• No results found

Operating the stacked photoanode at the thickness of exciton diffusion length enhances the efficiency of photoelectrochemical water splitting

N/A
N/A
Protected

Academic year: 2022

Share "Operating the stacked photoanode at the thickness of exciton diffusion length enhances the efficiency of photoelectrochemical water splitting"

Copied!
10
0
0

Loading.... (view fulltext now)

Full text

(1)

REGULAR ARTICLE

Operating the stacked photoanode at the thickness of exciton diffusion length enhances the efficiency of photoelectrochemical water splitting

ASHOK KUMAR UMMIREDDIaand RAJ GANESH S PALAa,b,*

aDepartment of Chemical Engineering, Indian Institute of Technology Kanpur, Kanpur, Uttar Pradesh 208016, India

bMaterial Science Programme, Indian Institute of Technology Kanpur, Kanpur, Uttar Pradesh 208016, India E-mail: rpala@iitk.ac.in

MS received 25 September 2020; revised 5 January 2021; accepted 18 February 2021

Abstract. In photoelectrochemical water splitting, every impinging photon should not only be absorbed but also be utilized towards reaction. While increasing photoelectrode thickness in photoelectrochemical reactors facilitates photon absorption, it has a debilitating effect on efficiency if the thickness required for complete photon absorption is much more than the exciton diffusion length, which is a property determined by the material and its processing. To address this issue, we demonstrate a general experimental methodology with a stack of cadmium selenide photoanodes wherein the thickness of each photoelectrode is of the order of exciton diffusion length and which improves overall photocurrent by about 50%.

Keywords. Photoelectrochemical water splitting; Photoelectrode thickness; Cadmium selenide; Stack;

Exciton diffusion length; Optical absorption.

1. Introduction

Realizing sustainable generation of fuels is a critical endeavor, and immense effort has gone towards solar water splitting.1–15 While photocatalysis offers a simpler photoreactor design, photoelectrochemistry provides separation of combustible products and a systematic approach to augment photo-potential with external electric potential.16–25 There are three major reactor designs used to perform a photoelectrochemi- cal (PEC) water splitting reaction, namely (1) particle- based PEC reactor, (2) tubular reactor, and (3) flat plate reactor.26 One design variable for greater absorption of solar flux is to operate both the photo- catalytic bed reactor and particle-based PEC reactor at depths wherein there are no photocatalytic particles in the ‘dark zone.’27 A tubular PEC reactor uses a solar concentrating cover plate and heterojunction photo- electrode (coupling two or more semiconductors) to increase solar absorption. Further, photoelectrode with a single semiconductor material having a greater

thickness, photoelectrode is formed via coupling two or more semiconductor materials (heterojunction photoelectrode), and both anode and cathode are made up of semiconductor materials (tandem configuration) are commonly employed strategies to enhance solar absorption in a flat plate photoelectrochemical reac- tor.28 Later two designs enhance solar absorption by utilizing light from different portions of solar spectra (such as UV, Vis, and IR). Still, an increase of each semiconductor’s thickness is needed for materials with a low absorption coefficient to enhance the absorption of light from the same spectra. However, increasing photon absorption via increasing photoelectrode thickness often decreases the efficiency due to increased charge-carrier recombination.28–30

To address this issue, we demonstrate a strategy wherein a single photoelectrode is replaced by a stack of photoelectrodes with the thickness of each photo- electrode is of the order of exciton diffusion length.

This strategy, not restricted to a particular class of materials, enables the utilization of transmitted light

*For correspondence

Supplementary Information: The online version contains supplementary material available athttps://doi.org/10.1007/s12039-021- 01893-7.

https://doi.org/10.1007/s12039-021-01893-7Sadhana(0123456789().,-volV)FT3](0123456789().,-volV)

(2)

from the preceding photoelectrodes at succeeding photoelectrodes. Albeit the stacked photoelectrodes were used for dye-sensitized photocells31 and PEC water splitting reaction,32–36 none of these works operated the individual electrodes with thickness as same as exciton diffusion length. Itoh et al., first demonstrated effective utilization of photons for PEC water splitting reaction by stacking Fe2O3 photoan- odes. Photoanode of 60 nm thick spray-coated Fe2O3 on SnO2 deposited glass produced photocurrent den- sity of 0.3 mA/cm2at 0.6 V vs. SCE (pH = 13) and the stack of four electrodes of the same thickness pro- duced photocurrent density of 1.1 mA/cm2. However, a single electrode of 600 nm thick produced a pho- tocurrent density of 0.5 mA/cm2 only under similar conditions.34Kimet al.,used dual photoanodes of 300 nm thick BiVO4 and 450 nm thick a-Fe2O3 to use photons effectively and enhance the photocurrent density of the PEC water-splitting reaction. Under direct light illumination, BiVO4 and Fe2O3 photoan- odes produced photocurrent densities of 5.0 and 4.5 mA/cm2 at 1.23 VRHE, respectively, and stack con- figuration (Fe2O3 behind BiVO4) produced 7.0 mA/

cm2 under similar conditions.32 When Wang et al., used single 250 nm thick BiVO4coated with FeOOH/

NiOOH cocatalysts as photoanode for PEC water splitting reaction and produced photocurrent density 5.13 mA/cm2 at 1.23 V vs. RHE. However, the dual photoanode stack enhanced the photocurrent density to 5.87 mA/cm2.35 Niu et al., effectively recycled the photons by stacking two photoanodes made of the same semiconductor material (Cu2O) but synthesized them using the thermal oxidation and electrodeposi- tion methods. In this report, the researchers configured anodes such that electrodeposited Cu2O (ED-Cu2O) in the front and thermally oxidized Cu2O (TO-Cu2O) at the back. The front ED-Cu2O produced a photocurrent density of 4 mA/cm2, back TO-Cu2O produced 3 mA/

cm2, and dual Cu2O photoanode produced 7 mA/cm2 at 0 V vs. RHE.33 Ahmetet al., used the combination of dual photoanodes and photovoltaic cells in the stack to use photons effectively. Stack of cobalt phosphate- coated hydrogenate tungsten-doped BiVO4 (CoPi/ W:BiVO4) anodes in the front and silicon hetero- junction (SHJ) solar cell at the back produced 5.12 mA/cm2at 1.23 V vs. RHE.36

The thermodynamic potential of 1.23 V is needed for the photoelectrochemical water-splitting reaction.

However, when considering the losses associated with charge carrier transportation and overpotential required for surface reactions, materials with a mini- mum bandgap of 1.7 eV are needed. Since solar intensity in the solar spectrum is dropped rapidly

below 390 nm, materials with a bandgap greater than 3.2 eV deliver less solar to hydrogen efficiency. These arguments suggest that materials with a bandgap between 1.7 eV and 3.2 eV are suitable for PEC water- splitting reactions. Among all pristine photoanodes, TiO2, a-Fe2O3, BiVO4, CdS, and CdSe possess suit- able bandgap (1.7 eV-3.2 eV) for the water-splitting reaction.37–39 Though TiO2 is stable under both strongly acidic and alkaline conditions, a high bandgap of 3.2 eV only absorbs the UV light (i.e., 5% of the solar spectrum), limits its use.40 a-Fe2O3 is having a suitable bandgap (2.2 eV), but it has a short diffusion length (2-4 nm) and low absorption coefficient, which requires at least 400-500 nm thick film to absorb full solar intensity.41Around 100-200 electrodes should be used in the stack to achieve maximum current density.

BiOV4has a diffusion length of 10 nm and needs 100 nm thick film to absorb full solar intensity, which requires ten electrodes in the stack to achieve maxi- mum photocurrent density.42,43 Controlling the thick- ness of films in the nanometer range and difficulty in stacking tens of electrodes limit us to use botha-Fe2O3

and BiOV4 hotoanodes. Further, CdS having micrometer range diffusion length and reasonable absorption coefficient does not need stacking, and only one electrode is sufficient to absorb full light inten- sity.44Among all photoanodes discussed above, CdSe possesses a moderate band gap (1.74 eV), reasonable diffusion length (200-400 nm), and absorption coeffi- cient.44–46 A total of four electrodes are sufficient to absorb light intensity completely. Besides, CdSe can be deposited electrochemically,47control the thickness of 250 nm, and possess diffusion length less than total optical absorption length. Due to the properties men- tioned earlier, CdSe is chosen as a model system to demonstrate our idea instead of exploring new mate- rials to provide a photoreactor design criterion appli- cable to any material via the exciton diffusion length.

However, CdSe suffers from photo corrosion, and hole scavenger such as methanol is used to prevent photo corrosion. Other strategies like depositing protective layers such as titanium dioxide48 and aluminium- doped ZnO49of angstrom length scale on the electrode surface by atomic layer deposition can be used to prevent photo corrosion and further integrated into this approach.

In this study, the electrodeposited cadmium selenide (CdSe) on indium doped tin oxide coated glass (ITO) was used as an electrode in the stack. ITO glass was chosen because of its high transmittance and good electrical conductivity.50–53 A stack of four photoan- odes with CdSe film thickness of 250 nm on ITO substrates was used in 20% (v/v) methanol in an

(3)

aqueous 1M KCl solution.54Light intensity decreases with an increase in the overall optical length (which is the sum of the thickness of individual photoelectrodes that are part of the stack), and the photocurrent at different photoelectrodes scales with its insolation.

The photocurrent per unit solar flux of a single elec- trode is 0.72 mA/cm2and increases by about 50% (to 1.09 mA/cm2) after stacking four electrodes under the same solar insolation. Further, we have proposed a model to estimate the total current from the stack and the current from individual electrodes in the stack. The calculated current densities compare well experimen- tal current densities.

2. Experimental 2.1 Materials

CdCl2.H2O (99%, Lot # GM0055A1604) was pur- chased from Loba Chemie., India; selenium powder (-100 mesh, 99.5%, Lot # MKBB5843) and nitrilo- triacetic acid trisodium salt (NTA) (98%, Lot # BCBS0649) were purchased from Sigma-Aldrich, India; sodium sulfite anhydrous (98% Lot

#DE6D661216) and methanol (99%, Lot # SD3F630217) were purchased from Merck, India; KCl (99.5%, Lot # 2205010817) and H2SO4 (98%, Lot # 1031430516) were purchased from Fisher Scientific, India; ITO coated glass substrates (10 X/sq.) were purchased from Techinstro, India; Milli-Q water (18.2 MX) was obtained from Merck Millipore water purification system, electrolytes were prepared with Milli-Q water (18.2 MX-cm), and all the chemicals were used as received.

2.2 Electrodeposition of CdSe film on ITO substrate

Before electrodeposition of CdSe, the ITO coated glasses were sonicated in soap solution for 90 min and washed with tap water and Milli-Q water. CdSe was then electrodeposited onto the pre-cleaned ITO coated glass substrates in a 50 mL electrochemical cellviaan electrochemical workstation (Metrohm Autolab from the Netherlands, Multi Autolab/M101). 3.0 g of Na2- SO3was dissolved in Milli-Q water to prepare 30 mL of selenosulfite solution, and pH was adjusted to 9.0 by adding 1 M H2SO4 solution dropwise. Selenium powder (0.12 g) was added to the solution mentioned above and heated until the solution was clear (90 °C and 4 h) under reflux condition. 0.265 g of NTA was

dissolved in Mill-Q water to prepare 10 mL of Cd- nitrilotriacetic acid trisodium (NTA) complex solu- tion. pH was adjusted to 8.0 by adding 3 M H2SO4

solution dropwise. 0.177 g of CdCl2.H2O added to the solution. We observed that the reproducibility of results is achieved better with freshly prepared Cd- NTA solution. After cooling the selenosulfite solution to room temperature, Cd-NTA solution was added to it. The electrochemical deposition was carried out in the solution at -1.0 V vs. Ag/AgCl on scotch tape masked ITO glasses (5.5 mm and 12.7 mm diameter) for the required charge; the platinum mesh was used as the counter electrode.50–52Different thicknesses of the CdSe films were prepared by varying the deposition charge densities passed through the working electrode (ranging from -75 mC/cm2 to -176 mC/cm2). A schematic of the experimental procedure is shown in Scheme 1.

2.3 Material characterization

X-ray diffraction (XRD, PANalytical, Germany) was performed by scanning from 10°to 80° with the scan rate of 2°/min. Morphology of the CdSe film was determined by FESEM (Carl Zeiss, Supra 40 VP), and chemical composition analysis was performed by the energy dispersive X-ray spectrometer (Oxford, UK;

spatial resolution 0.5 lm radius), which was annexed to FESEM. The absorbance of CdSe film was mea- sured by UV-Vis-NIR spectrophotometer (Cary 5000, Agilent Technologies). Thickness measurements were performed by an optical profilometer (NanoMap-D, Aep Technology, USA). The elemental composition of the surface was analyzed by X-ray photoelectron spectroscopy (XPS) in addition to EDX. XPS mea- surements were done by using a PHI Versa Probe II Scanning XPS Microprobe and analysis was done by CasaXPS software (version 2.3.22PR1.0). The XPS spectra were calibrated by setting C-1s peak (used as a reference) at 284.6 eV.

2.4 Photoelectrochemical measurements

The photoelectrochemical characterization was undertaken in a three-electrode set-up consisting of Ag/AgCl (sat. KCl) reference electrode, working electrode, and counter-electrode (platinum mesh).

Such a three-electrode set-up is sufficient as an applied bias to a photon to current efficiency is not a metric of interest here.9 A stack of photoelectrodes was used as a working electrode (the stack of photoelectrodes

(4)

replacing the single working photoelectrode in the conventional set-up). The stack is comprised of four photoelectrodes with CdSe film thickness of 250 nm on ITO glass substrates was used as the working electrode-stack. An ohmic contact was established via stainless steel foil on ITO glass, and it was ensured that the charge neutrality between the electrode com- partments by making the holes in the spacer between the photoelectrodes (Figure S1, Supplementary Infor- mation). 20% methanol (v/v) in 1 M KCl (aq) solution (pH = 7) was used as the electrolyte, where methanol acts as a hole scavenger.54An Oriel Sol3A coupled to an AM 1.5G filter (Newport) was used as the light source. The illumination intensity was calibrated using an energy meter (Newport, Model 842 PE), and the intensity measured on the surface of the first electrode in the stack from the solar simulator was 100 mW/cm2. All the potentials specified at the working electrode are mentioned in the RHE scale.

3. Results and Discussion

CdSe films of different thicknesses on ITO coated glass substrate were fabricated by electrodeposition at -1.0 V vs. Ag/AgCl for different charge densities.

X-ray diffraction (XRD) pattern of CdSe film elec- trodeposited on ITO coated glass for 1 h in Fig- ure1(a) confirms the formation of cubic CdSe (JCPDS card no: 00-019-0191). However, the broadness of the

peaks suggests that the degree of crystallinity is less.

The electrodeposited CdSe films’ morphology was probed by scanning electron microscopy (SEM) (Fig- ure1(b-c)). The electron microscopic image of smooth ITO glass (Figure1(b)) and CdSe electrodeposited on smooth ITO glass (Figure 1(c)) is shown. Figure 1c shows that the CdSe film consists of spherical particles ranging from 75 nm to 150 nm. Further, chemical composition analysis (Figure 1(d)) was performed by the energy dispersive X-ray spectrometer (EDX) annexed to SEM reveals the presence of Cd and Se, and the ratio of Cd and Se on the surface as almost unity. The survey X-ray photoelectron spectroscopy (Figure 1e) further confirms that the presence of both Cd and Se elements and the elemental ratio of Cd and Se on the surface is almost unity. The core-level spectra of Cd 3d split into two peaks, i.e., Cd 3d5/2and Cd 3d3/2(Figure1f). The peak at 404.44 eV is ascribed to Cd 3d5/2, and 411.22 eV is ascribed to Cd 3d3/2, which is in agreement with previous reports.55 The energy difference between the two peaks is 6.78 eV, which suggests that the valence state of Cd in the CdSe is ?2.56 Peak of Se 3d core-level spectra (Figure 1g) located at 53.40 eV as a single convoluted peak. The peak is deconvoluted into two peaks Se 3d3/2 and Se 3d5/2. Both Se 3d3/2and Se 3d5/2 peaks are separated by an energy difference of 0.85 eV and positioned at 53.97 eV and 53.12 eV, respectively, suggest that the valence of Se in CdSe is -2.56 Further, the energy difference between Cd 3d and Se 3d peaks is Scheme 1. Schematic illustration of the experimental protocol of electrochemical deposition of CdSe film on ITO glass substrate.

(5)

351.04 eV which is similar to the value of bulk CdSe.55,57These results further confirm that CdSe film is synthesized.

Each film’s thickness formed after passing different charge densities is measured by using a profilometer, and the calibration curve is plotted (Figure S2, Sup- plementary Information). The thickness of the CdSe film is linearly increased with the charge passed through ITO glass. UV-Vis spectra of CdSe films of varying thickness shown in Figure S3(a) (Supplemen- tary Information) suggest that all films show a similar trend of absorption across different wavelengths.

When absorbance is measured at a particular

wavelength (550 nm), absorbance scales linearly with the thickness of the CdSe film (Figure S3(b), Supple- mentary Information). This wavelength is chosen because 550 nm is the average wavelength in the UV- Vis range (300-700 nm). Figure S4 (Supplementary Information) shows the transmittance spectra of 254 nm thick CdSe film on the ITO substrate. The area under the curve was also calculated in the UV-Vis range (300-700 nm) to estimate the average transmittance.

The photocurrent density-voltage curves of CdSe films of different thickness (140 nm, 254 nm, 375 nm, and 434 nm) were shown in Figure 2(a). The thinnest Figure 1. (a) XRD of ITO glass and CdSe film electrodeposited on ITO glass for 1 h (b) SEM of ITO substrate (c) SEM of CdSe electrodeposited over smooth ITO substrate at -1.0 V vs. Ag/AgCl and total charge passed was -126 mC/cm2and (d) EDX analysis of electrodeposited CdSe. XPS spectra of (e) survey scan and elemental analysis of 254 nm thick CdSe film electrodeposited on ITO glass (f) core level spectra of Cd 3d and (g) Se 3d. Filled red circles represent raw XPS data.

(6)

film (140 nm) among all films shows the least pho- toresponse, and the thickest film (434 nm) shows less photoresponse compared to other films (254 nm and 375 nm) but slightly more than 140 nm thick film over a broad range of potentials. It is also observed that the 375 nm thick film exhibits the highest current density among all thicknesses in the lower potential regime.

The current observed in Figure 2(a) is a combination of both Faradaic photocurrent and non-Faradaic capacitative current. The I vs. V shows distinctive behavior in two regions. There is a region beyond which there is a monotonous exponential rise in the photocurrent and an onset potential (which is a com- bination of photo-potential and applied potential), as indicated by a dotted vertical line in the above figure, can be associated with this region. The current is predominantly capacitative below the onset potential, although some part of the current can be due to the Faradaic photocurrent.

The capacitative currents critically depend on the roughness of the material and roughness/electro- chemical surface area is expected to increase mono- tonously with the film’s thickness. The Faradaic photocurrent in these potential regimes (\ onset potential) will have a non-monotonous curve peaking at around the exciton diffusion length (which is*250 nm for this material). We hypothesize that these two currents’ convolution gives rise to that 375 nm thick film exhibiting the highest current density below onset potential. To compare current densities where the contribution from photocurrent is more, we have plotted the photocurrent densities vs. thickness of CdSe film at high potential bias, i.e., -0.95 vs. RHE (Figure 2(b)).

The photocurrent observed is the trade-off between optical absorption and charge carrier separation. As optical absorption increases with the thickness of the film and more photons are used for charge carrier generation, photocurrent increases with the thickness.

Moreover, charge carrier separation is constant up to the thickness as same as charge carrier diffusion length. It decreases thereafter due to electrons com- bining with holes before reaching the current collector.

The photocurrent density as a function of CdSe films of different thickness shows a peak in current density an intermediate thickness (Figure 2(b)) as the pho- tocurrent is presumably limited by optical absorption at small film thickness and by high electron-hole recombination rates at large thickness.50 We have not measured the exciton diffusion length in the conven- tional method (ex-situ measurements like photolumi- nescence),58,59 but measured the optimal thickness under photoelectrochemical reactor conditions, which would reflect the exciton diffusion length of the pro- cessed material under in-situ condition. The CdSe film that shows maximum photocurrent has a thickness

*250 nm and this thickness is in the range of charge carrier diffusion length of CdSe.44 250 nm thick film has been prepared by the passage of-126 mC/cm2of charge during electrodeposition, and CdSe films of this thickness were utilized in each of the photoelectrodes that formed a unit of the photoelectrode-stack.

The photocurrent density-voltage curves of indi- vidual photoanodes in the stack and stack of four photoelectrodes were obtained in the dark and under light illumination (Figure 3(a)), and under chopped light illumination (on/off cycles: 5 s) (Figure 3(b)). A sharp increase (decrease) in current is observed when Figure 2. (a) LSV of CdSe films with varying thickness under 20 mW/cm2light illumination; scan rate 20 mV/s (b) Peak in photocurrent at 0.95 V vs. RHE as a function of CdSe film thickness.

(7)

light is switched on (off) confirms the increase in current is solely from light (Figure 3(b)). In each measurement, linear sweep voltammetry (LSV) was done at a scan rate of 20 mV/s from 0 V to 1.0 V (vs.

RHE). Methanol was used as a hole scavenger, and reactions associated with it are mentioned in the literature.60,61

We have calculated the % transmittance of stacks containing two, three, and four electrodes as we know the transmittance of one electrode. The transmittance of light from a stack of n electrodes is

%T ¼100ð Þfs n ð1Þ Here, n is the number of electrodes in the stack.

f and s are the transmittances of film and substrate, respectively.

The corresponding photocurrent density of the stack of n electrodes was calculated from the photocurrent data in Figure3(a) by adding photocurrent densities up to the nth electrode. The photocurrent density as a function of %T from a stack of n electrodes is plotted in Figure 4. Photocurrent density linearly decreases with the increase in % transmittance from the stack consists of n electrodes. Less % transmittance from the stack indicates more photons are absorbed and utilized for water splitting reaction and vice versa.

The intensity of light transmitted to individual electrodes in the stack follows a geometric sequence.

Since current is directly proportional to the intensity of light under similar conditions, current from individual electrodes in the stack also follows a geometric

sequence. Current from the nthelectrode in the stackin

is

in¼i fsð Þn1 ð2Þ

and total current from the stack having n electrodes iTis

iT ¼ið1ð Þfs nÞ 1fs

ð Þ ð3Þ

Here, n is the number of electrodes in the stack.

f and s are the transmittances of film and substrate, respectively. This model is a slight variant of the previous work.34

Figure 3. (a) Photocurrent-voltage curves of electrodes at different positions in the stack and overall stack photocurrent across four electrodes (b) chopped light photocurrent-voltage characteristics of electrodes at different positions in the stack and stack of four electrodes. The thickness of CdSe film on each electrode used in the stack is *250 nm. All the measurements were done under AM 1.5 G light illumination (100 mW/cm2). At scan rate of 20 mV/s.

0 5 10 15 20 25 30 35 40

700 800 900 1000 1100 1200

1 2

3

Photo-CurrentDensity(μ$/cm2 )

Transmittance (%) 4

Figure 4. Photocurrent density as a function of %T from a stack of n electrodes. The numbers indicated on the plot suggest the number of photoanodes in the stack.

(8)

In the dark, all the electrodes in the stack show the same current-voltage response. At 1.0 V (vs. RHE) under simulated solar illumination, the first electrode from the solar simulator generates a photocurrent density of 0.72 mA/cm2. The average transmittance of CdSe film on the ITO substrate is calculated from Figure S4 (Supplementary Information) as f*s = 0.35 (i.e., 35% of incident photons are wasted as trans- mittance from each electrode). The photocurrent den- sity-voltage curve of the second electrode clearly showed photoresponse and generates a photocurrent density of 0.26 mA/cm2at 1.0 V (vs. RHE). It utilizes a fraction of light transmitted from the first electrode.

The third electrode shows little photoresponse com- pared to the second electrode and generates a pho- tocurrent density of 0.11 mA/cm2at 1.0 V (vs. RHE).

Since the intensity of light transmitted from the third electrode is significantly less, the photoresponse of the fourth electrode is not significant. Figure 5shows the correlation between experimental and calculated pho- tocurrent densities from equation 2. The total pho- tocurrent density from the stack at 1.0 V (vs. RHE) is increased by 47% to 1.06 mA/cm2 when compared to the photocurrent obtained from the single photoelec- trode (which is *0.72 mA/cm2) and 97% of its the- oretical maxima (when n??, iT= 1.09 mA/cm2) has been achieved by using four electrodes in the stack.

4. Conclusions

Typically, the thickness required for solar absorption is larger than distances that bulk-generated electron- hole pair can travel to reach the surface, wherein they

can be utilized for electrochemical reactions. Due to this reason, electrons and holes are annihilated, thereby causing a drop-in efficiency despite good optical absorption. To overcome this issue, we show that instead of photoreactor with a single photoelec- trode, if the photoreactor is formed via a stack of photoelectrodes such that the number of photoelec- trodes in a stack (n) is equal to the film thickness required for complete solar absorption divided by exciton diffusion length, an optimal design can be achieved. We have demonstrated this strategy by forming the stack with four 250 nm thick CdSe pho- toelectrodes. This design showed that the total pho- tocurrent generated can be increased by 50%, and the unutilized photons can be decreased from 35% to 3%.

As this strategy is not constrained by any particular class of materials, it can be adopted not only for promising materials presently available but also towards materials discovered in the future.

Supplementary Information (SI)

Schematic of the photoelectrochemical cell with the stack as working electrode, plots of calibration curve between charge passed and thickness of the film, UV-Vis for deter- mination of optical absorption/transmittance, and transmit- tance curve of 254 nm thick CdSe film on ITO substrate (Figures S1–S4) are mentioned in Supplementary Infor- mation. Supplementary Information is available atwww.ias.

ac.in/chemsci.

Acknowledgement

We acknowledge the support from the Science and Engi- neering Research Board, Department of Science and Technology, Government of India via project SERB/F/

11147/2017-2018.

References

1. Nocera D G 2012 The artificial leafAcc. Chem. Res.45 767

2. Shaner M R, Atwater H A, Lewis N S and McFarland E W 2016 A comparative technoeconomic analysis of renewable hydrogen production using solar energy Energy Environ. Sci. 92354

3. Pinaud B A, Benck J D, Seitz L C, Forman A J, Chen Z, Deutsch T G, et al. 2013 Technical and economic feasibility of centralized facilities for solar hydrogen production via photocatalysis and photoelectrochem- istryEnergy Environ. Sci.61983

4. Memming R 2015 Semiconductor Electrochemistry (New York: John Wiley & Sons) p.55

5. Behara D K, Sharma G P, Upadhyay A P, Gyanprakash M, Pala R G S and Sivakumar S 2016 Synchronization

1 2 3 4

0 100 200 300 400 500 600 700 800

Photo-currentdensity(μ$/cm2 )

Position of the electrode in stack Experimental Simulated

Figure 5. Correlation between experimental photocurrent density at 1.0 V vs. RHE and calculated photocurrent density from equation 2.

(9)

of charge carrier separation by tailoring the interface of Si–Au–TiO2 heterostructures via click chemistry for PEC water splittingChem. Eng. Sci.154 150

6. Grimes C, Varghese O and Ranjan S 2008Light, Water, Hydrogen: The Solar Generation of Hydrogen by Water Photoelectrolysis(US: Springer) p.135

7. Miller E L 2010On Solar Hydrogen and Nanotechnol- ogy(New York: John Wiley & Sons) p.31

8. Upadhyay A P, Behara D K, Sharma G P, Bajpai A, Sharac N, Ragan R, et al. 2013 Generic process for highly stable metallic nanoparticle-semiconductor heterostructures via click chemistry for electro/photo- catalytic applications ACS Appl. Mater. Interfaces 5 9554

9. Pala R G S 2017 Solar Hydrogen as a ‘‘Renewable Reductant’’: Points and Counterpoints R P Chhabra (Ed.) (Boca Raton, FL: CRC Press) p.1369

10. Rani S and Rajalakshmi N 2015 Effect of nanotube diameter on photo-electro-chemical properties of carbon quantum dot functionalized TiO2 nanotubes J. Clean Energ. Technol.3 367

11. Rawool S A, Pai M R, Banerjee A M, Arya A, Ningthoujam R S, Tewari R, et al. 2018 pn Hetero- junctions in NiO:TiO2 composites with type-II band alignment assisting sunlight driven photocatalytic H2 generationAppl. Catal. B221 443

12. Rai S, Ikram A, Sahai S, Dass S, Shrivastav R and Satsangi V R 2017 CNT based photoelectrodes for PEC generation of hydrogen: a review Int. J. Hydrogen Energy423994

13. Rajaambal S, Sivaranjani K and Gopinath C S 2015 Recent developments in solar H2generation from water splittingJ. Chem. Sci.127 33

14. Van de Krol R 2012 Photoelectrochemical Hydrogen Production(US: Springer) p.285

15. Alagarasi A, Rajalakshmi P U, Shanthi K and Selvam P 2019 Solar-light driven photocatalytic activity of mesoporous nanocrystalline TiO2, SnO2, and TiO2- SnO2compositesMater. Today Sustain.5100016 16. Tseng C-J and Tseng C-L 2011 The reactor design for

photoelectrochemical hydrogen production Int. J. Hy- drogen Energy366510

17. Castro S, Albo J and Irabien A 2018 Photoelectrochem- ical reactors for CO2 utilization ACS Sustain. Chem.

Eng.615877

18. Ong C K, Dennison S, Hellgardt K and Kelsall G 2011 Evaluation and modeling of a photoelectrochemical reactor for hydrogen production operating under high photon fluxECS Trans. 3511

19. Wang Q, Zhu H and Li B 2019 Synergy of Ti-O-based heterojunction and hierarchical 1D nanobelt/3D micro- flower heteroarchitectures for enhanced photocatalytic tetracycline degradation and photoelectrochemical water splittingChem. Eng. J.378 122072

20. Hou J, Cheng H, Takeda O and Zhu H 2015 Unique 3D heterojunction photoanode design to harness charge transfer for efficient and stable photoelectrochemical water splittingEnergy Environ. Sci.81348

21. Wang Q, Zhang B, Lu X, Zhang X, Zhu H and Li B 2018 Multifunctional 3D K2Ti6O13nanobelt-built archi- tectures towards wastewater remediation: selective adsorption, photodegradation, mechanism insight and

photoelectrochemical investigationCatal. Sci. Technol.

8 6180

22. Singh A P, Arora P, Basu S and Mehta B R 2016 Graphitic carbon nitride based hydrogen treated disor- dered titanium dioxide core-shell nanocatalyst for enhanced photocatalytic and photoelectrochemical per- formanceInt. J. Hydrogen Energy415617

23. Bhandary N, Singh A P, Kumar S, Ingole P P, Thakur G S, Ganguli A K and Basu S 2016 In situ solid-state synthesis of a AgNi/g-C3N4 nanocomposite for enhanced photoelectrochemical and photocatalytic activityChemSusChem 9 2816

24. Bhandary N, Singh A P, Ingole P P and Basu S 2017 Enhancing the photoelectrochemical performance of a hematite dendrite/graphitic carbon nitride nanocompos- ite through surface modification with CoFeOx ChemPhotoChem170

25. Seetharaman S, Balaji R, Ramya K, Dhathathreyan K S and Velan M 2014 Electrochemical behaviour of nickel- based electrodes for oxygen evolution reaction in alkaline water electrolysis Ionics20713

26. Hydrogen Production: Photoelectrochemical Water Splitting. www.energy.gov/eere/fuelcells/hydrogen-pro duction-photoelectrochemical-water-splitting. (Accessed 13th March).

27. Dijkstra M F J, Buwalda H, de Jong A W F, Michorius A, Winkelman J G M and Beenackers A A C M 2001 Experimental comparison of three reactor designs for photocatalytic water purificationChem. Eng. Sci.56547 28. Jiang C, Moniz S J A, Wang A, Zhang T and Tang J 2017 Photoelectrochemical devices for solar water splitting: materials and challenges Chem. Soc. Rev. 46 4645

29. Behara D K, Ummireddi A K, Aragonda V, Gupta P K, Pala R G S and Sivakumar S 2016 Coupled optical absorption, charge carrier separation, and surface elec- trochemistry in surface disordered/hydrogenated TiO2 for enhanced PEC water splitting reactionPhys. Chem.

Chem. Phys. 188364

30. Viswanathan B and Scibioh M A 2014 Photoelectro- chemistry: Principles and Practices (London, United Kingdom: Alpha Science) p. 123

31. Armstrong N R and Shepard V R 1982 Photoelectrol- ysis using linear arrays of chemically modified semi- transparent electrodes J. Electroanal. Chem. Interf.

Electrochem.131113

32. Kim J H, Jang J-W, Jo Y H, Abdi F F, Lee Y H, van de Krol R and Lee J S 2016 Hetero-type dual photoanodes for unbiased solar water splitting with extended light harvestingNat. Commun.713380

33. Niu W, Moehl T, Cui W, Wick-Joliat R, Zhu L and Tilley S D 2018 Extended light harvesting with dual Cu2O-based photocathodes for high efficiency water splitting Adv. Energy Mater.8 1702323

34. Itoh K and Bockris J O M 1984 Stacked thin-film photoelectrode using iron oxide J. Appl. Phys.56874 35. Wang S, Chen P, Bai Y, Yun J-H, Liu G and Wang L

2018 New BiVO4 dual photoanodes with enriched oxygen vacancies for efficient solar-driven water split- ting Adv. Mater.301800486

36. Ahmet I Y, Ma Y, Jang J-W, Henschel T, Stannowski B, Lopes T, et al. 2019 Demonstration of a 50 cm2

(10)

BiVO4 tandem photoelectrochemical-photovoltaic water splitting deviceSustain. Energy Fuels3 2366 37. Janu V C, Bahuguna G, Laishram D, Shejale K P,

Kumar N, Sharma R K and Gupta R 2018 Surface fluorination ofa-Fe2O3using select fluor for enhance- ment in photoelectrochemical properties Sol. Energy Mater. Sol. Cells174240

38. Kumar M P, Jagannathan R and Ravichandran S 2020 Photoelectrochemical system for unassisted high-effi- ciency water-splitting reactions using N-doped TiO2 nanotubesEnergy Fuels349030

39. Venkatkarthick R, Davidson D J, Ravichandran S, Vasudevan S and Sozhan G 2017 a-Fe2O3/TiO2

heterostructured photoanode on titanium substrate for photoelectrochemical water electrolysis Mater. Chem.

Phys.199 249

40. Xu H, Ouyang S, Liu L, Reunchan P, Umezawa N and Ye J 2014 Recent advances in TiO2-based photocatal- ysisJ. Mater. Chem. A2 12642

41. Sivula K, Le Formal F and Gra¨tzel M 2011 Solar water splitting: progress using hematite (a-Fe2O3) photoelec- trodesChemSusChem4432

42. Hilliard S, Friedrich D, Kressman S, Strub H, Artero V and Laberty-Robert C 2017 Solar-water-splitting BiVO4 thin-film photoanodes prepared by using a sol-gel dip- coating techniqueChemPhotoChem1 273

43. Seabold J A, Zhu K and Neale N R 2014 Efficient solar photoelectrolysis by nanoporous Mo:BiVO4 through controlled electron transport Phys. Chem. Chem. Phys.

161121

44. Novikov B V, Ilinskii A V, Lieder K F and Sokolov N S 1971 Determination of exciton diffusion length from photoconductivity low-temperature spectraPhys. Status Solid. B48473

45. Amelia M, Lincheneau C, Silvi S and Credi A 2012 Electrochemical properties of CdSe and CdTe quantum dotsChem. Soc. Rev.415728

46. Biswal N and Parida K M 2013 Enhanced hydrogen production over CdSe QD/ZTP composite under visible light irradiation without using cocatalystInt. J. Hydro- gen Energy381267

47. Shaikh A V, Mane R S, Joo O-S, Han S H and Pathan H M 2017 Electrochemical deposition of cadmium selenide films and their properties: a review J. Solid State Electrochem.212517

48. Sivula K 2014 Defects give new life to an old material:

electronically leaky titania as a photoanode protection layerChemCatChem6 2796

49. Paracchino A, Laporte V, Sivula K, Gra¨tzel M and Thimsen E 2011 Highly active oxide photocathode for photoelectro- chemical water reductionNature Mater.10456

50. Miao J, Yang H B, Khoo S Y and Liu B 2013 Electrochemical fabrication of ZnO–CdSe core–shell nanorod arrays for efficient photoelectrochemical water splitting Nanoscale511118

51. Kutzmutz S, La´ng G and Heusler K E 2001 The electrodeposition of CdSe from alkaline electrolytes Electrochim. Acta47955

52. Cerdeira F, Torriani I, Motisuke P, Lemos V and Decker F 1988 Optical and structural properties of polycrystalline CdSe deposited on titanium substrates Appl. Phys. A 46107

53. Ikram A, Sahai S, Rai S, Dass S, Shrivastav R and Satsangi V R 2014 Synergistic effect of CdSe quantum dots on photoelectrochemical response of electrode- posited a-Fe2O3filmsJ. Power Sources267664 54. Frame F A and Osterloh F E 2010 CdSe-MoS2: A

Quantum Size-Confined Photocatalyst for Hydrogen Evolution from Water under Visible Light J. Phys.

Chem. C11410628

55. Katari J E B, Colvin V L and Alivisatos A P 1994 X-ray photoelectron spectroscopy of CdSe nanocrystals with applications to studies of the nanocrystal surface J.

Phys. Chem.984109

56. Kim J W, Shim H-S, Ko S, Jeong U, Lee C-L and Kim W B 2012 Thorny CdSe nanotubes via an aqueous anion exchange reaction process and their photoelectrochem- ical applications J. Mater. Chem.2220889

57. Raut V S, Lokhande C D and Killedar V V 2017 Synthesis and studies on effect of indium doping on physical properties of electrodeposited CdSe thin films J. Mater. Sci. Mater. Electron.283140

58. Xing G, Mathews N, Sun S, Lim S S, Lam Y M, Gra¨tzel M, et al. 2013 Long-range balanced electron- and hole- transport lengths in organic-inorganic CH3NH3PbI3 Science342344

59. Lee E M Y and Tisdale W A 2015 Determination of exciton diffusion length by transient photoluminescence quenching and its application to quantum dot films J.

Phys. Chem. C1199005

60. Shen M and Henderson M A 2011 Identification of the active species in photochemical hole scavenging reac- tions of methanol on TiO2J. Phys. Chem. Lett.2 2707 61. Guzman F, Chuang S S C and Yang C 2013 Role of methanol sacrificing reagent in the photocatalytic evo- lution of hydrogenInd. Eng. Chem. Res.5261

References

Related documents

SaLt MaRSheS The latest data indicates salt marshes may be unable to keep pace with sea-level rise and drown, transforming the coastal landscape and depriv- ing us of a

Although a refined source apportionment study is needed to quantify the contribution of each source to the pollution level, road transport stands out as a key source of PM 2.5

These gains in crop production are unprecedented which is why 5 million small farmers in India in 2008 elected to plant 7.6 million hectares of Bt cotton which

INDEPENDENT MONITORING BOARD | RECOMMENDED ACTION.. Rationale: Repeatedly, in field surveys, from front-line polio workers, and in meeting after meeting, it has become clear that

3 Collective bargaining is defined in the ILO’s Collective Bargaining Convention, 1981 (No. 154), as “all negotiations which take place between an employer, a group of employers

Harmonization of requirements of national legislation on international road transport, including requirements for vehicles and road infrastructure ..... Promoting the implementation

Angola Benin Burkina Faso Burundi Central African Republic Chad Comoros Democratic Republic of the Congo Djibouti Eritrea Ethiopia Gambia Guinea Guinea-Bissau Haiti Lesotho

The PC at various light intensities at photon energy of 1.88 eV together with dark background current is shown in figure 9.. Both the structure of the spectrum and the PC peak