A sol–gel biotemplating route with cellulose nanocrystals to design a photocatalyst for improving hydrogen generation

Cong Wang a, Jian Li a, Erwan Paineau b, Abdelghani Laachachi c, Christophe Colbeau-Justin a, Hynd Remita ad and Mohamed Nawfal Ghazzal *a
aUniversité Paris-Saclay, Institut de Chimie Physique, UMR 8000 CNRS, 91405 Orsay, France. E-mail: mohamed-nawfal.ghazzal@universite-paris-saclay.fr
bUniversité Paris-Saclay, Laboratoire de Physique du Solide, UMR 8502 CNRS, 91405 Orsay, France
cMaterials Research and Technology Department (MRT), Luxembourg Institute of Science and Technology (LIST), 5 Rue Bommel - ZAE Robert Steichen, 4940 Käerjeng, Luxembourg
dCNRS, Institut de Chimie Physique, UMR 8000, 91405 Orsay, France

Received 18th November 2019 , Accepted 23rd February 2020

First published on 25th February 2020

Light harvesting capability and charge carrier lifetime play critical roles in determining the photoefficiency of a photocatalyst. Herein, a one-pot method is proposed to design mesostructured TiO2 materials by taking advantage of the ability of cellulose nanocrystals (CNCs) to self-assemble into chiral nematic structures during solvent evaporation. After the xerogel formation, the as-obtained CNC/TiO2 hybrid films exhibit a chiral nematic structure and tunable Bragg peak reflection, generating a lamellar TiO2 mesostructure after the biotemplate removal by calcination. More prominently, this straightforward method can be extended to couple TiO2 with other metal oxides, improving the light-harvesting and charge carrier separation of these photocatalysts, in particular for improving hydrogen generation. This foolproof approach opens new doors for the development of nanostructured materials for solar energy conversion and catalysis.

1. Introduction

Known as the bottleneck for solar energy conversion efficiency, the capability of light harvesting and the separation of photogenerated charge carriers have attracted great interest.1 Structuring photoactive materials into photonic crystals or hierarchical structures is a promising approach for improving the efficiency of photoactive materials. A photonic crystal can be used either as a reflector element separated from the photoactive layer for light accumulation,2–4 or as a photoactive absorber to slow down the velocity of light.5,6 Hierarchical structures, like in green leaves, can improve light scattering for optimal light management toward an efficient conversion of photons into chemical energy.7 In this context, structuring inorganic materials on the mesoscale using a self-biotemplate assembly process appears as a sustainable method for the synthesis of functional materials. Notably, cellulose nanocrystals (CNCs, as a biotemplate) have the specificity of forming a chiral nematic (CN) structure, which can be replicated into inorganic materials, a structure never achieved before with classical aqueous sol–gel approaches using surfactants as templates.8,9 Freestanding silica films with a CN structure were obtained by the polycondensation of tetraethyl orthosilane or methyl triethoxisilane precursors dissolved in the CNC aqueous solution.10,11 This breakthrough has led to promising methods for the development of functional materials and the design of novel photonic crystals. However, the method is restricted only to mesoporous silica or organosilica films, which limits the application to the sensing and separation of chiral molecules.12,13 This limitation stems from the high sensitivity of the other precursors (metal chlorides or alkoxides) to moisture leading to their fast hydrolysis and condensation, destabilising the chiral nematic structure of CNCs into the isotropic phase. Acid catalysis in a sol–gel process is one method for stabilizing titania precursors in an aqueous solution containing CNCs by promoting slow polycondensation but with the expense of limited hydrolysis.14–16 These infructuous attempts lead to a colorless mesoporous TiO2 film with the complete loss of the CN structures of CNCs. The CN organization in CNCs is very sensitive to the solvent polarity, ionic strength and pH of the aqueous solution.17 Other alternatives, such as hard templating or impregnation methods, have been proposed to transfer the helical structure into titanium dioxide films.18–20 Despite their efficiency and versatility, all these methods require multistep impregnation/drying to achieve the final material. Recently, a successful self-assembly of CNC/TiO2 was proposed.21 The procedure requires complex steps to dissolve TiO2 nanoparticles and use the sodium titanate precursor (unsuitable for photocatalysis or dye sensitized solar cells) to elaborate a water-soluble peroxotitanate in the CNC aqueous solution. The remaining sodium ion is known to have a detrimental effect on the photoefficiency of TiO2, as it acts as a recombination center for the photogenerated charge carriers.22,23 Clearly, it is challenging and highly desirable to develop an effective yet simple one-pot sol–gel self-biotemplating approach that can be extended to TiO2 and other functional metal oxides.

Here, we propose a straightforward one-pot method enabling both the structural design of TiO2 films and the coupling with metal oxides to improve the photon conversion efficiency. The one-pot self-biotemplating sol–gel approach enables the synthesis of CNC/TiO2 and CNC/mixed oxide hybrid films with tunable Bragg peak reflection. A stable mixture of titanium diisopropoxide bis(acetylacetonate) (TAA) and an aqueous solution of CNCs was used for the elaboration of the films. After removing the CNC biotemplate, the final TiO2 films exhibit a birefringent lamellar structure producing an increase of the photogenerated charge carrier lifetime. The charge carrier lifetime and the photocatalytic activity for hydrogen production were found to be significantly improved, compared to those of commercial TiO2-P25 (employed here as an internal standard). To validate the reliability and universality of our method, the procedure was extended to TiO2 mixed with different metal oxides (Cu, Ni, Bi, and V) to increase the charge carrier separation, improving the photocatalytic efficiency for hydrogen generation.

2. Experimental section

2.1 Synthesis of mesoporous lamellar TiO2 films

A CNC aqueous suspension (10 mL, 4.0 wt%, pH = 6.3) was sonicated for 1 h. Different volumes of TAA solutions (Table S1) were added to 2 mL ethanol and stirred for 10 min, then the ethanolic TAA solutions were mixed with the sonicated CNC suspension. After a slight shaking and stirring for 1 h a homogeneous bright yellow mixture was obtained, which was poured into 55 mm diameter polystyrene Petri dishes and dried at room temperature for about 48 h to obtain iridescent CNC/TiO2 hybrid films. The resulting hybrid films were then calcined in air at a heating rate of 1 °C min−1 to 500 °C for 2 h and cooled down to room temperature yielding white mesoporous lamellar TiO2 films.

2.2 Synthesis of mixed oxide films with a lamellar structure

The same CNC aqueous suspension was sonicated for 1 h. 1 mg of copper(II) acetate or nickel(II) acetylacetonate was added to 1 mL ethanol (in the case of bismuth(III) nitrate and vanadium(V) oxide, they were dissolved in deionized water) and sonicated until clear solutions were obtained. 0.2 mL TAA solution was added to 1 mL ethanol and stirred for 10 min, then mixed with metal oxide solutions already prepared and stirred for another 10 min. Then, the solution containing the precursors is added to the CNC suspension and stirred for 1 h. The mixtures were poured into polystyrene Petri dishes to form the hybrid films. Finally, the same calcination procedure as detailed above is performed.

3. Results and discussion

The ester sulfated surface of CNCs (CNC-OSO3H) ensures the stability of the nanorods by electrostatic repulsion in aqueous suspensions. The ability of the negatively charged CNC nanorods to form CN phases relies on several parameters, such as the concentration of the sulfate groups, the volume fraction of CNCs and the pH of the medium.24,25 4 wt% CNC is the critical concentration corresponding to the onset of isotropic to chiral nematic phase transition.26 The cross-polarized optical microscopy (POM) image of this solution is shown in Fig. 1a. At this concentration, the suspension is weakly birefringent and tactoids (liquid crystalline droplets) with different sizes and shapes dispersed in a continuous isotropic phase have been observed (Fig. 1c). Fig. 1c shows the typical POM image of the CNC suspension containing tactoids with spherical and ellipsoidal shapes. These tactoids are considered as the primitive components of liquid crystals. They have parallel birefringent bands, similar to the chiral nematic structure observed for the bulk CNC suspension (Fig. 1d). Tactoids spontaneously merge with each other and form an iridescent chiral nematic structure in the films.26 Both the CNC and hybrid TAA/CNC suspensions exhibit non-persistent strong shear-induced polychromatic birefringence between crossed polarizers, related to pretransitional effects (Fig. 1a and b).
image file: c9ta12665a-f1.tif
Fig. 1 Photographs of (a) pure CNC and (b) TAA/CNC (weight ratio = 0.5) suspensions observed between crossed polarizers before and during agitation (shear-induced polychromatic birefringence). Typical POM images of (c) CNC and (d) TAA/CNC suspensions during slow evaporation at room temperature (scale bar = 50 μm).

CNC/TiO2 hybrid mesostructured films are prepared by evaporation induced self-assembly (EISA) of the solution containing the ethanolic solution of the Ti(acac)2OiPr2 precursor (TAA) and CNC suspension (4 wt%). The mixture is highly miscible after agitation, and a stable water-based TiO2-precursor solution (yellow and transparent) can easily be obtained (Fig. 1b). However, the solution does not show any birefringence, indicating isotropic phase transition due to the dilution induced by the addition of the precursor. In the TAA precursor, acetylacetonate acts as a chelating agent, which limits the precursors' reactivity towards hydrolysis and polycondensation, making the solution stable enough during the EISA process. The pH of the hybrid solution varies from 6.26 to 6.04 when the TAA ratio increases from 0.5 to 2.5, corresponding to the initial pH of the CNC suspension (pH = 6.3). Droplet evaporation of the TAA/CNC hybrid suspensions observed using a POM is presented in Fig. 1d and S1. The formation of typical fingerprint textures clearly indicates the presence of a CN structure during the formation of CNC/TiO2 hybrid films. This feature is confirmed by performing time-lapse photography of the TAA/CNC suspension during the EISA process, showing the formation of an iridescent film with time due to the local ordering of CNCs in the CN structure (Fig. S2). Hence, our straightforward one-pot method enables elaborating the CNC/TiO2 photonic hybrid films with tunable colors and a variable maximum of Bragg peak reflection (from the UV to the entire visible range). Although the CNC films are flexible with limited cracks, the photonic hybrid films are likely to crack into fragments whose size gets smaller as the concentration of the TAA precursor increases. This is related to the water evaporation during EISA, which induces a shrinkage through the polycondensation of the titania precursor, generating capillary stress gradients within the films. The iridescent color rising from the films under natural light indicates the formation of the periodic structure reflecting a selective wavelength range. The color of the films turns from blue, green to yellow (Fig. 2a), in agreement with the red-shifted maximum wavelength of the Bragg reflection (Fig. 2b). Increasing the TAA/CNC weight ratio induces a red-shift of the Bragg peak reflection as evidenced by UV-vis-NIR spectroscopy (Fig. 2b). The maximum wavelength reflected by the hybrid photonic films (λmax) is related to the effective refractive index “neff”, the pitch “P” and the angle of the film axis to the emitted light following the equation:27

λmax = neffP[thin space (1/6-em)]sin(θ)(1)

image file: c9ta12665a-f2.tif
Fig. 2 (a) Photographs and (b) UV-vis spectra of the CNC/TiO2 hybrid films. SEM images of the CNC/TiO2 hybrid films: (c) weight ratio = 0.5, (d) weight ratio = 1.5, and (e) weight ratio = 2.5 (scale bar = 1 μm). POM images of the CNC/TiO2 hybrid films: (f) weight ratio = 0.5, (g) weight ratio = 1.5, and (h) weight ratio = 2.5 (scale bar = 50 μm).

Since the refractive index of the pure titania xerogel and CNCs remains unchanged (nTiO2 = 1.9 (ref. 28) and nCNC = 1.51 (ref. 29)), the red-shift would probably increase with the increase of the pitch “P”. To shed light on our assumption, the cross-sections of the films are analysed by SEM.

The resulting CN structures from the CNC/TiO2 hybrid films are further evidenced by the SEM imaging of the film cross-sections (Fig. 2c–e, S3a and b). A twisted structure, periodically repeated along the film axis, is observed whatever the composition of the film. This is a typical feature of the chiral nematic organization, in which the distance between two layers corresponds to the half-helical pitch “P/2” of the CN structure.30 Values of the half-helical pitch estimated from these SEM images are found to increase with the titanium loading, confirming the effect of the TAA/CNC ratio on the observed λmax red-shift (Table S1). Indeed, as suggested above, increasing of the P value, likely due to the increasing of the titania wall thickness (increase of the TAA concentration), induces a red-shift of the maximum Bragg peak reflection, in agreement with the UV-visible spectroscopic analysis (Fig. 2b). The photonic films have been further characterized using POM (Fig. 2f–h, S3c and d). All the films show a strong birefringence but a random distribution of the dark and bright regions, suggesting a different orientation of the helical axis towards the up-coming light (random orientation of the liquid crystal domains). It should be noted that typical fingerprint textures are observed in POM images, which definitively confirm the preservation of the chiral nematic ordering in such as-prepared hybrid photonic films regardless of the titanium loading.

The inorganic TiO2 films are obtained after removal of the CNC template by annealing the samples at 500 °C under air following the procedure described in Fig. 3a. Calcination can effectively remove the CNC from the hybrid films, as evidenced by FTIR spectroscopy (only the band corresponding to the Ti–O bond is observed, Fig. S4) producing mesoporous TiO2 films and ensuring the crystallisation of TiO2 into its photoactive anatase phase. Nitrogen physisorption measurements confirm the mesoporous structure of the calcined films. Isotherms are reversible and the type IV characteristic hysteresis loop (obvious for a higher TAA concentration) suggests interconnected pores with a cylindrical shape (Fig. S5a). A well-defined progressive uptake measure at a low relative pressure (p/p0 = 0.05–0.40) is characteristic of micropores, followed by a sudden increase due to the filling of the pores in the mesoporous range. The BET specific surface area reaches 64.5–72.6 m2 g−1 and the pore volume ranges from 0.35 to 0.14 cm3 g−1. The narrower pore size distribution implies identical pores for all the films, with an average pore diameter of 10 nm, a value in agreement with the CNCs' width (Table S2 and Fig. S5b).14 Wide-angle X-ray scattering (WAXS) patterns performed on hybrid photonic films before calcination display several peaks at Q ∼ 1.1 Å−1 [(1[1 with combining macron]0), (110)] and 1.6 Å−1 (200), characteristics of the crystalline form of CNCs (Fig. 3b).31 After calcination, these peaks are no longer present in the calcined films but instead, we can observe the occurrence of broad diffraction peaks corresponding to the nanosized TiO2 crystallites in the anatase form (Fig. 3c). The calcined films observed by POM as presented in Fig. 3d show remarkable birefringent textures implying a preserved anisotropic structure (Fig. S6). In agreement with this statement, SEM images of the cross-sections of TiO2 films reveal a transfer of the alignment layers to mesostructured lamellar TiO2 (referenced as L-TiO2 hereafter) with a long-range order (Fig. 3e and f). The layers parallel to the substrate are aligned predominantly with highly anisotropic structural features and the thickness of the layer increases with the TAA loading. The replica of the CN structure could not be confirmed from the SEM images, probably due to the crystal growth and sintering of the TiO2, or the movement of atoms occurring during the calcination step leading to the change of the chiral nematic structure to a 1D lamellar one. These results evidence the change of the CN structure in CNC/TiO2 hybrid films to inorganic TiO2 films with a lamellar mesostructure.

image file: c9ta12665a-f3.tif
Fig. 3 (a) Schematic illustration of the fabrication of a lamellar TiO2 structure (L-TiO2) after removal of the biotemplate by calcination. WAXS patterns of (b) the CNC/TiO2 hybrid film and (c) L-TiO2 films. (d) POM image of the L-TiO2-5 film showing strong birefringence (scale bar = 100 μm). SEM images of (e) the L-TiO2-2 and (f) L-TiO2-5 films (scale bar = 1 μm). (g) Hydrogen generation and (h) TRMC signals of the L-TiO2 films and TiO2-P25 powder.

The photocatalytic activity of the periodic long-range L-TiO2 films is evaluated through hydrogen generation. The L-TiO2 films show higher photoefficiency compared to that of standard TiO2-P25 whatever the TAA/CNC ratio of the films (Fig. 3g). The H2 generation rate of L-TiO2-1 is 287.2 μmol h−1 g−1, as nearly twice as high as TiO2-P25 (153.1 μmol h−1 g−1). This result indicates that structuring the TiO2 material in a periodic long-range lamellar structure can considerably improve its photoefficiency for hydrogen generation. Interestingly, the TAA/CNC ratio of the starting material is inversely related to the photocatalytic efficiency of hydrogen generation. Consideration should be given to the interaction of light with these structured photocatalysts to understand deeply the relationship underlying between these results. Indeed, the light reflection in a periodically structured material is reported to be an efficient strategy to improve the photon-to-charge carrier production.18 For instance, the photon velocity slows down significantly at the photonic band gap edges (blue and red) in the photonic structure, increasing the absorbance light factor of TiO2.32 Bioinspired porous TiO2 materials replicating the chloroplast structure obtained from plant leaves show enhanced light harvesting capability.7,33

The electronic properties of variable L-TiO2 films are assessed by using time-resolved microwave conductivity (TRMC). The TRMC is a non-destructive and a contactless technique that enables tracking the dynamics, the density and the lifetime of the photogenerated charge carriers on such periodic long-range lamellar mesostructured TiO2 films.34,35 After illumination, the TRMC signals rise following an ultrafast charge separation process (<200 fs),36 which is mainly due to mobile electrons at the surface of TiO2, since it is usually assumed that holes with a limited mobility remain on the bulk. The time-dependent change of microwave power transmission ΔP(t)/P is proportional to the photoconductivity Δσ(t) variation and therefore proportional to the density of mobile charge carriers, which can be expressed as:37,38

image file: c9ta12665a-t1.tif(2)
where Δni(t) defines the density of photogenerated charge carriers “i” at known time “t”, and μi defines their mobility. In the present study, all TRMC measurements are recorded under 360 nm UV illumination, and with a light energy density received by the samples set at 1.4 mJ cm−2. According to the TRMC results, the photogenerated charge carriers associated with the TRMC signal gradually increase as the TAA/CNC ratio increases (Fig. 3h). Assuming an equal number of absorbed photons for all samples, we surmise that the improvement in the density of the photogenerated charge carriers may be related to the light scattering in the long-range lamellar TiO2 structure. As depicted in Fig. 3e and f, the lamellar structure could be defined as alternating TiO2 layers separated by voids, having a variable refractive index. The thickness of the TiO2 layers increases with the ratio of TAA/CNC, which improves the confinement time of photons and their reflection. Therefore, the light propagation in such a lamellar mesostructure improves the light harvesting and consequently the population of photogenerated charge carriers. The TRMC results show an increase in the signal amplitude (which is correlated with the density of charge carriers e/h+) with the TAA/CNC ratio (Fig. 3h), corresponding to an inverse correlation of H2 production efficiency. For example, L-TiO2-1 displays twice the hydrogen production efficiency of L-TiO2-5, while the population of the photogenerated charge carriers is ten times less than that of L-TiO2-5. This result suggests that the density of charge carriers is probably not the limiting parameter for the H2 generation.

Thus, we turn to investigate the TRMC signals' decay, which typically occurs after the e/h+ production during the 8 ns laser pulse, giving useful information about the photocharge lifetime (Fig. 3h). The decay of the TRMC signals of photocharges can be divided into three decay stages: charge trapping (τtrap), charge recombination (τrec) and surface reactions (τsurf). The capture time of electrons and holes occurs in less than 30 ps, and it is accompanied by rapid recombination.39,40 Both of the trapping and recombination phenomena are considered as the main processes that can explain the TRMC signals' decay.41,42 The time constant of electron trapping and recombination can be quantitatively estimated from the signal decay using the following equation:37

image file: c9ta12665a-t2.tif(3)
where An and τn are the proportionality and time constants of each of the three decay processes, respectively. The photocharge lifetime presented in the inset of Fig. 3h is found to be correlated with the wall thickness of the lamellar structure. The low charge carrier number produced at the surface of the L-TiO2-1 sample has the longest lifetime, which could be translated by low trapping and recombination yield. This trend evidences that L-TiO2-1 synthesized at a lower TAA/CNC ratio shows weak electron trapping capability, thereby enhancing the proportion of active charges for H2 generation. The interfacial charge recombination is likely to occur in the boundaries (interfaces) of the long-range layers with thicker walls where defects could be localized.

Owing to the interesting photoactivity of L-TiO2-1 for hydrogen generation, several coupling schemes (Cu, Ni, Bi, and V) have been performed to improve the photoefficiency yield even further. Metal oxide coupled to TiO2 can trap electrons, which impedes the recombination of e/h+ resulting in a significant enhancement of photocatalytic activity.43,44 The solubility of the metal precursors combined with the one-pot self-assembly method enables elaborating biotemplated coupled metal oxides. The photonic colors rising from the light reflection at the surface of hybrid films (CNC/coupled metal oxides) do not show a notable variation (Fig. S7), which is consistent with the slight red-shift in the Bragg peak reflection (Fig. 4a). SEM images evidence the remaining of the twisted periodic structure, as already observed for the CNC/TiO2 hybrid films (Fig. 4b). This shows for the first time that the proposed one-pot self-assembly method could be extended to the elaboration of coupled metal oxides with a mesostructured architecture.

image file: c9ta12665a-f4.tif
Fig. 4 (a) UV-vis spectra and (b) SEM images of the CNC/coupled metal oxide hybrid films (scale bar = 1 μm).

The compositions of the mixed oxides with a lamellar structure are confirmed by X-ray photoelectron spectroscopy (XPS). The general XPS spectra of the films show the presence of carbon and oxygen elements (Fig. S8). The films show the Ti 2p peak deconvoluted into two curves, which can be assigned to the spin–orbit components 2p3/2 and 2p1/2, respectively. The separations between the Ti 2p3/2 and Ti 2p1/2 peaks are measured to be 5.7 eV for all the films, in agreement with the binding energy (BE) separation observed for anatase TiO2 (Fig. 5a).45,46 The XPS spectra of the target metal oxides are presented in Fig. 5b–e. In the case of bismuth oxide coupled with L-TiO2-1, symmetric peaks of the Bi 4f state are observed. The Bi 4f7/2 and Bi 4f5/2 asymmetric bands are both resolved into two bands, at 159 eV and 165 eV, respectively (Fig. 5b).47 These orbitals correspond to Bi3+ oxide state in Bi2O3, and the slight shift to lower binding energy could be due to oxygen vacancies.48 The XPS spectra of L-TiO2/Cu show characteristic peaks of the Cu 2p3/2 and Cu 2p1/2 levels at BEs 932.5 eV and 952.1 eV, respectively, corresponding to Cu(I) species in cuprous oxide.49,50 The weak satellite observed at 945 eV also demonstrated the presence of Cu(I) (Fig. 5c). In the case of Ni oxides (L-TiO2/Ni), two peaks at BE of 873.7 eV and 855.1 eV correspond to Ni 2p1/2 and Ni 2p3/2, respectively, and their corresponding satellites at 880 eV and 861 eV (Fig. 5d). These peaks are assigned to Ni2+ in Ni(OH)2 (derived from the oxidation of NiO).51 As seen in Fig. 5e, the BEs of V 2p3/2, V 2p1/2 and O 1s are 517.4, 524.8 and 530 eV, respectively, which agree well with those of V5+ and O2− in V2O5.52 The films are observed using transmission electron microscopy (TEM) and the results are shown in Fig. S9. The TEM micrograph shows a porous structure in agreement with the BET analysis. However, the lamellar structure of the films is not observed mainly because of the layers are parallel to the substrate. The UV-vis absorption spectra of L-TiO2 and different metal oxides coupled to TiO2 are shown in Fig. S10. The L-TiO2 is observed at about 400 nm, characteristic of the absorption of the anatase form. After coupling L-TiO2 with different metal oxides, a redshift is observed confirming the contribution of the mixed oxides. The maximum red-shift is observed for copper oxide and vanadium oxide compared to nickel and bismuth oxides.

image file: c9ta12665a-f5.tif
Fig. 5 (a) XPS spectra of Ti 2p of different coupled metal oxides. XPS spectra of (b) L-TiO2/Bi, (c) L-TiO2/Cu, (d) L-TiO2/Ni and (e) L-TiO2/V.

Metal oxides coupled to L-TiO2-1 show variable photoefficiencies (Fig. 6a). For instance, L-TiO2/Cu exhibits the highest H2 generation rate, 7.88 mmol h−1 g−1, which is 40-times higher than that of solely L-TiO2-1. The H2 generation rates using Ni and Bi are, 1.0 mmol h−1 g−1 and 0.9 mmol h−1 g−1, respectively, twice that obtained from L-TiO2-1. However, when V is involved, the photocatalytic activity deteriorates, implying the inactivation of L-TiO2. According to the TRMC investigations of these photocatalysts, L-TiO2-1 mixed with either Cu or V shows no signal (Fig. 6b). In the first case, Cu acts as an electron collector for better charge carrier separation and accelerates the reaction rate of hydrogen reduction, while the V5+ state is considered as a recombination center that inhibits the H2 production.53,54 The position of the Cu2O conduction band enables the thermodynamic transfer of electrons, resulting in charge carriers separation.55–57 The recombination of e/h+ in TiO2 is prevented by efficient inter-particle charge transfer leading to improve the hydrogen generation efficiency of L-TiO2/Cu.58,59 Furthermore, the BE of Ti 2p shifts to a lower energy (0.3 eV) in L-TiO2/Cu, suggesting a partial surface reduction through the formation of oxygen vacancies and a lower Ti oxidation state (Ti3+), which was reported to improve the photocatalytic activity of TiO2.60,61 Conversely, L-TiO2/V exhibits a slightly higher BE of Ti 2p (0.6 eV) forming a higher average oxidation state of Ti ions (Fig. 5a).62 NiO and Bi2O3 (2.8 eV (ref. 63)), can form a p–n junction with TiO2, generating an internal electric field, thereby inhibiting the e/h+ recombination in a similar way.64 A similar conclusion can be drawn from the TRMC signal (Fig. 6c), where the increase of the decay rate evidences efficient electron capture.

image file: c9ta12665a-f6.tif
Fig. 6 (a) Hydrogen generation of mixed oxide films with a lamellar structure. TRMC signals of (b) L-TiO2/Cu and L-TiO2/V, (c) L-TiO2/Bi and L-TiO2/Ni.

4. Conclusions

In summary, we report the first demonstration of a simple and reliable one-pot sol–gel self-assembly method for the design of a layered photocatalyst with a controlled architecture. The co-assembly of the water-soluble titanium diisopropoxide bis(acetylacetonate) precursor and the lyotropic cellulose nanocrystal dispersion allows the formation of a CNC/titania hybrid material containing a chiral nematic structure. Successful removal of the biotemplate generates a lamellar mesostructure whose birefringence increases due to its anisotropic nature. This kind of long-range ordered lamellar structure can improve the capability of light harvesting and extension of the charge carrier lifetime. The straightforward self-assembly method can be easily applied to synthesize metal oxide/TiO2 heterojunction with lamellar structure to further enhance its photoefficiency, resulting in higher hydrogen generation. This work offers an easy way to design a lamellar mesostructured photocatalyst with the ability to extend the charge carrier lifetime, which opens a new avenue for photocatalytic and solar energy conversion.

Conflicts of interest

There are no conflicts to declare.


Cong Wang acknowledges the China Scholarship Council (CSC) for his fellow research position. The authors thank Marie-Claire Schanne-Klein for full access to the polarized optical microscope and François Brisset for collecting the SEM and TEM images. Jian Li acknowledges the public grant overseen by the French National Research Agency (ANR) as part of the “Investissements d'Avenir” program (Labex NanoSaclay, reference: ANR-10-LABX-0035) for his post-doc position.

Notes and references

  1. J. B. Sambur, T. Y. Chen, E. Choudhary, G. Chen, E. J. Nissen, E. M. Thomas, N. Zou and P. Chen, Nature, 2016, 530, 77–80 CrossRef CAS PubMed.
  2. S. Nishimura, N. Abrams, B. A. Lewis, L. I. Halaoui, T. E. Mallouk, K. D. Benkstein, J. Lagemaat and A. J. Frank, J. Am. Chem. Soc., 2003, 125, 6306–6310 CrossRef CAS PubMed.
  3. S. Guldin, S. Huttner, M. Kolle, M. E. Welland, P. M. Buschbaum, R. H. Friend, U. Steiner and N. Tetreault, Nano Lett., 2010, 10, 2303–2309 CrossRef CAS PubMed.
  4. J. T. Park, J. H. Prosser, S. H. Ahn, S. J. Kim, J. H. Kim and D. Lee, Adv. Funct. Mater., 2013, 23, 2193–2200 CrossRef CAS.
  5. C. T. Dinh, H. Yen, F. Kleitz and T. O. Do, Angew. Chem., Int. Ed. Engl., 2014, 53, 6618–6623 CrossRef CAS PubMed.
  6. H. Zhao, Z. Hu, J. Liu, Y. Li, M. Wu, G. Van Tendeloo and B.-L. Su, Nano Energy, 2018, 47, 266–274 CrossRef CAS.
  7. X. Li, T. Fan, H. Zhou, S.-K. Chow, W. Zhang, D. Zhang, Q. Guo and H. Ogawa, Adv. Funct. Mater., 2009, 19, 45–56 CrossRef CAS.
  8. M. Giese, L. K. Blusch, M. K. Khan and M. J. MacLachlan, Angew. Chem., Int. Ed., 2015, 54, 2888–2910 CrossRef CAS PubMed.
  9. H. Kargarzadeh, J. Huang, N. Lin, I. Ahmad, M. Mariano, A. Dufresne, S. Thomas and A. Gałęski, Prog. Polym. Sci., 2018, 87, 197–227 CrossRef CAS.
  10. K. E. Shopsowitz, H. Qi, W. Y. Hamad and M. J. Maclachlan, Nature, 2010, 468, 422–425 CrossRef CAS PubMed.
  11. K. E. Shopsowitz, W. Y. Hamad and M. J. MacLachlan, J. Am. Chem. Soc., 2012, 134, 867–870 CrossRef CAS PubMed.
  12. M. K. Khan, M. Giese, M. Yu, J. A. Kelly, W. Y. Hamad and M. J. MacLachlan, Angew. Chem., Int. Ed., 2013, 52, 8921–8924 CrossRef CAS PubMed.
  13. K. E. Shopsowitz, J. A. Kelly, W. Y. Hamad and M. J. MacLachlan, Adv. Funct. Mater., 2014, 24, 327–338 CrossRef CAS.
  14. A. Ivanova, D. Fattakhova-Rohlfing, B. E. Kayaalp, J. Rathousky and T. Bein, J. Am. Chem. Soc., 2014, 136, 5930–5937 CrossRef CAS PubMed.
  15. A. Ivanova, M. C. Fravventura, D. Fattakhova-Rohlfing, J. Rathouský, L. Movsesyan, P. Ganter, T. J. Savenije and T. Bein, Chem. Mater., 2015, 27, 6205–6212 CrossRef CAS.
  16. J. Xue, F. Song, X.-W. Yin, Z.-L. Zhang, Y. Liu, X.-L. Wang and Y.-Z. Wang, ACS Sustainable Chem. Eng., 2017, 5, 3721–3725 CrossRef CAS.
  17. J. A. Kelly, M. Giese, K. E. Shopsowitz, W. Y. Hamad and M. J. MacLachlan, Acc. Chem. Res., 2014, 47, 1088–1096 CrossRef CAS PubMed.
  18. G. D. Gesesse, C. Li, E. Paineau, Y. Habibi, H. Remita, C. Colbeau-Justin and M. N. Ghazzal, Chem. Mater., 2019, 31, 4851–4863 CrossRef CAS.
  19. T. D. Nguyen, J. Li, E. Lizundia, M. Niederberger, W. Y. Hamad and M. J. MacLachlan, Adv. Funct. Mater., 2019, 29, 1904639–1904648 CrossRef.
  20. K. E. Shopsowitz, A. Stahl, W. Y. Hamad and M. J. MacLachlan, Angew. Chem., Int. Ed., 2012, 51, 6886–6890 CrossRef CAS PubMed.
  21. T. D. Nguyen, E. Lizundia, M. Niederberger, W. Y. Hamad and M. J. MacLachlan, Chem. Mater., 2019, 31, 2174–2181 CrossRef CAS.
  22. M. N. Ghazzal, D. P. Debecker and E. M. Gaigneaux, Thin Solid Films, 2016, 611, 117–124 CrossRef CAS.
  23. M. N. Ghazzal, E. Aubry, N. Chaoui and D. Robert, Beilstein J. Nanotechnol., 2015, 6, 2039–2045 CrossRef CAS PubMed.
  24. L. Zhong, S. Fu, X. Peng, H. Zhan and R. Sun, Carbohydr. Polym., 2012, 90, 644–649 CrossRef CAS PubMed.
  25. C. Li, J. Evans, N. Wang, T. Guo and S. He, Sci. Rep., 2019, 9, 11290 CrossRef PubMed.
  26. P. X. Wang, W. Y. Hamad and M. J. MacLachlan, Nat. Commun., 2016, 7, 11515 CrossRef PubMed.
  27. H. D. Vries, Acta Crystallogr., 1951, 4, 219–226 CrossRef.
  28. S. Lisinski, J. Krause, D. Schaniel, L. Ratke and T. Woike, Scr. Mater., 2008, 58, 553–555 CrossRef CAS.
  29. J. D. Kittle, X. Du, F. Jiang, C. Qian, T. Heinze, M. Roman and A. R. Esker, Biomacromolecules, 2011, 12, 2881–2887 CrossRef CAS PubMed.
  30. J. Majoinen, E. Kontturi, O. Ikkala and D. G. Gray, Cellulose, 2012, 19, 1599–1605 CrossRef CAS.
  31. A. D. French, Cellulose, 2013, 21, 885–896 CrossRef.
  32. N. Abrams, S. Nishimura, B. A. Lewis, L. I. Halaoui, K. D. Benkstein, T. E. Mallouk, J. Lagemaat and A. J. Frank, J. Am. Chem. Soc., 2003, 125, 6306–6310 CrossRef PubMed.
  33. H. Zhou, X. Li, T. Fan, F. E. Osterloh, J. Ding, E. M. Sabio, D. Zhang and Q. Guo, Adv. Mater., 2010, 22, 951–956 CrossRef CAS PubMed.
  34. M. Kunst and K. M. Schindler, J. Phys. Chem., 1990, 94, 8222–8226 CrossRef.
  35. T. J. Savenije, J. E. Kroeze and J. M. Warman, J. Am. Chem. Soc., 2004, 126, 7608–7618 CrossRef PubMed.
  36. D. B. Liston, N. J. Cherepy, J. A. Lovejoy, H. Deng and J. Z. Zhang, Appl. Catal., B, 1998, 102, 770–776 Search PubMed.
  37. N. Delegan, R. Pandiyan, S. Johnston, A. Dirany, S. Komtchou, P. Drogui and M. A. El Khakani, J. Phys. Chem. C, 2018, 122, 5411–5419 CrossRef CAS.
  38. R. Katoh, A. Furube, K.-i. Yamanaka and T. Morikawa, J. Phys. Chem. Lett., 2010, 1, 3261–3265 CrossRef CAS.
  39. N. Serpone and A. Salinaro, Pure Appl. Chem., 1999, 71, 303–320 CAS.
  40. S. Nakajima and R. Katoh, J. Mater. Chem. A, 2015, 3, 15466–15472 RSC.
  41. T. J. Savenije, A. Huijser, M. J. W. Vermeulen and R. Katoh, Chem. Phys. Lett., 2008, 461, 93–96 CrossRef CAS.
  42. M. Kunst, F. Goubard, C. Colbeau-Justin and F. Wünsch, Mater. Sci. Eng., C, 2007, 27, 1061–1064 CrossRef CAS.
  43. Z. Hai, N. E. Kolli, D. B. Uribe, P. Beaunier, M. Jose-Yacaman, J. Vigneron, A. Etcheberry, S. Sorgues, C. Colbeau-Justin, J. Chen and H. Remita, J. Mater. Chem. A, 2013, 1, 10829–10835 RSC.
  44. G. D. Gesesse, T. Le Neel, Z. Cui, G. Bachelier, H. Remita, C. Colbeau-Justin and M. N. Ghazzal, Nanoscale, 2018, 10, 20140–20146 RSC.
  45. M. N. Ghazzal, O. Deparis, J. De Coninck and E. M. Gaigneaux, J. Mater. Chem. C, 2013, 1, 6202–6209 RSC.
  46. M. N. Ghazzal, H. Kebaili, M. Joseph, D. P. Debecker, P. Eloy, J. De Coninck and E. M. Gaigneaux, Appl. Catal., B, 2012, 115–116, 276–284 CrossRef CAS.
  47. W. Luo, F. Li, Q. Li, X. Wang, W. Yang, L. Zhou and L. Mai, ACS Appl. Mater. Interfaces, 2018, 10, 7201–7207 CrossRef CAS PubMed.
  48. Y. Liang, J. Manioudakis, J. R. Macairan, M. S. Askari, P. Forgione and R. Naccache, ACS Omega, 2019, 4, 14955–14961 CrossRef CAS PubMed.
  49. J. Fan, Y. H. Dai, Y. L. Li, N. F. Zheng, J. F. Guo, X. Q. Yan and G. D. Stucky, J. Am. Chem. Soc., 2009, 131, 15568–15569 CrossRef CAS PubMed.
  50. J. Li, L. Liu, Y. Yu, Y. Tang, H. Li and F. Du, Electrochem. Commun., 2004, 6, 940–943 CrossRef CAS.
  51. L. Chen, J. Zhang, X. Ren, R. Ge, W. Teng, X. Sun and X. Li, Nanoscale, 2017, 9, 16632–16637 RSC.
  52. G. Silversmit, D. Depla, H. Poelman, G. B. Marin and R. D. Gryse, Surf. Sci., 2006, 600, 3512–3517 CrossRef CAS.
  53. A. D. Paola, G. Marcì, L. Palmisano, M. Schiavello, K. Uosaki, S. Ikeda and B. Ohtani, J. Phys. Chem. B, 2002, 106, 637–645 CrossRef.
  54. S. Stoyanov, D. Mladenova and C. Dushkin, React. Kinet., Mech. Catal., 2006, 88, 277–283 CrossRef CAS.
  55. S. Chen and L. W. Wang, Chem. Mater., 2012, 24, 3659–3666 CrossRef CAS.
  56. M. E. Aguirre, R. Zhou, A. J. Eugene, M. I. Guzman and M. A. Grela, Appl. Catal., B, 2017, 217, 485–493 CrossRef CAS.
  57. A. Visible, R. B. Wang, A. Vertova, S. Rondin, A. Minguzzi, E. Ahlberg and M. Busch, Chem. Mater., 2019, 31, 4787–4792 CrossRef.
  58. S. Xu and D. D. Sun, Int. J. Hydrogen Energy, 2009, 34, 6096–6104 CrossRef CAS.
  59. N. Lakshmana Reddy, S. Emin, M. Valant and M. V. Shankar, Int. J. Hydrogen Energy, 2017, 42, 6627–6636 CrossRef CAS.
  60. S. Wang, L. Pan, J. J. Song, W. Mi, J. J. Zou, L. Wang and X. Zhang, J. Am. Chem. Soc., 2015, 137, 2975–2983 CrossRef CAS PubMed.
  61. R. Ren, Z. Wen, S. Cui, Y. Hou, X. Guo and J. Chen, Sci. Rep., 2015, 5, 10714 CrossRef CAS PubMed.
  62. W. Choi, A. Termin and M. R. Hoffmann, J. Phys. Chem., 1994, 98, 13669–13679 CrossRef.
  63. Y. Peng, M. Yan, Q.-G. Chen, C.-M. Fan, H.-Y. Zhou and A.-W. Xu, J. Mater. Chem. A, 2014, 2, 8517–8524 RSC.
  64. S. A. Rawool, M. R. Pai, A. M. Banerjee, A. Arya, R. S. Ningthoujam, R. Tewari, R. Rao, B. Chalke, P. Ayyub, A. K. Tripathi and S. R. Bharadwaj, Appl. Catal., B, 2018, 221, 443–458 CrossRef CAS.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ta12665a

This journal is © The Royal Society of Chemistry 2020