Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

NNadir

(33,368 posts)
Wed Jan 9, 2019, 07:59 PM Jan 2019

Metal Free Thermochemical Water Splitting at Unusually Mild Conditions.

The paper I'll discuss in this post is this one: Phosphorus-Doped Graphene as a Metal-Free Material for Thermochemical Water Reforming at Unusually Mild Conditions (Garcia et al ACS Sustainable Chem. Eng., 2019, 7 (1), pp 838–846.

Recently in this space I discussed the thermochemical splitting of carbon dioxide (into CO and O2 gases) using a cerium oxide based catalyst in which the oxygen evolution reaction took place at 1400C, showing that there is - as currently operated using "simulated solar energy" - not enough cerium on earth to split one billion tons of carbon dioxide, using either solar thermal or nuclear energy (although nuclear is considerably less onerous in terms of putative cerium demands). One billion tons of carbon dioxide about 3% of what we currently dump each year into the planetary atmosphere.

Here's that post: Cerium Requirements to Split One Billion Tons of Carbon Dioxide, the Nuclear v Solar Thermal cases

From my perspective, the thermochemical splitting of either carbon dioxide or water is probably the only serious manner in which climate change can be reversed, and even if taken seriously - there are few people on this planet left or right who are serious about addressing climate change - it would still be a long shot, but, as the only shot with a modicum of probable success, one worth taking.

Scientists however, continue to work on the problem.

I have spent many years considering thermochemical cycles for splitting either water or carbon dioxide using nuclear energy (or less seriously solar thermal energy), and most, with a few exceptions, involve metals - the main exception being the famous sulfur iodine cycle (which has metal based modifications however) - my personal favorite being the zinc oxide cycle for reasons I won't go into here. The one I'll discuss here - it's really a half reaction, not a full cyclic reaction - is new to me, I must admit. It clearly is not scalable or even worthy of consideration of scale, but the research is extremely interesting and certainly comes under the rubric of "a good lead," particularly since the required temperatures for hydrogen evolution are unusually low, about 900 C.

This involves an interesting material, graphene, which has been the subject of huge amounts of research in materials science.

From the introduction:

Among the most general ways to obtain graphene-related materials, the one starting with graphite that is submitted to deep chemical oxidation to graphite oxide, followed by subsequent exfoliation to graphene oxide (GO), and final chemical reduction provides a graphene material denoted as reduced graphene oxide (r-GO). r-GO is among the most widely studied graphene materials because it can be prepared in a reliable way in gram scale (Scheme 1).(1,2)

figure

Scheme 1. Process of Preparation of r-GO from Graphite Involving Oxidation to Graphite Oxide and Exfoliation to GOa


a(i) Chemical oxidation, (ii) exfoliation, and (iii) chemical reduction.


The above process to perform graphite exfoliation by conversion of graphene (G) into GO is based on the possibility of carrying out the oxidation and reduction of G/GO, increasing the oxygen content to above 50 wt % from G to GO, with a certain degree of control, and then, subsequently decreasing this oxygen content from 50 to about 10 wt %, which is characteristic of r-GO. This ability to increase and decrease the oxygen content on G sheets is reminiscent of the so-called Mars van Krevelen oxidation/reduction of nonstoichiometric transition metal oxides, in where the oxygen content of the inorganic oxide can be varied to a certain extent, generally much lower than the one commented in the case of G/GO/r-GO.(3) This Mars van Krevelen mechanism has been, however, advantageously used to promote catalytic oxidations/reductions, and more related to the present work, this swing between the two related materials with different oxygen contents is at the base of the thermochemical cycles for water splitting or steam reforming.

In steam reforming, a substrate (S) promotes the reduction of water, resulting in the generation of hydrogen (eq 1) and substrate oxidation. If the oxidized form of the substrate, most frequently inorganic oxides (for instance ceria, perovskites, or spinel ferrites) due to the required thermal stability (T = 1300–1500 °C), can subsequently be thermally reduced by oxygen evolution (eq 2), then the two steps can serve to perform cyclically the overall water splitting.(4,5) It has been reported, that one of the main challenges in thermochemical water reforming is the development of materials able to promote efficiently thermochemical transformations at low temperatures (<1100 °C), especially for large scale production.(5?7)


Graphene is a form of carbon in which all of the carbon atoms are bonded together in a plane, which is also characteristic of graphite, but unlike graphite, the graphene is exactly one atom thick. The layers are not connected.

What is interesting here is that the carbon source for the graphene is biomass, as opposed to a dangerous fossil fuel source, meaning that it is possible that this approach is sustainable, at least on a moderate scale.

One source is alginic acid, which is obtained from brown algae, many species of which are believed to be excellent tools for carbon capture from the atmosphere. The other is phytic acid, which is per-phosphorylated inositol, which is found in beans, and notably in manure, where it is responsible for the environmentally problematic concentration of phosphorous.

Graphene in the presence of steam is reformed normally, yielding carbon dioxide and hydrogen - and the reforming of biomass is probably an excellent approach to carbon capture as well as thermochemical splitting - however there are certain mineral considerations that represent significant hurdles.

In order to prevent the reformation of graphene, the authors here have phosphorylated graphene oxide.

Some pictures from the paper, first the synthesis of the graphene (and its oxide):



The caption:


Scheme 1. Process of Preparation of r-GO from Graphite Involving Oxidation to Graphite Oxide and Exfoliation to GO


(i) Chemical oxidation, (ii) exfoliation, and (iii) chemical reduction.



Next, the xray photoelectron spectrum (XPS) of the graphene:



The caption:

Figure 1. XPS survey spectrum (a) and C 1s (b), O 1s (c), and P 2p (d) high-resolution peaks recorded for Phy-G and their corresponding best deconvolution fits.


The chemical nature of the phosphorous attached to the graphene can be discerned from nuclear magnetic resonance spectrometry (NMR) since the only isotope of phosphorous that occurs naturally, 31P, is magnetically active. The 31P spectrum:



The caption:

Figure 2. Solid state 31P NMR spectrum of Phy-G, with indication of the assignment based on the literature.(31?34)


"Phy-G" is phosphorous doped graphene.

High resolution tunneling electron microscope images:



Atomic force microscope images:

The caption:

Figure 4. AFM images of Phy-G samples. (a) General wide-field image of Phy-G samples showing a 2D sheet on which smaller particles are supported. (b) 3D image of a wide-field region of the same Phy-G sample. (c) Image corresponding to a part of a 2D sheet, where the blue, green, and red lines indicate the height measurements. (d) Height measurement along the lines indicated with the same colors in image (c).


The hydrogen evolution over 21 cycles:

The caption:

Figure 6. H2 evolution upon 21 consecutive activation-oxidation cycles (red). The temperature cycles have been included in blue.


The authors do some in silico calculations. Here's some fun details of their approach:

The potential energy calculations were performed using spin polarized DFT with the VASP 5.4.1 code (Vienna ab inito simulation program) developed at the Fakultät für Physik of the Universität Wien.(20) We used the projector augmented wave (PAW) scheme(21) with the Perdew–Burke–Ernzerhof (PBE)(22) exchange and correlation (xc)-functional and a plane-wave energy cutoff of 400 eV. The system was modeled by a hexagonal 5 × 5 unit cell containing 50 atoms with a P atom substituting a C atom (2% doping),(23) with an optimized C–C bond separation of 1.429 Å and a 14 Å separation between graphene sheets. ?-point sampling of the reciprocal space was used in the optimizations and the nudged elastic band (NEB)(24) method calculations.


Here's what they found:



The caption:

Figure 8. (a) Calculated PBE free energy profile (kcal/mol) at 650 °C for the stepwise thermochemical water splitting reaction on P-doped graphene (2%). The approximate transition structures TS1 and TS2 are the highest points on the NEB profiles (see Computational Details section). The structures include the most significant bond lengths in Å and angles in °. (b) Calculated PBE free energy in kcal/mol (relative to the R structure) for the intermediates formed in three successive hydrolysis steps (addition of a H2O molecule and cleavage of a P–C bond at every step) resulting in formation of phosphoric acid. Note that, in both figures, only the carbon atoms of the unit cell in the vicinity of the catalytic center are displayed.


Of course the main problem with this system is that oxygen is not evolved, the reduction of water to hydrogen is first accomplished by the oxidation of phosphorous, and finally, after a number of cycles, to the oxidation of the graphene, that is, ultimate reformation.

The authors write:

Lack of O2 Evolution
It is worth noting, that evolution of O2 was not detected in any step in these experiments, either using Phy-G or G, indicating that eq 2 does not take place. However, since H2 evolves in the hydrolysis steps, it is clear that the O atoms present in H2O must remain attached in the Phy-G catalyst or could promote some decomposition. In order to address the nature of the oxygenated groups being formed on Phy-G, Raman spectroscopy and XPS analysis of the Phy-G catalyst after extensive use in the thermochemical H2O reactions were carried out.

The XPS P 2p peaks of Phy-G, after its use in steam reforming and its best deconvolution fit, are presented as Figure 7, which also provides a comparison with the P 2p peak of the fresh sample. The first information provided by XPS was a decrease in the proportion of P quantified by the decrease of the P/C atomic ratio from the initial 0.072 value for the fresh Phy-G material to the 0.021 ratio determined for the Phy-G sample after its use in the thermochemical H2 generation from H2O. Comparison of P 2p spectra of fresh and used Phy-G confirms a shift in the P 2p peak of the used Phy-G toward higher binding energies, indicating the increasing presence of oxidized P in the catalyst composition. In addition, as it can be observed in Figure 7, the P 2p peak of Phy-G after the reaction presents only two main components instead of three. In this case, the component at 132 eV, related to the P–C bond, is no longer present, while components at 134 and 136 eV in relative percentages of 74.5 and 25.5%, respectively, are related to the formation of the P–O bonds...

...The solid-state 31P NMR spectra of fresh and used Phy-G have been similarly recorded, and they are compared in Figure S7. As it can be seen there, the contribution of peaks corresponding to triphenylphosphine and triphenylphosphine oxide has considerably decreased, while the peaks attributed to phosphate and other P oxide groups have undergone a notable increase in good agreement with the information provided by XP and Raman spectroscopies. Therefore, the incorporation of O atoms in P-doped G as phosphate groups is confirmed by three different techniques, and thus, the lack of O2 gas in the stream can be attributed to the oxophilic nature of P and also, to some degree, of graphenic C oxidation during reaction. Observation of CH4 and CO in the thermochemical cycles clearly indicates this gradual oxidation of G, since it is the most likely origin of CH4 is methanation of CO2.


Nevertheless, a cool paper, and quite interesting for the development of future catalytic systems.

An excerpt of the paper's conclusion:

It has been found experimentally that defective G obtained from biomass pyrolysis undergoes steam reforming at temperatures above 400 °C forming H2 and CO2. Grafting of P atoms on the G sheet increases considerably its stability under conditions of steam reforming. A graphenic material doped with P was obtained by pyrolysis of phytic acid. Characterization of this material shows that together with the expected P-doped G, the other nanoparticulated component is also present in much lesser proportions. Although the stability of Phy-G is notably higher than that of G, and H2 evolution is observed, no oxygen evolution could be achieved under the conditions tested. It seems that oxygen becomes too strongly attached to P atoms and also some degree of oxidation of the graphenic material to CO and CO2 (converted to CH4) is occurring...


Have a nice day tomorrow.
Latest Discussions»Issue Forums»Environment & Energy»Metal Free Thermochemical...