This work involves the long term electrooxidation of glycerol in a custom autoclave setup. Experiments are usually performed over approximately 48 hours. The custom autoclave setup was adapted from work done by Thomas Holm [1,2]. Within the autoclave a typical three electrode cell is built, wherein all glass pieces have been replaced with alkaline stable counterparts. This is largely done with Teflon, however a Nafion membrane is used to separate the working and reference electrode volumes from the bulk volume at the counter electrode to avoid contamination at the reference, and to keep the oxidation volume small. A small oxidation volume leads to higher concentrations of products, allowing them to be more easily detected using HPLC.

Analysis of the oxidized products is performed via HPLC using an Aminex HPX-87H column. Using these techniques we have been able to show that the temperature at which electrooxidation occurs significantly alters the selectivity of products, as well as causing some new products to form (See Figure 1). Calibration curves have been made up for each of these products using known standards, and so we are able to quantitatively determine how much of each product is formed. However, the carbonate peak is not quantitative, and so carbonate was separately analysed by gravimetric or titrimetric methods.

With this information we have been attempting to carry out a carbon balance between products and reactants to ensure all products has been accounted for. Though some reasonable values have been attained for this, problems arise when an electron balance is attempted. Much more oxidation is carried out than is expected from the number of electrons used during the experiment, and so we can conclude that some chemical interactions are occurring in solution. It is likely that some products are reacting with the alkaline solution, perhaps through chemical disproproportionation reactions. The existance of such reactions can be confirmed for glyceraldehyde, which has been shown to exist on a short time scale using PM-IRRAS [3] even though no glyceraldehyde is detected by HPLC. We find that HPLC glyceraldehyde standards decompose into other compounds that we see here. Work is ongoing to establish the identity and distribution of products as a function of the reaction conditions.

This research was conducted as part of the Engineered Nickel Catalysts for Electrochemical Clean Energy project administered from Queen’s University and supported by Grant No. RGPNM 477963-2015 under the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Frontiers Program.

1. T. Holm, P.K. Dahlstrøm, S. Sunde, D. A. Harrington, F. Seland, ECS Trans, 75, 1055, (2016).
2. T. Holm, P.K. Dahlstrøm, O.S. Burheim, S. Sunde, D.A. Harrington, and F. Seland, Electrochim. Acta., 222, 1792 (2016).
3. M.S.E. Houache, E. Cossar, S. Ntais, E.A. Baranova, J. Power Sources, on the web, doi:10.1016/j.jpowsour.2017.08.089.