The timing mismatch between the supply and demand for solar electricity will require energy storage,(1) for which conversion to a storable chemical fuel is one option. Beyond that, sectors such as marine navigation, aviation, etc. may still need liquid fuels, exported from the solar facility. These might also serve as one input into a future renewable chemicals industry. Dihydrogen is hard to store;(2) therefore, converting CO₂ into liquids such as methanol(3) or ethanol is a leading option. In “artificial photosynthesis”(4) one or more absorber “dyes” create a charge separation in a photoelectrochemical cell (PEC). The resulting holes drive anodic water oxidation to produce O₂ and protons (eq 1). The electrons drive a catalyst for CO₂ reduction leading to a fuel such as CO or MeOH, using protons that cross the cell from the anode. For simplicity, eq 2 shows the formation of CO where the overall process takes up 2e^– + 2H⁺; forming MeOH needs 6e^– + 6H⁺ (E₀ = −0.38 V).(1)(2) One important approach involves simple electrochemical (EC) reduction that could be driven by conventional renewable power.(5) Energy sources, no longer limited to solar, would now include the most available green electric power source, such as wind, hydroelectric, etc. We could even stretch to nuclear power, if this can be given pale green credentials. We could then build much bigger EC facilities adjacent to reliable CO₂ sources such as cement and steel plants(6) to gain economies of scale. Some of the fuel, say MeOH, may be retained within the EC facility for reconversion into power when demand is high. Older fuel cells gave poorer performance with MeOH than with H₂, but striking advances have recently been made in improving the direct methanol fuel cell.(7) Direct air capture using OH^– as base to give HCO₃^–,(8,9) though costly, is beginning to be considered more seriously as a CO₂ source, in which case carbon would be “borrowed” from the air, to be released on fuel oxidation. CO₂ capture from cement works etc. typically involves a liquid absorbent such as aqueous monoethanolamine (MEA).(10) However, much of the energy input in overall CO₂ capture–release goes into the thermal regeneration of the amine to release CO₂.(11) Maybe we can skip regeneration altogether and use the MEA/CO₂ adduct or HCO₃^– directly as the reactant for the electrochemical step, either in an EC or PEC system, hence the importance of recent reports on the EC reduction of MEA/CO₂ and of HCO₃^– to CO.(12) For example, on a Sn electrode at −0.8 V vs RHE, the product mixture from MEA/CO₂ consisted of H₂ (Faradaic efficiency: 60.8%), CO (4.1%), and formate (35.7%, E₀ = −0.61 V for CO₂/HCOOH), the H₂ coming from proton reduction. A Pd metal catalyst, again at −0.8 V, gives synthesis gas, H₂ + CO, a key industrial intermediate: H₂ (85.8%) and CO (12.4%) with minimal formate (1.3%). Likewise, several recent reports describe direct electrochemical reduction of bicarbonate, for example to CO.(13) The speciation of the typical 15–30% aqueous MEA with CO₂ has been followed by NMR(14) and IR(15) methods. At a MEA:CO₂ ratio of 2:1, the carbamate salt of eq 3 predominates, with bicarbonate and carbonate also present. The substrate for the reduction being itself an electrolyte, there is no need for any external electrolyte nor for the usual gas diffusion electrode needed to overcome mass transport limitations(16) associated with the poor solubility of CO₂. Schematically, MEA or OH^– could pick up CO₂ from the plant effluent gas flow in a capture module, then flow into the electrochemical module for reduction, then to an evaporator module which would separate any volatile liquid product before returning to the CO₂ capture module to continue the cycle (Figure 1). Bicarbonate has the advantage here because the amine could cross to the anode and be oxidized and might also give interfering reactions.(17) To avoid this, the amine could be a liquid siloxane oligomer, retained in the cathode compartment by a proton-permeable membrane, but this goes beyond the scope of the present discussion.(3) Figure 1. Possible cyclic process involving CO₂ capture from a source such as a cement works, followed by electrochemical reduction of MEA/CO₂ or HCO₃^–. Gaseous products separate, followed, if necessary, by passage through an evaporator to recover the volatile liquid products. The fact that aqueous CO₂ in MEA is largely present as [HOCH₂CH₂NHCO₂]⁻, along with some [HCO₃]⁻, means that either of these could be the immediate substrates undergoing reduction or else that CO₂ in equilibrium with these ions could play that role. Indeed, Dunwell et al. propose that in the reduction of aqueous CO₂ to CO, HCO₃^– ion enhances CO production on Au electrodes by increasing the effective CO₂ concentration through rapid equilibration of HCO₃^– with CO₂(aq).(18) Whether [HOCH₂CH₂NHCO₂]⁻ or HCO₃^– is the direct substrate or not, this provides a good reason to systematically include both MEA/CO₂ and HCO₃^– as substrates in EC and PEC solar fuels work. Having the reactant in liquid form means that any gaseous reduction products would be expected to self-separate more cleanly from the product mixture and not be diluted in excess gaseous reactant. If exportable liquid fuel is targeted, the production of H₂ + CO, synthesis gas,(12) could be useful because it can be readily converted either to MeOH or to hydrocarbons.(19) In contrast, if any EC or PEC process directly produced MeOH, it might well be more costly to separate it from the electrolyte and solvent. Separation processes are said to account for ∼45% of all process energy used in the chemical and petroleum refining industries,(20) but EC or PEC fuel research planning has not often taken this factor into account. As liquids, both of the common targets, MeOH and EtOH, may well pose problems in this respect. Separation problems can mar the viability of an otherwise highly commercially attractive transformation. In the Catalytica methane-to-methanol process,(21) the cost of product separation proved “prohibitively high”, so if ultimate application is a goal, separation deserves greater attention. EC or PEC production of H₂ or synthesis gas is convenient for product separation but problematic for storage while production of MeOH makes separation somewhat harder. One plausible compromise candidate is methyl formate (bp = 31 °C), which should self-separate under only slightly elevated temperature, yet be easily liquefied for storage. Its heat of vaporization is also low (29 versus 37.6 kJ/mol for MeOH)(22) and it is a good fuel for internal combustion engines (octane number 115) and a useful fuel additive.(23) In one rare case, Lucas et al. report the electroreduction of CO₂ over a Cu electrocatalyst to MeOCHO as the main product.(24) MeOCHO was also the principal product in a PEC reduction of CO₂ with lignin-stabilized Cu₂O under a Xe–Hg lamp (300–580 nm emission).(25) Neat methyl formate also undergoes thermocatalytic hydrogenolysis to MeOH (eq 4) by heterogeneous(26,27) or homogeneous(28,29) catalysts. For example, Iwasa et al. find that this occurs at 423 K over Pd/ZnO with 94% selectivity for MeOH, the balance being CO; Cu/ZnO is also effective but with lower selectivity (77%).(27) Alternatively, at 80 °C under 10 atmH₂, a ruthenium complex hydrogenolyzes neat MeOCHO to MeOH in 98% yield at 99% conversion, where solvent-free operation again simplifies product isolation.(28) With neat MeOCHO as reactant, product separations can be avoided if the selectivity and conversion are high enough or the tolerance for MeOCHO impurity in the product MeOH is sufficiently relaxed. Because H₂ is hard to avoid as a coproduct, usually undesired, of CO₂ reduction in aqueous media, the H₂ needed for hydrogenolysis should be available.(4)However, conversion of methyl formate to methanol may not even be required. Because MeOCHO is recognized as an intermediate in the operation of MeOH fuel cells(30) it is likely be a viable fuel cell substrate on its own, although methyl formate itself is not electrochemically active and initial hydrolysis to formic acid and methanol precedes electrooxidation.(30) If MeOCHO proves able to stand in for methanol in fuel applications, MeOCHO might have a wider role in a future fuel and solar energy storage economy. Comprehensive metrics(31) and techno-economic analyses(32) of competing systems will of course be needed to assess viability. In conclusion, electrochemical reduction of CO₂ to fuels may have some advantages over photoelectrochemical processes in certain applications. Amine adducts with CO₂, very rarely studied in this context today, as well as bicarbonate, should be included in such work. Self-separation of gaseous products such as synthesis gas may minimize separation costs where an exportable liquid fuel is the goal because either MeOH or hydrocarbons can be formed from it by conventional routes. Finally, the more volatile methyl formate may have advantages as a CO₂ reduction product in being relatively easily separated from the EC or PEC product mixture, yet be relatively easy to store as a liquid and use as a fuel. The author declares no competing financial interest. Views expressed in this Viewpoint are those of the author and not necessarily the views of the ACS. I thank the U.S. Department of Energy, Chemical Sciences, Geosciences, and Biosciences Division, Office of Basic Energy Sciences, Office of Science (DE-FG02-07ER15909) for support of our work in the area as well as Joe Hupp, Gary Brudvig, and Victor Batista for helpful discussions. This article references 32 other publications.

中文翻译：

---

二氧化碳太阳能燃料的替代策略

太阳能供需之间的时间不匹配将需要能量存储，（1）转换为可存储化学燃料是一种选择。除此之外，海上航行，航空等行业可能仍需要从太阳能设施出口的液体燃料。这些也可以作为未来可再生化学工业的一种投入。二氢很难储存；（2）因此，将CO ₂转化为液体，例如甲醇（3）或乙醇是最主要的选择。在“人工光合作用”（4）中，一种或多种吸收剂“染料”会在光电化学电池（PEC）中产生电荷分离。产生的空穴驱动阳极水氧化，产生O ₂和质子（eq 1）。电子驱动CO ₂的催化剂使用从阳极穿过电池的质子进行还原生成燃料，例如CO或MeOH。为简单起见，等式2显示了CO的形成，整个过程占2 e ^– + 2H ⁺；形成甲醇需要6 e ^– + 6H ⁺（E ₀ = -0.38 V）。（1）（2）一种重要的方法涉及可以通过常规可再生能源推动的简单电化学（EC）降低。（5）不再限于太阳能的能源现在将包括最可用的绿色电源，例如风能，水力发电等。如果可以给予淡绿色证书，我们甚至可以扩展到核电。然后，我们可以在水泥和钢铁厂等可靠的CO ₂来源附近建立更大的EC设施（6），以实现规模经济。当需求很高时，某些燃料（例如MeOH）可能会保留在EC设施内，以重新转化为动力。与H ₂相比，使用MeOH的旧燃料电池性能较差的，但显着的进展最近已经在改进直接甲醇燃料电池（7）构成的直接空气捕获使用OH。^-作为碱，得到HCO ₃ ^- ，（8,9），虽然昂贵，正开始被更认真考虑作为CO ₂源（在这种情况下，碳将从空气中“借入”）并通过燃料氧化释放。从水泥厂等处捕获的CO ₂通常涉及液体吸收剂，例如单乙醇胺水溶液（MEA）。（10）但是，在整个CO ₂捕获释放中输入的大部分能量都进入胺的热再生中以释放CO ₂（11）也许我们可以完全跳过再生，而使用MEA / CO ₂加合物或HCO₃ ^-直接作为反应物用于电化学步骤中，无论是在EC或PEC系统中，对EC减少的最近的报道因此重要性MEA / CO ₂和HCO的₃ ^- 。到CO（12）例如，在一个Sn电极在相对于RHE为-0.8 V的条件下，MEA / CO ₂的产物混合物由H ₂（法拉第效率：60.8％），CO（4.1％）和甲酸酯（35.7％，E ₀ = -0.61 V，CO _2）组成/ HCOOH），H ₂来自质子还原。钯金属催化剂的电压再次为-0.8 V，可生成合成气H ₂ + CO，这是一种关键的工业中间体：H ₂（85.8％）和一氧化碳（12.4％）和最低甲酸（1.3％）。同样，最近的一些报道描述了直接电化学还原碳酸氢盐，例如还原成CO。（13）NMR（14）和IR（15）方法跟踪了典型的15-30％的MEA与CO _2的水溶液形态。在MEA：CO _2之比为2：1时，等式3的氨基甲酸酯盐占主导地位，同时也存在碳酸氢盐和碳酸盐。用于还原的基质本身就是电解质，不需要任何外部电解质，也不需要常规的气体扩散电极即可克服与CO ₂溶解度差相关的传质限制（16）。从原理上讲，MEA或OH ^–可能吸收CO ₂气体从捕集模块中的工厂废水流中流出，然后流入电化学模块中进行还原，然后流向蒸发器模块，该蒸发器模块会分离出任何挥发性液体产物，然后再返回CO ₂捕集模块以继续循环（图1）。碳酸氢盐在此具有优势，因为胺可能会穿过阳极并被氧化并可能产生干扰反应。（17）为避免这种情况，胺可以是液态的硅氧烷低聚物，通过质子可渗透膜保留在阴极室中，但这超出了本讨论的范围。（3）图1.可能的循环过程，涉及从水泥厂等来源捕获CO ₂，然后电化学还原MEA / CO ₂或HCO ₃ ^–。分离出气态产物，然后，如果需要，通过蒸发器以回收挥发性液体产物。MEA中的CO ₂水溶液主要以[HOCH ₂ CH ₂ NHCO ₂ ] ^-的形式与某些[HCO ₃ ] ^-一起存在的事实，意味着这两种物质中的任何一种都可能是还原的直接底物，或者处于平衡状态的CO ₂这些离子可以起到这种作用。确实，Dunwell等人。建议将CO ₂水溶液还原为CO，HCO ₃ ^–离子增强对Au电极CO产量通过增加有效CO ₂通过HCO的快速平衡浓度₃ ^-与CO ₂（水溶液）（18）是否[HOCH ₂ CH ₂ NHCO ₂ ] ^-或HCO ₃ ^-是直接底物或不，这提供了一个系统地同时包含MEA / CO ₂和HCO ₃的充分理由^–作为EC和PEC太阳能燃料的基材。具有液体形式的反应物意味着将期望任何气态还原产物从产物混合物中更干净地自分离，并且不被过量的气态反应物稀释。如果将可出口液体燃料作为目标，则可生产H ₂+ CO，合成气（12）可能有用，因为它可以很容易地转化为MeOH或烃类。（19）相反，如果任何EC或PEC工艺直接生产MeOH，则将其分离的成本可能更高。来自电解质和溶剂。据说分离过程约占化学和石油精炼行业所用全部过程能量的约45％（20），但是EC或PEC燃料研究计划通常没有考虑这一因素。作为液体，两种常见目标甲醇和乙醇在这方面都可能造成问题。分离问题可能会损害原本具有高度商业吸引力的转型的可行性。在Catalytica甲烷制甲醇工艺中，（21）产品分离的成本被证明“过高”，因此，如果最终应用为目标，则分离应引起更多关注。₂或合成气便于分离产物，但在储存方面存在问题，而生产MeOH会使分离更加困难。一种可能的折衷候选物是甲酸甲酯（bp = 31°C），该甲酸甲酯应在仅略微升高的温度下自分离，但易于液化以储存。它的汽化热也很低（29相对于MeOH为37.6 kJ / mol）（22），是内燃机的良好燃料（辛烷值115）和有用的燃料添加剂。（23）在极少数情况下，卢卡斯（Lucas）等。报道了在铜电催化剂上将CO ₂电还原为MeOCHO作为主要产物。（24）MeOCHO也是用木质素稳定化的Cu ₂进行PEC还原CO ₂的主要产物。O在Xe-Hg灯下（300-580 nm发射）。（25）整齐的甲酸甲酯还通过多相（26,27）或均相（28,29）催化剂进行热催化氢解为MeOH（eq 4）。例如，Iwasa等。发现这发生在Pd / ZnO上于423 K处，对MeOH的选择性为94％，其余为CO。Cu / ZnO也是有效的，但选择性较低（77％）。（27）或者，在80°C下10 atmH _2下，钌络合物将纯的MeOCHO水解为MeOH，收率98％的甲醇转化率为99％，无溶剂操作再次简化了产物的分离。（28）使用纯净的MeOCHO作为反应物，如果选择性和转化率足够高或甲醇中的MeOCHO杂质耐受度足够宽松，则可以避免产物分离。因为H ₂由于水介质中CO _2的减少通常是不希望有的副产物，因此很难避免，氢解所需的H ₂应该是可用的。（4）但是，甚至不需要甲酸甲酯转化为甲醇。由于MeOCHO被认为是MeOH燃料电池运行中的中间体（30），因此尽管甲酸甲酯本身不具有电化学活性，并且在电氧化之前先水解为甲酸和甲醇，但它本身可能是可行的燃料电池基材。（ 30）如果MeOCHO被证明可以在燃料应用中代替甲醇，那么MeOCHO可能在未来的燃料和太阳能存储经济中发挥更大的作用。当然，需要使用竞争系统的综合指标（31）和技术经济分析（32）来评估生存能力。总之，在某些应用中，将CO ₂电化学还原为燃料可能比光电化学过程具有一些优势。与CO _2形成的胺加合物，今天在这种情况下很少进行研究的化合物，以及碳酸氢盐，都应包括在此类工作中。气态产物（例如合成气）的自分离可以最大程度地降低分离成本，在这种情况下，可出口的液体燃料是目标，因为可以通过常规路线从中形成MeOH或烃。最后，挥发性更强的甲酸甲酯可能具有作为CO _2的优势还原产物相对容易地从EC或PEC产品混合物中分离出来，但相对容易以液体形式存储和用作燃料。作者声明没有竞争性的经济利益。本观点中表达的观点是作者的观点，不一定是ACS的观点。我感谢美国能源，化学科学，地球科学和生物科学部，基础能源科学办公室，科学办公室（DE-FG02-07ER15909）对我们在该领域的工作提供的支持，以及Joe Hupp，Gary Brudvig，和Victor Batista进行有益的讨论。本文引用了其他32个出版物。