Multi Spectroscopiccharacterizationofbitumenanditspolarity Basedfractions

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/363288255

Multi-spectroscopic characterization of bitumen and its polarity-based


fractions

Article in Construction and Building Materials · October 2022


DOI: 10.1016/j.conbuildmat.2022.128992

CITATIONS READS

7 402

5 authors, including:

Stefan Werkovits Markus Bacher


TU Wien University of Natural Resources and Life Sciences Vienna
8 PUBLICATIONS 187 CITATIONS 130 PUBLICATIONS 2,688 CITATIONS

SEE PROFILE SEE PROFILE

Thomas Rosenau Hinrich Grothe


University of Natural Resources and Life Sciences Vienna TU Wien
638 PUBLICATIONS 13,937 CITATIONS 194 PUBLICATIONS 8,253 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Stefan Werkovits on 08 September 2022.

The user has requested enhancement of the downloaded file.


Construction and Building Materials 352 (2022) 128992

Contents lists available at ScienceDirect

Construction and Building Materials


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

Multi-spectroscopic characterization of bitumen and its


polarity-based fractions
Stefan Werkovits a, Markus Bacher b, Johannes Theiner c, Thomas Rosenau b, Hinrich Grothe a, *
a
Christian Doppler Laboratory for Chemo-Mechanical Analysis of Bituminous Materials, Institute of Materials Chemistry, TU Wien, Getreidemarkt 9/E-165-01-5, 1060
Vienna, Austria
b
University of Natural Resources and Life Sciences Vienna, Division of Chemistry of Renewables, Konrad-Lorenz-Straße 24, A-3430 Tulln, Vienna, Austria
c
University of Vienna, Microanalysis Services, Austria

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

Keywords: Bitumen contains a complex mixture of molecules, ranging from simple linear alkanes to complex polyaromatic
Bitumen systems. Used in road construction, those structures are permanently influenced by atmospheric processes, which
SARA fractions result in oxidation, degradation, loss of volatiles and structural rearrangements. In order to understand those
NMR
31 molecular changes, a profound description of the unaltered system is needed. In this study, structural and
P NMR
HSQC
chemical peculiarities of bitumen and its polarity-based fractions were analysed using nuclear magnetic reso­
HMBC nance (NMR) techniques in combination with infrared/fluorescence spectroscopy and elemental analysis. The
DOSY bitumen sample was separated into saturates, aromatics, resins and asphaltenes. Elemental analysis provided
Elemental analysis insights into variations of heteroatom content and hydrogen-carbon ratio among the fractions. Fluorescence
Infrared spectroscopy spectroscopy showed an increase in the aromatic condensation degree with rising polarity. The combination of
Fluorescence spectroscopy heteronuclear single quantum coherence (HSQC) and elemental analysis facilitated the assessment of length and
branching extent of aliphatic side chains, throughout the fractions. Another 2D NMR technique, heteronuclear
multiple bond coherence (HMBC) found evidence for minor contributions of carboxylic esters or amides within
the resins. Hydroxyl and amino groups were analysed by 31P NMR after derivatization, revealing minor amounts
of alcohols, amides and carboxylic acids in the most polar fractions. Additionally, diffusion ordered spectroscopy
was carried out to obtain estimations on average molecular sizes and weights. Applying in-depth NMR analysis to
bitumen and combining it with complementary analytical techniques, this study provides a unique basis for
future ageing studies and demonstrates how multi-spectroscopy can give new insights into bitumen structure and
chemistry.

1. Introduction and service life), their lifetime is shortened. In order to prevent deteri­
oration, extensive research has been implemented to assure the quality,
Throughout history, a diverse palette of construction materials has improve mechanical performance, and understand the fundamental
been loyal companion to mankind. Materials such as wood, glass, steel, physical and chemical processes [1,2] in the road construction
concrete, and asphalt became necessities in the founding of towns and materials.
the development of complex road networks connecting those centres of Regarding the chemical complexity of these materials, asphalt ranks
human activities. This facilitated the exchange of material goods, cul­ near the top. The predominant reasons are the varying origins and
tural values and knowledge, accelerating the development of modern production processes of the bituminous binder. Bitumen is a product of
societies. Those advances relied on the robustness and longevity of the crude-oil refining process and is obtained as the residue of the vac­
transport structures. However, as roads are permanently exposed to uum distillation. Only two chemical elements, hydrogen and carbon,
environmental influences, such as solar radiation, oxidative gases from account for over 90 wt% of the material [2]. Further elements present in
the atmosphere and huge temperature differences (during production considerable amounts are sulfur, oxygen and nitrogen, and also some

Abbreviations: NMR, nuclear magnetic resonance; HSQC, heteronuclear single quantum coherence; HMBC, heteronuclear multiple bond coherence; DOSY,
diffusion ordered spectroscopy; SARA, saturates aromatics resins asphaltenes.
* Corresponding author.

https://doi.org/10.1016/j.conbuildmat.2022.128992
Received 29 May 2022; Received in revised form 8 August 2022; Accepted 25 August 2022
Available online 5 September 2022
0950-0618/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

transition metals (V, Ni, Fe) are found in traces (ppm)[2]. On a molec­ compounds of oil mixtures are adsorbed on sandstones throughout an
ular level, several functional classes have been found, comprising ageing process. The adsorbed substances were compared to HSQC
aliphatic and (poly-)aromatic structures, oxygen-containing (carbonyl, spectra of the SARA fractions of bitumen (one part of the oil mixtures).
alcohol), sulfur-containing (sulfide, (benzo)thiophene) and nitrogen- There is less literature on application of HMBC spectra in bitumen
containing (pyridine, pyrrole, quinoline and porphyrin) moieties research available, one example is the study of Elbaz [28]. Another
[3–5]. These structural entities are not isolated but rather appear in approach was implemented by Durand [30], who focused on DOSY NMR
different combinations. The fascinating question that arises is how all to assess diffusion coefficients, molecular sizes and weights of hydro­
those functional groups are (inter)connected and how differences in carbon mixtures and asphaltenes. Based on DOSY NMR, Vukovic et al.
these connections affect the material behaviour. Especially the interre­ [31] focused extensively on different types of asphaltenes and analysed
lation between rheological and chemical behaviour and how both their aggregation behaviour in solution.
change during standardized ageing procedures was emphasized in All these examples illustrate how the wide landscape of NMR tech­
thorough studies in the recent literature, building a robust framework to niques can be utilized to analyse different structural features of the
link both worlds and showing the extraordinary importance of chemis­ highly complex hydrocarbon mixtures in bitumen. However, attempts to
try for the analysis and performance of the material [6–9]. The chemical combine all this information in one coherent picture are still largely
evaluation is, however, challenging, due to the almost unlimited number missing, and the present study attempts to advance in this direction,
of possible molecular structures. The task of determining a single com­ providing not only a more detailed description of the mixture of interest,
pound is both meaningless and (almost) impossible, and structural but also a promising and robust basis of future bitumen ageing studies.
characterization has to address the complex compound mixture by
suitable sum parameters. In this regard, structural characterization of 2. Materials and methods
bitumen is similar to that of other overly complex natural products, such
as lignins, tannins or lignocellulosic chromophores, which are also 2.1. Materials
characterized as combination of characteristic structural elements and
functional groups in the bulk [10–12]. A non-modified bitumen, as commonly used in road engineering,
To lower the complexity of the overall bitumen system, it is classified as a 70/100 penetration graded bitumen, was employed.
commonly separated into its polarity-based fractions, the so-called Detailed specification is given in Table 1.
SARA (saturates, aromatics, resins and asphaltenes) fractions [13].
This facilitates the qualitative analysis of the material and provides in­
2.2. SARA fractionation
formation on polarity distribution and ageing state of the bitumen
samples. For example, overlapping spectroscopic responses in various
SARA fractions of the bitumen sample were obtained using a stan­
spectroscopic measurements can be avoided by a preceding separation
dardized chromatographic technique [32]. For each SARA separation
step [4,5,8,9,14–16]. The fractionation provides, on the one hand, more
400 ± 40 mg of the binder (bitumen) was weighed in a 250 ml jar. 40 ml
detailed information on bitumen chemistry (e.g. presence and concen­
of HPLC-grade n-heptane was added and the mixture magnetically
tration of carbonyls, alcohols, aromatic structures, sulfur and nitrogen
stirred for 24 h ± 2 h at 200 rpm. The separation of the asphaltenes (n-
containing groups) and, on the other hand, the possibility to monitor
heptane insoluble) was achieved using syringe filters (Thermo Scienti­
ageing processes with improved accuracy.
fic™ Titan3™ PTFE (Hydrophobic) Syringe Filters, 25 mm diameter,
However, the complex structure and chemistry of bitumen is far from
0.2 µm pore size). The maltenes were collected as filtrate in four vials, of
being fully understood and a single separation or analysis technique
which two were evaporated to dryness on a heating plate at a temper­
alone cannot provide all the necessary information. Therefore, research
ature of 120 ◦ C, while being flushed with nitrogen. The maltene content
groups attempt to combine multiple spectroscopic techniques to validate
of the samples was calculated according to Eq. (1).
and increase the amount of obtainable information [17–19]. In this re­
gard, NMR spectroscopy stands out by the depth of molecular structural 4*mmaltenes,dried
mmaltenes % = *100 (1)
information attainable. It provides the possibility to detect certain nuclei mbitumen
(e.g. 1H, 13C, 31P) and resolves their chemical environment at a molec­
ular scale. This leads to information on directly bonded elements, mmaltenes%… percentual share of maltenes (%).
functional groups, structural motifs, molecular classes and their distri­ mmaltenes,dried … mean mass of the dried maltene fractions (g).
bution, in an ideal case providing the chemical structure of the whole
mbitumen… mass of the bitumen sample (g).
molecule. Other techniques address physicochemical molecular pa­
rameters, such as bond rotation, movements of molecular sections, size
Subsequently, the remaining maltenes were diluted to a concentra­
of molecules and how fast the molecular structures are able to move in
tion of 3.33 mg/ml. For the solid phase extraction (SPE) of the maltenes,
the respective medium. It can also be used to analyse how structural and
SPE cartridges (Thermo Scientific™ HyperSep™ Silica Cartridges, 25 ml
chemical features change during the ageing process, which is particu­
volume, 40–60 μm particle size) were mounted onto a vacuum manifold
larly valuable in bitumen research. In thorough studies, Pipintakos
and 15 ml of the maltene solution was added to the SPE cartridge. After
[17,20], Lu [20], Herrington [21], Michon [22,23] and Siddiqui [24]
the addition of 10 ml of n-heptane, the cartridges were moved to the next
showed how 1H and 13C NMR techniques can be used to assess structural
position on the manifold. In this position the mobile phase was changed
changes during simulated ageing procedures in the laboratory, resulting
and 25 ml of a mixture of toluene and n–heptane (80:20, v/v) was added.
in proposed ageing mechanisms. The main problem in analysing
The cartridges were moved to a third position, where 40 ml of a mixture
bitumen is that there is not one structure present but a rather innu­
of dichloromethane and methanol (90:10, v/v) was added. The obtained
merable quantity of different molecules, each producing a specific nu­
SAR-fractions were evaporated on a heating plate under N2 flushing
clear response by themselves, resulting in highly overlapping signals. In
recent years, researchers tried to solve this problem in two ways: 1)
Table 1
combining separation methods (see above) with NMR analysis [5,18,25]
Rheological parameters of the bitumen sample.
and 2) using 2D NMR techniques to better resolve overlapping signals
[26–28]. Silva et al. provided a comprehensive review on the usage of Rheologic parameter Obtained value

NMR techniques in the analysis of heavy crude oil mixtures [29]. An Penetration at 25 ◦ C 91 mm
interesting example is given in the work of Shikhov [26], which Performance grade 58–22
Softening point 47 ◦ C
employed HSQC as an advanced characterization method to track which

2
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

until constant masses were achieved. The remaining asphaltene frac­ excitation-emission mapping (FEEM) and emission scans were performed
tions were dissolved in toluene, and the solution transferred into glass with wavelength increments of 5 nm (1 nm for emission scans) and dwell
vials and evaporated in a similar way. times of 0.25 s. Further experimental parameters are given in Table 2.

2.3. Elemental analysis of bitumen and its fractions


2.6. NMR spectroscopy
Carbon, hydrogen, nitrogen and sulfur were analysed in a single run
All NMR spectra were recorded on a Bruker Avance II 400 (resonance
on an EA 3000 CHNS-O analyser (Eurovector) with flash combustion at
frequencies 400.13 MHz for 1H and 100.63 MHz for 13C) equipped with
1000 ◦ C. The combustion products N2, CO2, H2O and SO2 were analysed
a 5 mm N2-cooled cryo probe head (Prodigy) with z–gradients at room
by on line-gas chromatography with thermal conductivity detection. Data
temperature with standard Bruker pulse programs. The sample was
were recorded by the Callidus software. The analysis of oxygen is based
dissolved in 0.6 ml of CDCl3 (99.8 % D). Chemical shifts are given in
on high temperature pyrolysis of material in a HT-1500 high temperature
ppm, referenced to residual solvent signals (7.26 ppm for 1H, 77.0 ppm
oven (Hekatech) using granulated carbon as the reducing agent. Diges­
for 13C). 1H NMR data were collected with 32 k complex data points. All
tion at 1480 ◦ C creates only carbon monoxide from all kinds of oxygen in
two-dimensional experiments (HSQC, HMBC) were performed with 1 k
the samples. Data evaluation was performed with in-house software tools
× 256 data points, while the number of transients and the sweep widths
optimised for low-level contents of N, S and O. The LOQ (limit of quan­
were optimized individually. HSQC experiments were acquired in edited
tification) was 0.05 wt% (500 ppm) for C, H, N and O and 0.02 wt% (200
mode using an adiabatic pulse for inversion of 13C and GARP-sequence
ppm) for S, according to long term laboratory evaluations. The uncer­
for broadband 13C-decoupling, optimized for 1J(CH) = 145 Hz. For the
tainty of element quantification is smaller than 0.3 wt%. Quantitative
HMBC spectra a long-range coupling constant of JCH = 8 Hz was used.
manipulation of small amounts of high viscous and sticky bitumen ma­
terial might be unreliable and was therefore replaced by a two-step
2.6.1. DOSY NMR spectroscopy
sample preparation: The sample was dissolved in cyclohexane and
For measurement of the diffusion coefficients with diffusion-ordered
transferred into tin crucibles used for analysis by a micro-pipette. The
NMR spectroscopy (DOSY) a pulse sequence with bipolar gradient pulses
solvent was removed by evaporation at 105 ◦ C over 30 min. Attenuated
for diffusion and eddy current delays was used. The samples were dis­
total reflection infrared spectroscopy (ATR-IR) spectroscopy was used to
solved in 0.6 ml CDCl3 and for relative calibration of the spectra one
verify the complete removal of the solvent. The constant sample mass
drop of CH2Cl2 was added. CDCl3 itself was not suited as calibration
after drying was recorded on an ultra-micro-balance (Cubis MCA2,
reference because of overlap with aromatic signals leading to not clearly
Sartorius, ±0.1 µg) and used as 100 % reference. Reference measure­
defined peaks in the 2D DOSY spectra. The temperature during mea­
ments directly with the native material verified the validity of this sample
surement was set to 303 K and the diffusion specific parameters were
preparation. Triplicate analysis was performed. A sum of 100 ± 0.6 wt%
determined being Δ = 0.15 s and δ = 600 μs. Data were acquired with 16
confirmed complete coverage of elemental constituents.
k data points and a recycle delay of 2 s whereas the gradient strength
was incremented in 16 linear steps from 2 % to 98 % attenuation. Es­
2.4. Infrared spectroscopy timations on the hydrodynamic radius RH and the average molecular
weight were made according to Eqs. (2) and (3). The dynamic viscosity η
IR spectra of the fractions were obtained on a Vertex 80v FTIR was approximated with the viscosity of the solvent (CDCl3) at 303 K. The
spectrometer (Bruker, Karlsruhe) in attenuated total reflection mode diffusion coefficient D is strongly dependent on the analyte concentra­
with a diamond crystal. The spectral range was 680–4000 cm− 1, a tion [30]. To guarantee only minor deviations, the weight percentage
spectral resolution of 1 cm− 1 was chosen and 48 scans were accumulated was kept low, within a range of 0.5–1.5 wt% (manalyte/msolvent) for all
for each spectrum. The spectra were normalized at 2920 cm− 1. To samples. Finally, to obtain estimations of molecular weights throughout
measure the saturates, aromatics and resins, a small sample size of each the fractions the shape factor α (empirical exponent in equation (2)) was
fraction was applied directly on the crystal with a spatula. The asphal­ set to 0.6[30,33].
tenes were applied in solution (in dichloromethane) and IR spectra were
recorded after complete evaporation of the solvent (fading signal in the kB *T
RH = (2)
IR spectrum). For each fraction, four samples were applied on the crystal 6*π*η*Di
and each sample was measured four times.
Di… Measured diffusion coefficient of the sample i (m2/s).
2.5. Fluorescence spectroscopy kB… Boltzmann constant (J/K).
T… Temperature (K).
Photoluminescence spectroscopy was carried out with a fluorescence η… Dynamic viscosity of the sample at 303 K (Pa*s).
spectrometer (FLSP920, from Edinburgh Instruments) equipped with a RH… Hydrodynamic radius (Å).
XE900 xenon arc lamp, double Czerny-Turner monochromators (type ( )1
TMS300) and a S900 single photon photomultiplier (type R928) as the Mi = Mref /
Di α
(3)
detector. The measurements in solution were done in dichloromethane at Dref
concentrations of 0.5–1 mg/l in a quartz glass cuvette (500 µl, Hellma™
Suprasil™ quartz, Germany). A non-fluorescent metal substrate was used
Mi… Molecular weight of the sample i (g/mol).
for front-face measurements of non-dissolved samples. Fluorescence
Mref… Molecular weight of dichloromethane (g/mol).
Di… Measured diffusion coefficient of the sample i (m2/s).
Table 2 Dref… Diffusion coefficient of dichloromethane (m2/s).
Parameters of the fluorescence measurements. α… Average shape factor (-).
Method Excitation wavelength Emission wavelength Emission filter
(nm) (nm) (nm) 2.6.2. 31P NMR spectroscopy
Emission 270 340–600/850a 340 The measurement used a derivatization approach, adapted from a
scan standard method in lignin chemistry, in which acidic groups (OH, NH,
EEM 230–500/650a 340–600/440-850a 340/440a
COOH) are chemoselectively converted into phosphite esters [34]. A
a
solution/front-face set up. mixture of 15–20 mg sample and 3–4 mg of internal standard

3
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

31
Fig. 1. P NMR spectrum of derivatized asphaltenes with characteristic chemical shift regions indicated [35].

Table 3 3. Results and discussion


Yields of the SARA fractions.
Fraction Unaged (wt%) 3.1. SARA fractionation
Saturates 21.8 ± 0.6
Aromatics 45.6 ± 0.4 The SARA fractionation was repeated four times. The average weight
Resins 17.5 ± 0.2 percentages are shown in Table 3. The most abundant fraction was the
Asphaltenes 15.1 ± 0.2 aromatic one followed by saturates, resins and asphaltenes. The content
of low-polar fractions (saturates + aromatics) was approximately 30 wt
% higher than that of the more polar ones (resins and asphaltenes).
(N–hydroxy-5-norbornen-2,3-dicarboxylic acid imide, 97 %) were dis­
solved in 800 μl of a 1:1.6 (v/v) mixture of CDCl3 and pyridine (dried
over molecular sieve, 3 Å). After thorough mixing, 100 μl of phosphi­ 3.2. Elemental analysis of bitumen and its fractions
tylation reagent (2-chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane,
TMDP) and 100 μl of a solution of relaxation reagent (4 mg Cr(III)(acac)3 The elemental composition of the investigated bitumen sample and
/ml CDCl3) were added through a septum into the vial to avoid any its SARA fractions is shown in Table 4. Carbon and hydrogen were the
contact between reagents and moisture. After the addition of TMDP, the most abundant elements in the samples with more than 91 wt% in all
sample was shaken for 1 h at room temperature and subsequently fractions. The H/C-ratio, reflecting the degree of saturation, had the
transferred into a moisture-free NMR tube. For quantitative measure­ highest value in the saturates. The ratio decreased for the fractions with
ments, inverse gated 1H decoupled 31P NMR spectra were acquired with higher polarity. However, contrary to previous publications [4,18], the
64 k data points, a relaxation delay of 14 s and 160 scans. The charac­ order found in this study was: Saturates > Resins > Aromatics >
teristic 31P NMR spectrum of an asphaltene sample in the range between Asphaltenes. This divergence of order compared to other studies in
120 and 180 ppm is shown in Fig. 1. The educt (TMDP), the internal literature is probably correlated to the type of SARA fractionation and to
standard and characteristic chemical shifts [35] of specific OH- the origin of the analysed bitumen. To understand this observation on a
containing groups are highlighted. molecular basis, one has to emphasize at first that the amount of aro­
matic structures and their degree of condensation is increasing from
saturates to asphaltenes (according to literature [2]), resulting in a
saturation order of Saturates > Aromatics > Resins > Asphaltenes.
Assuming that this is also true for our binder, the higher H/C-ratio in
resins compared to aromatics points to the amount and length of

Table 4
Elemental content of the SARA fractions and bitumen.
Fraction C (wt%) H (wt%) N (wt%) O (wt%) S (wt%) H/C-Ratio

Saturates 85.01 ± 0.14 12.95 ± 0.04 < 0.05 0.79 ± 0.17 1.03 ± 0.09 1.81
Aromatics 83.63 ± 0.20 9.61 ± 0.11 0.58 ± 0.02 1.45 ± 0.08 4.54 ± 0.15 1.37
Resins 81.59 ± 0.50 9.96 ± 0.06 0.99 ± 0.03 2.61 ± 0.06 4.88 ± 0.01 1.45
Asphaltenes 82.37 ± 0.21 8.74 ± 0.07 1.01 ± 0.07 1.98 ± 0.03 5.42 ± 0.18 1.26
Bitumen 84.34 ± 0.24 10.24 ± 0.07 0.56 ± 0.03 1.13 ± 0.02 4.33 ± 0.27 1.45

4
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

saturated aliphatic side chains covalently bonded to aromatic structures. 1670 cm− 1 (resins), at 1735 cm− 1 (resins and asphaltenes) and from
From the heteroatom content in Table 4 it is evident that no broad bands above 3100 cm− 1, corresponding to quinolone [14] con­
detectable nitrogen groups and only rather little (≤1 wt%) of oxygen taining-, carboxylic acid [14]/ester containing-, and OH/NH-containing
and sulfur-containing moieties were present in the saturates fraction. structures, respectively. A detailed analysis of COOH/OH/NH structures
The higher N, O and S content in the other fractions is another was implemented using 31P NMR (see chapter 3.5.4).
contributor to their increased polarity. All three elements are (highly)
electronegative and thus will be bound in groups of high polarity that, in 3.4. Fluorescence spectroscopy
addition, engage in electrostatic, van der Waals and/or H-bond in­
teractions. The heteroatom content reached a maximum of approx. 9 wt The fluorescence measurements of the highly diluted fractions in DCM
% in resins and asphaltenes. The most abundant heteroatom in all are presented in Fig. 3. The position of the emission maximum shifted to
samples was sulfur followed by oxygen and nitrogen. Furthermore, the higher wavelengths from saturates to asphaltenes (Fig. 3, left). This
oxygen content in resins exceeded that in asphaltenes. characteristic red shift reflects differences of aromatic systems in the
fractions with an increasing degree of polycondensation of the aromatic
structures. The front-face measurements (Fig. 3, right) of the fractions
3.3. Infrared spectroscopy showed a red shift in emission for aromatics, resins and asphaltenes.
Those fractions contain high amounts of different aromatic structures
Infrared spectra of the fractions showed distinct chemical charac­ which interfere with each other by reabsorption processes (inner-filter
teristics (Fig. 2). In the saturates fraction almost all signals were effects) [36,37]. The reason for the almost unaltered saturates spectrum
apparently related to the aliphatic constituents. Minor aromatic bands was the low abundancy of aromatic structures in this fraction. The high
were observable at 1600 cm− 1 and 815 cm− 1 indicating the presence of content of saturated hydrocarbons, which acted as a non-fluorescent
aromatic structures in the saturates fraction. The signal intensity of the matrix in which the fluorescent aromatic molecules are diluted, pre­
major aromatic signal increased from saturates to asphaltenes. vents such inner-filter effects. Table 5 shows how the spectral regions of
Furthermore, sulfoxide signals at 1030 cm− 1 were found within the ARA maximum fluorescence intensity in the diluted samples are shifted to­
fractions, with the highest intensities in resins and asphaltenes. This is wards higher emission wavelength with increasing polarity.
evidently in agreement with the sulfur contents from elemental analysis.
Further indications of oxygen-containing groups came from bands at

Fig. 2. FTIR spectra of the SARA fractions. Left: Fingerprint area. Right: OH/NH region.

Fig. 3. Fluorescence emission (at 270 nm excitation) of the SARA fractions. Left: Dissolved in DCM; right: Front-face measurement of non-dissolved fractions.

5
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

Table 5 Table 6
Spectral regions with maximum fluorescence intensity and estimated aromatic Average diffusion coefficient D, hydrodynamic radius RH and molecular weight
condensation degree. Mw of the SARA fractions from 1H DOSY NMR experiments.
Fraction Emission (nm) Number of condensed aromatic rings [38] Sample D (m2/s) RH (Å) Mw (g/mol)

Saturates 350–370 ± 5 1–3 Saturates 4.77*10-10 9.15 1175 ± 38


Aromatics 370–405 ± 5 2–4 Aromatics 4.18*10-10 10.40 1452 ± 71
Resins 390–450 ± 10 2–5 Resins 3.74*10-10 11.69 1770 ± 150
Asphaltenes 445–500 ± 10 5–7 Asphaltenes 2.10*10-10 20.69 4570 ± 105
Bitumen 370–415 ± 5 – Bitumen 4.03*10-10 10.80 1545 ± 52

3.5. NMR spectroscopy than methylene (CH2) groups, resulting in either positive or negative
signals. This is represented in the 2D contour plot with different colours:
3.5.1
. 1H DOSY NMR CH3 and CH are given in blue and CH2 in red. Intrinsically, HSQC is
Fig. 4 shows a multiple display plot of the experimental 1H DOSY unable to report carbon atoms without H being directly attached, leav­
spectra of the fractions. The vertical axis corresponds to the chemical ing functional groups, such as ketones, carboxylic acids or esters, and
shift of protons (1H) and the horizontal axis to the logarithmic repre­ quaternary carbon moieties in aliphatic chains or condensed aromatic
sentation of the diffusion coefficient. Peaks at δH 5.31 and 7.26 ppm ring systems undetected. Table 7 gives a summary of important spectral
corresponded to the diffusion standard dichloromethane and the solvent areas [29] within the HSQC plots. The share of characteristic CH-
deuterochloroform, respectively. The signals in the aliphatic region structures in specified chemical environments was calculated by
(0.5–3 ppm) were the most intense and therefore used for the evaluation comparing the integrals in specific areas to the total integrals of all
of the average diffusion coefficients. These were extracted after cali­ signals.
bration of the spectra onto CH2Cl2 using the software Dynamics Center The HSQC spectra revealed that throughout all fractions aliphatic
2.8.0.1 (Bruker). Based on Eqs. (2) and (3), the average hydrodynamic structures exceed aromatic ones (shown in Fig. 6). The total methylene
ratio RH and the average molecular mass Mw were estimated (Table 6). content was greater than the methyl content and no evidence of olefinic
The calculated results revealed that both the hydrodynamic radius and structures was observed throughout all samples (in agreement with
the molecular weight increased with higher polarity. The measurement other publications [4,20]). The ratio between methylene and methyl
of the bituminous samples (Table 6) showed that both values were close groups decreased from saturates to resins, aromatics and asphaltenes.
to the values of the aromatic fraction. This result suggests that longer aliphatic chains are present in the
saturates and the shortest aliphatic structures are found in the
3.5.2. HSQC NMR asphaltenes. The highest aromatic CH content was found in the aro­
The HSQC spectra reveal information on carbon atoms directly matics and asphaltenes (~20 %) followed by the resins (~15 %) and
bound to protons within organic molecules. An exemplary HSQC spec­ the saturates (~5%). In order to understand this trend, one has to
trum of bitumen is shown in Fig. 5. The horizontal axis depicts the emphasize that only CHn groups are observable with HSQC. Conse­
chemical shifts of protons and the vertical axis those of carbons. The quently, not just the number of aromatic compounds but mainly the
intrinsic characteristic of the edited HSQC pulse sequence is that aromatic condensation degree and the number of substituents at the
methine (CH) and methyl (CH3) groups give rise to a different phase aromatic core have an impact on the amount of detectable aromatic CH
groups. The higher the degree of condensation or substitution the lower
is the share of measurable aromatic CH in the molecule, while the
‘HSQC-invisible’ quaternary carbons increase at the same time. Those
effects might overcompensate the numerical increase of aromatic
structures in resins and asphaltenes resulting in an overall lower aro­
matic CH percentage.
HSQC also provides the opportunity to differentiate aliphatic sub­
stituents directly attached to aromatic rings (see Fig. 7). The content of
aromatic methylene groups exceeded the aromatic methyl content in all
samples other than the resins. The saturates were the fraction with the
lowest intensities of aromatic CH which is primarily due to the lack of
aromatic molecules in general. The share of aromatic CH and directly
bonded CHn groups (aromatic methyl and methylene) was the highest in
the aromatics fraction which would indicate that this fraction comprises
the most aromatic molecules.
Tendencies for aliphatic side chain lengths and branching were
derived by comparing the share of aromatic alkyls (directly attached to
aromatic rings) to the total content of the respective alkyl (Table 8). The
decreasing share of aromatic methylene groups from aromatics,
asphaltenes, resins to saturates suggests that the chain length of
aliphatic substituents increases in the same order. The share of aromatic
methyl groups in the aromatics fraction exceeded the share in resins and
asphaltenes by a wide margin, indicating that a significant portion of the
methyl groups is directly attached to aromatic rings and/or that the
branching in the high-polar fractions is higher (the higher the branching
Fig. 4. 1H DOSY NMR spectra of the SARA fractions. The horizontal axis de­
picts the chemical shift (ppm) of protons and the vertical axis the diffusion of an aliphatic substituent the more CH3 groups are present). Addi­
coefficient D (m2/s). tionally, the obtained values for the bituminous sample are shown in

6
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

Fig. 5. Exemplary HSQC NMR spectrum of bitumen. Left: Aliphatic region. Right: Aromatic region. Blue coloured domains depict positive signals (CH3, CH groups)
and red domains negative signals (CH2-groups). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.)

Table 7
Characteristic 1H and 13
C chemical shift regions and structural assignments.
1 13
H-shift [ppm] C-shift [ppm] Structural assignments [29]

0.5–1.5 10–25 Terminal CH3 in aliphatic chains


0.9–1.9 24–45 CH2 of alkyl chains
1.3–1.5 31–34 CH group in aliphatic chains
2.0–3.0 10–24 Aromatic CH3
2.3–3.1 26–38 Aromatic CH2
4.5–6.0 120–135 CH in olefins
6.0–9.3 115–145 CH of aromatic compounds

Fig. 7. Share of aromatic CH and directly bonded aromatic methyl and meth­
ylene groups.

Table 8
Share of aliphatic groups directly attached to aromatic rings.
Region Saturates Aromatics Resins Asphaltenes Bitumen

Aromatic 4.92 ± 25.05 ± 15.51 ± 12.77 ± 18.50 ±


CH3/total 0.70 1.90 0.83 0.76 0.39
CH3
Aromatic 1.99 ± 8.12 ± 4.55 ± 5.14 ± 0.34 7.25 ±
CH2/total 0.68 0.09 0.84 0.67
Fig. 6. Ratio of methyl, methylene and aromatic groups in the fractions. CH2
Aromatic 2.50 ± 12.25 ± 7.12 ± 7.15 ± 0.11 9.87 ±
CHn/total 0.66 0.28 0.80 0.56
CHn
Table 8. Being a mixture of the above fractions, the contents of the
molecular groups ranged (as expected) between the obtained extrema of
the fractions.
Additionally, there were several unique signals within the resin in the 13C domain (see Fig. 8). Those signals had high intensities in
spectrum. Those signals appeared at 4.11/65.39 ppm (OCH2), 5.34/ the saturates, resins and asphaltenes but were hardly observable in
138.2 ppm and 5.57/101.0 ppm. The former was interpreted as an the aromatic fraction and the base binder. The shift values indicate
alkoxy moiety in an ether or ester, whereas the chemical shifts of the steric crowding [40], possible structures being CH2 groups that bridge
latter signals would suggest the presence of alkenes (olefins) [39]. branching points as highly substituted carbons; e.g. the bridging CH2
Another interesting feature were three methylene signals (CH2) with in 2,4 dimethyl pentyl or next to quaternary carbons like in
shifts of 1.19, 1.04 and 0.86 ppm in the proton domain and 45.5 ppm neopentyl.

7
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

Fig. 8. Aliphatic region of HSQC spectra (only CH2 signals). Black: bitumen. Fig. 10. HMBC spectrum of bitumen. Highlighted areas are distinguishable
Green: saturates. (For interpretation of the references to colour in this figure spectral regions. Multiple bond C–H correlations are shown for the aromatic
legend, the reader is referred to the web version of this article.) regions. Quaternary C-Atoms greater than 130 ppm.

3.5.3. HMBC NMR type structure with the quaternary carbon of this structure resonating at
The use of another 2D NMR technique, HMBC, enabled us to gain that frequency. There are three spectral regions for CH correlations
additional information not obtainable with HSQC. It detects carbon involving aromatic rings in HMBC: Cali–Haro, Caro–Haro and Caro–Hali (see
atoms which are not directly bound to protons but are separated from Fig. 10). The Caro-Hali region (bottom right in Fig. 10), with almost all
them through a number (mostly 2 – 4) of bonds. Especially quaternary cross-peaks with 1H shifts over 2 ppm, comprises aromatic alkyl groups,
carbons in aromatic structures and carbon atoms of carbonyls (e.g. ke­ i.e. aliphatic chains directly attached to the aromatic core. However,
tones, amides, esters) can be detected with this experiment. The there were also some signals between 1 and 1.5 ppm which have 13C
knowledge about those functionalities is crucial to improve the under­ shifts that exceeded 130 ppm. Those originated from quaternary aro­
standing of molecular structures within bitumen but is also important in matic carbons, which are correlated with the chain protons in β-position.
the analysis of ageing processes. The most intense carbonyl signals were A qualitative comparison of the fractions showed that the aromatic re­
observed in the resins at chemical shifts of 175.5 ppm, indicating the gions of resins and asphaltenes consisted almost exclusively of quater­
presence of carboxylic acids, esters or amides, and 2 different H-regions: nary carbon signals (confirming the higher degree of condensation, see
2.24 and 2.37 ppm, indicating aliphatic structures. This is in good above), whereas in the aromatics fraction additional correlations to ar­
agreement with the intense carbonyl band found in infrared spectros­ omatic CH were found. Finally, also correlations in the Caro-Haro region
copy at 1670 cm− 1 (Fig. 9). were found showing the correlations of aromatic protons with carbons of
Another interesting feature is that the 13C shift in the aliphatic region the aromatic ring systems.
was extended towards 57 ppm with the highest intensities in saturates,
but only minor contributions in the aromatics, asphaltenes and resins 3.5.4. 31P NMR
fraction. The 1H shift of this signal correlated with the frequent peak at Quantitative 31P NMR spectra of the derivatized fractions showed
45.4 ppm (for all fractions, see HSQC), providing evidence of neopentyl- minor contributions of accessible functional groups (OH, NH) in the

Fig. 9. Evidence for amide groups in HMBC and IR spectra of the resin fraction.

8
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

Table 9 4. Conclusions and outlook


OH–/NH-containing groups in the SARA fractions.
Sample Carboxylic acid (%) Alcohols + Amides (%) The focus of this study was to widen the knowledge basis regarding
molecular structures in bitumen. By pairing polarity-based separation
Saturates 0.00 0.00
Aromatics 0.00 ≤0.01 with a combination of spectroscopic techniques, distinct structural and
Resins ≤0.96 ≤0.91 chemical features were revealed. Such assignments would not have been
Asphaltenes ≤0.65 ≤0.32 possible for the base binder (non-fractionated bitumen). The use of 2D
Bitumen 0.00 0.00 NMR techniques provided important molecular features and how they
change with increasing polarity. Throughout the fractions notable dif­
resins and asphaltenes, never exceeding a total of 2 % for all detectable ferences in aromatic condensation and alkyl substituents (length, num­
compounds. The regions of the phosphorous ester derivate signals (see ber and branching) were revealed. DOSY NMR spectroscopy provided
Fig. 1) corresponded to aliphatic OH (alcohols, carboxylic acids) and NH another aspect of the structure of bitumen and its fractions, the average
(amines, amides) groups. The quantitative assessment (Table 9) showed molecular sizes and weights, showing an increase from saturates to
that only resins and asphaltenes contained trace amounts of those asphaltenes. Combining the information of NMR and IR spectroscopy
functional groups. The resins contained the highest share of carboxylic helped to validate functional groups and how they are distributed within
acids and alcohols/amides which is supported by the elemental the fractions. The use of a selective chemical reaction in combination
composition (highest O-content in resins) and the characteristic IR with 31P NMR showed great potential to classify and quantify OH– and
bands (amide and carboxylic acid band in resins). No corresponding NH– groups in the SARA fractions. However, with regard to general­
signals were detected in the bituminous sample, due to the high share of ization of the results, caution is advised as only one binder was analysed
saturates and aromatics, which pushed the concentration of the groups with the full set of analytical techniques in this study. Therefore, it is
below the detection limit. mandatory to emphasize, that varying the crude oil source and/or
changing the manufacturing process will influence the composition of
bitumen and the distribution of fractions and molecular species in the
3.6. Summary bitumen. Thus, while a limitation of generalized conclusions with regard
to the obtained numbers has to be acknowledged, the demonstration of
The main results of this study can be subdivided in structural and the analytical power of the combined spectroscopic approach for
chemical findings. The relative molecular sizes of the bitumen fractions bitumen characterization remains fully upright.
were estimated using DOSY experiments, revealing an increase from The sensitivity of the employed techniques to bitumina variability
saturates to asphaltenes. Furthermore, the combination of elemental and minor specific changes can also be used as an advantage, especially
analysis with HSQC NMR and fluorescence spectroscopy showed dif­ in ageing investigations. The gained knowledge in the assessment of
ferences in aromatic content throughout all fractions. The red-shifting molecular structures will be of great benefit for the elucidation of
emission wavelengths (from saturates to asphaltenes) in the fluores­ structural changes upon ageing processes. In particular, the comparison
cence spectra correlate with increasing degrees of aromatic condensa­ of naturally or artificially aged with unaltered bitumen samples will
tion. HSQC and elemental analysis revealed a change of order among provide new insights on a wide range of chemical parameters, going
aromatics and resins (lower saturation in aromatics and higher content beyond a mere evaluation of the main oxidation products. Application of
of aromatic compounds in aromatics), the main reasons for these dif­ the presented extended multi–spectroscopic approach to the ageing
ferences being number and length of aliphatic substituents. The peak- chemistry of bitumen will be the topic of an upcoming account that will
integration results of HSQC suggest more and longer side chains be reported in due course.
within the resins compared to the aromatics. Furthermore, information
of branching tendencies was gathered, with indications of neopentyl- CRediT authorship contribution statement
like structures (in HSQC). The abundance of these groups throughout
the fractions and the share of aliphatic CH3 groups (as an indirect in­ Stefan Werkovits: Conceptualization, Methodology, Investigation,
dicator) give first hints that the aromatics contain the aliphatic side Formal analysis, Validation, Data curation, Writing – original draft,
chains with the least extent of branching. Visualization. Markus Bacher: Methodology, Investigation, Formal
Looking at the chemical side, sulfur was the main heteroatom analysis, Writing – review & editing. Johannes Theiner: Methodology,
throughout all fractions, reaching a maximum of roughly 5.5 wt% in the Investigation, Writing – review & editing. Thomas Rosenau: Writing –
asphaltenes. Sulfur was observed in the infrared spectra of aromatics, review & editing. Hinrich Grothe: Resources, Supervision, Project
resins and asphaltenes as sulfoxides. The lack of sulfoxide signal in administration, Funding acquisition, Writing – review & editing.
saturates and its very low intensity within the aromatics (at elemental
analysis values comparable to resins and asphaltenes) suggest the Declaration of Competing Interest
presence of low-polar sulfur-containing compounds, such as aliphatic
and aromatic sulfides, disulfides and heterocyclic species, such as The authors declare that they have no known competing financial
(benzo)thiophenes. The oxygen content increases from saturates to interests or personal relationships that could have appeared to influence
resins (maximum), but interestingly decreases again from resins to the work reported in this paper.
asphaltenes. According to infrared spectroscopy we found oxygen
mainly as carboxylic acids and their derivatives (amides/esters) and Data availability
some alcohols. Carboxylic acids were found in resins and asphaltenes,
whereas aromatic amides were found only in the resins. Further evi­ Data will be made available on request.
dence of amide and ester structures were provided by NMR (HSQC/
HMBC). The analysis of OH/NH-containing groups by 31P NMR revealed Acknowledgements
low amounts of aliphatic alcohols and carboxylic acids within the resins
and asphaltenes. The lack of clear differentiation between NH and OH The financial support by the Austrian Federal Ministry for Digital and
groups with all these techniques is still a problem, and a presence of both Economic Affairs, the National Foundation for Research, Technology
types of functionalities is most likely. However, looking at the strikingly and Development and the Christian Doppler Research Association is
low content of nitrogen throughout all fractions (<1%), future attention gratefully acknowledged. Furthermore, the authors would also like to
should focus rather on carboxylic acid esters. express their gratitude to the CD laboratories company partners BMI

9
S. Werkovits et al. Construction and Building Materials 352 (2022) 128992

Group, OMV Downstream and Pittel + Brausewetter for their financial ageing mechanisms in bitumen, Constr. Build. Mater. 260 (2020), 119702, https://
doi.org/10.1016/j.conbuildmat.2020.119702.
support.
[20] X. Lu, H. Soenen, P. Sjövall, G. Pipintakos, Analysis of asphaltenes and maltenes
before and after long-term aging of bitumen, Fuel 304 (2021), 121426, https://doi.
References org/10.1016/j.fuel.2021.121426.
[21] P.R. Herrington, J.E. Patrick, G.F.A. Ball, Oxidation of roading asphalts, Ind. Eng.
[1] R.N.J. Saal, Asphalt Systems, in: L. Holliday, P.E. Evans (Eds.), Composite Chem. Res. 33 (11) (1994) 2801–2809, https://doi.org/10.1021/ie00035a033.
Materials, Elsevier Publishing Company, Amsterdam, 1966, pp. 453–474. [22] L. Michon, D. Martin, J.-P. Planche, B. Hanquet, Estimation of average structural
[2] D. Lesueur, The colloidal structure of bitumen: Consequences on the rheology and parameters of bitumens by 13C nuclear magnetic resonance spectroscopy, Fuel 76
on the mechanisms of bitumen modification, Adv. Colloid Interface Sci. 145 (1) (1) (1997) 9–15, https://doi.org/10.1016/S0016-2361(96)00184-6.
(2009) 42–82, https://doi.org/10.1016/j.cis.2008.08.011. [23] L. Michon, D.A. Netzel, B. Hanquet, D. Martin, J.-P. Planche, Carbon-13 Molecular
[3] J.C. Petersen, A Review of the Fundamentals of Asphalt Oxidation: Chemical, Structure Parameters of RTFOT Aged Asphalts: Three Proposed Mechanisms for
Physicochemical, Physical Property, and Durability Relationships, TRB Transp. Aromatization, Pet. Sci. Technol. 17 (3–4) (1999) 369–381, https://doi.org/
Res. Circ. (2009). 10.1080/10916469908949723.
[4] O.P. Strausz, E.M. Lown. The chemistry of Alberta oil sands, bitumen and heavy oil, [24] M.N. Siddiqui, NMR Fingerprinting of Chemical Changes in Asphalt Fractions on
Alberta Energy Research Institute, Canada, 2003. Oxidation, Pet. Sci. Technol. 28 (4) (2010) 401–411, https://doi.org/10.1080/
[5] J. Woods, J. Kung, D. Kingston, L. Kotlyar, B. Sparks, T. McCracken, Canadian 10916460903070751.
Crudes: A Comparative Study of SARA Fractions from a Modified HPLC Separation [25] J. Huang, Characterization of asphalt fractions by NMR spectroscopy, Pet. Sci.
Technique, Oil & Gas Sci. Technol. - Revue d’IFP Energies nouvelles 63 (1) (2008) Technol. 28 (6) (2010) 618–624, https://doi.org/10.1080/10916460903073797.
151–163, https://doi.org/10.2516/ogst:2007080. [26] I. Shikhov, D.S. Thomas, A. Rawal, Y. Yao, B. Gizatullin, J.M. Hook, S. Stapf, C.
[6] G. Pipintakos, C. Lommaert, A. Varveri, W. Van den bergh, Do chemistry and H. Arns, Application of low-field, 1H/13C high-field solution and solid state NMR
rheology follow the same laboratory ageing trends in bitumen? Mater. Struct. 55 for characterisation of oil fractions responsible for wettability change in
(5) (2022) https://doi.org/10.1617/s11527-022-01986-w. sandstones, Magn. Reson. Imaging 56 (2019) 77–85, https://doi.org/10.1016/j.
[7] S. Weigel, D. Stephan, The prediction of bitumen properties based on FTIR and mri.2018.10.004.
multivariate analysis methods, Fuel 208 (2017) 655–661, https://doi.org/ [27] B. Behera, S.S. Ray, I.D. Singh, Structural characterization of FCC feeds from Indian
10.1016/j.fuel.2017.07.048. refineries by NMR spectroscopy, Fuel 87 (10–11) (2008) 2322–2333, https://doi.
[8] J. Mirwald, S. Werkovits, I. Camargo, D. Maschauer, B. Hofko, H. Grothe, org/10.1016/j.fuel.2008.01.001.
Understanding bitumen ageing by investigation of its polarity fractions, Constr. [28] A.M. Elbaz, A. Gani, N. Hourani, A.-H. Emwas, S.M. Sarathy, W.L. Roberts, TG/
Build. Mater. 250 (2020), 118809, https://doi.org/10.1016/j. DTG, FT-ICR mass spectrometry, and NMR spectroscopy study of heavy fuel oil,
conbuildmat.2020.118809. Energy Fuels 29 (12) (2015) 7825–7835, https://doi.org/10.1021/acs.
[9] J. Mirwald, S. Werkovits, I. Camargo, D. Maschauer, B. Hofko, H. Grothe, energyfuels.5b01739.
Investigating bitumen long-term-ageing in the laboratory by spectroscopic analysis [29] S.L. Silva, A.M.S. Silva, J.C. Ribeiro, F.G. Martins, F.A. Da Silva, C.M. Silva,
of the SARA fractions, Constr. Build. Mater. 258 (2020), 119577, https://doi.org/ Chromatographic and spectroscopic analysis of heavy crude oil mixtures with
10.1016/j.conbuildmat.2020.119577. emphasis in nuclear magnetic resonance spectroscopy: a review, Anal Chim Acta
[10] M.Y. Balakshin, E.A. Capanema, X. Zhu, I. Sulaeva, A. Potthast, T. Rosenau, O. 707 (1–2) (2011) 18–37, https://doi.org/10.1016/j.aca.2011.09.010.
J. Rojas, Spruce milled wood lignin: linear, branched or cross-linked? Green Chem. [30] E. Durand, M. Clemancey, A.-A. Quoineaud, J. Verstraete, D. Espinat, J.-
22 (2020) 3985–4001, https://doi.org/10.1039/D0GC00926A. M. Lancelin, 1H diffusion-ordered spectroscopy (DOSY) nuclear magnetic
[11] M.Y. Balakshin, E.A. Capanema, I. Sulaeva, P. Schlee, Z. Huang, M. Feng, resonance (NMR) as a powerful tool for the analysis of hydrocarbon mixtures and
M. Borghei, O.J. Rojas, A. Potthast, T. Rosenau, New Opportunities in the asphaltenes, Energy Fuels 22 (4) (2008) 2604–2610, https://doi.org/10.1021/
Valorization of Technical Lignins, ChemSusChem 14 (4) (2021) 1016–1036, ef700775z.
https://doi.org/10.1002/cssc.202002553. [31] J. Parlov Vuković, P. Novak, T. Jednačak, M. Kveštak, D. Kovačević, V. Smrečki,
[12] P. Korntner, T. Hosoya, T. Dietz, K. Eibinger, H. Reiter, M. Spitzbart, T. Röder, I. Mikulandra, M. Djetelić Ibrahimpašić, S. Glanzer, K. Zangger, Magnetic field
A. Borgards, W. Kreiner, A.K. Mahler, H. Winter, Y. Groiss, A.D. French, influence on asphaltene aggregation monitored by diffusion NMR spectroscopy: Is
U. Henniges, A. Potthast, T. Rosenau, Chromophores in lignin-free cellulosic aggregation reversible at high magnetic fields? J. Dispersion Sci. Technol. 41 (2)
materials belong to three compound classes. Chromophores in cellulosics, XII, (2020) 179–187, https://doi.org/10.1080/01932691.2018.1561302.
Cellulose 22 (2) (2015) 1053–1062, https://doi.org/10.1007/s10570-015-0566-6. [32] N. Sakib, A. Bhasin, Measuring polarity-based distributions (SARA) of bitumen
[13] ASTM. ASTM D4124-01: Standard Test Methods for Separation of Asphalt into Four using simplified chromatographic techniques, Int. J. Pavement Eng. 20 (12) (2019)
Fractions, ASTM International, 2001. 1371–1384, https://doi.org/10.1080/10298436.2018.1428972.
[14] J.C. Petersen, R.V. Barbour, S.M. Dorrence, F.A. Barbour, R.V. Helm, Molecular [33] T. Cosgrove, P.C. Griffiths, Diffusion in bimodal and polydisperse polymer systems:
interactions of asphalt. Tentative identification of 2-quinolones in asphalt and their 2. Fully protonated bimodal and polydisperse polymer solutions, Polymer 36 (17)
interaction with carboxylic acids present, Anal. Chem. 43 (11) (1971) 1491–1496, (1995) 3343–3347, https://doi.org/10.1016/0032-3861(95)99434-V.
https://doi.org/10.1021/ac60305a012. [34] P. Korntner, I. Sumerskii, M. Bacher, T. Rosenau, A. Potthast, Characterization of
[15] F. Handle, M. Harir, J. Füssl, A.N. Koyun, D. Grossegger, N. Hertkorn, technical lignins by NMR spectroscopy: optimization of functional group analysis
L. Eberhardsteiner, B. Hofko, M. Hospodka, R. Blab, P. Schmitt-Kopplin, H. Grothe, by 31P NMR spectroscopy, Holzforschung 69 (6) (2015) 807–814, https://doi.org/
Tracking Aging of Bitumen and Its Saturate, Aromatic, Resin, and Asphaltene 10.1515/hf-2014-0281.
Fractions Using High-Field Fourier Transform Ion Cyclotron Resonance Mass [35] M. Li, C.G. Yoo, Y. Pu, A.J. Ragauskas, 31P NMR Chemical Shifts of Solvents and
Spectrometry, Energy Fuels 31 (5) (2017) 4771–4779, https://doi.org/10.1021/ Products Impurities in Biomass Pretreatments, ACS Sustain. Chem. Eng. 6 (1)
acs.energyfuels.6b03396. (2018) 1265–1270, https://doi.org/10.1021/acssuschemeng.7b03602.
[16] F. Handle, J. Füssl, S. Neudl, D. Grossegger, L. Eberhardsteiner, B. Hofko, [36] J.R. Lakowicz, in: Principles of Fluorescence Spectroscopy, 3rd, Springer US,
M. Hospodka, R. Blab, H. Grothe, The bitumen microstructure: a fluorescent Boston, MA, 2006.
approach, Mater. Struct. 49 (1–2) (2016) 167–180, https://doi.org/10.1617/ [37] A. Credi, L. Prodi, Inner filter effects and other traps in quantitative
s11527-014-0484-3. spectrofluorimetric measurements: Origins and methods of correction, J. Mol.
[17] G. Pipintakos, H. Soenen, H.Y.V. Ching, C. Vande Velde, S. Van Doorslaer, Struct. 1077 (2014) 30–39, https://doi.org/10.1016/j.molstruc.2014.03.028.
F. Lemiere, A. Varveri, W. Van den bergh, Exploring the oxidative mechanisms of [38] I.B. Berlman, in: Handbook of Fluorescence Spectra of Aromatic Molecules,
bitumen after laboratory short- and long-term ageing, Constr. Build. Mater. 289 Elsevier, 1971, pp. 107–415.
(2021), 123182, https://doi.org/10.1016/j.conbuildmat.2021.123182. [39] L.D. Field, H.L. Li, A.M. Magill, Organic Structures from 2D NMR Spectra, John
[18] M. Guo, M. Liang, Y. Fu, A. Sreeram, A. Bhasin, Average molecular structure Wiley & Sons Inc., New York, 2015.
models of unaged asphalt binder fractions, Mater. Struct. 54 (4) (2021), https:// [40] N.E. Jacobsen, Inverse Heteronuclear 2D Experiments: HSQC, HMQC, and HMBC,
doi.org/10.1617/s11527-021-01754-2. in: NMR Spectroscopy Explained: Simplified Theory, Applications and Examples
[19] G. Pipintakos, H.Y.V. Ching, H. Soenen, P. Sjövall, U. Mühlich, S. Van Doorslaer, for Organic Chemistry and Structural Biology, John Wiley & Sons, Inc, 2007,
A. Varveri, W. Van den bergh, X. Lu, Experimental investigation of the oxidative pp. 489–550.

10

View publication stats

You might also like