Electrocatalytic Nitrate Reduction to Ammonia at Atomically Dispersed Titanium Sites on Carbon Nanoflowers

Abstract

Active nitrogen species are widely acknowledged as pollutants that are both harmful to the environment as well as human health. Emissions from wastewater and agricultural applications contain nitrogen in many forms, but in soils and waterways nitrate (NO3 -) predominates due to the presence of nitrifying bacteria. In comparison to more reduced forms of nitrogen such as total ammonia (NH3/NH4 +), nitrate is not only a less versatile species from a chemical manufacturing and energy standpoint; it is also more difficult to selectively separate from wastewater streams. For instance, we have conducted considerable work on electrochemical ammonia stripping, achieving high selectivities and even enabling concentration of ammonia in the form of strong ammonium salt solutions.1 Moreover, it is estimated that recovery of active nitrogen species from municipal and agricultural wastewaters could offset the need for ~30% of Haber-Bosch ammonia synthesis. From this perspective, it is desirable to develop catalysts that are active and selective for the conversion of nitrate to ammonia. Titanium metal has been demonstrated as a relatively active and selective catalyst for nitrate reduction to ammonia in acid;2 however, similarly active and selective catalysts under neutral-to-alkaline conditions are a subject of active research.3–6 Here we report on the performance of atomically dispersed titanium sites on carbon nanoflowers7 for the reduction of nitrate to ammonia under alkaline conditions. The dispersity of these catalysts was supported by XAS (little/negligible Ti-Ti scattering, putative Ti-N and Ti-N-C scattering), STEM (visual confirmation of atoms), and XRD (no crystalline Ti phases observed). We contrast the performance of these atomically dispersed catalysts with that of bulk titanium foil, which shows comparably low activity and selectivity for nitrate reduction at similar pHs and potentials. In addition, we compare the atomically dispersed titanium catalysts with titanium nanoparticles, as a first-order check for catalyst aggregation under reductive reaction conditions. Moreover, the performances of all of these catalysts are compared with the baseline/controls of bare carbon paper, as well as carbon nanoflowers supported on carbon paper without titanium dopant. Additional control experiments include tests conducted in the absence of nitrate. These lines of evidence, together with the relatively high rate of ammonia production (>0.01 mmol / cmgeo 2 / h, >10 mmol / mg Ti / h), support the notion that atomically dispersed titanium is uniquely active and selective for NO3RR under alkaline conditions.Further, for the above catalysts at pH ~13, we show the impacts of potential on the product distribution, including nitrogen mass balance and Faradaic efficiency. As a best practice, we demonstrate consistent closure of both metrics. Potential was varied between -0.4 V and -0.85 V vs. RHE. Partial current toward NO3RR increases with increasing overpotential, as another sanity check supporting the findings. Notably, in addition to observing ammonia and nitrite as products of NO3RR, we also checked for other stable intermediates (in addition to N2) along the 8-electron reaction pathway – and as a result, we are able to report small amounts (FE ~5-10%) of hydroxylamine (NH2OH) product.In the specific case of atomically dispersed titanium on carbon nanoflowers, we also report on certain observed electrolyte effects, such as the incorporation of perchlorate vs. sulfate as a supporting electrolyte. We demonstrate the role of supporting salt both in modifying selectivity – e.g. sulfate suppressing HER, possibly via site-blocking – as well as in altering the carbon support itself. These observations contribute to a growing body of work that will enable the engineering of these catalysts for high-rate reduction of nitrate to ammonia, aiding both wastewater remediation and the decarbonization of ammonia synthesis.