O) (Fig. 5). The sign of the coupling constant (bij) was dependent modes coupled with amide 1 for para-electron-promoting
upon the geometry of the hydrogen-bonded DMF and phenol phenolic compounds. Only the C]C symmetric mode
derivatives (eqn (1)).55 If C]C and C]O modes are parallel coupled with amide 1 for para- and ortho-electron-withdrawing-
(Fig. 5a) and bij is positive, the intensity of the C]O vibrational substituted phenolic compounds. The symmetric and asym-
mode will be enhanced, with a transition dipole moment mi + mj. metric C]C modes of electron-withdrawing meta-substituted
In contrast, the intensity of the C]C vibrational mode will be phenolic compounds were coupled like para-electron-
weaker, with a transition dipole moment mi − mj. In the case of promoting phenolic compounds.
antiparallel C]C and C]O modes (Fig. 5b) and negative
coupling constant bij, the C]C mode will be weaker and
correspondingly the C]O mode will be stronger, with a transi- Conclusions
tion dipole moment of mi + mj and mi − mj, respectively. In the
Employment of FTIR spectroscopy and DFT calculations
head-to-tail (Fig. 5c) or head-to-head orientation (Fig. 5d) of
revealed the anomalous vibrational coupling between amide 1
C]C and C]O, the lower frequency mode of C]C will carry
and the C]C mode of phenol derivatives. For PC, IR absor-
more oscillator strength compared with the C]O mode. An
bance was enhanced markedly for symmetric and asymmetric
unperturbed intensity will result if the C]C mode and C]O
C]C modes upon gradual addition of DMF. For PCP, the
mode are perpendicular to each other (Fig. 5e). The geometries
symmetric C]C mode was enhanced signicantly, whereas the
of the DMF–PC complex and DMF–PCP complex are in-between
asymmetric C]C mode was not. Even though the frequency gap
the limits we modelled in Fig. 5. The symmetric and asymmetric
was less for the asymmetric C]C transition compared with that
C]C modes of PC and symmetric C]C mode of PCP are
for the C]O transition, and the distance between C]C and
coupled in the manner shown in Fig. 5c and d. Probably, the
C]O was constant for symmetric and asymmetric C]C modes
asymmetric C]C mode of PCP was coupled with the C]O
for PC and PCP, distinctive behaviour was observed for different
mode of DMF as shown in Fig. 5e or they were coupled very
phenol derivatives in the presence of DMF. Theoretical and
weakly according to the model shown in Fig. 5a–d.
  3 experimental observations revealed that, with alteration of
2
! ! mi !
rij  ! rij  !
mi  ! mj phenol substituents, the coupling pattern changed. The area of
mj
bij ¼ 1=4p30 4 3 5 (1) IR absorbance was enhanced z5.97 times for the C]C
r3ij r5ij
symmetric mode, whereas it was altered negligibly for the
asymmetric mode, in the case of PCP. For PC, the area of IR
where bij is the coupling constant, 30 is permittivity, mi and mj are
absorbance for symmetric and asymmetric modes was
transition dipole moments, and rij is the distance between two
enhanced signicantly (z1.4 times). Thus, we revealed the
vibration modes.
nature of vibrational coupling based on the structure of a guest
We wished to validate our hypotheses of coupling of the
molecule hydrogen-bonded with amide I. Our conclusions
symmetric and asymmetric C]C modes of PC and PCP with
could be valuable for depiction of the unusual dynamics of
amide 1. Hence, we carried out DFT calculations of para-ethyl-
coupled amide-I modes as well as the dependency of vibrational
phenol and para-nitrophenol. Surprisingly, we observed the
coupling on altered factors.
same results as PC for para-ethylphenol (AV5, AV6) and as PCP
for para-nitrophenol (AV7, AV8). We calculated the vibrational
coupling for meta-chlorophenol and ortho-chlorophenol
complexes with DMF. The symmetric and asymmetric C]C
Conflicts of interest
modes of meta-chlorophenol were coupled with amide I (AV9, The authors declare no competing nancial interest.

© 2023 The Author(s). Published by the Royal Society of Chemistry RSC Adv., 2023, 13, 1295–1300 | 1299

RSC Advances
Acknowledgements
AG thanks SNBNCBS, Kolkata, India, for instrumental facilities
and nancial support from the DST-India. SC thanks DST India
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
for a fellowship.
References
1 R. E. Dickerson and I. Geis, The Structure and Action of
Proteins, Harper and Row, New York, Evanston, London,
Open Access Article. Published on 05 January 2023. Downloaded on 1/6/2023 5:20:04 AM.
1969.
2 D. L. Nelson and M. M. Cox, Lehninger Principles of
Biochemistry, W. H. Freeman and Company, New York, 2005.
3 J. M. Berg, J. L. Tymoczko and L. Stryer, Biochemistry, W. H.
Freeman and Company, New York, 2007.
4 S. Krimm and J. Bandekar, Adv. Protein Chem., 1986, 38, 181–
364.
5 A. Barth and C. Zscherp, Q. Rev. Biophys., 2002, 35, 369–430.
6 S. Woutersen and P. Hamm, J. Phys.: Condens. Matter, 2002,
14, R1035–R1062.
7 P. Hamm, M. Lim and R. M. Hochstrasser, J. Phys. Chem. B,
1998, 102, 6123–6138.
8 J. Ma, I. M. Pazos, W. Zhang, R. M. Culik and F. Gai, Annu.
Rev. Phys. Chem., 2015, 66, 357–377.
9 R. Adhikary, J. Zimmermann and F. E. Romesberg, Chem.
Rev., 2017, 117, 1927–1969.
10 S. D. Fried and S. G. Boxer, Acc. Chem. Res., 2015, 48, 998–
1006.
11 H. Kim and M. Cho, Chem. Rev., 2013, 113, 5817–5847.
12 W. Huan, Z. Yanfei, Z. Fengtao, K. Zhengang, H. Buxing,
H. Junfeng, W. Zhenpeng and L. Zhimin, Sci. Adv., 2021,
7(22), 1–9.
13 A. Ghosh, B. Cohn, A. K. Prasad and L. Chuntonov, J. Chem.
Phys., 2018, 149(18), 184501.
14 S. Chakrabarty, A. Barman and A. Ghosh, J. Phys. Chem. B,
2022, 126, 5490–5496.
15 S. Chakrabarty, S. H. Deshmukh, A. Barman, S. Bagchi and
A. Ghosh, J. Phys. Chem. B, 2022, 126, 4501–4508.
16 M. K. Gilson and H. X. Zhou, Annu. Rev. Biophys. Biomol.
Struct., 2007, 36, 21–42.
17 R. U. Lemieux, Acc. Chem. Res., 1996, 29, 373–380.
18 Y. Levy and J. N. Onuchic, Annu. Rev. Biophys. Biomol. Struct.,
2006, 35, 389–415.
19 D. L. Mobley, E. Dumont, J. D. Chodera and K. A. Dill, J. Phys.
Chem. B, 2007, 111, 2242–2254.
20 T. H. Plumridge and R. D. Waigh, J. Pharm. Pharmacol., 2002,
54, 1155–1179.
21 N. S. Myshakina, Z. Ahmed and S. A. Asher, J. Phys. Chem. B,
2008, 112(38), 11873–11877.
22 J. Ma, I. M. Pazos, W. Zhang, R. M. Culik and F. Gai, Annu.
Rev. Phys. Chem., 2015, 66, 357–377.
23 R. Adhikary, J. Zimmermann and F. E. Romesberg, Chem.
Rev., 2017, 117, 1927–1969.
24 S. D. Fried and S. G. Boxer, Acc. Chem. Res., 2015, 48, 998–
1006.
25 H. Kim and M. Cho, Chem. Rev., 2013, 113, 5817–5847.
26 A. I. Ahmed and F. Gai, Protein Sci., 2017, 26(2), 375–381.
1300 | RSC Adv., 2023, 13, 1295–1300
View Article Online
Paper
27 S. Chakrabarty, S. Maity, D. Yazhini and A. Ghosh, Langmuir,
2020, 36, 11255–11261.
28 A. Ghosh, A. K. Prasad and L. Chuntonov, J. Phys. Chem. Lett.,
2019, 10, 2481–2486.
29 D. G. Brown and J. Bostrom, J. Med. Chem., 2016, 59, 4443–
4458.
30 V. R. Pattabiraman and J. W. Bode, Nature, 2011, 480, 471–479.
31 A. B. Hughes, Origins and Synthesis of Amino Acids, Amino
Acids, Peptides and Proteins in Organic Chemistry, Wiley-
VCH, Weinheim, Germany, 2009, vol. 1.
32 A. A. Kaspar and J. M. Reichert, Drug Discovery Today, 2013,
18, 807–817.
33 R. B. Dyer, F. Gai, W. H. Woodruff, R. Gilmanshin and
R. H. Callender, Acc. Chem. Res., 1998, 31, 709–716.
34 S. Woutersen, R. Pster, P. Hamm, Y. Mu, D. S. Kosov and
G. Stock, J. Chem. Phys., 2002, 117, 6833–6840.
35 N. Demirdöven, C. M. Cheatum, H. S. Chung, M. Khalil,
J. Knoester and A. Tokmakoff, J. Am. Chem. Soc., 2004, 126,
7981–7990.
36 M. M. Waegele, R. M. Culik and F. Gai, J. Phys. Chem. Lett.,
2011, 2, 2598–2609.
37 A. L. Serrano, M. M. Waegele and F. Gai, Protein Sci., 2012,
21, 157–170.
38 H. Kim and M. Cho, Chem. Rev., 2013, 113(8), 5817–5847.
39 P. Hamm, and R. M. Hochstrasser, Ultrafast Infrared, and
Raman Spectroscopy, 2001.
40 A. M. Cunha, E. Salamatova, R. Bloem, S. J. Roeters,
S. Woutersen, M. S. Pshenichnikov and T. L. C. Jansen, J.
Phys. Chem. Lett., 2017, 8, 2438–2444.
41 H. Li, R. Lantz and D. Du, Molecules, 2019, 24, 186.
42 W. M. Zhang, V. Chernyak and S. Mukamel, J. Chem. Phys.,
1999, 110, 5011–5028.
43 C. Scheurer, A. Piryatinski and S. Mukamel, J. Am. Chem.
Soc., 2001, 123, 3114–3124.
44 A. Moran and S. Mukamel, Proc. Natl. Acad. Sci. U. S. A., 2004,
101, 506–510.
45 E. G. Buchanan, W. H. James, S. H. Choi, L. Guo,
S. H. Gellman, C. W. Müller and T. S. Zwier, J. Chem. Phys.,
2012, 137, 094301.
46 M. Lima, R. Chelli, V. V. Volkov and R. Righini, J. Chem.
Phys., 2009, 130, 204518.
47 S. Krimm and Y. Abe, Proc. Natl. Acad. Sci. U. S. A., 1972, 69,
2788–2792.
48 S. Roy, T. L. C. Jansen and J. Knoester, Phys. Chem. Chem.
Phys., 2010, 12, 9347–9357.
49 C. Liang, M. Louhivuori, S. J. Marrink, T. L. C. Jansen and
J. Knoester, J. Phys. Chem. Lett., 2013, 4, 448–452.
50 J. Schaefer, H. G. Ellen, Y. N. Backus and M. Bonn, J. Phys.
Chem. Lett., 2016, 7, 4591–4595.
51 I. V. Rubtsov, J. Wang and R. M. Hochstrasser, J. Phys. Chem.
A, 2003, 107, 3384–3396.
52 J. V. Hatton and R. E. Richards, Mol. Phys., 1960, 3, 253–263.
53 M. Malathi, R. Sabesan and S. Krishnan, Curr. Sci., 2002, 86,
838–842.
54 A. Ghosh, J. Phys. Chem. B, 2019, 123, 7771–7776.
55 P. Hamm and M. Zanni, Concepts and Methods of 2D Infrared
Spectroscopy, Cambridge University Press, New York, 2011.
© 2023 The Author(s). Published by the Royal Society of Chemistry
 pubs.acs.org/JPCB Article

Anomalous Infrared Absorbance of S�O: A Perturbation Study of

α‑C−H/D
Suranjana Chakrabarty, Anjan Barman,* and Anup Ghosh*
Cite This: https://doi.org/10.1021/acs.jpcb.2c01374 Read Online

ACCESS Metrics & More Article Recommendations sı Supporting Information

*
ABSTRACT: Solvatochromic shifts of S�O vibrational probes describe
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Downloaded via S.N. BOSE NATL CTR BASIC SCIENCES on July 22, 2022 at 11:14:13 (UTC).

the strength of the surrounding electric fields and the hydrogen bonding
status. Herein, we demonstrated how the solvents alter the infrared (IR)
spectra of the S�O vibrating mode. The experimental measurement of the
involvement of α-H/D isotopic interactions with different solvents and
their effects on the IR absorbance spectra of the vibrational probe provides
detailed knowledge of the microsolvation environment despite the
complexity of overlapping bands in the spectra. Herein, we discover how
the solvents interact differently with DMSO and DMSO-d6, while being
electronically and structurally the same. Interestingly, the IR spectrum of
the S�O mode remains unaltered during α-isotopic replacement in the
presence of aprotic solvents (acetone, acetonitrile, and dichloromethane),
but in strongly coordinating polar solvents (D2O), it is altered remarkably.
There is a lack of quantitative information about the influence of the α-H atom or α-isotopic substitution on the vibrational probe in
the literature. Our experiments provide a detailed molecular understanding of the structure of DMSO in DMSO−solvent binary
mixtures. As DMSO plays an important role in virtually all subdisciplines of chemistry and biology, we believe that our work will be
of interest to a large diversity of studies in these fields.

■ INTRODUCTION
Interactions between molecules in the condensed phase
assessments become a challenging area of research. A proven
commanding tool is vibrational spectroscopy to unravel the
secret of molecular liquids. The infrared spectrum of a
molecule not only depends on the strength of a certain bond
and mass of the vibrating atoms but also can be markedly
influenced by medium factors. It has been reported that
solute−solvent interactions can have a reckonable influence on
the frequencies and intensities of vibrational bands in infrared
(IR) spectroscopy.1−7 For many years, the impact of the
solvent on the spectral properties of molecules, usually
mentioned as solvatochromism, has been investigated.8,9
Particularly, wavenumber values, intensities, and line shapes
of IR absorption spectra are influenced by the nature of the
solvent via dielectric effects and specific interactions. There are
lots of models that have already been projected to treat the
shifts in vibrational frequencies over varied ranges of liquid
densities.1,10−13 A net solvent-induced force invoked by these
models can disturb the normal-mode frequencies of an isolated
solute molecule. Thus, the obtained frequency shifts in the
stretching vibrational frequency are directly proportional to the
changes in bond lengths, which is induced by the surrounding
medium of the solutes.10,14 In this context, molecules bearing
the C�O moiety have been extensively investigated, as their
polarity and hydrogen bond-accepting nature make them very
sensitive to the solvent influence.15−26 However, being a very
good solvent/cosolvent and a pharmaceutically important
compound, recently, similar S�O bond-containing DMSO
molecules have gained interest to be explored.27−31 Although
numerous vibrational spectroscopic reports on DMSO have
been published, a consensus regarding the interaction with the
solvent is yet to be reached.32−44 Several vibrational studies
have been reported based on the IR absorbance peak
assignment of S�O and the corresponding mode of
DMSO.33−44 Unaltered S�O IR absorption in organic
solvents and altered S�O IR absorption in aqueous solvents
warn us to explore the α-H/D interaction of DMSO and
DMSO-d6 with the surrounding solvent molecules, as they are
structurally and electronically the same.33,43,44 Such an
observation cannot be explained based on the S�O bond
interaction of DMSO/DMSO-d6 with solvents; however,
mostly in the literature, only the interaction of the S�O
bond of DMSO with the solvent has been investigated
widely.31−37,40,41 In the binary mixture of DMSO−water, the
S�O bond exists in four structural forms, namely, self-
associated (aggregated), free monomer (0HB), singly hydro-
gen-bonded (1HB), and doubly hydrogen-bonded (2HB)
species.42,43 In this regard, our present report is an additive
study wherein we have explored the H-bond interaction not
Received: February 25, 2022
Revised: June 21, 2022

© XXXX American Chemical Society https://doi.org/10.1021/acs.jpcb.2c01374

A J. Phys. Chem. B XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry B
only with the S�O bond but also with α-H/D of DMSO/
DMSO-d6 in an aqueous solvent.
To identify structural and physicochemical behaviors, we
have performed IR absorption, density functional theory
(DFT) calculations, and molecular dynamics (MD) simu-
lations of DMSO dissolved in various solvents, namely acetone,
ACN, DCM, D2O, and hexafluoroisopropanol (HFIP).
However, a combined experimental and theoretical approach
offers the potential to understand the diverse interaction of the
DMSO mode with solvents, which motivates the present study.
Moreover, the nature of the interaction of DMSO in water has
been explored to understand α-H-bond effects on the IR
absorptions of S�O stretching modes. To distinguish between
different electrostatic environments and H-bond interactions
of DMSO, we turned to link experimental IR absorption with
MD simulation and DFT calculations. The important role of
this type of interaction is causing the condensed-phase
structure and dynamics and its significances on the different
vibrational spectra (Scheme 1).
Scheme 1. Chemical Structure of DMSO and DMSO-d6
■ MATERIALS AND METHODS
Sample Preparation. Dimethyl sulfoxide (DMSO 99.9%),
dimethyl sulfoxide d6 (DMSO-d6, 99.9%), and all solvents
(99.9%) are bought from Sigma-Aldrich and used without
further purifications. The linear IR absorption spectra were
recorded with an Fourier transform infrared (FTIR)
spectrometer (Bruker, Tensor 70). The 0.1 M solutions of
DMSO and DMSO-d6 molecules in acetone, ACN, DCM,
D2O, and HFIP were used throughout the study. The sample
solutions were placed between the 2 mm-thick CaF2 windows
separated using a Teflon spacer.
DFT Calculations. The molecular geometries of DMSO/
DMSO-d6 and their H-bond complexes were optimized by
density functional theory calculations, using the B3LYP hybrid
Figure 1. Black spectra denote the experimental IR absorption spectra of DMSO ((a) acetone, (b) ACN, (c) DCM, and (d) D2O) and DMSO-d6
((e) acetone, (f) ACN, (g) DCM, and (h) D2O). The experimental IR spectra are fitted to four peaks using Voigt lineshape functions. The red
peaks correspond to the S�O stretch, and the green, blue, and magenta peaks correspond to the CH3 (a−d) and CD3 (e−h) rocking modes. All
red dotted lines signify the S�O IR absorption peak position.
B https://doi.org/10.1021/acs.jpcb.2c01374
pubs.acs.org/JPCB Article
functional and a 6-311G +(d,p) basis set as implemented in
Gaussian 09.44−46
MD Simulation. MD simulations of DMSO in water and
organic solvents were carried out with the CHARMM general
forcefield using the GROMACS version 4.6.5 package.47−50
The forcefield parameters of DMSO and organic solvents were
obtained from virtualchemistry.org, which has been thoroughly
tested earlier.51 We have used the TIP4P water model for MD
simulation in water.52 A single DMSO molecule was placed in
the center of a 5 nm cubic box of the respective solvent
molecules. Then, energy minimizations of each MD simulation
were performed using the steepest-descent algorithm followed
by the equilibrium in the NVT ensemble using a velocity-
rescaling thermostat for 100 ps and equilibrium in NPT at 1
bar pressure using a Parrinnello−Rahman barostat for 1 ns at
300 K temperature.53,54 Finally, the coordinates were recorded
for a 1 ns MD simulation run and used for the calculation of
the electric field and hydrogen bond population of the S�O
bond of DMSO. Hydrogen bond populations were computed
using a 3.5 Å donor (D)-to-acceptor (A) distance cutoff,
between the oxygen atoms of the donor and acceptor, and a
30° D−H−A angle cutoff. The distance parameter was selected
to match the first minimum of the D−A radial distribution
function. The electric fields, which the surrounding solvent
molecules exerted on the S�O bond of DMSO and DMSO-
d6, were calculated by following similar protocol used in ref 55.
■ RESULTS AND DISCUSSION
To resolve the detailed molecular interaction of DMSO, in the
binary mixture solution (solute−solute and solute−solvent)
state, we have performed a series of infrared absorption
experiments on 0.1 M DMSO and DMSO-d6 in different
solvents, namely, acetone, ACN, DCM, and D2O. The
representative spectra are shown in Figure 1. It is assumed
that the electric field of S�O for these low concentrations (0.1
M) of both DMSO and DMSO-d6 is constant (Figure 2). The
second derivatives of the broad overlapping IR absorption
spectra of DMSO (Figure S1) and DMSO-d6 (Figure S2) in
the 950−1100 cm−1 region indicate four underlying peaks.
Each mentioned spectrum is fitted with four peaks using Voigt
lineshape functions, by fixing the frequencies corresponding to
the minima in the second derivative spectra (Figures S1 and
J. Phys. Chem. B XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry B
Figure 2. Center frequencies of the S�O stretching vibrations of
DMSO (black squares) and DMSO-d6 (red circles) versus the
calculated local electric field for different solvents. The solid lines are
the connection of these data.
S2). The fitted peaks in Figure 1 are assigned to the S�O
symmetric stretching mode (red) and three CH3/CD3 rocking
modes (green, blue, and magenta) based on previous
reports.34−39 Figure 1 shows that the IR absorption peak
position of S�O remained almost constant in an aprotic
organic solvent but surprisingly shifted toward a lower
frequency for DMSO-d6 (989 cm−1) than for DMSO (1013
cm−1) in D2O (Table S1). Quantum calculations also agreed
with the experimental results (Table S2).
To resolve the reason behind one S�O mode IR absorption
peak of DMSO in neat D2O, we have performed a MD
simulation with GROMACS (version 4.6.5) to calculate the
hydrogen-bonded population. The S�O mode of DMSO is
majorly (>96%) doubly hydrogen-bonded in neat water,
wherein single hydrogen-bonded (3%>) and free S�O (1%
>) is negligibly small. Therefore, single S�O IR absorption
peaks majorly come from the double hydrogen-bonded S�O
complex in neat water. Figures 1 and 2 show 24 cm−1 red-
shifted S�O IR absorbance of DMSO-d6 than DMSO,
although they are structurally and electronically similar except
the molecular weight for the additional isotopic mass. This
unusual IR absorption of the S�O mode of DMSO and
DMSO-d6 in D2O dictates us to investigate the effect of α-H/
D on S�O stretching modes. The similar IR absorption of the
S�O mode in aprotic solvents and dissimilar IR absorption in
D2O confirmed that D2O interacts differently with DMSO and
DMSO-d6, and probably, the different α-C−H/D interaction
with D2O is the reason (Figures 1 and 2). Among polar liquids,
D2O creates complicated situations because of the H-bond
donor (OD) and acceptor capability concurrently. Therefore,
possibly, the “O” atom of D2O form H-bonds with α-H/D of
DMSO/DMSO-d6 as a H-bond acceptor, and the “D” atom of
D2O form H-bonds with the S�O bond as a H-bond donor.
Normally, hydrogen atoms attached to a carbon atom cannot
participate in hydrogen bonding, but when the carbon atom is
bonded to electronegative atoms, it can form a hydrogen bond.
However, herein, H/D is attached with the α-carbon of DMSO
and DMSO-d6. The high polarizability of the S�O bond
makes α-H/D acidic and suitable to form H-bonds.41,42
A simple quantum treatment of this situation prescribed a
well-known solution that the energy of the α-C−H/D depends
upon the quantum number v (eq 1) where h, μ, and k refer to
Planck’s constant, reduced mass (eq 2), and force constant
(related to the stiffness of the bond), respectively. In most of
C https://doi.org/10.1021/acs.jpcb.2c01374
pubs.acs.org/JPCB Article
the chemical and biological interests, the system stays in its
ground state where vibrational quantum number v = 0.
However, it is important to note that the energy of the system
(Ev) cannot be zero in the ground state even at 0 K. Thus, the
C−H and C−D have a residual amount of vibrational energy,
which is called zero-point vibrational energy (ZPVE). k is
unaltered during the replacement of the “H” atom by “D”, as
the electronic structure of the molecule does not change. The
reduced mass is in the denominator of eq 1; the ZPVE of the
C−D bond is lower in position compared to the C−H bond
(Figure 3a). Thus, it is obvious that the C−D bond is stronger
Figure 3. Anharmonic potential energy curves with the indication of
vibrational energy levels and positions occupied by C−H and C−D.
‘-----’ represents the vibrating bond. The orange and blue colors
represent the zero-point energy of the respective bond. (a, b) Free
and hydrogen-bonded C−H/D bonds, respectively.
than the C−H bond. In the potential energy diagram shown in
Figure 3a, C−H and C−D stretchings are located close to the
center of the potential curve. However, the difference in both
cases is however considerably enough: the amplitude of C−H
(protonic) vibrations is greater than that of C−D (deuteronic)
ones, as reflected in Figure 3a.
The important questions of chemical and biological systems
concern how the strength of the H-bond is affected during
deuteration. Subsequently, how this α-C−H/D H-bond will
interrupt the S�O bond and its IR absorption? In Figure 3b,
we present the anharmonic vibrations of C−H---O/C−D---O.
As the amplitude of the C−H vibration is greater compared to
that of C−D, the C−H---O bond is less stable than the C−D---
O H-bond. On the zero-point energy level of vibrations, the
“C” atom is shifted toward the “D” and acceptor atom “O” of
D2O, which leads to strengthening of the H-bond bridge.
Furthermore, another important factor is the difference in the
amplitude of the bending vibrations of the C−H and C−D
bonds shown in Figure 4. Undoubtedly, due to the higher
amplitude of C−H bending vibrations, the “C−H---O” H-bond
is less stable than the C−D---O H-bond with D2O. Parallelly,
Figure 4. Amplitude of the bending vibration of (a) α-C−H and (b)
α-C−D of DMSO and DMSO-d6.
J. Phys. Chem. B XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry B
Figure 5. Illustration of (a) CH3 and (b) CD3 group rotation over the C−S axis from DMSO and DMSO-d6.
having a lower mass, −CH3 rotates faster than the −CD3
group, which makes the C−H---O H-bond weaker than the
C−D---O H-bond (Figure 5). The combination of the
mentioned effects leads to the situation where a deuterated
H-bond bridge is more stable.
i 1y h K
E v = jjjv + zzz
k 2{2 (1)
m1m2
= Parallelly, the S�O bond becomes weaker, as shown in
m1 + m2 (2)
Therefore, the plausible explanation of the S�O IR absorption
peak difference is the delocalization of the bond after the H-
bond formation of the α-H/D atom with D2O. As the α-D
forms a stronger H-bond than α-H with D2O, bond
delocalization is higher in DMSO-d6 than in DMSO (Scheme
S1). Therefore, unsurprisingly, the bond order decreases and
the bond distance increases in DMSO-d6. Subsequently, S�O
IR absorption of DMSO-d6 shifted toward a lower wave-
number compared to the DMSO one. The maximum no of
water molecules that can be accommodated by the α-H atom
of a single DMSO molecule is shown in Scheme S1. MD
simulation results of the α-C−H H-bond population validate
the arguments (Figure S5).
To support this viewpoint, DMSO and DMSO-d6 were
optimized by DFT calculations. Most importantly, the S�O
bond orders of DMSO and DMSO-d6 of the optimized
geometry are calculated, and their values are shown in Table 1.
Table 1. Bond Order (BO) of S�O in Free and after α-H-
Bond Formation of DMSO and DMSO-d6 with Water
(D2O)a
BO of BO of S�O (After α-H/D α-H-bond formation
S�O H-bond formation) energy (kJ/mol)
DMSO 1.1518 1.1116 −7.0798
DMSO- 1.1517 1.1061 −7.6310
d6
a
Hydrogen bond formation energy of α-C−H/D with D2O (column
four).
Parallelly, we have calculated the α-H/D H-bond energy by
optimizing the geometry of DMSO and DMSO-d6 with one
D2O molecule in neat D2O (Figure 4). The higher α-H-bond
formation energy of “C−D---O” (−7.631 kJ/mol) illustrated
the higher stability than that of “C−H---O” (−7.0798 kJ/mol).
Naturally, the α-D atom of DMSO-d6 forms a stronger H-bond
than the α-H atom of DMSO, which we have claimed in the
D https://doi.org/10.1021/acs.jpcb.2c01374
pubs.acs.org/JPCB Article
above-mentioned IR experiments. The lower bond order of
DMSO-d6 illustrated the lower IR absorbance frequency
compared to DMSO (Table 1). A close analysis indicated
that an increase of the H-bond strength correlated with a
decrease of the C−H/D bond strength. Due to the H-bond
formation of α-H/D, delocalization of the σ C−H/D bond
occurred. It caused an amplified overlap with the antibonding
orbital of the C−H/D bond, resulting in the reduction of the
α-C−H/D bond order, and the α-C−H/D bond became weak.
Scheme S1. As the C−D forms a stronger H-bond,
delocalization of the C−D bond is more, and thus, the S�
O bond of DMSO-d6 becomes weaker than the S�O bond of
DMSO. Consequently, the S�O IR absorption shifted more
for DMSO-d6 than for DMSO in D2O, wherein as there is no
chance to form H-bonds in aprotic organic solvents, the IR
absorption peak position remains unaltered. One “C−D---O”
H-bond with D2O of DMSO-d6 is 0.551 kJ/mol stronger than
the “C−H---O” H-bond formation energy of DMSO (Table
1). Naturally, six C−D H-bonds reduced the S�O bond order
more effectively compared to six C−H H-bonds. Based on this
apparent inconsistency of S�O IR absorption, we have
explained how DMSO and DMSO-d6 interact differently with
D2O instituted on α-H/D.
Furthermore, the dimerization of DMSO validates our
argument of the acidity of α-H, which we have proved by
quantum calculations (Figure S4).31−37 It is well established
that DMSO forms an antiparallel dimer by involving the α-C−
H H-bond and Coulombic interactions of the opposite S�O
dipole in neat and even in lower concentrations in organic
solvents.56 Herein, we have also designed the dimerization that
occurred due to the α-C−H hydrogen bonding with the “O”
atom of DMSO (Figure S4). We have used the MP2 (cc-
pVDZ) level of theory to optimize the antiparallel DMSO
dimer structure.56
The unusual influence of the fluorinated solvent (HFIP) is
observed on IR absorption of the S�O mode of DMSO
(Figure 6). Surprisingly, we have observed a lower frequency of
S�O IR absorbance of DMSO (0.1 M) in HFIP (black)
compared to that in D2O (red). The relative polarity of HFIP
(0.96) is lower than that of D2O (1), and the acidity (pKa =
7.97) is also lower than that of D2O (pKa = 7). As the acidity
of HFIP is lower, it is obvious that it will form a weak H-bond
compared to that of D2O with the S�O mode. However, the
H-bond energy of the α-H atom of DMSO with D2O (7.07 kJ/
mol) is lower compared to that of the F atom of HFIP (8.75
kJ/mol), which commands the S�O IR absorption peak in
lower frequency due to more delocalization of the C−H bond
J. Phys. Chem. B XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry B
Figure 6. IR absorption of DMSO in (a) D2O (blue) and (b) HFIP
(black).
over the S�O bond in HFIP. Subsequently, we have
calculated the electric field of the S�O mode in HFIP and
D2O. Surprisingly the higher electric field of S�O in HFIP
(125 MV/cm) than in D2O (80 MV/cm) prescribed a lower
frequency of S�O IR absorbance of DMSO in HFIP
compared to D2O. Therefore, combining these two effects
described the unusual S�O IR absorbances of DMSO in a
fluorinated solvent.
To check whether the α-C−H/D effect is true only for
DMSO-DMSO-d6, we have performed a similar experiment for
structurally same acetone and acetone d6 in acetonitrile and in
D2O, shown in Figure S3.
The evolution of these intrinsic α-H properties can reveal
intricate details about the nature of the interactions within the
system. These could be utilized to study a diverse range of
problems, including sulfoxide, carbonyl, ester, and amide I with
the α-H atom in liquid water or protic solvents.
The evaluation of the α-C−H H-bond effect on IR
absorption of the S�O mode is the indication of awareness
of the α-(N−H/D) H-bond effect on the IR absorption of
amide I in D2O. During the IR absorption experiment of
biological molecules at different pH values, mostly, D2O is
used as a solvent. Therefore, there are a lot of chances to
exchange α-N−H to α-N−D (Scheme S2).57 Thus, misinter-
pretation may occur on the observed amide I IR absorption, as
it is shifted toward a lower frequency not only because of
deuteration but also due to hydrogen bond formation of N−D
with water (Scheme S2). It is very clear from Scheme S2 that
when N−D forms a hydrogen bond with D2O, the amide I IR
absorption frequency shifted significantly more than in organic
solvents.
■ CONCLUSIONS
In this work, we have inspected the isotopic influence on the
structural properties of DMSO and DMSO-d6 in different
solvent binary mixtures. Due to the various biological and
chemical implications of DMSO and its binary mixture
solutions, the S�O probe has gained our attention for
research. Isotopic exchange of α-C−H to α-C−D altered the
IR absorption of the S�O mode of DMSO in protic solvents
but unaltered in an organic aprotic solvent. Herein, we have
exposed that α-C−D forms a stronger H bond than α-C−H in
D2O, which is the reason behind the IR absorption spectral
shift of the S�O mode toward a lower frequency. Parallelly,
we have clarified an unusual IR absorption of the S�O mode
in fluorinated solvents. This study increases our knowledge,
E https://doi.org/10.1021/acs.jpcb.2c01374
pubs.acs.org/JPCB Article
which triggers our ability to control the chemical reactions and
warns us to be aware of the α-hydrogen effect in water. This
report is best explanatory when analyzing the α-H/D isotope
effect on S�O IR absorption and the hazards of α-H-
containing amide and ketones.
■
*
ASSOCIATED CONTENT
sı Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.jpcb.2c01374.
Fitted IR absorption spectra and second-order derivative
spectra of DMSO and DMSO-d6 in different solvents;
solvated structure of DMSO and DMSO-d6; exper-
imental IR spectra of acetone and acetone d6 in different
solvents; schematic structure of N methyl acetamide and
deuterated N methyl acetamide; dimeric structure of
DMSO; fitting parameters of SO IR absorptions;
calculated SO IR absorption frequency of DMSO and
DMSO-d6; and histogram of α-C−H hydrogen bond
populations (PDF)
■ AUTHOR INFORMATION
Corresponding Authors
Anjan Barman − Department of Condensed Matter Physics
and Materials Sciences, S. N. Bose National Centre for Basic
Sciences, Kolkata 700106, India; orcid.org/0000-0002-
4106-5658; Email: abarman@bose.res.in
Anup Ghosh − Department of Condensed Matter Physics and
Materials Sciences, S. N. Bose National Centre for Basic
Sciences, Kolkata 700106, India; orcid.org/0000-0002-
1442-8740; Email: anupg86@gmail.com, anup.ghosh@
bose.res.in
Author
Suranjana Chakrabarty − Department of Condensed Matter
Physics and Materials Sciences, S. N. Bose National Centre
for Basic Sciences, Kolkata 700106, India
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.jpcb.2c01374
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
A.G. thanks SNBNCBS, Kolkata, for instrumental facilities and
financial support from the DST, India. A.G. also thanks Dr.
Suman Chakrabarty (SNBNCBS) for his valuable suggestions
during MD simulations.
■ REFERENCES
(1) Schweizer, K. S.; Chandler, D. Vibrational dephasing and
frequency shifts of polyatomic molecules in solution. J. Chem. Phys.
1982, 76, 2296−2314.
(2) Huggins, C. M.; Pimentel, G. C. Infrared Intensity of the C - D
Stretch of Chloroform-d in Various Solvents. J. Chem. Phys. 1955, 23,
895−898.
(3) Akins, D. L. Theory of Raman scattering by aggregated
molecules. J. Phys. Chem. A 1986, 90, 1530−1534.
(4) Mortensen, A.; Nielsen, O. F.; Yarwood, J.; Shelley, V.
Anomalies in the Isotropic Raman Spectra of Liquid Mixtures of
Isotopomers of Formamide. Intermolecular Interactions in the
Carbonyl Stretching Region. J. Phys. Chem. B 1995, 99, 4435−4440.
J. Phys. Chem. B XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry B
(5) Lord, R. C.; Nolin, B.; Stidham, H. D. Quantitative Study of the
Bonding of Chloroform-d in Various Solvents by Infrared
Spectrometry. J. Am. Chem. Soc. 1955, 77, 1365−1368.
(6) McHale, J. L. Intermolecular vibrational resonance coupling:
Intensity borrowing in polarized Raman spectroscopy. Int. J. Quantum
Chem. 1991, 40, 593−602.
(7) Veas, C. H.; McHale, J. L. Intermolecular resonance coupling of
solute and solvent vibrational modes. J. Am. Chem. Soc. 1989, 111,
7042−7046.
(8) Reichardt, C. Solvents and Solvent Effects in Organic Chemistry,
3rd ed.; Wiley-VCH, 2003.
(9) Buckingham, A. D. Solvent Effects in Infra-Red Spectroscopy.
Proc. R. Soc. London, Ser. A 1958, 248, 169−182.
(10) Benson, A. M., Jr.; Drickalner, H. G. Stretching Vibrations in
Condensed Systems: Especially Bonds Containing Hydrogen. J. Chem.
Phys. 1957, 27, 1164−1174.
(11) Pullin, A. D. E. The variation of infra-red vibration frequencies
with solvent. Spectrochim. Acta 1958, 13, 125−138.
(12) Buckingham, A. D. Solvent effects in vibrational spectroscopy.
Trans. Faraday Soc. 1960, 56, 753−760.
(13) Dijkman, F. G.; van der Maas, J. H. Inhomogeneous
broadening of Morse oscillators in liquids. J. Chem. Phys. 1977, 66,
3871−3878.
(14) Zakin, M. R.; Herschbach, D. R. Relation of vibrational
frequency shifts to molecular compression in liquid benzene. J. Chem.
Phys. 1985, 83, 6540−6541.
(15) Wang, H.; Wang, L.; Shen, S.; Zhang, W.; Li, M.; Du, L.;
Zheng, X.; Phillips, D. L. Effects of hydrogen bond and solvent
polarity on the C�O stretching of bis(2-thienyl) ketone in solution.
J. Chem. Phys. 2012, 136, No. 124509.
(16) Kolling, O. W. Effect of Hydrogen Bonding Solvents on the
Infrared Absorption Band for the Fundamental Vibration of the
Carbonyl Group in 1,1,3,3-Tetramethylurea. Trans. Kans. Acad. Sci.
1999, 102, 53−56.
(17) Almeida, L. C. J.; Santos, P. S. The anomalous solvent effect in
the vibrational spectrum of 2,3-diphenyl-cycloprop-2-enone: an
experimental and theoretical investigation. Spectrochim. Acta, Part A
2002, 58, 3139−3148.
(18) Nyquist, R. A.; Fouchea, H. A.; Hoffman, G. A.; Hasha, D. L.
Infrared Study of β -Propiolactone in Various Solvent Systems and
Other Lactones. Appl. Spectrosc. 1991, 45, 860−867.
(19) Nyquist, R. A.; Putzig, C. L.; Hasha, D. L. Solvent Effect
Correlations for Acetone: IR versus NMR Data for the Carbonyl
Group. Appl. Spectrosc. 1989, 43, 1049−1053.
(20) Nyquist, R. A.; Putzig, C. L.; Yurga, L. Solvent and
Concentration Effects on Carbonyl Stretching Frequencies: Ketones.
Appl. Spectrosc. 1989, 43, 983−991.
(21) Nyquist, R. A.; Clark, T. D.; Streck, R. Infrared study of alkyl
carboxylic acids in CCl4 and/or CHCl3 solutions. Vib. Spectrosc. 1994,
7, 275−286.
(22) Bruni, P.; Giorginy, E.; Maurelli, E.; Tosi, G. Fourier transform
infrared spectrometry investigation of solvent effect on NH and CO
stretching modes in N-acylaminopyridines. Vib. Spectrosc. 1996, 12,
249−255.
(23) Nyquist, R. A.; Putzig, C. L.; Clark, T. D. Infrared study of 1,3-
dimethyl-2-imidazolidinone in various solvents. Vib. Spectrosc. 1996,
12, 81−91.
(24) Engberts, J. B. F. N.; Famini, G. R.; Perjessy, A.; Wilson, L. Y. J.
Solvent effects on stretching frequencies of some 1-substituted 2-
pyrrolidinones. J. Phys. Org. Chem. 1998, 11, 261−272.
(25) Ghosh, A.; Cohn, B.; Prasad, A. K.; Chuntonov, L. Quantifying
conformations of ester vibrational probes with hydrogen-bond-
induced Fermi resonances. J. Chem. Phys. 2018, 149, No. 184501.
(26) Ghosh, A. Vibrational Coupling on Stepwise Hydrogen Bond
Formation of Amide I. J. Phys. Chem. B 2019, 123, 7771−7776.
(27) Jacob, S. W.; Herschler, R. Pharmacology of DMSO.
Cryobiology 1986, 23, 14−27.
F https://doi.org/10.1021/acs.jpcb.2c01374
pubs.acs.org/JPCB Article
(28) Wood, C. D.; Wood, J. Pharmacologic and Biochemical
considerations of Dimethyl Sulfoxide. Ann. N. Y. Acad. Sci. 1975, 243,
7−19.
(29) David, N. A. The Pharmacology of Dimethyl Sulfoxide. Annu.
Rev. Pharmacol. 1972, 12, 353−374.
(30) Santos, N. C.; Coelho, J. F.; Silva, J. M.; Saldanha, C.
Multidisciplinary utilization of dimethyl sulfoxide: pharmacological,
cellular, and molecular aspects. Biochem. Pharmacol. 2003, 65, 1035−
1041.
(31) Fawcett, W. R.; Kloss, A. A. Attenuated total reflection fourier-
transform IR spectroscopic study of dimethyl sulfoxide self-association
in acetonitrile solutions. J. Chem. Soc., Faraday Trans. 1996, 92,
3333−3337.
(32) Fawcett, W. R.; Kloss, A. A. Solvent-Induced Frequency Shifts
in the Infrared Spectrum of Dimethyl Sulfoxide in Organic Solvents. J.
Phys. Chem. C 1996, 100, 2019−2024.
(33) Sastry, M. I. S.; Singh, S. J. Self-association of dimethyl
sulphoxide and its dipolar interactions with water: Raman spectral
studies. J. Raman Spectrosc. 1984, 15, 80−85.
(34) Skripkin, M. Y.; Lindqvist-Reis, P.; Abbasi, A.; Mink, J.;
Persson, I.; Sandstrom, M. Basic materials physics of transparent
conducting oxides. Dalton Trans. 2004, 4038−4049.
(35) Perelygin, I. S.; Itkulov, I. G.; Krauze, A. S. Temperature
dependence of associative equilibria of DMSO according to raman
scattering spectra. Russ. J. Phys. Chem. 1991, 65, 410−414.
(36) Shikata, T.; Sugimoto, N. Dimeric Molecular Association of
Dimethyl Sulfoxide in Solutions of Nonpolar Liquids. J. Phys. Chem. A
2012, 116, 990−999.
(37) Figueroa, R. H.; Roig, E.; Szmant, H. H. Infrared study on the
self-association of dimethyl sulfoxide. Spectrochem. Acta 1966, 22,
587−592.
(38) Singh, S.; Srivastava, S. K.; Singh, D. K. Raman scattering and
DFT calculations used for analyzing the structural features of DMSO
in water and methanol. RSC Adv. 2013, 3, 4381−4390.
(39) Gajda, R.; Katrusiak, A. Electrostatic matching versus close-
packing molecular arrangement in compressed dimethyl sulfoxide
(DMSO) polymorphs. J. Phys. Chem. B 2009, 113, 2436−2442.
(40) Clark, T.; Murray, J. S.; Lane, P.; Politzer, P. Why are dimethyl
sulfoxide and dimethyl sulfone such good solvents? J. Mol. Model.
2008, 14, 689−697.
(41) Venkataramanan, N. S.; Suvithaa, A.; Kawazoea, Y. Density
functional theory study on the dihydrogen bond cooperativity in the
growth behavior of dimethyl sulfoxide clusters. J. Mol. Liq. 2018,
454−462.
(42) Oh, K. I.; Rajesh, K.; Stanton, J. F.; Baiz, C. R. Quantifying
hydrogen-bond populations in DMSO/water mixtures. Angew. Chem.,
Int. Ed. 2017, 56, 11375−11379.
(43) Oh, K. I.; Baiz, C. R. Crowding Stabilizes DMSO−Water
Hydrogen-Bonding Interactions. J. Phys. Chem. B 2018, 122, 5984−
5990.
(44) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J.
Ab Initio Calculation of Vibrational Absorption and Circular
Dichroism Spectra Using Density Functional Force Fields. J. Phys.
Chem. A 1994, 98, 11623−11627.
(45) Becke, A. D. Density-functional thermochemistry. III. The role
of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652.
(46) Chakrabarty, S.; Maity, S.; Yazhini, D.; Ghosh, A. Surface-
Directed Disparity in Self-Assembled Structures of Small-Peptide L-
Glutathione on Gold and Silver Nanoparticles. Langmuir 2020, 36,
11255−11261.
(47) Van Der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark,
A. E.; Berendsen, H. J. C. GROMACS: Fast, Flexible, and Free. J.
Comput. Chem. 2005, 26, 1701−1718.
(48) Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E.
GROMACS 4: Algorithms for Highly Efficient, Load-Balanced, and
Scalable Molecular Simulation. J. Chem. Theory Comput. 2008, 4,
435−447.
J. Phys. Chem. B XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

(49) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D.
A. Development and Testing of a General Amber Force Field. J.
Comput. Chem. 2004, 25, 1157−1174.
(50) Caleman, C.; van Maaren, P. J.; Hong, M.; Hub, J. S.; Costa, L.
T.; van der Spoel, D. Force Field Benchmark of Organic Liquids:
Density, Enthalpy of Vaporization, Heat Capacities, Surface Tension,
Isothermal Compressibility, Volumetric Expansion Coefficient, and
Dielectric Constant. J. Chem. Theory Comput. 2012, 8, 61−74.
(51) van der Spoel, D.; van Maaren, P. J.; Caleman, C. GROMACS
Molecule & Liquid Database. Bioinformatics 2012, 28, 752−753.
(52) Horn, H. W.; Swope, W. C.; Pitera, J. W.; Madura, J. D.; Dick,
T. J.; Hura, G. L.; Head-Gordon, T. Development of an Improved
Four-Site Water Model for Biomolecular Simulations: TIP4P-Ew. J.
Chem. Phys. 2004, 120, 9665−9678.
(53) Bussi, G.; Donadio, D.; Parrinello, M. Canonical Sampling
through Velocity Rescaling. J. Chem. Phys. 2007, 126, No. 014101.
(54) Parrinello, M.; Rahman, A. Polymorphic Transitions in Single
Crystals: A New Molecular Dynamics Method. J. Appl. Phys. 1981, 52,
7182−7190.
(55) Fried, S. D.; Bagchi, S.; Boxer, S. G. Measuring Electrostatic
Fields in Both Hydrogen-Bonding and Non-Hydrogen-Bonding
Environments Using Carbonyl Vibrational Probes. J. Am. Chem. Soc.
2013, 135, 11181−11192.
(56) Chakrabarty, S.; Deshmukh, S. H.; Barman, A.; Bagchi, S.;
Ghosh, A. On−Off Infrared Absorption of the S�O Vibrational
Probe of Dimethyl Sulfoxide. J. Phys. Chem. B 2022, 126, 4501−4508.
(57) Nielsen, S. O. Hydrogen-deuterium exchange in N-methyl-
acetamide. Biochim. Biophysics. Acta 1960, 37, 146−147.

Recommended by ACS
Cationic Effects on the Structural Dynamics of the
Metal Ion–Crown Ether Complexes Investigated by
Ultrafast Infrared Spectroscopy
Dexia Zhou, Hongtao Bian, et al.
NOVEMBER 11, 2021
THE JOURNAL OF PHYSICAL CHEMISTRY B READ

Far-Ultraviolet Spectroscopy and Quantum Chemical

Calculation Studies of the Conformational Dependence
on the Electronic Structure and Transitions of Cyclo...
Yusuke Morisawa, Yukihiro Ozaki, et al.
SEPTEMBER 10, 2021
THE JOURNAL OF PHYSICAL CHEMISTRY A READ

Strong Hydrogen Bonds of Coordinated Ammonia

Molecules
Jelena M. Živković, Snežana D. Zarić, et al.
NOVEMBER 19, 2021
CRYSTAL GROWTH & DESIGN READ

Hydrogen-Bonding in Liquid Water at Multikilobar

Pressures
Hendrik Vondracek, Martina Havenith, et al.
AUGUST 16, 2019
THE JOURNAL OF PHYSICAL CHEMISTRY B READ

Get More Suggestions >

G https://doi.org/10.1021/acs.jpcb.2c01374
J. Phys. Chem. B XXXX, XXX, XXX−XXX
* Unknown * | ACSJCA | JCA11.2.5208/W Library-x64 | manuscript.3f (R5.1.i4:5009 | 2.1) 2021/10/27 08:51:00 | PROD-WS-121 | rq_3200684 | 6/03/2022 22:22:29 | 8 | JCA-DEFAULT

pubs.acs.org/JPCB Article

1 On−Off Infrared Absorption of the SO Vibrational Probe of

2 Dimethyl Sulfoxide

3 Suranjana Chakrabarty,∥ Samadhan H. Deshmukh,∥ Anjan Barman, Sayan Bagchi,* and Anup Ghosh*
Cite This: https://doi.org/10.1021/acs.jpcb.1c10558 Read Online

ACCESS Metrics & More Article Recommendations *

sı Supporting Information

4 ABSTRACT: Dimethyl sulfoxide (DMSO), a polar solvent

5 molecule, is used in a wide range of therapeutic and
6 pharmacological applications. Different intermolecular interactions,
7 such as dimerization and hydrogen bonding with water, are crucial
8 to understanding the role of DMSO in applications. Herein, we
9 study DMSO in various solvation environments to decipher the
10 environment-dependent dimerization and hydrogen-bonding pro-
11 pensity. We use a combination of infrared spectroscopy, quantum
12 mechanical calculations, and molecular dynamics simulations to
13 reach our conclusions. Although DMSO can exist in a dynamic
14 equilibrium between monomers and dimers, our results show that
15 the relative intensity of the SO stretch and the CH3 rocking
16 modes is a spectroscopic indicator of the extent of DMSO
17 dimerization in solution. The dimerization (self-association) is seen to be maximum in neat DMSO. When dissolved in different
18 solvents, the dimerization propensity decreases with increasing solvent polarity. In the presence of a protic solvent, such as water,
19 DMSO forms a hydrogen bond with the solvent molecules, thereby reducing the extent of dimerization. Further, we estimate the
20 hydrogen-bond occupancy of DMSO. Our results show that DMSO predominantly exists as doubly hydrogen-bonded in water.

21

22
■ INTRODUCTION
Dimethyl sulfoxide (DMSO) is one of the polar aprotic
dictate molecular behaviors, enhancing polarizability in the
molecules.15−17 Having the propensity toward intermolecular
48
49

23 molecules with a large dipole moment.1 In the field of associations, either with itself or with other molecules (such as 50
24 pharmacology, toxicology, biochemistry, and other analytical water), DMSO occasionally exhibits abnormal physicochemical 51
25 applications, DMSO is extensively used as a prospective properties and can form DMSO−water aggregates with various 52
26 cosolvent in various chemical reactions.1,2 Being a cryopro- stoichiometries.18−21 At a higher DMSO concentration, the 53
27 tective agent, DMSO averts the development of intracellular hydrogen-bond network of water is ruptured and stable water− 54
28 and extracellular crystals during the cell and tissue freezing DMSO complexes are formed.22−24 For a detailed exploration 55
29 processes.3−7 In addition, DMSO has various therapeutic and of the structures of DMSO, infrared (IR) absorption and 56
30 pharmaceutical uses, such as anti-inflammatory, analgesic, Raman scattering spectroscopies were employed.25−30 These 57
31 antibacterial, antifungal, antiviral, and membrane penetration experiments revealed that DMSO can form antiparallel dimers 58
32 enhancement properties that remain underexplored by and chain-like associations both in the pure liquid state and in 59
33 pharmaceutical developers and ophthalmologists.8 In bio- solutions.25−31 60
34 chemical and biophysical assays, it is reported that DMSO Although several vibrational spectroscopic reports on 61
35 concentrations influence protein aggregation and stability.9,10 DMSO have been published, a consensus regarding the peak 62
36 Experimental and computational studies revealed that DMSO assignments has yet to be reached.26,30 The band at 1027 cm−1 63
37 induces structural and conformational changes in pro- is assigned to the SO stretch arising from parallel chain-like 64
38 teins.11−13 These wide ranges of applications have generated molecular associations. For antiparallel cyclic associations, the 65
39 a profound interest in the structural behavior of DMSO, either bands at 1058 and 1044 cm−1 are attributed to the out-of- 66
40 neat or in the solution state, in recent years.14−25
41 The structural properties and the dimerization strategies of
42 DMSO have been previously reported.25−30 The intermolec- Received: December 14, 2021
43 ular geometry of DMSO in the liquid state, determined from Revised: May 24, 2022
44 X-ray and neutron diffraction studies, showed that DMSO
45 mostly prefers to arrange in the antiparallel ordering of SO
46 dipoles in the liquid state.14 The formation of hydrogen bonds
47 between DMSO and water molecules has been shown to

© XXXX American Chemical Society https://doi.org/10.1021/acs.jpcb.1c10558

A J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

67 phase and in-phase SO stretch modes, respectively.31 The spectroscopic approach to elucidate the details of the 130
68 out-of-phase SO stretch of the cyclic dimer is detected at microsolvation environment of DMSO despite the complexity 131
69 1063 cm−1 in the IR spectra.30 Toshiyuki Shikata et al. of overlapping bands in the spectra. 132

■
70 reported that associations of DMSO dimers are formed in the
71 pure liquid and also in the presence of nonpolar solvents.32 MATERIALS AND METHODS 133
72 When dissolved in CCl4, antiparallel cyclic dimers of DMSO
73 are formed due to limited self-interaction,33 with peaks at 1060 Sample Preparation. Dimethyl sulfoxide (DMSO 99.9%), 134

74 and 1000 cm−1 arising from the SO stretch of monomers thiophene oxide (TPSO, 99.9%), and all solvents (99.9%) were 135

75 and dimers, respectively.33 Recently, it has been reported that bought from Sigma Aldrich. Karl Fischer titration (coulo- 136

76 DMSO dimerization happens even at low DMSO concen- metric) was performed to quantify the moisture present in 137

77 trations.34 Gajda and Katrusiak used single-crystal XRD at high used solvents. These titrations indicated negligible water 138

78 pressure to study the detailed dimeric structure of DMSO.35 content. Solutions of DMSO and TPSO at a 0.1M 139

79 The calculated electrostatic potentials revealed a large positive concentration in all solvents were prepared for IR absorption 140

80 partial charge on the “S” atom and a negative partial charge on measurements. 141

81 the “O” atom, through which a variety of intermolecular IR Spectroscopy. All IR absorption spectra were collected 142

82 electrostatic interactions can occur.36 Natarajan et al. on a Bruker 70S FTIR spectrometer equipped with a DTGS 143

83 investigated the structure and stability of DMSO molecules detector at 2 cm−1 resolutions. A 50 μL volume of each 144

84 using molecular electrostatic potential and atoms-in-molecules solution was placed between two CaF2 windows separated by a 145

85 (AIM) analysis with the employment of density functional 64 μm Teflon spacer, housed in a homemade sample cell. 146

86 theory (DFT) studies.37 Along with the SO stretch region, the OH stretch region was 147

87 Another interesting aspect of DMSO is the structural and also monitored to identify any small amount of water 148

physicochemical behavior difference in aprotic and protic contamination. A flat baseline in the OH stretch region 149
88
indicated that the samples were not contaminated by water. 150
89 solvents. To this end, similar behavioral changes can be found
DFT Calculations. All initial geometries of DMSO, TPSO, 151
90 in molecules containing a carbonyl (CO) moiety where the
and their water complexes were optimized preliminarily by 152
91 less polarizable C atom replaces the S atom. Furthermore,
density functional theory and B3LYP/6-311G(d) calculations 153
92 studies on hydrogen-bond occupancy have shown that SO-
using the Gaussian 09 suite of programs to estimate the dipole 154
93 and CO-containing molecules can form one or more
moments of the DMSO monomer and the dimer.50,51 Further, 155
94 hydrogen bonds in a protic solvent.38−49 Recently, a
to expose the dissociation energies of one hydrogen bond and 156
95 temperature- and concentration-dependent study on DMSO/
two hydrogen bond complexes of DMSO in water, the “scan” 157
96 water mixtures indicated the presence of four distinct forms of
analyses were performed on the B3LYP/6-311G(d) optimized 158
97 DMSO, namely self-associated (aggregated), free-monomer
structures.52 159
98 (0HB), singly hydrogen-bonded (1HB), and doubly hydrogen-
MD Simulations. Classical molecular dynamics simulation 160
99 bonded (2HB) species.38,39 On the contrary, a recent study has
was conducted on the GROMACS 2016.5 package.53 161
100 shown that CO-containing esters are exclusively doubly
Parameters for DMSO were derived from quantum chemical 162
101 hydrogen-bonded in water.40 The reported differences in
calculation followed by RESP charge fitting by ANTECHAM- 163
102 structures’ hydrogen-bond occupancies upon changing the C
BER software to produce the general AMBER force field.54 164
103 atom in CO to the S atom in SO warrants a detailed study
Simulation box was prepared by adding 7 DMSO molecules 165
104 of DMSO in the solution state. Moreover, a significant SO
with 4000 water molecules. Here, water molecules were 166
105 IR absorbance signature for DMSO dimers, either neat or in
modeled using the TIP4P-EW model. For short-range 167
106 the solution state, has been surmised in the previous
electrostatic interactions, a periodic boundary condition with 168
107 studies.25−38 Such an observation cannot be explained based
a cutoff distance of 16 Å was used. Before the simulation, an 169
108 on the symmetric structure of the DMSO dimer.
energy minimization using a steepest-descent algorithm was 170
109 To identify the structural conformers underlying the performed, followed by equilibrium in the NVT ensemble at 171
110 overlapping DMSO vibrational spectra and to determine 300 K for 1 ns using a velocity rescale thermostat and 172
111 their structural details and hydrogen-bonding nature in water, equilibrium for the NPT ensemble at 300 K and 1 bar using a 173
112 we performed a detailed investigation using IR absorption Parrinello−Rahman barostat for 5 ns. LINCS was used to 174
113 measurements, DFT calculations, and MD simulations on neat constrain all covalently bonded hydrogen atoms. MD run was 175
114 DMSO and DMSO dissolved in various solvents, namely carried out for 10 ns. A hydrogen-bond analysis was performed 176
115 carbon tetrachloride (CCl4), dichloromethane (DCM), and on GROMACS with the last 5 ns of the trajectory where a 177
116 water (D2O). We combined IR spectroscopy with ab initio cutoff distance of 3.5 Å and an angle cutoff of 30° were used. 178

■
117 quantum mechanical calculations to analyze the monomer−
118 dimer equilibrium in different solvents to show that the relative
119 intensity of the SO stretch and the CH3 rocking modes RESULTS AND DISCUSSION 179

120 provide an estimate of the DMSO dimer present in the The delicate balance in the noncovalent (solute−solute and 180
121 solution. Furthermore, in the presence of a protic solvent like solute−solvent) interactions drives the formation and dis- 181
122 water, we showed that the extent of dimerization is drastically ruption of intermolecular interactions of DMSO dissolved in 182
123 reduced due to the formation of DMSO−water hydrogen different solvents. To resolve the detailed molecular structure 183
124 bonds. We subsequently analyzed the hydrogen-bonded of DMSO, in both neat and solution states, we performed a 184
125 conformations of DMSO in water. Interestingly, we observed series of infrared absorption experiments on 0.1 M DMSO in 185
126 that DMSO interacts predominantly with two water molecules different solvents (Figures 1 and 2). The second derivatives of 186 f1f2
127 to form a double hydrogen-bonded conformer. Such the broad overlapping IR absorption spectra (Figures 1b, and 187
128 conformations are well known for molecules containing the 2b,2d,2f) in the 950−1100 cm−1 region indicate four 188
129 CO moiety.40 Overall, our findings suggest an efficient underlying peaks. Each spectrum is fitted with four peaks 189

B https://doi.org/10.1021/acs.jpcb.1c10558
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

(Table S1). The fitted peaks in Figures 1 and 2 are assigned to 193
the SO symmetric stretching mode and three CH3 rocking 194
modes based on previous reports.31−34,38,39 In neat DMSO 195
(Figure 1), three rocking modes and the symmetric SO 196
mode are observed at ∼1019, 1030, 1045, and 1059 cm−1, 197
respectively. Interestingly, the intensity of the SO mode is 198
comparable to that of the CH3 rocking modes even though the 199
transition dipole moment of SO is larger than those of the 200
rocking modes. One plausible explanation for the comparable 201
intensities in Figure 1 can be that neat DMSO predominantly 202
exists as symmetric dimers even at low concentrations, as 203
reported previously.34 Theoretical calculations (discussed 204
later) suggest that the SO mode of antiparallel dimers is 205
not IR active. However, when dissolved in another solvent, the 206
dimers should not be formed at a low DMSO concentration. 207
Indeed, the IR absorption spectra of a dilute solution of 208
DMSO in different solvents show that the intensity of the S 209
O stretch band is distinctly larger than those of the CH3 210
rocking modes (Figure 2). 211
We performed DFT calculations at the B3LYP/6-311G(d) 212
level of theory50,51 to estimate the dipole moments of the 213
DMSO monomer and the dimer. Our calculations suggest that 214
the dipole moment is zero for an antiparallel DMSO dimer, 215
whereas a considerable dipole moment (5.6D) is estimated for 216
the monomer. We surmise that the SO peak in the IR 217
Figure 1. (a) IR absorption spectrum of neat DMSO (black). The spectrum (Figure 1) arises from the monomeric species, which 218
experimental spectrum was fitted to four peaks corresponding to the are in equilibrium with DMSO dimers in neat DMSO. The 219
SO stretch (orange) and the CH3 rocking modes (green, blue, and self-association of the monomers to form antiparallel dimers is 220
magenta) using Voigt line shape functions. During the fitting protocol, quite likely due to the high polarity of the SO bond. In 221
the peak positions are assigned based on the (b) second derivative addition, the dynamic monomer−dimer equilibrium happens 222
spectrum.
due to the intrinsic molecular fluctuations. 223
We further evaluated the ratios of SO peak areas to the 224
190 using Voigt line shape functions. The fits are shown in Figures integrated peak areas of the rest of the bands in the 950−1100 225
191 S1 and S2. The peak positions are fixed to the frequencies cm−1 region (Figure 3a). The ratio has a minimum value for 226 f3
192 corresponding to the minima in the second derivative spectra neat DMSO and subsequently increases with the polarity of 227

Figure 2. Black curves denote the IR absorption spectra of DMSO in (a) CCl4, (c) DCM, and (e) D2O. The experimental IR spectra are fitted to
four peaks using Voigt line shape functions. The orange peaks correspond to the SO stretch, and the green, blue, and magenta peaks correspond
to the CH3 rocking modes. The peak positions during the fitting protocol are obtained from the second derivative spectra of DMSO in (b) CCl4,
(d) DCM, and (f) D2O.

C https://doi.org/10.1021/acs.jpcb.1c10558
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

Figure 3. (a) Ratio (R) of the IR absorption area of the SO stretching mode and the integrated area of the rocking mode at around 1045 cm−1.
(b) Ratio (R) of the integrated IR absorption area of the SO stretching mode and the integrated area of all overlapping rocking modes.

228 the solvent. In water, the ratio is remarkably enhanced by a To confirm our arguments, we performed control IR 254
229 factor of ∼6 from that in neat DMSO. The polarity and the absorption experiments and DFT calculations on TPSO. The 255
230 hydrogen-bonding capacity of the solvent play a key role in the DFT calculations suggest that the bulky five-membered ring of 256
231 monomer−dimer equilibrium of DMSO. A nonpolar aprotic TPSO provided steric hindrance when two TPSO monomers 257

232 solvent like CCl4 or DCM weakly interacts with the solute approached one another, making the dimerization unfeasible 258

233 (DMSO), but a polar protic solvent like water can form even in neat TPSO. We performed a series of IR absorption 259

234 hydrogen bonds with the SO moiety of DMSO (see Figure experiments of TPSO and analyzed the 950−1100 cm−1 260

S3 in the Supporting Information). Therefore, the propensity spectral region (same frequency range as DMSO). The IR 261
235
spectra of neat TPSO show three peaks (Figure 5a), one 262 f5
236 of the monomeric species is higher in water than in other
arising from the SO stretch and the other two arising from 263
237 organic solvents. In neat DMSO, two polar DMSO molecules
CH2 rocking modes. Interestingly the absorption of the SO 264
238 can form the antiparallel dimer, giving rise to an increase in the mode was stronger than that of the rocking modes. Similar 265
239 dimeric species. This effect is corroborated by the ratios shown results are obtained when TPSO was dissolved at a low 266
240 in Figure 3. The trend remains the almost same if the area of concentration in other solvents (Figure 5b−d). Herein, we 267
241 one of the rocking modes is considered (Figure 3a) instead of fitted TPSO IR absorption spectra based on the quality (Figure 268
242 the integrated area (Figure 3b). The small discrepancy in the S4) and second-order derivative spectra (Figure S5). Figure 6 269 f6
243 ratios for CCl4 and DCM solvents (Figure 3b) most plausibly shows the ratios of the SO peak area to that of the rocking 270
244 arises due to the effect of the coupling between the SO mode (∼990 cm−1). The similar ratios in TPSO, CCl4, and 271
245 stretch and the CH3 rocking modes (see the Supporting DCM corroborate the DFT prediction of the dimerization 272
246 Information). However, the overall trend remains the same. infeasibility in neat TPSO. The lower polarity of TPSO when 273
247 To elucidate this phenomenon, we calculated the dimeriza- compared with DMSO, along with the steric hindrance in 274
248 tion energy of DMSO and compared it with solvation energy TPSO to the formation of hydrogen bonds might have resulted 275
f4 249 in organic and aqueous solvents (Figure 4). When compared in a lower ratio in water. However, the control experiments in 276

250 to aprotic solvents, Figure 4 clearly shows that the stability TPSO provide additional support for the formation of an 277

251 obtained from the dimerization of DMSO is much smaller than antiparallel DMSO dimer in a neat solution. 278

252 that obtained from the solvation of DMSO by water molecules, To resolve the hydrogen-bonded populations of DMSO, we 279

253 thereby increasing the ratio in water (Figure 3). further performed MD simulation with GROMACS (version 280
4.6.5) in neat water. The hydrogen-bond occupancy of DMSO 281
has been explored by keeping the angle of H-donor−acceptor 282
(H−O−O) between 0 and 30° and the O (donor)−O 283
(acceptor) distance was below 3.5Å.40 The force-field 284
parameters for DMSO were obtained from virtualchemis- 285
try.org.41 The TIP4P-Ew model was used for water.42 The 286
simulations reveal the hydrogen-bond formation and breaking 287
mechanisms of DMSO in neat water. The hydrogen-bond 288
analysis from the MD simulation trajectories demonstrates that 289
SO can form zero, one, or two HB complexes with the 290
surrounding water molecules.44−48 Interestingly, the free 291
DMSO (0.35%) and 1HB complexes (3.5%) are found to be 292
negligibly small in comparison to 2HB (96.15%) populations. 293
Similarly, TPSO predominantly forms 2HB populations (92%) 294
compared to 1HB (8%) and 0HB (0%) in neat water. These 295
results support our experimental results wherein we mentioned 296
Figure 4. Solvation energy (orange) and dimerization energy (green) only one SO stretching IR absorption peak for both DMSO 297
of DMSO in different solvents. (Figure 2e) and TPSO (Figure 5d) in neat water. 298

D https://doi.org/10.1021/acs.jpcb.1c10558
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

Figure 5. IR absorption spectra (a) of neat TPSO (black) and TPSO in (b) CCl4 (c) DCM, and (d) D2O. The experimental spectra were fitted to
three peaks corresponding to the SO stretch (red) and the CH2 rocking modes (green, blue) using Voigt line shape functions. During the fitting
protocol, the peak positions are assigned based on the second derivative spectrum shown in Figure S5.

Figure 6. IR absorption ratio of the SO stretching frequency and

the CH2 rocking mode (≈990 cm−1) of TPSO.

299 To further understand the stability of HB complexes of

Figure 7. Morse potential diagram of one hydrogen bond and two
300 DMSO in neat water, we performed density functional theory
hydrogen bond complexes of DMSO in water.
301 (DFT) calculations by employing Gaussian 09 software.50,51
302 The hydrogen-bond energies of 1HB and 2HB complexes are
303 obtained from the B3LYP-6-311G-optimized structures. To mol, respectively. The similar values of formation energy 316
304 calculate the hydrogen-bond dissociation energies of 1HB and indicate the equal probability of 1HB and 2HB formation. We 317
305 2HB bonds, we used the Morse potential energy calculation calculated the change of Gibbs free energy of both the 318
306 method (eq 1)52 hydrogen-bond formation from DFT calculations and finally, 319

V = De(1 − e −a(r − re) 2

) used it to calculate the equilibrium constants. Subsequently, 320
307 (1)
after calculation of K1 and K2 (7942.63 and 718.38/mol, 321
308 where r is the distance between the atoms, re is the equilibrium respectively), we calculated the ratio of C1HB/C2HB from the 322
309 hydrogen-bond distance, and De is the dissociation energy. The eqs 1−6 of Scheme S1. The negligible ratio (2.50 × 10−5) of 323
310 Morse potential diagrams show that the HB dissociation C1HB/C2HB is the indication of negligible 1HB populations 324
311 energies of 1HB and 2HB of DMSO are almost comparable compared to 2HB (Table 1). Quantitative SO H-bond 325 t1
f7 312 (Figure 7), which signifies that DMSO can form 1HB and 2HB populations in water are extracted for 1HB and 2HB 326
313 with equivalent probability, as the estimated energy difference complexes with values of 0.0025 and 99.9975%, respectively. 327
314 is almost negligibly small. The calculated formation energy Thus, the DFT analysis made our assessment a concrete 328
315 values of 1HB and 2HB complexes are −24.33 and −22.72 kJ/ conclusion. From our experimental and theoretical studies, a 329

E https://doi.org/10.1021/acs.jpcb.1c10558
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

Table 1. Thermodynamical Data Obtained from DFT spectra of TPSO, details of coupling, and mathematical 373
Calculations calculations of H-bond populations (PDF) 374

complexes
1HB
formation
energy
(kJ/mol)
−24.33
change of Gibbs
free energy
(kJ/mol)
−22.25
equilibrium
constant
(mol)
7942.63
populations
(%)
0.0025
■ AUTHOR INFORMATION
Corresponding Authors
375

376

2HB −22.72 −16.29 718.38 99.9975 Sayan Bagchi − Physical and Materials Chemistry Division, 377
CSIR-National Chemical Laboratory, Pune 411008, India; 378

330 clear parity of H-bond populations is undoubtedly recognized. Academy of Scientific and Innovative Research (AcSIR), 379

331 Subsequently, we confirmed conclusive 2HB populations of Ghaziabad 201002, India; orcid.org/0000-0001-6932- 380

332 DMSO in neat water, which is also logically explained in 3113; Email: s.bagchi@ncl.res.in 381

333 Scheme S1. Especially, in the mixing process, hydrogen Anup Ghosh − Department of Condensed Matter Physics and 382

334 bonding is simply transferred from water−water interactions Materials Sciences, S. N. Bose National Centre for Basic 383

335 (18.5 kJ/mol) to water−DMSO interactions as the H-bond Sciences, Kolkata 700106, India; Physical and Materials 384

336 interaction of water−water is lower than the H-bond Chemistry Division, CSIR-National Chemical Laboratory, 385

interaction of water−DMSO in neat water (22.72 kJ/mol). Pune 411008, India; Academy of Scientific and Innovative 386

■
337
Research (AcSIR), Ghaziabad 201002, India; orcid.org/ 387
338 CONCLUSIONS 0000-0002-1442-8740; Email: anupg86@gmail.com 388

339 In this work, we inspected the structural arrangements of Authors 389

340 DMSO molecules in the neat form and in solution. We showed Suranjana Chakrabarty − Department of Condensed Matter 390
341 that the presence of dimeric associations in neat DMSO Physics and Materials Sciences, S. N. Bose National Centre 391
342 decreases the IR intensity of the SO stretch, making the for Basic Sciences, Kolkata 700106, India 392
343 intensity comparable with the CH3 rocking modes. We Samadhan H. Deshmukh − Physical and Materials Chemistry 393
344 demonstrated that the zero net dipole moment of the Division, CSIR-National Chemical Laboratory, Pune 394
345 antiparallel dimer, in contrast to that of the monomer, can 411008, India; Academy of Scientific and Innovative 395
346 affect the overall intensity of the SO peak arising from Research (AcSIR), Ghaziabad 201002, India 396
347 monomer−dimer equilibrium. We showed that the dimeric Anjan Barman − Department of Condensed Matter Physics 397
348 configurations are disrupted with the increase in solvent and Materials Sciences, S. N. Bose National Centre for Basic 398
349 polarity and therefore the SO absorbance enhances when Sciences, Kolkata 700106, India; orcid.org/0000-0002- 399
350 compared to that of the rocking mode. The relative intensity of 4106-5658 400
351 the SO stretch and the rocking mode reaches the highest in
352 water as all dimeric structures of DMSO are disrupted by the Complete contact information is available at: 401

353 formation of hydrogen-bonded complexes. A single HB https://pubs.acs.org/10.1021/acs.jpcb.1c10558 402

354 conformation (i.e., 2HB) is found to be present in neat

Author Contributions 403
355 water from MD simulations. This result is in contrast with the ∥
356 previously published reports.25−39,44,45 Our experiments S.C. and S.H.D. contributed equally. 404

357 provide a detailed molecular understanding of the structure Notes 405

of neat DMSO and DMSO−water binary mixtures. Finally, our The authors declare no competing financial interest. 406

■
358
359 results provided an overview of the influence of solvent polarity
360 in the dissociation of DMSO dimers. We envisage that our ACKNOWLEDGMENTS 407
361 findings will help in future studies based on the solvent effects A.G. thanks SNBNCBS, Kolkata, for instrumental facilities and 408
362 on dynamical configurations of biomolecules in DMSO−- financial support from the DST, India. S.B. thanks CSIR-NCL 409
s1 363 water mixtures (Scheme 1). and SERB, India, (SR/S2/RJN-142/2012 and EMR/2016/ 410
000576) for financial support. 411

■
Scheme 1. Chemical Structure of the Sulfoxide Compound
REFERENCES 412
(1) Vaisman, I. I.; Berkowitz, M. L. J. Local structural order and 413
molecular associations in water-DMSO mixtures. Molecular dynamics 414
study. J. Am. Chem. Soc. 1992, 114, 7889−7896. 415
(2) Chalaris, M.; Marinakis, S.; Dellis, D. Temperature effects on the 416
structure and dynamics of liquid dimethyl sulfoxide: A molecular 417
dynamics study. Fluid Phase Equilib. 2008, 267, 47−60. 418

■
(3) Lovelock, J. E.; Bishop, M. W. H. Prevention of Freezing 419
Damage to Living Cells by Dimethyl Sulphoxide. Nature 1959, 183, 420
364 ASSOCIATED CONTENT 1394−1395. 421
365 *
sı Supporting Information (4) Rall, W. F.; Fahy, G. M. Ice-free cryopreservation of mouse 422
366 The Supporting Information is available free of charge at embryos at −196 °C by vitrification. Nature 1985, 313, 573−575. 423
367 https://pubs.acs.org/doi/10.1021/acs.jpcb.1c10558. (5) Fahy, G. M. Cryoprotectant toxicity neutralization. Cryobiology 424
2010, 60, S45−53. 425
368 Fitted linear IR spectra of neat DMSO, fitted linear IR (6) Rammler, D. H.; Zaffaroni, A. Biological implications of DMSO 426
369 spectra of DMSO in different solvents, table of linear IR based on a review of its chemical properties. Ann. N. Y. Acad. Sci. 427
370 absorption peak positions of DMSO and TPSO, fitted 1967, 141, 13−23. 428
371 linear IR spectra of DMSO in DCM and D2O mixtures, (7) Jacob, S. W.; Herschler, R. Pharmacology of DMSO. Cryobiology 429
372 fitted linear IR spectra of TPSO, second-order derivative 1986, 23, 14−27. 430

F https://doi.org/10.1021/acs.jpcb.1c10558
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

431 (8) Hoang, C.; Nguyen, A. K.; Nguyen, T. Q.; Fang, W.; Han, B.; (28) Fine, G.; Mirone, P. Short-range orientation effects in dipolar 500
432 Hoang, B. X.; Tran, H. D. Application of Dimethyl Sulfoxide as a aprotic liquids–HI. Intermolecular coupling of vibrations in sulfoxides, 501
433 Therapeutic Agent and Drug Vehicle for Eye Diseases. J. Ocul. sulfones, nitriles and other compounds. Spectrochim. Acta, Part A 502
434 Pharmacol. Ther. 2021, 37, 441−451. 1976, 32A, 625−629. 503
435 (9) Chan, D.-H.; Madeline, E. K.; Kirsty, J. M.; Andrew, W. M.; (29) Sastry, M. I. S.; Singh, S. J. Self-association of dimethyl 504
436 Dijana, M. V.; Anthony, G. C.; Chris, A. Effect of DMSO on Protein sulphoxide and its dipolar interactions with water: Raman spectral 505
437 Structure and Interactions Assessed By Collision-induced Dissocia- studies. J. Raman Spectrosc. 1984, 15, 80−85. 506
438 tion and Unfolding. Anal. Chem. 2017, 89, 9976−9983. (30) Skripkin, M. Y.; Lindqvist-Reis, P.; Abbasi, A.; Mink, J.; 507
439 (10) Tjernberg, A.; Markova, N.; Griffiths, W. J.; Hallén, D. DMSO- Persson, I.; Sandstrom, M. Basic materials physics of transparent 508
440 Related Effects in Protein Characterization. SLAS Discovery 2006, 11, conducting oxides. Dalton Trans. 2004, 4038−4049. 509
441 131−137. (31) Perelygin, I. S.; Itkulov, I. G.; Krauze, A. S. Temperature 510
442 (11) Ghosh, S.; Chattoraj, S.; Chowdhury, R.; Bhattacharyya, K. dependence of associative equilibria of DMSO according to Raman 511
443 Structure and dynamics of lysozyme in DMSO−water binary mixture: scattering spectra. Russ. J. Phys. Chem. 1991, 65, 410−414. 512
444 fluorescence correlation spectroscopy. RSC Adv. 2014, 4, 14378− (32) Shikata, T.; Sugimoto, N. Dimeric Molecular Association of 513
445 14384. Dimethyl Sulfoxide in Solutions of Nonpolar Liquids. J. Phys. Chem. A 514
446 (12) Roy, S.; Jana, B.; Bagchi, B. Dimethyl sulfoxide induced 2012, 116, 990−999. 515
447 structural transformations and non-monotonic concentration depend- (33) Figueroa, R. H.; Roig, E.; Szmant, H. H. Infrared study on the 516
448 ence of conformational fluctuation around active site of lysozyme. J. self-association of dimethyl sulfoxide. Spectrochim. Acta, Part A 1966, 517
449 Chem. Phys. 2012, 136, No. 115103. 22, 587−592. 518
450 (13) Sterling, H. J.; Prell, J. S.; Cassou, C. A.; Williams, E. R. Protein (34) Singh, S.; Srivastava, S. K.; Singh, D. K. Raman scattering and 519
451 conformation and supercharging with DMSO from aqueous solution. DFT calculations used for analyzing the structural features of DMSO 520
452 J. Am. Soc. Mass Spectrom. 2011, 22, 1178−1186. in water and methanol. RSC Adv. 2013, 3, 4381−4390. 521
453 (14) Onthong, U.; Megyes, T.; Bakó, I.; Radnai, T.; Grósz, T.; (35) Gajda, R.; Katrusiak, A. Electrostatic matching versus close- 522
454 Hermansson, K.; Probst, M. X-ray and neutron diffraction studies and packing molecular arrangement in compressed dimethyl sulfoxide 523
455 molecular dynamics simulation of liquid DMSO. Phys. Chem. Chem. (DMSO) polymorphs. J. Phys. Chem. B 2009, 113, 2436−2442. 524
456 Phys. 2004, 6, 2136−2144. (36) Clark, T.; Murray, J. S.; Lane, P.; Politzer, P. Why are dimethyl 525
457 (15) Soper, A. K.; Luzar, A. Orientation of Water Molecules around sulfoxide and dimethyl sulfone such good solvents? J. Mol. Model. 526
458 Small Polar and Nonpolar Groups in Solution: A Neutron Diffraction 2008, 14, 689−697. 527
459 and Computer Simulation Study. J. Phys. Chem. A 1996, 100, 1357− (37) Venkataramanan, N. S.; Suvithaa, A.; Kawazoea, Y. Density 528
460 1367. functional theory study on the dihydrogen bond cooperativity in the 529
461 (16) Mancera, R. L.; Chalaris, M.; Samios, J. The Concentration growth behavior of dimethyl sulfoxide clusters. J. Mol. Liq. 2018, 249, 530
462 Effect on the ‘Hydrophobic’ and ‘Hydrophilic’ Behaviour around 454−462. 531
463 DMSO in Dilute Aqueous DMSO Solutions. A Computer Simulation (38) Oh, K. I.; Rajesh, K.; Stanton, J. F.; Baiz, C. R. Quantifying 532
464 Study. J. Mol. Liq. 2004, 110, 147−153. hydrogen-bond populations in DMSO/water mixtures. Angew. Chem. 533
465 (17) Kirchner, B.; Hutter, J. Solvent Effects on Electronic Properties Int. Ed. 2017, 56, 11375−11379. 534
466 from Wannier Functions in a Dimethyl Sulfoxide/Water Mixture. J. (39) Oh, K. I.; Baiz, C. R. Crowding Stabilizes DMSO−Water 535
467 Chem. Phys. 2004, 121, 5133−5142. Hydrogen-Bonding Interactions. J. Phys. Chem. B 2018, 122, 5984− 536
468 (18) Borin, I. A.; Skaf, M. S. Molecular association between water 5990. 537
469 and dimethyl sulfoxide in solution: A molecular dynamics simulation (40) Ghosh, A.; Cohn, B.; Prasad, A. K.; Chuntonov, L. Quantifying 538
470 study. J. Chem. Phys. 1999, 110, 6412−6420. conformations of ester vibrational probes with hydrogen-bond- 539
471 (19) Luzar, A.; Chandler, D. Structure and hydrogen bond dynamics induced Fermi resonances. J. Chem. Phys. 2018, 149, 184501−184511. 540
472 of water dimethyl sulfoxide mixtures by computer simulations. J. (41) Ghosh, A. Vibrational Coupling on Stepwise Hydrogen Bond 541
473 Chem. Phys. 1993, 98, 8160−8173. Formation of Amide I. J. Phys. Chem. B 2019, 123, 7771−7776. 542
474 (20) Vishnyakov, A.; Lyubartsev, A. P.; Laaksonen, A. Molecular (42) Kashid, S. M.; Jin, G. Y.; Chakrabarty, S.; Bagchi, S.; Kim, Y. S. 543
475 Dynamics Simulations of Dimethyl Sulfoxide and Dimethyl Two-Dimensional Infrared Spectroscopy Reveals Cosolvent Compo- 544
476 Sulfoxide−Water Mixture. J. Phys. Chem. A 2001, 105, 1702−1710. sition-Dependent Crossover in Intermolecular HydrogenBond. J. 545
477 (21) Kirchner, B.; Reiher, M. The Secret of Dimethyl Sulfoxide− Phys. Chem. Lett. 2017, 8, 1604−1609. 546
478 Water Mixtures. A Quantum Chemical Study of 1DMSO−nWater (43) Kashid, S. M.; Jin, G. Y.; Bagchi, S.; Kim, Y. S. Cosolvent 547
479 Clusters. J. Am. Chem. Soc. 2002, 124, 6206−6215. Effects on Solute−Solvent Hydrogen-Bond Dynamics: Ultrafast 2D 548
480 (22) Baker, E. S.; Jonas, J. Transport and relaxation properties of IR Investigations. J. Phys. Chem. B 2015, 119, 15334−15343. 549
481 dimethyl sulfoxide-water mixtures at high pressure. J. Phys. Chem. B (44) Borin, I. A.; Skaf, M. S. Molecular Association between Water 550
482 1985, 89, 1730−1735. and Dimethyl Sulfoxide in Solution: The Librational Dynamics of 551
483 (23) Chowdhuri, S.; Pattanayak, S. K. Pressure dependence on the Water. Chem. Phys. Lett. 1998, 296, 125−130. 552
484 single-particle dynamics and hydrogen bond structural relaxation of (45) Luzar, A.; Chandler, D. Structure and Hydrogen-Bond 553
485 water-DMSO mixtures under ambient and cold conditions. Mol. Phys. Dynamics of Water-Dimethyl Sulfoxide Mixtures by Computer- 554
486 2013, 111, 135−146. Simulations. J. Chem. Phys. 1993, 98, 8160−8173. 555
487 (24) Kirchner, B.; Hutter, J. The Structure of a DMSO-Water (46) Kashid, S. M.; Singh, R. K.; Kwon, H.; Kim, Y. S.; Mukherjee, 556
488 Mixture from Car-Parrinello Simulations. Chem. Phys. Lett. 2002, 364, A.; Bagchi, S. Arresting an Unusual Amide Tautomer Using Divalent 557
489 497−502. Cations. J. Phys. Chem. B 2019, 123, 8419−8424. 558
490 (25) Forel, M. -T.; Tranquille, M. Spectres de vibration du (47) Kashid, S. M.; Singh, R. K.; Kwon, H.; Seol, J. G.; Kim, Y. S.; 559
491 diméthylsulfoxyde et dn diméthylsulfoxyde-d6. Spectrochim. Acta, Mukherjee, A.; Bagchi, S. Reply to “Comment on ‘Arresting an 560
492 Part A 1970, 26, 1023−1034. Unusual Amide Tautomer Using Divalent Cations’”. J. Phys. Chem. B 561
493 (26) Fawcett, W. R.; Kloss, A. A. Attenuated total reflection fourier- 2021, 125, 479−483. 562
494 transform IR spectroscopic study of dimethyl sulfoxide self-association (48) Wulf, A.; Ludwig, R. Structure and Dynamics of Water 563
495 in acetonitrile solutions. J. Chem. Soc., Faraday Trans. 1996, 92, Confined in Dimethyl Sulfoxide. ChemPhysChem. 2006, 7, 266−272. 564
496 3333−3337. (49) Ludwig, R.; Farrar, T. C. Temperature Dependence of the 565
497 (27) Fawcett, W. R.; Kloss, A. A. Solvent-Induced Frequency Shifts Deuteron and Oxygen Quadrupole Coupling Constants of Water in 566
498 in the Infrared Spectrum of Dimethyl Sulfoxide in Organic Solvents. J. the System Wated Dimethyl Sulfoxide. J. Phys. Chem. D 1994, 98, 567
499 Phys. Chem. C 1996, 100, 2019−2024. 6684−6687. 568

G https://doi.org/10.1021/acs.jpcb.1c10558
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

569 (50) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J.

570 Ab Initio Calculation of Vibrational Absorption and Circular
571 Dichroism Spectra Using Density Functional Force Fields. J. Phys.
572 Chem. A 1994, 98, 11623−11627.
573 (51) Becke, A. D. Density-functional thermochemistry. III. The role
574 of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652.
575 (52) Chakrabarty, S.; Maity, S.; Yazhini, D.; Ghosh, A. Surface-
576 Directed Disparity in Self-Assembled Structures of Small-Peptide L-
577 Glutathione on Gold and Silver Nanoparticles. Langmuir 2020, 36,
578 11255−11261.
579 (53) Abraham, M. J.; Murtola, T.; Schulz, R.; Pall, S.; Smith, J. C.;
580 Hess, B.; Lindahl, E. GROMACS: High performance molecular
581 simulations through multi-level parallelism from laptops to super-
582 computers. SoftwareX 2015, 1−2, 19−25.
583 (54) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D.
584 A. Development and testing of a general amber force field. J. Comput.
585 Chem. 2004, 25, 1157−1174.

H https://doi.org/10.1021/acs.jpcb.1c10558
J. Phys. Chem. B XXXX, XXX, XXX−XXX
nmt00 | ACSJCA | JCA11.2.5208/W Library-x64 | manuscript.3f (R5.1.i4:5009 | 2.1) 2021/10/27 08:51:00 | PROD-WS-121 | rq_2820701 | 2/08/2022 22:22:35 | 7 | JCA-DEFAULT

pubs.acs.org/JPCB Article

1 Hydrophobicity-Dependent Heterogeneous Nanoaggregates and

2 Fluorescence Dynamics in Room-Temperature Ionic Liquids

3 Published as part of The Journal of Physical Chemistry virtual special issue “Kankan Bhattacharyya Festschrift”.
4 Chayan K. De, Anup Ghosh, and Prasun K. Mandal*
Cite This: https://doi.org/10.1021/acs.jpcb.1c08598 Read Online

ACCESS Metrics & More Article Recommendations *

sı Supporting Information

5 ABSTRACT: The hydrophobicity of room-temperature ionic liquids (RTILs) has been shown to have a very significant effect on
6 the optical and structural properties of and in RTILs. The average excited state lifetime of neat RTILs has been shown to be
7 increasing with increasing hydrophobicity of the RTILs. By employing pico-nanosecond-based fluorescence anisotropy decay, the
8 volume of the nanoaggregates in neat RTILs have been calculated. The volume of these nanoaggregates have been shown to be
9 decreasing with increase in hydrophobicity of the RTILs. Thus, hydrophobicity has been shown to have an important role, i.e.,
10 hydrophobicity can be used as a handle to tune the properties of RTILs as designer solvents. Moreover, the excited-state lifetime of
11 red-emitting fluorophores, i.e., whose fluorescence emission is not perturbed by the inherent emission of RTILs, has been shown to
12 increase with the increasing hydrophobicity of the RTILs. Highly hydrophobic RTILs have been shown to exhibit positive deviation
13 and highly hydrophilic RTIL has been shown to exhibit negative deviation from the linear correlation between average solvation time
14 (τs) versus viscosity/temperature (η/T).

1. INTRODUCTION others, makes RTILs unique and complex in comparison to 34

15 Hydrophobicity is a unique property in the case of ionic and normal solvents.5−12 Varying the physical and chemical 35
16 polar room-temperature ionic liquids (RTILs). There have properties of RTILs have been obtained by the iterative 36
17 been reports about the synthesis and applications of hydro- combinations of various cations and anions. Various 37
18 phobic and polar ionic liquids,1−4 however, effect of hydro- experimental techniques like small-angle X-ray scattering, 38
19 phobicity toward inherent structural properties and fluores- neutron scattering, optical spectroscopy, as well as simulation 39
20 cence dynamics have not been investigated in detail. In this analyses have shown the existence of different types of 40
21 context, we have chosen a few well-known hydrophobic ionic mesoscopic local heterostructures in liquid state.12−24 Thus, 41
22 liquids containing (hexafluorophosphate) PF6 and FAP RTILs are known as ‘“nanostructured fluids”’.19 Although 42
23 (tris(pentafluoroethyl)trifluorophosphate) anions. Generally, visibly transparent, however, imidazolium-based RTILs possess 43
24 the hydrophobicity follows the following for the anions: BF4 significant absorption in the UV region, and the tail of 44
25 (tetrafluoroborate) < PF6 < FAP. We have investigated and absorption extends up to a part of the visible region.25,26 45
26 compared the optical and structural properties of the
27 hydrophobic RTILs with hydrophilic ionic liquids containing
28 BF4 anion keeping the same cation (either bmim (1-butyl-3- Received: September 30, 2021
29 methylimidazolium) or hmim (1-hexyl-3-methylimidazolium), Revised: January 30, 2022
c1 30 Chart 1). A few related physicochemical properties of these
31 RTILs have been summarized in Table S1.
32 The existence of a multitude of interactions, such as
33 Coulombic, dipolar, van der Waals, and H-bonding, among

© XXXX American Chemical Society https://doi.org/10.1021/acs.jpcb.1c08598

A J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

Chart 1. Chemical Structures of RTILs and Fluorophores goodness of the fit. Viscosity of the RTILs at different 91
Explored in This Study temperatures were recorded on Brookfield DV3T rheometer. 92
2.3. Method. For fluorescence anisotropy decay, parallel 93
(∥) and perpendicular (⊥) polarized fluorescence decays were 94
measured until their peak difference reaches 5000 count. Then 95
the G factor was calculated. Then the time-resolved anisotropy 96
decay (r(t)) was calculated with the equation 97

IVV(t ) − GIVH(t )
r (t ) =
IVV(t ) + 2GIVH(t ) (1) 98

where IVV(t) and IVH(t) denote the polarized emission 99

intensity when excitation and emission polarizers are kept in 100
vertical−vertical and vertical−horizontal arrangements, respec- 101
tively. G is a correction factor of instrument and it is magnitude 102
is equal to IHV(t)/IHH(t), that is fluorescence intensity ratio 103

46 Imidazolium RTILs exhibit excitation wavelength-dependent keeping the excitation polarizer in horizontal position, 104

47 emission maximum and fluorescence decay.25−29 alternating the emission polarizer in vertical and horizontal 105

48 Hydrophobicity is an important physical parameter which positions, respectively. Suitable exponential decay functions 106

49 controls several physical processes such as solubility, were used to fit the fluorescence decay/anisotropy decay 107

50 association of suitable functional groups, interaction and curves. Fitting of these decay curves were done using iterative 108

51 binding kinetics of ligands to the protein, protein folding, reconvolution program (IBH). 109

52 solvation, drug delivery, and so on.30,31 Many of these

53 processes are monitored by either spectral fluorescence or 3. RESULTS AND DISCUSSION
54 temporal fluorescence or fluorescence dynamical investiga- 3.1. Hydrophobicity-Dependent Excited State Life- 110
55 tions. RTILs have been used as solvent/media for many time. The steady-state absorption behavior and excitation 111
56 chemical as well as biological processes.32 Hence, there is a wavelength-dependent emission behavior of six RTILs in the 112
57 strong quest toward understanding the effect of hydrophobicity UV−vis region have been shown in Figures S1 and S2. The 113
58 on fluorescence decay, rotational anisotropy decay, solvation absorption and emission spectra of these RTILs are similar to 114
59 phenomenon, and so on. what has been reported in literature.25,26 Fluorescence decays 115
60 In this article, we report the results of the hitherto of these neat RTILs have been depicted Figures S3 and S4. 116
61 unexplored investigations: (i) How does the fluorescence These kinds of excitation wavelength and monitoring wave- 117
62 emission maximum of RTILs depend on the hydrophobicity of length depenedent fluorescence decay behavior are well-known 118
63 the RTILs? (ii) Whether and how does the fluorescence in RTILs.25−28 119
64 emission maximum is correlated with the nanoaggregates All these decay profiles could be fitted with a triexponential 120
65 inherently present in RTILs, and how does the size of the decay equation and thus three lifetimes have been obtained. 121
66 nanoaggregates depend on the hydrophobicity of the RTILs? Similar behavior has been reported earlier.33−47 The time 122
67 (iii) How does the hydrophobicity of the RTILs affect the constants of these fitted decay profiles have been depicted in 123
68 excited state lifetime of red emitting dyes (LD700 and OX725, Tables S2 and S3. Interestingly, keeping the cation same, i.e., 124
69 Chart 1)? (iv) How does the hydrophobicity of the RTILs bmim, as the hydrophobicity increases (BF4 < PF6 < FAP) the 125
70 controls the solvation phenomenon of C153 (Chart 1) in average excited state lifetime increases (Figure 1). This result 126 f1
71 RTILs? has been obtained for 375 nm excitation. Similar results have 127
been obtained for λex = 402 nm (Figure S5) 128
2. EXPERIMENTAL SECTION 3.2. Hydrophobicity-Dependent Size of the Nano- 129
72 2.1. Synthesis and Characterization. BF4 and PF6 aggregates. Presence of nanoaggregates has already been 130
73 anion-containing RTILs have been prepared and purified shown earlier by simulation and different experimental 131
74 following literature procedures.33−35 Highly pure FAP anion studies.21,43,48 However, how the size of the nanoaggregates 132
75 containing RTILs have been procured from Merck and were changes throughout the emission range of the RTILs is still 133
76 used as it is. unknown. To decipher the size of nanoaggregates in RTILs, 134
77 2.2. Instrumentation. Steady-state absorption spectra time-resolved fluorescence anisotropy decay has been carried 135
78 were recorded with a UV−vis spectrophotometer (Cary100, out in these neat RTILs. The rotational relaxation time (τr) 136
79 Varian). Steady-state fluorescence spectra were obtained on a obtained from time-resolved fluorescence anisotropy decay has 137
80 fluorimeter (Fluoromax-3). Time-resolved fluorescence decay been used for extracting the information about the size of the 138
81 and fluorescence anisotropy decay experiments were carried nanoaggregates.49−53 For a particular monitoring wavelength, 139
82 out using a time-correlated single-photon counting (TCSPC) fluorescence anisotropy decay has been measured by varying 140
83 spectrometer (5000, IBH). Three different diode lasers (375 the temperature from 298 to 338 K, and a single exponential 141
84 nm, 402 and 635 nm) were used as the excitation sources. The decay function has been observed to be adequate to fit the 142
85 instrument response function (IRF) for each excitation rotational anisotropy decay (Figure S6). 143
86 wavelengths were recorded using a scatterer, and the fwhm The stability of the nanoaggregates in RTILs has been 144
87 of the IRF was ∼70−100 ps. A nonlinear least-squares iteration verified in the temperature range of 298−338 K. Temperature- 145
88 procedure using IBH DAS 6 (version 2.2) decay analysis independent emission maxima justifies (see Figure S7) that 146
89 software was used for the fitting all the decay curves. χ2 values fluorescence is coming from same species across this 147
90 and the plot of residuals were used as parameters to check the temperature range. With increasing temperature, faster rotation 148

B https://doi.org/10.1021/acs.jpcb.1c08598
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

Figure 1. Monitoring wavelength-dependent average excited-state

lifetime (τAvg) of different RTILs. (blue) bmim[BF4], (red)
bmim[PF6], (green) bmim[FAP], and λex = 375 nm. The error in
the excited-state lifetime is less than ±5%.

149 is observed as the viscosity of the RTILs decreases (Table S4).

150 The temperature-dependent variation of rotational correlation
151 time has been analyzed via the Stokes−Einstein-Debye (SED)
152 equation:
ηVh
τr =
153 kT (2)
154 where η, k, and T represent the viscosity, Boltzmann constant
155 and temperature, respectively. A linear correlation between τr
156 and η/T has been found (Figure S8), as predicted by SED
157 theory. Hydrodynamic volume (Vh) has been calculated from
158 this equation. For a neat RTIL, the hydrodynamic volume
159 represents the size of the nanoaggregates present in that neat Figure 2. Variation of the volume of the nanoaggregates with
160 RTIL. Temperature-dependent anisotropy decays have been monitoring wavelength in different RTILs. The error in the
161 analyzed for different monitoring wavelengths. Interestingly, it nanoaggregated volume is within ±3%.
162 has been found that the hydrodynamic volume or nano-
163 aggregated volume differs for different monitoring wavelengths. Table 1. Variation of Nanoaggregated Volume (Å3) with
164 More specifically volume of the nanoaggregates increases with Monitoring Wavelength in Different RTILsa
f2 165 increasing monitoring wavelength for each neat RTILs (Figure λmon (nm) bmim[BF4] bmim[FAP] hmim[BF4] hmim[FAP]
f2t1 166 2, Table 1). This kind of hitherto unreported interesting 400 184.33 134.28 186.32 137.63
167 observation indicates that as the monitoring wavelength 430 199.97 144.63 205.64 170.07
168 increases we are gradually monitoring larger nanoaggregates, 460 212.80 165.36 227.56 185.32
169 in other words more and more stable species present in the 500 247.49 184.36 261.67 250.48
170 heterogeneous RTIL. 530 261.06 194.73 279.48 289.44
171 In Figure 2, the volume of the nanoaggregates has been 560 263.07 211.77 334.28 291.95
172 shown to increase with an increase in the monitoring a
λex = 375 nm.
173 wavelength. However, the extent of the increase is different
174 for different RTILs. For bmim-based RTILs (both BF4 and
175 FAP), the magnitude of the change in the volume of the conclusion can be drawn here that the hydrophobicity of the 187
176 nanoaggregates is about 70−80 Å3. However, for hmim-based anion has very significant control over the formation of the 188
177 RTILs (both BF4 and FAP), the magnitude of the change in nanoaggregates in RTILs (Table 1). 189
178 the volume of the nanoaggregates is about 150−160 Å3. This is We are aware that in neat RTILs a phenomenon like 190
179 perhaps because of the change in the size of the cation. resonance energy transfer (RET) may interfere with 191
180 Quite interestingly, as the hydrophobicity increases (say fluorescence anisotropy decay.54−57 However, looking at the 192
181 from bmim[BF 4] to bmim[FAP]) the volume of the decay profiles, we note the following: (a) We did not observe 193
182 nanoaggregates decreases (Figure 2a). This interesting any rise time related to RET while monitoring fluorescence 194
183 observation cannot be just because of the “size of the anion” decay ranging from blue end to red end; (b) moreover, the 195
184 effect as FAP has a larger size than that of BF4. A similar fluorescence anisotropy decay was single exponential in nature. 196
185 observation has been noted as we move from hmim[BF4] to (c) We never observed a “dip-and-rise” type of fluorescence 197
186 hmim[FAP] as well (Figure 2b). Thus, a very important anisotropy decay. Thus, if RET is contributing at all, then its 198

C https://doi.org/10.1021/acs.jpcb.1c08598
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

199 time constants would be less than those of IRF, i.e., less than
200 100 ps. The rotational correlation times we obtained are in the
201 range of several nanoseconds, and these values are at least 10−
202 60 times higher than the RET time, if RET is present at all.
203 Thus, further analyses based on rotational correlation time are
204 free from any significant perturbation from RET. RTILs are
205 known as “designer solvents” as the property of the RTILs
206 could be tuned by changing the nature of cation and anion.43
207 Quite interesting are the observations that the (i) average
208 excited state lifetime increases and (ii) the size of the
209 nanoaggregates decreases with increasing hydrophobicity of
210 the RTILs. Thus, hydrophobicity can be considered as another
211 handle toward tuning properties of RTILs as designer solvents.
212 3.3. Hydrophobicity-Dependent Excited State-Life-
213 time of Red-Emitting Dyes. As hydrophobicity has been
214 shown to be a useful handle to control structural as well as
215 optical property of neat RTILs, it was necessary to probe the
216 optical behavior of red-emitting fluorophores in RTILs of
217 varying hydrophobicity. We have chosen red-emitting
218 fluorophores so that inherent fluorescence of RTILs do not
219 cause any perturbation in the investigation. We have explored
220 two red-emitting fluorophores, namely, LD700 and OX725 in
221 six different RTILs, viz., bmim[BF4], bmim[PF6], [bmim-
222 [FAP], hmim[BF4], hmim[PF6], and hmim[FAP] (Chart 1).
223 These dyes are positively charged hence varying the anion
224 (with the same cation) thus varying the hydrophobicity would
225 provide important insight about the nature of nanoscopic
226 location these dyes reside into as well as nature of fluorescence
227 decay pathway these dyes follow in these RTILs. Both the red
228 emitting fluorophores exhibit single exponential fluorescence
229 decay behavior in all six RTILs for all three different excitation
230 wavelengths (Figures S9 and S10).
231 Please recall the hydrophobicity increases in the following
232 order: BF4 < PF6 < FAP. Interestingly, as the hydrophobicity
233 increases from bmim[BF4] to bmim[PF6] to bmim[FAP] the
234 excited state lifetime of LD700 has been observed to increase
f3 235 from 2.7 to 2.9 to 4.6 ns (Figure 3a). Similarly as the
236 hydrophobicity increases from hmim[BF4] to hmim[PF6] to
237 hmim[FAP], the excited state lifetime of LD700 has been
238 observed to increase (Figure 3b). These observations have Figure 3. Variation of excited state lifetime of LD700 in (a) bmim-
239 been noted for all three excitation wavelengths, i.e., 375, 402, based RTILs (1) bmim[BF4], (2) bmim[PF6], and (3) bmim[FAP],
t2 240 and 635 nm (Figure 3 and Table 2). Such observations have and in (b) hmim-based RTILs, (4) hmim[BF4], (5) hmim[PF6], and
241 been noted for OX725 in six different RTILs and for all three (6) hmim[FAP]. λem = 680 nm. The error in the excited-state lifetime
242 excitation wavelengths as well. is less than ±5%.
243 However, as the alkyl chain length increases from bmim to
244 hmim keeping the anion same, the fluorescence lifetime of
245 LD700 or OX725 does not change at all (Table 2). Thus, we S12). The time constants of the solvation dynamics have been 262
246 can conclude that both LD700 and OX725 could sense the depicted in Table S5. As can be seen from this table, the 263
247 hydrophobicity of the anion. This means that being positively average solvation time has been found to be dependent on the 264
248 charged these dyes occupy a region much closer to the anion viscosity of the RTILs. 265
249 and quite far from the alkyl chain length of the cation. Such Maroncelli and co-workers have shown that the average 266
250 hitherto unreported consistent hydrophobicity-dependent solvation time (τs) depends on the size of the cation and 267
optical behavior of red-emitting fluorophores in RTILs is follows the eq 3.47
iηy
268

τs ∝ jjj zzz × R+q

251
quite interesting. p

kT {
252
253 3.4. Hydrophobicity-Dependent Solvation in RTILs.
(3) 269
254 Solvation has been one of the most studied dynamical
255 phenomenon in RTILs.27,29,33−40,58−60 In this work, we have where τs is average solvation time, η is the viscosity of the 270
256 investigated solvation dynamics of C153 in two mostly RTIL at temperature T, R+ is the radius of the cation, p ∼ 1, 271
257 hydrophobic RTILs (bmim[FAP] and hmim[FAP]). Nature and q ∼ 4. To check whether hydrophobicity of the RTILs has 272
258 of the fluorescence decay of C153 in both bmim[FAP] and any role in solvation, we have plotted the average solvation 273
259 hmim[FAP] are dependent on monitoring wavelength (Figure time (τs) versus viscosity/temperature (η/T) for bmim cation 274
260 S11), and the time-resolved emission spectra showed time- based RTILs with varying anions (Figure 4). The solvation 275 f4
261 dependent shift of the fluorescence emission maximum (Figure time of RTILs other than FAP is taken from literature.47 It is 276

D https://doi.org/10.1021/acs.jpcb.1c08598
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

Table 2. Excited-State Time Constant of Fluorescence 4. CONCLUSIONS

Decay of LD700 and OX725 in Six RTILsa In conclusion, the hydrophobicity of RTILs has been shown to 289
LD700 OX725 have a very significant effect on the optical and structural 290
properties of and in room-temperature ionic liquids (RTILs). 291
RTILs λex (nm) τ (ns) χ2
τ (ns) χ2
The average excited-state lifetime of neat RTILs (bmim[BF4], 292
375 2.69 1.15 1.36 1.18 [bmim[PF6], and [bmim[FAP]) has been shown to be 293
bmim[BF4] 402 2.74 1.12 1.27 1.14 increasing (from 4.50 to 11.30 ns) with increasing hydro- 294
635 2.67 1.07 1.29 1.08 phobicity of the RTILs. Employing pico-nanosecond-based 295
375 2.86 1.18 1.84 1.06 fluorescence anisotropy decay, the volume of the nano- 296
bmim[PF6] 402 2.89 1.00 1.82 1.05 aggregates in neat RTILs (bmim[BF4], [bmim[FAP] and 297
635 2.82 1.02 1.84 1.07
hmim[BF4], [hmim[FAP]) have been calculated. The value of 298
375 4.58 1.04 2.62 1.30
these nanoaggregates have been shown to be decreasing (from 299
bmim[FAP] 402 4.61 1.17 2.92 1.33
248 Å3 to 185 Å3) with the increase in hydrophobicity of the 300
635 4.63 1.13 2.64 1.16
RTILs. Thus, hydrophobicity has been shown to be having 301
375 2.73 1.12 1.61 1.09
important role, i.e., hydrophobicity can be used as a handle to 302
hmim[BF4] 402 2.67 1.13 1.57 1.00
tune the properties of RTILs as designer solvents. Moreover, 303
635 2.74 1.09 1.51 1.01
excited state lifetime of red emitting fluorophores (LD700 and 304
375 3.04 1.13 2.07 1.04
OX725), i.e., whose fluorescence emission is not perturbed by 305
hmim[PF6] 402 3.05 1.08 2.06 1.16
the inherent emission of RTILs, has been shown to increase 306
635 2.93 1.00 1.99 1.04
(from 2.70 to 4.70 ns for LD700 and from 1.30 to 2.80 ns for 307
375 4.62 1.13 2.72 1.09
OX725) with increasing hydrophobicity of the RTILs. Highly 308
hmim[FAP] 402 4.68 1.18 2.75 1.05
hydrophobic RTILs have been shown to exhibit positive 309
635 4.61 1.07 2.59 1.17
deviation and highly hydrophilic RTIL has been shown to 310
a
Fluorescence decay has been monitored at the fluorescence emission exhibit negative deviation from the linear correlation between 311
maximum. Error in excited state lifetime is less than ±5%. average solvation time (τs) versus viscosity/temperature (η/ 312
T). 313

■
277 reported that the hydrophobicity in RTILs follows the order
278 FAP− > Tf2N− > PF6−> BF4−> Cl−.61 ASSOCIATED CONTENT 314

*
sı Supporting Information 315
The Supporting Information is available free of charge at 316
https://pubs.acs.org/doi/10.1021/acs.jpcb.1c08598. 317

Steady-state and time-resolved spectroscopic data 318

(PDF) 319

■ AUTHOR INFORMATION
Corresponding Author
320

321
Prasun K. Mandal − Department of Chemical Sciences and 322
Centre for Advanced Functional Materials, Indian Institute of 323
Science Education and Research (IISER) Kolkata, 324
Mohanpur, West Bengal 741246, India; orcid.org/0000- 325
0002-5543-5090; Email: prasunchem@iiserkol.ac.in 326

Authors 327
Chayan K. De − Department of Chemical Sciences, Indian 328
Institute of Science Education and Research (IISER) Kolkata, 329
Mohanpur, West Bengal 741246, India 330

Figure 4. Correlation between average solvation time τs, with η/T in Anup Ghosh − Department of Chemical Sciences, Indian 331
bmim-containing RTILs displaying varying hydrophobicities. Institute of Science Education and Research (IISER) Kolkata, 332
Mohanpur, West Bengal 741246, India; orcid.org/0000- 333
0002-1442-8740 334

Complete contact information is available at: 335

279 The linear correlation of average solvation time (τs) versus https://pubs.acs.org/10.1021/acs.jpcb.1c08598 336
280 viscosity/temperature (η/T) is expected specially for RTILs
281 having same cation, i.e., bmim. For values, see Table S6. Notes 337
282 However, as can be seen from the Figure 4, the positive The authors declare no competing financial interest. 338
283
284
285
deviation from linearity has been observed for RTILs having a
highly hydrophobic anion, i.e., [Tf2N] and [FAP], and a
negative deviation has been observed for RTIL having
■ ACKNOWLEDGMENTS
P.K.M. thanks IISER-Kolkata for financial help and instru-
339

340
286 hydrophilic anion [Cl]. Thus, hydrophobicity (or hydro- mental facilities. Support from the Fast-Track Project (SR/ 341
287 philicity) of RTIL has been shown to exhibit a significant role FT/CS-52/2011) of DST-India is gratefully acknowledged. 342
288 over the solvation phenomenon as well. Financial support from the SERB-DST India, Project No. 343

E https://doi.org/10.1021/acs.jpcb.1c08598
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

344 CRG/2019/003605 is gratefully acknowledged. C.K.D. thanks (20) Urahata, S. M.; Ribeiro, M. C. C. Structure of ionic liquids of 1- 411
345 DST- INSPIRE, AG thanks UGC for Fellowship. alkyl-3-methylimidazolium cations: A systematic computer simulation 412

■
study. J. Chem. Phys. 2004, 120, 1855−1863. 413
(21) Wang, Y.; Voth, G. A. Unique Spatial Heterogeneity in Ionic 414
346 REFERENCES Liquids. J. Am. Chem. Soc. 2005, 127, 12192−12193. 415
347 (1) Fukaya, Y.; Ohno, H. Hydrophobic and polar ionic liquids. Phys. (22) Canongia Lopes, J. N.; Costa Gomes, M. F.; Pádua, A. A. H. 416
348 Chem. Chem. Phys. 2013, 15, 4066−4072. Nonpolar, Polar, and Associating Solutes in Ionic Liquids. J. Phys. 417
349 (2) Yao, W.; Wang, H.; Cui, G.; Li, Z.; Zhu, A.; Zhang, S.; Wang, J. Chem. B 2006, 110, 16816−16818. 418
350 Tuning the Hydrophilicity and Hydrophobicity of the Respective (23) Raabe, G.; Köhler, J. Thermodynamical and structural 419
351 Cation and Anion: Reversible Phase Transfer of Ionic Liquids. Angew. properties of imidazolium-based ionic liquids from molecular 420
352 Chem., Int. Ed. 2016, 55, 7934−7938. simulations. J. Chem. Phys. 2008, 128, 154509. 421
353 (3) Zhao, X.; Zhou, K.; Zhong, Y.; Liu, P.; Li, Z.; Pan, J.; Long, Y.; (24) de Andrade, J.; Böes, E. S.; Stassen, H. Liquid-Phase Structure 422
354 Huang, M.; Brakat, A.; Zhu, H. Hydrophobic ionic liquid-in-polymer of Dialkylimidazolium Ionic Liquids from Computer Simulations. J. 423
355 composites for ultrafast, linear response and highly sensitive humidity Phys. Chem. B 2008, 112, 8966−8974. 424
356 sensing. Nano Res. 2021, 14, 1202−1209. (25) Paul, A.; Mandal, P. K.; Samanta, A. How transparent are the 425
357 (4) Wang, Y.; Leng, W.; Gao, Y.; Guo, J. Thermo-Sensitive Polymer- imidazolium ionic liquids? A case study with 1-methyl-3-butylimida- 426
358 Grafted Carbon Nanotubes with Temperature-Controlled Phase zolium hexafluorophosphate, [bmim][PF6]. Chem. Phys. Lett. 2005, 427
359 Transfer Behavior between Water and a Hydrophobic Ionic Liquid. 402, 375−379. 428
360 ACS Appl. Mater. Interfaces. 2014, 6, 4143−4148. (26) Paul, A.; Mandal, P. K.; Samanta, A. On the Optical Properties 429
361 (5) Welton, T. Room-Temperature Ionic Liquids. Solvents for of the Imidazolium Ionic Liquids. J. Phys. Chem. B 2005, 109, 9148− 430
362 Synthesis and catalysis. Chem. Rev. 1999, 99, 2071−2083. 9153. 431
363 (6) Wasserscheid, P.; Keim, W. Ionic liquids - new solutions for (27) Hu, Z.; Margulis, C. J. Room-Temperature Ionic Liquids: Slow 432
364 transitions metal catalysts. Angew. Chem., Int. Ed. 2000, 39, 3772− Dynamics, Viscosity, and the Red Edge Effect. Acc. Chem. Res. 2007, 433
365 3789. 40, 1097−1105. 434
366 (7) Dupont, J.; de Souza, R. F.; Suarez, P. A. Z. Liquid (Molten Salt) (28) Ghosh, A.; Chatterjee, T.; Mandal, P. K. On the heterogeneity 435
367 Phase Organometallic Catalysis. Chem. Rev. 2002, 102, 3667−3692. of fluorescence lifetime of room temperature ionic liquids: onset of a 436
368 (8) Rogers, R. D.; Seddon, K. R. Ionic liquids - solvents of the future. journey for exploring red emitting dyes. Chem. Commun. 2012, 48, 437
369 Science 2003, 302, 792−793. 6250−6252. 438
370 (9) Seddon, K. R. Ionic liquids: A taste of the future. Nat. Mater. (29) Mandal, P. K.; Samanta, A. Fluorescence Studies in a 439
371 2003, 2, 363−365. Pyrrolidinium Ionic Liquid: Polarity of the Medium and Solvation 440
372 (10) WeinGaertner, H. Understanding ionic liquids at the molecular Dynamics. J. Phys. Chem. B 2005, 109, 15172−15177. 441
373 level: facts, problems and controversies. Angew. Chem., Int. Ed. 2008, (30) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.; 442
374 47, 654−670. Springer-Verlag: New York, 2006. 443
375 (11) Hallett, J. P.; Welton, T. Room-Temperature Ionic Liquids: (31) Reichardt, C.; Welton, T. Solvents and Solvent Effects in Organic 444
376 Solvents for Synthesis and catalysis. Chem. Rev. 2011, 111, 3508−
Chemistry, 4th ed.; Wiley-VCH Verlag GmbH & Co. KgaA: 445
377 3576.
Weinheim, 2010. 446
378 (12) Holbrey, J. D.; Reichert, W. M.; Rogers, R. D. Crystal structures
(32) Egorova, K. S.; Gordeev, E. G.; Ananikov, V. P. Biological 447
379 of imidazolium bis (trifluoromethanesulfonyl)imide ionic liquid’ salts:
Activity of Ionic Liquids and Their Application in Pharmaceutics and 448
380 the first organic salt with a cis-TFSI anion conformation. Dalton
Medicine. Chem. Rev. 2017, 117, 7132−7189. 449
381 Trans. 2004, 2267−2271.
(33) Karmakar, R.; Samanta, A. Solvation Dynamics of Coumarin- 450
382 (13) Chen, S.-H.; Yang, F.-R.; Wang, M.-T.; Wang, N.-n. Synthesis,
153 in a Room-Temperature Ionic Liquid. J. Phys. Chem. A 2002, 106, 451
383 characterization, and crystal structure of several novel acidic ionic
384 liquids based on the corresponding 1-alkylbenzimidazole with 4447−4452. 452

385 tetrafluoroboric acid. C. R. Chimie. 2010, 13, 1391−1396. (34) Karmakar, R.; Samanta, A. Steady-State and Time-Resolved 453

386 (14) Henderson, W. A.; Fylstra, P.; De Long, H. C.; Trulove, P. C.; Fluorescence Behavior of C153 and PRODAN in Room-Temperature 454
387 Parsons, S. Crystal structure of the ionic liquid EtNH3NO3− insights Ionic Liquid. J. Phys. Chem. A 2002, 106, 6670−6675. 455

388 into the thermal phase behavior of protic ionic liquids. Phys. Chem. (35) Mandal, P. K.; Sarkar, M.; Samanta, A. Excitation wavelength- 456
389 Chem. Phys. 2012, 14, 16041−16046. dependent fluorescence behavior of the room temperature ionic 457
390 (15) Triolo, A.; Russina, O.; Bleif, H.-J.; Di Cola, E. Nanoscale liquids and dissolved dipolar solutes. J. Phys. Chem. A 2004, 108, 458
391 Segregation in Room Temperature Ionic Liquids. J. Phys. Chem. B 9048−9053. 459
392 2007, 111, 4641−4644. (36) Mandal, P. K.; Paul, A.; Samanta, A. Excitation-Wavelength- 460
393 (16) Hardacre, C.; Holbrey, J. D.; Mullan, C. L.; Nieuwenhuyzen, Dependent Fluorescence Behavior of Some Dipolar Molecules in 461
394 M.; Youngs, T. G. A.; Bowron, D. T. Liquid Structure of the Ionic Room-Temperature Ionic Liquids. J. Photochem. Photobiol. A: Chem. 462
395 Liquid, 1-Methyl-4-cyanopyridinium Bis{(trifluoromethyl)sulfonyl}- 2006, 182, 113−120. 463
396 imide Determined from Neutron Scattering and Molecular Dynamics (37) Samanta, A. Dynamic Stokes Shift and Excitation Wavelength 464
397 Simulations. J. Phys. Chem. B 2008, 112, 8049−8056. Dependent Fluorescence of Dipolar Molecules in Room Temperature 465
398 (17) Hardacre, C.; Holbrey, J. D.; Mullan, C. L.; Youngs, T. G. A.; Ionic Liquids. J. Phys. Chem. B 2006, 110, 13704−13716. 466
399 Bowron, D. T. Small-angle neutron scattering from 1-alkyl-3- (38) Mandal, P. K.; Saha, S.; Karmakar, R.; Samanta, A. Solvation 467
400 methylimidazolium hexafluorophosphate ionic liquids ([Cnmim]- dynamics in room temperature ionic liquids Dynamic stokes shift 468
401 [PF6], n = 4, 6, and 8). J. Chem. Phys. 2010, 133, 074510. studies of fluorescence of dipolar molecules. Curr. Sci. 2006, 90, 301− 469
402 (18) Mele, A.; Tran, C. D.; De Paoli Lacerda, S. H. D. P. The 310. 470
403 structure of a room-temperature ionic liquid with and without trace (39) Samanta, A. Solvation Dynamics in Ionic Liquids: What We 471
404 amounts of water: The role of C-H···O and C-H···F interactions in 1- Have Learned from the Dynamic Fluorescence Stokes Shift Studies. J. 472
405 n-butyl-3-methylimidazolium tetrafluoroborate. Angew. Chem., Int. Ed. Phys. Chem. Lett. 2010, 1, 1557−1562. 473
406 2003, 42, 4364−4366. (40) Zhang, X.-X.; Liang, M.; Ernsting, N. P.; Maroncelli, M. 474
407 (19) Mele, A.; Romano, G.; Giannone, M.; Ragg, E.; Fronza, G.; Complete Solvation Response of Coumarin 153 in Ionic Liquids. J. 475
408 Raos, G.; Marcon, V. The local structure of ionic liquids: cation-cation Phys. Chem. B 2013, 117, 4291−4304. 476
409 NOE interactions and internuclear distances in neat [BMIM][BF4] (41) Ghosh, A.; Chatterjee, T.; Roy, D.; Das, A.; Mandal, P. K. On 477
410 and [BDMIM][BF4]. Angew. Chem., Int. Ed. 2006, 45, 1123−1126. the nanoscopic environment a neutral fluorophore experiences in 478

F https://doi.org/10.1021/acs.jpcb.1c08598
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 The Journal of Physical Chemistry B pubs.acs.org/JPCB Article

479 room temperature ionic liquids. J. Phys. Chem. C 2014, 118, 5051− (59) Roy, D.; Maroncelli, M. Simulations of Solvation and Solvation 547
480 5057. Dynamics in an Idealized Ionic Liquid Model. J. Phys. Chem. B 2012, 548
481 (42) Ghosh, A.; De, C. K.; Chatterjee, T.; Mandal, P. K. What type 116, 5951−5970. 549
482 of nanoscopic environment does a cationic fluorophore experience in (60) Maroncelli, M.; Zhang, X.; Liang, M.; Roy, D.; Ernsting, N. P. 550
483 room temperature ionic liquids? Phys. Chem. Chem. Phys. 2015, 17, Measurements of the complete solvation response of coumarin 153 in 551
484 16587−16593. ionic liquids and the accuracy of simple dielectric continuum 552
485 (43) Majhi, D.; Sarkar, M. Probing the microscopic structural predictions. Faraday Discuss. 2012, 154, 409−424. 553
486 organization of neat ionic liquids (ILs) and ionic liquid-based gels (61) Huddleston, J. G.; Visser, A. E.; Reichert, W. M.; Willauer, H. 554
487 through resonance energy transfer (RET) studies. Phys. Chem. Chem. D.; Broker, G. A.; Rogers, R. D. Characterization and Comparison of 555
488 Phys. 2017, 19, 23194−23203. Hydrophilic and Hydrophobic Room Temperature Ionic Liquids 556
489 (44) Arzhantsev, S.; Ito, N.; Heitz, M.; Maroncelli, M. Solvation Incorporating the Imidazolium Cation. Green Chem. 2001, 3, 156− 557
490 Dynamics of Coumarin 153 in Several Classes of Ionic Liquids: 164. 558
491 Cation Dependence of the Ultrafast Component. Chem. Phys. Lett.
492 2003, 381, 278−286.
493 (45) Ito, N.; Arzhantsev, S.; Maroncelli, M. The probe dependence
494 of solvation dynamics and rotation in the ionic liquid 1-butyl-3-
495 methyl-imidazolium hexafluorophosphate. Chem. Phys. Lett. 2004,
496 396, 83−91.
497 (46) Yan, T.; Wang, Y.; Knox, C. Solvation and Rotational Dynamics
498 of Coumarin 153 in Ionic Liquids: Comparisons to Conventional
499 Solvents. J. Phys. Chem. B 2010, 114, 6905−6921.
500 (47) Jin, H.; Baker, G. A.; Arzhantsev, S.; Dong, J.; Maroncelli, M.
501 Solvation and Rotational Dynamics of Coumarin 153 in Ionic
502 Liquids: Comparisons to Conventional Solvents. J. Phys. Chem. B
503 2007, 111, 7291−7302.
504 (48) Wang, Y.; Voth, G. A. Tail Aggregation and Domain Diffusion
505 in Ionic Liquids. J. Phys. Chem. B 2006, 110, 18601−18608.
506 (49) Das, A.; Roy, D.; De, C. K.; Mandal, P. K. Where does the
507 fluorescing moiety reside in a carbon dot?” - Investigations based on
508 fluorescence anisotropy decay and resonance energy transfer
509 dynamics. Phys. Chem. Chem. Phys. 2018, 20, 2251−2259.
510 (50) Das, S. K.; Sarkar, M. Rotational dynamics of coumarin-153
511 and 4-aminophthalimide in 1-ethyl-3-methylimidazolium alkylsulfate
512 ionic liquids: effect of alkyl chain length on the rotational dynamics. J.
513 Phys. Chem. B 2012, 116, 194−202.
514 (51) Das, S. K.; Majhi, D.; Sahu, P. K.; Sarkar, M. Investigation of
515 the influence of alkyl side chain length on the fluorescence response of
516 C153 in a series of room temperature ionic liquids. RSC Adv. 2015, 5,
517 41585−41594.
518 (52) Prabhu, S. R.; Dutt, G. B. Effect of Low Viscous Nondipolar
519 Solvent on the Rotational Diffusion of Structurally Similar Nondipolar
520 Solutes in an Ionic Liquid. J. Phys. Chem. B 2015, 119 (5), 2019−
521 2025.
522 (53) Ghosh, A.; De, C. K.; Chatterjee, T.; Das, A.; Roy, D.; Routh,
523 T.; Mandal, P. K. Correlation between size of nano-aggregates and
524 excitation wavelength dependent fluorescence emission in room
525 temperature ionic liquids: A case study with emim[FAP]. Chem. Phys.
526 Impact 2021, 3, 100054.
527 (54) Ghiggino, K. P.; Reek, J. N. H.; Crossley, M. J.; Bosman, A. W.;
528 Schenning, A. P. H. J.; Meijer, E. W. The Dynamics of Electronic
529 Energy Transfer in Novel Multiporphyrin Functionalized Dendrimers:
530 A Time-Resolved Fluorescence Anisotropy Study. J. Phys. Chem. B
531 2000, 104, 2596−2606.
532 (55) Silori, Y.; De, A. K. Tuning effect of local environment to
533 control mechanism of fluorescence depolarization: Rotational
534 diffusion and resonance energy transfer within homo-aggregates of
535 xanthenes. J. Photochem. Photobiol. A: Chem. 2019, 377, 198−206.
536 (56) Xu, Q.-H.; Wang, S.; Korystov, D.; Mikhailovsky, A.; Bazan, G.
537 C.; Moses, D.; Heeger, A. J. The fluorescence resonance energy
538 transfer (FRET) gate: A time-resolved study. Proc. Nat. Acad. Sci. USA
539 2005, 102, 530.
540 (57) Warren, S. C.; Margineanu, A.; Katan, M.; Dunsby, C.; French,
541 P. M. W. Homo-FRET Based Biosensors and Their Application to
542 Multiplexed Imaging of Signalling Events in Live Cells. Int. J. Mol. Sci.
543 2015, 16, 14695−14716.
544 (58) Roy, D.; Patel, N.; Conte, S.; Maroncelli, M. Dynamics in an
545 Idealized Ionic Liquid Model. J. Phys. Chem. B 2010, 114, 8410−
546 8424.

G https://doi.org/10.1021/acs.jpcb.1c08598
J. Phys. Chem. B XXXX, XXX, XXX−XXX
 Chemical Physics Impact 3 (2021) 100054

Contents lists available at ScienceDirect

Chemical Physics Impact

journal homepage: www.sciencedirect.com/journal/chemical-physics-impact

Correlation between size of nano-aggregates and excitation wavelength

dependent fluorescence emission in room temperature ionic liquids: A case
study with emim[FAP]
Anup Ghosh a, Chayan K. De a, Tanmay Chatterjee a, Ananya Das a, Debjit Roy a, Tapan Routh a,
Prasun K. Mandal a, b, *
a
Department of Chemical Sciences, Indian Institute of Science Education and Research (IISER) Kolkata, Mohanpur, West Bengal, India 741246
b
Centre for Advanced Functional Materials, Indian Institute of Science Education and Research (IISER) Kolkata, Mohanpur, West Bengal, India 741246

A R T I C L E I N F O A B S T R A C T

Keywords: Excitation wavelength dependent emission maximum (400 - 560 nm) of a room temperature ionic liquid has been
Ionic liquid investigated employing time-resolved fluorescence anisotropy based rotational dynamical analyses. Employing
Nano-aggregate Stokes-Einstein-Debye relationship we could show that nano-aggregate having aggregation volume of ~250 Å3,
Rotational dynamics
~350 Å3, ~400 Å3 and ~500 Å3 emit at ~400 nm, ~450 nm, ~500 nm, and ~560 nm respectively.
SED equation
Size dependent emission

Research on Room Temperature Ionic Liquids (RTILs) has been
increasing rapidly because of their interesting properties like low
melting point, high thermal stability, negligible vapor pressure, recy­
clability, high ionic conductivity etc. [1–7]. Thus RTILs have been
employed in separation, catalysis and energy related applications like
batteries, fuel cell photovoltaics, super-capacitors etc. [1–7]. Single
crystal XRD, neutron scattering, Raman spectroscopic study, small-wide
angle X-ray scattering, simulation studies etc. reveal microscopic het­
erogeneity in the structure of RTIL [8–28]. Structural heterogeneities
have been shown to be ranging from Angstrom (Å) to a few nanometre
(nm) in the spatial scale as a consequence of the segregation of the alkyl
tails into mesoscopic domains [29–31]. Such an organization might be
responsible for observed unique properties of RTILs like their ability to
dissolve both polar and not so polar compounds, anomalous trend for
the physical properties like viscosity upon increasing the alkyl chain
length of the imidazolium cation, [1–7] excitation wavelength depen­
dent emission behavior of neat RTILs [32–34] and complex solvation
dynamics [35–42] of dissolved dipolar solutes.
Out of all these properties, excitation wavelength dependent fluo­
rescence emission is still very intriguing and bear interesting analogy
with size dependent emission maximum of semiconductor quantum dots
(QDs) (say CdSe, InP etc.) [43–56]. Thus, using same chemical compo­
sition but by varying the size of the QDs it is possible to cover the entire
range of the UV-–Vis-NIR region of the electromagnetic spectrum.
Neat RTILs exhibit excitation wavelength dependent emission
behavior which is known as red edge effect [33,34]. It has been invoked
that different nano-aggregates behave as different species and their
ground state and excited state (electronic) energy levels are different
[33–34] Lack of energy transfer between these nanostructures causes
excitation wavelength dependent emission maximum in RTILs. It has
been shown that fluorescence decay behavior of neat RTILs are multi­
exponential in nature and dependent on both excitation and emission
wavelength [56.57] Average excited state lifetimes have been shown to
vary by a factor of three or more [57]. This observed behavior has been
ascribed to the absence of energy transfer among the different
nano-aggregating emitting unit inherently present in RTILs [51]. It has
been speculated that perhaps different (sized) nano-aggregates are
responsible for the said behavior. However, detailed effort in order to
quantify the size of the nano-aggregated species present in the liquid
state and it’s correlation with excitation wavelength dependent fluo­
rescence emission maximum is lacking. Although several rotational
anisotropic studies of the dissolved fluorophores have been carried out
in RTILs, however, the same measurements in neat RTILs are very few
[58–69].
Using fast rotational dynamics we have probed the existence of
different nano-aggregates in the liquid structure of RTILs and measured
the size of the nano-aggregating units. Moreover, whether there is any
correlation between the size of the nano-aggregating species with the

* Corresponding author.
E-mail address: prasunchem@iiserkol.ac.in (P.K. Mandal).

https://doi.org/10.1016/j.chphi.2021.100054
Received 15 September 2021; Received in revised form 31 October 2021; Accepted 9 November 2021
Available online 22 November 2021
2667-0224/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
A. Ghosh et al. Chemical Physics Impact 3 (2021) 100054

emission maximum was also probed into. In order to shed light in that
direction we have performed a case study using 1-Ethyl-3-methylimida­
zolium tris(pentafluoroethyl)trifluorophosphate (emim[FAP]) (see
Chart 1 and Table S0 for physico-chemical properties) as a model RTIL
due to its interesting photo-physical and physiochemical property [69,
70]. The choice of using this particular RTIL stems from the fact that this
RTIL is highly hydrophobic and less viscous in nature [28,56,57]. emim
[FAP] showed a wider electrochemical window, despite lower viscosity,
larger self-diffusion coefficients and higher ionic conductivity. emim
[FAP] has weaker inter-ionic interactions compared to the emim+ with
smaller anion like [PF6], [BF4], halides [70]. We have investigated
rotational dynamics of the neat RTIL, emim[FAP] as a function of tem­
perature. We have explored how the hydrodynamic volume of the
different nano-aggregating emitting unit in neat RTIL changes with
excitation and monitoring wavelengths.
As we have planned to perform temperature dependent fluorescence
dynamical experiments so we had to first do a steady state control
measurement in order to check that the RTIL structure remains intact in Fig. 1. Temperature independence of variation of emission maximum against
the entire temperature range (298 K to 338 K) of the experiment. As a excitation wavelength.
first step we investigated whether the absorption spectrum gets modified
with change in temperature. As can be seen (see Fig. S1) the absorption correlation time, and viscosity of the medium respectively.
spectrum remains same in the entire temperature range. A more sensi­
tive experiment is to check whether the excitation wavelength depen­ ηVh
τr = (1)
dent emission maximum of RTIL change with change in temperature. As kB T
can be seen from Fig. 1, the excitation wavelength dependent emission With increasing temperature rotational correlation time has been
maximum does not change as we vary the temperature from 298 K to observed to decrease gradually. These are not due to the breakdown of
338 K. We would like to recall a report based on MD computer simu­ the nano-aggregation because emission maximum does not change with
lation of temperature dependent behavior of RTIL [23]. It has been change temperature (Fig. 1& Supp Info Figure S1). Rather this obser­
shown that static dielectric constant of bmim[PF6] does not change vation is due to the decreasing the viscosity of the medium.
significantly with change in temperature [17]. The ultrafast inertial According to Eq. (1), rotational diffusion of a solute not only depends
component of solvation also remains unchanged. The cation-anion on the viscosity but also on the size of the solute. In order to investigate
structure and the potential of mean force relevant to the relative mo­ whether rotational anisotropy of neat emim[FAP] depends on moni­
tion of cations and anions in the radial direction, i.e. vibration as an ion toring wavelength or not we have probed different monitoring wave­
pair do not get altered substantially with change in temperature [23] lengths. Correlation between τr and η/T has been depicted in Fig. 2 and
Thus, our experimental result as well as reported MD simulation result the corresponding rotational correlation time constants have been
show that the nature of the nano-aggregates does not get altered in the depicted in Table 1. As can be seen from Fig. 2 as well as from Table 1,
temperature range of 298 K to 338 K. rotational anisotropy decay becomes more slower with increasing the
We have measured rotational anisotropy of neat emim[FAP] varying monitoring wavelength from 400 nm to 560 nm for a fixed excitation
both excitation and emission wavelengths over the temperature range wavelength (say 377 nm). This means that rotational correlation time
298 K to 338 K (see S2 to S8). Experimentally measured reorientation increases with increasing the monitoring wavelength. As can be seen
times of neat emim[FAP] have been analyzed using Stokes-Einstein- from linear correlation plot of τr and η/T depicted in Fig. 2, SED equa­
Debye (SED) hydrodynamic theory. According to the SED theory, the tion can depict the situation quite well for emim[FAP]. From Table 1 we
solute reorientation time is related to the bulk viscosity of the solvent could notice that at lower temperature (298 K) rotational correlation
according to eqn. (1) (shown below), where kB, T, Vh, τr, η are the time (θ) increases by a factor of ~1.8 (with change of monitoring
Boltzmann constant, absolute temperature, the hydrodynamic volume of wavelength from 400 nm to 560 nm), whereas, at higher temperature
the fluorescing unit, (nano-aggregates in the present case), rotational

Chart 1. Chemical structure of RTIL explored.

Fig. 2. Stokes-Einstein-Debye (SED) plot of emim[FAP] for different moni­
toring wavelengths (λex=377 nm).

2
A. Ghosh et al. Chemical Physics Impact 3 (2021) 100054

Table 1
Rotational correlational time (τr) at different temperatures, volume or radius of nano-aggregates.
λex (nm) λem (nm) τr 298K (ns) τr 308K (ns) τr 318K (ns) τr 328K (ns) τr 338K (ns) Volume of nano-aggregate (Å3) Radius of the nano-aggregate (Å)

377 400 3.78 2.99 2.05 1.61 1.14 238.85 3.84

420 4.27 3.17 2.24 1.70 1.26 271.98 4.01
450 5.38 3.81 2.50 1.96 1.51 356.20 4.39
500 5.90 4.07 2.68 2.12 1.66 389.34 4.53
530 6.41 4.60 2.97 2.16 1.70 437.66 4.71
560 6.90 4.87 2.98 2.24 1.71 483.27 4.86

(338 K) rotational correlation time has been increased by a factor of

~1.5 for the same monitoring wavelength range. Data shown in Fig. 2
could be fitted using eqn. (1) and the slope of the linear fit would provide
the hydrodynamic volume of the nano aggregates. It is well documented
in literature that the RTILs do form a nano-micro structures in liquid
state [14,22]. Assuming a spherical shape of the nano-aggregates, the
radius of the nanoaggregating units has been calculated for a particular
excitation wavelength (say 377 nm) the magnitude of the hydrodynamic
volume of the emitting species has been found to increase with the in­
crease in the monitoring wavelength (Table 1).
Having seen that the size of the nanoaggregates vary with the
monitoring wavelengths we wanted to probe whether the size of the
nanoaggregates vary with excitation wavelengths. Thus, the experi­
ments were carried out in three excitation wavelengths. In Fig. 3, we
have plotted rotational correlation time against η/T for three different
excitation and emission wavelengths and five different temperatures.
For lower T, η is high and differently sized nano-aggregates exhibit
different correlation time (for different excitation and monitoring
wavelengths). As T increases, η becomes lower, because of high thermal
energy difference in the rotational correlation time narrows down. As
can be seen (Fig. 3) the size of the nanoaggregates vary with the exci­
tation wavelengths keeping the respective emission maximum as the
monitoring wavelengths for each excitation wavelength. This interesting Fig. 4. Variation of emission wavelength with size of nano-aggregates in emim
observation points to the fact that with change in excitation wavelengths [FAP]. A linear fit has also been shown.
varying proportion of differently sized nano-aggregates get excited.
Finally we wanted to probe whether there is any correlation between and size (volume) of the nano-aggregates could be obtained.
the size of the nano-aggregates (obtained from SED equation) and the An aggregation volume of ~250 Å3, ~450 Å3, ~500 Å3 and ~560 Å3
emission maximum. Data (volume of the nano-aggregates) from Table 1 could be correlated with the fluorescence emission maximum of ~400
have been plotted in Fig. 4a. As can be seen from this figure as the size nm, ~450 nm, ~500 nm, and ~560 nm respectively (Fig. 4b).
(volume) of the nano-aggregating units increases emission maximum Computational and simulations results show that for imidazolium RTILs,
also increases and a near-linear correlation between emission maximum the cation-cation distance varies from 3 to 12 Å (Table S12 in SI). Thus
qualitative we can mention that for aggregation volume of ~250 Å3,
~350 Å3, ~400 Å3 and ~500 Å3 there are two, three, four, and five
imidazolium rings present respectively, within the aggregation volume,
which in turn emit at ~400 nm, ~450 nm, ~500 nm, and ~560 nm
respectively. Such nano-aggegated size dependent fluorescence emis­
sion maximum in RTILs is hitherto unreported.
Quite interestingly, in case of QDs the photoluminescence emission
maximum increases with increase in size of the QD. In the current case of
RTILs, we have observed a similar behavior that the emission wave­
length increases with increase in the size of the nano-aggregates. Such a
direct correlation is very interesting; however, it should be repeated for
more RTILs before concluding that this is a natural observation for
RTILs. Such experiments are currently underway.
There has been speculation that emission in RTIL is because of im­
purities. An impurity has a particular size and the size does not vary with
excitation or monitoring wavelengths. For an impurity reorientational-
correlation-time won’t vary with excitation or emission wavelengths.
Had the emission been because of an impurity the emission maximum
would not have shifted with excitation wavelengths as well. In this
Fig. 3. Stokes-Einstein-Debye (SED) plot of emim[FAP]. Excitation and emis­ detailed investigation, we have obtained different sizes of the emitting
sion wavelengths have been mentioned as inset. nanoaggregates and a near-linear correlation could be obtained between
the size (volume) of the nano-aggregates and the emission maximum.
Such regular behavior is never possible in case of an impurity. Thus, we
can mention that the emission from neat RTILs is not because of any

3
A. Ghosh et al.
impurity rather the excitation wavelength dependent emission is
because of inherently present nano-aggregates. The fact that different
reorientation times are because of different sized nano-aggregates pre­
sent in neat RTIL can be supported by another control experiment. If a
polar solvent such as acetonitrile is added to the neat RTIL then it is
expected that different nanoaggregates will break and form an average
sized moiety. It can be seen (see S9 and S10) that fluorescence anisot­
ropy decays of emim[FAP] in acetonitrile are much faster in comparison
to neat RTIL. These decay profiles are identical to each other i.e. inde­
pendent of monitoring wavelengths (for a fixed λex=377 nm). Moreover,
rotational correlation time is same for all monitoring wavelengths and
the magnitude (200 ps) is much less than what has been observed for
neat RTIL (~1 to ~7 ns). Thus, we can conclude that in acetonitrile
different nano-aggregating units have been broken down. This control
experiment further confirms that the emission from RTIL is not from any
impurity rather from the nano-aggregating units. Thus, using fast rota­
tion dynamical measurements we were able to measure the size of
different nano-aggregates inherently present in RTIL and correlate the
emission maximum with the size of the nano-aggregates. With increase
in size of the nano-aggregates the emission maximum has been shown to
be increased. In this regard RTIL behave in a similar way to that of the
QDs. Moreover, in case of QDs photolumiscent quantum yield depends
on excitation wavelength. [48,49,53] However, in order to understand
the underlying phenomenon in a much detailed way, theoret­
ical/computational analyses need to be performed, and these in­
vestigations need to be repeated for many such RTILs. These
experiments are currently underway.
In conclusion, we could show that emission maxima of RTILs do not
change with change in temperature (298 K to 338 K). This means that
the nature of nanoaggregates do not change significantly in the said
temperature range. For a particular temperature rotational correlation
time has been observed to increase with increase in monitoring wave­
length. For a particular monitoring wavelength the rotational correla­
tion time has been observed to decrease by a factor of 3 to 4 with
increase in temperature from 298 K to 338 K. Stokes-Einstein-Debye
model has been shown to be depict nano-spatial-information for emim
[FAP] quite well and thus size of the nano-aggregates could be calcu­
lated to vary from ~~500 Å3. Finally we could correlate the emission
maximum with the size of the nano-aggregates inherently present in an
RTIL namely emim[FAP]. We could show that nano-aggregate having
aggregation volume of ~250 Å3, ~350 Å3, ~400 Å3 and ~500 Å3 emit at
~400 nm, ~450 nm, ~500 nm, and ~560 nm respectively.
Author contributions
PKM envisaged the project. TC, AD, DR and TR purified the RTIL,
performed temperature dependent steady state optical spectroscopic
measurements, and measured the viscosity of the RTIL at different
temperatures. AG and CKD performed rotational anisotropy based time
resolved rotational dynamical measurements and analyzed the data. AG
and CKD made the figures and the tables. AG, CKD and PKM wrote the
manuscript.
Research data
Research Data can be provided if necessary.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgement
PKM thanks IISER Kolkata for instrumental facilities and financial
4
Chemical Physics Impact 3 (2021) 100054
help. Support from the Fast-Track Project (SR/FT/CS-52/2011) of DST-
India is gratefully acknowledged. AG thanks UGC, CKD and TR thank
IISER Kolkata, and AD, TC, DR thank CSIR for their respective
Fellowship.
Supplementary materials
Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.chphi.2021.100054.
References
[1] T. Welton, Room-temperature ionic liquids. Solvents for synthesis and catalysis,
Chem. Rev. 99 (1999) 2071–2083.
[2] P. Wasserscheid, W. Keim, Ionic liquids: new solutions for transitions metal
catalysts, Angew. Chem. Int. Ed. 39 (2000) 3772–3789.
[3] J. Dupont, R.F.D. Souza, P.A.Z. Suarez, Liquid (molten salt) phase organometallic
catalysis, Chem. Rev. 102 (2002) 3667–3692.
[4] R.D. Rogers, K.R. Seddon, Ionic liquids: solvents of the future, Science 302 (2003)
792–793.
[5] K.R. Seddon, Ionic liquids: a taste of the future, Nature 2 (2003) 363–365.
[6] H. WeinGaertner, Understanding ionic liquids at the molecular level: facts,
problems and controversies, Angew. Chem., Int. Ed. 47 (2008) 655–670.
[7] J.P. Hallet, T. Welton, Room-temperature ionic liquids: solvents for synthesis and
catalysis, Chem. Rev. 111 (2011) 3508–3576.
[8] A. Mele, C.D. Tran, S.H. De Paoli Lacerda, The structure of a room-temperature
ionic liquid with and without trace amounts of water: the role of C− H•••O and
C− H•••F interactions in 1-n-butyl- 3-methylimidazolium tetrafluoroborate,
Angew. Chem., Int. Ed. 42 (2003) 4364–4366.
[9] S.M. Urahata, M.C.C. Ribeiro, Structure of ionic liquids of 1-alkyl-3-methylimida­
zolium cations: a systematic computer simulation study, J. Chem. Phys. 120 (2004)
1855–1863.
[10] J. Dupont, Liquid (molten salt) phase organometallic catalysis, J. Braz. Chem. Soc.
15 (2004) 341–350.
[11] M.G.D. Pópolo, G.A. Voth, On the structure and dynamics of ionic liquids, J. Phys.
Chem. B 108 (2004) 1744–1752.
[12] Y. Wang, G.A. Voth, Unique spatial heterogeneity in ionic liquids, J. Am. Chem.
Soc. 127 (2005) 12192–12193.
[13] A. Mele, G. Romano, M. Giannone, E. Ragg, G. Fronza, G. Raos, V. Marcon, The
local structure of ionic liquids: cation-cation NOE interactions and internuclear
distances in neat [BMIM][BF4] and [BDMIM][BF4], Angew. Chem. Int. Ed. 45
(2006) 1123–1126.
[14] A. Triolo, O. Russina, H.-.J. Bleif, E.J. Di Cola, Segregation in room temperature
ionic liquids, Phys. Chem. B 111 (2007) 4641–4644.
[15] W. Jiang, Y. Wang, G.A. Voth, Molecular dynamics simulation of nanostructural
organization in ionic liquid/water, J. Phys. Chem. B 111 (2007) 4812–4818.
[16] H. Weingaertner, Understanding ionic liquids at the molecular level: facts,
problems and controversies, Angew. Chem. Int. Ed. 47 (2008) 654–670.
[17] E.W. Castner, C.J. Margulis, M. Maroncelli, J.F. Wishart, Ionic liquids: strctures and
photochemical reactions, Annu. Rev. Phys. Chem. 62 (2011) 85–105.
[18] O. Russina, A. Triolo, L. Gontrani, R. Caminiti, Mesoscopic structural
heterogeneities in room-temperature ionic liquids, J. Phys. Chem. Lett. 3 (2012)
27–33.
[19] M. Macchiagodena, F. Ramondo, A. Triolo, L. Gontrani, R. Caminiti, Liquid
structure of 1-ethyl-3-methylimidazolium alkyl sulfates by X-ray scattering and
molecular dynamics, J. Phys. Chem. B 116 (2012) 13448–13458.
[20] A. Triolo, O. Russina, R. Caminiti, H. Shirota, H.Y. Lee, C.S. Santos, N.S. Murthy, E.
W. Castner Jr, Comparing intermediate range order for alkyl- vs.ether-substituted
cations in ionic liquids, Chem. Commun. 48 (2012) 4959–4961.
[21] H.K. Kashyap, J.J. Hettige, H.V.R. Annapureddy, C.J. Margulis, SAXS anti-peaks
reveal the length-scales of dual positive–negative and polar–apolar ordering in
room-temperature ionic liquids, Chem. Commun 48 (2012) 5103–5105.
[22] S. Li, G. Feng, J.L. Banuelos, G. Rother, P.F. Fulvio, S. Dai, P.T. Cummings,
Distinctive nanoscale organization of dicationic versus monocationic ionic liquids,
J. Phys. Chem. C 117 (2013) 18251–18257.
[23] Y. Shim, H.J. Kim, Dielectric relaxation and solvation dynamics in a room-
temperature ionic liquid: temperature dependence, J. Phys. Chem. B 117 (2013)
11743–11752.
[24] T. Ishida, H. Shirota, Dicationic versus monocationic ionic liquids: distinctive ionic
dynamics and dynamical heterogeneity, J. Phys. Chem. B 117 (2013) 1136–1150.
[25] H.K. Kashyap, C.S. Santos, R.P. Daly, J.J. Hettige, N.J. Murthy, H. Shirota, Jr.E.
W. Castner, C.J. Margulis, How does the ionic liquid organizational landscape
change when nonpolar cationic alkyl groups are replaced by polar isoelectronic
diethers? J. Phys. Chem. B 117 (2013) 1130–1135.
[26] J.J. Hettige, H.K. Kashyap, H.V.R. Annapureddy, C.J. Margulis, Anions, the
reporters of structure in ionic liquids, J. Phys. Chem. Lett. 4 (2013) 105–110.
[27] K. Wei, L. Deng, Y. Wang, Z.-.C. Ou-Yang, G. Wang, Effect of side-chain length on
structural and dynamic properties of ionic liquids with hydroxyl cationic tails,
J. Phys. Chem. B 118 (2014) 3642–3649.
[28] A. Ghosh, T. Chatterjee, D. Roy, A. Das, P.K. Mandal, On the nanoscopic
environment a neutral fluorophore experiences in room temperature ionic liquids,
J. Phys. Chem. C. 118 (2014) 5051–5057.
A. Ghosh et al.
[29] Y. Wang, G.A. Voth, Tail aggregation and domain diffusion in ionic liquids, J. Phys.
Chem. B 110 (2006) 18601–18608.
[30] R. Hayes, G.G. Warr, Rob atkin, structure and nanostructure in ionic liquids, Chem.
Rev. 115 (13) (2015) 6357–6426.
[31] W. Yong-Lei, L. Bin, S. Sten, F. Mocci, Z.-.Y. Lu, J. Yuan, A. Laaksonen, M.D. Fayer,
Microstructural and dynamical heterogeneities in ionic liquids, Chem. Rev. 120
(13) (2020) 5798–5877.
[32] P.K. Mandal, M. Sarkar, A. Samanta, Excitation wavelength dependent
fluorescence behavior of the room temperature ionic liquids and dissolved dipolar
solutes, J. Phys. Chem. A 108 (2004) 9048–9053.
[33] A. Paul, P.K. Mandal, A. Samanta, How transparent are the imidazolium ionic
liquids? A case study with 1-methyl-3- butylimidazolium hexafluorophosphate,
[bmim][PF6], Chem. Phys. Lett. 402 (2005) 375–379.
[34] A. Paul, P.K. Mandal, A. Samanta, On the optical properties of the imidazolium
ionic liquids, J. Phys. Chem. B 109 (2005) 9148–9153.
[35] A. Samanta, Dynamic stokes shift and excitation wavelength dependent
fluorescence of dipolar molecules in room temperature ionic liquids, J. Phys.
Chem. B 110 (2006) 13704–13716.
[36] A. Samanta, Solvation dynamics in ionic liquids: what we have learned from the
dynamic fluorescence stokes shift studies, J. Phys. Chem. Lett. 1 (2010)
1557–1562.
[37] S. Daschakraborty, R. Biswas, Composition dependent stokes shift dynamics in
binary mixtures of 1-butyl-3-methylimidazolium tetrafluoroborate with water and
acetonitrile: quantitative comparison between theory and complete measurements,
J. Phys. Chem. B 118 (2014) 1327–1339.
[38] H.K. Kashyap, R. Biswas, Dipolar solvation dynamics in room temperature ionic
liquids: an effective medium calculation using dielectric relaxation data, J. Phys.
Chem. B 112 (2008) 12431–12438.
[39] H.K. Kashyap, R. Biswas, Solvation dynamics of dipolar probes in dipolar room
temperature ionic liquids: separation of ion− dipole and dipole− dipole interaction
contributions, J. Phys. Chem. B 114 (2010) 254–268.
[40] H.K. Kashyap, R. Biswas, Stokes shift dynamics in ionic liquids: temperature
dependence, J. Phys. Chem. B 114 (2010) 16811–16823.
[41] S. Daschakraborty, T. Pal, R. Biswas, Stokes shift dynamics of ionic liquids: solute
probe dependence, and effects of self-motion, dielectric relaxation window, and
collective intermolecular solvent modes, J. Chem. Phys. 139 (2013)
164503–164515.
[42] S. Daschakraborty, R. Biswas, Ultrafast solvation response in room temperature
ionic liquids: possible origin and importance of the collective and the nearest
neighbour solvent modes, J. Chem. Phys. 137 (2012) 114501–114513.
[43] C.B. Murray, D.J. Noms, M.G. Bawendi, Synthesis and characterization of nearly
monodisperse CdE (E = sulfur, selenium, tellurium) semiconductor
nanocrystallites, J. Am. Chem. Soc. 115 (1993) 8706–8715.
[44] Y. Masumoto, K. Sonobe, Size-dependent energy levels of CdTe quantum dots,
Phys. Rev. B 56 (1997) 9734–9737.
[45] A.A. Guzelian, J.E.B. Katari, A.V. Kadavanich, U. Banin, K. Hamad, E. Juban, A.
P. Alivisatos, R.H. Wolters, C.C. Arnold, J.R. Heath, Synthesis of size-selected,
surface-passivated InP nanocrystals, J. Phys. Chem. 100 (1996) 7212–7219.
[46] D. Roy, T. Routh, A.V. Asaithambi, S. Mandal, P.K. Mandal, Spectral and temporal
optical behavior of blue-, green-, orange-, and red-emitting cdse-based core/
gradient alloy shell/shell quantum dots: ensemble and single-particle investigation
results, J. Phys. Chem. C 120 (2016) 3483–3491.
[47] S. Mandal, D. Roy, C.K. De, S. Ghosh, M. Mandal, A. Das, P.K. Mandal,
Instantaneous, room-temperature, open-air atmosphere, solution-phase synthesis
of perovskite quantum dots through halide exchange employing non-metal based
inexpensive HCl/HI: ensemble and single particle spectroscopy, Nanoscale Adv 1
(2019) 3506–3513.
[48] D. Roy, A. Das, C.K. De, S. Mandal, P.R. Bangal, P.K. Mandal, Why does the
photoluminescence efficiency depend on excitation energy in case of a quantum
dot? A case study of CdSe-based core/alloy shell/shell quantum dots employing
ultrafast pump–probe spectroscopy and single particle spectroscopy, J. Phys.
Chem. C 123 (2019) 6922–6933.
[49] C.K. De, S. Mandal, D. Roy, S. Ghosh, A. Konar, P.K. Mandal, Ultrafast dynamics
and ultrasensitive single-particle intermittency in small-sized toxic metal free InP-
based core/alloy-shell/shell quantum dots: excitation wavelength dependency
toward variation of PLQY, J. Phys. Chem. C 123 (2019) 28502–28510.
[50] C.K. De, T. Routh, D. Roy, S. Mandal, P.K. Mandal, Highly photoluminescent inp
based core alloy shell QDs from air-stable precursors: excitation wavelength
5
Chemical Physics Impact 3 (2021) 100054
dependent photoluminescence quantum yield, photoluminescence decay
dynamics, and single particle blinking dynamic, J. Phys. Chem. C 122 (2018)
964–973.
[51] C.K. De, D. Roy, S. Mandal, P.K. Mandal, Suppressed blinking under normal air
atmosphere in toxic-metal-free, small sized, InP-based core/alloy-shell/shell
quantum dots, J. Phys. Chem. Lett. 10 (2019) 4330–4338.
[52] D. Roy, S. Mandal, C.K. De, K. Kumar, P.K. Mandal, Nearly suppressed
photoluminescence blinking of small-sized, blue–green–orange–red emitting single
CdSe-based core/gradient alloy shell/shell quantum dots: correlation between
truncation time and photoluminescence quantum yield, Phys. Chem. Chem. Phys.
20 (2018) 10332–10344.
[53] S. Mandal, S. Mukherjee, C.K. De, D. Roy, S. Ghosh, P.K. Mandal, Extent of
shallow/deep trap states beyond the conduction band minimum in defect-tolerant
CsPbBr3 perovskite quantum dot: control over the degree of charge carrier
recombination, J. Phys. Chem. Lett. 11 (2020) 1702–1707.
[54] S. Mandal, S. Ghosh, S. Mukherjee, C.K. De, D. Roy, T. Samanta, P.K. Mandal,
Unravelling halide-dependent charge carrier dynamics in CsPb(Br/Cl)3 perovskite
nanocrystals, Nanoscale 13 (2021) 3654–3661.
[55] S. Ghosh, S. Mandal, S. Mukherjee, C.K. De, T. Samanta, M. Mandal, D. Roy, P.
K. Mandal, Near-unity photoluminescence quantum yield and highly suppressed
blinking in a toxic-metal-free quantum dot, J. Phys. Chem. Lett. 12 (2020)
1426–1431.
[56] A. Ghosh, C.K. De, T. Chatterjee, P.K. Mandal, What type of nanoscopic
environment does a cationic fluorophore experience in room temperature ionic
liquids? Phys. Chem. chem. phys. 17 (2015) 16587–16593.
[57] A. Ghosh, T. Chatterjee, P.K. Mandal, On the heterogeneity of fluorescence lifetime
of room temperature ionic liquids: onset of a journey for exploring red emitting
dyes, Chem. Commun. 4 (2012) 6250–6252.
[58] V. Gangamallaiah, G.B. Dutt, Fluorescence anisotropy of a nonpolar solute in 1-
alkyl-3-methylimidazolium-based ionic liquids: does the organized structure of the
ionic liquid influence solute rotation? J. Phys. Chem. B 117 (2013) 5050–5057.
[59] S.R. Prabhu, G.B. Dutt, Rotational diffusion of nondipolar and charged solutes in
alkyl-substituted imidazolium triflimides: effect of C2 methylation on solute
rotation, J. Phys. Chem. B 118 (2014) 9420–9426.
[60] D.C. Khara, J.P. Kumar, N. Mondal, A. Samanta, Effect of the alkyl chain length on
the rotational dynamics of nonpolar and dipolar solutes in a series of N-alkyl-N-
methylmorpholinium ionic liquids, J. Phys. Chem. B 117 (2013) 5156–5164.
[61] S.K. Das, M. Sarkar, Rotational dynamics of coumarin-153 and 4-aminophthali­
mide in 1-ethyl-3-methylimidazolium alkylsulfate ionic liquids: effect of alkyl
chain length on the rotational dynamics, J. Phys. Chem. B 116 (2012) 194–202.
[62] S.K. Das, P.K. Sahu, M. Sarkar, Diffusion–viscosity decoupling in solute rotation
and solvent relaxation of coumarin153 in ionic liquids containing
fluoroalkylphosphate (FAP) anion: a thermophysical and photophysical study,
J. Phys. Chem. B 117 (2013) 636–647.
[63] P.K. Sahu, S.K. Das, M. Sarkar, Toward understanding solute− solvent interaction in
roomtemperature mono- and dicationic ionic liquids: a combined fluorescence
spectroscopy and mass spectrometry analysis, J. Phys. Chem. B 118 (2014)
1907–1915.
[64] S.R. Prabhu, G.B. Dutt, Effect of low viscous nondipolar solvent on the rotational
diffusion of structurally similar nondipolar solutes in an ionic liquid, J. Phys.
Chem. B 119 (2015) 2019–2025.
[65] S.K. Das, D. Majhi, P.K. Sahu, M. Sarkar, Investigation of the influence of alkyl side
chain length on the fluorescence response of C153 in a series of room temperature
ionic liquids, RSC Adv 5 (2015) 41585–41594.
[66] C. Lawler, M.D. Fayer, The influence of lithium cations on dynamics and structure
of room temperature ionic liquids, J. Phys. Chem. B 117 (2013) 9768–9774.
[67] S.R. Prabhu, G.B. Dutt, Rotational dynamics of imidazolium-based ionic liquids: do
the nature of the anion and the length of the alkyl chain influence the dynamics?
J. Phys. Chem. B 118 (2014) 13244–13251.
[68] S. Cha, T. Shim, Y. Ouchi, D. Kim, Characteristics of visible fluorescence from ionic
liquids, J. Phys. Chem. B 117 (2013) 10818–10825.
[69] S.K. Das, P.K. Sahu, M. Sarkar, Studies on electronic energy transfer (EET) on a
series of room temperature ionic liquids (RTILs): can the EET studies on RTILs be
exploited to predict their structural organization? RSC Adv. 4 (2014), 39184-3919.
[70] A. Nazet, S. Sokolov, T. Sonnleitner, T. Makino, M. Kanakubo, R. Buchner,
Densities, viscosities, and conductivities of the imidazolium ionic liquids [Emim]
[Ac], [Emim][FAP], [Bmim][BETI], [Bmim][FSI], [Hmim][TFSI], and [Omim]
[TFSI], J. Chem. Eng. Data 60 (2015) 2400–2411.
 pubs.acs.org/Langmuir Article

Surface-Directed Disparity in Self-Assembled Structures of Small-

Peptide L‑Glutathione on Gold and Silver Nanoparticles
Suranjana Chakrabarty, Swagata Maity, Darshana Yazhini, and Anup Ghosh*
Cite This: Langmuir 2020, 36, 11255−11261 Read Online

ACCESS Metrics & More Article Recommendations *

sı Supporting Information
Downloaded via S.N. BOSE NATL CTR BASIC SCIENCES on October 5, 2020 at 10:03:32 (UTC).
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ABSTRACT: Despite the key roles of L-glutathiones (GSHs) inbiology

and nano-biotechnology, understanding their labile structures and hydrogen
bond interactions with nanoparticles has posed a critical challenge to the
scientific community. The structural conformation of GSH as a capping
layer on gold nanoparticle (AuNP) and silver nanoparticle (AgNP) surfaces
is investigated. In this report, we attempt to explore the material-dependent
interaction of GSH with different spherical nanoparticle surfaces by
employing Fourier transform infrared (FTIR) spectroscopy. The infrared
signal of amide I of GSH is studied as a function of different materials’
spherical nanoparticles with comparable size. We revealed the β-sheet
secondary structure of GSH on AgNPs and the random structure on AuNPs
even though both the nanoparticles have comparable shapes and sizes and
belong to the same group of the periodic table. The GSH is firmly anchored
on the gold and silver surface via the thiol of the cys part. However, our experimental data designate a further interaction with the
AgNP surface via the carboxylic acid group of the gly- and glu- end of the molecule. It is observed that enhancement of IR
absorption of amide I of GSH is pronounced by a factor of 10 on AuNP but, in contrast, on the same-sized AgNP, the suppression is
perceived by a factor of 2, even though both are plasmonic materials with respect to free GSH. This study can be used as a point of
reference for understanding the structural conformation of the capping layer on nanoparticle surfaces as well as surface enhancement
of the IR absorption of amide I. We would like to emphasize that molecular self-assembly on the nanoparticle surfaces is definitely of
very broad interest for chemists working in nearly any subdiscipline, spanning from the nanoparticle-based medicine to surface-
enhanced spectroscopy to heterogeneous catalysis, etc.

■ INTRODUCTION
Recently, nanoparticle-oriented research has become an area of
interest due to its various potential scientific and industrial
applications, such as in chemical, medical, and biological
fields.1 The versatile properties of nanoparticles, such as
biocompatibility, drug delivery, and plasmonic activity, are
typically controlled by the molecular functionalization of their
surfaces.2−5 Thus, nanoparticle surface chemistry is one of the
most thriving modern research fields. Over the past few years,
metal nanoparticles have played a key role in modern
medicine, being a contrast carrier for drug delivery to target
cells.6,7 Constancy under appropriate physicochemical environ-
ments is a challenging requirement for such in vivo
applications.8−10 Nowadays, various coating agents, such as
organic molecules, peptides, proteins, DNA, etc., are
extensively used for stability and to increase the solubility of
nanomaterials in ionic solutions.11,12 Among these capping
agents, peptides provide invaluable support in material sciences
by interacting with the surface of noble metal nanoparticles.8,13
The inorganic nano−biointeractions are comprised of
exchanges of kinetic, thermodynamic, and physicochemical
interactions between the nanoparticle surfaces and peptides.14
It is reported that DNA-capped gold nanoparticles are used for
an anticancer drug, doxorubicin, delivery.15 Despite a large
number of investigations mainly on ligand-capped nanoparticle
synthesis with different shapes and sizes, structural details of
the capping layer are not explored in detail.16−19
The subject of molecular ordering and self-assembly on the
nanoparticle surface has been actively discussed during the last
decade. However, the efficacy of nanomedicines remains
inadequate, mainly because of a lack of information about
the surface structural properties of nano−biointeractions.17
The self-assembled monolayer structure of larger peptides on
the nanoparticle surfaces fully depends on the curvature of the
nanoparticles.18,19 In the presence of mercaptobenzoic acid,
peptides form the β-sheet structure on AuNPs.20 A detailed
investigation of ligand surface conformation is obligatory to
fully discover the broad prospects of the capping ligands in
Received: May 22, 2020
Revised: September 2, 2020
Published: September 3, 2020

© 2020 American Chemical Society https://dx.doi.org/10.1021/acs.langmuir.0c01527

11255 Langmuir 2020, 36, 11255−11261
Langmuir pubs.acs.org/Langmuir
nanobased drug delivery.21,22 The structural conformation of
the self-assembled monolayer has a massive effect on various
nanoparticles; for example, it can assist in proficient cell
membrane penetration.23 These findings are expected to serve
as a guideline for further development of biomedical
applications of noble metal nanoparticles.
Tripeptide L-glutathione (ç-glu-cys-gly) (GSH)-protected
AuNPs and AgNPs are extensively used in various applications,
including the biomedical applications, where GSH is found to
be useful as a cell-membrane-penetrating agent and in drug-
delivery systems.24−26 GSH is the only free thiol known to be
present inside the cells with a low millimolar concentration,
and it is known to be readily exchanged with various
nanoparticle ligands, once the functionalized nanoparticles
reach the cytoplasm.27 The concentration of GSH inside the
cell is higher than that of the extracellular environment, which
plays a crucial role in the transportation of drugs.24
Even though the structural characterization of GSH on the
nanoparticle surfaces is not fully explored, adsorption of GSH
on gold surfaces was comprehensively studied.28−34 It is
reported that GSH is attached via the thiol group to the
nanoparticle surfaces at a particular pH range in an aqueous
solution.27,34 A self-assembled monolayer of GSH on AgNP
has been reported recently.35 However, the understanding of
the detailed molecular interactions between GSH with various
inorganic noble nanoparticles is often challenging. In our work,
we have described GSH−nanoparticle interactions and try to
understand how GSHs interact with the nanoparticle surfaces
in terms of surface structure and binding mechanisms.
Parallelly, we have compared the function of AuNP and
AgNP on the capping layer structural conformation. This is
substantially compounded by the presumption that GSH
nanomaterial interaction is uniquely different and requires
individual analysis per nanoparticle basis. The general
significance of this report is the revealing of the structural
configuration of GSH on different nanoparticle surfaces. A
better understanding of the effect of environmental conditions
on the structural conformation of the surface GSH molecules
helps to elucidate the microscopic mechanisms of various
processes involving nanoparticles. Our report is the first to
demonstrate the disparity of the surface structural conforma-
tion of GSH on AuNPs and AgNPs. It is nevertheless clear that
the role of the GSH conformational structure can be very
important, with more evidence being accumulated in our
literature.
■ MATERIALS AND METHODS
All of the detailed procedures are given in the Supporting
Information.
■
RESULTS AND DISCUSSION
IR absorption spectroscopy is one of the widely used
techniques for characterizing the secondary structure of
proteins and peptides. The most intense IR absorbance band
(1600−1700 cm−1) of amide I is mostly used to investigate the
structures of proteins and peptides. The conformation-
sensitive amide I vibrational band directly obtained from the
backbone of proteins and peptides is associated with the CO
stretching vibration. All of the linear IR absorption spectra of
GSH are taken in D2O. The representative spectra are shown
in Figure 1. We have compared here the IR absorption
spectrum of amide I of free GSH with those of GSH capped on
AuNP and AgNP (comparable size ∼37 nm) to reveal their
11256
Article
Figure 1. (a) Amide I IR absorption spectra of free GSH (black),
GSH on AgNP (red), and GSH on AuNP (blue). (b) Linear IR
absorption of amide I of GSH on AgNP (black curve: experimental
data, red- and blue-fitted curves: two Lorentzian transitions of fibrils
and β-sheet, respectively, pink: cumulative fit data).
significant similarities or dissimilarities. The linear IR
absorption spectra of free GSH in D2O (Figure 1a) shows
two overlapping broad band transitions (1617 and 1645 cm−1),
which correspond to the two amide I bands.36
In contrast to the free GSH, the linear IR spectrum of amide
I (Figure 1, red curve) of GSH capped on AgNP shows two
narrow peaks at 1628 and 1604 cm−1. Herein, we have resolved
that the transition at 1628 cm−1 is the characteristic transition
of the amide I peak of the β-sheet structure, whereas at 1604
cm−1, the transition is a clear identification of the fibrils
structure on AgNP.37−40 Fitted Lorentzian linear IR data
suggested that GSH forms highly ordered parallel β-sheet
(80%) and fibril-like structures (20%) on AgNP by considering
the constant transition dipole of amide I in β-sheet and fibrils
(Figure 1b). On the other hand, the absence of the higher IR
absorption frequency of amide I (around 1680 cm−1) indicates
the lack of an antiparallel β-sheet structure.41 To determine
whether the secondary structure formation of GSH was specific
to AgNP, similar experiments of AuNP with comparable sizes
and shapes were carried out. The corresponding IR absorption
results for GSH on AuNP are shown in Figure 1 (blue). The
spectral position (1653 cm−1) dictates the completely distinct
random structure of GSH on AuNP, which is an opposite
occurrence of GSH on AgNP.
To confirm the structural conformation of GSH on AuNP
and AgNP, circular dichroism (CD) spectra were recorded in
the region 190−250 nm. An ordered structure (β-sheet) can
show both negative and positive peaks during CD analysis.
Most interestingly, the spectrum for all β-sheets has a negative
peak from 215 to 230 nm and a positive peak approximately at
195−210 nm.42 It can provide detailed structural information
about the molecular conformation of the capping-peptide
GSH. Using CD measurement, we compared and confirmed
the structural conformation of GSH on AgNP and AuNP with
our preliminary findings estimated from linear IR studies.
Figure 2a (red curve) shows a negative peak at 225 nm and a
positive peak at around 205 nm, which indicates the formation
of the β-sheet of GSH on AgNPs. From Figure 2b, it is clear
that the negative band at around 190 nm and a positive band at
around 213 nm revealed the random structure of GSH on
AuNPs. No spectral signature is observed for the ordered
structure of GSH on AuNPs. On the other hand, Figure 2a
(black curve) shows two intensely negative bands at 204 and
218 nm of free GSH. Hence, from the above studies, we have
demonstrated that after the coating of GSH on nanoparticles, it
adopts different molecular conformations on the surface of the
same-sized and -shaped AgNP and AuNP. Thus, the above
https://dx.doi.org/10.1021/acs.langmuir.0c01527
Langmuir 2020, 36, 11255−11261
Langmuir pubs.acs.org/Langmuir
Figure 2. (a) CD spectra of free GSH (black curve) and GSH on
AgNP (red curve) and (b) CD spectra of GSH on AuNP (black
curve).
linear IR and CD spectra display the nanoparticle-induced
structural changes and the secondary structure formation of
the tripeptide GSH after capping on 37 nm nanoparticles. It is
commonly known that to facilitate the arrangement of peptides
into β-sheets, sequences with at least three to six amide units
are needed.43−45 Remarkably, GSH has only two amide units
participating in the β-sheet formation, and the driving force for
its self-assembly is provided by the interaction with the
nanoparticle’s surface.
■
SURFACE-ENHANCED INFRARED ABSORPTION
(SEIRA)
IR absorption is enhanced when molecules are adsorbed on
metal surfaces (such as gold and silver) and the phenomenon
is known as surface-enhanced infrared absorption
(SEIRA).46−51 The strength of the SEIRA depends on the
metal type, surface morphology, as well as the orientation of
the molecular vibrations with the metal surfaces.52 The amide I
symmetric stretch is the strongest peak visible in the SEIRA
spectra. Figure 3a,b presents a comparison of the SEIRA
Figure 3. (a) Concentration-normalized IR absorption spectra of
amide I of free GSH (black curve) and GSH on AgNP (red curve).
(b) Concentration-normalized IR absorption spectra of amide I of
free GSH (black curve) and GSH on AuNP (red curve).
spectra of GSH on AgNPs and AuNPs. The spectra, such as
those in Figure 3a,b, allow the estimation of relative
enhancements. The IR absorption of amide I of GSH on
AgNPs (red color band) is suppressed to half of that of free
GSH (Figure 3a), while in contrast, enhancement of 10 times
is observed for the amide I IR absorption spectrum at 1653
cm−1 of GSH anchored on AuNPs (Figure 3b). Compared to
the free GSH, GSH on AgNP and AuNP display distinctive
properties that make it more attractive as a capping ligand. The
optical properties of GSH are radically changed after
adsorption on metal nanoparticles. To explain this anomalous
behavior of GSH on plasmonic nanomaterials of the same size,
shape, and group of the periodic table, a “surface selection
rule” was introduced. This rule states that the IR absorption of
11257
Article
a spectrum fluctuates due to the orientation of the vibrating
mode of the capping ligands on the metal surfaces: the parallel
stretching mode shows suppression, while the orthogonal
stretching modes cause surface enhancement by considering
the distance unchanged.49−51 Hence, probably the amide I
stretching mode of GSH is nearly parallel to the surface of the
AgNP (Figure 4a) and orthogonal to the surface of the AuNP
Figure 4. Orientation of amide I stretching mode of GSH on (a)
AgNP and (b) AuNP.
(Figure 4b). Thus, we showed that the GSH molecules on a
AgNP form β-sheet-like structures, in which amide I is almost
parallel to the surface of the AgNP, while on the same-sized
AuNP surface, they are randomly oriented where amide I is
perpendicular to the AuNP surface. This disparity of surface
structure of GSH on AgNP and AuNP undoubtedly describes
the suppression and enhancement of amide I.
In this work, we have demonstrated that GSH forms β-sheet-
like structures with an orientation nearly parallel to the AgNP
surface, thus “wrapping” the nanoparticle, and a random
structure on AuNP in an aqueous solution.
Now, the question as to why GSH forms β-sheet structures
on AgNP but not on AuNP arises. To achieve surface
structural disparity of GSH, we monitored the carboxylic
asymmetric stretching frequency. The asymmetric stretching
frequency of the carboxylic group of GSH on AgNP shifted
toward the blue region (1735 cm−1) compared to that of free
GSH (1725 cm−1), which is probably due to the bond
formation of the carboxylic group with the AgNP surface
(Figure S3). Therefore, GSH was probably attached to the
AgNP surface by three ends, one thiol and two other carboxylic
ends. Consequently, GSHs formed a β-sheet structure by
positioning parallelly to each other. In the case of GSH on the
AuNP, the IR peak position of carboxylic asymmetric
stretching modes is almost constant compared to that of free
GSH (Figure S3). These could be due to the attachment of the
thiol group to the surface of AuNPs, and the carboxylic groups
are freely present in the bulk with a greater distance from the
surface. To validate our assessment, we performed the
following theoretical calculation.
■
ESTIMATION OF BINDING ENERGY: DFT
CALCULATIONS
As there is a competition between the carboxylic group and the
thiol group of GSH to attach to the surface of the
nanoparticles, there is a need to define the exact interaction.
To gain exact knowledge about the interaction of GSH on the
surface of AuNP and AgNP in neat water, we performed
density functional theory (DFT) calculations by employing
Gaussian 09 software.53,54 In these calculations, we have used
two different basis sets, 631G for lower atomic weights (C, H,
N, O, S) and LanL2DZ is for higher atomic weights (Au and
Ag), for preliminary geometry optimization. To obtain the
binding energy of GSH on nanoparticles through thiol and
carboxylic oxygen bonds, the “scan” analyses are performed on
the B3LYP−LANL2DZ optimized structures. To calculate the
https://dx.doi.org/10.1021/acs.langmuir.0c01527
Langmuir 2020, 36, 11255−11261
Langmuir pubs.acs.org/Langmuir Article

binding energies of Au−S, Au−O, Ag−S, and Ag−O bonds, we

used the Morse potential energy calculation method (eq 1)
V = De ( 1 − e−a(r − re))2 (1)
where r is the distance between the atoms, re is the equilibrium
bond distance, and De is the dissociation energy.
From the Morse potential diagrams, in the case of silver
nanoparticles, the binding energies of GSH through S and O
atoms are almost comparable (Figure 5a), which signifies that

Figure 6. Model surface structure of GSH on AgNP (left) and AuNP

(right).

conformation on AgNPs and the random structure on

Figure 5. Morse potential diagram of (a) Ag−O (red curve) and Ag−
S (black curve) and (b) Au−O (red curve) and Au−S (black curve).
AuNPs of GSH (both with comparable concentrations) to
the solvent environment, we added either metal ions (CaCl2)
GSH can be attached on the AgNP surface through both S and or an alkali base (NaOH). The addition of salt causes a color
O atoms of the carboxylic group (Table 1). On the contrary, it change of GSH-coated AuNPs from reddish to blue. Parallelly,
the ultraviolet−visible (UV−vis) absorption spectra of AuNP
are red-shifted (Figure S4). This color change and red shift of
Table 1. Calculated Binding Energies for All Respective
the UV−vis spectra are caused by aggregation of the AuNPs.
Interactions
To test the limit of the salt that AuNPs and AgNPs can
molecular interaction binding energy (kJ/mol) tolerate, we considered UV−vis spectra with an increasing
Au−S−G 246 concentration of Ca2+ (Figure S4). The AgNPs remained
Au−O−G 182 stable up to 0.5 mM CaCl2 concentration, wherein 0.5 mM
Ag−S−G 142 Ca2+ induced aggregation instantaneously for AuNPs, which is
Ag−O−G 135 enough for demolishing the total UV−vis spectral signal of
AuNPs. The β-sheet structure of GSH was able to protect
AgNPs up to 5 mM concentration of CaCl2, which is
is clear that the binding interaction of GSH via S atom is much
approximately 10 times more concentrated than that in the
higher than the binding interaction via O atom on the AuNP
case of AuNPs. GSH-capped AuNPs easily aggregate with the
surface (Figure 5b), where the estimated energy difference is
treatment of CaCl2 as the electronegative carboxylic groups of
almost 64 kJ/mol (Table 1). These energy differences signify
GSH are free and present toward the bulk of the solution.
that GSH undoubtedly attaches through the S atom and not
Hence, they can easily chelate with Ca+2 ion and form salt
the carboxylic group on AuNP (Table 1).
bridges (Figure 7). But in the case of GSH-capped AgNPs,
Therefore, this study demonstrates that GSH forms
initially, there are no free carboxylic groups to chelate as both
molecular self-assembled structures on AgNPs through the
oxygen and sulfur are attached to the surface of the
attachment of both the S and O atoms but only through the
nanoparticles. On increasing the sufficient concentration of
attachment of S atom on the surface of AuNPs. DFT
CaCl2, the bond between the carboxylic group of GSH and the
calculation analysis made our assessment a concrete con-
surface of AgNPs is probably ruptured and chelated.
clusion. From our experimental and theoretical studies, a clear
disparity in the surface structure of GSH on the same-sized
AuNP and AgNP is undoubtedly recognized. Subsequently, we
have modeled a structure of GSH on AgNP and AuNP, where
the β-sheet structure and the random structure of GSH are
clearly shown, respectively (Figure 6).

■ TREATMENT OF SALT CACL2 AND PH OF

GSH-CAPPED NANOPARTICLES
It is well known that different physiological conditions have the
capability to manipulate the structural conformation of protein
and peptide fragments, which have incredible applications in
medicinal biotechnology. Therefore, understanding the de-
tailed structural conformations of tripeptide GSH fastened on
different spherical nanoparticles in the presence of different
physiological conditions is a stimulating topic for nanobased Figure 7. Model structure of the calcium-ion-bridge-facilitated AuNP
drug delivery. To test the sensitivity of the β-sheet aggregation.

11258 https://dx.doi.org/10.1021/acs.langmuir.0c01527
Langmuir 2020, 36, 11255−11261
Langmuir pubs.acs.org/Langmuir
To investigate the stability of the β-sheet of GSH on AgNPs,
linear IR absorption of amide I in the presence of different salt
concentrations was considered (Figure 8a). The secondary
Figure 8. (a) IR spectra of amide I of GSH on AgNP in the presence
of CaCl2 (black curve: 0 M, red curve: 0.2 M, blue curve: 0.5 M, and
pink curve: GS-0.6 M) and high pH (green curve). (b) CD spectra of
GSH (black curve: GSH on AgNP, red curve: GSH on AgNP in the
presence of CaCl2, and blue curve: GSH on AgNP at pH 10).
structure remains unaffected up to a certain concentration (0.4
M), but it was perturbed after the addition of excess CaCl2 (0.6
M). Here, the considered concentrations of CaCl2 are almost
100 times more than UV−vis experimental concentrations
because AgNP concentration is multiplied in the IR experi-
ment by almost 100 times. The disruption of β-sheets occurs
after the addition of an excess amount of Ca+2 ions because
Ca+2 ions compete with the positively charged AgNP surface as
they might be chelated by the carboxylic anions, forming a
calcium bridge (Figure 7). In a higher pD condition (pD =
10), there is no secondary structure signature peak of amide I
in linear IR spectra. Hydroxyl anions probably disrupt β-sheets
by altering the surface layer of the nanoparticles, which
probably leads to the carboxyl groups repulsing from the
surface. A lower pD (5.5) carboxylic group can be easily
attached, as the surface of the AgNPs is positively charged.55,56
Similar results were observed from CD spectra analysis (Figure
8b).
■ CONCLUSIONS
The present work provides a fundamental understanding of the
surface conformational structure of small tripeptide GSH on
silver and gold nanoparticles. The sulfur−silver bonds and the
interaction of the carboxylic anions with the silver surface
facilitate the formation of the parallel β-sheet and fibril-like
structures. On the contrary, no secondary structure is formed
on the AuNP surface even though they belong to the same
family of the periodic table and have almost the same size and
shape. Despite their similarities, they exhibit different behavior
with GSH to form different surface conformational structures.
In this study, we have reported the suppression of the amide I
IR absorption band of GSH on the AgNP surface because of
the nearly parallel orientation of the stretching mode with the
surface. However, IR absorption of amide I was enhanced
significantly for its nonparallel stretching mode with the AuNP
surface. In comparison with spherical AgNPs, superior
enhancement due to the GSH conformational structure was
confirmed for AuNPs. The obtained results could demonstrate
the utility of Au and Ag nanostructures in SEIRA applications
of amide I, which is considered to be relevant in biosensing
and biomarkers. GSH-capped AgNPs are more stable
compared to AuNP in the presence of calcium chloride. The
structural conformation study of GSH on nanoparticles and its
11259
Article
dependency on different physiological conditions (such as salt
and pH) is important for nano-biotechnological applications.
Our work is also important as GSH is an abundant molecule in
the cytoplasm and can be potentially exchanged with various
nanoparticle ligands during nanoparticle-based drug delivery.
Our findings of the surface conformational structure of GSH
would also be useful for other inorganic colloidal nanoparticle
studies.
■
*
ASSOCIATED CONTENT
sı Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.langmuir.0c01527.
Materials and methods (nanoparticle synthesis and
Fourier transform infrared (FTIR) sample preparation);
UV−visible spectra; size distribution of nanoparticles; IR
absorption of the carboxylic group of GSH; UV−vis and
CD spectra of GSH on AgNP (Figures S1−S4) (PDF)
■ AUTHOR INFORMATION
Corresponding Author
Anup Ghosh − Department of Condensed Matter Physics and
Materials Sciences, S. N. Bose National Centre for Basic
Sciences, Kolkata 700106, India; orcid.org/0000-0002-
1442-8740; Email: anupg86@gmail.com, anup.ghosh@
bose.res.in
Authors
Suranjana Chakrabarty − Department of Condensed Matter
Physics and Materials Sciences, S. N. Bose National Centre for
Basic Sciences, Kolkata 700106, India
Swagata Maity − Department of Condensed Matter Physics and
Materials Sciences, S. N. Bose National Centre for Basic
Sciences, Kolkata 700106, India
Darshana Yazhini − Department of Condensed Matter Physics
and Materials Sciences, S. N. Bose National Centre for Basic
Sciences, Kolkata 700106, India
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.langmuir.0c01527
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
This work was supported in part by the DST, Government of
India [DST/INSPIRE/04/2018/002031].
■ REFERENCES
(1) Khan, I.; Saeed, K.; Khan, I. Nanoparticles: Properties,
applications and toxicities. Arabian J. Chem. 2019, 12, 908−931.
(2) Jain, P. K.; Huang, X.; El-Sayed, I. H.; El-Sayed, M. A. Review of
Some Interesting Surface Plasmon Resonance-enhanced Properties of
Noble Metal Nanoparticles and Their Applications to Biosystems.
Plasmonics 2007, 2, 107−118.
(3) Hu, C.; Rong, J.; Cui, J.; Yang, Y.; Yang, L.; Wang, Y.; Liu, Y.
Fabrication of a graphene oxide-gold nanorod hybrid material by
electrostatic self-assembly for surface-enhanced Raman scattering.
Carbon 2013, 51, 255−264.
(4) Castellana, E. T.; Gamez, R. C.; Russell, D. H. Label-free
Biosensing with Lipid-Functionalized Gold Nanorods. J. Am. Chem.
Soc. 2011, 133, 4182−4185.
https://dx.doi.org/10.1021/acs.langmuir.0c01527
Langmuir 2020, 36, 11255−11261
Langmuir pubs.acs.org/Langmuir
(5) Yin, P. T.; Kim, T.-H.; Choi, J.-W.; Lee, K.-B. Prospects for
graphene-nanoparticle-based hybrid sensors. Phys. Chem. Chem. Phys.
2013, 15, 12785−12799.
(6) Murthy, S. K. Nanoparticles in modern medicine: State of the art
and future challenges. Int. J. Nanomed. 2007, 2, 129−141.
(7) Zou, L.; Wang, H.; He, B.; Zeng, L.; Tan, T.; Cao, H.; He, X.;
Zhang, Z.; Guo, S.; Li, Y. Current Approaches of Photothermal
Therapy in Treating Cancer Metastasis with Nanotherapeutics.
Theranostics 2016, 6, 762−772.
(8) Cobley, C. M.; Chen, J.; Cho, E. C.; Wang, L. V.; Xia, Y. Gold
nanostructures: a class of multifunctional materials for biomedical
applications. Chem. Soc. Rev. 2011, 40, 44−56.
(9) Han, G.; Martin, C. T.; Rotello, V. M. Stability of Gold
Nanoparticle-Bound DNA toward Biological, Physical, and Chemical
Agents. Chem. Biol. Drug. Des. 2006, 67, 78−82.
(10) Giljohann, D. A.; Seferos, D. S.; Daniel, W. L.; Massich, M. D.;
Patel, P. C.; Mirkin, C. A. Gold Nanoparticles for Biology and
Medicine. Angew. Chem., Int. Ed. 2010, 49, 3280−3294.
(11) Duchesne, L.; Gentili, D.; Comes-Franchini, M.; Fernig, D. G.
Robust Ligand Shells for Biological Applications of Gold Nano-
particles. Langmuir 2008, 24, 13572−13580.
(12) Han, G.; Ghosh, P.; Rotello, V. M. Functionalized Gold
Nanoparticles for Drug Delivery. Nanomedicine 2007, 2, 113−123.
(13) Slocik, J. M.; Naik, R. R. Probing peptide−nanomaterial
interactions. Chem. Soc. Rev. 2010, 39, 3454−3463.
(14) Nel, A. E.; Madler, L.; Velegol, D.; Xia, T.; Hoek;
Somasundaran, P.; Klaessig, F.; Castranova, V.; Thompson, M.
understanding biophysicochemical interactions at the nano-bio
interface. Nat. Mater. 2009, 8, 543−557.
(15) Alexander, C. M.; Maye, M. M.; Dabrowiak, J. C. DNA-capped
nanoparticles designed for doxorubicin drug delivery. Chem. Commun.
2011, 47, 3418−3420.
(16) Park, J.-W.; Shumaker-Perry, J. S. Structural Study of Citrate
Layers on Gold Nanoparticles: Role of Intermolecular Interactions in
Stabilizing Nanoparticles. J. Am. Chem. Soc. 2014, 136, 1907−1921.
(17) Wang, Y.; Cai, R.; Chen, C. The Nano-Bio Interactions of
Nanomedicines: Understanding the Biochemical Driving Forces and
Redox Reactions. Acc. Chem. Res. 2019, 52, 1507−1518.
(18) Shaw, C. P.; Middleton, D. A.; Volk, M.; Levy, R. Amyloid-
Derived Peptide Forms Self-Assembled Monolayers on Gold
Nanoparticle with a Curvature-Dependent β-Sheet Structure. ACS
Nano 2012, 6, 1416−1426.
(19) Mandal, H. S.; Kraatz, H.-B. Effect of the Surface Curvature on
the Secondary Structure of Peptides Adsorbed on Nanoparticles. J.
Am. Chem. Soc. 2007, 129, 6356−6357.
(20) Ho, J.-J.; Ghosh, A.; Zhang, T. O.; Zanni, M. T. Heterogeneous
Amyloid B-Sheet Polymorphs Identified on Hydrogen Bond
Promoting Surfaces Using 2D SFG Spectroscopy. J. Phys. Chem. A
2018, 122, 1270−1282.
(21) Mosquera, J.; Henriksen-Lacey, M.; García, I.; Martínez-Calvo,
M.; Rodríguez, J.; Mascareñas, J. L.; Liz-Marzán, L. M. Cellular
Uptake of Gold Nanoparticles Triggered by Host−Guest Interactions.
J. Am. Chem. Soc. 2018, 140, 4469−4472.
(22) Eibling, M. J.; MacDermaid, C. M.; Qian, Z.; Lanci, C. J.; Park,
S.-J.; Saven, J. G. Controlling Association and Separation of
GoldNanoparticles with Computationally Designed Zinc-Coordinat-
ing Proteins. J. Am. Chem. Soc. 2017, 139, 17811−17823.
(23) Pengo, P.; Ş ologan, M.; Pasquato, L.; Guida, F.; Pacor, S.;
Tossi, A.; Stellacci, F.; Marson, D.; Boccardo, S.; Pricl, S.; Posocco, P.
Gold Nanoparticles with Patterned Surface Monolayers for Nano-
medicine: Current Perspectives. Eur. Biophys. J. 2017, 46, 749−771.
(24) Ling, X.; Tu, J.; Wang, J.; Shajii, A.; Kong, Na.; Feng, C.;
Zhang, Y.; Yu, M.; Xie, T.; Bharwani, Z.; Aljaeid, B. M.; Shi, B.; Tao,
W.; Farokhzad, O. C. Glutathione-Responsive Prodrug Nanoparticles
for Effective Drug Delivery and Cancer Therapy. ACS Nano 2019, 13,
357−370.
(25) Cheng, R.; Feng, F.; Meng, F.; Deng, C.; Feijen, J.; Zhong, Z.
Glutathione-Responsive Nano-Vehicles as a Promising Platform for
11260
Article
Targeted Intracellular Drug and Gene Delivery. J. Controlled Release
2011, 152, 2−12.
(26) Lund, T.; Callaghan, M. F.; Williams, P.; Turmaine, M.;
Bachmann, C.; Rademacher, T.; Roitt, I. M.; Bayford, R. The
influence of ligand organization on the rate of uptake of gold
nanoparticles by colorectal cancer cells. Biomaterials 2011, 32, 9776−
9784.
(27) Berg, J. M.; Tymoczko, J. L.; Stryer, L. Biochemistry, 5th ed.;
Freeman & Company: New York, 2002.
(28) Hepel, M.; Tewksbury, E. Ion-Gating Phenomena of Self-
Assembling Glutathione Films on Gold Piezoelectrodes. J. Electro-
anal.Chem. 2003, 552, 291−305.
(29) Bieri, M.; Bürgi, T. L-Glutathione Chemisorption on Gold and
Acid/Base Induced Structural Changes: A PM-IRRAS and Time-
Resolved in Situ ATR-IR Spectroscopic Study. Langmuir 2005, 21,
1354−1363.
(30) Bieri, M.; Bürgi, T. Adsorption Kinetics of L-Glutathione on
Gold and Structural Changes During Self-Assembly: An in Situ
ATRIR and QCM Study. Phys. Chem. Chem. Phys. 2006, 8, 513−520.
(31) Maza, F. L.; Leo, L. M. D.; Rubert, A. A.; Carro, P.; Salvarezza,
R. C.; Vericat, C. New Insight into the Interface Chemistry and
Stability of Glutathione Self-Assembled Monolayers on Au(III). J.
Phys. Chem. C 2016, 120, 14597−14607.
(32) Vallée, A.; Humblot, V.; Méthivier, C.; Pradier, C.-M.
Glutathione Adsorption from UHV to the Liquid Phase at VariouspH
on Gold and Subsequent Modification of Protein Interaction. Surf.
Interface Anal. 2008, 40, 395−399.
(33) Vallée, A.; Humblot, V.; Méthivier, C.; Pradier, C.-M.
Adsorption of a Tripeptide, GSH, on Au (111) under UHVCondi-
tions; Pm-Rairs and Low T-XPS Characterisation. Surf. Sci. 2008, 602,
2256−2263.
(34) Moaseri, E.; Bollinger, J. A.; Changalvaie, B.; Johnson, L.;
Schroer, J.; Johnston, K. P.; Truskett, T. M. Reversible Self-Assembly
of Glutathione-Coated Gold Nanoparticle Clusters via pH-Tuna-
bleInteractions. Langmuir 2017, 33, 12244−12253.
(35) Ghosh, A.; Prasad, A. K.; Chuntonov, L. Two-Dimensional
Infrared Spectroscopy Reveals Molecular Self-Assembly on the
Surface of Silver Nanoparticles. J. Phys. Chem. Lett. 2019, 10,
2481−2486.
(36) Qian, W.; Krimm, S. Vibrational Analysis of Glutathione.
Biopolymers 1994, 34, 1377−1394.
(37) Wang, L.; Middleton, C. T.; Singh, S.; Reddy, A. S.; Woys, A.
M.; Strasfeld, D. B.; Marek, P.; Raleigh, D. P.; de Pablo, J. J.; Zanni,
M. T.; Skinner, J. L. 2DIR Spectroscopy of Human Amylin Fibrils
Reflects Stable B-Sheet Structure. J. Am. Chem. Soc. 2011, 133,
16062−16071.
(38) Hahn, S.; Kim, S.-S.; Lee, C.; Cho, M. Characteristic Two-
Dimensional IR Spectroscopic Features of Antiparallel and Parallel B-
Sheet Polypeptides: Simulation Studies. J. Chem. Phys. 2005, 123,
084905−084912.
(39) Moran, S. D.; Zanni, M. T. How to Get Insight into Amyloid
Structure and Formation from Infrared Spectroscopy. J. Phys.
Chem.Lett. 2014, 5, 1984−1993.
(40) Lomont, J. P.; Rich, K. L.; Maj, M.; Ho, J.-J.; Ostrander, J. S.;
Zanni, M. T. Spectroscopic Signature for Stable B-Amyloid Fibrils
Versus β-Sheet-Rich Oligomers. J. Phys. Chem. B 2018, 122, 144−153.
(41) Marsh, D. Dichroic Ratios in Polarized Fourier Transform
Infrared for Nonaxial Symmetry of Beta-Sheet Structures. Biophys. J.
1997, 72, 2710−2718.
(42) Greenfield, N. J. Using Circular Dichroism Spectra to Estimate
Protein Secondary Structure. Nat. Protoc. 2007, 1, 2876−2890.
(43) Panda, J. J.; Chauhan, S. V. Short peptide based self-assembled
nanostructures: implications in drug delivery and tissue engineering.
Polym. Chem. 2014, 5, 4418−4436.
(44) Nowick, S. J. Exploring B-Sheet Structure and Interactions with
Chemical Model Systems. Acc. Chem. Res. 2008, 41, 1319−1330.
(45) Graf, P.; Masic, A.; Thunemann, F. A.; Taubert, A.; et al.
Peptide-Coated Silver Nanoparticles: Synthesis, Surface Chemistry,
https://dx.doi.org/10.1021/acs.langmuir.0c01527
Langmuir 2020, 36, 11255−11261
Langmuir pubs.acs.org/Langmuir Article

and pH Triggered, Reversible Assembly into Particle Assemblies.

Chem. Eur. J. 2009, 15, 5831−5844.
(46) Osawa, M. Dynamic Processes in Electrochemical Reactions
Studied by Surface-Enhanced Infrared Absorption Spectroscopy
(SEIRAS). Bull. Chem. Soc. Jpn. 1997, 70, 2861−2880.
(47) Neubrech, F.; Pucci, A.; Cornelius, T. W.; Karim, S.; García-
Etxarri, A.; Aizpurua, J. Resonant Plasmonic and Vibrational Coupling
in a Tailored Nanoantenna for Infrared Detection. Phys. Rev. Lett.
2008, 101, No. 157403.
(48) Enders, D.; Nagao, T.; Pucci, A.; Nakayama, T.; Aono, M.
Surface-Enhanced Atr-Ir Spectroscopy with Interface-Grown Plas-
monic Gold-Island Films Near the Percolation Threshold. Phys. Chem.
Chem. Phys. 2011, 13, 4935−4941.
(49) Osawa, M.; Ataka, K.-I.; Yoshii, K.; Nishikawa, Y. Surface-
Enhanced Infrared Spectroscopy: The Origin of the Absorption
Enhancement and Band Selection Rule in the Infrared Spectra of
Molecules Adsorbed on Fine Metal Particles. Appl. Spectrosc. 1993,
47, 1497−1502.
(50) Ghosh, H.; Bürgi, T. Mapping Infrared Enhancement around
Gold Nanoparticles Using Polyelectrolytes. J. Phys. Chem. C 2017,
121, 2355−2363.
(51) Neubrech, F.; Huck, C.; Weber, K.; Pucci, A.; Giessen, H.
Surface-Enhanced Infrared Spectroscopy Using Resonant Nano-
antennas. Chem. Rev. 2017, 117, 5110−5145.
(52) Nishikawa, Y.; Nagasawa, T.; Fujiwara, K.; Osawa, M. Silver
Island Films for Surface-Enhanced Infrared Absorption Spectroscopy:
Effect of Island Morphology on the Absorption Enhancement. Vib.
Spectrosc. 1993, 6, 43−53.
(53) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J.
Ab Initio Calculation of Vibrational Absorption and Circular
Dichroism Spectra Using Density Functional Force Fields. J. Phys.
Chem. A 1994, 98, 11623−11627.
(54) Becke, A. D. Density-functional thermochemistry. III. The role
of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652.
(55) Adamczyk, K.; Simpson, N.; Greetham, G. M.; Gumiero, A.;
Walsh, M. A.; Towrie, M.; Parker, A. W.; Hunt, N. T. Ultrafast
Infrared Spectroscopy Reveals Water-Mediated Coherent Dynamics
in an Enzyme Active Site. Chem. Sci. 2015, 6, 505−516.
(56) Merga, G.; Cass, L. C.; Chipman, D. M.; Meisel, D. Probing
Silver Nanoparticles During Catalytic H2 Evolution. J. Am. Chem. Soc.
2008, 130, 7067−7076.

11261 https://dx.doi.org/10.1021/acs.langmuir.0c01527
Langmuir 2020, 36, 11255−11261
 Article

Cite This: J. Phys. Chem. B 2019, 123, 7771−7776 pubs.acs.org/JPCB

Vibrational Coupling on Stepwise Hydrogen Bond Formation of

Amide I
Anup Ghosh*
Department of Condensed Matter Physics and Materials Sciences, S. N. Bose National Centre for Basic Sciences, JD Block,
Sector-III, Salt Lake City, Kolkata 700 106, India
*
S Supporting Information

ABSTRACT: Despite the key roles of proteins and nucleic

Downloaded via S.N. BOSE NATL CTR BASIC SCIENCES on October 15, 2019 at 11:21:09 (UTC).
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

acids in biology, understanding their labile structures and

hydrogen bond interactions with guest molecules has posed a
critical challenge to the scientific community. In this report, I
have used dimethylformamide as a model amide to account
for amide hydrogen bond interactions of protein. To quantify
hydrogen bond conformation and the structural change, I
have monitored the amide I infrared (IR) stretching
frequencies while varying the pKa of phenol derivatives. For
all phenol derivatives, amide I has formed one hydrogen bond
and two hydrogen bond conformation. It has been observed
that the formation constant for one hydrogen bond is higher
than that of two hydrogen bonds for all phenol derivatives.
During the formation of hydrogen bond with amide I, IR absorbance of CC transition is enhanced for all phenol derivatives.
Enhancement of the IR absorbance of the CC transition indicates hydrogen bond-assisted vibrational coupling between the
amide I and phenol ring transition. The relative coupling constant is estimated to be higher for single hydrogen-bonded
conformer than the double hydrogen-bonded conformer. This is an intriguing result as the frequency difference between the
two coupled transitions predicts otherwise. Using IR absorption spectroscopy, a delicate interplay between hydrogen bonding
conformations and intermolecular vibrational coupling between amide I and H-bond donor phenol molecules has been shown.
This study can be used as a point of reference for understanding the structural information of proteins, peptides, and
nucleosides having hydrogen bond interaction with any drug or ligand molecules. My results as well provide an insight into the
vibrational coupling of carbonyl and CC transition of nucleobases.

■ INTRODUCTION
Hydrogen bond (hb) interactions are ubiquitous in protein
molecules and play key roles in numerous biological processes
like molecular association, catalysis, and signal transmission.1−7
Although peptide (−CONH−) units in proteins provide
several infrared (IR) signatures (amide I, amide II, amide III,
and amide A), the amide I band (1600−1700 cm−1) is mostly
studied in conjunctions to the protein structure and dynamics
because of its sensitivity to different secondary structures.
Moreover, the amide I mode is not strongly influenced by the
side-chain absorptions.5 Further, amide I is widely utilized as
IR probes because of their sensitivity to the local electrostatic
environment, solvation, and large molar extinction coeffi-
cient.8−12 In particular, IR absorption measurements of amide
I band is sensitive in a variety of biological conformations and
solvation in a way that it reflects vibrational couplings and
hydrogen bonding.4−7 A lot of experiments have been done to
investigate the relationships between amide I vibrational
couplings and peptide or protein conformations, and most
recently, peptide conformational dynamics has been stud-
ied.8−16 Theoretical two-dimensional IR (2D IR) spectro-
scopic approach has been taken to model the vibrational
coupling.17−19 Recently, vibrational couplings between carbon-
yl bonds located on DNA nucleobases have been revealed by
2D IR spectroscopy.20−22 Coupling between two isotope-based
nitrile groups (IR reporter) has been studied on DNA
templates.23 Parallel β-sheet vibrational couplings were
revealed by 2D IR spectroscopy.24 Vibrational dynamics and
vibrational modes of coupling of amide A and amide I/II
modes within the same amide unit and those connected by a
hb for several model dipeptides have been studied by two-color
2D IR spectroscopy.25 As hb’s of amide I make a very crucial
role in the structural conformation of protein and peptide,
vibrational dynamic studies of amide I is a recent utmost topic
of interest.26−32 To understand hydrogen bonding ability of
amide I with hb donor phenol derivatives, dimethyl formamide
and dimethyl acetamide have been studied.33,34 Recently, a
single peptide unit N-methylacetamide has been studied to
explore the nature of solute−solvent interactions on amide I
vibrational dynamics by using 2D IR experiments and
Received: May 29, 2019
Revised: August 24, 2019
Published: August 24, 2019

© 2019 American Chemical Society 7771 DOI: 10.1021/acs.jpcb.9b05118

J. Phys. Chem. B 2019, 123, 7771−7776
The Journal of Physical Chemistry B
molecular dynamics simulations.9,35−38 However, the amide I
transition is complex as it is generally delocalized.5
To simplify this challenging task and to gain molecular
insights of the amide I mode in the presence of intermolecular
hb’s, we have taken dimethylformamide (DMF) as a model
molecule to study the amide I vibrational frequency sensitivity
using linear IR spectroscopy. I have demonstrated the
sensitivity of the amide I IR absorption frequency in DMF
toward phenol derivatives (phenol, parachloro phenol, para
cresol, and anisole) and purine (basic skeleton of guanosine).
To understand the relationship between H-bonding con-
formation and hb dynamics of amide I, I have examined
correlations between the hb conformation formation constant
and populations with the phenol derivatives concentration and
also the vibrational coupling of amide I with CC transition
(phenol ring transition) of phenol derivatives. By using the
linear IR spectroscopy, I have shown that the hb’s are
ubiquitous with CO···HO in DMF in the presence of H-bond
donor phenol derivatives and how it has affected the amide I
transition and CC transition of guest molecules. The
formation constant of amide−phenol hb complexes is not
only described solely hydrogen bonding conformations but
also allows the estimation of contributions of 1 hb and 2 hb
conformations with variation of phenol concentrations. To do
so, a new method for predicting relative populations of 1 hb
and 2 hb complexes has been proposed by employing linear IR
spectroscopic analysis. Finally and most significantly, the
frequency correlation of amide I and CC IR absorbance has
been monitored for different hb conformations. In addition,
the effect of different chemical structures and pKa of the other
two phenol derivatives on the hb pairing and frequency−
frequency correlation with amide I has been studied in a
similar manner. Overall, this inclusive vibrational structural
information can be utilized in tandem to solve many biological
as well as chemical problems.
■ EXPERIMENTAL SECTION
Phenol, para-chlorophenol (p-CP), para-cresol (PC), anisole,
DMF, and methyl acetate (MA) are purchased from Sigma-
Aldrich and used without any further purification. A solution of
0.1 M DMF in CCl4 was used for Fourier transform infrared
(FTIR) experiments. The sample was placed in a homemade
FTIR sample cell with CaF2 windows and 60 μm Teflon
spacer. The linear IR absorption spectrum was measured using
an FTIR spectrometer (JASCO-FTIR-6300). The solvent
background of CCl4 was also measured and subtracted from all
the linear spectra. The same procedure was followed for MA−
phenol hb interaction in CCl 4 and purine−DMF in
acetonitrile.
■
RESULT AND DISCUSSION
A series of the IR absorption spectra of 0.1 M DMF solutions
in carbon tetrachloride with phenol concentrations ranging
from 0 to 0.5 M were collected. The representative spectra are
shown in Figure 1a. In the absence of phenol molecules, one
black single peak spectrum with frequency 1686 cm−1 of amide
I transition was obtained in Figure 1a. In the lower range of
concentrations, the phenol concentrations were increased in
steps of 0.005 M up to a concentration of 0.050 M (to avoid
self-aggregation of phenol). Absorption of amide I transition
was gradually decreased and new hydrogen-bonded amide I
transition was generated in the red-shifted region. The singular
7772
Article
Figure 1. (a) IR absorption spectra of DMF−phenol complexes of 0.1
M solutions of DMF in carbon tetrachloride with increasing
concentrations of phenol (0.000 Mblack, 0.005 Mred, 0.010
Mgreen, 0.020 Mblue, 0.030 Mcyan, 0.040 Mmagenta,
0.050 Myellow, 0.075 Mdark yellow, 0.10 Mnavy, 0.15 M
purple, 0.20 Mwine, 0.25 Molive, 0.30 Mdark cyan, 0.35 M
royal, 0.40 Morange, 0.45 Mviolet, 0.50 Mpink). (b)
Individual line shapes of 0 hb (black), 1 hb (red), and 2 hb
(green) conformations of DMF. (c) Relative population of different
H-bonded conformations for various concentrations of phenol. The
color code is the same as in (b). (d) Relative fraction of the H-bonded
DMF plotted for various concentrations of free phenol. The solid red
line shows a fit to the binding isotherm.
value decomposition (SVD) analysis,39−41 conducted on the
lower concentrations data, indicates the presence of two
significant components (6.81 and 0.872) out of the three major
components (6.81, 0.872, and 0.024), such that the third
component can be neglected undoubtedly. I have performed
multivariate curve resolution (MCR) analysis on the linear IR
data of amide I transition of DMF for the lower concentrations
range of phenol.40,41 During the MCR analysis on lower
concentrations data, two distinct components with different
line shapes of amide I were observed. For a higher
concentrations range, phenol concentration was increased by
a step of 0.05 M up to 0.5 M. The SVD analysis of the higher
concentrations of phenol indicates the presence of three
significant components, with the first four singular values of
7.37, 3.80, 0.807, and 0.111. Thus, the fourth component could
be easily neglected. The obtained line shapes for both 0 hb and
1 hb complexes at the lower concentrations of phenol (up to
0.050 M) was used as the initial guess in the MCR analysis of
the high phenol concentration spectral series (up to 0.5 M).
The final MCR results were checked for consistency such
that the line shapes obtained from the low and high
concentration series become identical. The obtained spectral
profiles of the individual components, bonding to either a
single phenol (1 hb) or to a pair of phenol molecules (2 hb)
with transition frequency 1673 and 1664 cm−1, respectively,
are presented in Figure 1b. Amide I absorption frequency shifts
to a different extent in each case because hb formation reduces
electron density from amide I (CO bond) and consequently
red-shifts of its stretching frequency. Ensemble populations of
all components were extracted from the individual peak area,
obtained from MCR analysis in the presence of different
DOI: 10.1021/acs.jpcb.9b05118
J. Phys. Chem. B 2019, 123, 7771−7776
The Journal of Physical Chemistry B
Figure 2. IR absorbance spectra of 0.1 M of (a) phenol, (b) PC, and (c) p-CP with the increasing concentration of DMF (0.00 Mblack, 0.02
Mred, 0.04 Mgreen, 0.06 Mblue, 0.08 Mcyan, 0.10 Mmagenta).
concentrations of phenol. Figure 1c shows the population
ensembles as a function of phenol concentrations. Three
distinct concentration regimes are observed for 0 hb, 1 hb, and
2 hb conformations. The molar integrated extinction
coefficient of 1 hb conformation of amide I was calculated
from lower concentrations data of phenol and that value was
used to calculate the concentrations of 2 hb conformation in
higher concentration series. As in lower concentrations of
phenol, only the 1 hb conformation formed with DMF, the
corresponding data were fitted to a 1:1 binding isotherm Y11 =
(K1X)/(1 + K1X), where Y is the fraction of the H-bonded
DMF, X is the concentration of the free phenol molecules in
solution, and K1 is the hb formation constant of the 1 hb
complex with value 26.41. At higher phenol concentrations, the
corresponding data were fitted to a 1:2 binding isotherm
(Figure 1d) given by Y12 = (K1X + K1K2X2)/(1 + K1X +
K1K2X2) where K2 is the stepwise formation constant of the 2
hb complex with the value of 6.98.42−44
Two CC antisymmetric and symmetric stretching modes
of the phenol ring have a Lorentzian line shape with transition
frequencies 1597 and 1605 cm−1 (Figure 2a). The data
presented in Figure 2a suggest that upon the formation of the
hb complexes, IR absorbance of CC transition of phenol
enhances gradually and frequency shifts slightly toward lower
frequency with the increasing concentrations of DMF. Now
the question arises, why the CC transition absorption has
enhanced. This enhancement could be the reason of n−π*
bonding of DMF and phenol ring or polarity change or altered
electron density of the vibrating bond or vibrational coupling.
To rule out the effect of n−π* bonding of lone pair of “O”
atom in DMF and “π” cloud of phenol ring, 1:1 concentration
of DMF (0.1 M) and anisole (0.1 M) were mixed and the IR
absorption spectrum was collected. The spectra in Figure 4a
clearly suggest that there was no interaction of nonbonded
electron of “O” atom in DMF with the “π” electron cloud of
anisole. This result also confirms that the CC IR transition
enhancements are not due to the polarity change of the
solvent, as the volume of additional DMF (9.4 μL) in anisole
solution (990.6 μL) and other phenol derivatives solution
(990.6 μL) are same. To check the altered electron density of
the vibrating bond, IR absorption spectra were collected from
MA and phenol in 1:1 concentration. The spectra shown in
Figure 4b speak the formation of hb conformation of MA (0.1
M) with phenol (0.1 M), but CC IR absorption enhance-
ment, position, and line shape were negligibly changed (Figure
4c). These results invalidate the arguments of enhancement
because of altered electron density of the vibrating bond. Thus,
the observed enhancement (Figure 2a) in IR absorption clearly
indicates the presence of vibrational coupling even though
amide I and CC transitions have ∼59 cm−1 (1 hb) and ∼68
cm−1 (2 hb) frequency gap. Density functional theory (DFT)
calculation videos have strengthened my argument of vibra-
7773
Article
tional coupling (see Videos S1 and S2). Tucker and co-workers
have shown the vibrational coupling of azide (2124 cm−1) and
nitrile (2242−1) reporters in RNA nucleoside even if they have
a 118 cm−1 frequency gap.45 Thus, their work also validates our
observation. The amide I and CC transition couples their
local modes to create delocalized normal modes. Hence, Figure
2a−c shows the clear agreement of vibrational coupling in
between amide I transition of DMF and CC transition of
phenol, PC, and p-CP, respectively. Thus, people can predict
coupled modes and therefore can identify the molecular
structures by either the frequency shift or the IR absorbance
change caused by delocalization of the vibrational modes. The
obtained results from Figure 4b,c clearly indicate that MA
(1748 cm−1) and phenol (1597, 1605 cm−1) have enough
frequency gap (∼140 cm−1) for not coupling or negligible
coupling. Now the question is the dependency of relative
coupling strength on individual hb conformation. As CC
transition for 0 hb, 1hb, and 2 hb conformations are very close
to each other, it is quite difficult to isolate the individual
transition peak and peak area contribution (Figure 2a). To do
so, 0.1 M of DMF has been taken, and the same experiment
like Figure 1a was repeated by monitoring the CC transition
(Figure 3a) from 1580 to 1620 cm−1. To calculate the
individual peak area of CC transition for free, 1 hb, and 2 hb
conformations, the integrated extinction coefficient of
respective conformations have been calculated. The integrated
extinction coefficient of CC transition of free phenol was
calculated from the standard plot, the concentration of free
phenol versus CC transition IR absorbance (Figure 3b). To
calculate the integrated extinction coefficient of 1 hb
conformation, lower concentrations (0−0.05 M) of phenol
series has been taken to avoid the formation of 2 hb
conformation. As the concentrations of free phenol and 1 hb
conformations were known in lower concentration series, the
area contribution of free phenol was calculated by multiplying
the integrated extinction coefficient with the concentration of
free phenol. The area contribution of CC transition of 1 hb
conformation was calculated by subtracting the area of CC
transition of free phenol from total ensemble area of lower
concentration series (0−0.05 M). The subtracted area versus
concentration of 1 hb conformation was plotted in Figure 3c to
calculate the integrated extinction coefficient CC transition
of 1 hb conformation. Similarly, to calculate the area
contribution of CC transition for 2 hb conformation, the
area contribution of free phenol and 1 hb conformation was
subtracted from the total CC transition area of ensembles in
higher concentration series. Area of CC transition versus 2
× concentration of 2 hb conformation was plotted in Figure 3d
to calculate the integrated extinction coefficient. By comparing
the slope of 0 hb, 1 hb, and 2 hb, it is clear that the slope
of Figure 3c is higher than that of Figure 3d, and consequently,
the slope of Figure 3d is higher than that of 3a, which
DOI: 10.1021/acs.jpcb.9b05118
J. Phys. Chem. B 2019, 123, 7771−7776
The Journal of Physical Chemistry B
Figure 3. (a) Spectra of CC transition of phenol in different
concentrations (0.000 Mblack, 0.005 Mred, 0.010 Mgreen,
0.020 Mblue, 0.030 Mcyan, 0.040 Mmagenta, 0.050 M
yellow, 0.075 Mdark yellow, 0.10 Mnavy, 0.15 Mpurple, 0.20
Mwine, 0.25 Molive, 0.30 Mdark cyan, 0.35 Mroyal, 0.40
Morange, 0.45 Mviolet, 0.50 Mpink) in the presence of 0.1 M
DMF. (b) Plot of CC IR absorbance area vs concentration of free
p-CP. (c) Plot of CC IR absorbance area vs concentration of 1 hb
complex. (d) Plot of CC IR absorbance area vs 2 × concentration
of 2 hb complex.
illustrates that CC transition in 1 hb conformation has
higher relative area enhancement than 2 hb conformation. The
relative area enhancement and average relative transition
dipole moment of CC transition of 1 hb and 2 hb
conformations with respect to free phenol have been extracted,
as shown in Table 1. The observed relative transition moment
revealed that amide I and CC transition coupled strongly for
1 hb conformation than 2 hb conformation.
The higher concentration of phenol leads to the formation
of 2 hb conformation, where the frequency of the amide I
mode was lowered than 1 hb conformation, and consequently,
frequency gap of amide I with CC transition frequency
becomes closer than 1 hb conformation (see Table 1).
Therefore, direct coupling between amide I and CC
transition dipole was expected to be more efficient for 2 hb
conformation than 1 hb conformation, but the experimental
result indicates the reverse. The resonant interaction of amide I
with the CC vibrational mode becomes even weaker for 2
hb conformation. The coupling of amide I and CC
transition dipole in H-bonded DMF and phenol could be
Figure 4. (a) IR absorbance of CC transition of anisole (0.1 M) in the absence (black) and presence (red) of DMF (0.1 M). (b) IR spectra of
MA (0.1 M) in the absence (black) and presence (red) of phenol (0.1 M) in CCl4. (c) IR spectra of CC transition of phenol (0.1 M) in the
absence (black) and presence (red) of MA (0.1 M).
7774
Article
due to the potential contribution of electrostatic, electro-
dynamics, and mechanical effects. However, to describe this
abnormal coupling behavior, the distance between amide I and
CC transition dipoles has been taken into account. By DFT
calculation, I have observed that amide I and CC transition
dipole distance is a bit higher for 2 hb conformation (3.58 Å)
than 1 hb conformation (3.25 Å). Therefore, amide I and C
C transition dipoles have been coupled strongly for 1 hb
conformation than 2 hb conformation by considering the same
angle between these two transition dipoles for both the
conformations. Another reason possibly is that the hb’s
mechanically couple the motions of amide I and CC
transition dipoles significantly for 1 hb conformation than 2 hb
conformation, as the 1 hb formation constant is stronger than
the 2 hb formation constant. This could also be a factor for this
abnormal behavior.
In order to determine whether the coupling observed in the
H-bonded conformations of DMF was specific to phenol,
similar experiments with p-CP and para cresol (PC) with
different chemical structures and pKa were carried out. The
corresponding IR absorption results for p-CP and PC are
shown in Figures S1 and S3, respectively. From the SVD and
MCR analysis, the individual line shape of 0 hb, 1hb, and 2 hb
conformations of DMF were obtained. From that, the
population of individual components was calculated and
plotted in Figures S1 and S3 for p-CP and PC, respectively.
The stepwise formation constant for p-CP (K1 = 60.9, K2 =
3.46) and PC (K1 = 12.36, K2 = 7.16) was calculated by fitting
the fractions of the hb complex formed with the free p-CP and
PC concentrations, respectively (see Figures S1 and S3). For
both the cases of p-CP and PC, K2 is much lesser than K1.
Similarly, relative area enhancement of CC transition from 1
hb and 2 hb conformations of p-CP and PC was calculated
(see Figures S2 and S4). The average relative transition dipole
moment suggests that amide I and CC transition coupled
strongly for 1 hb conformation than 2 hb conformation, as
shown Table 1 for both p-CP and PC also.
To quantify the information of hb mediated vibrational
coupling of amide I with CC transition in a nucleic acid
(DNA/RNA), purine (as hb donor) and DMF (as hb
acceptor) were studied in acetonitrile. Now, why purine?
Purine is the basic structural skeleton of guanine and adenine.
Nucleobases also have an amide I bond in their basic
structure. Figure 5 shows that IR absorbance of purine was
enhanced noticeably in the presence of DMF (1:1 concn ratio)
and while apparently for weak hb interaction, IR absorbance
frequency was unremarkably changed. This figure cleared the
presence of vibrational coupling of amide I and purine ring
transition during hb formation. This result indicates that
during the formation of hb in between complementary base
DOI: 10.1021/acs.jpcb.9b05118
J. Phys. Chem. B 2019, 123, 7771−7776
The Journal of Physical Chemistry B Article

Table 1. Analysis of H-Bond Formation Constant, Relative Integrated Area, and Average Relative Transition Dipole Momenta
H-bond donor pKa K1 (M−1) K2 (M−1) ≈
ΔCC−CO
1hb (cm−1) ≈
ΔCC−CO
2hb (cm−1) ICC CC
1hb/I0hb ICC CC
2hb/I0hb μCC CC
1hb/μ0hb μCC CC
2hb/μ0hb

p-CP 9.48 60.9 3.46 67 59 4.72/1 4.18/1 2.17/1 2.04/1

phenol 9.95 26.41 6.98 68 59 1.93/1 1.62/1 1.39/1 1.27/1
PC 10.18 12.36 7.16 55 48 2.97/1 1.88/1 1.72/1 1.37/1
anisole
a≈
ΔCC−CO
1hb and ≈ΔCC−CO
2hb are frequency gap in between CC and amide I transition, respectively. ICC CC CC CC
1hb/I0hb and I2hb/I0hb are the ratio of IR
1hb/μ0hb and μ2hb/μ0hb are the ratio of the relative transition dipole moment.
absorbance area of 1 hb and 2 hb with 0 hb, respectively. μCC CC CC CC

to quantitatively calculate the population of hb conformation

and qualitatively calculate the coupling constant.

■
*
ASSOCIATED CONTENT
S Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jpcb.9b05118.
Linear IR data of PC and p-CP (PDF)
Vibrational video of 1 hb conformation (from DFT
Figure 5. (a) Purine ring vibrational transition spectra in the absence calculation) (AVI)
(black) and presence (red) of DMF in acetonitrile with 1:1 Vibrational video of 2 hb conformation (from DFT
concentration ratio. (b) One hydrogen-bonded conformation of calculation) (AVI)

■
purine and DMF.
AUTHOR INFORMATION
pair of DNA, amide I and CC transition vibrationally may Corresponding Author
be coupled. As still there is no report of vibrational coupling *E-mail: anupg86@gmail.com, anup.ghosh@bose.res.in.
between carbonyl and ring transition of nucleobases, this study ORCID
could be a point of reference for future studies. Anup Ghosh: 0000-0002-1442-8740

■ CONCLUSIONS
Notes
The author declares no competing financial interest.
In conclusion, I have established experimentally the hydrogen-
bonded structural information of amide I in the presence of hb
donor phenol derivatives using linear IR spectroscopy. The
■ ACKNOWLEDGMENTS
This work was supported in part by DST Government of India
main spectroscopic features from all phenol derivatives [DST/INSPIRE/04/2018/002031].

approaches agree very well, enabling interpretation in terms
of hydrogen bonding population and stepwise hb formation-
based vibrational coupling. I found that amide I forms two
main hydrogen-bonded conformations, of which the vibra-
tional coupling of amide I and CC transition are different for
different hb conformations. Relative area and average transition
dipole moment enhancement of CC IR absorbance were
higher for 1 hb conformation than 2 hb conformation for all
the phenol derivatives, whereas the energy gap of the two
dipoles is less in the case of 2 hb conformation than 1 hb
conformation. For the first time, hb formation dependent
vibrational coupling of amide I where mechanical effects and
distance between the two coupled modes might play significant
roles has been reported. This observation definitely demon-
strates a delicate interplay between coupling effects and hb
formation constant of amide I. Thus, by linear IR spectroscopy,
I have unraveled the interplay between hydrogen bonding and
vibrational coupling. My study could be an important point of
reference for the protein and nucleic acid structure and
dynamics experiment to unravel the IR spectral signatures. I
believe that this report provides the most inclusive results into
the origins of the couplings in the host−guest hydrogen
bonding study in protein and peptides. The results presented
in this paper provide an insight into the quantification of hb
conformation of amide and quantification of the relative
coupling modes utilized to interpret the IR spectra of proteins,
peptides, and guest molecules. The detailed comparison among
all phenol derivatives strengthens our confidence in the ability
■
REFERENCES
(1) Dickerson, R. E.; Geis, I. The Structure and Action of Proteins; W.
A. Benjamin, Inc.: Menlo Park, CA, 1969.
(2) Nelson, D. L.; Cox, M. M. Lehninger Principles of Biochemistry;
W. H. Freeman and Company: New York, 2005.
(3) Berg, J. M.; Tymoczko, J. L.; Stryer, L. Biochemistry; W. H.
Freeman and Company: New York, 2007.
(4) Krimm, S.; Bandekar, J. Vibrational spectroscopy and
conformation of peptides, polypeptides, and proteins. Adv. Protein
Chem. 1986, 38, 181−364.
(5) Barth, A.; Zscherp, C. What vibrations tell us about proteins. Q.
Rev. Biophys. 2002, 35, 369−430.
(6) Woutersen, S.; Hamm, P. Nonlinear two-dimensional vibrational
spectroscopy of peptides. J. Phys.: Condens. Matter 2002, 14, R1035−
R1062.
(7) Hamm, P.; Lim, M.; Hochstrasser, R. M. Structure of the Amide
I Band of Peptides Measured by Femtosecond Nonlinear-Infrared
Spectroscopy. J. Phys. Chem. B 1998, 102, 6123−6138.
(8) Dyer, R. B.; Gai, F.; Woodruff, W. H.; Gilmanshin, R.; Callender,
R. H. Infrared Studies of Fast Events in Protein Folding. Acc. Chem.
Res. 1998, 31, 709−716.
(9) Woutersen, S.; Pfister, R.; Hamm, P.; Mu, Y.; Kosov, D. S.;
Stock, G. Peptide conformational heterogeneity revealed from
nonlinear vibrational spectroscopy and molecular-dynamics simu-
lations. J. Chem. Phys. 2002, 117, 6833−6840.
(10) Demirdöven, N.; Cheatum, C. M.; Chung, H. S.; Khalil, M.;
Knoester, J.; Tokmakoff, A. Two-Dimensional Infrared Spectroscopy
of Antiparallel β-Sheet Secondary Structure. J. Am. Chem. Soc. 2004,
126, 7981−7990.

7775 DOI: 10.1021/acs.jpcb.9b05118

J. Phys. Chem. B 2019, 123, 7771−7776
The Journal of Physical Chemistry B
(11) Waegele, M. M.; Culik, R. M.; Gai, F. Site-Specific
Spectroscopic Reporters of the Local Electric Field, Hydration,
Structure, and Dynamics of Biomolecules. J. Phys. Chem. Lett. 2011, 2,
2598−2609.
(12) Serrano, A. L.; Waegele, M. M.; Gai, F. Spectroscopic studies of
protein folding: Linear and nonlinear methods. Protein Sci. 2012, 21,
157−170.
(13) Hamm, P.; Hochstrasser, R. M. Structure and dynamics of
proteins and peptides: femtosecond two-dimensional infrared spec-
troscopy. In Ultrafast Infrared and Raman Spectroscopy; Fayer, M. D.,
Ed.; Marcel Dekker, Inc.: New York, 2001; Vol. 26, p 273.
(14) Gnanakaran, S.; Hochstrasser, R. M. Conformational
Preferences and Vibrational Frequency Distributions of Short
Peptides in Relation to Multidimensional Infrared Spectroscopy. J.
Am. Chem. Soc. 2001, 123, 12886−12898.
(15) Woutersen, S.; Hamm, P. Isotope-edited two-dimensional
vibrational spectroscopy of trialanine in aqueous solution. J. Chem.
Phys. 2001, 114, 2727−2737.
(16) Woutersen, S.; Mu, Y.; Stock, G.; Hamm, P. Subpicosecond
conformational dynamics of small peptides probed by two-dimen-
sional vibrational spectroscopy. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
11254−11258.
(17) Zhang, W. M.; Chernyak, V.; Mukamel, S. Multidimensional
femtosecond correlation spectroscopies of electronic and vibrational
excitons. J. Chem. Phys. 1999, 110, 5011−5028.
(18) Scheurer, C.; Piryatinski, A.; Mukamel, S. Signatures of β-
Peptide Unfolding in Two-Dimensional Vibrational Echo Spectros-
copy: A Simulation Study. J. Am. Chem. Soc. 2001, 123, 3114−3124.
(19) Moran, A.; Mukamel, S. The origin of vibrational mode
couplings in various secondary structural motifs of polypeptides. Proc.
Natl. Acad. Sci. U.S.A. 2004, 101, 506−510.
(20) Krummel, A. T.; Zanni, M. T. DNA vibrational coupling
revealed with two-dimensional infrared spectroscopy: Insight into why
vibrational spectroscopy is sensitive to DNA structure. J. Phys. Chem.
B 2006, 110, 13991−14000.
(21) Krummel, A. T.; Mukherjee, P.; Zanni, M. T. Inter and
Intrastrand Vibrational Coupling in DNA Studied with Heterodyned
2D-IR Spectroscopy. J. Phys. Chem. B 2003, 107, 9165−9169.
(22) Krummel, A. T.; Zanni, M. T. Interpreting DNA Vibrational
Circular Dichroism Spectra Using a Coupling Model from Two-
Dimensional Infrared Spectroscopy. J. Phys. Chem. B 2006, 110,
24720−24727.
(23) Krummel, A. T.; Zanni, M. T. Evidence for Coupling between
Nitrile Groups Using DNA Templates: A Promising New Method for
Monitoring Structures with Infrared Spectroscopy. J. Phys. Chem. B
2008, 112, 1336−1338.
(24) Woys, A. M.; Almeida, A. M.; Wang, L.; Chiu, C.-C.;
McGovern, M.; de Pablo, J. J.; Skinner, J. L.; Gellman, S. H.; Zanni,
M. T. Parallel β-Sheet Vibrational Couplings Revealed by 2D IR
Spectroscopy of an Isotopically Labeled Macrocycle: Quantitative
Benchmark for the Interpretation of Amyloid and Protein Infrared
Spectra. J. Am. Chem. Soc. 2012, 134, 19118−19128.
(25) Rubtsov, I. V.; Wang, J.; Hochstrasser, R. M. Vibrational
Coupling between Amide-I and Amide-A Modes Revealed by
Femtosecond Two Color Infrared Spectroscopy†. J. Phys. Chem. A
2003, 107, 3384−3396.
(26) Mirkin, N. G.; Krimm, S. Ab initio vibrational analysis of
hydrogen-bonded trans- and cis-N-methylacetamide. J. Am. Chem. Soc.
1991, 113, 9742−9747.
(27) Mirkin, N. G.; Krimm, S. Ab initio vibrational analysis of
isotopic derivatives of aqueous hydrogen-bonded trans-N-methyl-
acetamide. J. Mol. Struct. 1996, 377, 219−234.
(28) Torii, H.; Tatsumi, T.; Kanazawa, T.; Tasumi, M. Effects of
Intermolecular Hydrogen-Bonding Interactions on the Amide I Mode
ofN-Methylacetamide: Matrix-Isolation Infrared Studies and ab Initio
Molecular Orbital Calculations. J. Phys. Chem. B 1998, 102, 309−314.
(29) Herrebout, W. A.; Clou, K.; Desseyn, H. O. Vibrational
Spectroscopy ofN-Methylacetamide Revisited. J. Phys. Chem. A 2001,
105, 4865−4881.
7776
Article
(30) Kubelka, J.; Keiderling, T. A. Ab Initio Calculation of Amide
Carbonyl Stretch Vibrational Frequencies in Solution with Modified
Basis Sets. 1.N-Methyl Acetamide. J. Phys. Chem. A 2001, 105,
10922−10928.
(31) Ham, S.; Kim, J.-H.; Lee, H.; Cho, M. Correlation between
electronic and molecular structure distortions and vibrational
properties. II. Amide I modes of NMA-nD2O complexes. J. Chem.
Phys. 2003, 118, 3491−3498.
(32) Ham, S.; Cho, M. Amide I modes in the N-methylacetamide
dimer and glycine dipeptide analog: Diagonal force constants. J. Chem.
Phys. 2003, 118, 6915−6922.
(33) Hatton, J. V.; Richards, R. E. Solvent effects in the proton
resonance spectra of dimethyl-formamide and dimethyl-acetamide.
Mol. Phys. 1960, 3, 253−263.
(34) Malathi, M.; Sabesan, R.; Krishnan, S. IR carbonyl band
intensity studies in N, N-dimethyl formamide and N, N- dimethyl
acetamide on formation with phenols. Curr. Sci. 2002, 86, 838−842.
(35) Woutersen, S.; Mu, Y.; Stock, G.; Hamm, P. Hydrogen-bond
lifetime measured by time-resolved 2D-IR spectroscopy: N-methyl-
acetamide in methanol. Chem. Phys. 2001, 266, 137−147.
(36) Zanni, M. T.; Asplund, M. C.; Hochstrasser, R. M. Two-
dimensional heterodyned and stimulated infrared photon echoes of
N-methylacetamide-D. J. Chem. Phys. 2001, 114, 4579−4590.
(37) Kwac, K.; Cho, M. Molecular dynamics simulation study of N-
methylacetamide in water. I. Amide I mode frequency fluctuation. J.
Chem. Phys. 2003, 119, 2247−2255.
(38) Kwac, K.; Lee, H.; Cho, M. Non-Gaussian statistics of amide I
mode frequency fluctuation of N-methylacetamide in methanol
solution: linear and nonlinear vibrational spectra. J. Chem. Phys.
2004, 120, 1477−1490.
(39) Lawton, W. H.; Sylvestre, E. A. Self Modeling Curve
Resolution. Technometrics 1971, 13, 617−633.
(40) Borgen, O. S.; Kowalski, B. R. An extension of the multivariate
component-resolution method to three components. Anal. Chim. Acta
1985, 174, 1−26.
(41) Jaumot, J.; de Juan, A.; Tauler, R. MCR-ALS GUI 2.0: new
features and applications. Chemom. Intell. Lab. Syst. 2015, 140, 1−12.
(42) Connors, K. A. Binding Constants: The Measurement of
Molecular Complex Stability; Wiley: New York, 1987; Vol. 33.
(43) Thordarson, P. Determining association constants from
titration experiments in supramolecular chemistry. Chem. Soc. Rev.
2011, 40, 1305−1323.
(44) Hibbert, D. B.; Thordarson, P. The death of the Job plot,
transparency, open science and online tools, uncertainty estimation
methods and other developments in supramolecular chemistry data
analysis. Chem. Commun. 2016, 52, 12792−12805.
(45) Schmitz, A. J.; Hogle, D. G.; Gai, X. S.; Fenlon, E. E.; Brewer, S.
H.; Tucker, M. J. Two-Dimensional Infrared Study of Vibrational
Coupling between Azide and Nitrile Reporters in a RNA Nucleoside.
J. Phys. Chem. B 2016, 120, 9387−9394.
DOI: 10.1021/acs.jpcb.9b05118
J. Phys. Chem. B 2019, 123, 7771−7776
 Letter

Cite This: J. Phys. Chem. Lett. 2019, 10, 2481−2486 pubs.acs.org/JPCL

Two-Dimensional Infrared Spectroscopy Reveals Molecular Self-

Assembly on the Surface of Silver Nanoparticles
Anup Ghosh,† Amit K. Prasad,† and Lev Chuntonov*
Schulich Faculty of Chemistry and Solid State Institute, Technion − Israel Institute of Technology, Haifa 3200003, Israel
*
S Supporting Information

ABSTRACT: The conformation of molecules, peptides, and proteins, self-

assembled into structured monolayers on the surface of metal nanoparticles
(NPs), can strongly affect their properties and use in chemical or nanobiomedical
Downloaded by S.N. BOSE NATL CTR BASIC SCIENCES at 02:37:45:730 on June 03, 2019

applications. Elucidating molecular conformations on the NP surface is highly

challenging, and the microscopic details mostly remain elusive. Using polarization-
selective third-order two-dimensional ultrafast infrared spectroscopy, we revealed
the highly ordered intermolecular structure of γ-tripeptide glutathione on the surface
of silver NPs in aqueous solution. Glutathione is an antioxidant thiol abundant in
from https://pubs.acs.org/doi/10.1021/acs.jpclett.9b00530.

living cells; it is extensively used in NP chemistry and related research. We identified conditions where the interaction of
glutathione with the NP surface facilitates formation of a β-sheet-like structure enclosing the NPs. A spectroscopic signature
associated with the assembly of β-sheets into an amyloid fibril-like structure was also observed. Remarkably, the interaction with
the metal surface promotes formation of a fibril-like structure by a small peptide involving only two amino acids.

F unctionalized nanomaterials have recently been the focus

of extensive scientific research directed toward better
understanding the interaction of their surfaces with (bio)-
molecules.1−4 Having a detailed understanding of factors
affecting the self-assembly of molecules, peptides, and proteins
on the surface of noble-metal nanoparticles (NPs) is important
for numerous applications, including nano-optics,5 biosensors,6
heterogeneous catalysis,7 NP-based drug delivery,8,9 etc.10,11
The microscopic structure of the self-assembled molecular
monolayer (SAM) can have an enormous effect on the NP
properties, for example, facilitating efficient cell membrane
penetration.12 Glutathione (GSH) is a tripeptide (γ-Glu-Cys-
Gly) whose thiol group acts as an antioxidant in cells, where
GSH is abundant; it is extensively used as a NP capping
ligand.13 The uptake of the GSH-protected NPs by cancer cells
was found to be superior to NPs with mixed SAMs involving
GSH molecules.14,15 Small GSH-coated gold NPs were
proposed as an efficient cell delivery system16 and as amyloid
fibril inhibiting agents.11 Presumably, these special properties
of the GSH-coated NPs stem from the multiple possibilities
available for intermolecular hydrogen bonding in GSH,
allowing for the formation of the organized capping layer by
this ligand.17−20
A detailed structural characterization of intermolecular and
molecule−NP surface interactions is needed to fully explore
opportunities for the rational design of SAMs in NP
applications;21−23 however, achieving such a characterization
is often challenging.23−27 GSH monolayers on gold surfaces
were extensively studied by various methods.28−33 It was
revealed that in aqueous solutions GSH attaches to the metal
via the cysteine, and that at a pH range of 4−9, it appears in
anionic (Gly) and zwitterionic (Glu) forms.19,20,28−33 How-
ever, despite the extensive investigations, a detailed micro-
scopic picture of the GSH SAM structure has remained elusive.
Herein, we meet this challenge by applying a full arsenal of
two-dimensional ultrafast infrared spectroscopy (2DIR)
tools.34−37 In 2DIR, the spectrum is spread in two dimensions
allowing for separation of the line shape components and
resolving molecular conformations with congested spectra. The
cross-peaks indicate the coupling between the vibrational
modes and the associated transfer of the vibrational
excitation.38−40
In studies of biomolecules, infrared spectroscopy of amide
carbonyl stretching vibrational mode (amide-I) is particularly
informative.41 Indeed, 2DIR spectroscopy of free GSH in
solution indicates coupling between its two amide-I modes
with the corresponding cross-peaks, facilitating their assign-
ment. On the other hand, GSH SAMs on the NP surface
reveals spectroscopic signatures of the molecular organization,
typical of the β-sheets formed in proteins and peptides.41−44
Here, the coupling between multiple amide-I modes of the
amino acids comprising the β-sheet results in the appearance of
a collective delocalized exciton mode, red-shifted from their
individual transitions.45,46 Recently, Lomont et al.47 found that
in contrast to β-sheet-rich oligomers, when amyloid fibrils are
formed, stronger coupling between the amide-I modes leads to
the emergence of an additional red-shifted transition. Similar
signatures were observed in our experiments, suggesting that
GSH SAM is organized in β-sheet-like and fibril-like structures.
The GSH molecules were self-assembled on the surface of
the 40 nm silver NPs;48 the pH of the solution was 5.5. Linear
absorption of the amide-I mode of free GSH in D2O (Figure
1A) shows several overlapping broadband transitions. 2DIR
Received: February 24, 2019
Accepted: April 12, 2019
Published: April 12, 2019

© 2019 American Chemical Society 2481 DOI: 10.1021/acs.jpclett.9b00530

J. Phys. Chem. Lett. 2019, 10, 2481−2486
The Journal of Physical Chemistry Letters
Figure 1. Infrared spectroscopy of GSH. Linear absorption (A) and
2DIR spectrum (B) of a 0.1 M solution of free GSH in D2O, pD 5.5.
Linear absorption (C) and 2DIR spectrum (D) of SAM GSH on the
surface of 40 nm silver NPs. GSH molecules are illustrated above with
the transition dipole moments of the amide-I modes indicated by red
arrows. Both parallel and antiparallel β-sheet-like conformations are
illustrated; see text for discussion.
spectroscopy (Figure 1B) resolves the congested spectra into
four peaks assigned to amide-I transitions in Cys and Gly (ωex
Figure 2. Polarization-selective 2DIR spectroscopy of free GSH. The polarization of the excitation pulses and the waiting times corresponding to
the spectra in each column are indicated at the top. The pD values for each row appear on the right.
2482
Letter
= 1665 and 1645 cm−1) and to the antisymmetric stretching of
the deprotonated carboxylic groups of the Glu and Gly units
(ωex = 1618 and 1595 cm−1).29,49 This assignment is
supported by the ability of the polarization-selective 2DIR to
identify the intermode coupling.38,39 A series of waiting time
(Tw)-dependent spectra recorded for pD values of 2, 5.5, and
12 are shown in Figure 2. For the all-parallel polarization of the
excitation pulses, ⟨XXXX⟩, the cross-peak between the coupled
amide-I modes is not seen clearly at Tw = 0.3 ps; however, it
becomes noticeable at Tw = 1.5 ps. For the cross-polarized
pulse sequence, ⟨YYXX⟩, the cross-peak becomes clearly visible
at ωex = 1645 cm−1; ωdet = 1670 cm−1, especially at Tw = 1.5
ps.39 This trend was observed for all pD values. At low pD
values, both carboxyl groups in glutathione are protonated
(pKa = 2 and 3.4 for Glu and Gly, respectively), and the
carboxylic stretching transitions appear at ωex = 1730 cm−1.
This transition is gradually replaced by bands at ωex = 1618
and 1595 cm−1 when the pD value rises; the latter is especially
intensified after the deprotonation of the amine on the Glu
unit (pKa = 9.5), as seen in Figure 2.29 Finally, the cross-peak
at ωex = 1618 cm−1; ωdet = 1675 cm−1 suggests mode
connectivity because of the short intramolecular distance
between the carboxylic stretching and the amide-I modes of
Gly.50,51
In contrast to the free GSH, the linear spectrum of GSH
SAM (Figure 1C) shows two narrowed transitions at ωex =
1628 and 1604 cm−1. The former is the well-known marker of
the vibrational exciton of β-sheets,45,52,53 whereas the latter
suggests a fibril-like structure.47 The corresponding 2DIR
spectrum in Figure 1D remarkably resembles that of the
amyloid fibrils in ref 47. To quantitatively analyze the spectrum
we fitted the linear spectrum of the SAM GSH and the
diagonal slice of its 2DIR spectrum to two peaks, as shown in
Figure S2. Assuming similar strengths of the transition dipole
moments of the β-sheet-like and fibril-like structures, the ratio
DOI: 10.1021/acs.jpclett.9b00530
J. Phys. Chem. Lett. 2019, 10, 2481−2486
The Journal of Physical Chemistry Letters
of the peak areas for both linear and 2DIR spectra represents
the ratio of the concentrations of the two conformations. We
obtained that ca. 15% of the sample adopts fibril-like
conformation, as compared to the 25% reported in ref 47 for
amyloid peptides.
Reduction in both the inhomogeneous and the homoge-
neous line width of the SAM GSH, as compared to the free
GSH, is reflected in the diagonal and the antidiagonal width of
the 2DIR peak, respectively. The line narrowing is consistent
with the formation of the ordered intermolecular structure
constrained to the NP surface.23,54 Narrower line shapes
generally indicate slower molecular dynamics. The correlation
function of the frequency fluctuations (FFCF), reflecting
SAMs’ ultrafast dynamics, decays slower, compared with that
of free molecules in solution.55,56 For a single-oscillator
transition, the Tw-dependence of the nodal line slope (NLS)
of the 2D peak serves as a measure of the spectral diffusion.57
In the cases of coupled vibrational modes (free GSH) and
excitonic transitions (GSH SAM), equilibration of the
excitation between the coupled modes can scramble the NLS
signature of the FFCF decay. However, even though the NLSs
in our data cannot be strictly associated with the FFCF, it can
still serve as a parameter qualitatively describing the very
different spectral evolution observed in the free GSH and GSH
SAM.58,59
The evolution of the NLS for free and SAM GSH evaluated
at the lower amide-I transition frequency is compared in Figure
3. The NLS of the free GSH signal decays exponentially (τ =
Figure 3. Spectral diffusion of GSH. Circles, experimental values of
the nodal line slope; lines, their fit to the data. Black, free GSH
molecules, pD 5.5; red, SAM GSH; orange, SAM GSH with BaCl2
added to the sample, pD 3.5; blue, SAM GSH with KOD added to the
sample, pD 12.
0.35 ± 0.1 ps) to an offset amplitude representing static
inhomogeneity of the transition. Note that this decay is
significantly faster than the population transfer between the
coupled amide-I modes leading to growth of the cross-peaks in
Figure 2.40 On the other hand, the NLS of SAM GSH appears
as a straight line on the experimental time scale (limited by the
vibrational relaxation of the amide-I mode). Although its decay
is so slow that the time constant cannot be estimated, it is clear
that the static inhomogeneity is smaller for the SAM GSH.60
The lower initial value of the NLS of SAM GSH, compared to
that of free GSH, indicates a larger relative contribution of the
homogeneous component to the total line width. However, the
static inhomogeneity is smaller, the overall width is reduced,
and the slowing of the Tw-dependent dynamics of the SAM
GSH transition is evident. These observations suggest a highly
ordered molecular structure of the SAM GSH.
2483
Letter
In the pD 5.5 solutions of free GSH, the signal associated
with the protonated fraction of the carboxylic groups is not
observed. Regarding SAM GSH, the corresponding transition
appears 6 times weaker than that of the amide-I band in the
linear spectrum (Figure S3) and 40 times weaker in 2DIR. The
presence of this transition provides interesting information on
the GSH SAM structure. Because the pKa value of Gly carboxyl
is higher than that of Glu, it is more prone to protonation, for
example, by the Glu amine of the neighboring strand, available
for hydrogen bonding in the β-sheet arrangement. At the same
time, in contrast to free GSH, no transition associated with the
anionic carboxyl groups is seen for GSH SAM. This can be
rationalized by the sensitivity of the SAM conformation to the
silver surface charge. At high pD values, the surface of the silver
NPs61 is negatively charged by the adsorbed hydroxyl anions,
whereas at low pD values the surface charge is positive, and the
point-of-zero-charge is at pD 7, as estimated by Merga et al.62
Thus, at pD 5.5 the carboxyl anions are prone to interact with
surface cations. This interaction leads to elimination of the
carboxyl stretching transition from the spectral region
monitored in our experiment either because of the strong
deformation of the electronic potential, leading to a dramatic
shift of the transition frequency, or because of the orientation
of the transition dipole parallel to the surface and manifestation
of the surface selection rules.63
In order to test the sensitivity of the β-sheet conformation to
the solvation environment, we added to the GSH-capped NP
solution either metal ions (BaCl2) or alkali base (KOD).
Barium ions compete with the positively charged NP surface,
because they are chelated by the carboxylic anions,28 leading to
a conformational change in GSH SAM and disruption of β-
sheets.64 The corresponding spectra in Figure 4A,B show
Figure 4. Infrared spectroscopy of disrupted GSH SAM. Linear
spectroscopy (A) and 2DIR (B) of GSH SAM on the NP surface,
when BaCl2 was added into the solution. Linear spectroscopy (C) and
2DIR of GSH SAM on the NP surface, when KOD was added into
the solution.
transitions at ωex = 1665 and 1645 cm−1, similar to those of
free GSH at low pD values. Analysis of the NLS of the amide-I
transition (Figure 3) shows that FFCF decays at a rate similar
to that of free GSH (τ = 0.24 ± 0.12 ps); however, the
contribution of the homogeneous component to the line shape
is smaller and the inhomogeneity increases. Subsequently,
deuteryl anions disrupt β-sheets by passivating the NPs, which
leads to the carboxyl groups repulsing from the surface. The
DOI: 10.1021/acs.jpclett.9b00530
J. Phys. Chem. Lett. 2019, 10, 2481−2486
The Journal of Physical Chemistry Letters
corresponding spectra in Figure 4C,D resemble those observed
for solutions of free GSH at high pD values, with emphasized
transition at ωex = 1595 cm−1. Analysis of the NLS (Figure 3)
indicates that also here static inhomogeneity exceeds that of
free GSH.
Recently, Mandal et al.65 observed that leucine-rich peptide,
fully helical in solution, assembles into an antiparallel β-sheet
on the surface of 5 nm NPs (ωex = 1625 cm−1) but not on the
surface of larger 20 nm NPs and on flat surfaces. In contrast,
Shaw et al.54 observed that the fraction of the parallel β-sheet
formed by the amyloid-derived peptide (ωex = 1665 and 1640
cm−1) increased with NP size. Our own results for GSH SAM
on the surface of 2 nm silver NPs66 indicate that β-sheets are
not formed and that the corresponding spectra (Figures S3 and
S5) qualitatively resemble those of free GSH at pD 5.5,
illustrating the critical role of surface flatness in promoting the
formation of GSH β-sheets. Di Gregorio et al.67 noted that
chiroptical signal of the GSH-capped silver nanocubes vanishes
upon raising the pH. Indeed, in our samples we observed that
the β-sheet’s circular dichroism, a positive transition at 203 nm,
disappears upon addition of the barium or deuteryl ions
(Figure S6), confirming disruption of β-sheets.
In 2DIR spectroscopy, the parallel and antiparallel β-sheets
are distinguished by the cross-peaks appearing in the latter
case.45,46,52,68 On metal NPs, the electromagnetic boundary
conditions require that direction of the electric field is normal
to the surface69,70 and surface selection rules are implied.63
The low-frequency amide-I exciton mode of the β-sheet has
transition dipole moment perpendicular to the strand axis.71
For the parallel β-sheet, the transition dipole is inclined to the
plane of the sheet, such that if the sheet lies parallel to the
metal surface, its out-of-plane component would survive the
selection rules. On the other hand, for the antiparallel β-sheet,
the corresponding transition dipole moment is in-plane with
the sheet71 and would not be observed. The weaker high-
frequency exciton mode of the antiparallel β-sheets has
transition dipole moment parallel to the strand, in-plane with
the sheet.71 Thus, any signal from the antiparallel β-sheet
whose plane is parallel to the NP surface would vanish.
Interestingly, we have not detected high-frequency transition
associated with the antiparallel beta-sheet nor the correspond-
ing cross-peaks in the polarization-selective measurements of
SAM GSH (Figure S4), suggesting that if any antiparallel β-
sheet SAM is formed, its plane is parallel to the NP surface and
it is not observed in our measurements. Therefore, the signal
reported in the present work represents the parallel β-sheet
SAM. Interestingly, it is in contrast to conclusions of NMR
studies of the oxidized disulfide GSH dimer, self-assembled
into gels, where antiparallel β-sheets are stabilized by the
restricted rotation about the disulfide bond.17
In conclusion, using polarization-selective 2DIR spectrosco-
py, we revealed molecular conformations of GSH on the
surface of silver NPs. The thiol−silver bonds and the
interaction of the carboxylic anions with silver surface cations
facilitate formation of the β-sheet-like and fibril-like structures
enclosing the NP. Remarkably, the fibril-like structure is
formed by a small peptide with only two amino acids. A weak,
yet visible signal of the protonated carboxyl is also seen, when
β-sheets are formed. The conformation of the β-sheet appears
predominantly parallel; however, we cannot rule out the
presence of the antiparallel β-sheets, whose plane is parallel to
the NP surface. The structural response of β-sheets to a change
in the environment is important not only for the chemical and
2484
Letter
nanotechnological applications of the GSH-capped NPs but
also for NP-based medicine, because GSH is abundant in the
cytosol and can potentially exchange with various NP ligands.
Internalized NPs with GSH SAM eventually reach the acidic
lysosome, where the fibril-like structure on their surface may
initiate various processes.72
■
*
ASSOCIATED CONTENT
S Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jp-
clett.9b00530.
Experimental methods and additional experimental
results (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: chunt@technion.ac.il.
ORCID
Lev Chuntonov: 0000-0002-2316-4708
Author Contributions
†
A.G. and A.K.P. contributed equally to this work.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
This research is supported by the Israel Science Foundation
(grant 1118/15) and the Russell Berrie Nanotechnology
Institute, Technion.
■
REFERENCES
(1) Sapsford, K. E.; Algar, W. R.; Berti, L.; Gemmill, K. B.; Casey, B.
J.; Oh, E.; Stewart, M. H.; Medintz, I. L. Functionalizing
Nanoparticles with Biological Molecules: Developing Chemistries
That Facilitate Nanotechnology. Chem. Rev. 2013, 113, 1904−2074.
(2) Nel, A. E.; Mädler, L.; Velegol, D.; Xia, T.; Hoek, E. M. V.;
Somasundaran, P.; Klaessig, F.; Castranova, V.; Thompson, M.
Understanding Biophysicochemical Interactions at the Nano−Bio
Interface. Nat. Mater. 2009, 8, 543.
(3) Vallee, A.; Humblot, V.; Pradier, C.-M. Peptide Interactions with
Metal and Oxide Surfaces. Acc. Chem. Res. 2010, 43, 1297−1306.
(4) Slocik, J. M.; Naik, R. R. Probing Peptide−Nanomaterial
Interactions. Chem. Soc. Rev. 2010, 39, 3454−3463.
(5) Haran, G.; Chuntonov, L. Artificial Plasmonic Molecules and
Their Interaction with Real Molecules. Chem. Rev. 2018, 118, 5539−
5580.
(6) Howes, P. D.; Chandrawati, R.; Stevens, M. M. Colloidal
Nanoparticles as Advanced Biological Sensors. Science 2014, 346,
1247390.
(7) Mikolajczak, D. J.; Koksch, B. Peptide-Gold Nanoparticle
Conjugates as Sequential Cascade Catalysts. ChemCatChem 2018, 10,
4324−4328.
(8) Han, G.; Ghosh, P.; Rotello, V. M. Functionalized Gold
Nanoparticles for Drug Delivery. Nanomedicine 2007, 2, 113−123.
(9) Dykman, L. A.; Khlebtsov, N. G. Uptake of Engineered Gold
Nanoparticles into Mammalian Cells. Chem. Rev. 2014, 114, 1258−
1288.
(10) Gladytz, A.; Abel, B.; Risselada, H. J. Gold-Induced Fibril
Growth: The Mechanism of Surface-Facilitated Amyloid Aggregation.
Angew. Chem., Int. Ed. 2016, 55, 11242−11246.
(11) Gao, G.; Zhang, M.; Gong, D.; Chen, R.; Hu, X.; Sun, T. The
Size-Effect of Gold Nanoparticles and Nanoclusters in the Inhibition
of Amyloid-Β Fibrillation. Nanoscale 2017, 9, 4107−4113.
DOI: 10.1021/acs.jpclett.9b00530
J. Phys. Chem. Lett. 2019, 10, 2481−2486
The Journal of Physical Chemistry Letters
(12) Pengo, P.; Ş ologan, M.; Pasquato, L.; Guida, F.; Pacor, S.;
Tossi, A.; Stellacci, F.; Marson, D.; Boccardo, S.; Pricl, S.; Posocco, P.
Gold Nanoparticles with Patterned Surface Monolayers for Nano-
medicine: Current Perspectives. Eur. Biophys. J. 2017, 46, 749−771.
(13) Hong, R.; Han, G.; Fernández, J. M.; Kim, B.-j.; Forbes, N. S.;
Rotello, V. M. Glutathione-Mediated Delivery and Release Using
Monolayer Protected Nanoparticle Carriers. J. Am. Chem. Soc. 2006,
128, 1078−1079.
(14) Lund, T.; Callaghan, M. F.; Williams, P.; Turmaine, M.;
Bachmann, C.; Rademacher, T.; Roitt, I. M.; Bayford, R. The
Influence of Ligand Organization on the Rate of Uptake of Gold
Nanoparticles by Colorectal Cancer Cells. Biomaterials 2011, 32,
9776−9784.
(15) Amarnath, K.; Mathew, N. L.; Nellore, J.; Siddarth, C. R. V.;
Kumar, J. Facile Synthesis of Biocompatible Gold Nanoparticles from
Vites Vinefera and Its Cellular Internalization against Hbl-100 Cells.
Cancer Nanotechnol. 2011, 2, 121−132.
(16) Sousa, A. A.; Morgan, J. T.; Brown, P. H.; Adams, A.;
Jayasekara, M. S.; Zhang, G.; Ackerson, C. J.; Kruhlak, M. J.;
Leapman, R. D. Synthesis, Characterization, and Direct Intracellular
Imaging of Ultrasmall and Uniform Glutathione-Coated Gold
Nanoparticles. Small 2012, 8, 2277−2286.
(17) Lyon, R. P.; Atkins, W. M. Self-Assembly and Gelation of
Oxidized Glutathione in Organic Solvents. J. Am. Chem. Soc. 2001,
123, 4408−4413.
(18) Matsuura, K.; Matsuyama, H.; Fukuda, T.; Teramoto, T.;
Watanabe, K.; Murasato, K.; Kimizuka, N. Spontaneous Self-Assembly
of Nanospheres from Trigonal Conjugate of Glutathione in Water.
Soft Matter 2009, 5, 2463−2470.
(19) Lim, I.-I. S.; Mott, D.; Ip, W.; Njoki, P. N.; Pan, Y.; Zhou, S.;
Zhong, C.-J. Interparticle Interactions in Glutathione Mediated
Assembly of Gold Nanoparticles. Langmuir 2008, 24, 8857−8863.
(20) Moaseri, E.; Bollinger, J. A.; Changalvaie, B.; Johnson, L.;
Schroer, J.; Johnston, K. P.; Truskett, T. M. Reversible Self-Assembly
of Glutathione-Coated Gold Nanoparticle Clusters via pH-Tunable
Interactions. Langmuir 2017, 33, 12244−12253.
(21) Mosquera, J.; Henriksen-Lacey, M.; García, I.; Martínez-Calvo,
M.; Rodríguez, J.; Mascareñas, J. L.; Liz-Marzán, L. M. Cellular
Uptake of Gold Nanoparticles Triggered by Host−Guest Interactions.
J. Am. Chem. Soc. 2018, 140, 4469−4472.
(22) Eibling, M. J.; MacDermaid, C. M.; Qian, Z.; Lanci, C. J.; Park,
S.-J.; Saven, J. G. Controlling Association and Separation of Gold
Nanoparticles with Computationally Designed Zinc-Coordinating
Proteins. J. Am. Chem. Soc. 2017, 139, 17811−17823.
(23) Ho, J.-J.; Ghosh, A.; Zhang, T. O.; Zanni, M. T. Heterogeneous
Amyloid Β-Sheet Polymorphs Identified on Hydrogen Bond
Promoting Surfaces Using 2D SFG Spectroscopy. J. Phys. Chem. A
2018, 122, 1270−1282.
(24) Kozlowski, R.; Ragupathi, A.; Dyer, R. B. Characterizing the
Surface Coverage of Protein−Gold Nanoparticle Bioconjugates.
Bioconjugate Chem. 2018, 29, 2691−2700.
(25) Colangelo, E.; Comenge, J.; Paramelle, D.; Volk, M.; Chen, Q.;
Lévy, R. Characterizing Self-Assembled Monolayers on Gold
Nanoparticles. Bioconjugate Chem. 2017, 28, 11−22.
(26) Ong, Q.; Luo, Z.; Stellacci, F. Characterization of Ligand Shell
for Mixed-Ligand Coated Gold Nanoparticles. Acc. Chem. Res. 2017,
50, 1911−1919.
(27) Smith, A. M.; Johnston, K. A.; Crawford, S. E.; Marbella, L. E.;
Millstone, J. E. Ligand Density Quantification on Colloidal Inorganic
Nanoparticles. Analyst 2017, 142, 11−29.
(28) Hepel, M.; Tewksbury, E. Ion-Gating Phenomena of Self-
Assembling Glutathione Films on Gold Piezoelectrodes. J. Electroanal.
Chem. 2003, 552, 291−305.
(29) Bieri, M.; Bürgi, T. L-Glutathione Chemisorption on Gold and
Acid/Base Induced Structural Changes: A PM-IRRAS and Time-
Resolved in Situ ATR-IR Spectroscopic Study. Langmuir 2005, 21,
1354−1363.
2485
Letter
(30) Bieri, M.; Bürgi, T. Adsorption Kinetics of L-Glutathione on
Gold and Structural Changes During Self-Assembly: An in Situ ATR-
IR and QCM Study. Phys. Chem. Chem. Phys. 2006, 8, 513−520.
(31) Lobo Maza, F.; Méndez De Leo, L.; Rubert, A. A.; Carro, P.;
Salvarezza, R. C.; Vericat, C. New Insight into the Interface Chemistry
and Stability of Glutathione Self-Assembled Monolayers on Au(111).
J. Phys. Chem. C 2016, 120, 14597−14607.
(32) Vallée, A.; Humblot, V.; Méthivier, C.; Pradier, C.-M.
Glutathione Adsorption from UHV to the Liquid Phase at Various
pH on Gold and Subsequent Modification of Protein Interaction. Surf.
Interface Anal. 2008, 40, 395−399.
(33) Vallée, A.; Humblot, V.; Méthivier, C.; Pradier, C.-M.
Adsorption of a Tripeptide, GSH, on Au (1 1 1) under UHV
Conditions; Pm-Rairs and Low T-XPS Characterisation. Surf. Sci.
2008, 602, 2256−2263.
(34) Donaldson, P. M.; Hamm, P. Gold Nanoparticle Capping
Layers: Structure, Dynamics, and Surface Enhancement Measured
Using 2D-IR Spectroscopy. Angew. Chem. 2013, 125, 662−666.
(35) Li, J.; Qian, H.; Chen, H.; Zhao, Z.; Yuan, K.; Chen, G.;
Miranda, A.; Guo, X.; Chen, Y.; Zheng, N.; Wong, M. S.; Zheng, J.
Two Distinctive Energy Migration Pathways of Monolayer Molecules
on Metal Nanoparticle Surfaces. Nat. Commun. 2016, 7, 10749.
(36) Bian, H.; Li, J.; Chen, H.; Yuan, K.; Wen, X.; Li, Y.; Sun, Z.;
Zheng, J. Molecular Conformations and Dynamics on Surfaces of
Gold Nanoparticles Probed with Multiple-Mode Multiple-Dimen-
sional Infrared Spectroscopy. J. Phys. Chem. C 2012, 116, 7913−7924.
(37) Petti, M. K.; Lomont, J. P.; Maj, M.; Zanni, M. T. Two-
Dimensional Spectroscopy Is Being Used to Address Core Scientific
Questions in Biology and Materials Science. J. Phys. Chem. B 2018,
122, 1771−1780.
(38) Hamm, P.; Lim, M.; DeGrado, W. F.; Hochstrasser, R. M. The
Two-Dimensional IR Nonlinear Spectroscopy of a Cyclic Penta-
Peptide in Relation to Its Three-Dimensional Structure. Proc. Natl.
Acad. Sci. U. S. A. 1999, 96, 2036−2041.
(39) Woutersen, S.; Hamm, P. Structure Determination of
Trialanine in Water Using Polarization Sensitive Two-Dimensional
Vibrational Spectroscopy. J. Phys. Chem. B 2000, 104, 11316−11320.
(40) Woutersen, S.; Mu, Y.; Stock, G.; Hamm, P. Subpicosecond
Conformational Dynamics of Small Peptides Probed by Two-
Dimensional Vibrational Spectroscopy. Proc. Natl. Acad. Sci. U. S. A.
2001, 98, 11254−11258.
(41) Baiz, C. R.; Reppert, M.; Tokmakoff, A. Amide I Two-
Dimensional Infrared Spectroscopy: Methods for Visualizing the
Vibrational Structure of Large Proteins. J. Phys. Chem. A 2013, 117,
5955−5961.
(42) Ganim, Z.; Chung, H. S.; Smith, A. W.; DeFlores, L. P.; Jones,
K. C.; Tokmakoff, A. Amide I Two-Dimensional Infrared Spectros-
copy of Proteins. Acc. Chem. Res. 2008, 41, 432−441.
(43) Hamm, P.; Lim, M.; Hochstrasser, R. M. Structure of the
Amide I Band of Peptides Measured by Femtosecond Nonlinear-
Infrared Spectroscopy. J. Phys. Chem. B 1998, 102, 6123−6138.
(44) Wang, L.; Middleton, C. T.; Singh, S.; Reddy, A. S.; Woys, A.
M.; Strasfeld, D. B.; Marek, P.; Raleigh, D. P.; de Pablo, J. J.; Zanni,
M. T.; Skinner, J. L. 2DIR Spectroscopy of Human Amylin Fibrils
Reflects Stable Β-Sheet Structure. J. Am. Chem. Soc. 2011, 133,
16062−16071.
(45) Hahn, S.; Kim, S.-S.; Lee, C.; Cho, M. Characteristic Two-
Dimensional IR Spectroscopic Features of Antiparallel and Parallel Β-
Sheet Polypeptides: Simulation Studies. J. Chem. Phys. 2005, 123,
084905.
(46) Moran, S. D.; Zanni, M. T. How to Get Insight into Amyloid
Structure and Formation from Infrared Spectroscopy. J. Phys. Chem.
Lett. 2014, 5, 1984−1993.
(47) Lomont, J. P.; Rich, K. L.; Maj, M.; Ho, J.-J.; Ostrander, J. S.;
Zanni, M. T. Spectroscopic Signature for Stable Β-Amyloid Fibrils
Versus Β-Sheet-Rich Oligomers. J. Phys. Chem. B 2018, 122, 144−
153.
DOI: 10.1021/acs.jpclett.9b00530
J. Phys. Chem. Lett. 2019, 10, 2481−2486
The Journal of Physical Chemistry Letters Letter

(48) Evanoff, D. D.; Chumanov, G. Size-Controlled Synthesis of Dimensional Infrared Spectroscopy and Atomic Force Microscopy.
Nanoparticles. 1. “Silver-Only” Aqueous Suspensions Via Hydrogen Sci. Rep. 2017, 7, 41051.
Reduction. J. Phys. Chem. B 2004, 108, 13948−13956. (69) Chuntonov, L.; Haran, G. Maximal Raman Optical Activity in
(49) Qian, W.; Krimm, S. Vibrational Analysis of Glutathione. Hybrid Single Molecule-Plasmonic Nanostructures with Multiple
Biopolymers 1994, 34, 1377−1394. Dipolar Resonances. Nano Lett. 2013, 13, 1285−1290.
(50) Rubtsov, I. V. Relaxation-Assisted Two-Dimensional Infrared (70) Cohn, B.; Engelman, B.; Goldner, A.; Chuntonov, L. Two-
(RA 2DIR) Method: Accessing Distances over 10 Å and Measuring Dimensional Infrared Spectroscopy with Local Plasmonic Fields of a
Bond Connectivity Patterns. Acc. Chem. Res. 2009, 42, 1385−1394. Trimer Gap-Antenna Array. J. Phys. Chem. Lett. 2018, 9, 4596−4601.
(51) Müller-Werkmeister, H. M.; Li, Y.-L.; Lerch, E.-B. W.; Bigourd, (71) Marsh, D. Dichroic Ratios in Polarized Fourier Transform
D.; Bredenbeck, J. Ultrafast Hopping from Band to Band: Assigning Infrared for Nonaxial Symmetry of Beta-Sheet Structures. Biophys. J.
Infrared Spectra Based on Vibrational Energy Transfer. Angew. Chem., 1997, 72, 2710−2718.
Int. Ed. 2013, 52, 6214−6217. (72) Stern, S. T.; Adiseshaiah, P. P.; Crist, R. M. Autophagy and
(52) Cheatum, C. M.; Tokmakoff, A.; Knoester, J. Signatures of Β- Lysosomal Dysfunction as Emerging Mechanisms of Nanomaterial
Sheet Secondary Structures in Linear and Two-Dimensional Infrared Toxicity. Part. Fibre Toxicol. 2012, 9, 20.
Spectroscopy. J. Chem. Phys. 2004, 120, 8201−8215.
(53) Abramavicius, D.; Zhuang, W.; Mukamel, S. Peptide Secondary
Structure Determination by Three-Pulse Coherent Vibrational
Spectroscopies: A Simulation Study. J. Phys. Chem. B 2004, 108,
18034−18045.
(54) Shaw, C. P.; Middleton, D. A.; Volk, M.; Lévy, R. Amyloid-
Derived Peptide Forms Self-Assembled Monolayers on Gold
Nanoparticle with a Curvature-Dependent Β-Sheet Structure. ACS
Nano 2012, 6, 1416−1426.
(55) Kraack, J. P.; Hamm, P. Surface-Sensitive and Surface-Specific
Ultrafast Two-Dimensional Vibrational Spectroscopy. Chem. Rev.
2017, 117, 10623−10664.
(56) Rosenfeld, D. E.; Gengeliczki, Z.; Smith, B. J.; Stack, T.; Fayer,
M. Structural Dynamics of a Catalytic Monolayer Probed by Ultrafast
2D IR Vibrational Echoes. Science 2011, 334, 634−639.
(57) Kwac, K.; Cho, M. Molecular Dynamics Simulation Study of N-
Methylacetamide in Water. II. Two-Dimensional Infrared Pump−
Probe Spectra. J. Chem. Phys. 2003, 119, 2256−2263.
(58) Kuroda, D. G.; Vorobyev, D. Y.; Hochstrasser, R. M. Ultrafast
Relaxation and 2D IR of the Aqueous Trifluorocarboxylate Ion. J.
Chem. Phys. 2010, 132, 044501.
(59) King, J. T.; Baiz, C. R.; Kubarych, K. J. Solvent-Dependent
Spectral Diffusion in a Hydrogen Bonded “Vibrational Aggregate. J.
Phys. Chem. A 2010, 114, 10590−10604.
(60) Kwak, K.; Park, S.; Finkelstein, I. J.; Fayer, M. D. Frequency-
Frequency Correlation Functions and Apodization in Two-Dimen-
sional Infrared Vibrational Echo Spectroscopy: A New Approach. J.
Chem. Phys. 2007, 127, 124503.
(61) Adamczyk, K.; Simpson, N.; Greetham, G. M.; Gumiero, A.;
Walsh, M. A.; Towrie, M.; Parker, A. W.; Hunt, N. T. Ultrafast
Infrared Spectroscopy Reveals Water-Mediated Coherent Dynamics
in an Enzyme Active Site. Chemical Science 2015, 6, 505−516.
(62) Merga, G.; Cass, L. C.; Chipman, D. M.; Meisel, D. Probing
Silver Nanoparticles During Catalytic H2 Evolution. J. Am. Chem. Soc.
2008, 130, 7067−7076.
(63) Greenler, R. G.; Snider, D. R.; Witt, D.; Sorbello, R. S. The
Metal-Surface Selection Rule for Infrared Spectra of Molecules
Adsorbed on Small Metal Particles. Surf. Sci. 1982, 118, 415−428.
(64) Edington, S. C.; Gonzalez, A.; Middendorf, T. R.; Halling, D.
B.; Aldrich, R. W.; Baiz, C. R. Coordination to Lanthanide Ions
Distorts Binding Site Conformation in Calmodulin. Proc. Natl. Acad.
Sci. U. S. A. 2018, 115, E3126−E3134.
(65) Mandal, H. S.; Kraatz, H.-B. Effect of the Surface Curvature on
the Secondary Structure of Peptides Adsorbed on Nanoparticles. J.
Am. Chem. Soc. 2007, 129, 6356−6357.
(66) Farrag, M.; Thämer, M.; Tschurl, M.; Bürgi, T.; Heiz, U.
Preparation and Spectroscopic Properties of Monolayer-Protected
Silver Nanoclusters. J. Phys. Chem. C 2012, 116, 8034−8043.
(67) di Gregorio, M. C.; Ben Moshe, A.; Tirosh, E.; Galantini, L.;
Markovich, G. Chiroptical Study of Plasmon−Molecule Interaction:
The Case of Interaction of Glutathione with Silver Nanocubes. J. Phys.
Chem. C 2015, 119, 17111−17116.
(68) Roeters, S. J.; Iyer, A.; Pletikapić, G.; Kogan, V.; Subramaniam,
V.; Woutersen, S. Evidence for Intramolecular Antiparallel Beta-Sheet
Structure in Alpha-Synuclein Fibrils from a Combination of Two-

2486 DOI: 10.1021/acs.jpclett.9b00530

J. Phys. Chem. Lett. 2019, 10, 2481−2486
Quantifying conformations of ester vibrational probes with hydrogen-bond-induced
Fermi resonances
Anup Ghosh, Bar Cohn, Amit K. Prasad, and Lev Chuntonov

Citation: J. Chem. Phys. 149, 184501 (2018); doi: 10.1063/1.5055041

View online: https://doi.org/10.1063/1.5055041
View Table of Contents: http://aip.scitation.org/toc/jcp/149/18
Published by the American Institute of Physics

Articles you may be interested in

Perspective: Computational chemistry software and its advancement as illustrated through three grand
challenge cases for molecular science
The Journal of Chemical Physics 149, 180901 (2018); 10.1063/1.5052551

Low scaling EOM-CCSD and EOM-MBPT(2) method with natural transition orbitals
The Journal of Chemical Physics 149, 184103 (2018); 10.1063/1.5045340

Perspective: Dynamics of confined liquids

The Journal of Chemical Physics 149, 170901 (2018); 10.1063/1.5057759

The quantum mechanics-based polarizable force field for water simulations

The Journal of Chemical Physics 149, 174502 (2018); 10.1063/1.5042658

Nanoscale domains in ionic liquids: A statistical mechanics definition for molecular dynamics studies
The Journal of Chemical Physics 149, 184502 (2018); 10.1063/1.5054999

The activation energy for water reorientation differs between IR pump-probe and NMR measurements
The Journal of Chemical Physics 149, 164504 (2018); 10.1063/1.5050203
 THE JOURNAL OF CHEMICAL PHYSICS 149, 184501 (2018)

Quantifying conformations of ester vibrational probes

with hydrogen-bond-induced Fermi resonances
Anup Ghosh,a) Bar Cohn,a) Amit K. Prasad, and Lev Chuntonovb)
Schulich Faculty of Chemistry and Solid State Institute, Technion—Israel Institute of Technology,
Haifa 3200003, Israel
(Received 6 September 2018; accepted 22 October 2018; published online 8 November 2018)

Solvatochromic shifts of local vibrational probes report on the strength of the surrounding electric
fields and the probe’s hydrogen bonding status. Stretching vibrational mode of the ester carbonyl group
is a popular solvatochromic reporter used in the studies of peptides and proteins. Small molecules,
used to calibrate the response of the vibrational probes, sometimes involve Fermi resonances (FRs)
induced by inter-molecular interactions. In the present work, we focus on the scenario where FR does
not appear in the infrared spectrum of the ester carbonyl stretching mode in aprotic solvents; however,
it is intensified when a hydrogen bond with the reporter is established. When two molecules form
hydrogen bonds to the same carbonyl oxygen atom, FR leads to strong hybridization of the involved
modes and splitting of the absorption peak. Spectral overlap between the Fermi doublets associated
with singly and doubly hydrogen-bonded carbonyl groups significantly complicates quantifying dif-
ferent hydrogen-bonded conformations. We employed a combination of linear and third-order (2DIR)
infrared spectroscopy with chemometrics analysis to reveal the individual line shapes and to estimate
the occupations of the hydrogen-bonded conformations in methyl acetate, a model small molecule.
We identified a hydrogen-bond-induced FR in complexes of methyl acetate with alcohols and water
and found that FR is lifted in larger molecules used for control experiments—cholesteryl stearate
and methyl cyanoacetate. Applying this methodology to analyze acetonitrile-water solutions revealed
that when dissolved in neat water, methyl acetate occupies a single hydrogen-bonding conformation,
which is in contrast to the conclusions of previous studies. Our approach can be generally used when
FRs prevent direct quantification of the hydrogen bonding status of the vibrational probe. Published
by AIP Publishing. https://doi.org/10.1063/1.5055041

I. INTRODUCTION in the dry interior of packed peptides.7,11 The frequency shifts,

induced by the local electric field, are typically calibrated
Identification and quantification of hydrogen (H)-bonded
with small molecules, which serve as model systems that
conformations using vibrational reporters are extensively used
predict the response of the transition frequency to the field
when studying the molecular structure and its environment.1–4
exerted by the probe’s environment, when it is incorporated
Experimental methods used in these studies include vibrational
into the biomolecule.12–14 Selection of solvents is often used
solvatochromic spectroscopy,3 where one monitors shifts in
to span the range of the local field strengths, which for the
the transition frequency induced by the local electric field
C==O group is as broad as 0-50 MV/cm. A higher local
around the vibrational probe (vibrational Stark effect) as well
field strength induces a larger red-shift of the C==O transi-
as by specific solute-solvent interactions5,6 and ultrafast spec-
tion. Interestingly, vibrational frequency shifts of carbonyls
troscopy of solvation dynamics, where processes such as spec-
were found to correlate to observables in other spectroscopic
tral diffusion and chemical exchange are monitored.4 Carbonyl
techniques.14,15
ester moieties can serve as particularly useful chromophores
Even higher frequency shifts, compared to those induced
in such studies because they are highly sensitive to the local
by the local fields, are observed in C==O transitions upon the
field strength and feature a significant shift in the transition fre-
formation of H-bonds with the carbonyl oxygen.7,12,14,16–20
quency of the carbonyl stretching vibration.7–9 It was recently
Interestingly, in protic solvents, ester groups can attain sev-
shown that carbonyl esters can be site-specifically incorpo-
eral different H-bonded conformations, including bonding to
rated into the peptides and proteins using chemical biology
either a single hydrogen-donating solvent molecule (1hb) or to
methods.10
a pair of these solvent molecules (2hb). The two distinct states,
Recently, Gai and co-workers used solvatochromic mea-
1hb and 2hb, can be distinguished spectroscopically since the
surements in ester carbonyl stretching vibrational mode
C==O stretching frequency shifts to a different extent in each of
(C==O), combined with the Onsager model and molecular
these cases. In the extensively studied model molecule, methyl
dynamics simulations to estimate the local dielectric functions
acetate (MA),21–23 which is among the smallest molecules
involving the carbonyl ester abundant in natural fats and
a)A. Ghosh and B. Cohn contributed equally to this work. lipids,9,24 there is a strong peak in the C==O frequency region
b)Email: chunt@technion.ac.il at 1748.5 cm−1 in the aprotic non-polar carbon tetrachloride

0021-9606/2018/149(18)/184501/11/$30.00 149, 184501-1 Published by AIP Publishing.

184501-2 Ghosh et al.
solvent corresponding to the 0hb conformation.25 In methanol,
this transition does not shift appreciably; however, two addi-
tional peaks appear at 1729.9 cm−1 and 1705.0 cm−1 .16 In
water (D2 O), only two peaks are observed at 1726.0 cm−1 and
1704.2 cm−1 .20
Based on the extensive experimental studies, comple-
mented by theoretical analysis, Banno et al.16 assigned these
lower-frequency transitions to the 1hb and 2hb conforma-
tions of MA, which can coexist simultaneously in protic
solvents, especially in alcohols, and can interconvert on the
picosecond time scale. Later, this assignment was supported
by the studies of Righini and co-workers.19,20 However, despite
the apparent agreement between the experimentally observed
shifts of C==O transitions in different H-bonded conforma-
tions and their interpretation, based on the quantum chemistry
calculations,19,20 these studies suggest that two very differ-
ent solvation states, 1hb and 2hb, can be simultaneously
populated by a small and highly hydrophilic molecule such
as MA in the bulk of aqueous solution; this might appear
counterintuitive.
MA can be readily compared with another extensively
studied small molecule, N-Methylacetamide (NMA),26–28
which is the smallest molecule involving the amide group, a
building block of peptides and proteins.29 It is well known that
in diluted aqueous solutions, NMA has a single broad transi-
tion of its amide I vibrational mode that reflects a nearly con-
tinuous distribution of multiple H-bonded configurations.30–32
The asymmetry of the C==O line shape of NMA in water
was initially attributed to Fermi resonances (FRs).33 Two-
dimensional vibrational spectroscopy (2DIR) is well known
for its ability to reveal FRs because of its high sensitivity
to the inter-mode coupling.34–37 However, there is no clear
indication of the presence of FRs in the 2DIR spectrum of
NMA.27,31 The ultrafast chemical exchange between different
H-bonded conformations was first demonstrated using 2DIR
spectroscopy in NMA–methanol complexes.26 Stimulated by
the opportunity provided by this experimental advance, in
recent 2DIR experiments and molecular dynamics simulation
studies, researchers focused on elucidating the rates of inter-
conversion between different H-bonded conformations of MA
in aqueous and methanol solutions.16–20
Because of an apparent similarity in the molecular struc-
tures of NMA and MA and of their relatively high solubility in
water,38 the existence of the 1hb and 2hb conformers of MA
in the diluted aqueous solution requires additional experimen-
tal verification. In the present work, we address this aspect
via a detailed study of the H-bonded conformations between
the ester C==O group of MA and several H-bond-donating
molecules, including a series of substituted phenols, trifluo-
roethanol, and water (D2 O). We conducted a series of infrared
absorption and 2DIR experiments and found that two low-
frequency peaks observed in the infrared spectrum of MA in
the region of the C==O transition involve FRs, associated with
both 1hb and 2hb conformations; this effect is particularly
strong in the latter case.
It is well known that accidental degeneracies between the
infrared-active vibrational mode and a dark overtone or a com-
bination band of a different mode can lead to FR. Such occur-
rences are extensively discussed in the literature, mainly in the
J. Chem. Phys. 149, 184501 (2018)
context of the spectroscopy of small molecules.34,36 Recent
examples include the work of Kwon et al.,37 who observed
that molecule–ion interactions can significantly enhance the
strength of FR. Rodgers et al.39 noted that the presence of FR
could limit the ability of the vibrational probe to report on
the H-bonded conformation. Earlier studies by Wassermann
et al.40 and Greve et al.35 on the N−−H stretching modes in
pyrazole and aniline molecules, respectively, have shown that
the H-bond can dramatically enhance the coupling strength
between states participating in FR. In situations where differ-
ent H-bonded conformations possess overlapping FR doublets
that are not observed in the free (non-bonded) state, the iden-
tification and quantification of the H-bonded configurations
using vibrational probes become especially challenging.
In the present work, we revealed the presence of FR in
the ester C==O transition of MA using infrared absorption
measurements in solutions with increasing concentrations of
the H-bond-donating molecules. The paper is organized as
follows: First, the corresponding data for para-chlorophenol
(p-CP) as the H-bond donor in the carbon tetrachloride sol-
vent are presented. This substituted phenol is the most acidic
homologue from the series of homologues that we investi-
gated, including para-tert-butylphenol, para-cresol, phenol,
and para-cyanophenol.41 Because its H-bond formation con-
stant is the highest among the homologues, p-CP is the most
appropriate choice for the scope of the present discussion. A
series of the infrared absorption spectra are analyzed by sin-
gular value decomposition (SVD) to identify the number of
possible H-bonded conformations present in solution, as well
as by self-similar modeling42 and multivariate curve resolu-
tion (MCR) analysis43–45 to reveal their individual line shapes
and the corresponding contribution to the signal in each sam-
ple of the series. In order to support the assignment of the
lower-frequency C==O peaks to the FRs, we conducted 2DIR
spectroscopy, where the coupled vibrational modes feature the
corresponding cross-peaks.
In order to exclude the possibility that the observed FR
arises from the interaction with vibrational modes specific
to phenols, we conducted control experiments with a differ-
ent H-bond-donating molecule, trifluoroethanol (TFE). The
individual infrared spectra obtained for the 1hb and 2hb con-
formations allowed us to extract an unperturbed shift of the
C==O transition in MA upon H-bond formation, invoking the
model of Refs. 39 and 46. We compared the obtained shifts
to those of C==O in the cholesteryl stearate (CS), which is a
convenient choice of molecule for the additional control exper-
iments: In CS, which is a larger molecule than MA, the FRs
are lifted; CS is well soluble in carbon tetrachloride, which is a
convenient solvent for studies of the corresponding complexes
with alcohols.
Finally, we apply the combination of FTIR, MCR, and
2DIR methods to analyze hydrogen-bonded conformations of
MA in acetonitrile-water mixtures with gradually changing
water volume fractions ranging from 0 to 1.47–49 Interestingly,
we found that in neat water, MA occupies only a single H-
bonding conformation, where carbonyl oxygen of the ester
group in MA is bound to two hydroxyl groups of water (2hb).
Similar results are obtained in a series of the control exper-
iments with the homologue molecule methyl cyanoacetate
184501-3 Ghosh et al.
(MCA), where FR is lifted. Overall, our results suggest an effi-
cient approach to elucidate the details of the micro-solvation
environment of the vibrational probes despite the spectral
complexity introduced by the presence of FRs.
II. EXPERIMENTAL METHODS
The linear infrared absorption spectra were recorded with
an FTIR spectrometer (Bruker, Tensor 27). The 0.1M solutions
of molecules involving the ester group, MA (in carbon tetra-
chloride, TFE, and acetonitrile-D2 O mixtures), CS (in carbon
tetrachloride), and MCA (in acetonitrile-D2 O mixtures), were
used throughout the study. The concentration of the H-bond-
donating molecules was systematically varied, as described
below. The sample solutions were placed between the 2 mm-
thick CaF2 windows separated by the Teflon spacer. 2DIR data
were acquired with the home-built spectrometer (Solstice Ace,
Spectra Physics; Topas, Light Conversion) using four ca. 70 fs
laser pulses at 4 kHz, which were focused on the sample in
the BOXCARS configuration. All-parallel pulse polarization
was ensured by wire-grid polarizers. Spectral interferometry
was performed with the wobbling Brewster window50 on a 64-
element array detector (Infrared Associates, Infrared System
Development). The instrument response of 150 fs was mea-
sured with the help of the third-order non-resonant signal from
carbon tetrachloride.
III. RESULTS AND DISCUSSION
A. H-bonded complexes of methyl acetate
with para-chlorophenol
1. Infrared absorption
Several series of the infrared absorption spectra of 0.1M
MA solutions in carbon tetrachloride with p-CP concentrations
ranging from 0 to 1.5M were collected. The representative
spectra are shown in Fig. 1(a). In the first series, the p-CP con-
centration was increased in steps of 2 mM up to a concentration
of 40 mM. It is assumed that for these low concentrations, the
majority of the MA molecules will be in the 0hb and 1hb
conformations. Indeed, the SVD analysis conducted on these
data indicated the presence of only two significant components
despite the clear observation of three peaks in the absorption
spectrum, as seen, for example, in the red and blue lines in
Fig. 1(a) showing spectra for 20 mM and 40 mM of p-CP,
respectively. The singular values of the first three components
obtained were 0.88, 0.11, and 0.003 such that the third compo-
nent can be neglected. The second series of absorption spectra
included p-CP concentrations rising in steps of 20 mM up to
a concentration of 0.5M. The SVD analysis of these data indi-
cated the presence of three significant components, with the
first four singular values of 0.74, 0.22, 0.033, and 0.005.
Since the SVD analysis provides an orthogonal set of sin-
gular vectors, this method does not allow one to obtain the true
(positive) line shapes associated with the individual H-bonded
conformations. The latter task can be performed (to the limit of
the experimental resolution) with the help of methods based on
the self-modeling curve resolution.45 We have performed such
an analysis using procedures described in Refs. 42 and 43 as
well as using an open-source toolbox MCR-ALS 2.0 developed
J. Chem. Phys. 149, 184501 (2018)
FIG. 1. Infrared absorption of MA–p-CP complexes. (a) Representative spec-
tra of 0.1M solutions of MA in carbon tetrachloride with p-CP concentrations
of 0M, 0.02M, 0.04M, 0.08M, 0.14M, 0.24M, 0.5M, and 1.5M. (b) Individ-
ual line shapes obtained from the MCR analysis associated with 0hb (black),
1hb (blue), and 2hb (red) conformations of MA. (c) Occupation of different
H-bonded conformations for various concentrations of p-CP. The color code
is the same as in (b). (d) The relative fraction of the H-bonded MA plotted for
various concentrations of free p-CP hydroxyl groups. The solid line shows a
fit to the binding isotherm (see the text for details).
by Tauler and co-workers.44 First, the low p-CP concentration
spectra, where the SVD shows the presence of only two singu-
lar components, were analyzed. Because the line shape of the
non-H-bonded MA is generally known, matching between one
of the decomposed line shapes to that of the 0hb molecules was
used as a criterion in the evaluation of the MCR results. The
obtained line shapes for the 0hb and 1hb molecules were then
used as the initial guess in the MCR analysis of the high p-CP
concentration spectra. The final MCR results were checked
for consistency such that the line shapes obtained from the
low and high concentration data were identical. The variance
in the data explained by the MCR is R2 = 0.9999.
The obtained spectral profiles of the individual compo-
nents presented in Fig. 1(b) show that both low-frequency
components have double-peak line shapes, which typically
appear in the cases of FRs. The corresponding transition fre-
quencies of each Fermi-doublet (ω1 and ω2 ) are summarized
in Table I (see below). In Fig. 1(c), the relative fractions
of these components, as obtained by the MCR, are plot-
ted for increasing concentrations of the p-CP molecules in
solution.
The data presented in Fig. 1 suggest the following picture:
upon the formation of the 1hb complex, the energy of the first
excited state of the C==O stretching mode is lowered, which
reduces the energy gap and increases the coupling strength
with the dark overtone of the low-frequency mode. The result-
ing FR gives rise to the asymmetric doublet, shown as a blue
line in Fig. 1(b). Further increases in the p-CP concentration
lead to the formation of 2hb complexes, where the energy of
the C==O mode is lowered further. Here, the resonant inter-
action with the dark mode becomes even stronger, and mode
hybridization leads to a nearly fully symmetric doublet shown
as a red line in Fig. 1(b).
184501-4 Ghosh et al. J. Chem. Phys. 149, 184501 (2018)

TABLE I. Analysis of FR in MA.

ω 1 (cm 1 )a ω 2 (cm 1 )a Rb W (cm 1 )c ∆0 (cm 1 )d ω10 (cm 1 )e ω20 (cm 1 )e

Methyl acetate–p-CP

0hb 1748.5
1hb 1727.5 1709.9 3.1 7.6 9.0 1723.2 1714.2
2hb 1723.8 1705.7 1.5 8.9 3.6 1716.6 1712.9

Methyl acetate–TFE

0hb 1748.2
1hb 1730.9 1705.0 5.0 9.7 17.3 1726.6 1709.3
2hb 1725.9 1709.0 1.1 8.4 0.8 1717.9 1717.0

Cholesteryl stearate–p-CP

0hb 1732.5
1hb 1709.8
2hb 1696.1

Methyl acetate–acetonitrile-D2 O

0hb 1741.9
1hb 1729.9 1707.1 5.0 8.5 15.2 1726.1 1710.9
2hb 1726.0 1704.2 0.79 10.8 2.6 1716.4 1713.8

Methyl cyanoacetate–acetonitrile-D2 O

0hb 1755.7
1hb 1751.6
2hb 1742.1
a Transition frequency, as obtained from fitting the individual line shapes of the H-bonded molecules.
b Ratio between the areas of the transitions associated with the FR in H-bonded molecules as obtained from fitting the individual
line shapes.
c The strength of coupling between the modes involved in the FR, as obtained from fitting to the model of Ref. 46.
d Frequency difference between the transitions associated with the unperturbed modes involved in the FR, as obtained from fitting

to the model of Ref. 46.

e Unperturbed transition frequencies associated with modes involved in the FR, as obtained from fitting to the model of Ref. 46.

The contribution of the individual spectral component
(i.e., of the line shape of the particular H-bonded conforma-
tion) to the experimental spectrum obtained by MCR analysis
involves a product of its concentration in the sample and the
excitation probability, which scales with µ2 , where µ is the
transition dipole moment of the C==O stretching mode of the
H-bonded complex. Transition dipole moments are expected to
differ with the H-bonded conformation.20 The concentrations
of the H-bonded complexes in our samples were obtained by
the following method: First, the extinction coefficient of the
0hb MA solution in carbon tetrachloride (without p-CP) was
measured using Beer’s law. Second, the concentration of the
0hb molecules and, consequently, that of the 1hb molecules
were determined in a series of samples with low concentrations
of p-CP, where only two singular components were present.
Once the concentrations of the 1hb species are known, the
160 ± 15 km/mol, which is in a good agreement with the
known values.52,53 The relations between the transition dipole
moments of the H-bonded conformers are in agreement with
that of Candelaresi et al.’s theoretical analysis of the electronic
structure of the H-bonded MA.20
In order to check the results of Fig. 1(b) for consistency,
we fitted the data to the binding isotherm. As evident from
Fig. 1(c), with up to ca. 0.1M concentration of p-CP, only
the 1hb complex with MA is formed. The corresponding data
were fitted to a 1:1 binding isotherm Y11 = 1+K K1 X
, where Y
1X
is the fraction of the H-bonded MA, X is the concentration
of the free p-CP molecules in solution, and K 1 is the equilib-
rium constant of the 1hb complex formation. At higher p-CP
concentrations, the 2hb complex is also formed and the cor-
responding data are fitted to a 1:2 binding isotherm given by
K1 X+K1 K2 X 2

corresponding extinction coefficient is obtained and used to Y12 = 1+K 2 , where K 2 is the stepwise formation equi-
1 X+K1 K2 X

determine the concentrations and the extinction coefficient
of the 2hb species in samples with higher concentrations of
p-CP. Our results suggest the following relations between the
transition dipole moments: µ0hb /µ1hb = 0.85 and µ0hb /µ2hb
= 0.75, where the value of µ0hb = 0.3 D was estimated
following Ref. 51. The integrated intensity associated with
the 0hb transition of MA measured in our experiments is
librium constant of the 2hb complex.54 The values of K 1 = 17.1
l/mol and K 2 = 4.3 l/mol were obtained, which correspond to a
non-cooperative binding process with α = 4K 2 /K 1 ≈ 1, where
α is the interaction parameter,55,56 suggesting that a stacking
interaction between the phenol rings does not affect the H-
bonding equilibrium. The corresponding results are shown in
Fig. 1(d).
184501-5 Ghosh et al.
2. 2DIR spectroscopy
The FR character of the double-peak line shapes obtained
in infrared absorption measurements is further supported by
2DIR spectroscopy. In 2DIR, the coupled vibrational modes
can be identified via the associated cross-peaks.57 In this
case, the waiting time dependence of the cross-peak ampli-
tudes reflects the transfer of the vibrational excitation between
the coupled modes.58 However, chemical exchanges between
two different conformations,59 e.g., 0hb and 1hb, or 1hb and
2hb, will also manifest in the appearance of the cross-peaks,
where the growth of their amplitudes reflects the rate of the
exchange process. The rate of the chemical exchange can
potentially be very high in a highly fluctuating solvation shell
(e.g., in aqueous solutions) such that the time interval it takes
for molecules to switch their configuration can be shorter
than the response time of the ultrafast spectrometer. In such
cases, the cross-peaks associated with the chemical exchange
will appear in the 2DIR spectrum even at very early waiting
times.20
In Fig. 2, the waiting time series of the 2DIR spectra for
MA samples with various concentrations of p-CP are shown.
In the case of the sample with 0.08M p-CP, we expect, based
on the results of Fig. 1, that two H-bonded conformations,
0hb and 1hb, are present in solution with the corresponding
population fractions of 0.6 and 0.4, respectively. Two dom-
inant peaks are observed in the 2DIR spectrum: the peak at
the excitation frequency of 1748 cm−1 , which corresponds
to the 0hb conformation of MA, and the one at 1727 cm−1 ,
which corresponds to the high-frequency peak of the FR dou-
blet associated with the 1hb conformation. The ratio between
the amplitudes of the high-frequency and the low-frequency
peak amplitudes of the doublet of the 1hb component is
ca. 3.5, as obtained from the spectrum shown in Fig. 1(b).
It is therefore expected that the corresponding peak ratio in
the 2DIR spectrum will be ca. 12, which explains why, in the
case of the 0.08M sample, only two strong peaks are observed,
whereas the low-frequency transition of the 1hb conforma-
tion is rather weak. Nevertheless, the cross-peak amplitudes
between the 1727 cm−1 and the 1707 cm−1 transitions are
FIG. 2. 2DIR spectroscopy of MA–p-CP complexes. First row: the p-CP con-
centration is 0.08M, second row—0.5M, and third row—1.5M. Left column:
the waiting time is 0.3 ps, middle column—1.5 ps, and right column—3 ps.
All-parallel polarization of the excitation pulses was used.
J. Chem. Phys. 149, 184501 (2018)
clearly visible in the 2DIR spectrum collected at a waiting
time of 0.3 ps, as seen in Fig. 2 (first row, left column). Impor-
tantly, no cross-peaks appear between the 0hb and 1hb signals
up to the waiting times of 3 ps, indicating that a chemical
exchange between these conformations is not observed on the
time scale discussed in the present study, which is in agreement
with the relatively large value of the equilibrium constants (see
above).18
In the sample with a 0.5M concentration of p-CP, about
half of the MA molecules are expected to be in the 2hb
conformation. Here, two strong peaks are observed in the
2DIR spectrum at the excitation frequencies of 1727 cm−1
and 1706 cm−1 . The former peak includes signals from both
the 1hb and 2hb conformations of MA, whereas the latter rep-
resents predominantly the 2hb molecules. Strong cross-peaks
that appear at the waiting time as early as 0.3 ps indicate that
the two quantum states observed in the spectrum are coupled.
Note again that the 2DIR spectrum collected at a waiting time
of 3 ps indicates no exchange between the 0hb and 1hb con-
formations. This suggests that the waiting time dependence of
the cross-peaks’ amplitudes observed between the peaks with
excitation frequencies of 1727 cm−1 and 1706 cm−1 reflects
the relaxation of the vibrational excitation between the modes
involved in the FR of the 2hb complex. When the concentra-
tion of the p-CP molecules is raised to 1.5M, most of the MA
molecules are expected to be in the 2hb conformation. The
corresponding 2DIR data, shown in Fig. 2, are fully consistent
with this expectation.
B. H-bonded complexes of MA with trifluoroethanol
In order to determine whether the FR observed in the
H-bonded complexes of MA and p-CP is specific to the
p-CP molecule or, for example, is typical of the aryl moiety,
we conducted similar experiments with trifluoroethanol (TFE)
as the H-bond-donating molecule. The corresponding results
of linear infrared absorption experiments are summarized in
Fig. 3(a). In Fig. 3(b), the individual line shapes of the 0hb,
1hb, and 2hb components of MA are shown as obtained from
the MCR analysis (explained data variance R2 = 0.9998). The
line shapes of both the 1hb and 2hb conformations involve two
peaks; the ratio of their amplitudes shows a stronger hybridiza-
tion of the vibrational modes in the 2hb case. The occupation
of the H-bonded conformations by the MA molecules is plot-
ted in Fig. 3(c) for various concentrations of TFE. Fitting the
obtained data to the binding isotherm results in values of the
stepwise formation equilibrium constants of K 1 = 3.3 l/mol
and K 2 = 1.9 l/mol such that α ≈ 2.30. The equilibrium rate
of the formation of the MA–TFE complex is generally lower
than that of the MA–p-CP complex. At the TFE concentration
of 1M, we found that the 0hb, 1hb, and 2hb conformations
of the MA molecules are populated with approximately equal
probability.
The 2DIR data in Fig. 4 show that, similar to the previous
case, no exchange between the 0hb and 1hb conformations in
the MA–TFE complex can be observed for the time interval
of several picoseconds. For the sample with a 1M concentra-
tion of TFE, the cross-peaks between the two low-frequency
transitions associated with the FR of the 1hb conformers are
weak yet clearly visible. When the concentration of TFE is
184501-6 Ghosh et al.
FIG. 3. H-bonded complexes of MA with TFE. (a) Infrared absorption of
0.1M solutions of MA in carbon tetrachloride with 0M, 0.1M, 0.2M, 0.3M,
0.6M, 0.9M, 1M, and 13.2M (neat) of TFE. (b) Individual line shapes obtained
from the MCR analysis associated with 0hb (black), 1hb (blue), and 2hb (red)
conformations of MA. (c) Occupation of the different HB conformations for
various concentrations of TFE. The color code is the same as in (b). (d) The
relative fraction of the H-bonded MA plotted for various concentrations of the
free TFE hydroxyl groups. The solid line shows the fit to the binding isotherm
(see the text for details).
raised to ca. 13M, corresponding to the neat TFE liquid, the
cross-peaks associated with the FR of the 2hb conformation of
MA become prominent. The corresponding inter-mode energy
relaxation leads to the growth of the cross-peak amplitudes,
compared to the diagonal peaks for the later waiting times of
the experiment, as seen in Fig. 4. In addition, the signal of the
small fraction of the 0hb MA molecules observed at a waiting
time of 3 ps is seen in Fig. 4(d). Note again the absence of
the cross-peaks associated with the chemical exchange even
at this relatively late waiting time.
Overall, the results shown in Figs. 3 and 4 are in good
agreement with those, where p-CP was chosen as a H-bond-
donating molecule, supporting the assumption that the FR in
MA is not strongly affected by the identity of the H-bond-
donating molecule; however, it is the property of the MA
molecule itself.
FIG. 4. 2DIR spectroscopy of the H-bonded complexes of MA with TFE. (a)
The concentration of TFE is 1M; the waiting time is 0.3 ps. (b) The same as
in (a), the waiting time is 3 ps. (c) The concentration of TFE is 13.2M (neat);
the waiting time is 0.3 ps. (d) The same as in (c), the waiting time is 3 ps.
J. Chem. Phys. 149, 184501 (2018)
C. H-bonded complexes of cholesteryl stearate
with para-chlorophenol
As we have shown above, the FRs involving the C==O
stretching mode in the H-bond complexes of MA are not
very sensitive to the characteristics of the H-bond-donating
molecule and appear to be an intrinsic property of the MA
molecule. Accidental degeneracies leading to FRs can be
removed by isotope substitutions. Unfortunately, the isotope-
modified MA is not available commercially. To overcome this
limitation, we conducted measurements of the linear absorp-
tion and 2DIR in a different molecule involving the C==O
ester moiety–cholesteryl stearate (CS). We used the same
experimental strategy as in the experiments described above:
linear absorption for a series of solutions with gradually
increasing concentrations of p-CP was measured, and indi-
vidual line shapes of the 0hb, 1hb, and 2hb conformations
of CS were obtained by MCR analysis (explained data vari-
ance R2 = 0.9999). The corresponding results are summarized
in Fig. 5. Fitting the data to the binding isotherm gives val-
ues of the stepwise formation equilibrium constants that are
similar to the MA–p-CP complexes: K 1 = 20.7 l/mol and
K 2 = 6.6 l/mol such that α ≈ 1.3, showing essentially a
non-cooperative binding.
In CS, the C==O mode of the 0hb conformation is red-
shifted, compared to MA, and the corresponding transition
appears at 1732.5 cm−1 . Interestingly, as opposed to the case of
MA, for CS, the line shapes of both the 1hb and 2hb conform-
ers appear as single peaks at 1709.8 cm−1 and 1696.1 cm−1
and do not show the presence of FR. Consistent with the
results of Fig. 5, the 2DIR spectra of the H-bonded CS–p-CP
molecules lack the corresponding cross-peaks. In Fig. 6, the
2DIR spectra are shown for 0.1M CS solutions with 0.1M and
FIG. 5. H-bonded complexes of CS with p-CP. (a) Infrared absorption of
0.1M solutions of CS in carbon tetrachloride with 0M, 0.05M, 0.1M, 0.15M,
0.2M, 0.25M, and 0.5M of p-CP. (b) Individual line shapes obtained from
the MCR analysis associated with 0hb (black), 1hb (blue), and 2hb (red)
conformations of CS. (c) Occupation of the different HB conformations for
various concentrations of p-CP. The color code is the same as in (b). (d) The
relative fraction of the H-bonded CS plotted for various concentrations of the
free p-CP hydroxyl groups. The solid line shows a fit to the binding isotherm
(see the text for details).
184501-7 Ghosh et al.
FIG. 6. 2DIR spectroscopy of H-bonded complexes of CS with p-CP. (a) The
concentration of p-CP is 0.1M; the waiting time is 0.3 ps. (b) The same as
in (a), the waiting time is 3 ps. (c) The concentration of p-CP is 0.5M; the
waiting time is 0.3 ps. (d) The same as in (c), the waiting time is 3 ps.
0.5M concentrations of p-CP. In the former sample, the pop-
ulation of the H-bonded CS molecules is dominated by the
1hb conformation, whereas in the latter, the 2hb conforma-
tion dominates, as can be seen in Fig. 5(c). The 2DIR spectra
were collected for both early (0.3 ps) and later (3 ps) waiting
times of the experiment to check for the possible appearance
of the cross-peak following the relaxation of vibrational exci-
tation. Clearly, 2DIR line shapes of both the 1hb and 2hb
conformations do not indicate the presence of the coupled
modes.
D. H-bonded complexes of MA and MCA with water
After the methodology to quantify H-bonded confor-
mations of MA was established as described above, we
applied it to quantify H-bonded conformations of MA dis-
solved in water. To this end, we analyzed solutions of MA
in acetonitrile-water mixtures with gradually changing water
volume fractions ranging from 0 to 1.47–49 It is important to
note here that the MCR analysis is valid only when the line
shapes of the individual components remain invariant to the
microscopic heterogeneity of the acetonitrile-water mixture,
whose effects on the spectroscopic observables were recently
J. Chem. Phys. 149, 184501 (2018)
discussed by Mukherjee et al.49 The corresponding results of
linear spectroscopy are shown in Figs. 7(a)–7(c). Similar to
the previous cases, the SVD revealed the presence of three
major components (singular values of 0.65, 0.25, and 0.075)
that are associated with the 0-2hb conformations. The presence
of only three singular components in our dataset suggests that
this assumption holds such that the obtained vibrational line
shapes are either not strongly affected by this heterogeneity
or represent observables effectively averaged over the hetero-
geneity.48 The data variance explained by the MCR results is
R2 = 0.9993.
From the MCR analysis, we obtained, upon the increase
in water content and, thus, the availability of the H-bond-
donating hydroxyl groups, that 1hb conformations are formed
before the 2hb ones; however, the latter dominates when
approaching the solvent composition of bulk water. Both indi-
vidual line shapes, associated with the 1hb and 2hb con-
formations of MA, involve transitions at ca. 1730 cm−1
and 1704 cm−1 that stem from FR induced by forming H-
bonds with water molecules, as shown in Fig. 7(b). Impor-
tantly, the relative fractions of the H-bonded conformers in
Fig. 7(c) indicate that in neat water, MA occupies only a single
experimentally distinguishable H-bonded conformation, the
2hb one.
More support for the existence of FR was obtained
with 2DIR spectroscopy,57 where the coupling between the
involved modes is manifested as cross-peaks observed already
at the earliest waiting times of the experiment.34–37 This is
in contrast to the case of the chemical exchange between
different H-bonded conformations, where the cross-peaks typ-
ically do not appear at early waiting times.26,59–61 Indeed,
the corresponding cross-peaks are observed in the 2DIR
spectrum of MA in pure water collected at T = 0.3 ps, as
shown in Fig. 8(a). In cases where the chemical exchange
is faster than the instrument response, the cross-peaks are
expected to also appear at early waiting times; their ampli-
tudes typically rise with a time constant of ca. 0.5 ps or faster
and decay with vibrational lifetime.20 In the present experi-
ment, the waiting time dependence of the {ωτ = 1730 cm−1 ;
ωt = 1709 cm−1 } cross-peak in Fig. 8(c) has a rise time
FIG. 7. H-bonded complexes of MA
and MCA with water. [(a) and (d)]
Infrared absorption of MA and MCA
in D2 O-acetonitrile mixtures. The water
fraction is indicated. [(b) and (e)] Indi-
vidual line shapes of H-bonded con-
formers of MA and MCA. [(c) and (f)]
Relative fraction of the H-bonded con-
formers in different mixtures. The col-
ors match those of line shapes in panels
(b) and (d).
184501-8 Ghosh et al.
FIG. 8. 2DIR spectroscopy of H-bonded complexes of MA and MCA with water. [(a) and (b)] 2DIR spectra of MA and [(d) and (e)] that of MCA in pure D2 O.
Waiting times are indicated. (c) The waiting time dependence of the 2DIR peak amplitudes of MA. The lines show the bi-exponential and the rise-and-decay fits.
Diagonal peaks: green—{1730; 1743}, τ 1 = 0.3 ps (0.7), τ 2 = 1.7 ps (0.3); black—{1705; 1710}, τ 1 = 0.2 ps (0.6), τ 2 = 1.4 ps (0.4). Cross-peaks: blue—{1707;
1743}, τ 1 = 0.2 ps (0.6), τ 2 = 1.8 ps (0.4); red—{1730; 1709}, τ rise = 2.4 ps, τ decay = 1.8 ps. (f) The waiting time evolution of the nodal line slope in 2DIR
spectra of MCA. The red line shows fitting to an exponential decay, τ = 0.86 ps.
of 2.4 ps and a decay time of 1.8 ps, whereas the ampli-
tude of the second cross-peak {ωτ = 1707 cm−1 ; ωt = 1743
cm−1 } does not rise, but decays bi-exponentially, with a fast
time constant of 0.2 ps and a similar slow time constant of
1.7 ps. In addition, we noted that the amplitudes of both diag-
onal peaks decay with the fast time constants of 0.2-0.3 ps
and slow time constants of 1.4-1.7 ps. Bi-exponential decay
of the C==O stretching in MA was previously observed in
both aprotic and protic solvents, with the fast component
attributed to the equilibration of the vibrational excitation
between the coupled modes, whereas the slower one to popula-
tion relaxation.18,25,41 The evolution of the cross-peak ampli-
tudes, which is different from what is expected in the case
of chemical exchange, supports the assignment of the origin
of the observed cross-peaks in FR. Note also the qualitative
similarity between the 2DIR spectrum in Figs. 8(a) and 8(b)
and the 2DIR spectrum of FR in benzoylchloride, discussed in
Ref. 34.
In order to alter the resonance conditions between the
modes involved in FRs also under conditions of aqueous solu-
tions, we focused on a substituted homologue of MA–methyl
cyanoacetate (MCA). The corresponding infrared absorption
spectra are shown in Fig. 7(d). Analogous to the case of MA,
we found that MCA has three distinct H-bonded conform-
ers in acetonitrile-water mixtures, which can be distinguished
experimentally and assigned to 0-2hb C==O groups (MCR
R2 = 0.9999). However, as shown in Fig. 7(e), each con-
former has a single C==O transition with the corresponding
frequencies of 1755.7 cm−1 , 1751.6 cm−1 , and 1742.1 cm−1
for the 0hb, 1hb, and 2hb complexes, respectively. In MCA,
the C==O transition frequencies are shifted to the blue, as
compared to those of MA and CS, and the corresponding
spectra indicate that the FR is removed by the substitution.
Analysis of the component fractions revealed behavior sim-
ilar to MA, indicating that only a single 2hb conformation
J. Chem. Phys. 149, 184501 (2018)
is observed in neat water, as seen in Fig. 7(f). The 2DIR
spectrum of MCA in water [see Figs. 8(d) and 8(e)] also
has a single peak that undergoes spectral diffusion with a
characteristic time of 0.86 ps, as obtained from the analy-
sis of its nodal line slope in Fig. 8(f). The latter reflects the
loss of correlation between the fluctuating vibrational frequen-
cies,62,63 which in aqueous solutions is broadly attributed to
the reconfiguration of H-bonds.64–66
E. Analysis of Fermi resonance in MA
The interaction between the ester C==O modes and vari-
ous combination bands in aromatic acetate molecules is well
known;67,68 however, the aliphatic acetates were generally
considered not to exhibit such an effect.69,70 Although the
occurrence of FR in the C==O stretching mode of H-bonded
MA molecules has not been discussed before, it is not very
surprising since small molecules are generally well known to
be highly susceptible to this phenomenon. Careful inspection
of the 0hb spectrum of MA reveals the presence of a very weak
peak at 1710 cm−1 , which can be possibly assigned to the over-
tone of the skeleton deformation mode at 850 cm−1 , which can
couple to the C==O stretching.21
In order to obtain more details on the coupling between
the modes, we used the modeling of Refs. 39 and 46, where
the coupling to a mode with zero transition dipole moment
is assumed. The individual line shapes of the 1hb and 2hb
conformers possessing FRs from Figs. 1, 3, and 7 were fit-
ted to pairs of Lorentzian profiles with central frequencies
ω1 and ω2 and peak areas I 1 and I 2 . Based on the parame-
ters obtained from fits, we calculated the frequency difference
between the coupled modes, ∆ = ω1 − ω2 , as well as the ratios
of the peak intensities R = I 1 /I 2 , as summarized in Table I.
The strength of the inter-mode coupling W and the energy
difference between the unperturbed modes ∆0 were extracted
184501-9 Ghosh et al.
based on the expression derived in Ref. 46: R = ∆−∆ ∆+∆0
0
, where
∆ = ∆0 + 4W .
2 2 2 stretching transition is free of FR. In the case of aqueous solu-
In addition, using the parameters obtained with the model
of FR, we were able to estimate the transition frequencies that
would be associated with the unperturbed modes ω10 and ω20 .
Interestingly, for the case of the MA–p-CP complex, our results
show very close excitation frequencies of the dark modes of
1714.2 cm−1 and 1712.9 cm−1 for the 1hb and 2hb confor-
mations, respectively. By contrast, the frequency of the bright
C==O stretching mode shifts to the red by 6.6 cm−1 upon the
formation of the 2hb complex, compared to the 1hb one, bring-
ing the modes into a near-resonant condition. In the case of the
MA–TFE complex, our analysis predicts a different behavior:
the unperturbed overtone frequency is highly sensitive to the
H-bonded conformation. In the 1hb complex, its frequency is
1709.3 cm−1 and the hybridization with the C==O mode is
weak. However, in the case of the 2hb complex, in addition
to the ca. 9 cm−1 red-shift of the C==O transition, the over-
tone blue shifts by ca. 8 cm−1 , which leads to on-resonance
coupling with the C==O mode. A qualitatively similar effect
was observed for the MA-D2 O complex—formation of the
second H-bond leads not only to the red-shift of the C==O
transition but also to the blue shift of the dark mode, although
only by ca. 3 cm−1 , as well as to an increase in their coupling
strength, and, thus, to a stronger degree of mode hybridiza-
tion, which manifests in the two transitions of the nearly equal
amplitudes.
In order to obtain more information on the possible origin
of the mode participating in FR with the C==O stretching in
MA, we collected infrared absorption spectra in the frequency
range of 800-900 cm−1 . As we mentioned earlier, MA has a
skeleton deformation vibrational mode corresponding to the
transition at 845 cm−1 , shown in Fig. 9. Upon the addition of
the H-bond-donating molecules to the MA solution, a new peak
appears at 856.5 cm−1 in both p-CP and TFE solutions. On the
other hand, no peak in this range is observed in the case of the
FIG. 9. Infrared absorption of H-bonded MA and MCA in the low-frequency
range. Blue lines—free MA molecules in carbon tetrachloride solution [panels
(a) and (b)], in acetonitrile [panel (c)], and free MCA molecules [panel (d)].
The concentration of MA and MCA is 0.1M in all samples. Red line—MA
H-bonded molecules to p-PC (concentration 1.5M) in panel (a), TFE (con-
centration 1M) in panel (b), D2 O (neat) in panel (c), and MCA H-bonded
to D2 O (neat) in panel (d). Black lines show absorption in samples with the
H-bond-donating molecules only. The magenta line in panel (b) shows the
spectrum of the CS molecules (concentration 0.1M) lacking any transitions in
this spectral region.
J. Chem. Phys. 149, 184501 (2018)
CS sample [see the magenta line in Fig. 9(b)], where the C==O
tions, the peak associated with the skeleton deformation mode
of MA blue shifts from 847 cm−1 in acetonitrile to 858 cm−1
in water, whereas in the case of the MCA, the corresponding
frequency does not change, as shown in Figs. 9(c) and 9(d).
This experimental observation indicating either a blue shift
of the 845 cm−1 mode or an appearance of a new mode at
ca. 856-858 cm−1 in all the cases where FRs are observed
is in contrast to the simple analysis by the two-state model
discussed above. The model captures a significant blue shift
of the low-frequency mode only for the case of TFE, as
seen in the results summarized in Table I. The experimen-
tal data in Fig. 9 suggest a possible connection between the
low-frequency modes and those involved in the FR of MA.
Interestingly, comparing the frequency shifts of the unper-
turbed C==O transitions in different H-bonded complexes
shows that shifts for the 1hb conformations of the MA–p-CP,
MA–TFE, and CS–p-CP complexes are 25.3 cm−1 , 21.6 cm−1 ,
and 22.7 cm−1 , respectively. For the 2hb conformations, the
shifts are 31.9 cm−1 and 30.3 cm−1 for the MA–p-CP and
MA–TFE complexes, respectively, whereas it is 36.4 cm−1
in the case of the CS–p-CP complex. The unperturbed tran-
sition frequencies of the C==O stretching in MA H-bonded
to water are very similar to those of the MA–TFE complex.
The frequency shift obtained in the case of the 2hb MCA-D2 O
complexes as compared to the 1hb ones of 9.5 cm−1 is very
similar to 9.7 cm−1 obtained in the case of MA. We antici-
pate that quantum chemistry calculations employing a proper
theoretical description of the dispersion forces needed to cap-
ture both H-bonding and possible pi-pi interactions between
the aryl groups, as well as anharmonic couplings between the
vibrational modes of the H-bonded complex, could help to
identify the quantum-mechanical origin of the results observed
in our experiments.71 We noted that our preliminary analysis
with standard quantum chemistry tools indicates an impor-
tant role played by the low-frequency “floppy” vibrational
modes of the H-bonded complex whose reliable analysis of
their anharmonic coupling is challenging.
IV. CONCLUSIONS
An infrared study of H-bonded complexes of MA indi-
cates that the ester C==O stretching vibrational mode is
involved in FR, where the strength of the coupling with the dark
mode strongly depends on the H-bonded configuration. The
formation of H-bonds with the C==O group of MA generally
red shifts the frequency of the C==O transition, can blue shift
the frequency of the dark mode, and enhances the coupling
strength between them; all together, it dramatically increases
the degree of hybridization between the involved modes. This
effect is much stronger for 2hb conformers than for 1hb con-
formers. Following this finding, we have demonstrated that
the two transitions in the region of C==O stretching observed
in aqueous solutions of MA are associated with the H-bond-
induced FR that appears in a single H-bonded conformer. The
emergence of FR induced by intermolecular interactions has
been previously reported in cases of N−−H and C−−N stretch-
ing; it appears to be a general phenomenon. This effect is likely
184501-10 Ghosh et al.
to characterize also other small molecules, which are gener-
ally known to be prone to FR, whereas this is not observed, for
example, in the ester C==O stretching in larger CS and MCA
molecules studied in this work. The identity of the H-bond
donating molecules used in the present study (p-CP, TFE, and
D2 O) was found not to affect the general phenomena of the
emergence of FR in MA. In the view of the present results, the
assignment of the C==O transitions in small ester molecules to
different H-bonded conformations should be taken with cau-
tion. Interestingly, the 2DIR spectra of the MA complex with
cyanophenol reported by us earlier 41 did not show a clear cross-
peak pattern characteristic to the FR, and the cross-peaks’
growth that was attributed to the chemical exchange among
the H-bonded conformations was relatively slow. However,
our present findings suggest that it is necessary to consider the
anharmonic interactions in great detail. Substituted phenols
provide a convenient platform to study the H-bond-induced
FR, where the anharmonic interactions induced by H-bonding
can be systematically varied via the alternation of the substi-
tution group on the phenol ring. We are currently investigating
the role of the strength of the H-bonding interaction on the
anharmonic coupling in the complexes of substituted phenols
with different ester molecules.
In the context of solvatochromic vibrational probes of
(bio-)molecular environment, where the probe’s response is
assessed with the help of small molecules, it is very impor-
tant to be able to recognize FRs. Because of its accidental
nature, FR induced by intermolecular interactions may com-
plicate the development of a general approach for describing
molecular vibrational solvatochromism. Here, the unperturbed
vibrational transitions could be considered in constructing the
frequency-field maps for the vibrational probe. The system-
atic approach presented in this work to identify FRs involves
linear infrared spectroscopy as well as 2DIR and MCR anal-
ysis. We believe that this approach is general and therefore
can be broadly used to quantify the occupations of different
molecular conformations.
ACKNOWLEDGMENTS
This research is supported by Grant No. 1118/15 from the
Israel Science Foundation.
1 J. Ma, I. M. Pazos, W. Zhang, R. M. Culik, and F. Gai, Annu. Rev. Phys.
Chem. 66, 357 (2015).
2 R. Adhikary, J. Zimmermann, and F. E. Romesberg, Chem. Rev. 117, 1927
(2017).
3 S. D. Fried and S. G. Boxer, Acc. Chem. Res. 48, 998 (2015).
4 H. Kim and M. Cho, Chem. Rev. 113, 5817 (2013).
5 S. D. Fried, L.-P. Wang, S. G. Boxer, P. Ren, and V. S. Pande, J. Phys.
Chem. B 117, 16236 (2013).
6 B. Błasiak, C. H. Londergan, L. J. Webb, and M. Cho, Acc. Chem. Res. 50,
968 (2017).
7 I. M. Pazos, A. Ghosh, M. J. Tucker, and F. Gai, Angew. Chem. 126, 6194
(2014).
8 S. H. Schneider and S. G. Boxer, J. Phys. Chem. B 120, 9672 (2016).
9 P. Stevenson and A. Tokmakoff, J. Am. Chem. Soc. 139, 4743 (2017).
10 A. I. Ahmed and F. Gai, Protein Sci. 26, 375 (2017).
11 I. M. Pazos, J. Ma, D. Mukherjee, and F. Gai, Ultrafast hydrogen-bonding
dynamics in amyloid fibrils, J. Phys. Chem. B (to be published).
12 S. D. Fried, S. Bagchi, and S. G. Boxer, J. Am. Chem. Soc. 135, 11181
(2013).
J. Chem. Phys. 149, 184501 (2018)
13 S. D. Fried, S. Bagchi, and S. G. Boxer, Science 346, 1510 (2014).
14 S. M. Kashid and S. Bagchi, J. Phys. Chem. Lett. 5, 3211 (2014).
15 T. Haldar, S. M. Kashid, P. Deb, S. Kesh, and S. Bagchi, J. Phys. Chem.
Lett. 7, 2456 (2016).
16 M. Banno, K. Ohta, and K. Tominaga, J. Raman Spectrosc. 39, 1531
(2008).
17 M. Banno, K. Ohta, S. Yamaguchi, S. Hirai, and K. Tominaga, Acc. Chem.
Res. 42, 1259 (2009).
18 L. Chuntonov, I. M. Pazos, J. Ma, and F. Gai, J. Phys. Chem. B 119, 4512
(2014).
19 M. Pagliai, F. Muniz-Miranda, G. Cardini, R. Righini, and V. Schettino,
J. Phys. Chem. Lett. 1, 2951 (2010).
20 M. Candelaresi, M. Pagliai, M. Lima, and R. Righini, J. Phys. Chem. A 113,
12783 (2009).
21 J. Wilmshurst, J. Mol. Spectrosc. 1, 201 (1957).
22 J. Dybal and S. Krimm, J. Mol. Struct. 189, 383 (1988).
23 M. L. Senent, R. Domı́nguez-Gómez, M. Carvajal, and I. Kleiner, J. Chem.
Phys. 138, 044319 (2013).
24 V. V. Volkov, R. Chelli, W. Zhuang, F. Nuti, Y. Takaoka, A. M. Papini,
S. Mukamel, and R. Righini, Proc. Natl. Acad. Sci. U. S. A. 104, 15323
(2007).
25 M. Lim and R. M. Hochstrasser, J. Chem. Phys. 115, 7629 (2001).
26 S. Woutersen, Y. Mu, G. Stock, and P. Hamm, Chem. Phys. 266, 137
(2001).
27 M. F. DeCamp, L. DeFlores, J. M. McCracken, A. Tokmakoff, K. Kwac,
and M. Cho, J. Phys. Chem. B 109, 11016 (2005).
28 M. T. Zanni, M. C. Asplund, and R. M. Hochstrasser, J. Chem. Phys. 114,
4579 (2001).
29 A. Ghosh and R. M. Hochstrasser, Chem. Phys. 390, 1 (2011).
30 E. Salamatova, A. V. Cunha, R. Bloem, S. J. Roeters, S. Woutersen, T. L.
C. Jansen, and M. S. Pshenichnikov, J. Phys. Chem. A 122, 2468 (2018).
31 S. Woutersen, R. Pfister, P. Hamm, Y. Mu, D. S. Kosov, and G. Stock,
J. Chem. Phys. 117, 6833 (2002).
32 S. Ham, J.-H. Kim, H. Lee, and M. Cho, J. Chem. Phys. 118, 3491 (2003).
33 P. Hamm, M. Lim, and R. M. Hochstrasser, J. Phys. Chem. B 102, 6123
(1998).
34 J. Edler and P. Hamm, J. Chem. Phys. 119, 2709 (2003).
35 C. Greve, E. T. J. Nibbering, and H. Fidder, J. Phys. Chem. B 117, 15843
(2013).
36 M. J. Tucker, Y. S. Kim, and R. M. Hochstrasser, Chem. Phys. Lett. 470,
80 (2009).
37 Y. Kwon, C. Lee, and S. Park, Chem. Phys. 445, 38 (2014).
38 P. Howard and W. Meylan, Handbook of Physical Properties of Organic
Chemicals (Lewis Publishers, Boca Raton, FL, 1997).
39 J. M. Rodgers, R. M. Abaskharon, B. Ding, J. Chen, W. Zhang, and F. Gai,
Phys. Chem. Chem. Phys. 19, 16144 (2017).
40 T. N. Wassermann, C. A. Rice, M. A. Suhm, and D. Luckhaus, J. Chem.
Phys. 127, 234309 (2007).
41 L. Chuntonov, Phys. Chem. Chem. Phys. 18, 13852 (2016).
42 W. H. Lawton and E. A. Sylvestre, Technometrics 13, 617 (1971).
43 O. S. Borgen and B. R. Kowalski, Anal. Chim. Acta 174, 1 (1985).
44 J. Jaumot, A. de Juan, and R. Tauler, Chemom. Intell. Lab. Syst. 140, 1
(2015).
45 J.-H. Jiang, Y. Liang, and Y. Ozaki, Chemom. Intell. Lab. Syst. 71, 1 (2004).
46 J. F. Bertran, L. Ballester, L. Dobrihalova, N. Sánchez, and R. Arrieta,
Spectrochim. Acta, Part A 24, 1765 (1968).
47 J. E. Bertie and Z. Lan, J. Phys. Chem. B 101, 4111 (1997).
48 T. Takamuku, M. Tabata, A. Yamaguchi, J. Nishimoto, M. Kumamoto,
H. Wakita, and T. Yamaguchi, J. Phys. Chem. B 102, 8880 (1998).
49 D. Mukherjee, L. I. Ortiz Rodriguez, M. R. Hilaire, T. Troxler, and F. Gai,
Phys. Chem. Chem. Phys. 20, 2527 (2018).
50 R. Bloem, S. Garrett-Roe, H. Strzalka, P. Hamm, and P. Donaldson, Opt.
Express 18, 027067 (2010).
51 N. J. Turro, Modern Molecular Photochemistry (University Science Books,
1991).
52 R. Heess and H. Kriegsmann, Spectrochim. Acta, Part A 24, 2121 (1968).
53 W. Seth-Paul, Spectrochim. Acta, Part A 30, 1817 (1974).
54 K. A. Connors, Binding Constants: The Measurement of Molecular Complex
Stability (Wiley, New York, 1987), Vol. 33.
55 P. Thordarson, Chem. Soc. Rev. 40, 1305 (2011).
56 D. Brynn Hibbert and P. Thordarson, Chem. Commun. 52, 12792 (2016).
57 P. Hamm and M. T. Zanni, Concepts and Methods of 2D Infrared
Spectroscopy (Cambridge University Press, Cambridge, 2011).
58 L. Chuntonov and J. Ma, J. Phys. Chem. B 117, 13631 (2013).
184501-11 Ghosh et al.
59 Y. S. Kim and R. M. Hochstrasser, Proc. Natl. Acad. Sci. U. S. A. 102,
11185 (2005).
60 J. Zheng, K. Kwak, J. Asbury, X. Chen, I. R. Piletic, and M. D. Fayer,
Science 309, 1338 (2005).
61 K. Kwak, J. Zheng, H. Cang, and M. Fayer, J. Phys. Chem. B 110, 19998
(2006).
62 K. Kwac and M. Cho, J. Chem. Phys. 119, 2256 (2003).
63 K. Kwak, S. Park, I. J. Finkelstein, and M. D. Fayer, J. Chem. Phys. 127,
124503 (2007).
64 J. D. Eaves, J. J. Loparo, C. J. Fecko, S. T. Roberts, A. Tokmakoff, and
P. L. Geissler, Proc. Natl. Acad. Sci. U. S. A. 102, 13019 (2005).
J. Chem. Phys. 149, 184501 (2018)
65 R. Kumar, J. R. Schmidt, and J. L. Skinner, J. Chem. Phys. 126, 204107
(2007).
66 S. Park and M. D. Fayer, Proc. Natl. Acad. Sci. U. S. A. 104, 16731
(2007).
67 R. Nyquist and S. Settineri, Appl. Spectrosc. 44, 1629 (1990).
68 H. Lee and J. Wilmshurst, J. Chem. Soc. 1965, 3590.
69 S. M. Kashid, G. Y. Jin, S. Chakrabarty, Y. S. Kim, and S. Bagchi, J. Phys.
Chem. Lett. 8, 1604 (2017).
70 S. C. Edington, J. C. Flanagan, and C. R. Baiz, J. Phys. Chem. A 120, 3888
(2016).
71 L. Chuntonov and U. Peskin, Chem. Phys. 482, 93 (2017).
 ChemComm
View Article Online
COMMUNICATION View Journal | View Issue
Published on 25 August 2016. Downloaded by Technion - Israel Institute of Technology on 27/02/2017 10:08:09.

Chiral carbon dots derived from guanosine

Cite this: Chem. Commun., 2016,
52, 11159
Received 19th July 2016,
Accepted 8th August 2016
5 0 -monophosphate form supramolecular
hydrogels†
Anup Ghosh,‡ Bibudha Parasar,‡ Tanima Bhattacharyya and Jyotirmayee Dash*

DOI: 10.1039/c6cc05947c

www.rsc.org/chemcomm

Guanosine 5 0 -monophosphate, (5 0 -GMP), is a self-assembling natural

nucleotide that has unique potential to form ordered supramolecular
structures. We herein describe an intriguing property of Na2(50 -GMP)
to form blue emitting chiral carbon dots (G-dots) that exhibit excita-
tion dependent down-conversion and up-conversion fluorescence
signature and self-assemble to form fluorescent hydrogels.

The self-assembly of nucleic acids is a versatile method for

generating functional nanostructures, which include catalytic
hybridization, stimuli elicited assembly and molecular scale
computation.1 The nucleic acid subunit, guanine and its derivatives
can self-assemble to form a wide variety of supramolecular archi-
tectures such as ribbons, columnar and helical G-quartet stacks
(Scheme 1).2,3 In recent years, guanosine self-assembled structures
have also found applications in both nano- and biotechnology.2
The G-quartet is the most common structural motif found in DNA
and RNA quadruplexes, which are widespread in the telomeres and
the promoter region of DNA and in the 50 -untranslated regions of
the RNA.4 The macrocyclic G-quartet consists of a planar arrange-
ment of four guanines, stabilized by Hoogsteen type hydrogen
bonding and mono/divalent cations.3
On the other hand, zero-dimensional carbon based nano-
particles (carbon dots)5–11 of size less than 10 nm have attracted
immense interest over the past decade. They have shown promising
applications in bio-imaging,6 biosensors,7 biomedicine delivery
systems,8 dye sensitized solar cells,9 and organic solar cells,10 due
to their biocompatibility and exciting optical properties. Although a
lot of studies has been done on carbon dots (CDs), the chemical
structure12 and photoluminescence (PL) mechanism5,13 of CDs
still remain unknown. This can be partly attributed to the use
of multiple components in the synthesis of CDs, which makes it
Department of Organic Chemistry, Indian Association for the Cultivation of Science,
Jadavpur, Kolkata-700032, India. E-mail: ocjd@iacs.res.in; Fax: +91-33-2473-2805;
Tel: +91-33-2473-4971, ext. 1405
† Electronic supplementary information (ESI) available: Experimental proce-
dures, characterization data of compounds, and the copies of 1H NMR and 13C
NMR data. See DOI: 10.1039/c6cc05947c
‡ These authors contributed equally.
Scheme 1 Formation of 5 0 -GMP carbon dots (G-dots).
challenging for mechanistic investigations. While some studies
highlight the reason behind PL properties of CDs to be perfect
core carbon crystals with less defects14 and different surface
trap states14b formed by thermal pyrolysis and carbonization,
others propose that the PL mechanism of CDs is governed
solely by organic fluorophores.5 It has been extremely difficult
to address the specific chemical structure of CDs and reason its
excitation dependent PL properties.
Carbon dots can be made from carbohydrates, peptides and
other organic compounds.5 In continuation with our studies on
self-assembly of guanosine derivatives,15 we have used guanosine
5 0 -monophosphate (5 0 -GMP) as the starting material for prepara-
tion of CDs presuming that, (i) the presence of an intrinsic
aromatic system in 5 0 -GMP might require less rearrangements
enabling the system to be less challenging to analyze, (ii) 5 0 -GMP
being a self-assembling natural nucleotide would facilitate self-
assembly by hydrogen bonds,2,3,15 (iii) the phosphate group
of 5 0 -GMP would catalyze CD formation and impart water-
solubility. Using 5 0 -GMP as the component for the formation
of CDs, we have (i) addressed the structure of the 5 0 -GMP
derived CDs and (ii) explored the self-assembly nature of
guanosine to form CD based hydrogels, without addition of
any external agents.

This journal is © The Royal Society of Chemistry 2016 Chem. Commun., 2016, 52, 11159--11162 | 11159

Communication
The disodium salt of 5 0 -GMP was irradiated under micro-
wave conditions at 160 1C for 5 minutes. The resulting solution
exhibited intense fluorescence (Scheme 1) upon exposure to
UV-light thereby confirming the transformation of molecular
5 0 -GMP to 5 0 -GMP nano dots (G-dots). It should be pointed out
that 5 0 -GMP does not exhibit any fluorescence. The condition of
Published on 25 August 2016. Downloaded by Technion - Israel Institute of Technology on 27/02/2017 10:08:09.
irradiation was important because exposure to lower tempera-
ture or lesser time led to incomplete conversion, while exposure
to higher temperatures for prolonged time led to precipitation.
The morphology of the G-dots was confirmed by high resolution-
transmission electron microscopy (HR-TEM) and atomic force
microscopy (AFM) (Fig. 1a and b). HR-TEM images showed
spherical particles with an average diameter less than 5 nm
without any graphitic lattices (Fig. 1a and Fig. S1, ESI†). The
absence of the lattice pattern suggests them to be highly
amorphous. The AFM height profile was in well agreement with
the HR-TEM analysis (Fig. 1b).
UV/Vis spectra and PL studies were carried out to character-
ize the optical properties of G-dots. The UV/Vis spectrum of
5 0 -GMP exhibited two well defined absorption peaks at 219 nm
and 293 nm and a weak shoulder at 345 nm regions, where the
absorption peak at 219 nm and 293 nm are attributed to the
p - p* transition of the guanine ring and the n - p* transition
of the carbonyl group (Fig. 1c). The weak shoulder at 345 nm is
attributed to the weak propensity of the overall p - p*
conjugation of the 5 0 -GMP quartet helix formed in the solution.
Upon exposure to high energy microwave irradiation, the
absorption features at 219 nm and 293 nm diminished along
Fig. 1 HR-TEM (a) and AFM (b) image of G-dots. (c) Absorption spectra
of G-dots, (d) fluorescence spectra of G-dots at different excitation
wavelengths (lex = 310 to 420 nm with increment of 10 nm) (inset-
normalised one). (e) UP-conversion spectra of G-dots (lex = 600 to
850 nm with increment of 25 nm) (inset-normalised one). (f) Fluorescence
decay at different monitoring wavelengths for a particular excitation
wavelength (lex = 375 nm).
11160 | Chem. Commun., 2016, 52, 11159--11162
View Article Online
ChemComm
with intensification of a broad peak centered at 340 nm,
suggesting an increase in electron delocalization efficiency
(Fig. 1c). Like any other CDs, G-dots also exhibited the char-
acteristic excitation wavelength dependent fluorescence emis-
sion. Upon increasing the excitation wavelength from 310 nm
to 420 nm, the photoluminescence peaks of G-dots shifted from
447 nm to 462 nm (Fig. 1d). At an excitation wavelength of
360 nm, the photoluminescence quantum yield (QY) of G-dots
is determined to be 32.5%. In addition to PL spectra, G-dots
exhibited up-conversion luminescence with their maxima in
the region of 430 nm to 462 nm, when excited in the higher
wavelength region (600 nm to 825 nm) (Fig. 1e). Surprisingly,
time resolved photoluminescence spectra (lex = 375 nm) of the
5 0 -GMP dots monitored at different emission decayed differently
for different monitoring wavelengths and the average decay
constant decreased with an increase of monitoring wavelength
(Fig. 1f and Table S1, ESI†), which indicate the variation in
shapes, sizes and band gaps of G-dots. We have observed that
the fluorescence of G-dots was quenched in acidic pH and
remain unaffected at basic pH (Fig. S2, ESI†).
The PXRD patterns of the G-dots (Fig. S3, ESI†) displayed a
broad peak centered at 27.41 (d = 0.33 nm), which could be
related to the G-quadruplex type stacking between two adjacent
vertical G–G stacks. The peaks in the lower Bragg’s angle
regime were nearly absent for the G-dot, thus suggesting a
spherical symmetry of the resulting G-dot. For further surface
structural elaboration, Fourier transform infrared (FTIR) spectros-
copy of the G-dots was performed and compared with that of
5 0 -GMP (Fig. S4, ESI†). The peaks at 1693, 1524 and 1484 cm 1
are attributed to the amide I (CQO), amide II (N–H) and amide
III (C–N) stretching frequency. The broad-bands centered at
3423 and 3313 cm 1 are assigned to the O–H (Ws) and N–H
(Ws) stretching frequency. The sharp peak at 1086 cm 1 corre-
sponds to the PQO stretching frequency and that at 1365 cm 1
corresponds to C–H bending vibration. The phosphate group
renders the colloidal stability of the G-dots, with a measured
zeta potential value of 19.4 mV (Fig. S5, ESI†).
Surprisingly, the G-dots did not show the characteristic
graphitic D band and G band in the RAMAN spectra (Fig. S6,
ESI†), suggesting the presence of a highly disordered (amorphous)
state of carbon in the core corroborating our observation from
HR-TEM. We propose that the presence of a graphitic layer is not
a necessary condition to form G-dots. We believe that the inherent
aromatic nature of the guanine nucleotide is sufficient to exhibit
CDs like the photoluminescence property. The circular dichroism
spectrum of G-dots showed intense positive peaks at 218, 270 nm
and a shoulder at B300 nm. It also showed negative peaks
at B230 and B260 nm, suggesting the presence of continuous
self-assemblies of a 50 -GMP like structure (Fig. S7, ESI†). Moreover,
the XPS spectra of carbon (1s), oxygen (2p), nitrogen (2p), and
phosphorous (p,s) (Fig. S8, ESI†) also showed peaks corres-
ponding to 5 0 -GMP.
To understand the structure of G-dots, the 1H NMR spectra
of the G-dots were recorded in both D2O (Fig. 2a) and DMSO-d6
(Fig. S9, ESI†), and the 31P NMR spectrum was recorded in D2O
(Fig. 2b). Interestingly, all the protons were intact in the G-dots.
This journal is © The Royal Society of Chemistry 2016

ChemComm
Published on 25 August 2016. Downloaded by Technion - Israel Institute of Technology on 27/02/2017 10:08:09.
Fig. 2 (a) 1H NMR spectra of 5 0 -GMP and (b) G-dots. (c) MALDI-TOF
spectra of G-dots.
However, in comparison to 5 0 -GMP,16 all the proton peaks for
the G-dots gave sharp lines and up-fielded to some extent,
suggesting that G-dots contain highly symmetrical species in an
electron-rich environment. The NMR spectra thus indicate a
complete conversion of 5 0 -GMP to G-dots. In the spectra
recorded in D2O, the methylene protons attached to the 5 0
carbon on the sugar ring appeared at 3.65 and 3.73 as two
doublets of doublets suggesting that the methylene protons are
in a structurally rigid environment (Fig. 2).
In the phosphorous (31P) NMR spectra of 5 0 -GMP, a single
peak was observed at 4.39 corresponding to the ortho-
phosphate monoester, whereas 31P NMR spectra of G-dots showed
a peak at 3.17, suggesting the formation of orthophosphate
diester bonds (Fig. 2b). A similar comparison for 1H NMR spectra
recorded in DMSO-d6 suggested that all the protons were intact
and were slightly up-fielded. A new peak was observed at 5.06,
which was attributed to the –OH group in the phosphate.
Matrix assisted laser desorption ionization-time of flight
(MALDI-TOF) confirms the existence of a dimer, a tetramer
and other higher order oligomeric structures of 5 0 -GMP (Fig. 2c
and Fig. S10, S11, ESI†).
All the above observations can be collectively put-together to
elaborate the structure of G-dots. We believe that the spontaneous
self-assembly of guanosine drives the formation of G-dots. Moreover,
stacking of guanosine quartets allows self-assembly and the quartets
are further cross-linked by the orthophosphate diester bonds.
The other oxygen (O ) of the phosphate gets protonated and is
neutralized to reduce repulsion between adjacent stacking.
Continued polymerization of 5 0 -GMP through formation of
phosphodiester bonds produces guanosine-polymers of variable
length, which contribute to the rigidity of the G-dots. Moreover,
we believe that the presence of a wide range of polymeric length
of guanosine forms surface traps to contribute to the excitation
dependent emission property of the G-dots.
Guanosine derivatives are known to form supramolecular
hydrogels, which have found applications in tissue engineering,
birefringent materials and sensors.2 5 0 -GMP is known to
This journal is © The Royal Society of Chemistry 2016
View Article Online
Communication
Fig. 3 (a) Solution after microwave treatment of 5 0 -GMP. (b) Inverted
image showing the formation of the gel. (c) Fluorescence image of the
hydrogel, when excited at 365 nm. (d) HR-TEM image of the G-dot
hydrogel. (e) G 0 and G00 versus angular frequency at a constant strain
(0.01%) for the G-dot gel and G-dot gel + K+ hydrogel.
self-assemble to form a G-quartet based short range columnar
self-assembled structure, where sodium stabilizes the stacking
of quartets to some extent. It is water soluble and forms
hydrogels at acidic pH or in the presence of selective cations
such as K+, Ag+, etc. It also forms binary hydrogels with addition
of guanosine derived insoluble components.17 It is interesting
to observe that the G-dots are sparingly soluble in water. We
envisaged that a high concentration of G-dots might form
hydrogels, where the presence of a small number of guanine
motifs of 5 0 -GMP would facilitate the self-assembly of the
G-dots with the help of sodium to form extended structures.
Accordingly, microwave irradiation of a 140 mM 50 -GMP in Milli-Q
water (1.5 mL) resulted in a transparent solution, which formed
an opaque hydrogel after cooling (Fig. 3a–c). TEM analysis of the
dispersed hydrogel showed a network of fibrillar structures of
lengths extending to several micrometres (Fig. 3d). The fibrils
were possibly formed due to the super-assembly of G-dots, which
induces heterogeneity and renders stability to the gel. The CD
spectrum of the G-dot derived hydrogel showed features similar
to the G-dots (Fig. S7, ESI†) and suggested that the hydrogel
retains the chirality.
To investigate the visco-elastic behavior of the covalently
cross-linked G-dot hydrogels, oscillatory rheological measure-
ments were performed in frequency sweep modes at room
temperature. The storage modulus (G) was higher than the loss
modulus (G00 ) and they did not intersect throughout the applied
frequency range, suggesting the elastic nature of the hydrogel
(Fig. 3e). Importantly, the hydrogels retain the structural
similarity and elasticity even after terminating the shear-stress.
The thermal properties of the G-dot hydrogel examined using
differential scanning calorimetry (DSC) gave a gel–sol transition
temperature of 58 1C (Fig. S12, ESI†). The potassium cation
is known to drive the G-quartet assembly.1–3 We have then
examined the effect of K+ ions on the G-dot self-assembly to
form stable hydrogels. The G-dot gel prepared in the presence
Chem. Commun., 2016, 52, 11159--11162 | 11161

Communication
of K+ was found to exhibit a fibrillar structure with thinner
diameter (Fig. S1b, ESI†) with a higher thermal stability (gel–sol
transition temperature of 70 1C) and higher elasticity compared
to the G-dot gel (Fig. 3). All the above observations suggest that
the self-assembly drives the formation of G-dot hydrogels.
In summary, we have demonstrated the unprecedented
Published on 25 August 2016. Downloaded by Technion - Israel Institute of Technology on 27/02/2017 10:08:09.
formation of highly photoluminescent G-dots from the molecular
precursor 5 0 -guanosine monophosphate. The G-dots consist of
5 0 -GMP oligomers that form G-quartet like self-assemblies, which
is in sharp contrast to the conventional graphitic core based
carbon dots. The spectroscopic evidence collectively suggests that
guanine and sugar component do not undergo rearrangement
during the microwave treatment. RNA oligomers are believed to
be formed in a high energy environment under the prebiotic
conditions,18 and such a self-assembled structure formation of
G-dots from the 5 0 -GMP monomer may provide evidence for
that. The G-dots utilize the inherent self-assembly property of
50 -GMP to form fluorescent hydrogels without any externally added
mono-positive cations indicating that these nano-dots can be used
as a design platform for creating complex nano-structured
materials.19 Moreover, bio-compatible and biodegradable hydrogels
can find potential applications in long term cell tracking, drug
delivery and biomedical applications.20
We thank Department of Atomic Energy (BNRS) and the
Department of Science and Technology (DST) Nanomission
for research funding. TB thanks CSIR, India, for a research
fellowship. We thank Mr Pavan Kumar and Rabindra Nath Das
for useful suggestions.
References
1 (a) F. A. Aldaye, A. L. Palmer and H. F. Sleiman, Science, 2008,
321, 1795; (b) F. Wang, C.-H. Lu and I. Willner, Chem. Rev., 2014,
114, 2881; (c) F. Wang, X. Liu and I. Willner, Angew. Chem., Int. Ed.,
2015, 54, 1098.
2 (a) L. E. Buerkle and S. J. Rowan, Chem. Soc. Rev., 2012, 41, 6089;
(b) J. Dash and P. Saha, Org. Biomol. Chem., 2016, 14, 2157;
(c) G. M. Peters and J. T. Davis, Chem. Soc. Rev., 2016, 45, 3188.
3 (a) I. Bang, Biochem. Z., 1910, 26, 293; (b) M. Gellert, M. Lipsett and
D. Davies, Proc. Natl. Acad. Sci. U. S. A., 1962, 48, 2013; (c) J. T. Davis,
Angew. Chem., Int. Ed., 2004, 43, 668; (d) J. T. Davis and G. P. Spada,
Chem. Soc. Rev., 2007, 36, 296.
4 (a) Y. Qin and L. H. Hurley, Biochimie, 2008, 90, 1149; (b) A. De Cian,
L. Lacroix, C. Douarre, N. Temime-Smaali, C. Trentesaux, J.-F. Riou
and J.-L. Mergny, Biochimie, 2008, 90, 131; (c) A. Bugaut and
S. Balasubramanian, Nucleic Acids Res., 2012, 40, 4727; (d) P. Murat
and S. Balasubramanian, Curr. Opin. Genet. Dev., 2014, 25, 22.
5 (a) G. Hong, S. Diao, A. L. Antaris and H. Dai, Chem. Rev., 2015,
115, 10816; (b) S. Y. Lim, W. Shen and Z. Gao, Chem. Soc. Rev., 2015,
44, 362; (c) C. Ding, A. Zhu and Y. Tian, Acc. Chem. Res., 2013, 47, 20;
(d) L. Cao, M. J. Meziani, S. Sahu and Y. P. Sun, Acc. Chem. Res., 2013,
46, 171; (e) C. T. Nguyen, Y. Zhu, X. Chen, G. A. Sotzing, S. G. Focil
and R. M. Kasi, J. Mater. Chem. C, 2015, 3, 399.
6 (a) A. B. Bourlinos, A. Bakandritsos, A. Kouloumpis, D. Gournis,
M. Krysmann, E. P. Giannelis, K. Polakova, K. Safarova, K. Hola and
R. Zboril, J. Mater. Chem., 2012, 22, 23327; (b) P. G. Luo, S. Sahu,
S.-T. Yang, S. K. Sonkar, J. Wang, H. Wang, G. E. LeCroy, L. Cao and
Y.-P. Sun, J. Mater. Chem. B, 2013, 1, 2116.
11162 | Chem. Commun., 2016, 52, 11159--11162
View Article Online
ChemComm
7 (a) W. Shi, X. Li and H. Ma, Angew. Chem., Int. Ed., 2012, 51, 6432;
(b) S. Zhu, Q. Meng, L. Wang, J. Zhang, Y. Song, H. Jin, K. Zhang,
H. Sun, H. Wang and B. Yang, Angew. Chem., Int. Ed., 2013,
125, 4045; (c) M. Lin, H. Y. Zou, T. Yang, Z. X. Liu, H. Liu and
C. Z. Huang, Nanoscale, 2016, 8, 2999; (d) M. Lan, Y. Di, X. Zhu,
T.-W. Ng, J. Xia, W. Liu, X. Meng, P. Wang, C.-S. Lee and W. Zhang,
Chem. Commun., 2015, 51, 15574.
8 (a) J. Kim, J. Park, H. Kim, K. Singha and W. J. Kim, Biomaterials,
2013, 34, 7168; (b) M. Zheng, S. Liu, J. Li, D. Qu, H. Zhao, X. Guan,
X. Hu, Z. Xie, X. Jing and Z. Sun, Adv. Mater., 2014, 26, 3554;
(c) J. Wang and J. Qiu, J. Mater. Sci., 2016, 51, 4728; (d) M. Zheng,
S. Ruan, S. Liu, T. Sun, D. Qu, H. Zhao, Z. Xie, H. Gao, X. Jing and
Z. Sun, ACS Nano, 2015, 9, 11455.
9 (a) P. Mirtchev, E. J. Henderson, N. Soheilnia, C. M. Yip and
G. A. Ozin, J. Mater. Chem., 2012, 22, 1265; (b) Z. Ma, Y.-L. Zhang,
L. Wang, H. Ming, H. Li, X. Zhang, F. Wang, Y. Liu, Z. Kang and
S.-T. Lee, ACS Appl. Mater. Interfaces, 2013, 5, 5080.
10 (a) A. Y.-Y. Tam and V. W.-W. Yam, Chem. Soc. Rev., 2013, 42, 1540;
(b) D. K. Smith, Nat. Chem., 2010, 2, 162.
11 (a) F. Wang, Y.-H. Chen, C.-Y. Liu and D.-G. Ma, Chem. Commun.,
2011, 47, 3502; (b) X. Zhang, Y. Zhang, Y. Wang, S. Kalytchuk,
S. V. Kershaw, Y. Wang, P. Wang, T. Zhang, Y. Zhao and H. Zhang,
ACS Nano, 2013, 7, 11234; (c) W. Kwon, S. Do, J. Lee, S. Hwang,
J. K. Kim and S.-W. Rhee, Chem. Mater., 2013, 25, 1893.
12 (a) J. C. Vinci, I. M. Ferrer, S. J. Seedhouse, A. K. Bourdon,
J. M. Reynard, B. A. Foster, F. V. Bright and L. A. Coloón, J. Phys.
Chem. Lett., 2013, 4, 239; (b) Q. Xu, Y. Liu, C. Gao, J. Wei, H. Zhou,
Y. Chen, C. Dong, T. S. Sreeprasad, N. Li and Z. Xia, J. Mater.
Chem. C, 2015, 3, 9885; (c) S. Hu, W. Zhang, Q. Chang, J. Yang and
K. Lin, Carbon, 2016, 103, 391; (d) F. Arcudi, L. Ðord-ević and
M. Prato, Angew. Chem., Int. Ed., 2016, 55, 2107.
13 (a) S. K. Das, Y. Liu, S. Yeom, D. Y. Kim and C. I. Richards, Nano
Lett., 2014, 14, 620; (b) Y. Lu, L. Zhang and H. Lin, Chem. – Eur. J.,
2014, 20, 4246; (c) V. Strauss, J. T. Margraf, C. Dolle, B. Butz,
T. J. Nacken, J. Walter, W. Bauer, W. Peukert, E. Spiecker, T. Clark
and D. M. Guldi, J. Am. Chem. Soc., 2014, 136, 17308; (d) Y. H. Yuan,
Z. X. Liu, R. S. Li, H. Y. Zou, M. Lin, H. Liu and C. Z. Huang,
Nanoscale, 2016, 8, 6770.
14 (a) L. Wang, S. J. Zhu, H. Y. Wang, S. N. Qu, Y. L. Zhang, J. H. Zhang,
Q. D. Chen, H. L. Xu, W. Han, B. Yang and H. B. Sun, ACS Nano,
2014, 8, 2541; (b) H. P. Liu, T. Ye and C. D. Mao, Angew. Chem.,
Int. Ed., 2007, 46, 6473; (c) M. J. Krysmann, A. Kelarakis, P. Dallas
and E. P. Giannelis, J. Am. Chem. Soc., 2012, 134, 747.
15 (a) J. Dash, A. J. Patil, R. N. Das, F. L. Dowdall and S. Mann,
Soft Matter, 2011, 7, 8120; (b) R. N. Das, Y. P. Kumar, S. Pagoti,
A. J. Patil and J. Dash, Chem. – Eur. J., 2012, 19, 6008; (c) Y. P. Kumar,
R. N. Das, S. Kumar, O. M. Schutte, C. Steinem and J. Dash,
Chem. – Eur. J., 2014, 20, 3023; (d) R. N. Das, Y. P. Kumar, O. M.
Schutte, C. Steinem and J. Dash, J. Am. Chem. Soc., 2015, 137, 34;
(e) Y. P. Kumar, R. Das, O. M. Schütte, C. Steinem and J. Dash,
Nat. Protoc., 2016, 11, 1039.
16 (a) A. Wong, R. Ida, l. Spindler and G. Wu, J. Am. Chem. Soc., 2005,
127, 6990; (b) G. Wu and I. C. M. Kwan, J. Am. Chem. Soc., 2009,
131, 3180.
17 (a) Y. Yu, D. Nakamura, K. DeBoyace, A. W. Neisius and L. B. McGown,
J. Phys. Chem. B, 2008, 112, 1130; (b) L. E. Buerkle, Z. Li, A. M. Jamieson
and S. J. Rowan, Langmuir, 2009, 25, 8833; (c) Z. Li, L. E. Buerkle,
M. R. Orseno, K. A. Streletzky, S. Seifert, A. M. Jamieson and S. J. Rowan,
Langmuir, 2010, 26, 10093; (d) L. E. Buerkle, H. A. von Recum and
S. J. Rowan, Chem. Sci., 2012, 3, 564.
18 A. C. Rios, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 643.
19 (a) J.-M. Lehn, Science, 2002, 295, 2400; (b) G. M. Whitesides and
B. Grzybowski, Science, 2002, 295, 2418.
20 (a) R. Xing, K. Liu, T. Jiao, N. Zhang, K. Ma, R. Zhang, Q. Zou, G. Ma
and X. Yan, Adv. Mater., 2016, 28, 3669; (b) K. Liu, R. Xing, Q. Zou,
G. Ma, H. Möhwald and X. Yan, Angew. Chem., Int. Ed., 2016,
55, 3036.
This journal is © The Royal Society of Chemistry 2016
 PCCP
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

View Article Online

PAPER View Journal | View Issue

What type of nanoscopic environment does a

cationic fluorophore experience in room
Cite this: Phys. Chem. Chem. Phys.,
temperature ionic liquids?†
Open Access Article. Published on 26 May 2015. Downloaded on 1/31/2023 9:54:00 AM.

2015, 17, 16587

Anup Ghosh, Chayan K. De, Tanmay Chatterjee and Prasun K. Mandal*

Received 8th April 2015,
Accepted 26th May 2015
DOI: 10.1039/c5cp02036k
www.rsc.org/pccp
In the presence of a cationic fluorophore (rhodamine 6G) whose absorption has a significant spectral
overlap with the emission of a room temperature ionic liquid (RTIL), the emission of the latter gets
quenched, and the quenching has been shown to be dynamic in nature. It has been shown that
resonance energy transfer (RET) indeed happens between the RTIL (donor) and rhodamine 6G (cationic
acceptor), and RET is the reason for the quenching of the RTIL emission. The spectral and temporal
aspects of the RET (between neat RTILs as the donors and rhodamine 6G as the acceptor) were closely
studied by steady-state and picosecond time-resolved fluorescence spectroscopy. The influence of the
alkyl chain length of the cation, size of the anion, excitation wavelength and concentration of the
acceptor on the RET dynamics were also investigated. The energy transfer time (obtained from the rise
time of the acceptor) was noted to vary from 2.5 ns to 4.1 ns. By employing the Förster formulation, the
donor–acceptor distance was obtained, and its magnitude was found to vary between 31.8 and 37.1 Å.
The magnitude of the donor–acceptor distance was shown to be independent of the alkyl chain length of
the cation but dependent on the size of the anion of the RTIL. Moreover, the donor–acceptor distance
was observed to be independent of the excitation wavelength or concentration of the acceptor. It was
shown that the Förster formulation can possibly account for the mechanism and hence can explain the
experimental observables in the RET phenomenon. Following the detailed experiments and rigorous
analysis, a model has been put forward, which can successfully explain the nanoscopic environment that
a cationic fluorophore experiences in an RTIL. Moreover, the nanoscopic environment experienced by
the cationic probe has been noted to be different from that experienced by a neutral fluorophore.

Introduction applications, such as batteries, fuel cell photovoltaics, super-

The versatility of a solvent depends on its ability to adjust its
nanoscopic environment so as to incorporate the incoming solute.1
In this regard, room temperature ionic liquids (RTILs) are quite
special as they can dissolve a range of compounds ranging from
nonpolar hexane to highly polar water, as well as from neutral to
cationic to anionic probes.1,2 Thus, remarkable interest in RTILs as a
class of non-volatile environmentally-green benign solvents has been
generated. Moreover, RTILs have other interesting properties like
very low vapor pressure, low melting point, high thermal stability,
inflammability, recyclability, high ionic conductivity, etc.3–9 Thus,
RTILs have been actively used in synthesis, (bio)catalysis, material
science, and chemical engineering specially in energy-related
Department of Chemical Sciences, Indian Institute of Science Education and
Research (IISER) – Kolkata, Mohanpur, West-Bengal, 741246, India.
E-mail: prasunchem@iiserkol.ac.in
† Electronic supplementary information (ESI) available: Steady state absorption,
fluorescence, time resolved fluorescence decay, time constants of fluorescence
decay, Stern–Volmer plot, etc. See DOI: 10.1039/c5cp02036k
capacitors, etc.3–9 However, what kind of nanoscopic environ-
ment a dissolved solute (be it nonpolar/polar/neutral/cationic/
anionic) experiences in an RTIL is still not known in detail.
RTILs are special in the sense that, unlike conventional organic
solvents, they possess structural nano-domains both in the solid
and liquid state.10–26 From crystal structure studies it has been
shown that nanostructural heterogeneities in neat RTILs can
range up to a few nanometers, with the actual magnitude
depending on the alkyl chain length of the imidazolium ring.
The existence of p–p interactions and intermolecular hydrogen
bonding between imidazolium rings have been reported.10,13,14
The existence of local mesoscopic heterogeneities, i.e. in the
nanoscale segregation of local structures, has been shown to
prevail in RTILs by neutron scattering studies, as well as by
simulation studies, and the size of the nonpolar domain has
been shown to increase with increases in the alkyl chain length
of the imidazolium ring.15–29
The unusual steady-state excitation wavelength dependent
emission behaviour (known as the red edge effect (REE))

This journal is © the Owner Societies 2015 Phys. Chem. Chem. Phys., 2015, 17, 16587--16593 | 16587

Paper
exhibited by neat RTILs has been assigned to the existence of
the different energetically associated forms and to the lack of
solvation/energy transfer between those species.30,31 Recently,
it has been shown that neat RTILs exhibit excitation, as well as
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
monitoring, wavelength dependent fluorescence decay behaviour.32
No rise time in picoseconds or longer time regimes could be
observed.32 This proves that no energy transfer among different
nano-domains takes place in picoseconds or longer timescales.
Thus, the reason for the observation of REE in neat RTILs can be
understood. It has also been shown that some dipolar solutes
show REE in an RTIL perhaps because of incomplete solvation or
Open Access Article. Published on 26 May 2015. Downloaded on 1/31/2023 9:54:00 AM.
the existence of specific interactions (like H-bonding) between
the dipolar probe and RTIL.33–35 Fluorescence dynamical studies
have shown that probe-dependent solvation in RTILs is biphasic
in nature, where one component is typically of a few hundred
picoseconds and the other component is in the range of a few
nanoseconds.35–38 However, because of the complex structural
aspect of RTILs, the nature of the solvation structure is still not
fully understood. Hence, it is very much necessary to understand
the nanoscopic environment a probe experiences while being
dissolved in an RTIL.
Resonance energy transfer studies in RTILs have been reported
earlier.39–42 Very recently a modest but significant effort has been
made towards understanding the nanoscopic environment a
fluorophore experiences while being dissolved in an RTIL by
exploring this with the resonance energy transfer (RET) techni-
que.43 The probe chosen was a neutral one.43 Förster formulation44
has been shown to explain the fluorescence dynamical parameters
related to the RET phenomenon with reasonable accuracy. The
rise time (varying from 3.00 to 4.00 ns) or rate constant of energy
transfer has been shown to be independent of the excitation
wavelength and the concentration of the acceptor. However, the
magnitude of the rise time has been shown to vary with
the nature of the cation (alkyl chain length) and with the size
of the anion. By employing Förster formulation, the donor–
acceptor distance inside nanostructural RTIL media could be
obtained, and were seen to vary from 30.35 to 36.48 Å.43 The size
of the nanostructural cage has been shown to be dependent on
the size of the alkyl chain length of the cation, as well as on the
size of the anion of the RTILs.43 It was shown that the size of the
nanoscopic cages does not depend on the excitation wavelength
or concentration of the acceptor. Following detailed experiments
and rigorous analysis, a model has been put forward that can
successfully explain the nanoscopic environment that a neutral
fluorophore experiences in an RTIL.43
RTILs are composed of cations and anions. Hence the kind of
environment an ionic probe will experience should be different
from the environment experienced by a neutral probe. In order
to gain an insight in this direction, we used rhodamine 6G as a
reference cationic probe to help us explore the RET phenomenon
between RTILs (as a donor) and rhodamine 6G (as an acceptor).
We employed RTILs with the same cation but different anion
and vice versa in order to understand the effect of the cation
and anion on the nanocages that are formed surrounding the
cationic probe. We used different concentrations of the reference
cationic probe (acceptor) in order to understand the effect of
16588 | Phys. Chem. Chem. Phys., 2015, 17, 16587--16593
View Article Online
PCCP
Chart 1 Chemical structure of 5 RTILs and rhodamine 6G.
the concentration of the acceptor on the RET dynamics. Moreover,
since different fluorescence dynamics in RTILs have been shown
to be excitation and monitoring wavelength dependent, we
probed different excitation and monitoring wavelengths to probe
their effect on RET dynamics. All these rigorous experiments, are
helpful in order to understand whether the Förster method44 can
account for the experimentally observable RET phenomenon.
The chemical structure of the RTILs and rhodamine 6G are
depicted in Chart 1.
Experimental
Steady-state absorption and fluorescence spectra were recorded
with a UV-vis spectrophotometer (Cary100, Varian) and a spectro-
fluorimeter (Fluoromax-3), respectively. The experimental tempera-
ture was maintained at 298 K. Time resolved fluorescence decay
experiments were carried out using a time-correlated single-photon
counting (TCSPC) spectrometer (5000, IBH). Two diode lasers
(377 nm and 402 nm) were used as the excitation sources, and an
MCP photomultiplier was used as the detector. The instrument
response function (IRF) was detected by placing a scatterer, and
the FWHM of the IRF was B70 ps. A nonlinear least-squares
iteration procedure using IBH DAS 6 (Version 2.2) decay analysis
software was used for the fitting of all the decay curves. w2 values
and the plot of residuals were used as parameters to check the
goodness of the fit. The fluorescence quantum yields were
calculated using coumarin 153 as a reference.
Results and discussion
The RTILs show (see the ESI†) typical broad absorption spectra
in UV region, with a long tail extending in to the visible region.
The emission behavior of these five RTILs is quite broad in
the UV-vis region (see ESI†). A representative plot emission of
hmim[BF4] is shown in Fig. 1. Also shown in the same figure is
the absorption spectrum of rhodamine 6G. The significant
spectral overlap between the emission spectrum of the donor
(RTIL) and the absorption spectrum of the acceptor (rhodamine
6G) fulfils the prerequisite condition to observe RET between
the donor and the acceptor.45,46
With increasing the concentration of the acceptor (rhodamine
6G), the emission intensity of the donor (hmim[BF4]) decreases
This journal is © the Owner Societies 2015
 View Article Online

PCCP Paper
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
Open Access Article. Published on 26 May 2015. Downloaded on 1/31/2023 9:54:00 AM.

Fig. 1 Emission spectrum of the donor (hmim[BF4]) [lex = 377 nm] and the
absorption spectrum of the acceptor (rhodamine 6G).
gradually (Fig. 2a). This observation indicates that more and more
quenching of the donor emission is happening with increase in
the concentration of the acceptor. From the steady-state emission
of the donor, a Stern–Volmer plot was obtained and is depicted
in Fig. 2b. A linear Stern–Volmer plot (Fig. 2b) indicates that the
nature of the quenching is either static or dynamic.45,46
In order to know whether static or dynamic quenching is
happening, the time resolved fluorescence decay of the donor was
measured. As can be seen from Fig. 3, the donor (hmim[BF4])
Fig. 3 Fluorescence decay curve of RTIL (hmim[BF4]) in the absence and
in the presence of the acceptor (lex = 377 nm and lem = 450 nm)
(The decay behavior has been fitted with a tri-exponential decay function
P P
y = y0 + Bi* exp(t/ti). (Note that Bi = 100)).
decay becomes faster with increasing the concentration of the
acceptor (rhodamine 6G). This observation clearly indicates
that the nature of the quenching is dynamic.
The excited state time constants of the donor decay are
depicted in Table 1. As can be seen from Table 1, for excitation
at 377 nm, the excitation average lifetime of the donor (RTIL)
decreases from 4.87 ns to 4.08 ns on changing the concentration of
the acceptor (rhodamine 6G) from 0 to 236 mM. Thus, a reduction
of B16% of the average lifetime of the donor was noted. Also,
for excitation at 402 nm, the reduction of the average lifetime of
the donor was noted to be B16% (from 6.14 ns to 5.16 ns).
Thus, the reduction of the average lifetime of the donor was
noted to be independent of the excitation wavelength.
However, that still does not prove that the quenching occurs
because of the resonance energy transfer between the RTIL
as a donor and rhodamine 6G as an acceptor. It has been
documented in literature that it could only be the quenching of
the donor fluorescence, without resonance energy transfer to
acceptor.45–48 We would like to emphasise that observation of
the Stern–Volmer plot or the faster fluorescence decay of the
donor in the presence of the acceptor does not necessarily

Table 1 Excited state time-constants of the donor (hmim[BF4]) in the

presence various concentrations of acceptor (rhodamine 6G)

lex lem Conc. t1 t2 t3 hti

(nm) (nm) (mM) (ns) B1 (ns) B2 (ns) B3 (ns) w2
377 450 0 0.34 9.20 2.07 39.79 7.89 51.01 4.87 1.09
88 0.47 12.00 2.23 36.77 7.71 51.23 4.84 1.20
110 0.51 10.05 1.95 34.75 7.08 55.20 4.63 1.35
142 0.50 9.69 1.90 34.34 6.94 55.97 4.58 1.36
176 0.37 10.8 2.00 43.44 7.60 45.75 4.38 1.27
236 0.45 13.54 2.02 43.73 7.36 42.73 4.08 1.22

402 470 0 0.88 13.25 3.31 43.84 10.67 42.91 6.14 1.06
88 0.34 7.82 2.42 39.25 9.10 52.93 5.79 1.22
Fig. 2 (a) Variation of the RTIL (hmim[BF4]) emission intensity (lex = 377 nm) 110 0.34 8.63 2.36 39.86 8.78 51.51 5.49 1.21
w.r.t. the increasing concentration of rhodamine 6G; (b) Stern–Volmer 142 0.26 8.37 2.20 38.73 8.45 52.90 5.34 1.21
plot showing the concentration-dependent change of fluorescence of 176 0.27 7.80 2.21 38.28 8.42 53.91 5.26 1.18
the donor. 236 0.26 6.92 1.96 36.61 7.85 56.47 5.16 1.28

This journal is © the Owner Societies 2015 Phys. Chem. Chem. Phys., 2015, 17, 16587--16593 | 16589
 View Article Online

Paper PCCP

Table 2 Excited state time-constants of rhodamine 6G (acceptor) (dif-

ferent concentrations) in hmim[BF4]a

lex lem Conc. t1 t2

(nm) (nm) (mM) (ns) B1 (ns) B2 w2
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

377 550 88 2.60 1.88 4.99 101.88 1.24

110 2.62 2.52 5.16 102.52 1.14
142 2.64 4.17 5.45 104.17 1.24
176 2.68 5.23 5.53 105.23 1.27
236 2.67 1.87 5.59 101.88 1.25

402 560 88 2.50 4.79 5.19 104.79 1.07

110 2.49 5.56 5.34 105.56 1.08
3.81
Open Access Article. Published on 26 May 2015. Downloaded on 1/31/2023 9:54:00 AM.

142 2.57 5.55 103.81 0.97

176 2.54 5.98 5.52 105.98 1.08
236 2.49 3.69 5.84 103.69 1.02
a
Fig. 4 Fluorescence decay curve of the acceptor (rhodamine 6G) (lex = The decay behaviour
P has been fitted with P a tri-exponential decay
377 nm, lem = 550 nm, concentration of acceptor = 264 mM). A clear rise is function y = y0 + Bi* exp(t/ti). (Note that Bi = 100). (A negative
shown in the inset. value of Bi means a rise and a positive value of Bi means decay).

The magnitude of spectral overlap J(l) could be calculated

mean that RET is happening. To check whether RET is indeed
happening or not we measured the fluorescence decay of the
acceptor. The excitation wavelength for the fluorescence decay
of the acceptor (following RET) was judiciously chosen in such
a way that mostly donors will be excited at the excitation
wavelength and the extent of direct excitation of the acceptor
is at a minimum. If RET is happening between the RTIL (donor)
and rhodamine 6G (acceptor) then we should observe a rise
in the fluorescence decay profile of the acceptor in a shorter
timescale followed by a decay at the longer timescales.45,46 The
fluorescence decay behavior of rhodamine 6G (acceptor) is
depicted in Fig. 4.
The observation of a rise in fluorescence decay means that at
the time of excitation the acceptor molecules are not directly
excited, rather they are excited because of the RET and the rise
time is the time taken for the RET to happen. A clear rise (see
inset of Fig. 4) followed by decay of the acceptor proves that
RET is indeed happening from the RTIL (hmim[BF4]) (as donor)
to the rhodamine 6G (acceptor) (for the other four RTILs see the
ESI†). This proves that the reason for the quenching of the
donor fluorescence is RET from the donor to the acceptor. To
probe whether there is any excitation wavelength dependent
RET dynamics, we performed time resolved fluorescence decay
measurements of the acceptor monitoring at 560 nm following
excitation with 377 nm and 402 nm diode lasers. To probe
whether there is any concentration dependent rise time of RET,
we chose different concentrations of the acceptor (88, 110, 142,
176, 236 mM). The time constants of the RET phenomenon and
that of the acceptor decay are depicted in Table 2 (for the other
four RTILs see the ESI†). As can be seen from Table 2, the (rise)
time of RET is B2.60 ns to 2.68 ns at 377 nm excitation and
2.49 ns to 2.57 ns at 402 nm excitation. Thus, the magnitude
of rise time (hence the rate of RET) is noted to be independent of
the excitation wavelength. More interestingly, the magnitude of
the rise time does not change significantly (2.60 to 2.68 ns) with
increase in the acceptor concentration (from 88 to 236 mM).
Thus, the magnitude of the rise time (hence the energy transfer
time of RET) is noted to be independent of concentration.
following Förster theory44 using eqn (1):
Ð1
FD ðlÞeA l4 dl
JðlÞ ¼ 0 Ð 1 (1)
0 FD ðlÞdl
where, FD(l) is the normalized fluorescence intensity of the
donor in the absence of acceptor and eA is the molar extinction
coefficient of the acceptor. The Förster distance44 (R0) could be
calculated using eqn (2):
R0 = 0.211[k2n4QDJ(l)]1/6 (2)
2
where, k is the orientation factor, and its value generally varies
from 0 to 4. But here we considered the value of k2 to be 2/3 for
random orientation of the donor and acceptor molecules. QD is
the quantum yield of the donor in the absence of acceptor.
The distance between the donor and acceptor (RDA) could be
calculated following Förster theory44 using the following eqn (3):
 
1 1 R0
kFRET ¼ A ¼ 0 (3)
trise tD RDA
where, kFRET is the rate of RET (which could be obtained
experimentally from the risetime of the acceptor, tArise), t0D is
the life time of the donor in the absence of any acceptor, and R0
and RDA are the Förster distance and the actual distance
between the donor and the acceptor, respectively.
The magnitude of the parameters associated with the
Förster formulation are depicted in Table 3. Using the Förster
formulation, we calculated the donor–acceptor distance (RDA).
The magnitude of RDA was found to be independent on the
nature of the cation of the RTIL. With the same anion (FAP),
the magnitude of RDA was found to remain unchanged (37.13–
37.26 Å) with a change in the alkyl chain length (from ethyl to
hexyl). However, for the same cationic chain length (hmim), the
magnitude of RDA was found to increase (31.84 Å to 37.26 Å) as
the size of the anion increases (from BF4 to PF6 to FAP). Thus,
the distance between the donor and the acceptor remains
unchanged with changes in the alkyl chain length of the cation
but increases with increase in the size of the anion.

16590 | Phys. Chem. Chem. Phys., 2015, 17, 16587--16593 This journal is © the Owner Societies 2015
 View Article Online

PCCP Paper

Table 3 Magnitude of different physical parameters related to RET

(obtained experimentally and using the Förster formulation)

lex t0D J(l) (1015) trise R0 RDA kFRET

(nm) RTILs (ns) j (M cm1 nm4) (ns)
1
(Å) (Å) (108 s1)
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

377 emim[FAP] 8.48 0.21 1.63 4.14 41.83 37.13 2.42

bmim[FAP] 8.61 0.24 1.81 3.34 43.54 37.19 2.99
hmim[FAP] 4.82 0.23 1.11 3.28 39.72 37.26 3.04
hmim[PF6] 4.33 0.08 1.83 2.64 36.15 33.28 3.78
hmim[BF4] 4.87 0.09 1.36 2.62 34.79 31.84 3.82

402 emim[FAP] 6.37 0.13 1.98 4.14 39.81 37.17 2.41 Scheme 1 Nanocage model structure of RTILs in the presence of a
bmim[FAP] 7.30 0.19 2.03 3.24 42.63 37.10 3.08 positively charged fluorophore.
Open Access Article. Published on 26 May 2015. Downloaded on 1/31/2023 9:54:00 AM.

hmim[FAP] 4.02 0.18 1.21 3.26 38.44 37.14 3.06

hmim[PF6] 5.13 0.08 2.17 2.71 37.04 33.29
hmim[BF4] 6.14 0.10 1.74 2.50 36.91 31.77
The magnitude of RDA, i.e. the distance between the donor
and acceptor, was found to remain unchanged for different
excitation wavelengths say for 377 nm and 402 nm (for a particular
RTIL, hmim[BF4] the values are 31.84 Å and 31.77 Å, respectively).
Different photophysical processes in a particular RTIL have
been shown to be dependent on the excitation wavelength.35–37
However, a recent communication has shown that the dynamical
fluorescence behaviour of a fluorophore, (dissolved in an RTIL)
whose emission is beyond that of the RTIL, is excitation wave-
length independent.32 In the present report, we used rhodamine
6G, whose emission is beyond that of the RTIL, and the
dynamics of RET process was found to be excitation wavelength
independent. A similar excitation wavelength independent magni-
tude of RDA had been noted earlier using a neutral fluorophore.43
No matter what methodology we use, physically the donor–acceptor
distance (RDA) is fixed and independent of the method we employ.
For different excitation wavelengths, the magnitude of spectral
overlap ( J(l)), and hence the Förster distance (R0), will be different
as the emission of these RTILs is excitation wavelength dependent.
However, using the Förster formulation, even for different
excitation wavelengths the magnitude of the donor–acceptor
distance (RDA) was found to be very close for a particular RTIL.
This observation points to the fact that perhaps the Förster
formulation can account for the mechanism of RET dynamics
in RTILs. A similar observation had been noted earlier using a
neutral fluorophore.43 However, what is still intriguing is the fact
that the dynamics of RET i.e. the risetime of RET is independent
of the concentration of the acceptor. In a homogeneous medium,
with an increase in concentration of the acceptor, the distance
between the donor and acceptor is expected to decrease and
hence so will the energy transfer time. However, in the present
case, the energy transfer time has been found to be indepen-
dent of the concentration of acceptor. This means that with
increases in the acceptor concentration, the distance between
the donor–acceptor pair remains the same. In order to explain
these experimental observations we have put forward a model
shown in Scheme 1. The concentration dependent change in
RET dynamics is true for homogeneous media. However, RTILs
are known to be inherently heterogeneous. According to our
model (see Scheme 1), the cation and anion of the RTIL remain
side by side and the tail of the cation remains outside the cage.
3.69
4.00
It has been shown earlier that the emission of the RTIL is
because of the imidazolium moiety.30,31 As the tail of the cation
remains outside, the distance between the donor (imidazolium
ring) and the acceptor remains unchanged with changes in the
alkyl chain length of the cation of the RTIL. Thus, for a particular
anion, the distance between the donor and the acceptor should
remain unchanged. This is exactly what we observed experimen-
tally. This behavior is exactly opposite what had been observed
using a neutral fluorophore.43 Moreover, as the size of the anion
increases from BF4 to PF6 to FAP (for the same cation, hmim)
the distance between the imidazolium moiety (donor) and the
rhodamine 6G (acceptor) should increase (Scheme 1). This is
also exactly what we observed experimentally.
According to our model (Scheme 1), the acceptor rhodamine
6G will go to these nanocages and, since the concentration of
neat RTIL (BM) is a few orders higher than the concentration
of the fluorophore (Ba few hundred mM), with increase in the
acceptor concentration, the fluorophore will go to different nano-
cages and as a result the distance between the donor and acceptor
will remain the same for a particular RTIL–rhodamine 6G pair.
Hence, the rise time or rate constant of the RET phenomenon
should remain the same, even with increase in the concentration
of the acceptor. This is what we observed experimentally. Thus,
rigorous experiments and analyses provide evidence that nano-
cages do form in RTIL surrounding the acceptor. Thus, the
optical spectroscopic results mentioned in the manuscript help
us understand the nanoscopic environment that a cationic
probe (rhodamine 6G) experiences inside an RTIL. However,
the nanoscopic environment experienced by a cationic probe
(rhodamine 6G) is quite different from the nanoscopic environ-
ment experienced by a neutral probe. This observation makes it
more interesting to probe the nanoscopic environment experi-
enced by the anionic probe. These experiments are currently
underway and soon the results of these investigations will
be reported.
It has been shown that fluorescence from an RTIL is inherent
and not due to any impurity.30,31,41,43 Earlier, from spectroscopic
characterization (NMR and MS), it was clearly shown that RTILs
are sufficiently pure and any impurities present, if at all, should
be of sub-micromolar concentration.43 Researchers working on
RET are well aware that in order to observe RET the distance
between the donor and acceptor should be below 200 Å. In order
to achieve such a distance (assuming a hard sphere model), the

This journal is © the Owner Societies 2015 Phys. Chem. Chem. Phys., 2015, 17, 16587--16593 | 16591

Paper
concentration of the donor or acceptor should be B100 mM or
higher. Moreover, different RET dynamics should be observed
for different types of acceptor molecules (neutral and cationic
probe) using the same RTILs. Thus, it could be shown beyond
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
any doubt that it is not the impurity but rather the inherent
emission of imidazolium that is involved in the RET process.
Conclusions
In conclusion, we have shown that in the presence of a cationic
fluorophore (rhodamine 6G) as an acceptor whose absorption
Open Access Article. Published on 26 May 2015. Downloaded on 1/31/2023 9:54:00 AM.
has significant spectral overlap with the emission of the RTIL,
the emission of the latter can be quenched and the nature of
the quenching has been shown to be dynamic. It has been
shown that RET is indeed happening between the RTIL (donor)
and the rhodamine 6G (cationic acceptor) and RET is the
reason for quenching of the RTIL emission. By employing a
Förster formulation, we obtained the donor acceptor distance
(31.8 Å to 37.1 Å). The magnitude of the donor–acceptor
distance was shown to be independent of the alkyl chain length
of the cation but dependent on the size of the anion of the
RTILs. Moreover, the donor–acceptor distance was noted to be
independent of the excitation wavelength or concentration of
the acceptor. It has been shown that the Förster formulation
can perhaps account for the mechanism and hence can explain
the experimental observables in the RET phenomenon. Finally,
we put forward a model that can explain the experimental
observables with reasonable accuracy. Also, the nanoscopic
environment experienced by the cationic probe was noted to
be different from that of the neutral probe.
Notes
The authors declare no competing financial interests.
Acknowledgements
PKM thanks IISER – Kolkata for financial help and instrumental
facilities. Support from the Fast-Track Project (SR/FT/CS-52/2011)
of DST-India is gratefully acknowledged. AG thanks UGC, CKD
thanks IISER – Kolkata, TC thanks CSIR – India for Fellowship.
References
1 T. Welton, Chem. Rev., 1999, 99, 2071–2083.
2 P. Wasserscheid and W. Keim, Angew. Chem., Int. Ed., 2000,
39, 3772–3789.
3 J. Dupont, R. F. D. Souza and P. A. Z. Suarez, Chem. Rev.,
2002, 102, 3667–3692.
4 R. D. Rogers and K. R. Seddon, Science, 2003, 302, 792–793.
5 K. R. Seddon, Nat. Mater., 2003, 2, 363–365.
6 H. Weingaertner, Angew. Chem., Int. Ed., 2008, 47, 654–670.
7 J. P. Hallet and T. Welton, Chem. Rev., 2011, 111, 3508–3576.
8 J. D. Holbrey, W. M. Reichert and R. D. Rogers, Dalton
Trans., 2004, 2267–2271.
16592 | Phys. Chem. Chem. Phys., 2015, 17, 16587--16593
View Article Online
PCCP
9 S.-H. Chen, F.-R. Yang, M.-T. Wang and N.-N. Wang,
C. R. Chim., 2010, 13, 1391–1396.
10 W. A. Henderson, P. Fylstra, H. C. D. Long, P. C. Trulove and
S. Parsons, Phys. Chem. Chem. Phys., 2012, 14, 16041–16046.
11 A. Triolo, O. Russina, H.-J. Bleif and E. Di Cola, J. Phys.
Chem. B, 2007, 111, 4641–4644.
12 C. Hardacre, J. D. Holbrey, C. L. Mullan, M. Nieuwenhuyzen,
T. G. A. Youngs and D. T. Bowron, J. Phys. Chem. B, 2008,
112, 8049–8056.
13 C. Hardacre, J. D. Holbrey, C. L. Mullan, T. G. A. Youngs and
D. T. Bowron, J. Chem. Phys., 2010, 133, 074510.
14 A. Mele, C. D. Tran and S. H. D. P. Lacerda, Angew. Chem.,
Int. Ed., 2003, 42, 4364–4366.
15 A. Mele, G. Romano, M. Giannone, E. Ragg, G. Fronza,
G. Raos and V. Marcon, Angew. Chem., Int. Ed., 2006, 45,
1123–1126.
16 S. M. Urahata and M. C. C. Ribeiro, J. Chem. Phys., 2004, 120,
1855–1863.
17 Y. Wang and G. A. Voth, J. Am. Chem. Soc., 2005, 127,
12192–12193.
18 J. N. C. Lopes, M. F. C. Gomes and A. A. H Padua, J. Phys.
Chem. B, 2006, 110, 16816–16818.
19 O. Borodin and G. D. Smith, J. Phys. Chem. B, 2006, 110,
11481–11490.
20 W. Jiang, Y. Wang and G. A. Voth, J. Phys. Chem. B, 2007,
111, 4812–4818.
21 Z. Hu and C. J. Margulis, Acc. Chem. Res., 2007, 40, 1097–1105.
22 G. Raabea and J. Köhler, J. Chem. Phys., 2008, 128, 154509.
23 J. D. Andrade, E. S. Böes and H. Stassen, J. Phys. Chem. B,
2008, 112, 8966–8974.
24 T. Yan, Y. Wang and C. Knox, J. Phys. Chem. B, 2010, 114,
6905–6921.
25 S. N. V. K. Aki, J. F. Brennecke and A. Samanta, Chem.
Commun., 2001, 413–414.
26 P. K. Mandal and A. Samanta, J. Phys. Chem. B, 2005, 109,
15172–15177.
27 C. Wakai, A. Oleinikova and H. Weingaertner, J. Phys. Chem.
Lett., 2005, 109, 17028–17030.
28 F. V. Bright and G. A. Baker, J. Phys. Chem. B, 2006, 110,
5822–5823.
29 C. Wakai, A. Oleinikova and H. Weingaertner, J. Phys. Chem. B,
2006, 110, 5824.
30 A. Paul, P. K. Mandal and A. Samanta, Chem. Phys. Lett.,
2005, 402, 375–379.
31 A. Paul, P. K. Mandal and A. Samanta, J. Phys. Chem. B, 2005,
109, 9148–9153.
32 P. K. Mandal, M. Sarkar and A. Samanta, J. Phys. Chem. A,
2004, 108, 9048–9053.
33 P. K. Mandal, A. Paul and A. Samanta, J. Photochem.
Photobiol., A, 2006, 182, 113–120.
34 A. Samanta, J. Phys. Chem. B, 2006, 110, 13704–13716.
35 P. K. Mandal, S. Saha, R. Karmakar and A. Samanta,
Curr. Sci., 2006, 90, 301–310.
36 A. Samanta, J. Phys. Chem. Lett., 2010, 1, 1557–1562.
37 X.-X. Zhang, M. Liang and N. P. Ernsting, J. Phys. Chem. B,
2013, 117, 4291–4304.
This journal is © the Owner Societies 2015

PCCP
38 A. Ghosh, T. Chatterjee and P. K. Mandal, Chem. Commun.,
2012, 48, 6250–6252.
39 A. Adhikari, D. K. Das, D. K. Sasmal and K. Bhattacharyya,
J. Phys. Chem. B, 2009, 113, 3737–3743.
This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.
40 D. K. Das, A. K. Das, T. Mondal, A. K. Mandal and
K. Bhattacharyya, J. Phys. Chem. B, 2010, 114, 13159–13166.
41 H. Izawa, S. Wakizono and J. Kadokawa, Chem. Commun.,
2010, 46, 6359–6361.
42 V. G. Rao, S. Mandal, S. Ghosh, C. Banerjee and N. Sarkar,
J. Phys. Chem. B, 2012, 121, 12021–12029.
Open Access Article. Published on 26 May 2015. Downloaded on 1/31/2023 9:54:00 AM.
This journal is © the Owner Societies 2015
View Article Online
Paper
43 A. Ghosh, T. Chatterjee, D. Roy and A. Das, J. Phys. Chem. C,
2014, 118, 5051–5057.
44 Th. Förster, Ann. Phys., 1948, 2, 55–75.
45 B. Valeur, Molecular Fluorescence Principles and Applications,
Wiley-VCH Verlag GmbH, Weinheim, 2002.
46 J. R. Lakowicz, Principles of Fluorescence Spectroscopy,
Plenum Press, New York, 3rd edn, 1999.
47 S. Saha and A. Samanta, J. Phys. Chem. A, 1998, 102, 7903–7912.
48 K. Santhosh, S. Patra, S. Soumya, D. C. Khara and
A. Samanta, ChemPhysChem, 2011, 12, 2735–2741.
Phys. Chem. Chem. Phys., 2015, 17, 16587--16593 | 16593
 Research Article

pubs.acs.org/acscatalysis

Phenalenyl in a Different Role: Catalytic Activation through the

Nonbonding Molecular Orbital
Sudipta Raha Roy,† A. Nijamudheen,‡ Anand Pariyar,† Anup Ghosh,† Pavan K. Vardhanapu,†
Prasun K. Mandal,*,† Ayan Datta,*,‡ and Swadhin K. Mandal*,†
†
Department of Chemical Sciences, Indian Institute of Science Education and Research, 741252 Kolkata, Mohanpur-741246, India
‡
Department of Spectroscopy, Indian Association for the Cultivation of Science, 2A and 2B Raja S. C. Mullick Road, Jadavpur, 700032
Kolkata, West Bengal, India
*
S Supporting Information

ABSTRACT: We have demonstrated that the nonbonding molecular

orbital (NBMO) of the phenalenyl (PLY) cation can be used as a
Lewis acid catalyst for different organic transformations. Detailed
computational and spectroscopic studies for the aminolysis reaction
of epoxide reveal that this catalysis works through a different
mechanism, and the phenalenyl cation activates the amine moiety
using its empty NBMO, which triggers the epoxide ring-opening
reaction. It has been shown that the energy of the NBMO of PLY
cation plays a key role in modulating the catalytic activity. This study
establishes that the cationic state of phenalenyl unit is useful not only
for construction of the spin memory device by external spin injection
using its NBMO, but in addition, the same NBMO can act as an
organic Lewis acceptor unit to influence the catalytic outcome of a
homogeneous reaction.
KEYWORDS: catalysis, organocatalysis, organic Lewis acid, phenalenyl (PLY) cation, nonbonding molecular orbital (NBMO),
electrophilic activation, noncovalent interaction

■ INTRODUCTION
The phenalenyl (PLY) unit has played an intriguing role in
different fields of research spanning from chemistry, to material
chemistry, to device physics, acting as a key electronic reservoir.
Phenalenyl is a well-known odd alternant hydrocarbon with
high symmetry (D3h) which has the ability to form three redox
species: cation, neutral radical, and anion.1 Formation of this
redox triad involves the use of the formally nonbonding
molecular orbital (NBMO) of the phenalenyl molecule and
hence does not greatly affect the stability of the resulting
species.2,3 The transformation from cation to radical to anion of
phenalenyl unit progresses through successive electron accept-
ance in an accessible NBMO as predicted from an early Hückel
molecular orbital (HMO) calculations in 1960s.2 Later, the
concept of using this NBMO of phenalenyl was proposed for
designing neutral free radical based molecular conductors by
Haddon.3 Subsequently, this idea led to the discovery of the
best neutral organic conductor at room temperature.4 The
radical state of phenalenyl molecules has been used as a
building block to prepare intriguing materials for exploring new
conjugated electronic systems, such as multifunctional elec-
tronic and magnetic materials exhibiting simultaneous bist-
ability in multiple physical channels.4a,5 Until recently, most
studies focused on the radical state of phenalenyl based
molecules, and recent articles by Morita, Takui, and Hicks
describe the present status of phenalenyl based radical
materials.6
For some time, we started working on another possible
electronic state of phenalenyl moiety, where it stays in its
cationic state possessing an empty NBMO. Recently, we have
demonstrated that the cationic state can be generated by metal
ion coordination, and organometallic phenalenyl complexes can
act as catalysts, in which the phenalenyl unit played a significant
role in modulating the catalytic outcome.7 Furthermore, we
recently demonstrated the potential of cationic phenalenyl
moiety in molecular spintronics.8 A zinc ion coordinated
cationic phenalenyl unit in an organometallic compound acted
as a spin trap, leading to the development of the phenalenyl
based spin memory device. 8 When the nonmagnetic
phenalenylorganozinc compound is deposited over a ferro-
magnetic substrate, the cationic phenalenyl accepts electron
from the substrate and becomes magnetic, resulting in an
unexpectedly large magnetoresistance of 20% at near room
temperature which can be manipulated by external stimuli.8
This result could provide a platform for developing molecular
memory devices of next generation with superior strength,
Received: July 24, 2014
Revised: October 7, 2014
Published: October 22, 2014

© 2014 American Chemical Society 4307 dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis
which is based on the phenalenyl’s ability to accept an electron
utilizing its empty NBMO.
In the present study, we used metal-free cationic phenalenyl
compounds to design an organic Lewis acid catalyst which
utilizes the empty NBMO of phenalenyl for electron accept-
ance. In this regard, we posed a question whether the presence
of this readily available NBMO in the phenalenyl system in its
cationic state can be utilized to design well-defined molecular
catalysts. It was postulated that the vacant NBMO of
phenalenyl unit may have a great influence in catalytic
reactions. The initial interaction of the electron density of the
approaching nucleophile could be supported by the presence of
an energetically accessible orbital of the ligand system (Chart
1). It is now well documented that dramatic enhancement of
Chart 1. (a) Cationic Phenalenyl Compounds 1 and 2; (b)
Schematic Description of Molecular Orbitals of Cationic
PLY1a
a
The schema above show that NBMO of PLY cation can support
interaction with nucleophilic substrate (Nu).
the catalytic activity can be achieved by modulating the Lewis
acidity of the catalyst.9 In fact, the utilization of a Lewis acid is
considered as one of the most versatile ways to facilitate
catalysis. By definition, a Lewis acid has a low-lying LUMO that
is capable of accepting an electron pair.10 The majority of Lewis
acids are based on metal ions or elements that are electronically
unsatisfied and have easily accessible orbitals (for example,
boron- and aluminum-based Lewis acids).9 Moreover, a Lewis
acid derived from purely organic molecules is difficult to realize
as the LUMO of most neutral organic molecules is antibonding
in nature, which is difficult to access energetically during
catalysis. Very few conjugated electron greedy molecules have
been reported in catalysis.11
In this approach, we design organic Lewis acid catalysts using
the vacant NBMO of PLY cation which might have specific
advantage due to two reasons: the accessibility of the accepting
orbital and the use of NBMO will not necessitate
compromization on the stability of the transition state unlike
in a case when the molecule utilizes a formally antibonding
orbital as a major pathway toward the transition state. In the
present study, we have chosen a model reaction, namely, the
epoxide ring-opening reaction involving amine in the presence
of organic phenalenyl cations as catalysts. To the best of our
knowledge, not a single organocatalyst system is reported in
4308
Research Article
which the NBMO has been used as the major pathway of
interaction, until the contribution of our study.
■ RESULTS AND DISCUSSION
The phenalenyl (PLY)-based cations (Chart 1a) 9-methoxy-1-
ethoxyphenalenium tetrafluoroborate (1, O,O-system) and 9-
methylamino-1-ethoxyphenalenium tetrafluoroborate (2, N,O-
system) were synthesized following the reported procedure.12
For the present study, we have chosen the epoxide ring-
opening reaction with amine as a model organic reaction. The
epoxide ring-opening reaction is an important one as it can lead
to the formation of β-hydroxy amine, which has been key
intermediate for the synthesis of several drugs, pharmaceuticals,
and natural products.13 The classical approach for the synthesis
of β-hydroxy amine involves Lewis acid based14 (preferably a
metal salt) activation of epoxides followed by nucleophilic
attack with amines. As a result, various Lewis acid catalysts15,16
have been developed for activating the epoxides. As proof of
concept, we began the study by testing the catalytic activity of 1
(2.5 mol %) for the model reaction of 2-(phenoxymethyl)-
oxirane (3a) with N-methylaniline (4a) under neat condition at
40 °C. To our expectation, the desired β-hydroxy amine, 1-
(methyl(phenyl)amino)-3-phenoxypropan-2-ol (5) was ob-
tained in 84% yield (Table 1, entry 1). Subsequently, the
Table 1. Standardization of Reaction Condition for the
Epoxide Ring Openinga
a
3a (2 mmol) was treated with 4a (2 mmol, 1 equiv) in the presence
of catalyst 1 (2.5 mol %) at 40 °C under neat condition for 11 h unless
otherwise specified. bConversion was determined by 1H NMR
spectroscopy.
amount of catalyst loading was varied to determine the most
effective catalytic condition. The use of lesser amount of
catalyst 1 (1.25 mol %) has a slight detrimental effect on the
overall conversion of the reaction (Table 1, entry 2). The use of
greater amounts (5 mol %) of 1 did not provide any beneficial
effect in terms of reduction of the reaction time or conversion
(Table 1, entry 3). It may be argued that the reaction of the
PLY-based catalyst with amine may generate the HBF412a and
the in situ generated acid can catalyze the reaction. When we
mixed the catalyst 1 and amine 4a (in 1:1 ratio, see the
following text for further details) and recorded the 1H NMR
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis Research Article

Table 2. Aminolysis Reaction of Epoxide in the Presence of Catalysts 1 and 2a

entry amine epoxide yield (%)b,c

catalyst 1 catalyst 2
1 4a 3a 84 (80) 46
2 4a 3b 85 72
3 4a 3c 99 (91) 76
4 4a 3d 85 60
5 4a 3e 85 (70) 63
6 4b 3a 99 (93) 80
7 4b 3b 99 (95) 86
8 4b 3d 83 (75) 65
9 4c 3a 99 72
10 4c 3b 83 76
11 4c 3d 98 (90) 82
12 4d 3a 98 (85) 85
13 4d 3b 96 (90) 75
14 4d 3d 94 (80) 73
15 4e 3a 55 40
16 4f 3a 93 (75) 74
17 4f 3b 92 (80) 78
18 4f 3d 98 (89) 73
a
The epoxide (2 mmol) was treated with amine (2 mmol) in the presence of catalyst 1 or 2 (2.5 mol %) at 40 °C. bConversion was determined by
1
H NMR spectroscopy. cParentheses represent the isolated yield.

spectrum, we did not observe any significant change in
chemical shift from either of these two individual moieties,
which ruled out the possibility of HBF4 catalyzed reaction. The
parent 9-hydroxyphenalenone, when used as a catalyst, resulted
in poor conversion (11%, Table 1, entry 4). In another control
experiment, we used NaBF4 as catalyst which also led to very
poor catalytic activity (19%, Table 1, entry 5) revealing that the
presence of the PLY-based cationic moiety as catalyst has
prominent effect on the catalysis. In another control experi-
ment, when this reaction was performed in the absence of
catalyst 1, only 8% (Table 1, entry 6) conversion was observed.
This observation confirms the indispensable role of the PLY-
based cation during catalysis.
In order to check the versatility of this catalytic process, we
investigated the scope of catalyst 1 (2.5 mol %) for the
aminolysis reaction of several functionalized epoxides. We also
used another PLY-based cation, 9-methyiamino-1-ethoxy-
phenalenium tetrafluoroborate (2, N,O-system), as catalyst to
check the change in efficacy of the catalyst depending on the
nature of PLY cation used. Under the optimized procedure,
reactions proceed smoothly and gave the corresponding β-
hydroxy amine in good to excellent yield. The regioselective
nucleophilic attack by different amines took place almost
4309
exclusively on the less-hindered side of the terminal epoxide
(Table 2, entries 1−2, 4−5, etc.). When the cyclohexene oxide
(3c) was treated with 4a (Table 2, entries 3), it gave the
diastereoselective trans-β-hydroxy amine exclusively. The
reaction of epichlorohydrin 3e (Table 2, entry 5) provided
excellent example of chemo and regioselectivity and afforded
85% yield of the β-hydroxy amine corresponding to
nucleophilic attack at the terminal carbon of the epoxide
moiety. Different electronically and sterically diverse amines
were reacted in the presence of catalysts 1 and 2 at 40 °C under
solvent-free condition. The reaction rate appeared to be
influenced by the nature of the amine (Table 2, entries 1 and
6) as evident from the reaction performed with 4a (an aromatic
amine), resulting in less conversion than with an aliphatic
amine 4b. Treatment of terminal epoxides 3a and 3b with
morpholine 4c (Table 2, entries 9−11) and piperidine 4d
(Table 2, entries 12−14) as a congener of cyclic aliphatic
amine, gives the corresponding β-hydroxy amine in good to
excellent yield. A sterically hindered amine, such as di-
isopropylamine 4e (Table 2, entry 15), also works under the
optimized condition. We tried to synthesize the β-hydroxy
amine with long chain acyclic amine. The regioselective
nucleophilic attack took place almost exclusively on the less
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis
hindered side of the terminal epoxide (Table 2, entry 16−18)
upon the treatment of dibutylamine (4f) with the terminal
epoxides (4a, 4b, and 4d). This methodology is equally
effective for aromatic as well as aliphatic amines. The catalytic
results reveal that irrespective of amines, reaction proceeds with
excellent conversion in case of catalyst 1. Though there are only
very few reports for the epoxide ring-opening reaction in the
presence of metal-free catalysts,16a,c the present catalysts show
comparable or better catalytic activity than the reported metal-
free catalysts. Moreover, this catalyst system can stay live for a
number of consecutive catalytic cycles without losing the
catalytic activity significantly (see Supporting Information for
details). As a general observation, it was noticed from Table 2,
in all the reactions tested in the present study, the catalyst 1
outperforms catalyst 2 in terms of activity under the identical
reaction condition.
We have performed a comparative kinetic study to
understand the relative efficacy of catalysts 1 and 2 for the
model reaction involving 3a and 4a via 1H NMR spectroscopy
over the course of the first 6 h (Figure 1). From the kinetic
analysis, it is evident that catalyst 1 is superior than that of
catalyst 2, as also observed consistently from the data presented
in Table 2.
Figure 1. Kinetic studies revealing that catalyst 1 has better efficacy
than catalyst 2.
The difference in reactivity between the catalysts 1 and 2
intrigues us to find its relationship in the molecular level, and
therefore, we have carried out density functional theory (DFT)
study at the M06-2X/6-31+G(d) level of theory. An analysis of
the charge density distributions is shown in Figure 2a. The
calculated orbital energy diagrams of 1 and 2 show that the
LUMO (see Figure 2b) of these catalysts are largely
nonbonding in nature and is highly separated from the
LUMO−1 and LUMO+1(Figure 2c). The presence of the
NBMO of PLY cation as the LUMO has earlier been predicted
for the cationic state of parent phenalenyl molecule.1,3 The
LUMO of catalyst 1 (Figure 2c) is well separated from its
nearest molecular orbitals with a calculated LUMO−(LUMO
+1) gap of 2.52 eV and a LUMO−1 − LUMO gap of 5.37 eV.
4310
Research Article
At the same time, catalyst 2 has a comparatively smaller LUMO
− LUMO+1 gap of 2.34 eV and LUMO−1 − LUMO gap of
5.22 eV. Thus, the DFT calculation supports the presence of
energetically accessible empty molecular orbital for electron
acceptance in these PLY-based catalysts. Interestingly, the
calculation reveals that the relative energy of the LUMO largely
depends on the nature of the substitutions in the PLY. In other
words, the LUMO energy levels in these PLY units can be
tuned by changing the substitution from O,O (in catalyst 1) to
N,O (in catalyst 2) in the present study (Figure 2c).
To understand the electron-acceptance property of this
organic Lewis acid and to further support our DFT-calculation-
based findings, we carried out a cyclic voltammetric study.
These experiments indeed revealed the presence of two one-
electron acceptance processes in both catalyst 1 and 2 (Figure
3a and b, respectively). The E1/2 values of 1 (E11/2 = −0.408 V
and E21/2 = −1.486 V) and 2 (E11/2 = −0.931 V and E21/2 =
−1.820 V) indicate that electron acceptance from the first to
the second one becomes progressively difficult as a result of
increasing Coulombic repulsion in the system. A similar
observation was also noted by Haddon and co-workers with
the 9-ethoxy-1-ethoxyphenalenium tetrafluoroborate system.17
Earlier EPR studies on spirobiphenalenyl boron salts4,18 had
shown that the first reduction leads to the formation of an EPR
active radical species and the second reduction results in the
EPR silent anionic species, supporting our result that the
LUMO of a PLY system can accommodate two electrons. The
cyclic voltammetry thus in the present case can be explained by
considering the generation of a neutral radical and an anionic
species (Figure 3c). Comparing first reduction potential data of
1 and 2 [1 (E11/2 = −0.408 V) and 2 (E11/2 = −0.931 V)], it is
clear that the reduction of 1 is much more facile than 2. This
result points out that the LUMO of 1 is more accessible for
electron acceptance as compared to that of 2, which is in
consonance with our DFT calculation (Figure 2c).
Next, we attempted to envisage the molecular level
interaction of the catalyst with the substrates to understand
the mechanistic insight of the epoxide ring opening with amine.
A general perception for the Lewis acid catalyzed aminolysis
reaction of epoxide ring-opening reaction is that a Lewis acid
(preferably a metal salt) can activate the epoxide substrate by
forming a strong coordinate bond with the oxygen atom of the
epoxide ring. This type of interaction increases the electro-
philicity of the carbon atoms of the epoxide ring which assists
in the opening of epoxide ring.14−16 However, when catalyst 1
was mixed with amine 4a in a ratio of 1:1, there was a sharp
visual change in color from orange to dark brown. On the other
hand, no such sharp color change was observed when we
physically mix catalyst 1 and epoxide 3a. This observation
clearly highlights that NBMO-based organic Lewis acid works
in a different pathway than the traditional Lewis acid catalyzed
aminolysis reaction of epoxide and further prompted us to
record a detailed absorption and emission spectra of these
mixtures.
We followed the interaction between amine/epoxide (see
Supporting Information for epoxide) with the catalyst
spectroscopically. After a careful stepwise addition of amine
4a to a solution of 1 (25 μM), the absorbance of 1 was found to
be gradually decreased with increasing the concentration of 4a.
Presence of an isosbestic point (Figure 4a) at 468 nm for 4a
proves beyond any doubt that a ground state interaction has
indeed happened with catalyst 1 and 4a.
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis Research Article

Figure 2. (a) Mulliken charge densities on selected carbon atoms in catalyst 1. (b) Computed LUMO of the catalyst 1. (c) Molecular orbital energy
profile diagrams of catalysts 1 and 2. Energy of LUMO level (in eV) is shown.

Figure 3. (a,b) Cyclic voltammogram of catalysts 1 and 2 in acetonitrile, referenced to SCE via internal ferrocene revealing two successive one-
electron reduction processes. (c) Successive electron acceptance by catalysts 1 and 2 generates a neutral radical and an anionic species.

In order to obtain the stoichoimetry of the amine−catalyst
adduct we have carried out Job’s plot (ΔOD vs mole fraction of
4a; where ΔOD = I0 − I; I0 = absorbance of 1 in absence of 4a,
and I = absorbance of 1 in the presence of quencher 4a) which
clearly demonstrates a 1:1 association of 1 and 4a (Figure 4b).
Further, we have monitored the emission spectroscopy of 1 at
different concentration of amine to understand the exact
interaction of 1 with 4a. Sequential addition of 4a to a solution
of 1 (25 μM) resulted in a gradual decrease in the fluorescence
intensity of the catalyst 1 (Figure 4c), further confirms the
interaction between amine and catalyst. The extent of the
interaction has been followed by the Stern−Volmer plot (plot
of F0/F against the concentration of 4a; where F0 =
fluorescence intensity in absence of any quencher, F =
fluorescence intensity in the presence of quencher) (Figure
4d) as well as by Benesi−Hildebrand plot (plot of 1/[F − F0]
versus 1/[4a]) (Figure 4e). As it can be seen from Figure 4d, a
strict linear behavior of the quenching (obtained from Stern−
Volmer plot) confirms that the quenching is static in nature.19a
4311
We have also estimated the association constant of this process
between the 1 and 4a from emission spectroscopic measure-
ments by using Benesi−Hildebrand equation. From the
Benesi−Hildebrand plot (Figure 4e), the cumulative associa-
tion constant has been calculated to be 13.3 × 103 LM−1 for the
catalyst−amine pair and that for the catalyst−epoxy pair is 4.8
× 103 LM−1 (see Supporting Information for details). These
values suggest that the association between catalyst and amine
is 2−3 times much stronger than that between catalyst epoxide.
Furthermore, to understand the nature of association between
catalyst and amine, we mixed the catalyst 1 and amine 4a (in
1:1 ratio) and recorded the 1H NMR spectrum. We did not
observe any significant change in the chemical shift from either
of these two individual starting moieties (see Supporting
Information for NMR spectra). This NMR experiment clearly
suggests that there is no strong association (covalent bond
formation) between catalyst 1 and amine 4a. However, from
the absorption and emission spectroscopic studies, we can
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis Research Article

Figure 4. Spectroscopic signature of the binding of catalyst 1 with amine 4a. (a) Absorption spectra in MeCN. (b) Job’s Plot (plot of ΔOD vs mole
fraction of amine 4a). (c) Steady state emission spectra in MeCN. (d) Stern-Volmer plot (plot of F0/F vs 4a concentration). (e) Benesi Hildebrand
plot (plot of 1/[F − F0] vs 1/[4a]). For details see text.

Figure 5. Optimized geometries of catalyst 1 with (a) a model epoxide 3a and (b) amine 4a. Selected bond distances (in Å) are shown. (c)
Molecular orbital energy profile diagrams of catalyst 1, amine 4a, and epoxide 3a. Energy gaps between the LUMO level of catalyst 1 and HOMO
levels of reactants (in eV) are shown.

4312 dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319

ACS Catalysis
Figure 6. Kinetic studies of epoxide ring-opening reaction for the formation of 5 monitored by 1HNMR spectroscopy in CDCl3. (a,b) Plots showing
the change of epoxide concentration with time for the reaction of 3a (2 mmol), 4a (2 mol) with varying catalyst 1 concentration. (c,d) Plots of rate
(mmol/h) versus epoxide concentration (mmol).
conclude that there is a loose association (noncovalent)
between catalyst 1 and amine 4a.
The spectroscopic results were rationalized on the basis of
computations on the interaction of catalyst 1 with both the
amine and epoxide reactants. Ethylene oxide was used as the
simplified model to mimic the epoxide reactants. Calculations
show that epoxide could form a complex with the planar surface
of catalyst 1 through the lone pair−π interaction resulting from
an interaction of 1:1 mixture of 1 and ethylene oxide as shown
in Figure 5a. The distance of 2.83 Å between the aromatic
surface of catalyst 1 and epoxide oxygen is within the range of
conventional lone pair−π interactions.19b,c However, the
binding energy of this complex was found to be small. Binding
energies obtained for adduct of 1 with epoxide was of ΔE
(BSSE corrected) = −5.5 kcal/mol. Also, the Gibb’s free energy
changes were found to be clearly positive; +2.5 kcal/mol. This
suggests a poor association of 1 with epoxide and possibly
eliminates the well-documented electrophilic activation of
epoxide in the present case.15,16
Alternatively, it may be considered that amine might form a
more stable adduct with 1 as the nucleophilicity of the nitrogen
of the amine is more than that of the oxygen atom of the
epoxide. Interactions of catalyst 1 with the amine reactants
4313
Research Article
were then computed using N-methylaniline as the model amine
(Figure 5b). It is important to note that catalyst 1 can bind with
N-methylaniline through a large number of stacked and
nonstacked orientations in addition to a lone pair−π interaction
model. We have optimized all important geometries that could
be formed from the binding of 4a with different binding sites
on catalyst 1 to identify the most stable complex (details of the
optimized structures and their energetics are tabulated in the
Supporting Information). Among them, the most suitable
reactive intermediate (Figure 5b) is stabilized by a binding
energy of ΔE (BSSE corrected) = −11.1 kcal/mol due to
favorable π−π stacking interactions.
Furthermore, it was observed from the calculations that the
LUMO of catalyst 1 remains 0.72 eV above the HOMO of
amine 4a while for the HOMO of epoxide 3a, the LUMO of 1
is placed 1.62 eV above (Figure 5c). Thus, the comparatively
shorter HOMO−LUMO gap in the case of adduct formed
between 1 and 4a results in stronger donor−acceptor
interaction and initiate the nucleophilic activation. These
computational studies are also in agreement with our
spectroscopic results.
Detailed kinetic studies were performed following the
literature method reported earlier20 to gain further insight
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis
Figure 7. Computed potential energy surface (PES) for the reaction between 4a and ethylene oxide in the presence of catalyst 1. Selected bond
distances (in Å) and relative energies with respect to the reactants (in kcal/mol) are shown.
into the aminolysis process for the model reaction between 3a
and 4a under neat condition using 1 as catalyst. Kinetic studies
of all the reactions were performed by following the standard
optimized reaction condition. The evolution of the specific
resonances of the product β-hydroxy amine 5 was monitored by
1
H NMR spectroscopy relative to the concentration of either
starting amine or epoxide. From this conversion, the effective
concentration (mmol) of the 4a/3a over the courses of the first
9 h was determined, which in turn was used for the calculation
of rate (mmol/h) of the reaction.
To determine the order of the reaction with respect to
catalyst concentration, the aminolysis reaction was carried out
in two different concentrations of catalyst 1 (Run 1: with 5 mol
% catalyst loading and Run 2: with 7.5 mol% catalyst loading
separately) and keeping the concentration of epoxide 3a (2
mmol) and amine 4a (2 mmol) fixed. The progress of the
reaction was monitored during the first 9 h (Figure 6a,b) which
shows gradual decrease of epoxide concentration with time.
The amine and epoxide concentration at a given time can be
quantified with the help of 1H NMR spectroscopy, and the rate
of the reaction can be calculated by determining the conversion
at a given time. A plot of rate (mmol/h) versus different
epoxide concentrations was determined for two sets of
reactions (using two different catalyst concentration), which
reveals a linear increase of the reaction rate with epoxide
concentration (Figure 6c,d). Comparing the rate of reaction for
two sets of catalyst concentration, one can calculate (see
Supporting Information for detail) and confirm the first-order
dependency of the reaction rate with respect to the catalyst
concentration.
4314
Research Article
Efforts were also given to determine the order of the reaction
with respect to the amine concentration as well as with respect
to epoxide concentration, respectively, in a similar way (see
Supporting Information for details). It shows a second-order
dependency of the reaction rate with respect to the amine
concentration and a first-order dependency with epoxide
concentration.
Thus, from the present study, the overall rate law for the
aminolysis reaction of 3a with 4a catalyzed by 1 can be
summarized as shown in eq 1.
Rate = k[Catalyst]1 × [Amine]2 × [Epoxide]1 ............ (1)
Subsequently, in order to understand the complete
mechanistic cycle and the role of catalyst, we have modeled
the reaction between ethylene oxide and 4a in the presence (or
absence) of catalysts 1 or 2. Calculations at M06-2X/6-
31+G(d) level of theory predicted that, in the absence of any
catalyst, a very high activation energy (ΔΔG‡ = 35.5 kcal/mol)
is required for this chemical transformation (see Supporting
Information for detail).
The computed potential energy surface for this reaction in
the presence of catalyst 1 is shown in Figure 7. The π−π
stacked reactive intermediate 6 can readily undergo two
intramolecular rearrangements, first to 7 with strengthening
of catalyst−nitrogen bond and then to an unfolded
conformation in 8. Small activation barriers of ΔG ‡ = 13.0
kcal/mol (TS1) and 5.1 kcal/mol (TS2) were found for these
rearrangements. Both of these rearrangements occur with
strengthening of newly formed C−N bond between the catalyst
and reactant as well as with the activation of N−H bond. The
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis
Scheme 1. Schematic Presentation of Plausible Mechanism for the Phenalenyl Cation Catalyzed Reaction between Amine with
Epoxide Forming β-Hydroxy Aminea
a
A generalized presentation has been adopted for the sake of simplicity where A, B, C, and D represent 6, 9, 10, and 11 of Figure 7, respectively.
C−N bond distance decreases from 3.05 Å in 6 to 1.63 Å in 7
and subsequently shortens to 1.61 Å in the nonstacked,
unfolded conformer 8. At the same time, the N−H bond
weakens from 1.01 Å in free 4a to 1.03 Å in 8. In addition, the
DFT calculations show that the formation of 7 and 8 with
shorter C··· N bond distances (1.61−1.63 Å) are endothermic
processes, and the reverse reaction to return to the noncovalent
adduct 6 should be a rapid processes. This suggests that the
amine catalyst adduct stays in a dynamic equilibrium between 6,
7, and 8 with low energy barrier. Both the calculations and
spectroscopic experiments show that the stable catalyst···amine
complex is noncovalent in nature.
In the following step, the epoxide molecule forms a N−H···
O hydrogen bond by interacting with 8 resulting in the
intermediate 9. A second molecule of amine (4a) would further
activate the reaction by binding with the epoxide carbon atom
to form the weakly bound complex 10. In the succeeding step,
the proton abstraction by the epoxide and the formation of a
carbon−nitrogen bond between the epoxide and second
molecule of amine (4a) occurs through the concerted transition
state TS3 to form the complex 11. In the presence of epoxide,
the catalyst amine interaction switches form loose association
to a stronger association, and in 11, the C−N bond distance
approaches to 1.5 Å, indicating a covalent type bonding
interaction as shown in Figure 7. 11 can undergo a proton
transfer process forming the desired product and regenerating
6.21 Similar to the uncatalyzed reaction, this step was identified
as the rate-determining step. Importantly, the activation energy
of 25.7 kcal/mol (ΔΔG‡) required to cross the barrier TS3,
which is significantly lower compared to that of 35.5 kcal/mol
required for the uncatalyzed reaction (see Supporting
Information for detail).
Thus, our calculations justify the role of catalyst in the
reaction profile. Further, our computational results have been
corroborated by the experimental kinetic studies, which show
4315
Research Article
that the reaction is second order with respect to the amine and
first order with respect to the epoxide and the catalyst (see eq
1).
From the rate law of the reaction between 3a and 4a in the
presence of catalyst 1, it suggests the involvement of two
molecules of amine with one molecule of each of catalyst and
epoxide during the rate-determining step of the aminolysis
reaction. This is also reflected in the DFT calculated reaction
profile which revealed that the rate-determining step involves
one epoxide molecule, two amine molecules, and one catalyst
molecule (during the transformation of 10 to 11, Figure 7).
Combining all these results, we propose a schematic
presentation of plausible mechanism for the phenalenyl cation
based organic Lewis acid catalyzed reaction of amine with
epoxide (Scheme 1).
The greater nucleophilicity of the nitrogen atom of the amine
moiety in comparison to the oxygen atom of the epoxide
prefers stronger interaction between the catalyst and amine-
generating catalytically active species A. The primary
interaction between amine and catalyst in A turned out to be
a noncovalent type. This is evident from spectroscopic studies
and the DFT calculation which have shown that the minimum
energy geometry of the initial complex between 1 and 4a (6 in
Figure 7) is purely noncovalent in nature,22 and it can be
ascertained that there is a loose association between catalyst 1
and amine 4a. Interestingly, this 1:1 amine−catalyst association
in turn activates the hydrogen atom of N−H functionality of
the amine and prompts the formation of hydrogen bond with
oxygen of the epoxide molecule forming B. This hydrogen
bonding with epoxide in B activates the epoxide molecule by
increasing its electrophilicy. Then, the second molecule of
amine approaches B and forms the intermediate C by binding
with the epoxide carbon atom at the sterically and electronically
favorable site. Subsequently the C−O bond of epoxide breaks,
and the new C−N bond between epoxide and second amine
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis Research Article

Figure 8. Reaction coordinate for the rate-determining step for the reaction between N-methylaniline and ethylene oxide in the presence of catalyst
2. Selected bond distances (in Å) and relative energies in kcal/mol are shown.

molecule is formed in intermediate D. Finally, a proton transfer attracted to utilize these cationic PLY molecules as versatile
from D via E regenerates the catalytically active species A. DFT organic Lewis acids for different aldehyde based reactions, and
calculation reveals that the Gibb’s free energy change is we planned to use them as catalysts for the synthesis of
thermodynamically feasible (−23.9 kcal/mol, see Supporting bis(indolyl)methanes (see Supporting Information for plausible
Information for details) for the proton transfer process mechanism). The development of new methodologies23 for the
transforming D to A forming desired β-hydroxy amine via synthesis of bis(indolyl)methanes is a subject of continuous
epoxide ring opening and regenerating the catalyst−amine interest to synthetic organic/medicinal chemists as indole and
adduct A. their derivatives have versatile biological activities24 and are
Furthermore, to understand the superior activity of catalyst 1 widely present in various biologically active natural products.25
over catalyst 2, it was found by DFT calculation that the free To examine the effect of the cationic PLY-based molecule, the
energy of activation required for the rate-determining step
reaction of indole with 4-methoxybenzaldehyde (Scheme 2)
(shown in Figure 8) in the presence of catalyst 2 is higher in
under neat conditions and at room temperature was considered
energy than the respective barrier TS3 in the presence of
for a model study. An initial screening was done to obtain
catalyst 1 by 1.7 kcal/mol.
Thus, although catalyst 2 shows significant catalytic activity maximum conversion to the product in short period. The
compared to the uncatalyzed reaction, it is less efficient than
the catalyst 1. Hence, our computational study fully Scheme 2. Synthesis of Bisindolylmethanes in the Presence
substantiates the experimental findings. The enhanced activity of PLY Cations (1 and 2) as Organic Lewis Acid Catalysts
of catalyst 1 could be explained from a comparison of HOMO
energy of N-methylaniline with LUMO energies of both Lewis
acid catalysts 1 and 2. The LUMO of catalyst 1 lays 0.72 eV
above the HOMO of N-methylaniline, whereas for catalyst 2, it
is placed 1.07 eV above the HOMO of N-methylamine,
resulting in a larger HOMO−LUMO gap for catalyst 2 as
compared to catalyst 1. Such significant changes in the catalytic
activity were understood on the basis of the fact that the
substitution on PLY system can tune its LUMO energy. In
general, the substitution of an electron-withdrawing group
stabilizes the LUMO, whereas the substitution of an electron-
donating group can lead to the reverse effect.
To establish that our concept of using the NBMO of the PLY
cation is not limited to only epoxide ring-opening reaction and
that the same concept can be extended and generalized for
more organic transformations, we tried to check the efficacy of
catalysts 1 and 2 in different organic transformations. We can
anticipate that NBMO of the PLY cation can also be able to
activate the aldehyde. Keeping this goal in mind, we were
4316 dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis
progress of the reactions was monitored by TLC taking into
consideration the complete consumption of indole. The
reaction was completed in 40 min using catalyst 1. In a control
experiment when the model reaction was performed in the
absence of catalyst, only 10% desired product was formed. This
observation clearly highlights the role of NBMO based catalyst
for this chemical transformation.
To establish the generality of the organic Lewis acid
catalyzed bisindolylmethane formation, various functionalized
aldehydes were treated with indole (Scheme 2). The reaction is
compatible with a variety of functional groups as well as with
heterocyclic aldehyde. In all cases, the reaction proceeded
smoothly at room temperature to afford the corresponding
bisindolylmethanes in excellent yields (85−95%). To compare
the catalytic activity between the catalysts 1 and 2, few parallel
experiments were performed. Here we also found that catalyst 1
is more effective than catalyst 2.
Furthermore, we extend this concept of using NBMO based
organic Lewis acid concept in multicomponent reactions
(MCRs) as well. In recent years, several new criteria have
been brought forward in the synthetic process to solve the
challenging synthetic problem. Among them, MCRs26 have
been recognized as a new synthetic support. The perpetual
interest of synthetic chemists toward the development of newer
methodologies for the synthesis of naphthylbenzamide27
moiety relies on its broad range of biological activities.28 To
confirm the efficiency of the organic Lewis acid catalysts 1 and
2 for MCR transformation, we used a model reaction that
comprises a three-component reaction (3-MCR) involving an
benzaldehyde, β-napthol, and benzamide in the presence of
catalyst 1 (2.5 mol %) at 100 °C under neat condition (Scheme
3). To our delight, the desired naphthylbenzamide was formed
in 95% yield after 1 h. However, in the absence of catalyst, very
poor conversion was observed.
Encouraged with this preliminary result, we applied this
present methodology with the various functionalized aldehydes.
Irrespective of functionality, the final product was formed in
excellent yield (Scheme 3).
Scheme 3. Synthesis of Naphthylbenzamide in the Presence
of PLY-Based Organic Lewis Acid Catalysts 1 and 2
4317
Research Article
In these MCR reactions, we also observed the consistent
catalytic activity ordering between 1 and 2 when the reactions
were carried out under identical condition, which shows
catalyst 1 is more active in comparison to catalyst 2.
■
CONCLUSIONS
In summary, we have demonstrated that the NBMOs of the
phenalenyl cation can be used as highly active Lewis acid
catalysts for organic transformations such as the epoxide ring-
opening reaction with amine in the present case. The chemistry
of phenalenyl system started more than six decades back, and a
number of phenalenyl-based neutral free radicals have already
been realized in the solid state; their promise as fascinating
materials has been very well documented. However, with this
study, we show that the empty NBMO of organic phenalenyl
cation can be utilized as Lewis acid catalyst. This work
represents a major step in phenalenyl chemistry, dealing with
cationic phenalenyl states to design molecular catalysts using its
easily accessible electron-acceptance capability into the empty
NBMO. This particular concept was also further successfully
extended in different types of organic transformations, showing
the versatility of cataionic PLY as organic Lewis acid.
Furthermore, we have shown that by chemically tuning the
energy of NBMO of PLY cation, one can alter the Lewis acidity,
and eventually, it can change the catalytic outcome of the
reaction. This phenomenon clearly describes the importance of
the NBMO mediated organocatalysis.
■
EXPERIMENTAL SECTION
General Considerations. All solvents were distilled from
Na/benzophenone prior to use. All chemicals were purchased
from Aldrich and Fluka Chemicals and used as received. The
1
H and 13C NMR spectra were recorded on 400 and 500 MHz
spectrometer in CD3CN/CDCl3/DMSO-d6 with residual
undeuterated solvent (CD3CN: 1.94/1.3, 118.2, CDCl3: 7.26/
77.0, DMSO-d6: 2.5/39.5) as an internal standard. Chemical
shifts (δ) are given in ppm and J values are given in Hz. The
HR-MS data were obtained using a Q-TOF Micromass. Open-
column chromatography and thin-layer chromatography (TLC)
were performed on Silica gel [60−120 mesh]. Evaporation of
solvents was performed at reduced pressure using a rotary
evaporator.
Experimental Procedure for the Cyclic Voltammetry.
Cyclic Voltammetry was performed using a PAR potentiostat at
room temperature in dry acetonitrile under argon atmosphere
with n-Bu4NClO4 (0.1 M) as supporting electrolyte. Potentials
were scanned with respect to the quasi-reference electrode in a
single compartment cell fitted with Pt electrodes (working,
auxiliary and reference) and referenced to the Fc/Fc+ couple of
ferrocene at 0.38 V versus SCE. The Epa-Epc separation of the
reversible couples was within 10% of that of the Fc/Fc+ couple.
Typical Procedure for the Formation of β-Hydroxy
Amines (Table 2, Entry 1). A mixture of 2-(phenoxymethyl)-
oxirane 3a (0.30g, 2 mmol) and N-methylaniline 4a (0.214g, 2
mmol, 1 equiv) were stirred magnetically at 40 °C in the
presence of catalyst 1 (18 mg, 2.5 mol %). After 11 h, the
reaction mixture was diluted with EtOAc (5 mL) followed by
addition of water (2 mL). The supernatant EtOAc solution was
decanted. The aqueous layer/portion was extracted with EtOAc
(2 × 5 mL). The combined EtOAc extracts were dried
(Na2SO4) and concentrated under vacuum. The conversion of
the reaction was checked by 1H NMR spectroscopy. When the
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis
crude product was estimated to be greater than 80% pure, flash
chromatography (1:0 to 7:3 hexane/EtOAc) was then
performed to get the isolated yield of the corresponding pure
β-hydroxy amines.
Procedure for the Synthesis of Bisindolylmethane
(Scheme 2). A mixture of aldehyde (2 mmol, 1 equiv) and
indole (0.464 g, 4 mmol, 2 equiv) was stirred at 40 °C in the
presence of catalyst 1 (18 mg, 2.5 mol %) for 40 min. After
completion of the reaction (TLC), the mixture was diluted with
water (2 mL) and EtOAc (5 mL). The supernatant EtOAc
solution was decanted off. The aqueous layer/portion was
extracted with EtOAc (2 × 5 mL). The combined EtOAc
extracts were dried over Na2SO4. The organic phase was
concentrated in vacuo, and the purity of the residue was
checked by 1H NMR spectroscopy. Flash chromatography was
performed with eluant EtOAc/petroleum ether (1:9) unless
crude product was estimated to be greater than 95% pure.
Procedure for the Synthesis of Naphthylbenzamide
(Scheme 3). A mixture of aldehyde (1 mmol, 1 equiv), 2-
napthol (0.144 g, 1 mmol, 1 equiv), and benzamide (0.121 g, 1
mmol, 1 equiv) was stirred at 100 °C in the presence of catalyst
1 (9 mg, 2.5 mol %) for 1 h. After completion of the reaction
(monitored by TLC), the mixture was diluted with water (2
mL) and EtOAc (5 mL). The supernatant EtOAc solution was
decanted. The aqueous layer/portion was extracted with EtOAc
(2 × 5 mL). The combined EtOAc extracts were dried
(Na2SO4) and concentrated under vacuum. The crude reaction
mixture was recrystallized from EtOH to obtain the pure pale
yellow solid product.
Computational Methods. The ground-state and tran-
sition-state structures were optimized using the hybrid meta
exchange correlation functional M06-2X with 6-31+G(d) basis
set.29 Truhlar’s M06-2X functional has been successful for the
calculation of accurate thermochemistry, reaction barriers, and
noncovalent interactions present in the reactions involving
main group elements.30 Vibrational frequencies were calculated
to ensure that the located minima and transition states were
characterized by zero and one imaginary frequencies
respectively and to obtain the zero point energy (ZPE)
corrections to the energies. Relaxed intrinsic reaction
coordinate scans were performed to confirm that the located
saddle points do connect the reactants and products (or
intermediates).31 Basis set super position errors (BSSE) were
calculated using a counterpoise correction scheme, whenever
necessary.32 All structural calculations were done using standard
quantum chemistry programs.33
Experimental Details on Spectroscopy. The steady-state
absorption and emission spectra were recorded in a UV
spectrophotometer (Cary100, Varian) and FluoroMax-3
spectrofluorimeter, respectively. All experiments were carried
out at room temperature (298 K). We have taken 25 μM
concentration of catalyst 1 or 2 in each measurement.
■
ASSOCIATED CONTENT
* Supporting Information
S
The following files are available free of charge on the ACS
Publications website at DOI: 10.1021/cs5010695.
Experimental procedure, spectroscopic data, and scanned
spectra; Cartesian coordinates, vibrational frequencies,
and energies for all of the optimized geometries; and
additional results from the spectroscopy and calculations
(PDF).
4318
■
Research Article
AUTHOR INFORMATION
Corresponding Authors
*E-mail: swadhin.mandal@iiserkol.ac.in.
*E-mail: spad@iacs.res.in.
*E-mail: prasunchem@iiserkol.ac.in.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
S.K.M. thanks SERB grant No. SR/S1/IC-25/2012 (DST),
New Delhi for financial support. S.R.R. and A.P. thank CSIR
and SERB (DST) for RA fellowships, respectively. P.K.V.
thanks UGC for a research fellowship. A.D. is grateful to the
DST and INSA for partial support. The authors also
acknowledge the NMR facility of IISER-Kolkata.
■
REFERENCES
(1) Reid, D. H. Quart. Rev. 1965, 19, 274−302.
(2) Reid, D. H. Tetrahedron 1958, 3, 339−352.
(3) (a) Haddon, R. C. Nature 1975, 256, 394−396. (b) Haddon, R.
C. Aust. J. Chem. 1975, 28, 2343−2351.
(4) (a) Itkis, M. E.; Chi, X.; Cordes, A. W.; Haddon, R. C. Science
2002, 296, 1443−1445. (b) Pal, S. K.; Itkis, M. E.; Tham, F. S.; Reed,
R. W.; Oakley, R. T.; Haddon, R. C. Science 2005, 309, 281−284.
(c) Mandal, S. K.; Itkis, M. E.; Chi, X.; Samanta, S.; Lidsky, D.; Reed,
R. W.; Oakley, R. T.; Tham, F. S.; Haddon, R. C. J. Am. Chem. Soc.
2005, 127, 8185−8196. (d) Mandal, S. K.; Samanta, S.; Itkis, M. E.;
Jensen, D. W.; Reed, R. W.; Oakley, R. T.; Tham, F. S.; Donnadieu, B.;
Haddon, R. C. J. Am. Chem. Soc. 2006, 128, 1982−1994. (e) Bag, P.;
Pal, S. K.; Itkis, M. E.; Sarkar, S.; Tham, F. S.; Donnadieu, B.; Haddon,
R. C. J. Am. Chem. Soc. 2013, 135, 12936−12939.
(5) Miller, J. S. Angew. Chem., Int. Ed. 2003, 115, 27−29.
(6) (a) Morita, Y.; Suzuki, S.; Sato, K.; Takui, T. Nat. Chem. 2011, 3,
197−204. (b) Hicks, R. G. Nat. Chem. 2011, 3, 189−191.
(7) (a) Mukherjee, A.; Sen, T. K.; Ghorai, P. K.; Mandal, S. K.
Organometallics 2013, 32, 7213−7224. (b) Sen, T. K.; Mukherjee, A.;
Modak, A.; Mandal, S. K.; Koley, D. Dalton Trans. 2013, 42, 1893−
1904. (c) Mukherjee, A.; Sen, T. K.; Ghorai, P. K.; Mandal, S. K. Sci.
Rep. 2013, 3, 2821. (d) Mukherjee, A.; Sen, T. K.; Ghorai, P. K.;
Samuel, P. P.; Schulzke, C.; Mandal, S. K. Chem.Eur. J. 2012, 18,
10530−10545. (e) Sen, T. K.; Mukherjee, A.; Modak, A.; Ghorai, P.
K.; Kratzert, D.; Granitzka, M.; Stalke, D.; Mandal, S. K. Chem.Eur.
J. 2012, 18, 54−58.
(8) Raman, K. V.; Kamerbeek, A. M.; Mukherjee, A.; Atodiresei, N.;
Sen, T. K.; Lazic, P.; Caciuc, V.; Michel, R.; Stalke, D.; Mandal, S. K.;
Blugel, S.; Munzenberg, M.; Moodera, J. S. Nature 2013, 493, 509−
513.
(9) (a) Ishihara, K. Lewis acids in organic synthesis; Yamamoto, H.,
Ed.; Wiley-VCH: Weinheim, Germany, 2000; Vol 1, p 135. (b) Carey,
F. A.; Sundberg, R. J. Advanced Organic Chemistry: Part A: Structure
and Mechanisms, 5th ed.; Springer: Berlin, 2007.
(10) Sereda, O. Tabassum, S.Wilhelm, R. Lewis Acid Organocatalysts,
in Asymmetric Organocatalysis, Topic In Current Organic Chemistry; List,
B., Ed.; Springer: Berlin, 2009; pp 86−117.
(11) Bah, J.; Franzen, J. Chem.Eur. J. 2014, 20, 1066−1072.
(12) (a) Franz, K. D.; Martin, R. L. Tetrahedron 1978, 34, 2147−
2152. (b) Haddon, R. C.; Rayford, R.; Hirani, A. M. J. Org. Chem.
1981, 46, 4587−4588.
(13) (a) Corey, E. J.; Zhang, F. Angew. Chem., Int. Ed. 1999, 38,
1931−1934. (b) O’Brien, P. Angew. Chem., Int. Ed. 1999, 38, 326−329.
(c) Johannes, C. W.; Visser, M. S.; Weatherhead, G. S.; Hoveyda, A. H.
J. Am. Chem. Soc. 1998, 120, 8340−8347. (d) Ager, D. J.; Prakash, I.;
Schaad, D. R. Chem. Rev. 1996, 96, 835−876. (e) Li, G.; Chang, H. T.;
Sharpless, K. B. Angew. Chem., Int. Ed. 1996, 35, 451−454.
(14) (a) Posner, G. H.; Rogers, D. Z. J. Am. Chem. Soc. 1977, 99,
8208−8214. (b) Carre, M. C.; Houmounou, J. P.; Caubere, P.
Tetrahedron Lett. 1985, 26, 3107−3110. (c) Cossy, J.; Bellosta, V.;
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
ACS Catalysis
Hamoir, C.; Desmurs, J. R. Tetrahedron Lett. 2002, 43, 7083−7086.
(d) Papini, A.; Ricci, I.; Taddei, M.; Seconi, G.; Dembach, P. J. Chem.
Soc., Perkin Trans. 1984, 2261−2265. (e) Fujiwara, M.; Imada, M.;
Baba, A.; Matsuda, H. Tetrahedron Lett. 1989, 30, 739−740. (f) Fan, R.
H.; Hou, X. L. J. Org. Chem. 2003, 68, 726−730. (g) Chakraborti, A.
K.; Rudrawar, S.; Kondaskar, A. Org. Biomol. Chem. 2004, 2, 1277−
1280. (h) Bonollo, S.; Fringuelli, F.; Pizzo, F.; Vaccaro, L. Synlett 2008,
10, 1574−1578 DOI: 10.1055/s-2008-1078410.
(15) (a) Yamada, J. I.; Yumoto, M.; Yamamoto, Y. Tetrahedron Lett.
1989, 30, 4255−4258. (b) Yamamoto, Y.; Asao, N.; Meguro, M.;
Tsukade, N.; Nemoto, H.; Adayari, N.; Wilson, J. G.; Nakamura, H. J.
Chem. Soc. Chem. Commun. 1993, 1201−1203. (c) Chini, M.; Crotti,
P.; Favero, L.; Machhia, F.; Pineschi, M. Tetrahedron Lett. 1994, 35,
433−436. (d) Meguro, M.; Asao, N.; Yamamoto, Y. J. Chem. Soc.,
Perkin Trans. 1994, 1, 2597−2601. (e) Auge, J.; Leroy, F. Tetrahedron
Lett. 1996, 37, 7715−7716. (f) Sekar, G.; Singh, V. K. J. Org. Chem.
1999, 64, 287−289. (g) Sagava, S.; Abe, H.; Hase, Y.; Inaba, T. J. Org.
Chem. 1999, 64, 4962−4965. (h) Pujala, B.; Rana, S.; Chakraborti, A.
K. J. Org. Chem. 2011, 76, 8768−8780.
(16) (a) Bonnet-Delpon, D.; Begue, J.-P. J. Org. Chem. 2000, 65,
6749−6751. (b) Yadav, J. S.; Reddy, B. V. S.; Basak, A. K.; Narsaiah, A.
V. Tetrahedron Lett. 2003, 44, 1047−1050. (c) Surendra, K.;
Krishnaveni, N. S.; Rao, K. R. Synlett 2005, 506−510
DOI: 10.1055/s-2005-862359. (d) Azizi, N.; Saidi, M. R. Org. Lett.
2005, 7, 3649−3651.
(17) Sarkar, A.; Tham, F. S.; Haddon, R. C. J. Mater. Chem. 2011, 21,
1574−1581.
(18) Chi, X.; Itkis, M. E.; Patrick, B. O.; Barclay, T. M.; Reed, R. W.;
Oakley, R. T.; Cordes, A. W.; Haddon, R. C. J. Am. Chem. Soc. 1999,
121, 10395−10402.
(19) (a) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.;
Plenum Press: New York, 1999. (b) Egli, M.; Sarkhel, S. Acc. Chem.
Res. 2007, 40, 197−205. (c) Mooibroek, T. J.; Gamez, P.; Reedijk, J.
CrystEngComm. 2008, 10, 1501−1515.
(20) Sharma, U.; Togati, N.; Maji, A.; Manna, S.; Maiti, D. Angew.
Chem., Int. Ed. 2013, 52, 12669−12673.
(21) For examples of proton transfer mechanisms assisted by
counterion or other cooperative mechanisms, see (a) Dudnik, A. S.;
Xia, Y.; Li, Y.; Gevorgyan, V. J. Am. Chem. Soc. 2010, 132, 7645−7655.
(b) Xia, Y.; Dudnik, A. S.; Gevorgyan, V.; Li, Y. J. Am. Chem. Soc.
2008, 130, 6940−6941. (c) Shao, Z.; Zhang, H. Chem. Soc. Rev. 2009,
38, 2745−2755. (d) Masuda, Y.; Mori, Y.; Sakurai, K. J. Phys. Chem. A
2013, 117, 10576−10587. (e) Nijamudheen, A.; Jose, D.; Datta, A. J.
Phys. Chem. C 2011, 115, 2187−2195. (f) Patil, N. T.; Nijamudheen,
A.; Datta, A. J. Org. Chem. 2012, 77, 6179−6185.
(22) The calculations predict that the formation of 7 and 8 (in Figure
7) are endothermic processes and the reverse reactions to go back to
the weakly bound noncovalent adduct 6 should be very fast (ΔG‡ =
0.5 and 4.8 kcal/mol only). Therefore, in a mixture of 1 and 4a, the
much smaller populations of the structures 7 and 8 compared to 6
makes the detection of 7 and 8 by spectroscopy intractable.
(23) (a) Chakraborti, A. K.; Raha Roy, S.; Kumar, D.; Chopra, P.
Green Chem. 2008, 10, 1111−1118. (b) Wang, S. Y.; Ji, S.-J. Synth.
Commun. 2008, 38, 1291−1298. (c) Kamble, V. T.; Kadama, K. R.;
Joshia, N. S.; Muleya, D. B. Catal. Commun. 2007, 8, 498−502.
(d) Firouzabadi, H.; Iranpoor, N.; Jafari, A. A. J. Mol. Catal. A: Chem.
2006, 244, 168−172. (e) Deb, M. L.; Bhuyan, P. J. Tetrahedron Lett.
2006, 47, 1441−1443. (f) Wang, L.-M.; Han, J.-W.; Tian, H.; Sheng,
J.; Fan, Z.-Y.; Tang, X.-P. Synlett 2005, 2, 337−340. (g) Zhang, Z. H.;
Yin, L.; Wang, Y.-M. Synthesis 2005, 12, 1949−1954 DOI: 10.1055/s-
2005-869959. (h) Mo, L.-P.; Chuan, Z.; Zhang, Z.-H. Synth. Commun.
2005, 35, 1997−2004. (i) Singh, P. R.; Singh, D. U.; Samant, S. D.
Synth. Commun. 2005, 35, 2133−2138. (j) Ji, S.-J.; Wang, S.-Y.; Zhang,
Y.; Loh, T.-P. Tetrahedron 2004, 60, 2051−2055.
(24) (a) Casapullo, A.; Bifulco, G.; Bruno, I.; Riccio, R. J. Nat. Prod.
2000, 63, 447−451. (b) Garbe, T. R.; Kobayashi, M.; Shimizu, N.;
Takesue, N.; Ozawa, M.; Yukawa, H. J. Nat. Prod. 2000, 63, 596−598.
(c) Bao, B.; Sun, Q.; Yao, X.; Hong, J.; Lee, C. O.; Sim, C. J.; Im, K. S.;
Jung, J. H. J. Nat. Prod. 2005, 68, 711−715.
4319
Research Article
(25) Sundberg, R. J. The Chemistry of Indoles; Academic Press: New
York, 1996.
(26) (a) Corey, E. J.; Clark, D. A.; Goto, G.; Marfat, A.; Mioskowski,
C.; Sameulson, B.; Hammerson, S. J. Am. Chem. Soc. 1980, 102, 1436−
1439. (b) Luly, J. R.; Yi, N.; Soderquist, J.; Stein, H.; Cohen, J.; Perun,
T. J.; Plattner, J. J. J. Med. Chem. 1987, 30, 1609−1616. (c) Hashiyama,
T. Med. Res. Rev. 2000, 20, 485−501. For review: (d) Domling, A.
Chem. Rev. 2006, 106, 17−89. (e) Zhu, J.Multicomponent Reactions;
Bienayme, H., Ed.; Wiley-VCH: Weinheim, Germany, 2005. (f) Wipf,
P.; Stephenson, C. R. J.; Okumura, K. J. Am. Chem. Soc. 2003, 125,
14694−14695.
(27) (a) Das, B.; Laxminarayana, K.; Ravikanth, B.; Rao, B. R. J. Mol.
Catal. A: Chem. 2007, 261, 180−183. (b) Nagarapu, L.; Baseeruddin,
M.; Apuri, S.; Kantevari, S. Catal. Commun. 2007, 8, 1729−1734.
(c) Khodaei, M. M.; Khosropour, A. R.; Moghanian, H. Synlett 2006,
6, 916−920. (d) Patil, S. B.; Singh, P. R.; Surpur, M. P.; Samant, S. D.
Ultrason. Sonochem. 2007, 14, 515−518. (e) Lei, M.; Ma, L.; Hu, L.
Tetrahedron Lett. 2009, 50, 6393−6397.
(28) (a) Juaristi, E. Enantioselective Synthesis of β-Amino Acids; Wiley:
New York, 1997. (b) Knapp, S. Chem. Rev. 1995, 95, 1859−1876.
(c) Wang, Y.-F.; Izawa, T.; Kobayashi, S.; Ohno, M. J. Am. Chem. Soc.
1982, 104, 6465−6466. (d) Dingermann, T.; Steinhilber, D.; Folkers,
G.Front Matter in Molecular Biology in Medicinal Chemistry; Wiley-
VCH: Weinheim, Germany, 2004. (e) Shen, A. Y.; Tsai, C. T.; Chen,
C. L. Eur. J. Med. Chem. 1999, 34, 877−882. (f) Shen, A. Y.; Chen, C.
L.; Lin, C. I. Chin. J. Physiol. 1992, 35, 45−52.
(29) (a) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Theory
Comput. 2006, 2, 364−382. (b) Zhao, Y.; Truhlar, D. G. Theor. Chem.
Acc. 2008, 120, 215−241. (c) Hariharan, P. C.; Pople, J. A. Theor.
Chim. Acta 1973, 28, 213−222.
(30) (a) Jissy, A. K.; Datta, A. J. Phys. Chem. Lett. 2013, 4, 1018−
1022. (b) Nijamudheen, A.; Jose, D.; Shine, A.; Datta, A. J. Phys. Chem.
Lett. 2012, 3, 1493−1496. (c) Pieniazek, S. N.; Clemente, F. R.; Houk,
K. N. Angew. Chem., Int. Ed. 2008, 47, 7746−7749. (d) Wheeler, S. E.;
Houk, K. N.; Schleyer, P. V. R.; Allen, W. D. J. Am. Chem. Soc. 2009,
131, 2547−2560.
(31) (a) Fukui, K. Acc. Chem. Res. 1981, 14, 363−368. (b) Hratchian,
H. P.; Schlegel, H. B. J. Chem. Phys. 2004, 120, 9918−9924.
(32) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553−566.
(33) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; et al. Gaussian 09; Gaussian, Inc.: Wallingford,
CT, 2009.
dx.doi.org/10.1021/cs5010695 | ACS Catal. 2014, 4, 4307−4319
 Article

pubs.acs.org/JPCC

On the Nanoscopic Environment a Neutral Fluorophore Experiences

in Room Temperature Ionic Liquids
Anup Ghosh, Tanmay Chatterjee, Debjit Roy, Ananya Das, and Prasun K. Mandal*
Department of Chemical Sciences, Indian Institute of Science Education and Research (IISER)Kolkata, Mohanpur Campus,
West-Bengal 741252, India
*
S Supporting Information

ABSTRACT: The nanoscopic environment experienced by a fluorophore while

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Downloaded via S.N. BOSE NATL CTR BASIC SCIENCES on April 19, 2019 at 11:34:28 (UTC).

being dissolved in a complex nanoscopic structure of room temperature ionic liquid

(RTIL) is not fully known. Here we have explored resonance energy transfer (RET)
in order to have an insight in this direction in a noninvasive manner. We have varied
the alkyl chain length of the cation and size of the anion in order to probe the
nanoscopic environment within an RTIL. Other factors like different excitation
wavelengths and different concentrations of the probe has been varied in order to
understand the nature of the nanoscopic environment more accurately. Förster
formulation has been shown to explain fluorescence dynamical parameters related to
the RET phenomenon with reasonable accuracy. The rise time (varying from 3.00 to
4.00 ns) or rate constant of energy transfer has been shown to be independent of the
excitation wavelength and the concentration of the acceptor. However, the
magnitude of rise time varies with the nature of the cation (alkyl chain length)
and size of the anion. By employing Förster formulation, we have obtained the
donor−acceptor distance inside nanostructural RTIL media. The magnitude of donor−acceptor distance (varying from 30.35 to
36.48 Å) and hence the size of the nanostructural cage have been shown to be dependent on the size of the alkyl chain length of
the cation as well as the size of the anion of the RTILs. Quite expectedly the size of the nanoscopic cages does not depend on the
excitation wavelength or concentration of the acceptor. After detailed experiments and rigorous analysis, we have put forward a
model that can successfully explain the nanoscopic environment that a fluorophore experiences in an RTIL.

Room temperature ionic liquids (RTILs) have obtained a
significant place as a class of nonvolatile environmentally green
solvents; however, what kind of nanoenvironment a molecular
probe experiences while being dissolved in an RTIL is still not
known. RTILs are composed of mostly an organic cation and
an inorganic anion.1−7 Stupendous interest in RTILs has been
generated because of the interesting properties like vanishing
vapor pressure, low melting point, high thermal stability,
inflammability, recyclability, high ionic conductivity, etc.1−7 Out
of different RTILs, those based on imidazolium cationic moiety
are maximum explored. Thus, RTILs have made its place in
synthesis, (bio)catalysis, material sciences, and chemical
engineering, specially in energy related applications, such as
batteries, fuel cell photovoltaics, supercapacitors, etc.1−7 RTILs
are different from conventional solvents in many respects. For
example, unlike conventional solvents, RTILs possess structural
nanodomains both in solid and liquid states.8−24 From crystal
structure studies it has been shown that there exists
nanostructural heterogeneities in neat RTILs. These spatial
heterogeneities can range up to a few nanometers, and the
magnitude depends on the alkyl chain length of the imidazole
ring.8−10 Existence of π−π interaction and intermolecular
hydrogen bonding between imidazole rings have been
documented.8 It has also been reported that the hydrogen
atom attached to the carbon atom located in between two
nitrogen atoms of the imidazole ring can take part in hydrogen
bonding.11,12 Neutron scattering as well as simulation studies
have shown that RTILs do possess local mesoscopic
heterogeneities, i.e., nanoscale segregation of local structures
is prevailing, and the size of the nonpolar domain increases with
increase in alkyl chain length of the imidazole ring.13−24
RTILs are highly polar in nature, and the polarity is in the
range of short to long chain alcohol.25,26 However, there exists
report of much lower polarity of RTILs from the dielectric
constant point of view.27−29 From the optical spectroscopic
studies it has been shown that the absorption of RTILs is due
to imidazolium moiety, and by using different control
experiments, it has been shown beyond any scientific doubt
that the absorption as well as emission from RTILs are inherent
to RTILs and not due to any impurity.30,31 Proper rigorous
purification procedure needs to be followed in order to remove
the impurity. The purity of thus produced RTIL is confirmed
by NMR and MS measurements. As can be seen (see
Supporting Information), the RTILs produced are of very
high purity grade and do not contain any impurity (see
discussion part for detail). It has been reported that neat RTILs
exhibit excitation wavelength-dependent emission behavior, a
phenomenon commonly known as red edge effect (REE).30,31
Received: February 7, 2014
Published: February 11, 2014

© 2014 American Chemical Society 5051 dx.doi.org/10.1021/jp501342g | J. Phys. Chem. C 2014, 118, 5051−5057
The Journal of Physical Chemistry C Article

This unusual behavior has been assigned to the existence of the Chart 1. Chemical Structures of RTILs and the 4NBD Dye
different energetically associated forms and to the lack of
solvation/energy transfer between those species. It has also
been shown that some dipolar solutes do show excitation
wavelength-dependent emission while being dissolved in an
RTIL.32−34 Incomplete solvation or the existence of specific
interaction (like H-bonding) between the dipolar probe and
RTIL has been speculated to be responsible for the observed
REE for the dipolar solutes.32−34 Fluorescence dynamical
studies have shown that solvation34−37 in RTILs is biphasic in
nature, one component is typically of a few hundred
picoseconds and the other component is in the range of a
few nanoseconds. However, there exists different views
regarding the mechanism (especially regarding the faster
component) of solvation in different RTILs.34−37 Recently, it
has been shown that neat RTILs do exhibit excitation as well as
monitoring wavelength-dependent fluorescence decay behav-
ior.38 However, no rise time in picosecond or longer time
regime could be observed.38 This proves that no energy transfer
among different nanodomains take place in picosecond or
longer time scale. Thus, the reason for the observation of red
edge effect in neat RTILs could be understood. wavelengths, viz. 377 and 402 nm. In addition to those
Although a lot of photophysical measurements have been elaborate and rigorous experiments we have also performed
performed in RTILs, however, experiments based on resonance acceptor concentration-dependent RET dynamics in order to
energy transfer (RET) have not been explored in great verify whether RET dynamics depends on the concentration of
detail.39−42 Specially the dynamics of RET in neat RTILs has the acceptor. Results from all these experiments have been
not been explored at all. RET is a phenomenon in which the elaborated. On the basis of all these rigorous experimental
excitation energy of the donor is nonradiatively transferred to results a model has been put forward that could explain the
the acceptor provided some electrodynamic conditions are nanoscopic environment experienced by a neutral probe while
fulfilled and donor and acceptor are in close proximity (20 to being dissolved in an RTIL.
200 Å).43−45 The acceptor should be chosen in such a way that Steady-state absorption and emission: Absorption and
there exists significant spectral overlap between emission emission spectra of hmim[FAP] and that of neutral dye
spectrum of donor and the absorption spectrum of acceptor 4NBD have been shown in Supporting Information The
,and the extent of direct excitation of the acceptor is minimal. absorption and emission spectra of RTIL (donor) are well
The application of RET phenomenon is that it enables us to separated from that of the acceptor (4NBD). Absorbance at the
measure the distance between donor and acceptor in excitation wavelengths (377 and 402 nm) of the acceptor
heterogeneous matrices and environments like live cell, etc., (4NBD) is very low compared to that of the donor (RTIL)
in a noninvasive way. It has been shown recently41 that RTILs (see Supporting Information). This is a prerequisite condition
can engage in RET. We have explored RET to probe the in order to keep direct excitation of the acceptor at the minimal
distance between the imidazolium moiety (donor) and the level. In order to observe RET between any donor−acceptor
neutral dye (as acceptor), thus having an insight regarding the hybrid, the emission spectrum of the donor should have
nanoscopic environment that a neutral probe experiences in an spectral overlap with the absorption spectrum of the accept-
RTIL. or.43−45 It is evident from Figure 1 that there is significant
Hereby in this article we report the RET phenomenon in spectral overlap between the emission spectrum of RTIL
neat RTILs and have measured the dynamics of RET using
FAP, PF6, and BF4 anion based imidazolium RTIL as donor
and 4NBD (neutral dye) as an acceptor (see Chart 1). The
choice of the dye stems from the fact that the direct excitation
of the dye at 377 and 402 nm is minimal (see Supporting
Information).46 In order to check whether RET dynamics
depends on the alkyl chain length of the cation we have carried
out experiments in FAP anion based RTILs with three different
alkyl chain length of the cation, viz., ethyl, butyl, and hexyl. In
order to check whether the RET dynamics depend on the
nature of anion we have performed experiments in three RTILs
with varying anions, viz. FAP, BF4, and PF6 anion with same
cation (hmim). There exists a few literature reports that have
shown that different spectral and temporal fluorescence
behavior of neat RTILs or dissolved solutes do depend on
excitation wavelength or on the chain length of the cation or
the nature of the anion. In order to check whether the
dynamics of RET depends on the excitation wavelength, we Figure 1. Emission spectrum of hmim[PF6] (λex = 377 nm) (black, ---)
have performed the experiments for two different excitation and absorption spectrum of 4NBD (red, ).

5052 dx.doi.org/10.1021/jp501342g | J. Phys. Chem. C 2014, 118, 5051−5057

The Journal of Physical Chemistry C
(hmim[FAP], donor) and the absorption spectrum of 4NBD
(acceptor) (for the other four RTILs, see Supporting
Information). Thus, the necessary condition in order to
observe resonance energy transfer between the donor (RTIL)
and the acceptor (4NBD) is fulfilled.
On increasing the concentration of the acceptor (4NBD), we
have noticed that the emission of the donor (hmim[FAP])
decreases and that of the acceptor increases. This observation
has been depicted in Figure 2a (for the other four RTILs, see
Figure 2. (a) Variation of RTIL (hmim[FAP]) emission intensity (λex
= 377 nm) w.r.t. increasing concentration of 4NBD; (b) Stern−
Volmer plot showing a concentration variation change of fluorescence
of donor.
Supporting Information). As the fluorescence of donor gets
quenched with an increase in concentration of the acceptor, we
have plotted F0/F against acceptor concentration. This plot has
been shown in Figure 2b. As can be seen from Figure 2b, a
linear response of the Stern−Volmer plot44,45 indicates that
either static or dynamic quenching is taking place for the
experimental range of concentration (for the other four RTILs,
see Supporting Information).
In order to check whether static or dynamic quenching is
taking place, we have performed fluorescence decay measure-
ment of the donor using time-correlated single photon
counting (TCSPC) technique. The results of this experiment
have been shown in Figure 3 and the corresponding time
constants have been depicted in Table 1. As can be seen from
5053
Article
Figure 3. Fluorescence decay curve of RTIL (hmim[FAP]) without
and in the presence of acceptor (λex= 377 nm and λem = 450 nm)
(concentration of acceptor is depicted in the plot).
Figure 3 as well as from Table 1, fluorescence decay of the
donor (hmim[FAP], monitored at 450 nm, emission maximum
for 377 nm excitation) becomes faster as the concentration of
the acceptor is increased (for the other four RTILs, see
Supporting Information). The average lifetime has been found
to decrease by 12.50−18.75% when the concentration of the
acceptor is about 350 μM (from 4.82 to 4.22 ns (for 377 nm
excitation) and from 4.01 to 3.25 ns (for 402 nm excitation))
(for the other four RTILs, see Supporting Information). Thus,
the decrease of the average fluorescence lifetime of the donor
(hmim[FAP]) with increase in concentration of the acceptor
has been found to be dependent on the excitation wavelength.
Earlier from the linear Stern−Volmer plot we had concluded
that the quenching is either static or dynamic. The fact that the
fluorescence decay of the donor becomes faster in the presence
of increasing concentration of acceptor points to the fact that
dynamic quenching is taking place. Thus, so far we can
conclude that only one type of quenching is taking place and
that the nature of the quenching is dynamic.
Decrease of fluorescence lifetime of the donor with increase
in acceptor concentration tells that the quenching is dynamic in
nature; however, it does not conclusively prove that the
resonance energy transfer (RET) is happening. It has been
shown in the literature that it could only be the quenching of
the donor fluorescence without resonance energy transfer to
quencher/acceptor.45,47 To check whether RET is happening
or not we have measured fluorescence decay of the acceptor.
The excitation wavelength for the fluorescence decay of the
acceptor (following RET) has been judiciously chosen in such a
way that mostly donor will be excited at the excitation
wavelength, and the extent of direct excitation of the acceptor is
minimum. If RET is happening between RTIL and 4NBD, then
we should observe a rise in fluorescence decay profile of the
acceptor at shorter time scale followed by a decay at the longer
time scale.43−45 Fluorescence decay behavior of the acceptor
has been depicted in Figure 4. A clear rise (see inset of Figure
4) followed by a decay of acceptor proves that RET is indeed
happening from the RTIL donor to the 4NBD acceptor (for
the other four RTILs, see Supporting Information). The time
constant of the RET phenomenon and that of the acceptor
decay has been depicted in Table 2 (for the other four RTILs,
see Supporting Information).
dx.doi.org/10.1021/jp501342g | J. Phys. Chem. C 2014, 118, 5051−5057
The Journal of Physical Chemistry C Article

Table 1. Time Constants of Fluorescence Decay of the Donor (hmim[FAP]) in the Presence of Acceptor (4NBD) (Different
Concentrations)
λex (nm) λem (nm) concentration (μM) τ1 (ns) B1 τ2 (ns) B2 τ3 (ns) B3 ⟨τ⟩ (ns) χ2
377 450 0 0.26 11.56 2.63 50.12 8.95 38.33 4.82 1.03
176 0.78 14.59 2.56 45.93 8.26 39.48 4.55 1.18
220 0.26 6.21 1.97 44.57 7.08 49.22 4.37 1.14
264 0.28 7.23 1.97 47.49 7.30 45.27 4.26 1.15
352 0.41 10.07 2.13 47.24 7.45 42.69 4.22 1.10
402 450 0 0.26 9.08 2.25 50.57 7.07 40.35 4.01 1.13
176 0.27 11.06 2.09 52.79 6.89 36.14 3.62 1.15
220 0.23 11.19 1.94 50.87 6.43 37.93 3.45 1.19
264 0.24 11.28 1.85 49.50 6.34 37.44 3.31 1.14
352 0.21 11.05 1.78 49.53 5.95 39.42 3.25 1.15

the other four RTILs, see Supporting Information). Interest-

ingly, with increase in acceptor concentration (from ∼260 to
∼350 μM) the energy transfer time (rise time) does not change
at all for both excitation wavelengths. A similar phenomenon
has been observed for the other four RTILs (see Supporting
Information) This is a bit surprising because it is expected that
with increase in the concentration of the acceptor the distance
between the donor moiety and the acceptor moiety will
decrease, and hence, the time constant of the RET should
decrease. Instead here we have observed that energy transfer
time remains the same with an increase in concentration of the
acceptor.
We have tried to understand the RET phenomenon using
Förster formulation.43−45 According to Förster formulation, the
necessary condition for RET is that there should be overlap
between the emission spectrum of donor and absorption
spectrum of acceptor. The magnitude of spectral overlap can be
Figure 4. Fluorescence decay curve of the acceptor (λex = 377 nm and calculated using eq 1.
λem = 560 nm; concentration of acceptor = 352 μM). A clear rise has ∞
been shown in the inset. ∫0 FD(λ)εA (λ)λ 4 dλ
J (λ ) = ∞
Table 2. Time Constants of Fluorescence Decay of the ∫0 FD(λ)dλ (1)
Acceptor (4NBD) (Different Concentrations) in
hmim[FAP] where FD(λ) denotes the normalized fluorescence intensity of
the donor in the absence of acceptor. εA(λ) is the molar
λex λem concentration of rise time lifetime extinction coefficient of the acceptor. Förster distance R0 has
(nm) (nm) acceptor (μM) (ns) (ns) χ2
been calculated using eq 2.
377 560 264 3.97 7.20 1.01
352 3.97 7.40 1.08 R 0 = 0.211[k 2n 4Q DJ(λ)]1/6 (2)
402 560 264 4.00 7.30 1.12
352 4.00 7.30 1.04 where k2 is the orientation factor, its value generally varies from
0 to 4. However, here we take it as 2/3 for random orientation
of the donor and acceptor molecules. n is the refractive index of
As can be seen from Table 2, the time constant for RET is the medium, and QD is the quantum yield of the donor in the
around 3.95 to 4 ns (for both 377 and 402 nm excitation) (for absence of acceptor. The rate of the energy transfer as well as

Table 3. Magnitude of Different Physical Parameters Obtained Using Förster Formulation

λex (nm) RTILs τ0D (ns) ϕD J(λ) (1015) (M−1 cm−1 nm4) τrise (ns) R0 (Å) RDA (Å) KFRET (108 s−1) E
377 emim[FAP] 8.48 0.21 1.19 3.00 39.68 33.37 3.33 73.45
bmim[FAP] 8.61 0.24 1.03 3.73 39.54 34.39 2.68 69.79
hmim[FAP] 4.82 0.23 0.68 3.97 37.48 36.30 2.51 54.79
hmim[PF6] 4.33 0.08 1.02 4.00 32.73 31.93 2.50 58.13
hmim[BF4] 4.87 0.09 0.91 3.19 32.68 30.35 3.13 61.34
402 emim[FAP] 6.37 0.13 1.42 3.04 37.64 33.27 3.28 67.71
bmim[FAP] 7.30 0.19 1.15 3.72 38.75 34.63 2.68 66.25
hmim[FAP] 4.02 0.18 0.87 4.00 36.48 36.48 2.50 50.00
hmim[PF6] 5.13 0.08 1.15 4.00 33.33 31.80 2.50 57.14
hmim[BF4] 6.14 0.10 0.96 3.44 33.43 30.35 2.90 64.10

5054 dx.doi.org/10.1021/jp501342g | J. Phys. Chem. C 2014, 118, 5051−5057

The Journal of Physical Chemistry C
the distance between donor and acceptor have been calculated
(using the rise time) following eq 3.
1 ⎛ R0 ⎞
6
1 anion, but independent of the concentration. This means that
kFRET = A = 0⎜ ⎟ with an increase in acceptor concentration, the distance
τrise τD ⎝ RDA ⎠
where τArise is the rise time of the acceptor. τ0D is the lifetime of
the donor in the absence of any acceptor. R0 and RDA are the
Förster distance and the actual distance between the donor and
the acceptor. The magnitude of J(λ), R0, KFRET, RDA, etc., have
been depicted in Table 3.
As can be seen from Table 3, the magnitude of the spectral
overlap (and Förster radius R0) are dependent on the nature of
the RTIL. For RTILs having the same cation with different
anions, the value of the spectral overlap has varied. The value of
spectral overlap has varied for RTIL having same anion but
varied cation chain length. Moreover, the magnitude of RET
time (τrise) (or the rate constant of RET) has been found to
vary for different RTILs with the same cation and different
anions as well as for RTILs with the same anion but different
cations.
Using Förster formulation we have calculated the donor−
acceptor distance (RDA). The magnitude of RDA has been found
to be dependent on the nature of the RTIL. With the same
anion (FAP), the magnitude of RDA has been found to increase
as the chain length of the cation increases. When the chain
length of the cation increases from ethyl to hexyl, the
magnitude of RDA has been found to increase from 33.37 to
36.30 Å. Moreover, with the same cation hmim, as the size of
the anion increases from BF4 to PF6 to FAP the magnitude of
RDA has been found to increase from 30.54 to 36.30 Å. Thus,
the magnitude of RDA has been shown to be dependent both on
the nature of the cation (alkyl chain length) and size of the
anion of the RTIL.
The magnitude of the rise time or the rate constant of energy
transfer for a particular RTIL has been noted to be same for
402 nm excitation wavelength in comparison to 377 nm as
excitation wavelength. Different photophysical processes in a
particular RTIL have been shown to be dependent on the
excitation wavelength.32−34 However, a recent communication
has shown that dynamical fluorescence behavior of a
fluorophore (dissolved in an RTIL), whose emission is beyond
that of the RTIL, is excitation wavelength independent.38 In the
present report, we have 4NBD whose emission is beyond that
of the RTIL, and the dynamics of RET process has been found
to be excitation wavelength independent.
Quite interestingly for different excitation wavelengths for a
particular RTIL, the magnitude of donor−acceptor distance
(RDA) has been found to be exactly the same. No matter what
methodology we use, the donor−acceptor distance (RDA)
should be fixed and independent of the method we employ. For
different excitation wavelengths the magnitude of spectral
overlap (J(λ)) as well as Förster distance (R0) are different as
the emission of these RTILs is excitation wavelength
dependent. However, using Förster formulation, even for
different excitation wavelengths, the magnitude of donor−
acceptor distance (RDA) has been found to be exactly the same
for a particular RTIL. This observation points to the fact that
perhaps Förster formulation can explain the mechanism of RET
dynamics in RTILs.
However, what is still intriguing is that the RET dynamics
(rise time) is independent of concentration. With an increase in
concentration of the acceptor the distance between donor and
Article
acceptor is expected to decrease and hence the energy transfer
time. However, in the present case the energy transfer time has
been found to be dependent on the nature of the cation and the
(3)
between D−A pair remains same. In order to explain these
experimental observations we have put forward a model shown
in Scheme1.
Scheme 1. Nanocage Assembly with Varying Anions and the
Acceptor Inside the Nanocage
The concentration-dependent change in RET dynamics is
true for a homogeneous media. However, RTILs are known to
form nanocages, and the size of the nanocages has been shown
to extend up to a few nanometers depending on size of the alkyl
chain length of the cation.15,17,18,20,22 According to our model
(see Scheme 1), cation and anion of the RTIL remain side by
side, and the tail of the cation remains inside the cage.
Formation of different nanocages within RTIL has been
invoked earlier.15,17,18,20,22 It has been shown earlier that the
emission of the RTIL is because of the imidazolium moiety.30,31
As the tail of the cation remains inside the cage, thus the
distance between the donor (imidazolium ring) and the
acceptor increases with increase of the alkyl chain length.
Thus, for a particular anion, the distance between the donor
and the acceptor should increase with an increase in the chain
length. This is exactly what we have observed experimentally.
Moreover, as the size of the anion increases from BF4 to PF6 to
FAP (for the same cation, hmim) the bulkier the anion more
will be the cation pushed outward (Scheme 1), and thus, the
distance between the donor (imidazolium ring) and the
acceptor will increase with increase in size of the anion. This
is exactly what we have observed experimentally. We have put
forward a model (Scheme 1) to explain the experimentally
observed phenomenon.
According to our model (Scheme 1), 4NBD (acceptor) will
go to these nanocages, and since the concentration of neat
RTIL (∼M) is a few order higher than the concentration of the
dye (∼a few hundred μM), with increase in acceptor
concentration the dye will go to different nanocages, and as a
result, the distance between donor and acceptor will remain the
same for a particular RTIL-4NBD pair. Hence, the rise time or
rate constant of the RET phenomenon should remain the same
even with an increase in concentration of the acceptor. This is
what we have observed experimentally. Thus, rigorous
experiments provide evidence that nanocages do form in
RTIL surrounding the acceptor. Thus, the optical spectroscopic
results mentioned in the manuscript help us understand the
nanoscopic environment that the probe experiences inside
RTIL. This first report of dynamics of RET in neat RTIL as
donor and a neutral dye as acceptor will prompt similar other
experiments using cationic and anionic dyes. These experiments
are currently underway, and very soon the results of these
investigations will be reported.
5055 dx.doi.org/10.1021/jp501342g | J. Phys. Chem. C 2014, 118, 5051−5057
The Journal of Physical Chemistry C
It has been shown that fluorescence from RTIL is inherent
and not due to any impurity.30,31,41 From spectroscopic
characterization (NMR and MS) it has been shown clearly
that RTILs are sufficiently pure, and impurities, if present at all,
should be of submicromolar concentration. Researchers
working on RET are well aware that in order to observe
RET the distance between donor and acceptor should be below
200 Å. From a physical chemistry point of view assuming hard
sphere model one can calculate that in order to achieve such a
distance concentration of either donor or acceptor should be of
millimolar concentration in order to achieve such a distance.
Thus, we could prove that it is not the impurity rather the
inherent emission of imidazolium is involved in the RET
process.
In conclusion, we have shown that in the presence of an
acceptor whose absorption has spectral overlap with the
emission of the RTIL, the emission of the latter can be
quenched and the nature of the quenching has been shown to
be dynamic in nature. It has been shown that RET is indeed
happening between RTIL (donor) and 4NBD dye (acceptor),
and we have calculated necessary kinetic and dynamic time
constants. By employing Förster formulation, we have obtained
the donor−acceptor distance. The magnitude of donor−
acceptor distance has been shown to be dependent on the
alkyl chain length of the cation as well as the size of the anion of
the RTILs but not on the excitation wavelength or
concentration of the acceptor. The rise time or rate constant
of energy transfer has been shown to be independent of the
excitation wavelength and the concentration of the acceptor.
However, its magnitude varies with the nature of the cation
(alkyl chain length) and anion. By using different excitation
wavelengths and thus different spectral overlap and Förster
distance, we have calculated that the donor−acceptor distance
be the same for both excitation wavelengths. This observation
hints toward the fact that Förster formulation can perhaps
explain the mechanism and hence can explain the experimental
observables in the RET phenomenon. In the end, we have put
forward a model that can explain the experimental observables
with reasonable accuracy.
■
*
ASSOCIATED CONTENT
S Supporting Information
Experimental details, characterization of synthesized RTILs,
steady-state absorption, fluorescence, time-resolved fluores-
cence decay, time constants of fluorescence decay, Stern−
Volmer plot, etc. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*(P.K.M.) E-mail: prasunchem@iiserkol.ac.in.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
P.K.M. thanks IISER-Kolkata for financial help and instrumen-
tal facilities. Support from the Fast-Track Project (SR/FT/CS-
52/2011) of DST-India is gratefully acknowledged. A.G. thanks
UGC, and T.C., D.R., and A.D. thank CSIR for their respective
Fellowships.
5056
■
Article
REFERENCES
(1) Welton, T. Room-Temperature Ionic Liquids. Solvents for
Synthesis and catalysis. Chem. Rev. 1999, 99, 2071−2083.
(2) Wasserscheid, P.; Keim, W. Ionic Liquids: New Solutions for
Transitions Metal Catalysts. Angew. Chem., Int. Ed. 2000, 39, 3772−
3789.
(3) Dupont, J.; Souza, R. F. D.; Suarez, P. A. Z. Liquid (Molten Salt)
Phase Organometallic Catalysis. Chem. Rev. 2002, 102, 3667−3692.
(4) Rogers, R. D.; Seddon, K. R. Ionic Liquids: Solvents of the
Future. Science 2003, 302, 792−793.
(5) Seddon, K. R. Ionic Liquids: A Taste of the Future. Nat. Mater.
2003, 2, 363−365.
(6) WeinGaertner, H. Understanding Ionic Liquids at the Molecular
Level: Facts, Problems and Controversies. Angew. Chem., Int. Ed. 2008,
47, 654−670.
(7) Hallet, J. P.; Welton, T. Room-Temperature Ionic Liquids:
Solvents for Synthesis and catalysis. Chem. Rev. 2011, 111, 3508−3576.
(8) Holbrey, J. D.; Reichert, W. M.; Rogers, R. D. Crystal Structures
of Imidazolium Bis(trifluoromethanesulfonyl)imide Ionic Liquid Salts:
The First Organic Salt with a cis-TFSI Anion Conformation. Dalton
Trans. 2004, 2267−2271.
(9) Chen, S.-H.; Yang, F.-R.; Wang, M.-T.; Wang, N.-N. Synthesis,
Characterization, and Crystal Structure of Several Novel Acidic Ionic
Liquids Based on the Corresponding 1-Alkylbenzimidazole with
Tetrafluoroboric Acid. C. R. Chim. 2010, 13, 1391−1396.
(10) Henderson, W. A.; Fylstra, P.; Long, H. C. D.; Trulove, P. C.;
Parsons, S. Crystal Structure of the Ionic Liquid EtNH3NO3: Insights
into the Thermal Phase Behavior of Protic Ionic Liquids. Phys. Chem.
Chem. Phys. 2012, 14, 16041−16046.
(11) Mele, A.; Tran, C. D.; Lacerda, S. H. D. P. The Structure of a
Room-Temperature Ionic Liquid with and without Trace Amounts of
Water: The Role of C−H···O and C−H···F Interactions in 1-n-Butyl-
3-methylimidazolium Tetrafluoroborate. Angew. Chem., Int. Ed. 2003,
42, 4364−4366.
(12) Mele, A.; Romano, G.; Giannone, M.; Ragg, E.; Fronza, G.;
Raos, G.; Marcon, V. The Local Structure of Ionic Liquids: Cation−
Cation NOE Interactions and Internuclear Distances in Neat
[BMIM][BF4] and [BDMIM][BF4]. Angew. Chem., Int. Ed. 2006,
45, 1123−1126.
(13) Hardacre, C.; Holbrey, J. D.; Mullan, C. L.; Nieuwenhuyzen, M.;
Youngs, T. G. A.; Bowron, D. T. Liquid Structure of the Ionic Liquid,
1-Methyl-4-cyanopyridinium Bis{(trifluoromethyl)sulfonyl}imide De-
termined from Neutron Scattering and Molecular Dynamics
Simulations. J. Phys. Chem. B 2008, 112, 8049−8056.
(14) Hardacre, C.; Holbrey, J. D.; Mullan, C. L.; Youngs, T. G. A.;
Bowron, D. T. Small Angle Neutron Scattering from 1-Alkyl-3-
methylimidazolium Hexafluorophosphate Ionic Liquids ([Cnmim]-
[PF6], n = 4, 6, and 8). J. Chem. Phys. 2010, 133, 074510.
(15) Triolo, A.; Russina, O.; Bleif, H.-J.; Di Cola, E. Nanoscale
Segregation in Room Temperature Ionic Liquids. J. Phys. Chem. B
2007, 111, 4641−4644.
(16) Urahata, S. M.; Ribeiro, M. C. C. Structure of Ionic Liquids of 1-
Alkyl-3-methylimidazolium Cations: A Systematic Computer Simu-
lation Study. J. Chem. Phys. 2004, 120, 1855−1863.
(17) Wang, Y.; Voth, G. A. Unique Spatial Heterogeneity in Ionic
Liquids. J. Am. Chem. Soc. 2005, 127, 12192−12193.
(18) Lopes, J. N. C.; Gomes, M. F. C.; Padua, A. A. H. Nonpolar,
Polar, and Associating Solutes in Ionic Liquids. J. Phys. Chem. B 2006,
110, 16816−16818.
(19) Borodin, O.; Smith, G. D. Structure and Dynamics of N-Methyl-
N-propylpyrrolidinium Bis(trifluoromethanesulfonyl)imide Ionic
Liquid from Molecular Dynamics Simulations. J. Phys. Chem. B
2006, 110, 11481−11490.
(20) Jiang, W.; Wang, Y.; Voth, G. A. Molecular Dynamics
Simulation of Nanostructural Organization in Ionic Liquid/Water. J.
Phys. Chem. B 2007, 111, 4812−4818.
(21) Hu, Z.; Margulis, C. J. Room-Temperature Ionic Liquids: Slow
Dynamics, Viscosity, and the Red Edge Effect. Acc. Chem. Res. 2007,
40, 1097−1105.
dx.doi.org/10.1021/jp501342g | J. Phys. Chem. C 2014, 118, 5051−5057
The Journal of Physical Chemistry C Article

(22) Raabea, G.; Kö hler, J. Thermodynamical and Structural Showing Emissions by Excitation at Wide Wavelength Areas. Chem.
Properties of Imidazolium Based Ionic Liquids from Molecular Commun. 2010, 46, 6359−6361.
Simulations. J. Chem. Phys. 2008, 128, 154509. (42) Rao, V. G.; Mandal, S.; Ghosh, S.; Banerjee, C.; Sarkar, N. Study
(23) Andrade, J. D.; Böes, E. S.; Stassen, H. Liquid-Phase Structure of of Fluorescence Resonance Energy Transfer in Zweitter Ionic Micelle:
Dialkylimidazolium Ionic Liquids from Computer Simulations. J. Phys. Ionic Liquids-Induced Changes in FRET Parameters. J. Phys. Chem. B
Chem. B 2008, 112, 8966−8974. 2012, 121, 12021−12029.
(24) Yan, T.; Wang, Y.; Knox, C. On the Structure of Ionic Liquids: (43) Förster, T. Intermolecular Energy Migration and Fluorescence
Comparison between Electronically Polarizable and Nonpolarizable (Trans1 RS Knox). Ann. Phys. 1948, 2, 55−75.
Models. J. Phys. Chem. B 2010, 114, 6905−6921. (44) Valeur, B. Molecular Fluorescence Principles and Applications;
(25) Aki, S. N. V. K.; Brennecke, J. F.; Samanta, A. How Polar Are Wiley-VCH Verlag GmbH: Weinheim, Germany, 2002.
Room-Temperature Ionic Liquids? Chem. Commun. 2001, 413−414. (45) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.;
(26) Mandal, P. K.; Samanta, A. Fluorescence Studies in a Plenum Press: New York, 1999.
Pyrrolidinium Ionic Liquid: Polarity of the Medium and Solvation (46) Saha, S.; Samanta, A. Photophysical and Dynamic NMR Studies
Dynamics. J. Phys. Chem. B 2005, 109, 15172−15177. on 4-Amino-7-nitro-2-oxa-1,3-diazole Derivatives: Elucidation of the
(27) Wakai, C.; Oleinikova, A.; Weingaertner, H. How Polar Are Nonradiative Deactivation Pathway. J. Phys. Chem. A 1998, 102,
Ionic Liquids? Determination of the Static Dielectric Constant of an 7903−7912.
Imidazolium-Based Ionic Liquid by Microwave Spectroscopy. J. Phys. (47) Santhosh, K.; Patra, S.; Soumya, S.; Khara, D. C.; Samanta, A.
Chem. Lett. 2005, 109, 17028−17030. Fluorescence Quenching of CdS Quantum Dots by 4-Azetidinyl-7-
(28) Bright, F. V.; Baker, G. A. Comment on “How Polar Are Ionic nitrobenz-2-oxa-1,3-diazole: A Mechanistic Study. ChemPhysChem
Liquids? Determination of the Static Dielectric Constant of an 2011, 12, 2735−2741.
Imidazolium-Based Ionic Liquid by Microwave Dielectric Spectrosco-
py. J. Phys. Chem. B 2006, 110, 5822−5823.
(29) Wakai, C.; Oleinikova, A.; Weingaertner, H. How Polar Are
Ionic Liquids? Determination of the Static Dielectric Constant of an
Imidazolium-Based Ionic Liquid by Microwave Spectroscopy. Reply. J.
Phys. Chem. B 2006, 110, 5824−5824.
(30) Paul, A.; Mandal, P. K.; Samanta, A. How Transparent Are the
Imidazolium Ionic Liquids? A Case Study with 1-Methyl-3-
butylimidazolium Hexafluorophosphate, [bmim][PF6]. Chem. Phys.
Lett. 2005, 402, 375−379.
(31) Paul, A.; Mandal, P. K.; Samanta, A. On the Optical Properties
of the Imidazolium Ionic Liquids. J. Phys. Chem. B 2005, 109, 9148−
9153.
(32) Mandal, P. K.; Sarkar, M.; Samanta, A. Excitation Wavelength
Dependent Fluorescence Behavior of the Room Temperature Ionic
Liquids and Dissolved Dipolar Solutes. J. Phys. Chem. A 2004, 108,
9048−9053.
(33) Mandal, P. K.; Paul, A.; Samanta, A. Excitation-Wavelength-
Dependent Fluorescence Behavior of Some Dipolar Molecules in
Room-Temperature Ionic Liquids. J. Photochem. Photobiol. A: Chem.
2006, 182, 113−120.
(34) Samanta, A. Dynamic Stokes Shift and Excitation Wavelength
Dependent Fluorescence of Dipolar Molecules in Room Temperature
Ionic Liquids. J. Phys. Chem. B 2006, 110, 13704−13716.
(35) Mandal, P. K.; Saha, S.; Karmakar, R.; Samanta, A. Solvation
Dynamics in Room Temperature Ionic Liquids: Dynamic Stokes Shift
Studies of Fluorescence of Dipolar Molecules. Curr. Sci. 2006, 90,
301−310.
(36) Samanta, A. Solvation Dynamics in Ionic Liquids: What We
Have Learned from the Dynamic Fluorescence Stokes Shift Studies. J.
Phys. Chem. Lett. 2010, 1, 1557−1562.
(37) Zhang, X.-X.; Liang, M.; Ernsting, N. P. Complete Solvation
Response of Coumarin 153 in Ionic Liquids. J. Phys. Chem. B 2013,
117, 4291−4304.
(38) Ghosh, A.; Chatterjee, T.; Mandal, P. K. On the Heterogeneity
of Fluorescence Lifetime of Room Temperature Ionic Liquids: Onset
of a Journey for Exploring Red Emitting Dyes. Chem. Commun. 2012,
48, 6250−6252.
(39) Adhikari, A.; Das, D. K.; Sasmal, D. K.; Bhattacharyya, K.
Ultrafast FRET in a Room Temperature Ionic Liquid Microemulsion:
A Femtosecond Excitation Wavelength Dependent Study. J. Phys.
Chem. B 2009, 113, 3737−3743.
(40) Das, D. K.; Das, A. K.; Mondal, T.; Mandal, A. K.;
Bhattacharyya, K. Ultrafast FRET in Ionic Liquid-P123 Mixed
Micelles: Region and Counterion Dependence. J. Phys. Chem. B
2010, 114, 13159−13166.
(41) Izawa, H.; Wakizono, S.; Kadokawa, J. Fluorescence Resonance-
Energy-Transfer in Systems of Rhodamine 6g with Ionic Liquid

5057 dx.doi.org/10.1021/jp501342g | J. Phys. Chem. C 2014, 118, 5051−5057

 RSC Advances
View Article Online
COMMUNICATION View Journal | View Issue

Chemical tweaking of a non-fluorescent GFP

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

chromophore to a highly fluorescent coumarinic

Open Access Article. Published on 14 October 2013. Downloaded on 27/02/2017 10:12:14.

Cite this: RSC Adv., 2013, 3, 24021

fluorophore: application towards photo-uncaging and
Received 31st July 2013 stem cell imaging†
Accepted 10th October 2013
Tanmay Chatterjee, Debjit Roy, Ananya Das, Anup Ghosh, Partha Pratim Bag
DOI: 10.1039/c3ra44034f
and Prasun K. Mandal*
www.rsc.org/advances

Three step chemical tweaking of nonfluorescent parent GFP chro- important towards bioimaging.12 There are quite a few number
mophore has yielded a three hundred times as bright coumarinic of photoactivable compounds, however, there is only one report
fluorophore. A fluorogenic compound has been prepared which so far where a photoactivable compound has been made with
after photo-uncaging yields the same highly bright fluorophore that GFP chromophore analogue.13
could be used for live stem cell imaging. In this manuscript we describe how three step chemical
tweaking of non-uorescent parent GFP chromophore (p-HBDI)
could lead to a highly uorescent coumarinic uorophore
The quest to enhance the uorescence quantum yield of (Scheme 1). Three modications that has been incorporated are
nonuorescent GFP (green uorescent protein)1,2 chromo- (i) p-hydroxy has been replaced by o-hydroxy, (ii) N,N-diethyl-
phores (p-HBDI, see Scheme 1) has engaged many researchers amino group has been introduced at the p-position, and (iii) sp3
from biology and chemistry for more than a decade. Strong nitrogen in imidazole ring has been replaced by oxygen. In this
deviation of uorescence quantum yield (f) of p-HBDI (f < direction we have made four compounds: 4-(2-hydroxy-4-N,
0.001) from that of wild type GFP (f
0.8) has been noted in N-diethylamino-benzylidene)-1,2-dimethyl-1H-imidazol-5(4H)-
solution.3,4 Different mechanisms have been put forward as one (OHIM), 4-(2-methoxy-4-N,N-diethylamino-benzylidene)-1,
plausible reasons for the low uorescence quantum yield of GFP
chromophore.5,6 Several successful attempts7,8 to increase f of
GFP chromophore by restricting the torsional motion include
mimicking GFP barrel encapsulation in human serum albumin,
in octaacid cavitand, and using truncated or split GFP tech-
nique, and selective complexation with RNA aptamers,
complexation with metal salts (such as Zn2+) or BF2 moiety.
However, concerns have been raised regarding their usage due
to lability of metal–BF2 complexes in solution.9 Different
chemical modications/substitutions in benzene and imidazole
ring of p-HBDI have been tried but so far a maximum of only

10% uorescence quantum yield could be achieved in GFP
chromophore analogues.10 Thus, the quest for chemically
modied GFP chromophore analogue with enhanced uores-
cence quantum yield is still very much active.
Photoactivation process provides spatial and temporal
control over the release of desired chemical11 and thus photo-
activable compounds and uorescent proteins are quite

Department of Chemical Sciences, Indian Institute of Science Education and Research

(IISER)-Kolkata, Mohanpur Campus, West-Bengal, 741252, India. E-mail:
prasunchem@iiserkol.ac.in Scheme 1 (a) Three step chemical tweaking (in red) of p-HBDI. (b) Chemical
† Electronic supplementary information (ESI) available. See DOI: structure and synthesis of all four dyes. (c) Possible pathway for conversion of
10.1039/c3ra44034f OHBO to cOHBO.

This journal is ª The Royal Society of Chemistry 2013 RSC Adv., 2013, 3, 24021–24024 | 24021

RSC Advances
Table 1 Photophysical properties of OHIM and cOHBO
1 1
Dye Solvent lmaxa (nm) 3 (M cm ) f s
OHIM Toluene 484(448) 29 075 2.2  103 0.76, 32.8 ps
Methanol 505(451) 30 580 1.4  103 0.73, 13.9 ps
cOHBO Toluene 444(377) 19 030 0.90 3.27 ns
Methanol 485(392) 19 000 0.70 3.88 ns
a
Emission maximum, absorption maximum (nm) in parentheses.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.
Open Access Article. Published on 14 October 2013. Downloaded on 27/02/2017 10:12:14.
2-dimethyl-1H-imidazol-5(4H)-one (OMIM) and two of their
respective point atom mutation analogues (OHBO and OMBO
respectively) where the nitrogen atom at 1-position have been
replaced by oxygen atom (see Scheme 1b) (for synthesis and
characterisation see ESI†). Absorption and emission spectra of
OHIM and OHBO in different solvents are shown in ESI† and
the photophysical parameters are depicted in Table 1.
(For OMIM and OMBO see ESI†).
Surprisingly, we have noticed that f as well as the uores-
cence lifetime (s) of OHBO are drastically different from that of
OHIM (or other two derivatives). To the best of our knowledge
this is the rst example of chemical modication of GFP chro-
mophore leading to a highly bright uorescent dye with near
unity uorescence quantum yield, i.e. more than 300 fold
uorescence enhancement from the GFP chromophore in
solution. It was observed that the uorescence lifetime of
OHIM, OMIM, and OMBO are in femtosecond or picosecond
time scale, whereas uorescence lifetime of OHBO is in
nanosecond time scale. Thus an enhancement of 100 to 4000
times of s is noted for OHBO.
Such an enhancement of uorescence quantum yield and
uorescence lifetime made us rethink what could be the
possible reason and hence we had to look deeper into the
spectroscopic characterisation. Based on careful analysis we
propose here that although OHBO is produced (previous
literature reports suggest that hydrolysis of methoxy group with
BBr3 yields hydroxy group)10b in the reaction but it undergoes an
intramolecular rearrangement and hence formation of a
coumarinic structure is a possibility (Scheme 1c). So, OHBO
plausibly remains not in hydroxyl derivative state but in
coumarinic structure (cOHBO). This proposition has been
supported by IR, and NMR spectroscopic results. For
Fig. 1 Molecular structure of cOHBO and ONBYOHBO. Thermal ellipsoids were
drawn at the 50% probability level.
24022 | RSC Adv., 2013, 3, 24021–24024
View Article Online
Communication
rearrangement, presence of –OH group at ortho position and
oxygen in oxazole ring are necessary conditions. For OHIM,
OMIM or OMBO these conditions are not fullled and thus no
intramolecular rearrangement is possible. Thus, uorescence
properties of OHIM, OMIM, and OMBO are similar to each
other but quite different from that of cOHBO. Absorption
maximum of cOHBO matches with that of many other
coumarinic probes that absorb at around 375–400 nm.14 The
best proof in support of coumarinic structure stems from the
single crystal X-ray structure of cOHBO (see Fig. 1). (For lattice
parameters etc. see ESI†).
It is quite interesting to note that at room temperature
intramolecular rearrangement (Scheme 1c) is spontaneous
(i.e. not externally photoinduced or thermo-induced). Thus, it is
evident that three step tweaking especially the last step of point
atom mutation in OHIM (N being replaced by O) followed by
rearrangement and thus forming cOHBO is responsible for the
much improved uorescence properties. Most of the bright dyes
available are either coumarin, rhodamine based. Hereby we
have converted nonuorescent GFP chromophore to uorescent
coumarinic dye. This methodology can be used for making
many other nonuorescent uorescent protein chromophore to
a bright uorescent one. Work in this direction is currently
underway.
Photoactivable uorophores and uorescent proteins11,12 are
quite important towards understanding the cellular processes
using different imaging techniques. Exploiting the idea of
spontaneous rearrangement of OHBO leading to cOHBO
(see Scheme 1c) and hence highly bright uorescence; we have
introduced an o-nitrobenzoyl group (see Scheme 2) in OHBO.
Details of synthesis, and characterization of ONBYOHBO is
given in ESI.† This photoactivable compound could be cleaved
by light (at
370 nm) leading to formation of highly uorescent
cOHBO.
The fact that indeed cOHBO is formed because of photo-
cleavage of ONBYOHBO (see Fig. 1) is conrmed by uorescence
spectroscopy. Steady state emission spectrum of cOHBO
matches quite well with that of the photo-irradiated product
(see ESI†). A better proof is shown in Fig. 2. Fluorescence decay
curve and the decay parameters of cOHBO (see ESI†) and that of
the photoirradiated product match quite well.
This conclusively proves that cOHBO is formed from
ONBYOHBO. The decay pattern of cOHBO is clearly different
from photoactivable ONBYOHBO, thus, the former could be
spectroscopically distinguished easily from the latter. Earlier
Scheme 2 Synthesis of ONBYOHBO and its photocleavage to cOHBO.
This journal is ª The Royal Society of Chemistry 2013

Communication
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.
Open Access Article. Published on 14 October 2013. Downloaded on 27/02/2017 10:12:14.
Fig. 2 Fluorescence decay of ONBYOHBO, irradiated photoproduct, and cOHBO
in DMF. lex ¼ 377 nm.
reports of introduction of o-nitrobenzoyl group to GFP chro-
mophore yielded nonuorescent product aer photocleavage.13
Our approach has yielded more than 300 times brighter uo-
rescent product (cOHBO) than that in the reported literature.
Imaging of live cancerous cells like HeLa, U2OS with uo-
rescent dyes is quite common. However, non-invasive in vivo
stem cell imaging15 is more modern and quite important
towards regenerative medicinal therapy towards curing incur-
able diseases like cancer, Alzheimer disease, Parkinson disease
etc. Fluorescence imaging of stem cells have so far been done
using uorescent proteins, quantum dots etc. Application of
uorescent dyes for live stem cell imaging is at its nascent stage.
Thus we thought of using cOHBO for this purpose. To our
delightful surprise we could image live stem cells using cOHBO
(Fig. 3). (Details of imaging can be found in ESI†). We strongly
believe that it is an important step towards using uorescent
dyes for stem cell imaging. In recent future many more dyes will
be tried to improve the applicability of uorescent dyes towards
stem cell imaging.
To conclude, we have shown that by three point chemical
tweaking parent GFP chromophore can be converted to
coumarinic uorophore with more than three hundred times
enhanced uorescence quantum yield and lifetime. This
coumarinic probe could be successfully employed in live cell
imaging. Employing this idea a photocleavable compound has
been made which aer photocleavage yields same coumarinic
uorophore. Employing our methodology many more nonu-
orescent protein chromophores can be converted to highly
uorescent uorophores. Moreover, using the photoactivation
strategy highly bright uorophores emitting at different wave-
length ranging from blue to red can be prepared.
Fig. 3 Live stem cell imaging with cOHBO. Left one is fluorescence, middle one is
DIC and the right one is the merged image.
This journal is ª The Royal Society of Chemistry 2013
View Article Online
RSC Advances
PKM thanks IISER-Kolkata for instrumental facilities and
nancial help. Support from the Fast-Track Project (SR/FT/
CS-52/2011) of DST-India is gratefully acknowledged. TC, DR,
AD, PPB thank CSIR, and AG thanks UGC for their fellowship.
PKM acknowledges Professor P. Ramamurthy (NCUFP),
Malancha Ta (IISER-K) for Femtosecond decay measurement,
and Stem cell imaging respectively.
Notes and references
1 (a) R. Y. Tsien, Annu. Rev. Biochem., 1998, 67, 509–544; (b)
Green uorescent protein: properties, application, and
protocols, ed. M. Chale and S. R. Kein, Wiley Interscience,
New Jersey, 2nd edn, 2006.
2 (a) M. Zimmer, Chem. Rev., 2002, 102, 759–781; (b) R. N. Day
and M. W. Davidson, Chem. Soc. Rev., 2009, 38, 2887–2921;
(c) M. Zimmer, Chem. Soc. Rev., 2009, 38, 2823–2832.
3 K. Y. Chen, Y. M. Cheng, C. H. Lai, C. C. Hsu, M. L. Ho,
G. H. Lee and P. T. Chou, J. Am. Chem. Soc., 2007, 129,
4534–4535.
4 (a) M. Vengris, I. H. M. van Stokkum, X. He, A. F. Bell,
P. J. Tonge, R. van Grondelle and D. S. Larsen, J. Phys.
Chem. A, 2004, 108, 4587–4598; (b) S. R. Meech, Chem. Rev.,
2009, 38, 2922–2934.
5 (a) W. Weber, V. Helms, J. A. McCammon and
P. W. Langhoff, Proc. Natl. Acad. Sci. U. S. A., 1999, 96,
6177–6182; (b) P. Altoe, F. Bernardi, M. Garavelli,
G. Orlandi and F. Negri, J. Am. Chem. Soc., 2005, 127, 3952–
3963; (c) A. Usman, O. F. Mohammed, E. T. J. Nibbering,
J. Dong, K. M. Solntsev and L. M. Tolbert, J. Am. Chem.
Soc., 2005, 127, 11214–11215; (d) S. S. Stavrov,
K. M. Solntsev, L. M. Tolbert and D. J. Huppert, J. Am.
Chem. Soc., 2006, 128, 1540–1546; (e) J. Dong,
K. M. Solntsev, O. Poizat and L. M. Tolbert, J. Am. Chem.
Soc., 2007, 129, 10084–10085; (f) J. S. Yang, G. J. Huang,
Y. H. Liu and S. M. Peng, Chem. Commun., 2008, 1344–
1346; (g) S. Olsen and S. C. Smith, J. Am. Chem. Soc., 2008,
130, 8677–8689; (h) A. Baldridge, S. R. Samanta, N. Jayaraj,
V. Ramamurthy and L. M. Tolbert, J. Am. Chem. Soc., 2010,
132, 1498–1499; (i) A. Baldridge, S. R. Samanta, N. Jayaraj,
V. Ramamurthy and L. M. Tolbert, J. Am. Chem. Soc., 2011,
133, 712–715.
6 (a) N. M. Webber, K. L. Litvinenko and S. R. Meech, J. Phys.
Chem. B, 2001, 105, 8036–8039; (b) K. L. Litvinenko,
N. M. Weber and S. R. Meech, J. Phys. Chem. A, 2003,
107, 2616–2623; (c) D. Mandal, T. Tahara and
S. R. Meech, J. Phys. Chem. B, 2004, 108, 1102–1108; (d)
S. Raq, B. K. Rajbongshi, N. N. Nair, P. Sen and
P. Ramanathan, J. Phys. Chem. A, 2011, 115, 13733–
13742.
7 (a) L. Wu and K. Burgess, J. Am. Chem. Soc., 2008, 130, 4089–
4096; (b) K. P. Kent, L. M. Oltrogge and S. G. Boxer, J. Am.
Chem. Soc., 2009, 131, 15988–15989; (c) A. Baldridge,
S. R. Samanta, N. Jayaraj, V. Ramamurthy and
L. M. Tolbert, J. Am. Chem. Soc., 2010, 132, 1498–1499; (d)
A. Baldridge, K. M. Solntsev, C. Song, T. Tanioka,
J. Kowalik, K. Hardcastle and L. M. Tolbert, Chem.
RSC Adv., 2013, 3, 24021–24024 | 24023

RSC Advances
Commun., 2010, 46, 5686–5688; (e) K. P. Kent and S. G. Boxer,
J. Am. Chem. Soc., 2011, 133, 4046–4052; (f) K. Do and
S. G. Boxer, J. Am. Chem. Soc., 2011, 133, 18078–18081; (g)
A. Baldridge, S. Feng, Y. T. Chang and L. M. Tolbert, ACS
Comb. Sci., 2011, 13, 214–217; (h) J. S. Paige, K. Y. Wu and
S. R. Jaffrey, Science, 2011, 333, 642–646.
8 M. S. Baranov, K. A. Lukyanov, A. O. Borissova, J. Shamir,
D. Kosenkov, L. V. Slipchenko, L. M. Tolbert,
I. V. Yampolsky and K. M. Solntsev, J. Am. Chem. Soc.,
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.
2012, 134, 6025–6032.
9 J. Kang, G. Zhao, J. Xu and W. Yang, Chem. Commun., 2010,
Open Access Article. Published on 14 October 2013. Downloaded on 27/02/2017 10:12:14.
46, 2868–2870.
10 (a) P. E. Ivashkin, I. V. Yampolsky and K. A. Lukyanov, Russ.
J. Bioorg. Chem., 2009, 35, 726–743; (b) W. T. Chuang,
C. C. Hsieh, C. H. Lai, C. H. Lai, C. W. Shih, K. W. Chen,
W. Y. Hung, Y. H. Hsu and P. T. Chou, J. Org. Chem., 2011,
76, 8189–8202.
11 (a) D. Maurel, S. Banala, T. Laroche and K. Johnsson, ACS
Chem. Biol., 2010, 5, 507–516; (b) P. Klan, T. Solomek,
G. C. Bochet, A. Blanc, R. Givens, M. Rubina, V. Popik,
A. Kostikov and J. Wirz, Chem. Rev., 2013, 113, 119–191; (c)
Special issue, Photochem. Photobiol. Sci., 2012, 11,
433–600.
24024 | RSC Adv., 2013, 3, 24021–24024
View Article Online
Communication
12 (a) G. H. Patterson and J. Lippincott-Schwartz, Science, 2002,
297, 1873–1877; (b) R. Ando, H. Hama, M. Yamamoto-Hino,
H. Mizun and A. Miyawaki, Proc. Natl. Acad. Sci. U. S. A.,
2002, 99, 12651–12656; (c) H. Tsutsui, S. Karasawa,
H. Shimizu, N. Nukina and A. Miyawaki, EMBO Rep., 2005,
6, 233–238; (d) S. Habuchi, R. Ando, P. Dedecker,
W. Verheijen, H. Mizuno, A. Miyawaki and J. Hoens,
Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 9511–9516; (e)
N. Gagey, M. Emond, P. Neveu, C. Benbrahim, B. Goetz,
I. Aujard, J. B. Baudin and L. Jullien, Org. Lett., 2008, 10,
2341–2344; (f) E. Betzig, G. H. Patterson, R. Sougrat,
O. W. Lindwasser, S. Olenych, J. S. Bonifacino,
M. W. Davidson, J. Lippincott-Schwartz and H. F. Hess,
Science, 2006, 313, 1642–1645.
13 D. Groff, F. Wang, S. Jockusch, N. J. Turro and P. J. Schultz,
Angew. Chem., Int. Ed., 2010, 49, 7677–7679.
14 U. Brackmann, Lambdachrome Laser Dyes Lambda Physik,
Goettingen, Germany, 1986.
15 (a) Y. Gao, J. K. Y. Chan and C. Xu, Am. J. Nucl. Med. Mol.
Imaging, 2013, 3, 232–246; (b) M. R. Porcel, J. C. Wu and
S. S. Gambhir, Molecular imaging of stem cells (July 30,
2009), StemBook, ed. The Stem Cell Research Community,
StemBook, DOI: 10.3824/stembook.1.49.1.
This journal is ª The Royal Society of Chemistry 2013

ChemComm
Cite this: Chem. Commun., 2012, 48, 6250–6252
www.rsc.org/chemcomm
On the heterogeneity of fluorescence lifetime of room temperature ionic
liquids: onset of a journey for exploring red emitting dyesw
Published on 01 May 2012. Downloaded by North Dakota State University on 22/10/2014 02:55:34.
Anup Ghosh, Tanmay Chatterjee and Prasun K. Mandal*
Received 20th February 2012, Accepted 30th April 2012
DOI: 10.1039/c2cc32177g
An excitation and emission wavelength dependent non-exponential
fluorescence decay behaviour of room temperature ionic liquids
(RTILs) has been noted. Average fluorescence lifetimes have been
found to vary by a factor of three or more. Red emitting dyes
dissolved in RTILs are found to follow hitherto unobserved single
exponential fluorescence decay behaviour.
Room temperature ionic liquids (RTILs) have been considered
to be green alternative solvents for various applications for over
a decade, owing to their better solvent properties in comparison
to the volatile organic solvents.1–7 Coexistence of different types
of interactions, such as coulombic, dipolar, van der Waals, and
H-bonding etc., makes these RTILs unique and complex in
comparison to normal solvents.8–18
Various optical spectroscopic studies as well as simulation
analyses have shown the existence of heterogeneous cation–cation,
cation–anion mesoscopic local ordering in liquid state.8–18 Thus,
RTILs are considered to be ‘‘nano-structured fluids’’.15 Optical
studies on RTILs have shown that imidazolium-based RTILs
possess significant absorption in the UV region, extending into a
part of the visible region.19,20 Imidazolium RTILs are known to
exhibit excitation wavelength dependent emission behaviour.19,20
It has been shown unambiguously that both absorption and
emission are inherent to RTIL itself and not due to any
impurity.19,20 However, although there exists detailed investi-
gation on the spectral behaviour of RTILs, there is no
rigorous analysis of the fluorescence decay behaviour of neat
RTILs. An unusual excitation wavelength dependent fluores-
cence behaviour (known as Red Edge Effect or REE) has been
reported for neat RTILs.19,20 REE has also been observed for
some dipolar solutes dissolved in RTILs.21 It has been speculated
that different associated species are responsible for the observed
phenomenon, perhaps either because of inefficiency of the excita-
tion energy-transfer between them or an incomplete solvation
process.19–21 It has also been mentioned that REE is observed
Department of Chemical Sciences, Indian Institute of Science
Education and Research (IISER)-Kolkata, Mohanpur Campus,
West Bengal 741252, India. E-mail: prasunchem@iiserkol.ac.in
w Electronic supplementary information (ESI) available: Experimental
details, absorption and emission behavior of neat RTILs, fluorescence
decay curves of RTILs excited at different wavelengths, fluorescence
lifetime charts of neat RTILs, fluorescence decay curves of dyes
dissolved in RTILs and their lifetime chart. See DOI: 10.1039/
c2cc32177g
6250 Chem. Commun., 2012, 48, 6250–6252
View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
COMMUNICATION
because of longer-time trapping of solutes in quasistatic local
solvent cages in comparison to their fluorescence lifetime.22,23
Fluorescence decay of dissolved dyes (e.g. C153, AP etc.) has
been reported to be non-exponential in nature. It was speculated
that these dyes trapped in different solvent cages behave as
different species and thus decay differently.21,24,25
The fluorescence quantum yield of imidazolium RTILs has
been shown to be 0.003.19,20 This means that RTILs are very
weakly fluorescent. However, while performing single molecule
spectroscopic studies with bright dyes it was observed that it is
not possible to see single dyes unless we excite beyond 550 nm.
This surprising observation made us believe that although
fluorescence quantum yield of RTILs is very low, significant
background fluorescence is generated as the number of fluorescing
species around the single dye is high.
In this communication, we report hitherto unknown monitoring
wavelength dependent heterogeneity of the fluorescence lifetime of
most hydrophobic FAP anion based imidazolium RTILs (Chart 1).
We show spectroscopic evidence for the reason behind the
observation of REE in neat RTILS. In contrast to previously
reported nonexponential decay behaviour of dyes in RTILs, we
demonstrate and provide a possible reason for the single
exponential fluorescence decay behaviour of two red emitting
dyes (LD700 and OX725 (Chart 1)).
The steady state absorption behaviour of these RTILs resembles
that of other imidazolium based RTILs (see S1, ESIw). These
RTILs also show an excitation wavelength dependent emission
behaviour (see S2, ESIw) similar to what has been observed for
other imidazolium based RTILs.19,20
However, monitoring the fluorescence decay of both RTILs
when excited at 375 nm, we have noticed hitherto unobserved
monitoring wavelength dependent emission decay behaviour.
As can be seen from Fig. 1, when the monitoring wavelength is
changed from 390 nm to 440 nm, the fluorescence lifetime
increased (see Table 1) (for complete chart see S3, ESIw). With
a further increase in monitoring wavelength (to 660 nm),
Chart 1 Chemical structures of ionic liquids and the red emitting dyes.
This journal is c The Royal Society of Chemistry 2012
 View Article Online

For example, when the fluorescence decay was monitored at

450 nm (or 620 nm) while exciting at 377 nm and 402 nm,
clearly distinct decay time constants were noted (S3, ESIw).
This strongly suggests the lack of ultrafast (subpicosecond)
energy transfer among various nano-structured subensembles
that are inherently present in RTILs. Different time constants
suggest that with different excitation wavelengths different
subensembles are excited and thus they emit and decay
Published on 01 May 2012. Downloaded by North Dakota State University on 22/10/2014 02:55:34.

differently.
Several fluorescent dyes have so far been explored in RTILs.24,25
Fluorescence decay behaviour of these dyes has been characterized
Fig. 1 Fluorescence decay curves of [emim][FAP]; lex = 377 nm. by a non-exponential decay function. It has been proposed that
Monitoring wavelengths are shown as inset. these dyes get trapped in different nanostructural cages and thus
decay differently. In our study, we have measured the fluorescence
Table 1 Monitoring wavelength dependent fluorescence lifetime of decay curve of two red emitting dyes namely LD700 and OX725
[emim][FAP]. lex = 377 nm in [emim][FAP] (see Fig. 2). Interestingly, we have observed
lem (nm) t1 (ns) (B1) t2 (ns) (B2) t3 (ns) (B3) hti w2 a single exponential fluorescence decay behaviour. This obser-
vation is in stark contrast to literature reports to date (e.g.
390 0.12 (5.42) 1.25 (29.81) 10.85 (64.77) 7.41 1.24
non-exponential decay of C153, AP, PRODAN etc.).
430 0.84 (19.31) 3.25 (25.22) 13.30 (55.44) 8.35 1.05
440 1.01 (18.52) 4.25 (27.35) 14.50 (54.13) 9.20 1.15 Decay curves of both red emitting dyes in other RTILs are
620 0.41 (13.36) 2.20 (36.98) 8.90 (49.66) 5.28 1.09 shown in the ESIw (see S6 and S7). From all these decay
660 0.45 (12.62) 1.75 (57.64) 6.10 (29.69) 2.88 1.04 curves, a clear single exponential decay behaviour is evident.
The decay time constants are given in S8 (ESIw). It might be
the average fluorescence lifetime continued to decrease. A argued that the excitation wavelength is 635 nm, which is
triexponential decay function was necessary to fit the decay much beyond the tail of absorption spectra of RTILs, so single
function at all wavelengths. Strikingly, a three-fold change in exponential decay behaviour is expected. However, in order to
the average fluorescence lifetime was noted on changing the prove our claim we have also used 377 nm and 402 nm as the
monitoring wavelength (Table 1). excitation wavelengths while keeping the monitoring wave-
We repeated the fluorescence decay measurements for both length the same. Even under these experimental conditions, we
RTILs by exciting at 402 nm. Even for this particular excita- have observed a clear single exponential decay for both dyes
tion wavelength, we observed a two-fold increase in average in both RTILs (see S9–S12, ESIw). Fluorescence lifetimes
lifetime (see S3, ESIw). A similar observation was noted for the for different excitation wavelengths are found to be almost
[hydemim][FAP] (see S4 and S5, ESIw). Thus, a non-exponential identical (S8, ESIw). This observation is in stark contrast to
fluorescence decay behaviour was quite evident for different what is currently known about the fluorescence decay behaviour
excitation wavelengths and for different monitoring wavelengths. of the dissolved dyes in RTILs. It is believed by the RTIL
Therefore, for both RTILs, a heterogeneity of fluorescence life-
time is clearly demonstrated.
We would like to report here another interesting observation.
While studying REE of RTILs,19,20 it was speculated that the lack
of solvation and energy transfer between the subensembles could
be the reason for the observation of REE. On closer look we
found no rise-time after fitting decay curves of neat RTILs. This
observation has been noted for both excitation wavelengths and
for monitoring wavelengths ranging from the lower wavelength
end through to the extreme red edge of the emission curve. This
phenomenon has been observed for both RTILs. This is a clear
indication of the lack of any excited state phenomenon (like
solvation) in the neat RTILs. Moreover, it also tells that there
exists no resonance energy transfer among various heterogeneous
nano-structured subensembles inherently present in the RTILs.
The time-resolution of our setup is about 70 ps so we could
safely mention that there is no solvation or energy transfer
process slower than 35 ps. However, in principle, an ultrafast
solvation with subpicosecond time constant cannot be over-
ruled. If there has been an ultrafast energy transfer between
subensembles in neat RTILs then we should have observed
similar decay time constants when fluorescence decays were
monitored at the same wavelength. However, that is not what we Fig. 2 Fluorescence decay behaviour of LD700 (top) and OX725
observed. Instead, we have found different decay time constants. (bottom) in [emim][FAP]. lex = 635 nm and lem = 650 nm.

This journal is c The Royal Society of Chemistry 2012 Chem. Commun., 2012, 48, 6250–6252 6251

community that subensembles of dyes associated with a
particular local nanoenvironment absorb and emit at different
wavelengths. This was known to be the reason behind their
nonexponential fluorescence decay behaviour. Similar results
should have been observed in our case also. However, instead
we have observed a clear single exponential decay of these two
dyes in both RTILs for different excitation wavelengths. Thus,
it is evident that it is not the excitation wavelength rather the
Published on 01 May 2012. Downloaded by North Dakota State University on 22/10/2014 02:55:34.
monitoring wavelength which dictates the exponentiality of
the fluorescence decay behaviour.
In order to understand this observation we look back at the
steady state optical behaviour of the dyes in RTILs. A closer
look at the optical behaviour of most of these dyes (say, C153,
AP, PRODAN etc.)21,24,25 would reveal that these dyes have
an emission curve which overlaps with that of the RTILs.
However, the absorption and emission spectra for LD700 and
Oxazine725 are well beyond the tail of absorption and emission
curves of RTILs (see S13 and S14, ESIw). The sizes as well as
the electronic property of LD700 and Oxazine725 are not
drastically different from those of Coumarin 153 or PRODAN.
Thus, if Coumarin 153 or PRODAN can be trapped in
nanostructural cages then LD700 and Oxazine725 could also
go to similar cages.
Recent research on ionic dyes in RTILs has shown the
existence of ionic probe–RTIL counterion interaction that is
different from neutral probe–RTIL interaction.26–28 So it may
be surmised that the cationic dyes employed in this work
would go to specific regions and thus produce only one type
of species and hence would show single exponential decay.
However, earlier time-resolved fluorescence studies employing
ionic dyes have shown a non/stretched exponential fluores-
cence decay.29,30 Emission spectra of all those ionic dyes have
some magnitude of overlap with those of RTIL.29,30
It is thus quite evident that, because of absence of perturbation
of RTIL emission, these two dyes do follow a single mode of
fluorescence decay process. Thus, our observation points to the
fact that a reassignment for the nonexponential decay behaviour
of dyes, while dissolved in RTILs, is necessary in light of the
present results.
In conclusion, we have demonstrated hitherto unknown
excitation wavelength dependent fluorescence decay behaviour
and hence a heterogeneity of fluorescence lifetime of neat RTILs.
The evidence of the speculated reason for the observation of a
red edge effect, i.e. the lack of solvation or energy transfer
phenomenon among different nano-structured subensembles
inherently present within RTILs, is established. We have
shown a clear single exponential fluorescence decay behaviour
of red emitting dyes whose fluorescence is not perturbed by
that of RTILs. We strongly believe that these observations will
open a new domain of spectroscopic studies where perturbation
of RTIL emission is no longer present on the fluorescence
processes of the probes. Work towards understanding dynamic
phenomena like solvation, resonance energy transfer employing
6252 Chem. Commun., 2012, 48, 6250–6252
View Article Online
neutral and ionic red emitting dyes in RTILs is currently
underway.
Abbreviations used in the text: FAP, tris(pentafluoroethyl)-
trifluorophosphate; RTIL, room temperature ionic liquid;
emim, 1-ethyl-3-methyl-imidazolium; hydemim, 1-(2-hydroxyl-
ethyl)-3-methyl-imidazolium; C153, coumarin 153; AP, 4-amino-
phthalimide; PRODAN, 6-propionyl-2-dimethylaminonaphthalene.
PKM thanks IISER-Kolkata for financial help and instru-
mental facilities. AG and TC thank UGC and CSIR for their
respective Fellowship.
Notes and references
1 T. Welton, Chem. Rev., 1999, 99, 2071.
2 P. Wasserscheid and W. Keim, Angew. Chem., Int. Ed., 2000,
39, 3772.
3 J. Dupont, R. F. D. Souza and P. A. Z. Suarez, Chem. Rev., 2002,
102, 3667.
4 R. D. Rogers and K. R. Seddon, Science, 2003, 302, 792.
5 K. R. Seddon, Nat. Mater., 2003, 2, 363.
6 H. WeinGaertner, Angew. Chem., Int. Ed., 2008, 47, 655.
7 J. P. Hallet and T. Welton, Chem. Rev., 2011, 111, 3508.
8 J. Dupont, P. A. Z. Suarez, R. F. D. Souza, R. A. Burrow and
J. Kintzinger, Chem.–Eur. J., 2000, 6, 2377.
9 C. Chiappe and D. Pieraccini, J. Phys. Org. Chem., 2005, 18, 275.
10 Y. Wang and G. A. Voth, J. Am. Chem. Soc., 2005, 127, 12192.
11 J. N. C. Lopes and A. A. H. Padua, J. Phys. Chem. B, 2006,
110, 3330.
12 J. N. C. Lopes, M. F. C. Gomes and A. A. H. Padua, J. Phys.
Chem. B, 2006, 110, 16816.
13 D. Jeong, Y. Shim, M. Y. Choi and H. J. Kim, J. Phys. Chem. B,
2007, 111, 4920.
14 L. M. N. B. F. Santos, J. N. C. Lopes, J. A. P. Coutinho, J. M. S. S.
Esperanca, L. R. Gomes, I. M. Marrucho and L. P. N. Rebelo, J. Am.
Chem. Soc., 2007, 129, 284.
15 K. Iwata, H. Okajima, S. Saha and H. Hamaguchi, Acc. Chem.
Res., 2007, 40, 1174.
16 Y. Wang, W. Jiang, T. Yan and G. A. Voth, Acc. Chem. Res.,
2007, 40, 1193.
17 L. P. N. Rebelo, J. N. C. Lopes, J. M. S. S. Esperanca, H. J. R.
Guedes, J. Lachwa, V. N. Visak and Z. P. Visak, Acc. Chem. Res.,
2007, 40, 1114.
18 A. A. H. Padua, M. F. C. Gomes and J. N. A. C. Lopes, Acc.
Chem. Res., 2007, 40, 1087.
19 A. Paul, P. K. Mandal and A. Samanta, Chem. Phys. Lett., 2005,
402, 375.
20 A. Paul, P. K. Mandal and A. Samanta, J. Phys. Chem. B, 2005,
109, 9148.
21 P. K. Mandal, M. Sarkar and A. Samanta, J. Phys. Chem. A, 2004,
108, 9048.
22 Z. Hu and C. J. Margulis, Proc. Natl. Acad. Sci. U. S. A., 2006,
103, 831.
23 Z. Hu and C. J. Margulis, Acc. Chem. Res., 2007, 40, 1097.
24 A. Samanta, J. Phys. Chem. B, 2006, 110, 13704.
25 A. Samanta, J. Phys. Chem. Lett., 2010, 1, 1557.
26 M. Muramats, Y. Nagasawa and H. Miyasaka, J. Phys. Chem. A,
2011, 115, 3886.
27 H. Qiu, M. Takafuji, T. Sawada, X. Liu, S. Jiang and H. Ihara,
Chem. Commun., 2010, 46, 8740.
28 S. Bruzzone, M. Malvaldi and C. Chiappe, J. Chem. Phys., 2008,
129, 074509.
29 N. Ito, S. Arzhantsev and M. Maroncelli, Chem. Phys. Lett., 2004,
396, 83.
30 D. C. Khara and A. Samanta, Indian J. Chem., Sect. A, 2010,
49, 714.
This journal is c The Royal Society of Chemistry 2012