Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Journal of Colloid and Interface Science 581 (2021) 826–835

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science


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

Molecular design of dye-TiO2 assemblies for green light-induced


photocatalytic selective aerobic oxidation of amines
Xia Li, Huimin Hao, Xianjun Lang ⇑
Sauvage Center for Molecular Sciences, College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072, China

h i g h l i g h t s g r a p h i c a l a b s t r a c t

 Molecular inverse design was


adopted to identify the blueprint of
dye-TiO2 assemblies.
 The substituted –OH groups of
anthraquinone provide the inherent
visible light absorption and the
binding sites.
 The bi-dentate chelating binding
between dye molecules and TiO2 is
responsible for better photocatalytic
activity.
 The 1,2-dihydroxyanthraquinone-
TiO2 assembly can be powerful
photocatalyst for selective aerobic
oxidation of amines into imines.
 This work highlights the promise of
tailoring dye-semiconductor
assembly for visible light
photocatalysis.

a r t i c l e i n f o a b s t r a c t

Article history: Dye-semiconductor assemblies are very versatile visible light photocatalysts in terms of tunability by
Received 18 June 2020 tweaking either dye molecules or semiconductor materials. Here, we adopted a strategy of molecular
Revised 20 July 2020 inverse design of alizarin red S (ARS) to identify the blueprint underlying the superior photocatalytic
Accepted 23 July 2020
activity of ARS-TiO2 assembly. We discovered that the substituted –OH groups of anthraquinone provide
Available online 3 August 2020
visible light absorption and binding sites. Importantly, the molecular features of 1,2-
dihydroxyanthraquinone (1,2-DHA) contributes mostly to the unique photocatalytic activity after bind-
Keywords:
ing with TiO2 with broad visible light absorption which can be maintained at high concentration of
Visible light
Molecular design
amines. Moreover, the electron-withdrawing effect of -SO +
3 Na groups increase the acidities of substi-

Dye-semiconductor tuted –OH groups, leading to stronger binding and subsequent higher activity. Ultimately, in situ formed
Anthraquinone 1,2-DHA-TiO2 assembly can be a powerful photocatalyst for green light-induced selective oxidation of
Amines amines into imines with aerial O2. This work makes evident the promise of molecular design in tailoring
dye-semiconductor assemblies for visible light-induced photocatalytic selective chemical
transformations.
Ó 2020 Elsevier Inc. All rights reserved.

⇑ Corresponding author.
E-mail address: xianjunlang@whu.edu.cn (X. Lang).

https://doi.org/10.1016/j.jcis.2020.07.115
0021-9797/Ó 2020 Elsevier Inc. All rights reserved.
X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835 827

1. Introduction demonstration of this concept by selective aerobic oxidation of


amines [30]. Inspired by the successful blue light photocatalysis
Solar energy promises a more sustainable way in energy supply of small organic molecules decorated TiO2, ARS could act a blue-
and will play an even more important role, depending on newly print for molecular design in dye-TiO2 assemblies, in which the
unveiled harvesting pathways [1,2]. It can be converted into elec- reaction could proceed under green light irradiation. We adopted
tricity through photovoltaic technologies [3], which are relatively a strategy of molecular inverse design from ARS by varying the
mature with large-scale application. Meanwhile, solar energy can installed substituted groups to deepen our understanding of the
also be used for the transformation of organic chemicals through molecular underpinning for dye-TiO2 assemblies.
photocatalytic processes [4,5], which are still emerging with huge
potential to address the energy issues. In this context, visible light 2. Experimental section
within the solar spectrum should be adequately harvested to drive
photocatalytic reactions. But this target has been quite challenging 2.1. Reagents and solvents
for Earth-abundant transition-metal oxides like TiO2 which can
only absorb UV light due to its large band gaps of rutile or anatase All the reagents were obtained from commercial suppliers such
phase [6]. However, as the most-investigated semiconductor pho- as Sigma-Aldrich, Alfa Aesar, TCI, J & K Scientific and TRC, etc. The
tocatalyst, a plethora of knowledge has been established for the solvents were supplied by Merck, Fischer Scientific and Sigma-
metal oxide of TiO2 [7,8], which in turn could be a nice platform Aldrich. Benzyl-a,a-d2-amine was supplied by CDN Isotopes, Que-
in designing effective visible light photocatalysts. Thereby, dye- bec, Canada. Anatase TiO2 (Ishihara ST-01) was purchased from
TiO2 assemblies come to light in generating long-lived charge sep- Ishihara Sangyo Kaisha, LTD., Japan. ZnO was acquired from
aration state to enable redox reactions related to solar energy har- XFNANO, China. All the reagents are of analytical grades or above,
vesting [9]. and solvents were directly used without further purification.
Dye-TiO2 assemblies can potentially be photocatalysts
employed in areas including conversion of CO2 [10], production
2.2. Typical procedure for photocatalytic test
of H2 [11–13], treatment of water [14,15] and transformation of
organic chemicals [16–18]. In these regimes, different dyes in
In a typical procedure, 25 mg of TiO2, 6.67  104 mmol of
cooperation with TiO2 were used to harvest visible light to acceler-
ligand, 0.004 mmol of TEMPO [(2,2,6,6-tetramethylpiperidin-1-yl)
ate redox chemical reaction at room temperature. In particular, one
oxyl] and 0.2 mmol of amines were added into a 10 mL Pyrex reac-
efficient pathway of using visible light is photocatalytic selective
tor containing 1 mL of CH3CN as the solvent. Chlorobenzene was
transformation of organic chemicals. However, in real photocat-
used as an internal standard. After 5 min of ultrasonication, the
alytic reaction conditions especially with the involvement of O2
mixture was stirred for 30 min under the dark condition to reach
as the oxidant, the processes were quite complex and therefore
adsorption–desorption equilibrium. Next, the reaction mixture
the molecular understanding on the design of dyes remains elu-
was then magnetically stirred at 1500 rpm and illuminated by
sive. In this regard, one can use an already established dye-TiO2
green light-emitting diodes (LEDs) simultaneously with a single-
assembly photocatalyst to consider the principles of molecular
apertured rubber septum of Pyrex reactor to connect with aerial
design for organic dyes in these assemblies. We coin this strategy
O2. When the photocatalytic experiments were completed, the
as molecular inverse design for the dye-semiconductor assemblies.
photocatalyst nanoparticles were separated from the reaction mix-
The main skeleton of the dye molecule was kept and the functional
ture by centrifugation, and the reaction products were analyzed by
groups were installed in a stepwise fashion throughout the design
gas chromatography equipped with a flame ionization detector
process. Moreover, the position of the functional groups might be
(GC-FID). The structures of products were confirmed by compar-
shifted as a control molecule to reaffirm the underlying principle
ison with the retention time of authentic samples by GC-FID and
of molecular design.
further confirmed by gas chromatography-mass spectrometry
Selective aerobic oxidation of amines into corresponding imines
(GC–MS). Additional instrumental conditions and analysis details
is one of the most significant chemical transformations [19–26].
were offered in the supplementary data. Conversion and selectivity
There is unique challenge to carry out this transformation in a pho-
for selective oxidation of substrates to target products were
tocatalytic scenario by dye-TiO2 assemblies because the visible
defined as follows:
light absorption of many organic dyes is sensitive to pH. Even
though we did not perform the oxidation of amines in aqueous Conversionð%Þ ¼ ½ðC0  Cs Þ=C0   100
conditions, the basic nature of amines, particularly in high concen-
trations, could impact the visible light absorption of dyes in an Yieldð%Þ ¼ Cp =C0  100
inert organic solvent, leading to ill-favoured photochemical conse-
quents. Thus, the scope of organic dyes suitable for dye-TiO2 Selectivityð%Þ ¼ ½Cp =ðC0  Cs Þ  100
assembly photocatalysts is not as broad as we envisaged before-
hand. Therefore, the molecular insight for the successful design where C0 is the initial concentration of reactant, Cs and Cp are the
or selection of dye should be deepened to guide the future design concentrations of substrate and target product, respectively, at a
towards practical application. certain time during photocatalytic reaction.
Previously, we have decorated the surface of anatase TiO2 with
colourless small molecules including phenols [27], catechols [28] 2.3. Typical procedure for photoelectrochemical test
and salicylic acids [29] as visible light photocatalysts to induce
the selective aerobic oxidation of amines. The bindings give rise The photocurrent tests were carried out on a Metrohm Autolab
to visible light absorption and the substituted groups in these PGSTAT302N electrochemical workstation in a conventional three
small molecules influence ultimate photocatalytic activity. How- electrode configuration with Ag/AgCl (saturated KCl) as the refer-
ever, these surface complexes can only harvest blue light with con- ence electrode and Pt wire as the counter electrode. A 0.1 mol/L
siderable conversions. Dye-TiO2 assembly can take on this Na2SO4 aqueous solution was used as the electrolyte. Green LEDs
challenge by markedly extending the absorption in visible region. (k = 510 ± 15 nm, 3 W  4) served as the light source. The working
To our knowledge, ARS-TiO2 assembly was the only successful electrodes were prepared as follows: 6 mg of the as-prepared pho-
tocatalyst was suspended in 3 mL dilute Nafion dispersion to pro-
828 X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835

duce a slurry, which was then dropped 10 lL onto an indium tin this property (Fig. 1a). Besides, the UV–visible absorbance of dye-
oxide (ITO) glass electrode and the film was dried by an infrared TiO2 assemblies were also collected as one of the essential mea-
lamp, then the working electrode was obtained after five cycles. surements for visible light photocatalysts (Fig. 1b). It was obvious
Electrochemical impedance spectroscopy (EIS) measurements that AQs are able to form a assembly with the surface hydroxyl
were determined with the same three electrode configuration in sites of TiO2, which extended the absorption edge of TiO2 to visible
0.1 mol/L Na2SO4 aqueous solution. region and the visible light absorption range of AQs-TiO2 assem-
blies follow the same rule with AQs.
3. Results and discussion After characterizing the light absorption properties of different
AQs and AQs-TiO2 assemblies, we then investigated the influence
The material design of macroscopic structures and phase of TiO2 of different AQs as ligands for anatase TiO2 on the visible light pho-
can leverage light absorption, adsorption of dye and substrates, tocatalytic selective aerobic oxidation of benzylamine into imine
electron transfer and mass transfer of substrates and products (Table 1). The results indicate that the light absorption directly
and consequently dominates the eventual outcomes of photocat- affects the consequential photo-induced redox reactions. There-
alytic reactions [31–33]. However, these dimensions would not fore, the photocatalytic activity also increased with the installment
be considered and thereby a commercial TiO2 with high specific of –OH groups in AQ (Table 1, entries 3–5 vs. entry 1). In addition,
surface area was preferred throughout the investigation to avoid we found that the relative position of –OH groups effect the bind-
the discrepancy originated from the material aspect of TiO2. ing pattern (mon-dentate chelating binding or bi-dentate chelating
Herein, the molecular dimension of adsorbed dyes was focused binding) between TiO2 and ligands of 1,2-DHA and 1,4-DHA. Note
on to elucidate the principle for molecular design. Previously, that 1,2-DHA-TiO2 has apparently higher photocatalytic activity
ARS-TiO2 assembly was prepared beforehand as a photocatalyst. than 1,4-DHA-TiO2 (Table 1, entry 5 vs. entry 6). As we expected,
However, we discovered that visible light irradiation and basic the substituted -SO +
3 Na group of AQ has no obvious influence on

condition of amines could promote the adsorption of dyes and the conversion of benzylamine (Table 1, entry 2 vs. entry 1). But
the consequential photocatalytic activity. Subsequently, in situ due to electron-withdrawing effect of -SO +
3 Na group increased

formed dye-TiO2 assemblies were adopted as the applicable photo- the acidities of substituted –OH groups, which might lead to stron-
catalysts. Complicated prefabricating procedure for these photo- ger binding and subsequent higher activity (Table 1, entry 7 vs.
catalysts were circumvented and the activity was enhanced as entry 5). However, the presence of three hydroxyl groups in AQ
well. More importantly, this strategy allows for an immediate eval- skeleton could led to a significant decrease of conversion for ben-
uation of molecular design process. zylamine (Table 1, entry 8), indicative that two adjacent hydroxyl
Scheme 1 shows the complete process of molecular inverse groups can confer a much better conversion.
design for ARS. With the anthraquinone (AQ) remains unvaried, Besides, apart from anatase TiO2, other types of TiO2 like more
the substituted -SO + common AEROXIDE P25 TiO2 was also tested to verify whether
3 Na and –OH groups were installed in a step-
wise fashion to understand the specific importance of these installed the trend is general and the results were summarized in
groups. Fortunately, these molecules can all be procured commer- Table S4, which were in line with the results in Table 1. Besides,
cially, therefore leading to an easier evaluation process without the reaction results under irradiation of blue LEDs were listed in
tedious process for the synthesis of these molecules. By subtracting Table S5, suggesting the same principle hold under blue light irra-
the -SO + diation. But the results in Table S5 also emphasize the importance
3 Na group from ARS, 1,2-dihydroxyanthraquinone
(1,2-DHA) is attained. Then, subtracting one hydroxyl group of light absorption of AQs in determining the eventual selective
from 1,2-DHA, 2-hydroxylanthraquinone (2-HAQ) and 1- oxidation of benzylamine. Thus, more elaborated comparisons of
hydroxylanthraquinone (1-HAQ) can be obtained. Further subtract- 1,2-DHA-TiO2 and 1,4-DHA-TiO2 with similar visible light absorp-
ing one hydroxyl group from 2-HAQ or 1-HAQ leads to AQ, the major tion ranges were carried out to highlight the importance of relative
skeleton. Moreover, both 1,4-dihydroxyanthraquinone (1,4-DHA) positions of the two installed hydroxyl groups. Both 1,2-DHA-TiO2
and trihydroxy-9,10-anthracenedione (1,2,3-THA) were selected and 1,4-DHA-TiO2 have very broad visible light absorption overlap-
as control molecules for 1,2-DHA by shifting one of the 2-hydroxyl ping with that of light emitting spectrum of green LEDs (Fig. 1b).
group onto the 4-hydroxyl group position or adding an extra adja- Moreover, the results of photocurrent curves also indicated that
cent hydroxyl group located at AQ skeleton. 1,2-DHA-TiO2 in principle has superior performance for charge
The UV–visible diffuse reflectance spectroscopy was recorded in separation than 1,4-DHA-TiO2 (Fig. 2a). The impendence spectra
Fig. 1 for the solid samples of molecules in Scheme 1. We can indicated that both 1,2-DHA-TiO2 and 1,4-DHA-TiO2 have the
observe that the visible light absorption of AQs increased with smaller arc radius than that of TiO2 (Fig. 2b), which means they
the instalment of –OH groups whilst -SO + have lower resistance for faster charge transfer. Besides, the bind-
3 Na dose not influence

Scheme 1. Molecular inverse design for the ARS-TiO2 assembly.


X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835 829

Fig. 1. The diffuse reflectance UV–vis spectroscopy of (a) different AQs and (b) different AQs as ligands for surface of anatase TiO2.

Table 1
The influence of AQs on the green light-induced photocatalytic selective aerobic oxidations of benzylamine.[a]

Entry Anthraquinones Conv.[%][b] Sel.[%][b]


1 AQ 2 99
2 AQS 5 99
3 1-HAQ 38 99
4 2-HAQ 45 99
5 1,2-DHA 78 98
6 1,4-DHA 14 99
7 ARS 90 98
8 1,2,3-THA 13 99
[a]
Reaction conditions: benzylamine (0.2 mmol), TEMPO (0.004 mmol), anatase TiO2 (25 mg), ligand (6.67  104 mmol), CH3CN (1 mL), green LEDs (k = 510 ± 15 nm,
3 W  4), air (1 atm), 40 min.
[b]
Determined by GC-FID using chlorobenzene as the internal standard, conversion of benzylamine, selectivity of N-benzylidenebenzylamine.

Fig. 2. a) Transient photocurrent and b) electrochemical impedance results of 1,2-DHA-TiO2, 1,4-DHA-TiO2 and TiO2.
830 X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835

Table 2
The influence of TEMPO derivatives on the green light-induced photocatalytic
selective aerobic oxidations of benzylamine.[a]

Entry TEMPO derivatives Conv.[%][b] Sel.[%][b]


1 78 99

2 27 99

3 70 99

4 60 99

5 42 99

6 46 99 Fig. 3. The influence of the amount of TEMPO on the photocatalytic selective


oxidation of benzylamine with aerial O2 by 1,2-DHA-TiO2 assembly.

[a]
Reaction conditions: benzylamine (0.2 mmol), TEMPO (0.004 mmol), anatase to the significant function of TiO2 that can activate O2 for the pho-
TiO2 (25 mg), ligand (6.67  104 mmol), CH3CN (1 mL), green LEDs
tocatalytic selective aerobic oxidation of organic substrates, the
(k = 510 ± 15 nm, 3 W  4), air (1 atm), 40 min.
[b]
Determined by GC-FID using chlorobenzene as the internal standard, conver- influence of the amount TiO2 was also investigated (Table S2). In
sion of benzylamine, selectivity of N-benzylidenebenzylamine. addition, the photocatalytic experiment was also performed in
C2H5OH. Although a comparable conversion (73%) of benzylamine
was obtained, the selectivity (75%) for imine was much lower than
ing of 1,2-DHA with TiO2 with the catechol group is beneficial to that in CH3CN. Therefore, CH3CN was chosen as the solvent for sub-
the stability of the assembly even though 1,2-DHA is an organic sequent investigations.
dye sensitive to pH. After the establishment of the green light-induced selective oxi-
During visible light-induced selective aerobic oxidation of dation of benzylamine with air by 1,2-DHA-TiO2 assembly photo-
organic chemicals by TiO2 photocatalysis, whether designing a catalysis, a series of blank or control experiments (Fig. 4) were
dye-sensitized system or a surface complex system, TEMPO is an carried out to get some understanding of the possible reaction
indispensable component for its unique ability that reducing the mechanism. It can be clearly seen that under the standard protocol
charge stress upon photocatalyst by the catalytic cycle amongst of 0.33 mol% of 1,2-DHA, 25 mg TiO2 and 2 mol% of TEMPO, the
TEMPO, 2,2,6,6-tetramethylpiperidine-1-oxoammonium (TEMPO+) selective oxidation of benzylamine can reach 78% yield for desired
and 2,2,6,6-tetramethylpiperidin-1-ol (TEMPOH) in the overall imine (Fig. 4h). Moreover, TiO2, a visible light-absorbing dye, and
photocatalytic processes. Besides, its tunability in oxidation poten-
tials and electron transfer capability make it one of the best co-
catalysts in photocatalytic selective chemical transformations
[34,35]. After discovering 1,2-DHA-TiO2 assembly can be powerful
photocatalyst for selective oxidation of amines into imines with
aerial O2, we then selected a series of 4-substituted derivatives of
TEMPO to figure out whether these redox mediators can further
improve the green light-induced selective oxidation of benzy-
lamine by 1,2-DHA-TiO2 assembly photocatalysis. The results were
summarized in Table 2 in which TEMPO gave the best conversion
of benzylamine among these selected redox mediators. Specifi-
cally, it can be observed that 4-substituted derivatives of TEMPO
show only comparably lower conversions of benzylamine and have
no influence on the selectivities of N-benzylidenebenzylamine
(Table 2, entry 1 vs. entries 2–6).
Since TEMPO is the redox mediator that gave the best conver-
sion of benzylamine, further studies were implemented to assess
the influence of the amount of TEMPO on the photocatalytic selec-
tive aerobic oxidation of benzylamine. The results were shown in
Fig. 3. In the absence of TEMPO, the conversion of benzylamine
was only 14%. But significant increase of conversions was achieved Fig. 4. Control experiments for the photocatalytic selective oxidation of amines into
after catalytic amount of TEMPO was added. This suggests that imines. (a) blank reaction; (b) TiO2 only; (c) 1,2-DHA only; (d) TEMPO only; (e)
adding a redox mediator is necessary for higher photocatalytic without TiO2; (f) without 1,2-DHA; (g) without TEMPO; (h) standard conditions; (i)
activity. Nevertheless, further increase the amount of TEMPO, the dark condition. Standard conditions: benzylamine (0.2 mmol), TEMPO
(0.004 mmol), 1,2-DHA (6.667  104 mmol), anatase TiO2 (25 mg), CH3CN
improvement of conversion was not obvious. In consideration of (1 mL), green LEDs (k = 510 ± 15 nm, 3 W  4), air (1 atm), 40 min. [b] yield of imine
TEMPO can play as a radical scavenger in many photocatalytic (N-benzylidenebenzylamine) was determined by GC-FID using the chlorobenzene
reactions, we chose only 2 mol% TEMPO as the co-catalyst. Due as the internal standard.
X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835 831

or even completely ceased (Fig. 4b-g). It was noteworthy that green


light is a prerequisite for occurrence of reaction. Turn off the light
source, the reaction was completely closed (Fig. 4i). When anatase
TiO2 was replaced with another semiconductor metal oxide, ZnO,
no imine product could be observed (Fig. 4j), which indicates the
binding and interaction between dye molecule and semiconductor
surface is the key factor affecting the activity of dye-semiconductor
assemblies. Therefore, it could be concluded that this adsorption
provides a channel for adequate electron injection, which is the
process that initiates the consequential redox reactions.
For a better designed photocatalyst, the mechanistic aspects
have to be understood thoroughly. To gain more insights into the
reaction mechanism, we then conducted the kinetic studies under
optimal reaction conditions (Fig. 5). According to our previous
report [28], the photocatalytic reaction that visible light-induced
selective oxidation of amines to imines by catechol-TiO2 assembly
often observe the characteristics of zeroth-order reaction kinetic.
However, the kinetic curve of green light-induced photocatalytic
selective oxidation of benzylamine by 1,2-DHA-TiO2 assembly con-
Fig. 5. Kinetic studies on green light-induced photocatalytic selective oxidation of forms to the characteristics of pseudo first-order reaction kinetics
amines with aerial O2 by 1,2-DHA-TiO2 assembly. Triangle and square are for the with a reaction constant kH of 0.03586 molL1min1 (Fig. 5, blue
selective oxidation of benzylamine and benzyl-a,a-d2-amine respectively. (For
line). The same reaction pattern follows for the selective aerobic
interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.) oxidation of benzyl-a,a-d2-amine with a reaction constant kD of
0.01444 molL1min1 (Fig. 5, red line). The resulting kinetic iso-
tope effect (KIE) values is 2.48 for benzylamine oxidation, the typ-
ical values for primary kinetic isotope effect suggest Ca  H
Table 3 abstraction by the TEMPO+ indeed occurs and it is the rate deter-
Quenching experiments to determine the ROS for the green light-induced photocat- mining step in the catalytic cycle. This is to our knowledge a
alytic selective aerobic oxidation of benzylamine.[a] unique example for benzylamine oxidation with dye-TiO2 assem-
Entry Quencher (equiv.) Roles Yield [%][b] bly photocatalytic system. To further verify the reaction observes
pseudo first-order reaction kinetic, the influences of the initial con-
1 N2 (–) O2 replacement 0
2[c] p-BQ (0.2) O2 scavenger 5 centration of benzylamine on the reaction rate were then investi-
3 AgNO3 (1) ecb scavenger 16 gated (Table S3). The results were in good agreement with the
4[d] CD3CN (–) 1
O2 maintainer 36 kinetic results.
[a]
Reaction conditions: benzylamine (0.2 mmol), TEMPO (0.004 mmol), 1,2-DHA Reactive oxygen species (ROS) often play crucial roles in selec-
(6.667  104 mmol), anatase TiO2 (25 mg), CH3CN (1 mL), green LEDs tively oxidizing organic molecules into the desired products [36].
(k = 510 ± 15 nm, 3 W  4), air (1 atm), 40 min. Next, trapping experiments were carried out to explore the param-
[b]
Determined by GC-FID using chlorobenzene as the internal standard, conver- eter of ROS in producing imine during photocatalytic selective oxi-
sion of benzylamine, selectivity of N-benzylidenebenzylamine.
[c]
p-BQ, p-benzoquinone.
dation of benzylamine (Table 3). In N2 atmosphere, the
[d]
CD3CN (1 mL). photocatalytic reaction did not take place, indicating that O2 is
the terminal oxidant (Table 3, entry 1). Then we found the addition
of scavenger p-BQ (p-benzoquinone) for superoxide anion radicals
(O 
2 ) or scavenger AgNO3 for electrons (ecb) could induce the con-

TEMPO as redox mediator were the fundamental components in version of benzylamine decreased remarkably (Table 3, entries 2
the photocatalytic system. The high photocatalytic performance and 3). It is worth noting that although photogenerated electrons
of the system lies in the proper combination of TiO2, dye and cannot directly participate in the oxidation process, they can acti-
TEMPO. Without any of the three, the reaction went very slowly vate O2 to generate O 2 . Consequently, the conversion of benzy-

Fig. 6. EPR spectra recorded during green light-induced photocatalytic selective oxidation of benzylamine by 1,2-DHA-TiO2 assembly photocatalysis (a) spin trapping of
superoxide radicals (O2 ) with DMPO ; (b) conduction band electrons of 1,2-DHA-TiO2 assembly; (c) TEMPO. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
832 X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835

the photocatalytic oxidation of benzylamine. The results of trap-


ping experiments suggest that O 2 is the main ROS for this photo-
catalytic reaction.
To make our conclusion more convincing, in situ electron para-
magnetic resonance (EPR) experiments were performed with 5,5-
dimethyl-pyrroline-N-oxide (DMPO) as a spin trap in a CH3CN
solution under green light irradiation to further identify the forma-
tion of O
2 in our system (Fig. 6a). The low-temperature EPR spec-
trum indicates that O 2 was produced in the photocatalytic
selective oxidation of benzylamine and featured good stability.
After defining that O2 is the main ROS, we performed in situ EPR
experiments to investigate the reaction details taking place in
the photocatalytic cycle. Fig. 6b shows the electron transfer pro-
cess of photocatalyst under visible light irradiation. Without visible
light irradiation, the conduction band electron (e cb) signal is pro-
duced by the inherent oxygen vacancies of TiO2 (Fig. 6b, dark).
Fig. 7. The proposed mechanism for the green light-induced photocatalytic
Under visible light irradiation, the e
cb signals dramatically increase
selective oxidation of benzylamine with aerial O2 by 1,2-DHA-TiO2 assembly. (For
interpretation of the references to colour in this figure legend, the reader is referred due to ligand to metal charge transfer (LMCT) (Fig. 6b, 2 min and
to the web version of this article.) 4 min). However, the signal intensity of e cb significantly declined
to its original state once O2 was filled (Fig. 6b, O2 1 min), which
demonstrated that electron transferred to O2. The signal variation
lamine was markedly inhibited when AgNO3 was added into the of ecb of TiO2 confirmed that electron can inject from the ligand
photocatalytic reaction system. Furthermore, when singlet oxygen to the conduction band of TiO2 under the condition of visible light
(1O2) maintainer CD3CN was added into the reaction system as the irradiation and then transfer to O2 with the formation of O 2 . Then,
solvent, the conversion of benzylamine was decreased (Table 3, we recorded the dynamic change of TEMPO by EPR technology
entry 4), which indicates 1O2 did not produce significant effect in (Fig. 6c). TEMPO is an EPR active species. With the reaction went

Table 4
Green light-induced photocatalytic selective oxidation of primary amines into imines with aerial O2 by 1,2-DHA-TiO2 assembly.[a]

Entry Substrate Product T(h) Conv.[%][b] Sel.[%][b]


1 1.0 92 99

[c]
2 5.0 80 98

3 0.9 93 99

4 1.2 92 98

5 1.3 93 98

6 0.6 96 99

[d]
7 0.9 92 98

8 1.3 88 89

9 1.0 91 99

10 1.2 91 98

11 1.1 90 98

12 0.9 97 95

13 1.2 82 98

[a]
Reaction conditions: amine (0.2 mmol), TEMPO (0.004 mmol), 1,2-DHA (6.667  104 mmol), anatase TiO2 (25 mg), CH3CN (1 mL), green LEDs (k = 510 ± 15 nm, 3 W  4),
air (1 atm).
[b]
Determined by GC-FID using chlorobenzene as the internal standard.’ and begin as a paragraph.
[c]
benzylamine (1 mmol), TiO2 (25 mg), 1,2-DHA (6.667  104 mmol), TEMPO (0.02 mmol).
[d]
t-Bu, tert-butyl.
X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835 833

on, the signal intensity of TEMPO decreased partly, which suggest to get final product of N-benzylidenebenzylamine. At the same
TEMPO partly transformed into EPR inactive species, such as time, TEMPO+ turn into TEMPOH. Finally TEMPOH was restored
TEMPO+ and TEMPOH. to TEMPO by reacting with O 2 . This kind of dye-semiconductor
Based on the above investigation and the previous reports, a assembly containing catechol motif will experience a reaction
plausible mechanism was proposed to explain the green light- pathway like that of type-II dye-sensitized solar cells. The adsorp-
induced photocatalytic selective oxidation of benzylamine by 1,2- tion of 1,2-DHA on TiO2 led to the formation of a newly generated
DHA-TiO2 assembly (Fig. 7). Firstly, 1,2-DHA through its catechol visible light absorption band suitable for green light-induced selec-
group self-assembles onto the surface of anatase TiO2 to form tive oxidation of benzylamine. Electron from HOMO of adsorbed
1,2-DHA-TiO2 photocatalyst, which allows the reaction to occur 1,2-DHA was directly excited to the conduction band (cb) of TiO2
at the green light range. Secondly, the surface charge transfer com- without experiencing an excited state of charge transfer. Interest-
plex absorbed green light, and charge carriers were generated. ingly, the valence band (vb) of TiO2 was a spectator during the
Electrons were injected into the conduction band of TiO2 from entire processes for photocatalytic selective oxidation.
the highest occupied molecular orbital (HOMO) of chemically To demonstrate the general applicability of the designed dye-
adsorbated 1,2-DHA and then transfer to O2 resulting in the pro- TiO2 assembly for photocatalytic selective oxidation of amines.
duction of O2 . Simultaneously, holes were left at the 1,2-DHA- First, photocatalytic activity of this system toward the green
TiO2 surface complex in the form of radical cation. Thirdly, surface light-induced selective oxidation of primary amines was tested.
complex radical cation restores itself into original state by interact- The results were listed in Table 4. Primary benzylamine and its
ing with TEMPO. Meanwhile, TEMPO turns into TEMPO+. It is the derivatives bearing various functional groups (Table 4, entries 1–
key to link photocatalytic cycle and TEMPO catalytic cycle, which 13) were conveniently converted to corresponding imines during
can prevent 1,2-DHA-TiO2 assembly from being destroyed by reaction time of 1 h or so under the standard protocol. Electron-
ROS. Fourthly, TEMPO+ oxidize benzylamine with a step of direct rich benzylamines (–OCH3 and –CH3 substituted derivatives) were
two-electron transfer, resulting in the formation of benzylide- oxidized faster than these substrates with electron-withdrawing
neamine. Benzylideneamine couples with unreacted benzylamine substituent (–F, –Cl and –Br) (Table 4, entries 3, 6 and 9–11), indi-

Table 5
Green light-induced photocatalytic selective oxidation of secondary amines into imines with aerial O2 by 1,2-DHA-TiO2 assembly.[a]

Entry Substrate Product T(h) Conv.[%][b] Sel.[%][b]


1 2.0 84 73

2 1.3 86 80

3 1.7 86 78

4[c] 2.0 80 85

5 2.5 83 68

6 3.0 80 70

7 2.0 49 68

8 4.0 74 92

9 4.0 85 91

10 4.0 57 89

11 4.0 82 92

12 4.0 76 92

13 4.0 71 94

14 1.5 69 82

[a]
Reaction conditions: amine (0.2 mmol), anatase TiO2 (25 mg), 1,2-DHA (6.667  104 mmol), TEMPO (0.006 mmol for entries 1–7 and 14; 0.010 mmol for entries 8–13),
CH3CN (1 mL), green LEDs (k = 510 ± 15 nm, 3 W  4), air (1 atm).
[b]
Determined by GC-FID using chlorobenzene as the internal standard.
[c]
t-Bu, tert-butyl.
834 X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835

cating that electronic effect exists in this photocatalytic reaction. Acknowledgements


The conversions of the regioisomer increased in the order of
orth < meta < para isomer, indicating the presence of steric effect This work was funded by the National Natural Science Founda-
(Table 4, entries 3–5). In addition, it should be mentioned that tion of China (grant number 21773173) and the start-up fund of
the heterocyclic amine containing N or S atom which usually poi- Wuhan University.
son most transition-metal catalysts can also be efficiently trans-
formed into corresponding imines (Table 4, entries 12 and 13).
The performance for the photocatalytic selective oxidation of Appendix A. Supplementary data
primary amines were very encouraging. Next, the scope of sub-
strates was further explored to include secondary amines and the Supplementary data to this article can be found online at
results were listed in Table 5. It was found that secondary amines https://doi.org/10.1016/j.jcis.2020.07.115.
can also be well transformed into corresponding imines with
extended reaction time but the selectivities for products were
declined to some extent. The electronic effect and steric effect References
are also present in the process and follow the same rules with
selective oxidation of primary amines. In addition to the excellent [1] M. Bellardita, F. Parrino, E.I. García-López, G. Marcì, V. Loddo, L. Palmisano,
Heterogeneous photocatalysis for selective formation of high value-added
photocatalytic activity, heterogeneous photocatalyst should have molecules: some chemical and engineering aspects, ACS Catal. 8 (2018)
unique advantage that easy to separate from the reaction system 11191–11225.
compared to homogeneous photocatalyst. Therefore, the recycla- [2] V. Balzani, G. Bergamini, P. Ceroni, Light: A Very Peculiar Reactant and Product,
Angew. Chem. Int. Ed. 54 (2015) 11320–11337.
bility of 1,2-DHA-TiO2 was evaluated for the photocatalytic selec- [3] G. Segev, J.W. Beeman, J.B. Greenblatt, I.D. Sharp, Hybrid photoelectrochemical
tive oxidation of benzylamine. It was found that only a <10% and photovoltaic cells for simultaneous production of chemical fuels and
decrease in conversions was observed after three cycles, all with electrical power, Nat. Mater. 17 (2018) 1115–1121.
[4] Y.A. Wu, I. McNulty, C. Liu, K.C. Lau, Q. Liu, A.P. Paulikas, C.J. Sun, Z.H. Cai, J.R.
a selectivity of more than 99% (Fig. S1). Importantly, the powder Guest, Y. Ren, V. Stamenkovic, L.A. Curtiss, Y.Z. Liu, T. Rajh, Facet-dependent
X-ray diffraction (PXRD) patterns indicated that recycle experi- active sites of a single Cu2O particle photocatalyst for CO2 reduction to
ments did not change the crystal structure of TiO2 (Fig. S2). methanol, Nat. Energy 4 (2019) 957–968.
[5] L. Chen, J. Tang, L.N. Song, P. Chen, J. He, C.T. Au, S.F. Yin, Heterogeneous
photocatalysis for selective oxidation of alcohols and hydrocarbons, Appl.
Catal. B: Environ. 242 (2019) 379–388.
[6] K. Lee, H. Yoon, C. Ahn, J. Park, S. Jeon, Strategies to improve the photocatalytic
4. Conclusions activity of TiO2: 3D nanostructuring and heterostructuring with graphitic
carbon nanomaterials, Nanoscale 11 (2019) 7025–7040.
In conclusion, a strategy of molecular inverse design of ARS was [7] H. Guo, N. Jiang, H.J. Wang, K.F. Shang, N. Lu, J. Li, Y. Wu, Enhanced catalytic
performance of graphene-TiO2 nanocomposites for synergetic degradation of
adopted to identify the important factor determining the superior fluoroquinolone antibiotic in pulsed discharge plasma system, Appl. Catal. B:
photocatalytic activity for ARS-TiO2 assembly. The substituted –OH Environ. 248 (2019) 552–566.
groups of AQ provide the inherent visible light absorption and the [8] V. Kurnaravel, S. Mathew, J. Bartlett, S.C. Pillai, Photocatalytic hydrogen
production using metal doped TiO2: A review of recent advances, Appl. Catal.
binding sites. Ultimately, in situ formed 1,2-DHA-TiO2 assembly B: Envrion. 244 (2019) 1021–1064.
can also be powerful photocatalyst for green light-induced selec- [9] E. Sundin, M. Abrahamsson, Long-lived charge separation in dye-
tive oxidation of a wide range of amines into imines with aerial semiconductor assemblies: a pathway to multi-electron transfer reactions,
Chem. Commun. 54 (2018) 5289–5298.
O2. More importantly, the molecular features of 1,2-DHA con-
[10] S.J. Woo, S. Choi, S.Y. Kim, P.S. Kim, J.H. Jo, C.H. Kim, H.J. Son, C. Pac, S.O. Kang,
tribute mostly to the unique photocatalytic activity after binding Highly selective and durable photochemical CO2 reduction by molecular Mn(I)
with TiO2 with broad visible light absorption which can be main- catalyst fixed on a particular dye-sensitized TiO2 platform, ACS Catal. 9 (2019)
tained at high concentration of amines. In the end, 1,2-DHA, a nat- 2580–2593.
[11] J. Willkomm, K.L. Orchard, A. Reynal, E. Pastor, J.R. Durrant, E. Reisner, Dye-
ural organic dye with strong visible light absorption, turned out to sensitised semiconductors modified with molecular catalysts for light-driven
be an efficient visible light photocatalyst in cooperation with TiO2. H2 production, Chem. Soc. Rev. 45 (2016) 9–23.
This work makes evident the promise of molecular design in tailor- [12] X.H. Zhang, L.J. Yu, C.S. Zhuang, T.Y. Peng, R.J. Li, X.G. Li, Highly efficient visible/
near-IR-light-driven photocatalytic H2 production over asymmetric
ing dye-semiconductor assemblies. Future work to construct dye- phthalocyanine-sensitized TiO2, RSC Adv. 3 (2013) 14363–14370.
semiconductor assemblies based on AQs could deepen our under- [13] O. Bettucci, T. Skaltsas, M. Calamante, A. Dessi, M. Bartolini, A. Sinicropi, J.
standing of interfacial charge transfer between an organic molecu- Filippi, G. Reginato, A. Mordini, P. Fornasiero, L. Zani, Combining
dithienosilole-based organic dyes with a brookite/platinum photocatalyst
lar dye and a semiconductor and lead to more compelling results toward enhanced visible-light-driven hydrogen production, ACS Appl. Energy
towards visible light-induced selective chemical transformations. Mater. 2 (2019) 5600–5612.
[14] N. Bao, Y. Li, X.H. Yu, J.J. Niu, G.L. Wu, X.H. Xu, Removal of anionic azo dye from
aqueous solution via an adsorption-photosensitized regeneration process on a
TiO2 surface, Environ. Sci. Pollut. Res. 20 (2013) 897–906.
5. Author statement [15] Y.M. Cho, W.Y. Choi, C.H. Lee, T. Hyeon, H.I. Lee, Visible light-induced
degradation of carbon tetrachloride on dye-sensitized TiO2, Environ. Sci.
Technol. 35 (2001) 966–970.
X. Lang carried out the funding acquisition, conceptualization, [16] X. Li, J.L. Shi, H.M. Hao, X.J. Lang, Visible light-induced selective oxidation of
and supervision for the project. X. Li performed most of the inves- alcohols with air by dye-sensitized TiO2 photocatalysis, Appl. Catal. B: Environ.
232 (2018) 260–267.
tigation and formal analysis. H. H. assisted in the process of inves- [17] X.J. Lang, X.D. Chen, J.C. Zhao, Heterogeneous visible light photocatalysis for
tigation and formal analysis. X. Li wrote the original draft. X. Lang selective organic transformations, Chem. Soc. Rev. 43 (2014) 473–486.
reviewed and edited the final draft with input from the other [18] W.L. Sheng, J.L. Shi, H.M. Hao, X. Li, X.J. Lang, Selective aerobic oxidation of
sulfides by cooperative polyimide-titanium dioxide photocatalysis and
authors. triethylamine catalysis, J. Colloid Interface Sci. 565 (2020) 614–622.
[19] J.B. Shi, J.L. Zhang, T.L. Liang, D.X. Tan, X.N. Tan, Q. Wan, X.Y. Cheng, B.X. Zhang,
B.X. Han, L.F. Liu, F.Y. Zhang, G. Chen, Bipyridyl-containing cadmium-organic
frameworks for efficient photocatalytic oxidation of benzylamine, ACS Appl.
Declaration of Competing Interest Mater. Interfaces 11 (2019) 30953–30958.
[20] B. Chen, L. Wang, S. Gao, Recent advances in aerobic oxidation of alcohols and
The authors declare that they have no known competing finan- amines to imines, ACS Catal. 5 (2015) 5851–5876.
[21] X.J. Lang, H.W. Ji, C.C. Chen, W.H. Ma, J.C. Zhao, Selective formation of imines
cial interests or personal relationships that could have appeared by aerobic photocatalytic oxidation of amines on TiO2, Angew. Chem. Int. Ed.
to influence the work reported in this paper. 50 (2011) 3934–3937.
X. Li et al. / Journal of Colloid and Interface Science 581 (2021) 826–835 835

[22] D.R. Sun, L. Ye, Z.H. Li, Visible-light-assisted aerobic photocatalytic oxidation of [30] Z. Wang, X.J. Lang, Visible light photocatalysis of dye-sensitized TiO2: The
amines to imines over NH2-MIL-125(Ti), Appl. Catal. B: Environ. 164 (2015) selective aerobic oxidation of amines to imines, Appl. Catal. B: Environ. 224
428–432. (2018) 404–409.
[23] L. Ye, Z.H. Li, ZnIn2S4: a photocatalyst for the selective aerobic oxidation of [31] Z.Q. Zhang, L.L. Bai, Z.J. Li, Y. Qu, L.Q. Jing, Review of strategies for the
amines to imines under visible light, ChemCatChem 6 (2014) 2540–2543. fabrication of heterojunctional nanocomposites as efficient visible-light
[24] Z.B. Yu, E.R. Waclawik, Z.J. Wang, X.M. Gu, Y. Yuan, Z.F. Zheng, Dual catalysts by modulating excited electrons with appropriate thermodynamic
modification of TiNb2O7 with nitrogen dopants and oxygen vacancies for energy, J. Mater. Chem. A. 7 (2019) 10879–10897.
selective aerobic oxidation of benzylamine to imine under green light, J. Mater. [32] L.Q. Jing, W. Zhou, G.H. Tian, H.G. Fu, Surface tuning for oxide-based
Chem. A. 5 (2017) 4607–4615. nanomaterials as efficient photocatalysts, Chem. Soc. Rev. 42 (2013) 9509–
[25] P. Chen, L.H. Meng, L. Chen, J.K. Guo, S. Shen, C.T. Au, S.F. Yin, Double-shell and 9549.
flower-like ZnS-C3N4 derived from in situ supramolecular self-assembly for [33] R.F. Qian, H.X. Zong, J. Schneider, G.D. Zhou, T. Zhao, Y.L. Li, J. Yang, D.W.
selective aerobic oxidation of amines to imines, ACS Sustainable Chem. Eng. 7 Bahnemann, J.H. Pan, Charge carrier trapping, recombination and transfer
(2019) 14203–14209. during TiO2 photocatalysis: An overview, Catal. Today 335 (2019) 78–90.
[26] J.L. Shi, R.F. Chen, H.M. Hao, C. Wang, X.J. Lang, 2D sp2 carbon-conjugated [34] X.J. Lang, J.C. Zhao, Integrating TEMPO and its analogues with visible-light
porphyrin covalent organic framework for cooperative photocatalysis with photocatalysis, Chem. Asian J. 13 (2018) 599–613.
TEMPO, Angew. Chem. Int. Ed. 59 (2020) 9088–9093. [35] B.L. Ryland, S.S. Stahl, Practical aerobic oxidations of alcohols and amines with
[27] J.L. Shi, H.M. Hao, X.J. Lang, Phenol-TiO2 complex photocatalysis: visible light- homogeneous copper/TEMPO and related catalyst systems, Angew. Chem. Int.
driven selective oxidation of amines into imines in air, Sustainable Energy Ed. 53 (2014) 8824–8838.
Fuels 3 (2019) 488–498. [36] Y. Nosaka, A.Y. Nosaka, Generation and detection of reactive oxygen species in
[28] J.L. Shi, H.M. Hao, X. Li, X.J. Lang, Merging the catechol-TiO2 complex photocatalysis, Chem. Rev. 117 (2017) 11302–11336.
photocatalyst with TEMPO for selective aerobic oxidation of amines into
imines, Catal. Sci. Technol. 8 (2018) 3910–3917.
[29] X. Li, H. Xu, J.L. Shi, H.M. Hao, H. Yuan, X.J. Lang, Salicylic acid complexed with
TiO2 for visible light-driven selective oxidation of amines into imines with air,
Appl. Catal. B: Environ 244 (2019) 758–766.

You might also like