Photoreforming of food waste into value-added products over visible-light-absorbing catalysts

Food and mixed wastes are converted into H2 and organics over CdS and carbon nitride photocatalysts.


Oxidation intermediates survey, CdS/CdOx in 10 M NaOH
Acetate [b] 5.00 ± 0.25 124 ± 23 Formate [b] 147 ± 30 10700 ± 2200 Lactate [b] 290 ± 14 19800 ± 2000 Pyruvate [b] 0.0 ± 0.0 0.0 ± 0.0        Note that CH4 is used as a quantification reference, and is not a gaseous product of the system. The O2 observed in the CdS spectrum is atmospheric. In the case of PR in H2O, the H2 could be easily separated from CO2 by common industrial processes such as pressure swing adsorption. Figure S13. Zeta potential measurements of (a) CdS QDs (data from ref. [9]) and (b) H 2 N CNx with and without Ni2P over a range of pH.

Sugar hydrolysis in alkaline media
Sugars in alkaline media will react following Lobry de Bruynvan Ekenstein transformations (Scheme 1), which show reversible isomerisation between different sugars. For fructose, the formed isomers are glucose and mannose, which were detected in the pre-treated solutions.
However, glucose (e.g. from starch) will also engage in this reaction and the same isomers will be formed. The intermediates derived from the enediols can take part in a variety of reactions, which are responsible for the broad array of decomposition products observed after pre-treatment. The precise mechanism of these decomposition reactions has been the subject of many detailed studies, 11,12 and the type and amount of decomposition products can be influenced by reaction conditions such as temperature, base and sugar concentration.
In our case, we could detect formate, lactate as well as C5 and C4 sugars in the alkalinetreated fructose and starch samples. The formation of formate and the shorter chain sugars S28 occurs by carbon cleavage from a nucleophilic attack by an OH − anion (Scheme 3). The resulting Cx-1 sugar can participate in the same reaction. The formation of lactate from sugar hydrolysis has also been reported (Scheme 4). [11][12][13] Briefly, a C6 sugar is cleaved into two C3 units. A dehydration step then yields an α,β-dihydroxy compound, and a subsequent nucleophilic attack by an OH − anion yield lactate. The mechanism of photoreforming of sugars (fructose, glucose) has been studied by Sanwald,et al. (Scheme 5). 14 In brief, ring-opening (C-C α-scission) of the sugar generates formate species. Light-driven formate hydrolysis (path A) is very slow under neutral conditions, and the primary photoreforming pathway is therefore suggested to be oxidative C-C cleavage (path B) to shorter formates. This mechanism would account for the formate that we observed after photoreforming of fructose and starch in neutral conditions. Scheme 5. Photoreforming of sugars in neutral conditions, as reported in ref. [14].

Details of Carbon Footprint Calculations
For all cases, a raw material input of 1 kg fructose and 40 L H2O (with 22 kg KOH for case 1) was utilised. Experimentally measured conversions (see Table S9) were used, except for the 100% conversion cases. For simplicity, the following assumptions were made:  A lower H2 energy density of 120 × 10 6 J kg −1 was used;  The carbon footprint of fructose is assumed to be equal to that of real food waste;  The catalyst is re-usable and not included in the calculations;  Heat recovery of 80% is applied to the pre-treatment process;  Less formate is experimentally observed than we would expect from the stoichiometric conversion of fructose to formate and H2. The remainder of the carbon is assumed to be contained within CO2/CO3 2− . The quantities of CO2/CO3 2− utilised in the case studies below are estimations based on this assumption, rather than experimental values;  The energy required to extract formate was not included, as an estimated value for this process could not be found in the literature;  The carbon footprint of waste disposal is not included due to lack of data. . b this value is negative since we are producing formic acid rather than consuming it. c calculated assuming that pre-treatment occurs in a polypropylene tank (thermal conductivity 0.20 W m −1 K −1 , cross sectional area 0.75 m 2 , wall thickness 4.8 mm), initial water temperature and external air temperature are both 25 °C, and the carbon footprint of electricity consumption is 500 g CO2 / kWh. 17 d calculated assuming that stirring requires 1 kW m −3 , 16 and that the carbon footprint of electricity generation is 500 g CO2 / kWh. 17 e this value was obtained from ref.
[18]. f this value is negative since we are removing food waste.