WEBINAR Q&A summary

We got many excellent questions and comments during our live Webinars. Few were answered live, the others were addressed on-line. Below is the summary of the on-line Q&A. Thank You!

August 5, 2020

Renewable Hydrogen Generation - Basic Science and New Technologies

Q1 (Rody Stephenson): What is your most successful complete device developed so far?

A1: Depends how you define successful... efficiency? Cost per kg H2? Also depends on the class of devices considered....

A1: Practically in the next few years, large-scale solar arrays that coupled with large-scale electrolyzers are a clear technology path. Longer term, integrated devices are promising.

A1: Recent example for CO2 reduction: ACS Energy Lett., DOI: 10.1021/acsenergylett.9b02576 (2020) and references therein.

Q2 (Akhilender Singh): What are your thoughts on development of ternary and quaternary oxides for PEC applications? What are the major challenges?

A2: There is a huge variety of compositional space to explore. Another challenge is that the intrinsic electron and hole mobilities are usually quite low. This is often a feature of the d-states that have poor electronic overlap. How do we get around this?

A2: JCAP has compiled a large database called MEAD (>1 million expts.) on the electrochemical and photoelectrochemical on 100,000’s of compositions of ternary and quarternary metal oxides.  This is a potential resource for future researchers to analyze as they make hypotheses about transport and catalytic properties of oxides.

Q3 (Akanksha Menon): If 4% of hydrogen is currently produced via electrolysis of water, and sufficient penetration of renewables that will enable low electricity prices, this well-established method could replace steam reforming of methane. What role does artificial photosynthesis play in hydrogen production then? Or should the focus be on producing other fuels?

A3: As the panel discussion touched upon, integrated photoelectrochemical systems may ultimately have lower costs than indirect systems (PV + electrolysis) when the solar capacity factor and system-level costs are considered.  There has been a study looking at what is the return on investment for a large-scale (1 GW) PEC system (R. Sathre et.al. Energy and Env. Sci.).  The analysis indicates that the biggest levers for cost reduction are PEC solar to hydrogen efficiency and materials durability.  Balance of systems (compression, piping, storage) did not dominate at the system level.  This good news for science, as advances in PEC efficiency and durability do matter at the system level.

A3: We need electrolysis - scale it, make it cheaper, work out the technology and durability issues. Longer-term artificial photosynthetic concepts may be useful in scaling and lower costs further.

A3: Also, potentially artificial photosynthesis approaches might become important for smaller, decentralized rural systems. For example, steam reforming cannot be easily downscaled.

Q4 (Mohd Khan): if we have cheap electricity available, then we make green hydrogen at a competitive price versus SMR H2.  Where does that leave the whole field of integrated PEC devices and photo catalysis? Also what makes integrated PEC devices cheaper? They still have same light absorbers and electrocatalysts as PV-electrolysis setups.

A4: In order to make it competitive PEC systems must use different less expensive materials and balance of systems. That of course is a challenge. PEC systems have one major advantage - small crystals can work very well as reactions can be derived locally.

Q5 (Dharmesh Hansora): What is perception and scope of PV grade material and metal oxide material for fabrication view development of Large integrated photoelectrodes for Practical solar h2 generation?

A5: I view it as healthy that there are proponents of both which are planning and executing scale up projects. For metal oxides, the recent ca. m^2 demonstration from Domen and co-workers (U. Tokyo) at a few % solar to H2 is interesting and significant.

Q6 (Harun Tueysuez): With the current technology, the hydrogen production via steam reforming has production cost of below 1$/kg while green hydrogen through water electrolysis costs around 6-7 $/kg. I am curious how this gab can be closed? Would be sufficient to discover cheaper earth-abundant HER and OER catalysts?

A6: First we need low cost renewable electricity, probably $0.01/kWh - that is attainable with wind and solar in a decade. Then we need very low capital and operational expenses for electrolyzers. Electrolyzer production is currently at a very low scale - mainly by hand. This can improve in cost by a factor of 10, I suspect.

Q7 (Muhammad Sohail): Its provide opportunities to store PV energy at TW scale as compared to other storage technologies and for long term duration as well.

A7: Compared to other energy storage media (water pumping, compressed air, batteries), compressed hydrogen is quite efficient on a weight basis but not as great as liquid fuels on a volumetric basis.  Another factor to think about is the ‘round-trip’ efficiency of hydrogen generation (electrolyzer STH efficiency ~ 70%) and use to make electricity (fuel cell FTE efficiency ~ 70%; so round-trip efficiency is about 50%).

Q8 (Sebastian Bold): Do you think that - and if so, when - integrated photoelectrochemical systems for water splitting can overtake non-integrated PV+electrolysis systems in terms of efficiency and cost?

A8: At this time is not clear when integrated systems may take over from PV+EC systems. The former approach still has many technical obstacles to overcome and their efficiency is not yet competitive.

Q9 (Salvador Eslava): Do you think PEC can outperform PV+EC?

A9: I do believe that, in theory, PEC can outperform PV+EC: 1) the efficiency in PEC can be greater given that the current density might be lower, 2) the efficiency can be great given that we have shorter path lengths for ions/electrons plus the potential for thermal integration, and 3) finally in PV+EC you always need an DC-DC converter (some sort of power electronics, potentially also DC-AC and back), which introduces additional losses (not needed for PEC).

Q10 (Oscar Orantes): What would be the ideal environment are for Hydrogen Fuel cell powered vehicles? Are industrial applications more sensible or is there a possibility for the infrastructure to be so integrated and efficient that we can see FCEV dominate a significant portion of the automotive market?

A10: For the near future, batteries will be power source of choice for automobiles, but for long-haul trucks, FCs may be preferred because on range vs. weight basis they win.

A10: Another big area is ocean shipping... you can't address that with batteries. Air travel can also be run on hydrogen, though probably direct combustion instead of fuel cells is likely for large jets.

Q11 (Teja Venkatesa Perumal): How can we take inspiration from biological systems to build better light harvesting structures for integrated PEC devices to receive concentrated sunlight?

A11: I personally take two inspirations from the oxygen evolving complex in natural photosynthesis. (1) As I said during panel discussion, its ability to regenerate is something that we do not know how to emulate, yet, with our inorganic catalysts. (2) Proteins around the Mn-based active site rearrange between each electron transfer step (4 required total to split 2 waters to O2) in such a way to reduce the overpotential. We do not yet have a way to implement this type of functionality in an artificial structure.

A11: Biology does self-repair. It is not particularly efficient. So let's learn how to repair defects that form, passivate interfaces, etc. in artificial systems.

Q12 (Hariprasad Narayanan): Why not much of study concentrated on the real-time light-matter interaction? Once the light irradiates, they will be a rearrangement of the density of states in the semiconductor. Do you think there requires a paradigm shift in designing the material for solar fuel production?

A12: Hariprasad, I suggest you attend the next two webinars. This subject may be discussed in the context of plasmonic electrocatalysis, where the real-time interactions between non-equilibrium carrier populations in the catalyst with the available densities of states for electron transfer are of vital importance.

Q13 (Lei Wang): What are the opportunities and strategies to conduct long-term stability testing for a water splitting device? From a catalyst level to a system level.

A13: I would argue that what you want to do is develop a scientific understanding of durability. For catalysts this means what controls the reaction barriers to dissolution and the rates of corrosion and passivation? How do you control these in the context of also enhancing activity? Technology will follow from this science understanding.

A13: Long-term catalyst testing must be done, and should be done by industry, as it tries to develop a technology. This is not a valid objective for a national lab.

A13: As a strategy: once you understand the degradation mechanisms, you might be able to develop accelerated testing procedures, significantly reducing effort and expense. Additionally, there is a significant opportunity for theory here

A13: The durability questions are not yet resolved for the component materials - we do indeed see loss of efficiency over time at the prototype device level, but tracing this to root causes in catalyst activity, catalytic material loss, photoelectrode corrosion and efficiency loss are very much open questions.

Q14 (Jose Mendoza): 1. How can we think on the tradeoff between a dynamic catalytic structure that might be more efficient and stability (e.g. number of cycles before the catalyst decays)?  2. In a similar way, how can we think of a system that is earth-abundant that might not last many cycles vs. precious metal-based catalysts that might last long? In other words, lifetime of catalyst and cost compounded quantity.   3. The catalytic cycle of OEC in PSII is still being investigated as Joel was mentioning. How could you foresee knowing the full catalytic mechanism might help with water splitting? This might be obvious in many directions, but maybe there is a more important direction in your own point of view (e.g. knowing how reconstructing the OEC is still favorable overall).

A14: 1. I believe there is no physical law that states that activity and durability must be inversely related. We can design some bonds that are strong at the active site and some that are weak. A key challenge is that we can't put atoms where we want them typically for heterogeneous catalysts and molecular chemistry is difficult outside of single metal centers that are not particularly active for the reactions we care about usually.  2. This is an economic question really. My feeling is that the more durable system will always win out in cost. Labor and balance of system costs are just too high usually.  3. The OEC manages protons very effectively in neutral media. How this relates to high conductivity acid and base solutions is in my mind still an open question.

A14: Jose, 3 very good points. Let's think of what makes a "good" catalyst. It makes a reaction that is possible (negative \Delta G) turn over faster than a less good catalyst. Considering OER, we create a negative \Delta G situation by controlling the chemical potential of the electrons at our electrode (we make it very positive). So a good catalyst does the reaction we want (OER), without doing a reaction we don't want (H2O2 production or corrosion) for as many turnovers as possible. The cost of generation increases greatly if the catalyst deactivates and cannot be regenerated. Ultimately, the investors who will fund the "green" H2 plants will not be interested in how the catalysts work, but they will want to be as sure as possible that they do work.

Q15 (SK Yasir Hosen): Do you think hydrogen-based renewable energy will have greater impact than other renewable energy research such as biofuel in the coming decade? Why or Why not?

A15: Biofuels are not particularly promising - the efficiency of biological solar energy conversion is just too low and the land requirements are too high. Hydrogen is far superior and one creates a closed loop cycle without carbon!

Q16 (Joseph DuChene): Long-term stability (~10-20 yr.) of PEC devices is still a challenge for the field - what are your thoughts about how to benchmark material stability more quickly? For instance, is accelerated testing a reasonable prediction of real-world device stability? Or do we have to test a device under simulated conditions for a decade to have confidence in its stability?

A16: Joe, as with the development of halide perovskite solar cells we need to develop accelerated wear protocols now to better predict if PEC systems can last the required 10 years.

A16: I would also suggest that we need a ‘predictive science of durability’, i.e., one where rate laws and scaling relationships can be generated over multiple orders of magnitude in time.  For catalyst efficiency and selectivity, we have framed this search in terms of ‘descriptors’ — we need descriptors for materials durability.  Another frontier is developing active repair and regenerative schemes for inorganic electrochemical materials and interfaces.

A16: To develop accelerated testing, we have to understand the degradation mechanisms, so there is quite a bit of science to be discovered there. Additionally, I would argue that once we understand degradation of materials, we have to understand if in devices (where the local environment in terms of heat, mass and charge transport might be different, and locally even vary) the same degradation is observed / limiting.

Q17 (Matthew Mayer): Regarding obtaining detailed information of catalysts using in situ methods of microscopy and spectroscopy (e.g. XAS and TEM as mentioned by Prof. Bell), are there some key examples of where such information has actually fed back into the material design to produce predicted improvements based on the improved structural understanding?

A17: The best example is likely in the fuel cell catalysts for ORR. The understanding of the core shell structures and strains that develop and change the d-band center have led to new materials with much lower Pt utilization, high activity, good durability, and a path to commercialization.

Q18 (Victor Trinh): You mentioned the use of first principle studies (DFT) in facilitating the design of new materials. But conventional first principle calculations themself also have their own issues to reach the high accuracy of reflecting those systems and we might need to go beyond DFT and it is computational less effective then. How could we compromise between the effectiveness versus accuracy of computational calculations for those materials investigation?

A18: Excellent question. My group together with Martin Head-Gordon at UCB is working on this issue. The trade offs are accuracy vs. computational cost. At the moment there is not universal answer.

Q19 (Ricardo Silva Buarque): Regarding the photocatalytic route, with several nanoscale materials being studied in recent times (i.e.: monolayers), how do you balance the need for enough thickness for light absorption and the material's properties dependent on the thickness (ex: direct bandgap in monolayer transition-metal dichalcogenides)?

A19: The basic concept is you want to match your materials length scale to the electronic and ionic transport in the system. Nano and microscale materials allow for shorter collection lengths for excited carriers (good) and allow for integration of ion transport paths (good) but you need more of them to absorb all the light (bad) and they have higher surface area that mediate electron and hole recombination (bad). There are many experiments and models emerging that compare all these tradeoffs.... but basically if you don't have essentially PV-grade materials with near unity quantum efficiency for carrier collection and high photovoltage the system is not technologically interesting.

Q20 (Amitava Sarkar): Comment on cost: Spending all of my professional years in corporate R&D, all I can say is that it’s important to put aside cost when we focus scientific innovation (otherwise no novel/disruptive technology can be developed). Cost of H2 (electrolysis or SMR) is production cost (Class 1 cost + cost reduction over years of operational best practice) and current cost estimates for PECs are not even Class 5 category. When there is order of magnitude difference between Class 1 and Class 5 cost category, we better not compare production cost to that of Class 5 cost estimate!

A20: Agreed, at this moment the only cost competitive approach for H2 generation is PV plus EC. Today this approach is about 2-3 x more costly than SMR, but the cost of the former technology is dropping. PV+EC will also be greatly favored as the cost of electricity drops to 3 cents/kwh and below.

A20: I certainly agree that it is very difficult to compare cost of a very mature technology and one that is still in its infancy (this is why I think it is important to do this also more in a probabilistic way, i.e. expected cost with a certain standard deviation), especially also as it sometimes can smother innovation. However, techno-economic assessments allow us to better understand if - under some conditions / assumptions - such a technology could ever compete. And in this sense also provide guidance on where - from a cost perspective - most attention/innovation is needed.

Q21 (Andre Argenton): This is a fantastic discussion and specifically to Shannon's point on "there is no point to scale up before we know the science works", I offer an additional point: I agree that scaling up has to happen at the right time, but I've been a proponent to bring the engineering of scaling up to the mindset of researchers. Thinking ahead of time of the potential limitations in design of reactors and systems and limitations in transport phenomena might provide guidance to where to do science.

A21: I completely agree with you. Science and engineering need to relatively soon go hand in hand and will inspire each other, therefore accelerate the path to eventual deployment of a technology

A21: Andre - thanks for making this important point; indeed device prototyping has been a part of the JCAP culture and while we have not built systems at a large scale, we have be continually pressed the future reality of scaling in our thinking about the science.

Q22 (Bradley Brennan): Can stable semiconductor absorbers be embedded into membranes and then coated on each side with their respective catalysts and ancillary components to bypass low solar conversion of the electricity generation PV process? Lots of corrosion and other issues possible, but would provide a direct solar option, roll-to-roll processing possibilities, and built-in separation of gases.

A22: Brad - great question, and one that touches on interfacial science.  We have thought about direct integration of semiconductor photoelectrodes into membranes and there have been a few attempts to do this to date.  Key questions center on how to enable the membrane material to ‘wet’  the semiconductor so as to prevent H2 and O2 cross-over, while still exposing the cathodic and anodic sides of the photoelectrode to the electrolyte.

Q23 (Ponart Aoonratsameruang): Metal oxides seem to exhibit low efficiency compared to Si or III-V semiconductors. Can silicon-based photoelectrodes commercialized?

A23: Metal oxides have lower efficiency because they don't have long lifetimes or large mobilities. How can we get around that? This is a big challenge. Si-based photoelectrodes are challenged by stability... luckily the Si costs are now very low so IF the balance of systems for making a photochemical system can be made much lower then these approaches could be commercialized successfully.

Q24 (Lindsay): To what extent have you explored the trade off between protecting the membrane from corrosion or poisoning versus blocking off the hydrogen access to the catalyst in a fuel cell?

A24: Lindsay, good question and you identify a science and technology gap. Most stability/corrosion studies have focused on the catalyst or photoelectrocatalysts, usually for the OER reaction, as this one is the most difficult in water splitting. There has been less attention on the contribution of membrane to operational stability. As with the catalysts, if there is not a way to regenerate or repair it in-situ, it must be able to last for a long time, several years according to our Life Cycle Assessments, to have a positive ROI, both from a monetary and environmental point of view.

A24: this has much to do with the details of what is known as the three-phase region ... ionomer that is in contact with catalyst is susceptible to degradation - even Nafion degrades at the ORR potentials slowly. Hydrogen transport can be engineered by tuning the porosity.

Q25 (Justin Bui): Do you think it is feasible to attempt to perform electrolysis on non-pure water feedstocks, such as seawater or freshwater, which will contain microbial contaminants and other co- and counter-ions, or do you think the stability and resulting inefficiencies of the device in these environments would be too much of an issue?

A25: Yes, there have been studies of doing water splitting using seawater. However, there are problems with electrode corrosion and microbial contamination. These are not easy issues to overcome.

A25: Durability in the presence of impurities of feed water is critical for long-term use of an electrolyzer. Even the "purest" water will contain impurities - from the pipes, air, etc. So that science is really important. Economically it will likely be better, expect in perhaps very specialized situations, to always pre purify water prior to electrolysis. It is much less energy to distill water than electrolyze it.

Q26 (Lindsay): As a follow up I am researching ways to improve performance and durability of alkaline fuel cells. Can you point me to any new studies or contacts to collaborate with?

A26: Will Mustain (South Carolina) has done some very nice work recently.

A26: You may want to look at a review article by J. Song et al., Chem Soc. Rev. 2020, 49, 2196.

August 6, 2020

Reduction of Carbon Dioxide to Renewable Fuels and Chemicals - Accomplishments and Challenges

Q1 (Rody Stephenson): Can you use CO2 at 400 ppm, or does it have to be concentrated by several orders of magnitude?

A1: At these low concentrations, the total achievable reaction rate (diffusion-limited rate) is relatively low. There is some modeling work by Nathan Lewis' group that you may be interested in that shows quantitatively what these rates might be. Likely, the CO2 will need to be further concentrated even for low current density applications such as photocatalysis. We can think of innovative strategies to concentrate CO2 in situ near the surface of the catalyst.

A1: I agree with Chris. Additionally, investigating CO2 reduction in gas streams containing contaminants similar to what would be found in the outlet of power plants or other point sources is another open challenge.

Q2 (Phil De Luna): Beyond technology development, what are some policy recommendations to increase investment, de-risk technologies, and support adoption of CO2 reduction? How can the community raise the profile of CO2 reduction with lawmakers and financiers?

A2: That's a very important question! While making policy recommendations is outside my area of expertise, I want to emphasize that this is really a global challenge. We need global cooperation to achieve real progress in these areas. As citizens we should support candidates who recognize the scale and challenges of adapting of more sustainable infrastructure.

Q3 (Jose Mendoza): On the realm of selectivity and structure control for CO2RR, what is the primary structural model or motif has most information/studies as cubanes is to OER? Maybe some examples contain Copper atoms (heterogenous) or motif related to the catalytic site in proteins (homogeneous), however, in your point of view what could be the structural model that can teach us the most?

A3: Structure-activity relationships have been established primarily for elemental transition metals (e.g., Ag, Au, Cu, Sn, etc.) and heteroatom and transition metal doped carbons. For metals, step sites seem to be quite important for key reaction steps such as CO2 adsorption and CO-CO dimerization. Transition metal-doped carbons seem to share similar active sites as porphyrin-based molecular catalysts, showing that there are correlations between heterogeneous and molecular catalysis.

A3: Jose, this is a question I have been thinking quite a bit about. Looking at various computational studies that provide mechanistic information with respect to C-C bond formation on Cu, one needs to include multiple Cu atoms in the models. As with all small molecule modeling chemistry, making a structure that looks like the active site may not be sufficient to get the desired reactivity. It is nevertheless an interesting target, and might provide mechanistic insight. We are exploring related questions in our research.

Q4 (Trung): Copper is a well-know catalyst in producing higher C2 products. Is there any alternative that can drive CO2 reduction to higher C2 products? There is a lot of work on molecular metal complex as electrocatalyst, recently. Do you think that dinuclear metal complex or dual-atom catalyst will be promising candidate?

A4: There was some recent work from Charles Dismuke's group that showed transition metal phosphides could convert CO2 to C2+ (up to C5) at relatively low overpotentials and current densities. The presence of very different reaction intermediates and products indicates that the reaction mechanism on these catalysts is different, and is worth further exploring. There has been significant work on molecular complexes as electrocatalysts, and recently there have been demonstrations that some mononuclear complexes can produce methanol. Dinuclear complexes are interesting for promoting associative (and dissociative) mechanisms. They could be exciting in the context of CO2 reduction.

Q5 (Jose Mendoza): If we choose a catalyst for the CO2RR part, then we have to pick the source for protons (perhaps from water). In a multicomponent system, how the source for protons might influence overall performance?
In yesterday's discussion, it was discussed that the pH might influence the proton diffusion coefficients and thus the performance, including selectivity. This is more or less clear for water splitting but I was wondering what are your thoughts on the carbon dioxide reduction reaction (CO2RR).

A5: Jose, this is a question I have been thinking quite a bit about. Looking at various computational studies that provide mechanistic information with respect to C-C bond formation on Cu, one needs to include multiple Cu atoms in the models. As with all small molecule modeling chemistry, making a structure that looks like the active site may not be sufficient to get the desired reactivity. It is nevertheless an interesting target, and might provide mechanistic insight. We are exploring related questions in our research.

Q6 (Teja Venkatesa Perumal): Since adsorption of CO2 molecule onto the catalyst surface also plays a crucial role to improve activity apart from surface reactions, what are some effective surface modifications that can be adopted for different catalysts to adsorb stable CO2 molecule and make it reactive?

A6: There are a few examples of how you can utilize surface modifiers and promoters, such as organic ligands, ammonium and pyridinium salts, to drive selectivity. It is interesting also to look at heterostructures of metal/metal or metal/metal oxide materials to promote sustained production of higher hydrocarbons.

August 7, 2020

Solar Fuels - Future Science and Technology

Q1 (William Leighty): "Systems engineering" and "end use":  Thank you for reminding us.  Shouldn't "energy services" always be our focus, since nobody wants to buy a liter of fuel, hg of Hydrogen, or kWh -- they want the services those energy carriers can deliver?  Are we in danger of losing this focus, in our rush to "green" and "smarten" the electricity grid, assuming it is the default energy system?

A1: Thank you for the question, and I really like the term "energy service". I agree that we should have this in mind from the outset as that will dictate what an engineered device will look like; without an end use in mind, and even a target audience in mind, we may end up designing something that is not going to be impactful or of interest to industry to scale.

Q2 (Muhammad Sohail): Any comments on technoeconomical viability of the technology, if developed, as compared to existing fossil fuels.

A2: A technoeconomic comparison using current models between "a solar fuels device today" and "existing fossil fuel technologies" would result in the latter being the dominant solution under almost every scenario. But, this is not quite a fair comparison - as Jason described - because there are applications where integrated systems matter, we don't know what end uses may exactly look like or how they will particularly be impactful until we build a >1 cm2 prototype PEC device, and importantly, without including a cost on so-called externalities (e.g. damage done to the planet by releasing CO2), we don't know what the true technoeconomic comparison is.

Q3 (Teja Venkatesa Perumal): Why can't we directly incorporate existing Atmospheric Water Generator (AWG) technology into PEC devices to capture atmospheric humidity and produce water for water splitting and CO2RR?

A3: That's a great idea Teja.  This is exactly the point we are trying to make about bringing together technologies from across fields to design and optimize a monolithic device.  There are many exciting technologies out there; the challenging will be bringing them together in an efficient and cost effective way.  I'm excited to see ideas like this become reality.

Q4 (Sonia Kishore Bendre): Could you discuss the current prospects of sustainable engineering on an international level? Where do you see the most potential?

A4: Hi Sonia, Undoubtedly, sustainable technology and engineering will need to be deployed in almost if not all countries and regions before 2050 in order for us to avoid the adverse impacts of climate change. This will involve a deliberate working together and integration of policy, economic, industrial and scientific machinations.

A4: Each country, region, or locality will need to assess their own specific energy and environmental needs and this is an assessment and discussion that must be had between residents, business/industries, and governing institutions.

A4: Overall though, with the current and near-future plethora of sustainable technologies that are coming to the fore (in the areas of wind, solar, CO2 fuels, solar fuels, hydrogen fuels, sustainable water), I don't doubt that there'll be a custom mix people can implement for their specific needs.

Q5 (Sonia Kishore Bendre): As someone based in Nigeria, I'm curious to know if any of you can speak about my first question with reference to Nigeria?

A5: Hi Sonia, being from Nigeria, this is a question I've wondered about myself. As a major oil producing/exporting nation, a key first step would be to apply sustainable technologies to get to carbon-neutral modes of energy generation. At the same time, rapid investments and policies should be made for the rapid advancement of renewable and carbon negative technologies. The time factor is very important: increased collaboration between countries with similar energy mixes or demands is crucial and action needs to be taken as soon as possible.

Q6 (Matthew Kuchta): I agree with Jason about how do we scale-up? It is my belief that the real answer will come from the existing super-commercial scale companies...e.g. Exxon, Dow, BASF, Du Pont...they know how to build huge production plants and the reality of it...a dialog between 'basic science' researchers and large industry should be ongoing...

A6: This is an excellent point and industry-academic partnerships are absolutely critical to realizing our long-term visions. There is cause for optimism here as the level of interest and engagement from industry in solar fuels basic science was quite small 10 years ago and has been rapidly expanding. A few companies have solar fuel prototypes of their own and recognize the opportunity to capitalize on our scientific advancements.

Q7 (Joseph DuChene): Shane, what current density for the PEC were you assuming for that calculation you mentioned needing 1mm thick of water on the active device area? This is very interesting to consider!

A7: ~20 mA/cm2, as far as I recall from a calculation I did a while ago.

Q8 (Sonia Kishore Bendre): What is being done to assist developing countries with huge potential for solar energy systems but lack of economic stability to set these systems up?

A8: While I wish I had the sociopolitical solutions to these critically important questions that have implications for the sustainable advancement of human civilization, the answer I can provide now is that a key aspect of solar fuels technology compared to existing chemical and fuels industry is "distributed" vs. "centralized". The ability of solar fuels technology to deliver fuel on-site without a pipeline to a refinery really broadens the possible deployment of the technology and changes the technoeconomic landscape. So as scientists we need to communicate our findings to facilitate international energy and chemical planning.