“One-Pot” Solar Fuels

Abstract

Chemists have always taken great pride in simplifying a complex multistep chemical synthesis into a single step process, in just one reactor. The motivation for doing so, apart from demonstrating synthetic competency and elegance, is the pursuit of green and sustainable solutions to chemical problems, as single step strategies have the inherent advantage of being atomically, energetically and economically efficient, as well as environmentally friendly and responsible. Chemists have always taken great pride in simplifying a complex multistep chemical synthesis into a single step process, in just one reactor. The motivation for doing so, apart from demonstrating synthetic competency and elegance, is the pursuit of green and sustainable solutions to chemical problems, as single step strategies have the inherent advantage of being atomically, energetically and economically efficient, as well as environmentally friendly and responsible. This paradigm of converting a multiple step to single step synthesis is ubiquitous in the annals of organic, inorganic, organometallic, coordination, polymer, and biological chemistry as well as materials chemistry. Examples range from pharmaceutical and nanomaterials chemistry to homogeneous and heterogeneous catalysis. In this Preview, I will focus attention on an exciting breakthrough by Aldo Steinfeld and coworkers in the field of solar fuels.1Tou M. Michalsky R. Steinfeld A. Joule. 2017; 1 (this issue): 146-154Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar In brief, their advance has enabled a solar-driven thermochemical CO2 → CO + ½ O2 process to operate continuously, in a single step, at a single temperature, in a single reactor, with 100% selectivity. To place Steinfeld and coworkers’ finding in context, recall that state-of-the-art in the field of solar thermochemical fuels from H2O and CO2, which, incidentally, this group currently holds the record value of energy efficiency,2Marxer D. Furler P. Takacs M. Steinfeld A. Energy Environ. Sci. 2017; 10: 1142-1149Crossref Google Scholar involves a well-established two-step cyclic process. The procedure commences with the thermally induced loss of lattice O2− in the form of O2 from a stoichiometric redox active metal oxide, denoted as MOx. The result of this first reduction step is a non-stoichiometric metal oxide MOx-y containing vacant oxygen sites in the crystal lattice. This reaction follows the balanced equation:MOx → MOx-y + ½y O2 (first reduction step) The first reduction step requires high temperatures, typically around 1,500°C depending on the metal oxide. However, when the temperature is lowered to around 1,000°C, the so-formed oxygen vacancies in MOx-y can abstract oxygen from CO2 in a second oxidation step to form CO, while concurrently reinstating the starting MOx lattice. In this case, the reaction follows the balanced equation:MOx-y + y CO2 → MOx + y CO (second oxidation step) This step readies the MOx for the next cycle. After an initial reduction step, a second oxidation step driven by the reaction with H2O forms H2 according to the balanced equation:MOx-y + y H2O → MOx + y H2 (second oxidation step) The net reactions are CO2 = CO + ½ O2 and H2O = H2 + ½ O2, with fuel and oxygen generated in different steps. These CO2 and H2O splitting cycles can convert the entire spectrum of solar energy, when using a solar thermal receiver, into stored chemical energy in the form of CO and H2. Thus, the two approaches taken together can provide a solar thermochemical route to syngas, comprised of a mixture of H2 and CO. Via conventional catalytic processes, conversion of the syngas to high energy density hydrocarbon fuels can satisfy the needs of the maritime and aviation transportation sectors, without changing the existing global infrastructure. With this brief background of prior work in solar fuels, one can appreciate that while solar thermochemical carbon dioxide and water splitting is a most impressive and promising technology, its practice is not at all straightforward. This is because it necessitates multiple cycling of two redox reactions at two different high temperatures and sophisticated chemical reactors. Imagine now if this two-step, two-temperature process were reducible to a single-step, single-temperature process; it would clearly offer numerous advantages. These include consolidation of the reactor design, reduction of materials stresses and energy losses from temperature and pressure swings, and simplification of integrating the technology with large-scale solar concentration facilities. This single-step concept was proposed four decades ago,3Fletcher E.A. Moen R.L. Science. 1977; 197: 1050-1056Crossref PubMed Scopus (289) Google Scholar but the development of a viable solar reactor technology for splitting CO2 was only recently demonstrated experimentally. Published in this inaugural issue of Joule,1Tou M. Michalsky R. Steinfeld A. Joule. 2017; 1 (this issue): 146-154Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar Steinfeld and coworkers have reduced this single-step process into practice. In a tour de force of high-flux optics, materials science, reaction engineering, and chemical processing, illustrated in Figure 1, their group has elegantly demonstrated that solar radiative heating of a dense-packed, redox-active ceria membrane enables the thermochemical splitting of CO2 to CO and O2. The process has the added key advantage of being able to separate within the same reactor the CO from the O2 into single streams. Based on prior work on thermally induced splitting of H2O and CO2 in ceramic membrane reactors, I envision an extension to solar driven “one-pot” splitting of mixtures of H2O + CO2 into “designer syngas,” in which the CO/H2 ratio is under synthetic control and can therefore be tuned to be used in a wide range of downstream processes. Operating as a fully integrated modular unit, this technology offers the potential to make solar fuels in a single-step process a reality. The heart of the solar reactor shown in Figure 1 is a tube-shaped ceria membrane capped at one end and centered coaxially in an outer tube made of alumina. Concentrated solar irradiation causes thermal dissociation of CO2 on the tube inner surface to form CO and O2. The O2 permeates through the tubular membrane to the outer side, swept away with the inert Ar carrier gas flowing through the surrounding alumina tube, thereby achieving in situ separation of products CO and O2. Sophisticated materials, chemical engineering, and solar thermal reactor design have gone into making this reactor work. However, being a materials chemist, I would like to turn your attention to the remarkable solid-state physical-chemical properties of the ceria membrane that enable the splitting of CO2 to separate streams of CO and O2 occur in a single step. Let us begin with a little background into the structure and conductivity of stoichiometric CeO2 and non-stoichiometric CeO2-x forms of ceria. In its pristine form, CeO2 is pale yellow and a poor ionic and electrical conductor. Its fluorite crystal structure can tolerate high levels of reductive oxygen loss to form CeO2-x containing oxygen vacancies, without structural phase change from room to elevated temperatures. The reduced non-stoichiometric form of CeO2-x at low oxygen deficiencies is blue, originating from Ce(III) → Ce(IV) inter-valence charge transfer electronic excitations, where doubly ionized oxygen vacancies are thought to dominate. At high oxygen deficiencies, its color switches to black arising from conduction electron excitations and a transition to singly ionized vacancies. Non-stoichiometric CeO2-x are good oxide ionic conductors at elevated temperatures, as the oxygen vacancies facilitate oxide ion conductivity by a vacancy hopping mechanism. In its pure form, CeO2 is an n-type wide bandgap semiconductor. In its non-stoichiometric form, CeO2-x can be either a mixed valence localized Ce(III,IV) hopping semiconductor or a delocalized Ce(IV-x) metal. The former exists in the low temperature regime and transitions to the latter in the high temperature regime. This shift in electronic structure explains the increase in electronic conductivity observed above 1,000°C as it passes through a semiconductor-metal transition. This property would certainly be an asset for a redox active, solar thermochemical membrane reactor. With respect to the chemistry occurring on the surface and within the volume of the ceria membrane, when CO2 collides with the surface of the CeO2-x membrane at 1,500°C, presumably it adsorbs at an [O] and the C=O bond of CO2 cleaves. Thereby, the [O] vacancy is filled with an O atom, concurrent with desorption of the so-formed CO. In this context, atomistic modeling has shown how, in non-stoichiometric CeO2-x both oxygen vacancies and reduced Ce(III) centers accumulate at the surfaces, where they can promote catalytic activity.4Cheng Z. Sherman B.J. Lo C.S. J. Chem. Phys. 2013; 138: 014702Crossref PubMed Scopus (138) Google Scholar Notably, density functional theory calculations reveal how the reduced Ce(III) centers are not located at the nearest neighbor site to the oxygen vacancy, as was previously assumed, but are spatially separated.5Ganduglia-Pirovano M.V. Da Silva J.L.F. Sauer J. Phys. Rev. Lett. 2009; 102: 026101Crossref PubMed Scopus (462) Google Scholar Pertinently, adsorption and splitting of CO2 was favored at surface oxygen vacancies on non-stoichiometric CeO2-x compared to stoichiometric CeO2.4Cheng Z. Sherman B.J. Lo C.S. J. Chem. Phys. 2013; 138: 014702Crossref PubMed Scopus (138) Google Scholar At this point, the O atom needs two electrons to become lattice O2− and begin to diffuse through the membrane. The driving force for diffusion arises from the chemical potential gradient established by the difference in O2 partial pressure established on opposite sides of the membrane. However, where does the O atom get these two electrons? It turns out that at 1,500°C ceria is uniquely ambipolar. This unusual property allows electrons and oxide ions to diffuse in opposite directions across the membrane, the details described by the Zener double transfer mechanism, illustrated in Figure 2. This two-way transfer is similar to the Cabrera-Mott process for the oxidation of pure transition metals and depends on the mobility of electrons in the oxide as well as the energy level (chemical potential) alignments for O2 and electrons across the membrane.7Cabrera N. Mott N.F. Rep. Prog. Phys. 1948; 12: 163http://iopscience.iop.org/article/10.1088/0034-4885/12/1/308/metaCrossref Scopus (2355) Google Scholar To amplify this point, achieving high CO and O2 production rates from the splitting of CO2 requires positioning of the Fermi level between the redox potentials for CO2 reduction and O2 oxidation as well as high ambipolar conductivity of electrons and oxygen ions, involving counter diffusion through the body of the ceria membrane. Indeed, the ambipolar diffusion coefficients of ceria at T = 1,500°C, determined by the Steinfeld group,8Ackermann S. Scheffe J.R. Steinfeld A. J. Phys. Chem. C. 2014; 118: 5216-5225Crossref Scopus (104) Google Scholar revealed that electronic and ionic conduction across a 1 mm thickness occurs in seconds. That explains their selection of the membrane material. As seen in Figure 2, solar thermal energy enables the reaction CO2 + 2e− → O2− + CO. This occurs on the high oxygen partial pressure side of the membrane. The O2− diffuses through to the low oxygen partial pressure side of the membrane enabling the reaction O2− → O2 + 2e−. This process separates the O2 from the CO as the electrons diffuse to the other side of the membrane to complete the CO2 splitting reaction. One can envision a similar process occurring for the H2O splitting reaction to separate streams of H2 and O2. Another question that arises is that of the heating mechanism of the ceria membrane. Inspection of Figure 1 shows concentrated solar radiation at a flux intensity equivalent to 3,500 suns entering a thermally insulated cavity-receiver through a small aperture. The cavity, engineered for efficient radiative capture by multiple internal reflections, makes it a Planck’s blackbody absorber and emitter. Conventional wisdom says the cavity behaves as a “brute heat force,” whereby blackbody radiation emitted by the cavity heats the Al2O3 concentric tube. This in turn re-radiates toward the surface of the inner ceria membrane tube, allowing it to reach thermal equilibrium at 1,500°C with further conductive heat transfer across the membrane facilitating oxide ion and electron counter diffusion, as shown in Figure 2. However, I am also curious whether one should consider here the possibility that electronic transitions are being excited in the CeO2-x by the high-flux solar radiation. Could non-radiative relaxation processes contribute a photo-thermal effect, whereby some portion of the solar heating process originates from conduction electron, inter- or intra-or sub-bandgap electronic non-radiative relaxation? Furthermore, what wavelength spectral ranges would be responsible for these processes? In this respect, CeO2 is an indirect bandgap material, meaning that electrons, which are vertically excited from the valence band edge, deposit at a point in the Brillouin zone away from the conduction band minimum. As the excited electrons relax in the conduction band toward the minimum, they can release the excess energy as heat in the crystal lattice. The very low quantum yield for photoluminescence is consistent with this proposal. Investigations of the photochemistry and photophysics of CeO2-x should help resolve these issues. While the conversion rate and energy efficiency are currently not so high, what Aldo Steinfeld and coworkers have demonstrated is an important first step into a completely new world of making solar fuels. Perhaps the way to improve the performance metrics is to optimize the transport properties of the gaseous species through the construction of a solar membrane reactor containing a network of millimeter-size and micrometer-size channels. The larger channels would enable penetration of the whole spectrum of blackbody thermal radiation and facilitate efficient heat transfer, while the smaller channels would serve to increase the membrane surface area and its chemical reactivity for efficient mass transfer. Such a hierarchical structure should lead into higher mass conversion and energy efficiency. Presumably, at the high operating temperatures of solar fuels production, amorphous electrically resistive phases between the ceria grains do not deleteriously affect ionic and electronic charge transport in the membrane. The next logical step in the enhancement and enrichment of this research is a “one-pot” synthesis of syngas by simply co-feeding H2O and CO2 that couples the H2O and CO2 splitting reactions in a single solar membrane reactor. It will be interesting to see how these two molecules compete for the membrane surface active sites and whether the resulting H2/CO ratios are controllable. This exciting direction could lead to “one-pot” production of solar fuels and engineering to industrially significant scales! On a closing note, it is worth mentioning that ceria is currently widely used in oxygen gas sensors, high-temperature solid oxide fuel cells, and electrolysers, enabled by its high oxide ion conductivity while remaining an electrical insulator. In these devices, oxide ions diffuse through the ceria electrolyte and charge-balancing electrons move through the external circuit linking the contacting electrodes. This is true at low temperatures beginning around 500°C; however, as we go to higher temperatures, ceria becomes electrically conductive and shorts the circuit between surrounding electrodes. This situation is not conducive for the aforementioned applications, where ceria functions as a solid electrolyte. Notably, by the time, ceria reaches 1,500°C it is a very good electronic and ionic conductor, thereby providing exactly the properties needed for the successful operation of the solar-driven ambipolar membrane reactor, which is electrode-less and functions by allowing oxide ions and electrons to counter-diffuse across the membrane. While other ceramic materials have good ionic and electronic conductivity at much lower temperatures, the problem is that some of them form carbonates at high temperatures in the presence of CO2 and are unable to convert CO2 into CO.9Gálvez M.E. Jacot R. Scheffe J. Cooper T. Patzke G. Steinfeld A. Phys. Chem. Chem. Phys. 2015; 17: 6629-6634Crossref PubMed Google Scholar Another problem is that the material must be stable at very different oxygen partial pressures on both sides of the membrane in reducing versus oxidizing conditions. Clearly, ceria is currently the material of choice for the solar-driven ambipolar membrane reactor, as so elegantly demonstrated by Aldo Steinfeld and coworkers at ETH Zurich in their paper for the inaugural issue of Joule. G.A.O. is a Government of Canada Research Chair in Materials Chemistry and Nanochemistry. Financial support for this work has been provided by the following funding agencies: Ontario Ministry of Research Innovation (MRI); Ministry of Economic Development, Employment and Infrastructure (MEDI); Ministry of the Environment and Climate Change; Connaught Innovation Fund; Connaught Global Challenge Fund; and the Natural Sciences and Engineering Research Council of Canada (NSERC). Solar-Driven Thermochemical Splitting of CO2 and In Situ Separation of CO and O2 across a Ceria Redox Membrane ReactorTou et al.JouleAugust 9, 2017In BriefConversion of CO2 into fuels via a solar-driven thermochemical process enables the efficient production of sustainable transportation fuels. The solar reactor continuously splits CO2 across a redox membrane into separate streams of CO and O2 using concentrated solar radiation. The stable experimental results and the theoretical solar-to-fuel energy conversion efficiency support the viability of the modular solar reactor technology for converting CO2 to fuels at a large scale by integration with the established concentrating solar tower and dish systems. Full-Text PDF Open Access