Int. J. Hydrogen Energy, VoL 10, No. 9, pp. 571-576, 1985. Printed in Great Britain.

0360-3199/85 $3.00 + 0.00 Pergamon Press Ltd.

~) 1985 International Association for Hydrogen Energy.

HYDROGEN PRODUCTION FROM WATER: SUMMARY OF RECENT RESEARCH AND DEVELOPMENT PRESENTED AT THE FIFTH WHEC

T. OHTA*, J. E. FUNK'~, J. D. PORTER:~ and B. V. TILAK§

*Hydrogen Energy Research Laboratories, Yokahama National University, Yokohama, Japan. tDept, of Mechanical Engineering, University of Kentucky, Lexington, Kentucky, U.S.A. ,AT & T, Bell Laboratories, Murray Hill, New Jersey,

U.S.A. §Research Center, Occidental Chemical Corporation, Grand Island, New York, U.S.A.

(Received for publication 25 February 1985)

AbstraetnAbout 44% of the R & D papers presented at the 5th World Hydrogen Energy Conference (from 15 to 19 July 1984, Toronto, Canada) were concerned with methods of hydrogen production from water. Both traditional and innovative technologies for water electrolysis were reviewed in detail at this conference. The main installations of water electrolysis were introduced. Thermochemieal methods of water decomposition, which were flourishing at the 1st and 2nd WHECs, have declined in emphasis since then. Thirteen papers were presented on the thermochemical method, but no radical improvements were described. Photolytic methods of water decomposition, and especially photoassisted electrolysis using semiconductor electrodes, have proved to be one of the most popular areas of research, beginning about ten years ago. Theoretical treatment, as well as efficiency limits, were discussed in detail, but inexpensive, appropriate electrode materials, with better solar efficieneies than TiO2, although sharing its chemical stability, were not reported.

1. I N T R O D U C T I O N

Hydrogen Energy Systems comprise the ideal total energy system for the future. This means that the devel- opment of the elementary technologies of this system should not be influenced by short-term fluctuations in energy supply-demand relations. Rather technological innovation will select out the most appropriate tech- nologies over time. Every World Hydrogen Energy Conference has truly reflected the trends in R & D of the high profile and innovative technologies. Examples of such technologies are thermochemical water decom- position, the hydrogen jet airplane, and so on. Some of these technologies have become deeply rooted in active research fields and some of them have declined in emphasis. Hydrogen production is the most fun- damental part of the hydrogen energy system, and has always been the object of intense and vigorous research and development. A large body of knowledge con- cerning water decomposition has been accumulated so far. This will be a latent, but potent tool with which to solve the energy problems of the future.

In this paper, we shall review the present status of the science and technology of water decomposition, based upon the presented papers in the Hydrogen Pro- duction sessions of the 5th World Hydrogen Energy Conference, which took place 15-19 July 1984, in Toronto, Canada. A partial pre-printed proceedings for the whole conference has been published [1].

2. T H E R M O D Y N A M I C R E V I E W OF W A T E R D E C O M P O S I T I O N

When water in a liquid state is split into a gaseous state of hydrogen and oxygen, that is to say, when the

chemical equation:

H20(I) = H2(g) + ½02(g) (2.1)

holds, under the condition that the temperatures and the pressures of the initial and the final states are the same, the change of enthalpy is given by:

A H = A G + TAS, (2.2)

where AG, AS, and T denote the change of Gibbs' free energy, the change of entropy, and the absolute temperature, respectively. The second term TAS in the right hand side of equation (2.2) expresses the heat quantity needed to split water at the temperature T.

At T = 298 K, we have for the process of equation (2.1)

AH = 68.32 kcal mo1-1 1

AG = 56.69 kcal mol-1 ~ (2.3)

AS = 39.00 cal deg -1 mo1-1J

As the temperature rises, the entropy change AS decreases slightly. Figure 1 shows the enthalpy H vs temperature T diagram for both water (H > 0) and the gaseous states of the split hydrogen and oxygen (H > 0). Two points should be noted from the figure. The first point is that at the temperature Tc = A H / A S , where AG = 0, we get a gaseous mixture of hydrogen and oxygen. An exact calculation using the appropriate par- tition function gives the result that not only hydrogen and oxygen molecules, but also hydroxyl radicals and hydrogen and oxygen atoms can exist at such high temperatures. A major concern of the direct thermal water decomposition method is thus how to separate hydrogen from this mixed gas.

571

572

6o

l.O

A

\

-ZO

-60

-80 0

G' / HHz+ LOz

S00 1000 1500 Z000 ZS00 3000 TEMPERATURE ( K )

Fig. 1. Enthalpy-temperature diagram for water-splitting.

The second point is that more than 83% of the energy needed to split water at room temperature is the Gibbs' free energy. This is the reason why large quantities of energy in the form of electrical energy, photon energy or chemical energy, are necessary in order to carry out this process. Considering the alternatives, this is also the reason why water electrolysis is the traditional and practical technology for water splitting at room temperature.

At moderate temperatures considerably lower than Tc (=4150 K under 1 atm), we need the Gibbs' free energy corresponding to the distance between the broken line (horizontal) and the solid line of the enthalpy of HzO in Fig. (1). If this free energy is provided by electrical energy, the method of water splitting is called water electrolysis. The electrical energy co~esponding to the value of AG given by Eq. (2.3) is ecluivalent to electrical energy of AE = 2.46 V. However, in the water decomposition process two elec- trons can participate, so that the needed voltage for water electrolysis is ideally 1.23 V. In the practical elec- trolyzer, there will be overvoltage whose amount is typically about 0.7 - 0.8 V at room temperature. This overvoltage depends upon the current density, tem- perature, the electrode material, and so on; reducing its magnitude is the goal of much R & D effort in this field. Research on water electrolysis was well repre- sented at the 5th WHEC and the results will be described in the next section.

Heat energy producing temperatures around 1000 K can be provided by solar radiation and by nuclear fission and fusion processes without great difficulty. Therefore, if chemical energy can be made to supply the required Gibbs' free energy at such a temperature, then this could be an effective technique for water decomposition. The method of thermochemical water decomposition was born from such an idea.

T. OHTA et al.

Consider, for example, one set of three chemical equations:

A + B + H 2 0 - o A H 2 + B O , [

AH2 "--* A + H2 I (2.4) BO ~ B + ½Oz,

where A is an oxidizing agent and B is a reducing agent. The key reaction is the first in equation (2.4), which splits H20 into AH2 and BO. This means that chemical energy provides most of the necessary free energy in this scheme. The remaining two reactions can be advanced by heat, at a lower temperature, or by using electrical energy with a lower voltage. The process denoted by A B ' C ' D ' E ' F ' G ' H ' I ' D E in Fig. 1 expresses a thermochemical cycle for water decomposition. The compensation of the free energy such as B'C' , or E 'F ' G' are due to the released chemical energies and the change in Gibbs' energy with temperature.

In Fig. 2, we show a schematic diagram of a photo- electrochemical cell (PEC) which has an n-type semi- conductor (e.g. TiO2) electrode 1, platinum electrode 2, and ion-conducting diaphragm 3. If an ultraviolet beam is irradiated on 1, hydrogen and oxygen are evolved separately in the gas burettes 4, and an electric current flows in the circuit 5. This phenomenon can be interpreted as follows: The ultraviolet beam (hv) generates energetic electron (e-) and positive hole (p÷) pairs within the semiconductor electrode:

TiO2 + 2hv---~ 2e- + 2p-. (2.5)

In the case of n-type TiO2, the photogenerated holes are the minority carriers and so have a short lifetime within the semiconductor. Consequently, it is the majority carriers, the electrons, which move through the outer circuit to recombine with protons at the surface of the counter electrode:

2e- + 2H ÷ ~ H2. (2.6)

The protons H + come to the platinum electrode from the semiconductor electrode, through the semiper-

4 6 4

5

hi/

Fig. 2. Schematic diagram of electrochemical photocell. 1: TiO2 electrode, 2: Platinum, electrode, 3: Diaphragm (Agar salt bridge), 4: Gas burette, 5: Load resistance, 6: Voltmeter.

HYDROGEN PRODUCTION FROM WATER 573

meable diaphragm in the cell. Protons are formed at the semiconductor electrode surface by the process of water oxidation by the energetic photogenerated min- ority carriers:

2p* + H20---* ½0 + 2H ÷. (2.7)

An analogous set of processes can be considered for the case of a p-type semiconductor where the photo- generated minority carriers are energetic electrons. The most important challenge in this method of photo- assisted electrolysis is to find effective semiconductor electrode materials. Among the conditions imposed upon the materials are: (1) chemical stability in the solution, under irradiation, (2) appropriate width of the energy gap, which should be just greater than the water- splitting voltage, and (3) the conduction band edge should lie negative of the E ° for H"/H2. Devising appropriate schemes, defining the constraints imposed upon efficient materials and exploring solar conversion efficiencies were major concerns of the photoproduction sessions at the 5th WHEC. Typical n-type semi- conductors so far reported are TiO2, SrTiO3, WO3, SnO 2, BaTiO3, KTaO3, Bi203, and Fe203.

3. INDUSTRIAL WATER ELECTROLYSIS AND ELECTROCATALYSIS

The papers presented in the water electrolysis division represent significant activity and progress towards achieving an energy consumption of less than 4.5 kWh Nm -3 via improvements in water electrolysis, and much lower energy consumption via vapor phase electrolysis because of the thermodynamic and kinetic advantages associated with high temperature operations.

Justification for continued efforts to develop cost- effective high temperature electrolyzers was empha- sized by trade-off analyses and techno-economic analy- sis of water vapor electrolysis cells operating at 1000°C and integrated into process design involving electrical and thermal energy derived from coal. These later esti- mates projected production costs of 17¢ to 22¢ Nm -3 of H2, which are double the steam-reforming costs, half the cost of conventional water electrolysis and cost- comparable to the coal gasification process.

It is worth mentioning that some inroads have been made into understanding the mechanistic aspects of vapor phase electrolysis operations although more con- tributions would have been profitable.

Presentations were also devoted to asbestos separator development and use of WC cathodes for H2 production from acidic media which are worthy of further pursuits.

Tables 1 and 2 show the current developments in industrial water electrolysis operations in the world and the vapor-phase electrolysis. One of the most important factors in the water electrolysis is to optimize electro- catalysts which are primarily Ni and Co based compositions.

The catalytic factors involved in the H2 evolution reaction have been classically explained in terms of heats of adsorption of hydrogen and percentage d- character of the metals, and exhibited in terms of vol- cano plots. Parsons addressed the issue of achieving enhanced catalysis by combining metals on either side of the volcano curve and came to the conclusion that it is almost impossible to realize enhanced reaction rates assuming non-interaction between the dual sites and the presence of single adsorbed intermediates. However, studies reported in this symposium on several high surface-area nickel based catalysts such as intermetallic compounds of Ni, Ni-Mo, Ni-Co, Ni-Fe, Ni-Mo-Cd, NiSx coatings containing FeSx and MoSx, Raney nickel, Ni-Co spinel coatings formed by low-pressure plasma- spraying techniques have shown significant activity and, hence, voltage savings under water electrolysis oper- ating conditions. These results suggest that the mech- anism of H2 evolution involves more than one adsorbed intermediate as shown by an analysis of the EMF decay data. An alternate mechanism is also invoked in terms of Brewer-Engel theory. It is interesting to note that studies on Ag-Pd alloy show no direct relationship of i0 to percentage d-character. While several coatings have been developed exhibiting electrocatalytic activity, an understanding of the kinetics on high surface area composites is lacking and should be examined in detail to be in a position to predict and develop cost-effective catalysts.

Pathways involved in oxygen evolution reaction are more complex than those proposed for the hydrogen evolution reaction since discharge of 02 occurs on oxide covered or non-metallic surfaces. In recent years, a number of mixed transition metal oxide catalysts--- especially spinels and perovskites of Ni, Co and/or Fe were examined. Of these, NiCo204 and Li-doped Co304 appear promising. Oxygen evolution on these oxides appear to involve formation and subsequent decomposition to higher oxide, and good correlation between the redox oxide couple and the minimum potential for 02 evolution has been established. Never- theless, the oxygen overpotential is still high and iden- tification of pathways, and relationship of the kinetics to the electronic and steric properties requires deep studies to realize an understanding of the proposed interaction of reactants and absorbed intermediates with oxide surfaces.

One of the topics of applied concern addressed during the session is the deactivation mechanism of cathodes. Under open-circuit conditions, galvanic effects seem to be causal for the loss of catalytic activity and under operating conditions, mechanical erosion, chemical/ electrochemical dissolution and/or iron contamination appears to be the contributing factor for catalyst degra- dation.

Finally, two papers worthy of attention relate to electrode configuration involving teflon-bonded elec- trodes and integrated electrode/diaphragm structures comprised of alkaline-earth titanates with highly porous catalyst layers formed by sintering without any void-

574 T. OHTA et al.

Table 1. Industrial water electrolysis

Organization/Manufacturer Present status/Operating conditions Future developments

Asahi Glass • Perfluoro carboxylated membrane • Activated cathode/Re-Rh modified Raney nickel

anode • Zero gap • 1.7V at 7 0 A d m -2 at ll0*C and 30wt-% KOH

• Continued life testing

Julich's KFA Electrolyzer • Bipolar Allis-Chalmers type design • Raney Ni activated by Lye treatment • Ceramic NiO based diaphragm on Ni gauze • 1.5V at 4kA m -2 at -100°C and 10M KOH;

5000 hr life

• Scale-up from 200 cm: to 1 m 2 cells

Life Systems, Inc. • For stand-alone H2 production • 180 psia at 82°C; 1.5 V at 160 mA cm -2 for >27,000 h

with 0.1 sq. ft cells, and >2,500 h with I sq. ft. cell

• Scale-up to 15 kW electrolyzer

L'Autonome • Bipolar with modified asbestos diaphragm • Raney Ni cathode and LaSrCoO3 on Raney Ni as

anode with zero gap • 4 bar, ll0°C, 1.8 V at 4 kA m -z with 0.2 m 2 cells for

~300 h

• 20 cell stack with improved electrodes and design

Alsthom Atlantique • Bipolar with catalyzed sintered nickel electrodes • 120°C; 3bar; 2.1 V at 10kAm -2 and 40% KOH • Low capital cost technology

• 2.4 MW unit designed

Inorganic Membrane Technology • Bipolar with inorganic membranes based on polyantimonic acid

• NiS cathode, Co spinel on Ni as anode • 1.75 V at 10 kA m -2 at 120°C (4.17 kWh Nm -3) • Costs projected at $ 540 m -2 (½ of SPE technology?)

• 1 MW module development

NORSK Hydro • bipolar filter-press type • Asbestos diaphragm with catalytic cathodes • 1.8 V at 2.5 kA m -2 (4.3 kWh Nm-3); 80°C and 25%

KOH) • H2 Production Costs: 15¢ m -3 of H 2

• Large capacity (5-10 kA) New generation cells

Electrolyzer, Inc. • 100 kA unipolar generation 1 cells • Proprietary electrocatalysts • 1.85 V at 2.5 kA/m-2; 70°C and 25 wt-% KOH

(4.4 kWh Nm -3) • 7,000 h operation to-date

• Long-term performance Data for 35 MW Design and installation

Showa Denko and Mitsubishi • Porous PTFE impregnated with potassium titanate • High surface-area nickel based catalysts • 1.67-1.73 V at 0.4 A cm -2 at 120°C (4 kWh Nm -3)

with cells producing 20 Nm 3 of H2/h

• None

Public Service Electric & Gas Co.

• GE's SPE Technology for on-site production for cooling electric generators

• Performed well over 18 months of operations

• Economic assessment in progress

Teledyne Energy Systems • Field testing of HGS prototype electrolyzer under various load profiles (5.3 kWh Nm -3) for cooling electric generators

• Long-term testing for design improvements and system reliability

HYDROGEN PRODUCTION FROM WATER

Table 2. Vapor-phase electrolysis

575

Organization/Manufacturer Present status/Operating conditions Future developments

Dornier/Lurgi • Integrated modular system with serial and parallel • Design of 3.5 kW unit tubular cells with yttria stabilized zirconia

• 1.07 V at 0.3 A cm -2 (2.57 kWh Nm -3) with single cells

• Capital costs -1200 $ Nm -3 at 3.2 kWh Nm -3

Westinghouse • Solid oxide fuel cells in electrolyzer mode • Tubular cells with Sr modified LaMnO3 anode and

Ni cermet cathode • 1.23 V at 300 mA/cm -~ at 800--1050°C for ~1000 h

• Optimization of cell design and operations

fraction and/or ohmic penalties. This aspect should be a topic of major discussion in the next W H E C Conference along with long-term performance and economic aspects related to the electrocatalysts.

4. I N N O V A T I V E W A T E R SPLITTING T E C H N O L O G I E S

Thermochemical water splitting

Ten years ago, more than one hundred kinds of thermochemical cycles had been proposed and every energy research institute had its own thermochemical cycles to be developed. However, in the past few years, the development of thermochemical splitting of water has declined. Less than ten cycles are still being studied and they were discussed at the 5th WHEC. There were no new cycles proposed, but the successes of the long- studied systems were reviewed. There were also discus- sions on the coupling of a solar central receiver to a sulfuric acid decomposition process and the direct thermal decomposition of water process.

There was no discussion of new approaches to the search for better cycles. The absence of a strong end-use driving force and questions about the primary energy source seem to be slowing the pace of development in the thermochemical field.

Photon process

As has been stressed in the Introduction, one needs Gibbs' free energy as well as heat energy in order to split water. The free energy can be provided as electrical energy, photon energy, or chemical energy. Therefore, it stands to reason that it would be reasonable to try to utilize solar photon energy to split water, as nature carries out in one part of the photosynthesis process.

In the 5th WHEC, thirteen papers on the method of photoassisted hydrogen production using semicon- ductor electrodes were presented. This is one of the most daunting and challenging methods of hydrogen production, and the one most removed from practice at present. Massive semiconductor materials which can be used in photoelectrochemical systems were the subject of the bulk of the reports. TiOe was well-represented.

This venerable photoanode material remains the subiect of vigorous and wide-ranging materials research.

Fabrication of optimal polycrystalline material and assessments of present fabrication methods were described. Other transition metal oxides, variously incorporating Fe, Nb, Co, and V, were represented. Studies using CdS and MoS2, and I I I -V semiconductors GaAs and InP were also presented. Details concerning novel alkaline earth and transition metal diphosphides were presented for the first time, in a solar hydrogen context.

One paper emphasizing the importance of interfacial kinetics at the semiconductor-electrolyte junction was presented. There was considerable discussion con- cerning the calculation of solar-energy to hydrogen con- version efficiency. Microheterogeneous systems, vari- ously employing photosynthetic bacteria, blue-green algae, and surface modified TiO2 metal powders and CdS powders were also described. These systems are inherently inefficient, but their relatively simple engin- eering requirements make them attractive candidates in the long-term, and fundamental research remains active.

Further fundamental study of the semiconductor- electrolyte interface and especially of surface states, which remain particularly poorly characterized and understood, was cited as a short-term goal. Surface modification to achieve surface state 'passivation' and catalysis of productive reactions, the subject of several papers presented at the 5th WHEC, will receive increas- ing attention in the future.

5. CONCLUSION

Review of recent research and development on hydro- gen production from water is described, based on the 5th W H E C papers. Although more than 40% of the papers presented at this multidisciplinary conference were concerned with the production of hydrogen there was no radically innovative research described. This indicates that the field has achieved a measure of matur- ity, and that previous speculation can be replaced, in many instances, by realistic estimates of the possible. This can only be of benefit to those responsible for

576 T. OHTA et al.

defining future socio-economic policies, since the Hydrogen Energy System is becoming definable and quantifiable. In addition, a large body of fundamental knowledge concerning water-electrolysis, thermo- chemical, photochemical, photoelectrochemical, direct thermal, and other methods of hydrogen production has been accumulated. This is a powerful tool for developing future alternative energy schemes.

Acknowledgements--The authors would like to express their

thanks to Dr B. Bel~inger of Institute de Recherche d'Hydro- Quebec. Varennes, Canada, for his kind coordination of the preparation of the present report.

REFERENCES

1. T. N. Veziroglu and J. B. Taylor (eds) Hydrogen Energy Progress V, Proc. 5th World Hydrogen Energy Conference, Toronto, Canada 15-20 July 1984. Vols 1-4, Pergamon Press, New York (1984).