HomeLatest ThreadsGreatest ThreadsForums & GroupsMy SubscriptionsMy Posts
DU Home » Latest Threads » Forums & Groups » Topics » Science » Science (Group) » Revealing Mechanistic Pro...

Tue Sep 15, 2020, 10:57 PM

Revealing Mechanistic Processes in Gas-Diffusion Electrodes During CO2 Reduction to CO/Formate.

The paper I'll discuss in this post is this one: Revealing Mechanistic Processes in Gas-Diffusion Electrodes During CO2 Reduction via Impedance Spectroscopy (Bienen et al., ACS Sustainable Chem. Eng. 2020, 8, 36, 1375913768)

Electricity is decidedly not a primary source of energy, and as such, it is always a thermodynamically degraded form of energy. It is therefore intrinsically wasteful. Storing electricity - generally as chemical energy although there are limited opportunities to store it as gravitational energy or as compressed gases - further thermodynamically degrades electrical energy and also is intrinsically wasteful both at the point of storage and at the point of release. Batteries, for example, give off heat, charging and discharging. Despite vast enthusiasm for energy storage by the general public, and tens upon hundreds of thousands of papers written on this subject, all of this is generally true, as I often state in my posts in this space.

There is one caveat however that is possibly of some possible import - if one captures energy that would have been wasted anyway and stores it in a form that is usable - the thermodynamic losses, although there still will be losses, will be minimized. At the most extreme, the only form of primary energy is nuclear. The sun is famously a nuclear device, which drives the wind, provides light - which over hundreds of millions of years was converted into chemical energy, dangerous fossil fuels which we can't burn fast enough - and also into current inventories of biomass, including but not limited to food. Some of the very best minds of the 20th century learned how to harness nuclear energy in the form of fission, although strictly put, uranium and thorium and the elements made from them are actually stored energy from ancient supernovae. Indeed, geothermal energy, which is provided by heat released by nuclear decay is also stored energy from ancient supernovae, being released in a generally regular way, earthquakes and volcanoes not withstanding. Only tidal energy is seemingly divorced from the nuclear primacy.

Nevertheless in practical engineering terms, we tend to think of dangerous fossil fuels and related biomass as primary energy, light from the sun as primary energy, wind as primary energy, geothermal as primary energy, gravitational energy, which is used in tidal and hydroelectric systems (the latter still driven by the sun) as primary energy, and of course, nuclear fission as primary energy.

The paper listed at the outset of this post is about wasting energy to make electricity and then wasting even more of it to make chemical energy, in this case formate and carbon monoxide, the latter, via the "water gas" reaction, an equivalent of hydrogen.

Despite my contempt for the thermodynamics of electricity, I do sometimes imagine cases where it might be acceptable to waste some energy to make electricity, and even chemical energy via electricity.

To illustrate, I offer a diversion. Over the last week or so, at various times of the day, on various days of the week, I've been downloading the data and graphics on the CAISO systems status pages, in three areas, demand, supply, and emissions. CAISO monitors and oversees electricity in California, CAISO = CAlifornia Independent System Operators. They have these wonderful real time page with graphics and numbers that will tell you the state of affairs with respect to electricity at any moment of the day. It's here: CAISO Pages

Here are some graphics I downloaded a little while ago as of this writing, near the 17th hour of the day, Pacific Time:

Here is the demand for electricity in California; the shaded area is historical data for the day, the line beyond is the predicted electricity demand.

The 17th hour, 5 pm PST or PDT +/- an hour or two, as the case might be is consistently close to the peak demand for energy in California, weekends, work days, holidays, every day.

Most people know that the sun is "going down" around that time, and as a result the output of solar energy is declining. California is unique among those places invested in the "renewables will save us" fantasy, in that the output of solar energy can dominate, at times, the output of so called "renewable energy", at least around noon.

This graphic and data breaks down what percentage was around the 17th hour of September 15, 2020:

This graphic and data shows the data for the September 15, 2020 up to 17:10 (5:10 pm):

The absolutely flat grey line near the bottom, 2,241 MW, is California's last remaining nuclear reactor, Diablo Canyon, two nuclear reactors contained in a single building. From midnight until 7 am this morning, it was producing just about as much energy as all of the wind turbines, solar cells, biomass combustion plants (which almost certainly include garbage incinerators), and geothermal plants in the entire state.

The disturbing line, of course, is the "imports" line and the "natural gas" line.

Here are California's emissions, as of 17:10 (5:10 pm PDT).

Don't hold me to this, but my general impression is that if you were to average all the "per hour" emissions data I've seen over the course of pulling up and downloading 20 or 30 of these files, the average would be somewhere between 7,000 metric tons and 8,000 metric tons of carbon dioxide ruthlessly dumped by the State of California to provide electricity to its citizens.

There are 8,767.76 hours in a sideral year, suggesting that California dumps about 65,750,000 metric tons of carbon dioxide into the planetary atmosphere each year, this while being "green" and regularly passing, going back to the wonderful time I lived in that State, now decades ago, "by 2000" and "by 2020" and "by 2050" and "by 2060" energy bills claiming that the state will be "green," at least as they imagine "green."

To be honest, my contempt aside, this actually isn't bad, and on their emissions page they have one of those innumerate "percent talk" statements about how great they're doing on emissions.

The reality is however, that all this talk, half a century of it, has not addressed climate change. California, and Oregon, and Washington, recently Australia, this week the Pantanal, are all burning, and powerful hurricanes slice through the Southeastern and Northeastern United States like coupled freight trains.

We needed to be releasing zero carbon dioxide into the atmosphere to generate electricity, and we needed to be doing it years ago.

I just went on the CAISO site again: Today's peak electricity demand took place at 17:18 this afternoon, at 35,971 MWe.

Diablo Canyon is not very impressive in terms of its thermodynamic efficiency with respect to its "primary" energy. Like most nuclear power plants designed in its era, it's around 33% efficient, atrocious really, but similar to all historic Rankine cycle plants. Modern gas powered plants, which are combined Brayton/Rankine cycle plants are closer, and sometimes exceed 50% efficiency.

The reality is, that today, to have provided all the electricity for today's peak demand, 35,971/2,241 would require 17 buildings the size of Diablo Canyon of Diablo Canyon's precise design - which is, by the way, going to be shut in three or four years, because it's "too dangerous" in the minds of people who can't think very well, but the huge tracts of the state burning each year is not "too dangerous."

Of course, at around 3:40 am on September 15, 2020, the demand for electricity in California was less than 22,000 MWe, and the bulk of the putative imaginary nuclear power plants would not be necessary.

In general, my thermodynamic sense is that it is stupid to generate electricity to create potential chemical energy. It is much smarter to use direct thermal to chemical energy conversion if one were to store energy.

However, all chemical processes require cooling, and the rules of process intensification, which are rules to derive the most usable energy per unit of primary energy generated, can conveniently convert thermal energy to electricity with Rankin or Brayton or Stirling heat engines, and storing this mechanical energy may be justifiably be done with an electricity intermediate.

The chemical process in the paper that I promised to discuss at the outset uses carbon dioxide as a feedstock; if this carbon dioxide is obtained from the waste dump into which we've transformed our planetary atmosphere, that can help clean it up.

Here is the cartoon associated with the paper's abstract, which can be read at the link:

From the introduction to the paper:

All countries that signed the Paris agreement have committed to reduce greenhouse gas emissions according to the target to keep global warming well below 2 C with regard to preindustrial values. This entails significant efforts to reduce the carbon footprint of all sectors in the following decades. Because industrial processes significantly contribute to the overall emissions of industrialized countries, for example, in the EU-28 with 8%, the development of innovative routes with lower or negligible CO2 emissions that can substitute traditional processes based on fossil resources can make a contribution to achieve the desired targets.(1−3) The electrochemical conversion of CO2 using renewable energy not only substitutes conventional CO2-intensive processes based on fossil resources but also utilizes CO2 itself as a feedstock, thereby making it a valuable carbon source.(4−6) Depending on the catalysts and reaction conditions, several carbon-based products can be obtained by the cathodic electrochemical reduction of CO2.(7,8)

This work focuses on the tin-catalyzed conversion of CO2 to obtain formate as the target product. Typically, CO2 is not solely converted to formate (eq 1) on tin catalysts but also to carbon monoxide to a lower extent (eq 2), while the aqueous electrolyte is reduced to hydrogen (eq 3) in a parasitic side reaction, reducing the charge efficiency with respect to formate formation, according to the following equations (1) (2) (3)

Formate can be used as a deicing agent or drilling fluid as well as for tanning and silage when protonated to formic acid.(9) In a proof-of-concept study, it was also shown that formate obtained via CO2 reduction reaction (CO2RR) can be utilized as energy carrier and fed into a direct formate fuel cell to produce electricity or after catalytic decomposition to hydrogen with subsequent re-electrification in a typical polymer electrolyte membrane fuel cell.(10,11)

Unfortunately, the conversion of dissolved CO2 on planar electrodes is limited to a maximum current density of well below 10 mA cm2 because of mass transport constraints evoked by the low solubility of CO2 in the aqueous electrolyte (33 mM L1 in H2O, 25 C, 1 atm) and the diffusion of dissolved CO2 from the bulk electrolyte to the electrode surface.(7,12,13) This limitation can be circumvented by the use of so-called gas-diffusion electrodes (GDEs) providing a porous architecture and intensifying the contact between the gas-, liquid-, and solid phases. Their use entails a substantial increase in the number of active sites while the diffusion length of dissolved CO2 to the catalyst surface is reduced. Accordingly, gaseous CO2 can be employed as the substrate and, due to the above effects; the macroscopic mass transport of the reactant is substantially accelerated.(14−16) As a result, the achievable current densities when using GDEs for CO2 electrolysis can be increased by more than an order of magnitude compared to planar electrodes without sacrificing selectivity toward CO2RR products.(17−19) The so far achieved current densities which are already on industrially relevant orders are the reason why GDEs have gained increasing interest in recent years for the investigation of CO2 electrolysis systems.(15,19−21) Nevertheless, long-term stability of these GDEs is an under-represented research topic in the literature. Besides potential catalyst degradation, GDEs might suffer from a change of their hydrophobic properties over time, resulting in flooding, efficiency losses, and in a shift of the product selectivity toward the undesired hydrogen evolution.(22,23) In that respect, electrochemical impedance spectroscopy (EIS) is a powerful tool to deconvolute contributions of specific physical and (electro)chemical processes to the overall resistance during operation of electrochemical devices. The knowledge which process (e.g., charge-transfer, mass transport of reactants) determines the polarization resistance gives valuable insights required for the rational optimization of the employed GDEs and can aid in revealing possible degradation mechanisms. To use EIS as an electrochemical diagnostic tool, it is crucial to know which physical phenomena are observed in the measured impedance spectra...

About their approach using Electrochemical Impedence Spectroscopy:

Electrochemical Impedance Spectroscopy

Before discussing the results of the EIS measurement, it is useful to get a general understanding of CO2 reduction employing the GDEs manufactured in our laboratory. A polarization curve for the GDE operated with 100 vol % CO2 in 1.0 M KOH at 30 C is depicted in Figure 1a. The curve was recorded via a stepwise increase of the current and logging the corresponding potential after 15 min at each value. At the maximum current, −400 mA cm2, the potential is approx. −725 mV with a FE for CO and H2 of about 12 and 5%, respectively. The remaining 83% is attributed to the production of formate. A slight scattering of the potentials at higher currents is induced due to the gas evolution at the GDE surface making a precise potential determination impossible. The observed FEs for CO (∼12%) and H2 (∼5%) remain nearly constant for potentials of −450 mV to −725 mV. This fact indicates that even at a current density of −400 mA cm2, no mass transport limitation for CO2 is observed because otherwise the FE for H2 would increase. For 1 and 0.5 mA cm2, the FEs were not determined because the quantification error would be too high due to the low amount of charge involved.

Figure 1:

The caption:

Figure 1. (a) Polarization curve and FEs obtained during CO2 electrolysis using aqueous 1.0 M KOH as the electrolyte and (b) impedance spectra recorded in a galvanostatic operation mode employing pure CO2 and N2 as feed gas.

FE is Faradaic Efficiency, which is a function of how many electrons participate in the desired reaction as opposed to side reactions. It is not equivalent to thermodynamic efficiency.

More commentary:

Temperature- and Current-Dependent Behavior

In Figure 2, the current- and temperature-dependent Nyquist and imaginary impedance versus frequency plots are shown. The obtained FEs for H2 and CO during the EIS measurements are in good agreement with the observed values for the FE acquired during the measurement of the polarization curve. The medium-frequency process around 20 Hz shows a distinct dependence on the applied current. A higher current leads to an exponential decrease of the corresponding medium-frequency resistance (cf. Figure 2a). Additionally, Figure 2b reveals that the characteristic frequency of this resistance shifts to higher values when increasing the current. The exact same trends can be observed for the medium-frequency process for an increase of temperature (cf. Figure 2c,d). This temperature- and current-dependent behavior is commonly associated with a charge-transfer process and strongly indicates that the medium-frequency process displays a charge-transfer reaction.(34,37) A process in the low-frequency region (on the right in the Nyquist plot) can be observed for the lowest current density of −25 mA cm2 (cf. Figure 2a).

Figure 2:

Figure 2. Impedance spectra obtained during CO2 electrolysis for (a) current density and (c) temperature parameter series to identify current- and thermal-activated processes and their corresponding imaginary part vs frequency plots (b,d).

Some additional graphics:

The caption:

Figure 3. (a) Impedance spectra recorded during CO2 electrolysis for varying CO2 volume fractions in the feed gas and (b) corresponding imaginary part vs frequency plots.

The caption:

Figure 4. (a) Impedance spectra recorded during electrolysis for varying CO2 volume fractions in the feed gas and (b) time-dependent behavior of the impedance spectra for the operation with 20 and 10 vol % CO2 in the feed gas.

Some more commentary:

Kinetic Isotope Effect

In the literature, there is still ongoing debate about whether CO2 (aq) or HCO3 (aq) is the active species during the rate-limiting step of the CO2RR.(8,41−44) To gain further insights into this matter, we will present experiments related to kinetic isotope effects (KIEs) in the following in which D2O is employed as solvent for the electrolyte. Processes which are affected by hydrogen atoms are expected to be slowed down when substituting hydrogen with heavier deuterium which should shed light on the species involved in the reaction.(45)

The changes when substituting H2O with D2O become evident comparing the data in the imaginary part versus frequency part representation of the results. These plots for pure N2 and CO2 gas feed in H2O- and D2O-based KOH electrolytes are depicted in Figure 5 and partially reveal significant differences in the characteristic frequencies of the processes. The measured spectrum for the operation of the GDE with pure N2 will be determined by HER only and should be influenced when using D2O as solvent and is used as benchmark comparison to demonstrate how a spectrum changes when it is affected by the substitution of hydrogen with deuterium. Indeed, the characteristic frequency, or in other words the velocity of the observed charge-transfer process, is reduced from 6.6 to 1.5 Hz when using D2O as solvent (cf. Figure 5a). The peak height which is in first proximity proportional to the resistance is increased, and the iR corrected potential also increases from −1003 to −1072 mV. Unsurprisingly, the ratio of the deuterium and hydrogen gas volume fraction (0.73:1.0) in the product gas stream agreed very well with the ratio of the corresponding thermal conductivities (0.75:1.0, λD2 = 138 mW K1 cm1, λD2 = 185 mW K1 cm1)(46)...

The caption:

Figure 5. Imaginary part vs frequency plots obtained from EIS measurements during electrolysis using pure (a) N2 and (b) CO2 as feed gas and H2O- or D2O-based 1.0 M KOH as the electrolyte.

An excerpt from the conclusion:

We investigated carbon-supported tin-based GDEs employed for the electrochemical conversion of CO2 to formate and CO in aqueous 1.0 M KOH via EIS. The goal was to elucidate the underlying physical processes to help interpreting future EIS spectra and enabling targeted optimization of CO2 converting GDEs. We observed four features in the EIS spectrum which we tried to further specify by carrying out parameter series, namely, temperature, CO2 partial pressure, current density, and electrolyte. Careful analysis leads us to the following interpretation of the spectrum: the 45 line in the high-frequency domain is a typical feature observed for porous electrodes representing resistances arising because of the electrical conductivity of the GDE matrix and the ionic conductivity in the pore system. The subsequent process in the high-frequency range was identified to be the conversion of CO2 with OH to form bicarbonate or vice versa. Our results suggest that the medium-frequency process depicts the charge transfer from the electrode onto CO2 (aq) to form the CO2 anion radical...

I personally love radicals, and that carbon dioxide radical is a very interesting species.

I'm not particularly great with electrochemistry, but I rather enjoyed the effort to get my head around this paper.

It also gave me a chance to ruminate on when the deliberate generation, for storage, of the thermodynamically degraded (but useful) form of energy, electricity, is acceptable for energy storage.

I trust you will have a safe and pleasant day tomorrow and will sleep well tonight.

2 replies, 424 views

Reply to this thread

Back to top Alert abuse

Always highlight: 10 newest replies | Replies posted after I mark a forum
Replies to this discussion thread
Arrow 2 replies Author Time Post
Reply Revealing Mechanistic Processes in Gas-Diffusion Electrodes During CO2 Reduction to CO/Formate. (Original post)
NNadir Sep 15 OP
John ONeill Sep 17 #1
NNadir Sep 17 #2

Response to NNadir (Original post)

Thu Sep 17, 2020, 06:48 AM


This website, started by a 17 year old Danish kid, summarises the daily output of CAISO, but also of about a hundred other grids, ranging from whole countries down to little islands like supposedly 100% renewable El Hierro, in the Canary Islands. It conveniently computes the carbon footprint average for each point of the day, with a percentage of 'renewable' and 'non-renewable' (ie nuclear) non fossil generation. I've linked Kyushu, partly because my niece is living there, and also because it shows pretty dramatically what the effect on emissions is of all the solar built there since 2011, compared to the two reactors which have still been permitted to run. Actually, the bell curve for solar is usually a lot taller, but they have some nasty weather going through. Cleanest countries are, of course, the hydro and nuclear ones. https://www.electricitymap.org/zone/JP-KY?wind=false&solar=false

Reply to this post

Back to top Alert abuse Link here Permalink

Response to John ONeill (Reply #1)

Thu Sep 17, 2020, 09:11 PM

2. Well, thank you. I bookmarked it. CAISO, reflects an American Denmark or Germany...

...inasmuch it is committed to phasing out nuclear with no corresponding commitment to phase out fossil fuels.

The nice thing about this website as opposed to CAISO, is that the pricing is clearer than it is on CAISO, CAISO not listing bulk pricing but rather busbar prices for each facility by bringing the cursor over each facility, pricing apparently fluctuating as a function of demand.

Denmark, of course, is making a lot of money drilling for oil and gas in the North Sea. However, for electricity, in terms of carbon dioxide, it's rather clean, coming in at under 200 g/kWh, as opposed to Germany, which is well over 300 kWh, roughly in the area of a high efficiency combined cycle gas plant.

That offshore oil and gas drilling hellhole Denmark, as opposed to Germany is able to do be a little lower than a combined cycle gas plant by importing power, in West Denmark from Norway - a hydro driven country, except for a single gas plant - and Sweden, which is a hydro and nuclear powered country, effective "batteries" whenever the wind isn't blowing in Denmark or the North Sea.

The problem of course, with hydro, which in most places - except the tropics - approximates the carbon load of nuclear 10-25 g CO2/kWh, is the other currently largely ignored cost, which is land use. Land use changes, as I understand it, accounts for roughly 10 gigatons (metric) of the 45 gigatons (metric) added each year because people don't go nuclear about climate change.

Another variable, which does not show up in the electricity profile of that offshore oil and gas drilling hellhole, is the mass intensity. This, I think, is going to show up as the true Achilles heel of so called "renewable energy" when it is coupled to land use, which for wind is a disaster. There is not enough steel, concrete, aluminum, neodymium and dysprosium to make the wind industry sustainable, nor really, is there enough pure wilderness to despoil.

France, regrettably, is creeping up the carbon dioxide ladder, and is now over 50 g CO2/kWh. It is adding wind capacity which implies adding gas capacity, since wind depends on dispatchable gas generating capacity either as spinning or fast start up reserve.

It's too bad...

One of the interesting things I've noticed about CAISO over the last week or two, while downloading supply graphics at different times of the day, is that over the last two weeks, wind rarely gas risen over 1000 MWe instantaneous power. It must be a season of doldrums, which for now is a good thing, since wind would fan the fires. This may be very different at different weather conditions of course, but it does mean that if California was relying on millions of metric tons of batteries charged by wind, they'd be in bad shape right now.

Humanity abandoned wind energy in the 19th century for precisely this reason, it's unreliability over uncomfortably long periods.

In American history of course, the unreliability of the wind saved George Washington from being wiped out or captured in the Battle of Long Island, because the wind wasn't blowing and the British fleet couldn't go up the Hudson to cut off his escape from Manhattan using rowboats to New Jersey.

It is nice site, you've linked.

I'll stick with CAISO mostly for my current research focus, because California is an American anti-nuke "renewables will save us" nirvana. Regrettably its on fire, because all the talk about renewable nirvanas have not arrested climate change; in fact, it's only made it worse. Bad mistakes, unfortunately, in this case, have led to bad results.

Thanks again.

Reply to this post

Back to top Alert abuse Link here Permalink

Reply to this thread