N-amended graphene catalyses vinyl chloride dechlorination by iron minerals

Chlorinated solvents (CS) are a common class of contaminants in soils and groundwater that are difficult to degrade fully via natural processes, and their effective remediation is critical to ensuring groundwater safety. Reductive degradation of CS by reduced iron materials such as zerovalent iron (ZVI) is a well-recognized remediation strategy. However, ZVI shows limited reactivity with some of the more persistent CS, such as vinyl chloride (VC). Moreover, ZVI is non-selective and its reducing equivalents are quickly used up by anaerobic oxidation and other side reactions. Soil Fe(Ⅱ) minerals, such as green rust (GR), magnetite and iron sulfide, should, from a thermodynamic standpoint, be able to reduce VC, and do not react with water, i.e., no hydrogen production. However, many experimental studies have shown that VC reduction by these materials is limited.

Carbonaceous materials (CMs) are often used as sorbent material in the removal of organic pollutants from wastewater. Moreover, they have been shown to act as catalyst in dechlorination reactions with reduced iron materials. Based on these observations, this thesis aimed to develop a CM capable of catalyzing fast and full degradation of VC by reduced iron reactants. In a first instance, nanoscale ZVI (nZVI) was used as the reduced iron material to optimize the properties of the CM. Afterwards, the optimized CM was also tested on iron (II) soil minerals, and in real VC contaminated groundwaters.

A catalytically active CM was produced by pyrolysis of graphene nanoplatelets that were mixed with a nitrogen source (i.e., urea). N-doping has been shown to increase the formation of redox active N functional groups. Exploration of different pyrolysis temperatures (PT) showed that N-amended graphene (NG) materials prepared with PT between 600 and 950 °C had a significant catalytic effect on the reduction of the VC by nZVI, with NGs produced at 950 °C (NG950) yielding highest VC reduction extents and rates. For example, the addition of NG950 to nZVI increased the VC reduction extent from 1% to more than 69% within 68 h and gave a pseudo first order rate constant of 0.047 h -1 . Exploration of different urea dosages showed that NG catalytic activity was positively correlated with urea dosage when produced at PT of 600 °C, but urea dosage had no significant effect on NG catalytic activity when produced at PT of 950 °C.

A systematic analysis of the physicochemical properties of NG was completed, including carbon structure, specific surface area, C and N content, surface elemental composition and speciation. Correlation of these properties to the VC reactivity data suggested the content of pyridine-N-oxide and structural defects (both created by N-doping), to be key factors for high NG catalytic reactivity. Multiple linear regression fitting as well as principal components analysis were performed which further validated these interpretations. In terms of VC reduction mechanisms by NG amended nZVI systems, data from hydrogen yield and atomic hydrogenscavenger experiments suggest that NG enables storage of atomic hydrogen and thus increases its availability for VC reduction. However, NG likely also enables electron transfer from nZVI to VC.

Finally, NG was tested for catalyzing VC reduction by green rust, GR. NG prepared with an urea:graphene mass ratio of 0.25 and PT of 950 °C (NG950-0.25) was found to enhance VC reduction by GR (with an Fe(Ⅱ)/Fe(Ⅲ) ratio of 2.7) the most, while no VC reduction was observed with GR alone. Specifically, NG950-0.25 enabled more than 80% of VC conversion to ethylene within 7 days of mixing with GR, and comparison to the NG+nZVI system showed almost identical VC reduction extents and rates. Cyclic voltammetry, hydrogen yield and quenching experiments confirmed that atomic hydrogen is also forming in the here tested GR system, and responsible for parts of VC dechlorination. Additional CS competitionexperiments with either perchloroethylene, trichloroethylene, or cis-dichloroethylene added to VC reactors, demonstrated insignificant CS-VC competition effects and full degradation also of other CS by the NG-GR system. Lastly, VC dechlorination kinetics were also not affected by groundwater ions.

In conclusion, this thesis demonstrates that NG can be used as a very efficient catalyst to degrade CS by iron minerals, thus provides a very promising strategy for the reductive remediation of VC. This study could contribute an important reference for the exploration of the catalytic reduction mechanism of CMs, especially the importance of N doping to create N functional groups and defective structures. In addition, due to the atomic hydrogen storage and hydrogen production involved in this study, the research in the field of energy has a promising application.