Thermochemical water-splitting cycles producing hydrogen are reviewed, emphasizing those that have promise to be coupled with concentrated solar high-temperature heat available in "sunbelt regions". There are solar peculiarities in comparison to conventional thermochemical processes: high thermal flux density and frequent thermal transitions because of the fluctuating insolation. Therefore, conventional industrial thermochemical processes are generally not suitable for solar-driven processes. Thus, adaptation of the process to solar high-temperature heat is required. A chemical process should be operated most efficiently without interruption. However, the solar high-temperature heat supply is intermitten; that is, it is interrupted during the night and by cloud passage during the day. To solve this problem, it is ideal to store solar high-temperature heat in a thermal storage system and use the stored thermal energy continuously (24 h a day) for the process. However, thermal storage at temperatures above 1000 K is very difficult to achieve on an industrial scale. Generally speaking, even for a solar thermochemical process that is planned to be operated only during daytime, chemical engineers are not in agreement as to whether the chemical process can be stopped, for example, for 30 min by cloud passage, and then operated effectively again. The thermal inertia problems and startup difficulties in an intermittently operated system will be more severe if the chemical plant is larger, with larger heat capacities of the reactors, separators, piping, etc. It is also difficult for a complicated process with a number of steps (e.g. reaction, separation, and concentration) to be effectively operated with quick response to an intermittent heat supply from concentrated solar radiation. From this perspective, for the development of an effective solar thermochemical process, the solar reactor should have a quick response to the intermittent heat supply, and there should be as few steps as possible in the process. Section 2 reviews two-step water-splitting processes involving redox metal oxides. The chemistry of the processes is well discussed. The reactions are very simple, as the process contains only a few separation steps in which technical difficulty does not lie. However, the two-step thermochemical cycle requires heat above 1500 K, minimum. This presents a challenge to solar reactor concepts and has resulted in the research and development of several types of windowed solar reactors. Direct irradiation of the redox working material enables the reactor to respond very quickly to the intermittent heat supplied by concentrated solar radiation. For iron-based oxide processes, several types of windowed solar reactors are currently being developed - monolithic, foam, rotary-type, CR5, and internally circulating fluidized bed reactors. In these reactors, both the thermal reduction and hydrolysis steps are planned to be carried out in a single reactor. However, basic kinetic studies on the thermal reduction and hydrolysis reactions are not yet sufficient to prove the feasibility of the reactor. The kinetics of the reactions must be an important key factor for each reactor. Thus, further development of active redox materials and devices will have a very large impact on the improvement of these reactors. Further investigation of the kinetics of the reactions in the iron oxide based processes is urgently needed. For the ZnO process, the kinetics of the reactions are being investigated well. However, two reactors are required in the current operational concept: one for the thermal reduction of ZnO and one for the hydrolysis of Zn. The hydrolysis reactor still requires temperatures above 1023 K for Zn evaporation. In addition, the volatile oxide process requires an additional product quenching system. In the case of the ZnO process, the product recombination of zinc vapor and oxygen severely decreases the chemical conversion in the absence of efficient quenching. Some of the reactor concepts must use solar thermal energy for the hydrolysis step as well as for the thermal reduction step if the hydrolysis step requires high-temperature operation, and heat recovery from the high-temperature thermal reduction step is insufficient. The hydrolysis step in the metal oxide two-step cycles is exothermic, but, generally, the heat released is not particularly great. Heat losses are increased at higher temperature, and heating the steam feed to a higher temperature requires a higher temperature heat supply. This solar-driven hydrolysis step results in the heliostat field being utilized partially by the hydrolysis process. Normally, as heliostats make up well over 50% of the capital cost in a solar thermal hydrogen plant, this has a major effect on economics. Therefore, lower temperature operation of the hydrolysis step is an important factor. If the hydrolysis temperature is high, the heat recovery from the higher temperature thermal reduction step to the hydrolysis step must be efficiently conducted, as in the CR5 concept. The reactor concepts may be successful for a small-scale solar plant, for example, with a few hundred kilowatts of hydrogen production, although fundamental research, such as into active redox materials or device development, is still required in addition to engineering work on the reactor design and operation. However, in scaling up to multi-megawatts size, other engineering difficulties may be found, for example, in making a cluster of the reactors at the top of a solar tower. The scale of one reactor must be limited by the size limitations imposed by its quartz window and, thus, a cluster of the reactors will be required for multi-megawatt scaled application. Section 3 reviews multistep water-splitting cycles capable of working below 1200 K. They are thermochemical cycles with more than three steps and thermochemical-electrochemical hybrid cycles with more than two steps. However, the promising cycles all suffer from the use of corrosive reactants and difficulties in the separation steps. For cycles with more than three reaction steps, the efficiency may be extensively reduced when coupled to concentrated solar high-temperature heat, instead of nuclear heat. The case of the UT-3 process is a good example, as pointed out by Teo et al. From this point of view, the "two-step" Westinghouse cycle is the most promising process among the multistep water-splitting cycles capable of working below 1200 K, although it is not a pure thermochemical cycle. In hybrid cycles, an important criterion is the minimum voltage for electrochemical step efficiency. There is always a compromise between acid concentration and cell voltage. Materia selection for the acid decomposition step of the Westinghouse cycle is also still challenging, as well as the electrode selections for the electrochemical step. The DLR is developing a solar direct absorbing volumetric receiver-reactor, which may have quick response to the intermittent heat supply. However, to allow the reactor concept to be upscaled to multi-megawatts scale, a cluster of the reactors is required at the top of a solar tower because the reactors are windowed. © 2007 American Chemical Society.
