Journal ofMaterials Chemistry A

PAPER

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article OnlineView Journal | View Issue

Kerf loss silicon a

Department of Chemical Engineering, Nation

Taiwan. E-mail: hytuan@che.nthu.edu.tw

† Electronic supplementary informa10.1039/c6ta03657k

‡ These authors contributed equally to th

Cite this: J. Mater. Chem. A, 2016, 4,12921

Received 2nd May 2016Accepted 18th July 2016

DOI: 10.1039/c6ta03657k

www.rsc.org/MaterialsA

This journal is © The Royal Society of C

s a cost-effective, high-efficiency,and convenient energy carrier: additive-mediatedrapid hydrogen production and integrated systemsfor electricity generation and hydrogen storage†

Tzu-Lun Kao,‡ Wei-Hsiang Huang‡ and Hsing-Yu Tuan*

Base-catalyzed chemical etching of silicon in water can produce hydrogen and dissociated orthosilicic acid

(SiO2(OH)22�), suggesting that silicon can be regarded as an energy carrier. However, this process needs

a large amount of low-priced silicon as the essential reactive material for cost saving and faster and

high-yield hydrogen production agreeable for industrialization. In this study, high-performance hydrogen

production through wet chemical etching of micrometer-sized kerf loss silicon recovered from the

sawing process of solar-grade wafers is reported. Additives, including sodium metasilicate (Na2SiO3) and

metasilicic acid (H2SiO3), were employed to accelerate the water splitting reaction, resulting in an

optimized hydrogen production rate of 4.72 � 10�3 g(H2) per s per g(Si) and a yield of 92% that ranks as

the best performance in the reported literature on a micrometer-sized silicon basis. In addition,

a proof-of-concept example showing that kerf loss silicon is a convenient energy carrier was conducted

using a kerf loss silicon-based hydrogen production reactor in coordination with either a fuel cell, which

converted the supplied hydrogen to electricity, or a high-pressure tank for hydrogen storage.

Introduction

Increased greenhouse gas emission and limited fossil fuel reserveshave become an urgent world-wide environmental problem.1–7 Thequest to develop alternative energy sources is important. Hydrogenis an exceptional energy source, notable for its high heat value of141.79 MJ kg�1 and benign emissions aer combustion.8–16

Therefore, many well-known companies in automotive industry,such as Toyota Motor Corp., have invested in the development ofelectric vehicles powered by stacks of fuel cells, increasing thedemand for hydrogen. Hydrogen can be generated througha water-splitting reaction by various catalytic means that includeelectrochemical,17–19 photoelectrochemical,20 photocatalytic,21–24

and base-catalyzed chemical etching methods.25–30 However, theamount of hydrogen production per unit time is restricted by theelectrode surface area in the rst two catalyticmethods. For clarity,the hydrogen production rates were calculated by Faraday's law ofelectrolysis from two studies regarding electrochemical andphotoelectrochemical hydrogen generation.31,32 The rates are 9.85� 10�9 g(H2) per s and 9.92� 10�10 g(H2) per s, respectively. As forthe photocatalytic approach, the absorbing of light to generate an

al Tsing Hua University, Hsinchu 30013,

tion (ESI) available. See DOI:

is work.

hemistry 2016

electron–hole pair is essential for hydrogen production, sug-gesting the performance depends strongly on the crystallinityand the bandgap energy of the catalyst.33 A hydrogen productionrate of 1.89 � 10�7 g(H2) per s per g(Si) was obtained froma previous photocatalytic hydrogen generation study.34 Addi-tionally, hydrogen can be generated via ethanol steamreforming35–37 or bio-processes.38,39 Unfortunately, the prod-ucts of these methods are usually not pure hydrogen buta mixture of different gases, which is not easily separated, andthe latter method usually involves microbes generatinghydrogen very slowly, indicating a large reactor is needed.

The chemical etching reactions between specic metals andwater generate hydrogen associated with benign byproductssuch as metal-oxides and metal hydroxide.40–43 Earth abundantmaterials, including silicon,43 aluminum,44–46 and magnesium,have been studied as cost-effective and sustainable energycarriers.47–49 Silicon theoretically has a high gravimetrichydrogen yield of 0.144 g(H2) per g(Si) for hydrogen genera-tion. There is a signicant amount of high-purity siliconconsumed annually. In 2012, 157 Mtons of silicon with a purityhigher than 99.99% was consumed solely in the United States,which indicated that 11.3 Mtons of hydrogen might be theo-retically produced per year with that amount of silicon throughthis base-catalyzed chemical etching route. A general mecha-nism of silicon corrosion in potassium hydroxide has beenpreviously studied, and the gross reaction is summarized bythe following equation:50,51

J. Mater. Chem. A, 2016, 4, 12921–12928 | 12921

Journal of Materials Chemistry A Paper

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article Online

Si + 2OH� + 2H2O / SiO2(OH)2� + 2H2

To make the hydroxide ions diffuse into the oxide layer moreeasily, either the particle sizes must be reduced or porousstructures must be established to increase the specic surfacearea of silicon materials.

Traditionally, stain etching can be used as an alternative toelectrochemical etching to produce porous silicon on a largescale.28,52 However, the use of poisonous hydrouoric acid andthe high mass loss of silicon make this top-down methodineffective and costly. On the other hand, silicon nanoparticleswith an average diameter of 10 nm revealed that a maximumhydrogen releasing rate value as high as 3.0 � 10�3 g(H2) per sper g(Si) and a molar ratio of hydrogen released to siliconloaded as high as 2.58 can be obtained from the occurrence ofpseudo-isotropic etching and the existence of silicon hydridebonding (Si–H).53 However, using carbon dioxide to providea rapid heating source and employing ammable silane asa precursor may not be environmentally friendly. Dai et al.proposed a bottom-up synthesis of mesoporous silicon witha specic surface area of 580 m2 g�1, which is the highest valueever reported.34 Hydrogen was released at a ultrafast rate of 5.0� 10�2 g(H2) per s per g(Si) and around 1.6 to 1.8 moles ofhydrogen were released per mole of silicon. However, thismethod requires the use of an expensive alloy, NaK, and toxiccompound, SiCl4.

In this work, kerf loss silicon collected from wafer sawingwaste was used to rapidly generate hydrogen. Fig. 1 shows theschematics of the general concept of this hydrogen productionprocess, and the simplicity of this experimental apparatus can

Fig. 1 Schematic illustration of using kerf loss silicon as a hydrogen genestorage.

12922 | J. Mater. Chem. A, 2016, 4, 12921–12928

be recognized. The kerf loss silicon is retrieved from the massloss of solar-grade silicon wafers (up to 50 wt% of the siliconwafer) caused by the sawing process. A solar cell fabricationyields approximately thousands of kilograms of kerf loss siliconannually. We found that the additives, including sodium met-asilicate (Na2SiO3) and silicic acid (H2SiO3), can signicantlyenhance the chemical rates of the base-catalyzed silicon–waterreaction. Unlike previous studies, the whole additive-mediatedwet chemical etching process did not involve toxic anddangerous chemicals, and the resulting product was composedof only silicon oxide and metal–silicate. An optimized resultshowed a hydrogen production rate of 4.72 � 10�3 g(H2) per sper g(Si) and a yield of 92% for the process, which was ranked asthe best in the reported literature on a micrometer-sized siliconbasis. Finally, a silicon-based hydrogen generation batchreactor designed and linked to a fuel cell allowed the generatedhydrogen to either be converted to electricity or released intoa gas tank for hydrogen storage. These two cases demonstratethat kerf loss silicon is an efficient and convenient energycarrier.

ExperimentalMaterials

All reagents in this work are analytical grade and commerciallyavailable. Sodium metasilicate (Na2SiO3, SiO2: 44–47%), silicicacid (H2SiO3, 99.9%) and ethanol (ACS reagent grade; >99.5%)were purchased from Sigma-Aldrich corporation. Potassiumhydroxide (KOH, 45.5%) was purchased from Sigma-Aldrich.Kerf loss was purchased from AUO Crystal Corp.

ration reactant and integration for electricity production and hydrogen

This journal is © The Royal Society of Chemistry 2016

Paper Journal of Materials Chemistry A

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article Online

Characterization

Powder X-ray diffraction pattern (XRD) analysis was performedon a Rigaku Ultima IV X-ray diffractometer using Cu Ka radia-tion source (l ¼ 1.54 A). Scanning electron microscopy (SEM)images were obtained by Hitachi SU8010 to observe themorphology, and energy-dispersive X-ray spectroscopy was alsodone for composition analysis. The high resolution transmissionelectron microscopy (HRTEM) and selected area electrondiffraction (SAED) images were obtained by a JEOL ARM200F.Inductively coupled plasma-mass spectrometer (ICP-MS) anal-ysis was performed by an Agilent 7500ce. X-ray photoelectronspectra were obtained by using a ULVAC-PHI Quantera SXMhighresolution X-ray photoelectron spectrometer.

Hydrogen generation process using kerf loss silicon asa reactant

Centrifugation treatment was done at 2000 rpm using 100 mg ofkerf loss silicon obtained from the AUO Crystal Corp. with 30 mlof ethanol, and the supernatant was then centrifuged again at10000 rpm to obtain the actual reactant used for hydrogenproduction. The hydrogen generation process, hydrogenproduction rate and yield were measured with the instrumentshown in Fig. S1.† A 25 ml ask containing 10 mg of kerf losssilicon was immersed in a water bath controlled at a constanttemperature and was continuously stirred. The etching solutionwas prepared by dissolving the solutes in 10 ml of deionizedwater. A cold trap was additionally placed between the ask andthe custom-made cylinder-like glassware to reduce the volu-metric differences caused by the evaporation of water in theprocess of hydrogen production. Finally, volumetric differencesresulting from hydrogen generation were monitored by a real-time high resolution digital video camera once the etchingsolution was injected into the reaction ask.

Large-scale hydrogen generation process

In a large scale hydrogen generation process, 200 mg of siliconand 15 ml of etching solution were mixed and stirred continu-ously in a stainless steel reactor (100 ml) (Fig. S2†), which wasnecessary to withstand the high pressure produced by largeamounts of silicon. The dynamics of system pressure andtemperature were recorded by a digital video camera for the useof theoretical calculation for the quantitative evolution ofhydrogen.

Hydrogen generation system integration as an electricitygenerator

Silicon (200 mg) and 15 ml of etching solution were mixed andstirred continuously in a stainless steel reactor (100 ml). Aerthe reaction completed, the output of the hydrogen generationreactor was linked to a fuel cell (two-stacked polymer electrolytemembrane cell), a hydrogen energy monitor (Horizon FCJJ-24),and an electric model vehicle in series with the nal outputcurrent and voltage recorded by a digital video camera.

This journal is © The Royal Society of Chemistry 2016

Hydrogen generation system integration for hydrogen storage

Silicon (500 mg) and 30 ml of etching solution were mixed andstirred continuously in the reactor, which was connected toa gas tank (300 ml). Aer the reaction completed, the valve wasopened manually allowing the hydrogen to ow freely into thegas tank.

Results and discussion

Fig. 2a shows the scanning electron microscopy (SEM) image ofkerf loss silicon used for hydrogen production. Microake-likemorphology of the kerf loss silicon was observed with randomshapes, a size distribution around 0.5 to 1 mm in the lateraldimension and 100 nm thickness. The crystalline species withinwere identied bymeans of XRD. In Fig. 2b, the spectrumwas inclose agreement with face-centered cubic silicon (JCPDS no.89-5012), suggesting no other crystalline impurities existed.Furthermore, most kerf loss silicon was polycrystalline asconrmed by the HRTEM image (Fig. 2e), which showed cleargrain boundaries. The SAED pattern (Fig. 2f) shows diffractionrings that are identical to silicon. There were some larger piecesof silicon observed with single crystallinity, as shown inFig. S3.† The energy dispersive spectrum (EDS) showed that thekerf loss silicon was composed majorly of silicon with a minoroxygen signal (Fig. 2c). ICP-MS conrmed that no other metalcontaminants were detectable (Table S1†).

The etching mechanism of a silicon–water reaction can beillustrated as a series of reactions, as shown in Fig. S4.† Theetching mechanism of a silicon–water reaction involves twohydroxide ions reacting with one silicon atom on the crystalsurface, which causes the back bonds between the silicon atomand two inner silicon atoms to be weakened and further broken.A positively charged silicon–hydroxide complex is formed and isthen detached from the crystal surface, allowing the freehydroxide ions to react with the exposed inner silicon atoms.Since the oxide layer prohibits hydroxide ions from diffusinginto the surface of the silicon atoms, the breaking of the backbonds of hydroxide-bonded silicon is considered the ratelimiting step for the whole reaction process.50,51 To track thesilicon–water reaction process, a series of ex situ experimentswere carried out by mixing 10 ml 0.57 M KOH aqueous solutionwith 40 mg kerf loss silicon. As the reaction time went by, thecolor of the etching solution became gradually lighter (Fig. 3a–e).Moreover, X-ray photoelectron spectroscopy (XPS) spectra of thesolid residues collected at different reaction times showed thatthe oxide peak at 103.3 eV became relatively strong as the reac-tion time proceeded (Fig. 3f), indicating the concentration of thesilicate increased.

A hydrogen production testing systemwas constructed for themeasurement of hydrogen production yield and rate as shown inFig. 4a, where the yield was dened as the amount of hydrogenproduced divided by the maximum amount of hydrogen theo-retically generated, and the rate was dened as the mass ofhydrogen generated per unit time per unit mass of siliconloaded. In a hydrogen generation experiment, increased pres-sure pushed the water level downward (Fig. 4b and Video S1†).

J. Mater. Chem. A, 2016, 4, 12921–12928 | 12923

Fig. 2 (a) SEM image, (b) XRD pattern, (c) EDS spectrum and composition analysis results, (d) TEM image, (e) HRTEM image, and (f) SAED patternof kerf loss silicon.

Fig. 3 (a–e) Color evolution throughout the kerf loss silicon/waterreaction (f) XPS spectra of the solid residues recovered at differentreaction times.

Fig. 4 (a) Photograph of the hydrogen generation testing system.Inset: a higher resolution image focused on the water level. (b) Illus-tration of the system depicting the water level variation duringhydrogen generation.

Journal of Materials Chemistry A Paper

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article Online

However, the pressure difference is insignicant compared tothe ambient pressure, giving 1 atm equal to 1033.6 cm H2O.Hence, we simplied the calculation by the assumption that allgas behaved like an ideal gas and the process was isobaric (seemore detail and equations in the ESI†). Based on theseassumptions, the number of hydrogen molecules in molesgenerated equals the ambient pressure multiplied by the volu-metric change divided by the gas constant and temperature inspecic units with the volumetric change, which was calculated

12924 | J. Mater. Chem. A, 2016, 4, 12921–12928

from themultiplication product of the cross-sectional area of theinner tube and the water level difference with the volume ofetching solution excluded.

Fig. 5 shows hydrogen generation results from the based-catalyzed etching reactions. In a typical hydrogen productionexperiment, 10 mg kerf loss silicon was corroded in 10 ml of0.57 M KOH aqueous solution at 303 K, which provided a yieldof 73% and a reaction rate of 6.0 � 10�5 g(H2) per s per g(Si).Various amounts of Na2SiO3 and H2SiO3 were employed toaccelerate the reaction rate and improve the hydrogen genera-tion yield. First, the effects of individual additives were tested.When 5 g dm�3 Na2SiO3 was added, a yield of 71% anda hydrogen generation rate of 6.8 � 10�5 g(H2) per s per g(Si)were obtained, suggesting that Na2SiO3 had a positive effect onboth hydrogen yield and generation rate. On the other hand,when 0.5 g dm�3 H2SiO3 was added, a lower hydrogen yield(64%) with a higher generation rate (6.3 � 10�5 g(H2) per s perg(Si)) was obtained, meaning that H2SiO3 may only have

This journal is © The Royal Society of Chemistry 2016

Fig. 5 Hydrogen generation results of the based-catalyzed etching ofkerf loss silicon with and without Na2SiO3 and H2SiO3 at 303 K. KOH,Na2SiO3, and H2SiO3 are abbreviated as K, M and S, respectively.

Paper Journal of Materials Chemistry A

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article Online

a positive effect on the hydrolysis rate. When both additiveswere added, a higher hydrogen yield of 83% with a hydrogengeneration rate of 7.4 � 10�5 g(H2) per s per g(Si) was obtained.Additionally, simply mixing those chemicals together in waterconrmed the additives did not react with water to generatehydrogen. All these results are summarized in Table S2.† Basedon the etching mechanism, the increased hydrogen generationrate and the yield are likely a result of Na2SiO3 and H2SiO3

acting as silicate group nuclei, transferring the newly formedsilicon hydroxide from the silicon surface to the existing nuclei,which helps in breaking the back bonds between the hydroxide-bonded silicon atoms and the inner silicon atoms.54 Moreover,as the reaction temperature was raised from 303 K to 343 K, thehydrogen generation rate increased to 2.56 � 10�4 g(H2) per sper g(Si), around 2.3 times that seen at 303 K (Fig. 6).

Fig. 6 Hydrogen generation results of reacting kerf loss silicon withetching solution (0.57 M KOH + 50 g dm�3 Na2SiO3 + 0.5 g dm�3

H2SiO3) at 303 K and 343 K.

This journal is © The Royal Society of Chemistry 2016

Currently, we have unveiled a specic composition of kerfloss silicon along with two additives at 343 K that enhance thehydrogen generation yield and rate. In addition, it is reasonablethat higher concentrations of KOH aqueous solution can lead toan accelerated reaction rate as hydroxide ions are one of theoverall reaction species. Nevertheless, the etching reaction rateis proportional to the concentration of hydroxide ions to thepower of one quarter within a certain concentration range, butit is negatively correlated when the concentration is higher than20 wt% according to a previously reported study.50 This impliesthat changing the concentration of KOH in the aqueous solu-tion is not an effective way to increase the reaction rate.

Eight experiments were conducted to obtain optimizedperformance. Three chemicals were considered as factors byadding more (denoted as +) or less (denoted as �) of eachchemical to understand the interactions among these two addi-tives and hydroxide ions. For convenient data reading, KOH,Na2SiO3, and H2SiO3 are denoted as K, M, and S, respectively (seeTable S3† for the detailed recipes of these experiments). Theaverage hydrogen generation rate was dened based on a yieldaround 70%, meaning one data point where the yield reachedaround 70% was chosen and the average rate was calculatedaccordingly. Fig. 7a and b shows that Na2SiO3 had a positive effectin both 0.57 M and 2.14 M KOH aqueous solutions as 0.5 g dm�3

of H2SiO3 was added. However, Fig. 7c and d show that Na2SiO3

only had a positive effect in a 2.14 M KOH aqueous solution as4 g dm�3 of H2SiO3 was added. In brief, Na2SiO3 may affect thehydrogen production rate and yield positively as a small quantityof H2SiO3 was added. The optimized hydrogen production rateand yield were obtained as 4.72 � 10�3 g(H2) per s per g(Si) and92%, respectively. The optimized production rate is comparedwith the rates reported from current existing hydrogen produc-tion techniques compiled in Table 1, showing the superiority ofthis additives-mediated rapid hydrogen production method.

To exhibit the potential of kerf loss silicon as a convenientenergy carrier, a stainless steel reactor integrated with a fuel cell(two-stacked polymer electrolyte membrane cell) was designedto generate electricity for an electric model vehicle (Fig. 8a andVideo S2†). A multimeter was placed in between to monitor theelectricity output from the fuel cell. A large-scale experiment wasconducted by mixing 200 mg kerf loss silicon with 2.14 M KOHaqueous solution in a stainless steel reactor at 343 K. Uponinjection of the etching solution, the temperature of the reactorincreased because of the exothermic silicon–water reaction(Fig. 8b). The kinetics were around 3.8 times slower compared toonly using 10 mg kerf loss silicon under the same condition,which may be the consequence of a non-uniform reaction asa larger quantity of kerf loss silicon was used (Fig. 8b). A fuel cellconnected to an electric model vehicle was then supplied withthe generated hydrogen as the reaction completed (Fig. 8c). Theslightly increasing voltage over time might be because of theaccumulation of water vapor inside the fuel cell, which is similarto the behavior of a proton-exchange-membrane fuel cell(PEMFC) observed by Paci.55 As shown in Fig. 8c, stable workingvalues were quickly reached (1.1 V and 0.54 A) and kept forminutes. In addition, the process was used for hydrogen storageby connecting the stainless steel reactor to a gas tank (Fig. 8d).

J. Mater. Chem. A, 2016, 4, 12921–12928 | 12925

Fig. 7 Hydrogen generationwith various combinations of additives where KOH, Na2SiO3, and H2SiO3 are abbreviated as K, M, and S, respectively.The plus (+) sign represents a higher concentration of respective additives and vice versa. The detailed recipes are listed in the ESI.† Effects ofaltering the concentration of Na2SiO3 on hydrogen generation when the concentration of KOH in the etching solution is (a) 0.57 M and (b) 2.14 Mwith the concentration of H2SiO3 fixed at 0.5 g dm�3. Effects of altering the concentration of Na2SiO3 on hydrogen generation when theconcentration of KOH in the etching solution is (c) 0.57 M and (d) 2.14 M with the concentration of H2SiO3 fixed at 4 g dm�3. All results are basedon reacting kerf loss silicon with the respective etching solutions at 343 K.

Journal of Materials Chemistry A Paper

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article Online

When 500 mg of kerf loss silicon and 30 ml of etching solutionmixed in the reactor at 343 K, the ideal amount of generatedhydrogen was expected to be 72 mg. The pressure within thereactor increases and nally stabilized at 136.1 psi, which wasequivalent to approximately 44 mg hydrogen. Once the outputvalve was opened, allowing the hydrogen inside the reactor to

Table 1 Performance of various hydrogen production methods

Type Catalyst Te

Etching of Si KOH 34Etching of Si NaOH N/Etching of Si NaOH 37Etching of Si NH3 33Etching of Si NaOH 35Etching of Si KOH R.Electrochemical SiNWs/FeP N/Photoelectrochemical TiO2/RGO/Cu2O N/Photocatalytic Si 29Ethanol steam reforming Meso-cLaNiAl 87Fermentative Microbes 30

12926 | J. Mater. Chem. A, 2016, 4, 12921–12928

ow into the gas tank, a nal pressure of 41.5 psi was observed inVideo S3,† which was in good agreement with the estimatedpressure of 42 psi. In both cases, kerf loss silicon sufficientlyexhibited its quality as a direct and ready source for hydrogengeneration.

mp. (K) Rate (gH2per s per gSi) Reference

3 4.72 � 10�3 This workA 9.33 � 10�9 (gH2

per s) 263 2.03 � 10�4 273 6.48 � 10�5 283 5.56 � 10�4 29T. 1.48 � 10�4 30A 9.85 � 10�9 (gH2

per s) 31A 9.92 � 10�10 (gH2

per s) 328 1.89 � 10�7 343 N/A 373 N/A 39

This journal is © The Royal Society of Chemistry 2016

Fig. 8 (a) Instrument set-up for large scale hydrogen production and the application to generate electricity. (b) Amount of hydrogen productionand temperature profile for the reaction of 200mg kerf loss silicon with 2.14 M KOH aqueous solution at 343 K. (c) Profiles of voltage and currentsupplied by the fuel cell to power up the electric model vehicle. (d) Instrument set-up for hydrogen production and storage.

Paper Journal of Materials Chemistry A

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article Online

Conclusions

An approach for hydrogen production based on the reactionbetween water and kerf loss silicon collected from wafer sawingwaste was reported with a high hydrogen generation rate andyield. An additive-mediated hydrogen production can be madeby combining sodium metasilicate (Na2SiO3) and silicic acid(H2SiO3), reaching a hydrogen production rate as high as 4.72�10�3 g(H2) per s per g(Si) with a yield of 92%. The scale-up ofthis process can be easily achieved and integrated for many usesbecause of the simplicity of the experimental apparatus, thedirectness of the chemical reactions, and the low-cost of thereactive materials. A silicon-based hydrogen generation batchreactor was designed to be integrated with a fuel cell or a gastank for electricity generation or hydrogen storage, showing kerfloss silicon is an efficient and convenient energy carrier.

Acknowledgements

The authors acknowledge the nancial support by the Ministryof Science and Technology of Taiwan (NSC 102-2221-E-007-023-MY3, NSC 102-2221-E-007-090-MY2, NSC 101-2623-E-007-013-IT, and NSC 102-2633-M-007-002), the Ministry of EconomicAffairs, Taiwan (101-EC-17-A-09-S1-198), National Tsing Hua

This journal is © The Royal Society of Chemistry 2016

University (102N2051E1), and the assistance from Center forEnergy and Environmental Research, National Tsing-HuaUniversity.

Notes and references

1 I. A. Carbajal-Ramos, M. F. Gomez, A. M. Condo, S. Bengio,J. J. Andrade-Gamboa, M. C. Abello and F. C. Gennari,Appl. Catal., B, 2016, 181, 58–70.

2 M. C. Sanchez-Sanchez, R. M. Navarro and J. L. G. Fierro,Catal. Today, 2007, 129, 336–345.

3 L. Zhang, W. Li, J. Liu, C. Guo, Y. Wang and J. Zhang, Fuel,2009, 88, 511–518.

4 D. Zanchet, J. B. O. Santos, S. Damyanova, J. M. R. Gallo andJ. M. C. Bueno, ACS Catal., 2015, 5, 3841–3863.

5 F. E. Osterloh and B. A. Parkinson, MRS Bull., 2011, 36, 17–22.

6 R. Liu, Z. Zheng, J. Spurgeon and X. Yang, Energy Environ.Sci., 2014, 7, 2504–2517.

7 C. Zhang, W. Fan, H. Bai, X. Yu, C. Chen, R. Zhang andW. Shi, ChemElectroChem, 2014, 1, 2089–2097.

8 Y. Yavor, S. Goroshin, J. M. Bergthorson and D. L. Frost, Int. J.Hydrogen Energy, 2015, 40, 1026–1036.

9 W. M. Budzianowski, Energy, 2012, 41, 280–297.

J. Mater. Chem. A, 2016, 4, 12921–12928 | 12927

Journal of Materials Chemistry A Paper

Publ

ishe

d on

19

July

201

6. D

ownl

oade

d by

Nat

iona

l Tsi

ng H

ua U

nive

rsity

on

08/0

9/20

17 1

0:19

:50.

View Article Online

10 H. L. Maclean and L. B. Lave, Prog. Energy Combust. Sci.,2003, 29, 1–69.

11 G. W. Crabtree, M. S. Dresselhaus andM. V. Buchanan, Phys.Today, 2004, 57, 39–44.

12 S.-I. Fujita, H. Kawamori, D. Honda, H. Yoshida and M. Arai,Appl. Catal., B, 2016, 181, 818–824.

13 L. Gradisher, B. Dutcher and M. Fan, Appl. Energy, 2015, 139,335–349.

14 M. Cargnello, A. Gasparotto, V. Gombac, T. Montini,D. Barreca and P. Fornasiero, Eur. J. Inorg. Chem., 2011,2011, 4309–4323.

15 K. Shimura and H. Yoshida, Energy Environ. Sci., 2011, 4,2467–2481.

16 R. M. Navarro, M. C. Sanchez-Sanchez, M. C. Alvarez-Galvan,F. D. Valle and J. L. G. Fierro, Energy Environ. Sci., 2009, 2, 35–54.

17 D. Merki, S. Fierro, H. Vrubel and X. Hu, Chem. Sci., 2011, 2,1262–1267.

18 M. Wang, L. Chen and L. Sun, Energy Environ. Sci., 2012, 5,6763–6778.

19 R. Van De Krol, Y. Liang and J. Schoonman, J. Mater. Chem.,2008, 18, 2311–2320.

20 E. L. Miller, Energy Environ. Sci., 2015, 8, 2809–2810.21 K. Maeda, K. Teramura, D. Lu, T. Takata, N. Saito, Y. Inoue

and K. Domen, Nature, 2006, 440, 295.22 Y. Zhang, J. Liu, G. Wu and W. Chen, Nanoscale, 2012, 4,

5300–5303.23 G. Yang, W. Yan, Q. Zhang, S. Shen and S. Ding, Nanoscale,

2013, 5, 12432–12439.24 T. Zhou, Y. Du, A. Borgna, J. Hong, Y. Wang, J. Han,

W. Zhang and R. Xu, Energy Environ. Sci., 2013, 6, 3229–3234.25 P. Brack, S. E. Dann, K. G. U. Wijayantha, P. Adcock and

S. Foster, Energy Sci. Eng., 2015, 3, 535–540.26 Y. Hu, H. Yan, K. Liu, H. Cao and W. Li, Int. J. Hydrogen

Energy, 2015, 40, 8633–8641.27 E. R. Weaver, J. Ind. Eng. Chem., 1920, 12, 232–240.28 S. Litvinenko, S. Alekseev, V. Lysenko, A. Venturello,

F. Geobaldo, L. Gulina, G. Kuznetsov, V. Tolstoy,V. Skryshevsky, E. Garrone and D. Barbier, Int. J. HydrogenEnergy, 2010, 35, 6773–6778.

29 H.-S. Yoo, H.-Y. Ryu, S.-S. Cho, M.-H. Han, K.-S. Bae andJ.-H. Lee, Int. J. Hydrogen Energy, 2011, 36, 15111–15118.

30 G. Bernhard, K. Dmitry and S. Olga, Nanotechnology, 2011,22, 305402.

31 C. Lv, Z. Chen, Z. Chen, B. Zhang, Y. Qin, Z. Huang andC. Zhang, J. Mater. Chem. A, 2015, 3, 17669–17675.

32 W. Fan, X. Yu, H.-C. Lu, H. Bai, C. Zhang and W. Shi, Appl.Catal., B, 2016, 181, 7–15.

12928 | J. Mater. Chem. A, 2016, 4, 12921–12928

33 H. Ahmad, S. K. Kamarudin, L. J. Minggu and M. Kassim,Renewable Sustainable Energy Rev., 2015, 43, 599–610.

34 F. Dai, J. Zai, R. Yi, M. L. Gordin, H. Sohn, S. Chen andD. Wang, Nat. Commun., 2014, 5, 3605.

35 A. N. Fatsikostas, D. I. Kondarides and X. E. Verykios, Chem.Commun., 2001, 851–852.

36 M. Ni, D. Y. C. Leung and M. K. H. Leung, Int. J. HydrogenEnergy, 2007, 32, 3238–3247.

37 H. Ma, L. Zeng, H. Tian, D. Li, X. Wang, X. Li and J. Gong,Appl. Catal., B, 2016, 181, 321–331.

38 I. K. Kapdan and F. Kargi, Enzyme Microb. Technol., 2006, 38,569–582.

39 N. F. Azman, P. Abdeshahian, A. Kadier, H. Shukor,N. K. N. Al-Shorgani, A. A. Hamid and M. S. Kalil,Renewable Energy, 2016, 93, 700–708.

40 D. Wen, Energy Environ. Sci., 2010, 3, 591–600.41 E. I. Shkolnikov, A. Z. Zhuk and M. S. Vlaskin, Renewable

Sustainable Energy Rev., 2011, 15, 4611–4623.42 D. B. Beach, A. J. Rondinone, B. G. Sumpter, S. D. Labinov

and R. K. Richards, J. Energy Resour. Technol., 2006, 129,29–32.

43 N. Auner and S. Holl, Energy, 2006, 31, 1395–1402.44 Y. Yavor, S. Goroshin, J. M. Bergthorson, D. L. Frost, R. Stowe

and S. Ringuette, Int. J. Hydrogen Energy, 2013, 38, 14992–15002.

45 H. Z. Wang, D. Y. C. Leung, M. K. H. Leung and M. Ni,Renewable Sustainable Energy Rev., 2009, 13, 845–853.

46 T. Hiraki, M. Takeuchi, M. Hisa and T. Akiyama, Mater.Trans., 2005, 46, 1052–1057.

47 C.-Y. Cho, K.-H. Wang and J.-Y. Uan, Mater. Trans., 2005, 46,2704–2708.

48 M. H. Grosjean, M. Zidoune, L. Roue and J. Y. Huot, Int. J.Hydrogen Energy, 2006, 31, 109–119.

49 M.-H. Grosjean and L. Roue, J. Alloys Compd., 2006, 416, 296–302.

50 H. Seidel, L. Csepregi, A. Heuberger and H. Baumgartel, J.Electrochem. Soc., 1990, 137, 3612–3626.

51 E. D. Palik, V. M. Bermudez and O. J. Glembocki, J.Electrochem. Soc., 1985, 132, 871–884.

52 S. Limaye, S. Subramanian, B. Goller, J. Diener andD. Kovalev, Phys. Status Solidi A, 2007, 204, 1297–1301.

53 F. Erogbogbo, T. Lin, P. M. Tucciarone, K. M. Lajoie, L. Lai,G. D. Patki, P. N. Prasad and M. T. Swihart, Nano Lett., 2013,13, 451–456.

54 L. Soler, A. M. Candela, J. Macanas, M. Munoz and J. Casado,Int. J. Hydrogen Energy, 2009, 34, 8511–8518.

55 V. R. Albertini, B. Paci, F. Nobili, R. Marassi and M. DiMichiel, Adv. Mater., 2009, 21, 578–583.

This journal is © The Royal Society of Chemistry 2016