Journal of Catalysis 310 (2014) 51–56

Contents lists available at SciVerse ScienceDirect

Journal of Catalysis

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

Photocatalytic hydrogen production in a noble-metal-free systemcatalyzed by in situ grown molybdenum sulfide catalyst

0021-9517/$ - see front matter � 2013 Published by Elsevier Inc.http://dx.doi.org/10.1016/j.jcat.2013.04.006

⇑ Corresponding author. Fax: +61 7 33654199.E-mail address: l.wang@uq.edu.au (L. Wang).

Xu Zong, Zheng Xing, Hua Yu, Yang Bai, Gao Qing (Max) Lu, Lianzhou Wang ⇑ARC Centre of Excellence for Functional Nanomaterials, School of Chemical Engineering and AIBN, The University of Queensland, Qld 4072, Australia

a r t i c l e i n f o

Article history:Received 25 February 2013Revised 2 April 2013Accepted 5 April 2013Available online 19 June 2013

Keywords:PhotocatalysisHydrogen productionMolybdenum sulfideWater splitting

a b s t r a c t

The establishment of cost-effective photocatalytic system based on earth-abundant materials is crucialfor the practical utilizations of solar energy. To achieve this goal, it is highly desirable to employ photo-sensitizers and catalysts that are both derived from earth-abundant materials. Herein, we report an effi-cient noble-metal-free system by integrating the above two components. We found that functional MoSx

nanoparticles can be obtained with an in situ photoreduction manner during photocatalytic reactions inthe presence of inexpensive organic sensitizers. The thus-obtained MoSx catalysts demonstrated quitehigh efficiency for catalyzing H2 evolution under visible light. The factors influencing the performanceof the photocatalytic system was investigated and a two-step reaction mechanism was proposed. Theconcept of in situ formation of hydrogen evolution catalyst paves the way for investigating biomimeticmolybdenum sulfide catalysts for photocatalytic H2 production in systems without the presence of noblemetals.

� 2013 Published by Elsevier Inc.

1. Introduction

Conversion of solar energy to hydrogen via photocatalytic watersplitting offers an attractive solution toward energy and environ-mental problems [1–5]. To facilitate the liberation of hydrogenfrom water, the employment of catalysts that can offer low activa-tion energy for H2 evolution is crucial. In this regard, platinumgroup metals have been traditionally used and showed unrivalledactivity for catalyzing hydrogen evolution reaction (HER) [6]. How-ever, large-scale hydrogen production with platinum group metalswill be severely baffled by some inevitable challenges such as scar-city and high cost, which has inspired tremendous endeavor offinding alternative non-precious metal catalysts [7,8].

In biological systems, enzymes such as hydrogenases and nitro-genases catalyze hydrogen evolution and the responsible activesites are composed of earth-abundant elements such as Fe, Ni,Mo, and sulfur atoms [9,10]. Inspired by nature, much effort hasbeen devoted to the development of functional materials thatcan mimic the active sites of these enzymes for H2 production[7,11]. In this regard, the investigation of biomimetic molybdenumsulfide (MoSx), an inorganic analogue of nitrogenase, as a heteroge-neous electrocatalyst for catalyzing HER has attracted renewedinterest [6,11–21]. The work on electrochemical hydrogen evolu-tion on bulk MoS2 was pioneered by Tributsch and co-workers in

the 1970s [22]. In recent years, different MoSx materials such asMoS2, MoS3, and Mo3S4 particles or nanosheets in the amorphousor crystalline form were reported to show high efficiency for cata-lyzing HER in electrochemical and photocatalytic reactions, andMoSx-based materials have been supposed to be a potential alter-native for noble metals in catalyzing H2 production [20]. Despitethe aforementioned progresses, most of the efforts have been de-voted to the electrochemical hydrogen production with MoSx elec-trodes or photocatalytic hydrogen production in a heterogeneoussystem where semiconductors are used as the light absorber[12,13,20]. While much less attention has been paid to hydrogenproduction in a system sensitized with organic dyes that typicallyresembles the artificial photosynthesis process. Moreover, eventhough numerous systems have shown activity for H2 production,most of them are based upon photosensitizers consisting of noblemetals such as Ru, Ir, Pt, and Rh [23–25]. From a practical stand-point, the development of a noble-metal-free system for hydrogenproduction is still a great challenge to be met and thus is highlysignificant and important research topic [26–29].

Herein, we report that functional molybdenum sulfide nanopar-ticles can be obtained with an in situ approach during photocata-lytic reactions. The thus-obtained molybdenum sulfide catalystsdemonstrated quite high efficiency for catalyzing H2 evolution ina noble-metal-free system under visible light. To our knowledge,this is the first successful example of employing molybdenum sul-fide catalyst for catalyzing H2 production in a noble-metal-free sys-tem sensitized by organic dye, which could shed light on thedesign of ‘‘green’’ and economical chemistry cycles.

Fig. 1. TEM image of molybdenum sulfide nanoparticles.

52 X. Zong et al. / Journal of Catalysis 310 (2014) 51–56

2. Materials and methods

2.1. Catalysts preparation and chemicals

Triethanolamine (TEOA, Sigma–Aldrich), Erythrosin B (EB, Sig-ma–Aldrich), (NH4)6Mo7O24 (Sigma–Aldrich), and (NH4)2S solution(21%, Fluka) were of reagent quality and used as received. (NH4)2-

MoS4 was synthesized from (NH4)6Mo7O24 and (NH4)2S solutionaccording to a reported procedure [30]. In a typical synthesis,1.5 g of (NH4)6Mo7O24 was added to 20 ml of (NH4)2S solution atambient temperature under stirring conditions. The mixture wasthen heated at 353 K to obtain a deep red solution. The solutionwas stirred for 2 h and then transferred to a refrigerator with dura-tion of 12 h. The precipitated red crystals were thoroughly washedwith ethanol, dried, and stored under nitrogen. All solutions in theexperiments were prepared with deionized ultrapure water(18 M cm) that was obtained by an Elga water purification system.

2.2. Characterizations

The UV–visible absorption spectra of the samples were obtainedon an UV visible spectrophotometer (UV-2200, Shimadzu). X-rayphotoelectron spectroscopy (XPS) was recorded with an X-ray pho-toelectron spectrometer (Thermo Escalab 250, a monochromatic AlKR X-ray source). The morphologies of MoSx catalysts were charac-terized with transmission electron microscopy (TEM, Philips Tec-nei F20). X-ray diffraction (XRD) patterns of MoSx catalysts werecollected using an X-ray diffractometer (Rigaku Miniflex). Fluores-cence spectra were measured using a Jobin–Yvon Horiba Fluorologin steady-state mode using a xenon lamp as the excitation source.

2.3. Photocatalytic reactions

Photocatalytic hydrogen evolution reaction was performed in aTEOA/photosensitizer/MoSx three-component system. Typically,Erythrosine B (2 mM) and (NH4)2MoS4 (1 mM) were added to aquartz reaction cell containing 100 mL of 10 vol% TEOA aqueoussolution under vigorous stirring. The reaction cell was then con-nected to a closed gas circulation and evacuation system. The solu-tion was thoroughly degassed and irradiated by a 300 W Xe lamp(Beijing Trusttech Co. Ltd., PLS-SXE-300UV). The Xe lamp wasequipped with an optical cut-off filter (k > 420 nm) to eliminateultraviolet light, and the reactor was equipped with a water filterto remove infrared light. The amount of H2 produced was analyzedusing an on-line gas chromatography (GC-2014, Shimazu) with athermal conductivity detector (TCD). The temperature of the reac-tion solution was controlled by a flow of cooling water with ther-mo-controller. The pH value of the TEOA solution was adjusted todifferent values using concentrated HNO3 prior reaction.

Fig. 2. High-resolution XPS spectra of Mo 3d and S2p measured on molybdenumsulfide nanoparticles.

3. Results and discussion

Photocatalytic hydrogen evolution reaction was performed in athree-component system containing Erythrosin B (EB) as the pho-tosensitizer (PS), triethanolamine (10 vol%) as the sacrificial elec-tron donor, and (NH4)2MoS4 as the precursor for molybdenumsulfide catalyst. Upon visible light irradiation (k > 420 nm), vigor-ous gas bubbles were observed, and the gas was confirmed to beH2 using on-line gas chromatography. Control experimentsshowed that the absence of any of the three components led tonegligible H2 production.

The original reaction solution is bright red color and remainsunchanged when kept in dark condition. After irradiating the solu-tion with visible light (k > 420 nm), the color of the solution chan-ged from bright red to dark brown within 10 min. UV–Vis analysis

indicated the absence of (NH4)2MoS4 after the reaction (DetailedUV–Vis results will be discussed later). However, when EB wasadded to the reaction solution after the cessation of the reaction,efficient hydrogen production was still observed (Fig. S1). This isquite different from common three-component homogenous sys-tem, where the addition of both PS and catalyst are necessary to re-start hydrogen evolution after their decomposition [26,27].Therefore, it is reasonable to suppose that some active species

Fig. 3. Time course of photocatalytic hydrogen evolution from EB–TEOA solutionusing (NH4)2MoS4 precursor, MoSx powder or remaining reaction solution recov-ered after the cessation of the reaction. (light source: 300 W Xe lamp, k > 420 nm,100 mL solution, EB = 2 mM, temperature = 293 K).

Fig. 4. Time course of photocatalytic hydrogen evolution from EB–MoSx–TEOAsystem as a function of pH value. (light source: 300 W Xe lamp, k > 420 nm, 100 mLsolution, EB = 2 mM, MoSx = 1 mM, temperature = 283 K).

Fig. 5. Time course of photocatalytic hydrogen evolution from EB–MoSx–TEOAsystem at different temperature. (light source: 300 W Xe lamp, k > 420 nm, 100 mLsolution, EB = 2 mM, MoSx = 1 mM).

Fig. 6. Time course of photocatalytic hydrogen evolution from EB–MoSx–TEOAsystem as a function of (a) MoSx and (b) EB concentrations (light source: 300 W Xelamp, k > 420 nm, 100 mL solution, temperature = 293 K). The concentration of EBin (a) is fixed at 2 mM and that of MoSx in (b) is fixed at 1 mM.

X. Zong et al. / Journal of Catalysis 310 (2014) 51–56 53

derived from (NH4)2MoS4 are present in the solution and act ashydrogen evolution catalyst.

We then centrifuged the reaction solution after the cessation ofthe reaction to obtain a highly flocculated dark powder andremaining reaction solution. A careful analysis of the dark brownpowder showed the prevailing presence of agglomerated nanopar-ticles with average diameter of �50 nm (Fig. 1). The nanoparticleswere identified to be amorphous with XRD analysis. X-ray photo-electron spectroscopy (XPS) indicated the presence of Mo and Sin the surface region of the nanoparticles (Fig. 2). The Mo 3d5/2(Mo 3d3/2) peak is located at 229.3 eV (232.5 eV) along with theappearance of a S 2s peak at 226.4 eV, indicating the existence ofa Mo with an oxidation state of +4 [12]. This chemical state ofMo was also observed in the amorphous MoS3 as well as theMoS2 reported by Merki et al. [16,17]. The S 2p spectrum consistsof two doublets with S 2p2/3 energies of 162.1 and 163.6 eV. Thespectra are quite similar to that of amorphous MoS3, while muchdifferent from that of commercial MoS2 particles, indicating theexistence of both S2� and S2�

2 ligands [16,17]. Therefore, the activenanoparticles are supposed to be amorphous reduced MoSx (x > 2)species that consists of S2�

2 ligands [20]. It is reported that (NH4)2-

MoS4 can be reduced to MoSx by photogenerated electrons onsemiconductors [31]. In our reaction system, it is reasonable topropose that (NH4)2MoS4 is reduced to MoSx nanoparticles which

could act as efficient catalyst for catalyzing the subsequent hydro-gen evolution in the presence of organic dyes.

This hypothesis was then confirmed with the following experi-ment. The recovered MoSx powder was redispersed in a reactionsolution containing fresh Erythrosin B and triethanolamine. Afterirradiating this fresh solution containing recovered MoSx powder,vigorous H2 production was observed (Fig. 3), and the activity ob-tained with recovered MoSx powder accounted for more than 80%

54 X. Zong et al. / Journal of Catalysis 310 (2014) 51–56

of that starting from (NH4)2MoS4 precursor, therefore verifyingthat the in situ generated amorphous MoSx nanoparticles act asan efficient and relatively stable HER catalysts during the photocat-alytic reactions. However, there is still one possibility that theremaining reaction solution after recovering the MoSx powdermay contain active species derived from the decomposition of(NH4)2MoS4 which can also efficiently catalyze the HER reaction.Therefore, we also tested the photocatalytic activity of this solutionby adding fresh EB. It was found that only a small amount of H2

was produced. Therefore, we can conclude that in situ-generatedMoSx powder plays the main role in catalyzing the H2 evolution.

Various experimental parameters such as pH value, solutiontemperature, and the concentrations of EB and catalyst were thenstudied to investigate their influence on the photocatalytic activity.Fig. 4 shows the effect of pH value of the solution on the rate ofhydrogen evolution. It was found that the rate of hydrogen evolu-tion increased linearly with the decrease in pH value. The activityachieved at pH value of 8.1 is about 3 times of that achieved at pH10.6. The increase in activity at lower pH values is likely due to thehigher proton concentration in solution and the fact that hydrogengeneration becomes more thermodynamically favorable withdecreasing pH values. The temperature of the reaction solutionwas also found to drastically affect the rate of hydrogen evolution(Fig. 5). When the reaction solution temperature was increasedfrom 283 to 293 K, the activity of hydrogen evolution was drasti-cally increased by about 1.5 times, demonstrating the great poten-tial of utilizing the infrared range of the solar spectrum in

Fig. 7. UV–Vis absorption spectra of (NH4)2MoS4 (2 � 10�5 M, left) and EB(1 � 10�5 M, right) in TEOA (10 vol%) aqueous solution after irradiation(k > 420 nm).

combination with the photocatalytic processes for solar hydrogenproduction. At present, we are still unclear about the reason forthis unusual drastic influence of the reaction temperature on thephotocatalytic performance of the reaction system. We tentativelysuppose that the significantly enhanced activity is likely due to theimproved thermodynamic and/or kinetic conditions for hydrogenevolution with increased reaction temperature.

The influence of the catalyst and photosensitizer concentrationwas then investigated. Fig. 6a shows the effect of catalyst concen-tration on the rate of hydrogen evolution when the sensitizer con-centration is fixed at 2 mM. As the (NH4)2MoS4 precursor isreduced to MoSx catalyst during the reaction, the MoSx catalystconcentration is calculated based upon the (NH4)2MoS4 added tothe system. It is found that increasing the catalyst concentrationincreases the rate of hydrogen evolution and the total amount ofhydrogen linearly at low catalyst concentrations, indicating afirst-order dependence on the catalyst concentration (Fig. S2).When the concentration of catalyst is higher than 1 mM, the rateof hydrogen evolution is more than 1.5 mmol h�1 in the first hour,which is among one of the highest reported value for the molecularhydrogen systems in the visible light range [8]. However, at highercatalyst concentrations (>1 mM), while more hydrogen is evolved,the rate of H2 production does not scale-up linearly with catalystconcentration. The total amount of H2 produced from this systemis about 4.95 mmol in 5 h, which is much higher than the amountsof EB (0.2 mmol) and catalyst (0.4 mmol), indicating that the reac-tions proceed photocatalytically. When the catalyst concentration

Fig. 8. Fluorescence quenching of deoxygenated EB aqueous solutions (EB = 10 uM,pH = 10.0) by TEOA (a) and Mo salt (b), respectively. TEOA 10 eq and Mo 10 eq meanthat the amounts of TEOA and Mo salt are 10 times of that of EB in themeasurement. Likewise for the other legends.

Scheme 1. Proposed two-step reaction mechanism for photocatalyic H2 production in the Dye–MoSx–TEOA system.

X. Zong et al. / Journal of Catalysis 310 (2014) 51–56 55

is fixed at 1 mM and the sensitizer concentration is varied, similartrend is observed (Fig. 6b and S3). Moreover, the system is ob-served to have a longer lifetime at higher EB or catalystconcentrations.

The above results showed that the initial efficient hydrogenevolution gave way to cessation of H2 production with the reactiontime. To better understand the reason for the decreased activity,the reaction solution was monitored using UV–Vis spectroscopy.The UV–Vis spectra of (NH4)2MoS4 or EB solution in the presenceof TEOA remain unchanged in a wide pH range (Fig. S4). Irradiationof pure (NH4)2MoS4 solution does not lead to noticeable change inits spectra. However, under the same conditions, the absorptionintensity of EB decreased drastically together with blue shifts ofthe characteristic absorption peaks that are associated with thepartial loss of iodide of EB (Fig. 7). Therefore, (NH4)2MoS4 aloneis stable while EB itself is unstable in TEOA aqueous solution uponlight irradiation. The UV–Vis spectra of the reaction solution con-taining EB, (NH4)2MoS4, and TEOA were then investigated. It wasfound that the absorption spectra of the reactions solution weresimply the sum of the individual components before irradiation(Fig. S5, left). Upon light irradiation, the absorption for the decom-posed EB was present while the characteristic (NH4)2MoS4 absorp-tion disappeared, suggesting the decomposition of the (NH4)2MoS4

to form MoSx. Considering that color of the reaction solutionturned to black brown within 10 min during the reaction, it is rea-sonable to believe that MoSx can be efficiently produced from(NH4)2MoS4 during the reaction. Moreover, the lifetime of EB wasfound to be prolonged in the presence of MoSx as an electronacceptor (Fig. S5, right). It should be noted that after the cessationof the reaction, further addition of (NH4)2MoS4 in the reaction solu-tion did not lead to appreciable hydrogen production even thoughMoSx was present in the system (Fig. S1). Therefore, the degrada-tion of EB sensitizer is supposed to be the main reason for the dete-rioration of the activity. This point was further supported by thephotocatalytic reaction using fluorescein – a more stable dye asthe PS (Fig. S6). It was found that hydrogen can be produced at astable rate in 9 h under visible light using fluorescein, while hydro-gen evolution almost stopped in about 5 h even though much high-er initial activity was achieved using halogenated EB. Although wefind the degradation of Erythrosin B (EB) is a problem for the pres-ent simple photocatalytic system, the important role of the in situformed MoSx for H2 production is clearly demonstrated. The devel-opment a robust dye coupled with the present MoSx catalyst isanticipated to lead to more sustainable H2 production.

Previous studies on homogeneous artificial photosynthesis sys-tems have shown that hydrogen formation proceeds through theelectron transfer quenching of the 3PS excited state with either oxi-dative or reductive pathway, leading respectively to PS+ with elec-tron transfer from the dye to the catalyst or PS� with electrontransfer to the dye from the sacrificial electron donor TEOA[26,27,32]. In the present three-component system, reactions are

initiated by the excitation of EB2� to its 3�EB2� excited state uponlight irradiation. We found that the fluorescence of EB was difficultto be quenched by TEOA while can be efficiently quenched by theMo salt (Fig. 8 and Fig. S10). An oxidative quenching mechanism issubsequently proposed for the present system. As shown inScheme 1, following the excitation of EB2�, the excited electronin 3�EB2� is transferred to (NH4)2MoS4 to reduce (NH4)2MoS4 toMoSx in the first step. In the second step, as hydrogen evolutioncatalyst MoSx instead of (NH4)2MoS4 is present in the system, theelectron transfer from 3�EB2� to the in situ formed MoSx leads tothe formation of H2 and the regeneration of EB2� by TEOA as sac-rificial electron donor.

4. Conclusions

In conclusion, molybdenum sulfide catalysts produced throughin situ photoreduction demonstrated high efficiency for catalyzingH2 evolution in a noble-metal-free organic dye-sensitized system.The concept of the in situ growth of hydrogen evolution catalyst-MoSx suggests a new pathway for the design of low lost catalystfor photocatalytic H2 production in homogeneous systems. Ongo-ing efforts are focused on the modification of molybdenum sulfidecatalyst with Co and Ni elements toward enhanced catalytic activ-ity and the long-term stability of the sensitizers.

Acknowledgments

This project was supported by Australian Research Council(through its DP programs) and Queensland State GovernmentSmart State program (NIRAP).

References

[1] N.S. Lewis, D.G. Nocera, Proc. Natl. Acad. Sci. USA 103 (2006) 15729.[2] J.A. Turner, Science 305 (2004) 972.[3] A.J. Bard, M.A. Fox, Acc. Chem. Res. 28 (1995) 141.[4] M. Gratzel, Acc. Chem. Res. 14 (1981) 376.[5] A.J. Esswein, D.G. Nocera, Chem. Rev. 107 (2007) 4022.[6] T.F. Jaramillo, K.P. Jorgensen, J. Bonde, J.H. Nielsen, S. Horch, I. Chorkendorff,

Science 317 (2007) 100.[7] P. Du, R. Eisenberg, Energy Environ. Sci. 5 (2012) 6012.[8] W. Zhang, J. Hong, J. Zheng, Z. Huang, J. Zhou, R. Xu, J. Am. Chem. Soc. 133

(2011) 20680.[9] D.C. Rees, J.B. Howard, Science 300 (2003) 929.

[10] S.C. Lee, R.H. Holm, Proc. Natl. Acad. Sci. USA 100 (2003) 3595.[11] A.B. Laursen, S. Kegnaes, S. Dahl, I. Chorkendorff, Energy Environ. Sci. 5 (2012)

5577.[12] Z. Chen, D. Cummins, B.N. Reinecke, E. Clark, M.K. Sunkara, T.F. Jaramillo, Nano

Lett. 11 (2011) 4168.[13] F.A. Frame, F.E. Osterloh, J. Phys. Chem. C 114 (2010) 10628.[14] I. Hatay, P.Y. Ge, H. Vrubel, X. Hu, H.H. Girault, Energy Environ. Sci. 4 (2011)

4246.[15] J. Kristensen, J. Zhang, I. Chorkendorff, J. Ulstrup, B.L. Ooi, Dalton Trans. (2006)

3985.[16] D. Merki, S. Fierro, H. Vrubel, X. Hu, Chem. Sci. 2 (2011) 1262.[17] H. Vrubel, D. Merki, X. Hu, Energy Environ. Sci. 5 (2012) 6136.

56 X. Zong et al. / Journal of Catalysis 310 (2014) 51–56

[18] T.F. Jaramillo, J. Bonde, J. Zhang, B.-L. Ooi, K. Andersson, J. Ulstrup, I.Chorkendorff, J. Phys. Chem. C 112 (2008) 17492.

[19] B. Hinnemann, P.G. Moses, J. Bonde, K.P. Jorgensen, J.H. Nielsen, S. Horch, I.Chorkendorff, J.K. Norskov, J. Am. Chem. Soc. 127 (2005) 5308.

[20] X. Zong, H.J. Yan, G.P. Wu, G.J. Ma, F.Y. Wen, L. Wang, C. Li, J. Am. Chem. Soc.130 (2008) 7176.

[21] X. Zong, Y. Na, F.Y. Wen, G.J. Ma, J.H. Yang, D.G. Wang, Y. Ma, M. Wang, L. Sun,C. Li, Chem. Commun. (2009) 4536.

[22] H. Tributsch, J.C. Bennett, J. Electroanal. Chem. 81 (1977) 97.[23] J.I. Goldsmith, W.R. Hudson, M.S. Lowry, T.H. Anderson, S. Bernhard, J. Am.

Chem. Soc. 127 (2005) 7502.[24] H. Ozawa, M. Haga, K. Sakai, J. Am. Chem. Soc. 128 (2006) 4926.[25] P. Du, J. Schneider, P. Jarosz, R. Eisenberg, J. Am. Chem. Soc. 128 (2006) 7726.

[26] Z. Han, W.R. McNamara, M.-S. Eum, P.L. Holland, R. Eisenberg, Angew. Chem.Int. Ed. 51 (2012) 1667.

[27] T.M. McCormick, B.D. Calitree, A. Orchard, N.D. Kraut, F.V. Bright, M.R. Detty, R.Eisenberg, J. Am. Chem. Soc. 132 (2010) 15480.

[28] P. Zhang, M. Wang, J. Dong, X. Li, F. Wang, L. Wu, L. Sun, J. Am. Chem. Soc. 114(2010) 15868.

[29] T. Lazarides, T. McCormick, P. Du, G. Luo, B. Lindley, R. Eisenberg, J. Am. Chem.Soc. 131 (2009) 9192.

[30] D. Genuit, P. Afanasiev, M. Vrinat, J. Catal. 235 (2005) 302.[31] S. Kanda, T. Akita, M. Fujishima, H. Tada, J. Colloid Interface Sci. 354 (2011)

607.[32] W.R. McNamara, Z. Han, P.J. Alperin, W.W. Brennessel, P.L. Holland, R.

Eisenberg, J. Am. Chem. Soc. 133 (2011) 15368.