Opening the pathway towards a scalable electrochemical semi-hydrogenation of alkynols via earth-abundant metal chalcogenides

Electrosynthetic methods are crucial for a future sustainable transformation of the chemical industry. Being an integral part of many synthetic pathways, the electrification of hydrogenation reactions gained increasing interest in recent years. However, for the large-scale industrial application of electrochemical hydrogenations, low-resistance zero-gap electrolysers operating at high current densities and high substrate concentrations, ideally applying noble-metal-free catalyst systems, are required. Because of their conductivity, stability, and stoichiometric flexibility, transition metal sulfides of the pentlandite group have been thoroughly investigated as promising electrocatalysts for electrochemical applications but were not investigated for electrochemical hydrogenations of organic materials. An initial screening of a series of first row transition metal pentlandites revealed promising activity for the electrochemical hydrogenation of alkynols in water. The most active catalyst within the series was then incorporated into a zero-gap electrolyser enabling the hydrogenation of alkynols at current densities of up to 240 mA cm−2, Faraday efficiencies of up to 75%, and an alkene selectivity of up to 90%. In this scalable setup we demonstrate high stability of catalyst and electrode for at least 100 h. Altogether, we illustrate the successful integration of a sustainable catalyst into a scalable zero-gap electrolyser establishing electrosynthetic methods in an application-oriented manner.


Mechanochemical Synthesis of the Fe9-xNixS8 & MCo8S8 pentlandite catalysts
The synthesis of the catalysts was conducted through mechanochemical means. 1 The Pnmaterials were synthesized from the high-purity elemental powders. Under argon atmosphere, 5 g of a stoichiometric mixture of the elemental powder and 24 g of 2 mm zirconium oxide milling balls were mixed in a 20 mL zirconium oxide grinding vessel. In a Fritsch Planetary Micro Mill Pulverisette 7 premium line, the mixture was milled 120 min at 1100 rpm for two cycles with a 60 min break in between to obtain the pentlandite catalyst.

Physical Characterization
SEM imaging was either performed using a TESCAN Vega 3 microscope or a ZEISS Gemini2 Merlin HR-FESEM equipped with an OXFORD AZtecEnergy X-ray microanalysis system for energy dispersive X-ray spectroscopy (EDX).

PXRD Analysis
Powder X-ray diffraction (PXRD) measurements were performed on a Bruker D2 Phaser diffractometer equipped with a LynxEye detector operating at 30 kV acceleration voltage and 10 mA emission current using Cu K-α radiation (λ = 1.54184 Å). The data was recorded in a range from 10-70° 2θ.

Electrochemical investigation of pellet electrodes in a PEEK H-type cell
Fabrication of the employed electrode was performed similarly to previous investigations. 2 In short, approximately 50 mg of the obtained powder materials was pressed with a minimum pressure of 800 kg cm -2 into a 3 mm pellet with a geometric area of 0.071 cm 2 . The obtained pellets were integrated into electrodes consisting of a brass rod with screw threads and a PTFE housing. As a binder material carbon glue (3M) was employed. The prepared electrodes were placed in an oven overnight at 60°C to allow the glue to harden. Afterwards, the use of 12 μm, 3 μm and 0.3 μm sanding papers to polish the pellet ensured an even, smooth surface.
To remove any oxide species from the surface of the electrode, conditioning of the electrode was performed prior to electrolysis via linear sweep voltammetry. Specifically, a potential range from 0 to -0.5 V vs. RHE at a scan rate of 100 mV s -1 was applied, till a stable electrochemical response in the voltammogram was obtained. Electrolysis was performed under galvanostatic conditions at 100 mA cm -2 for 2 h under constant stirring of the electrolyte. A Ni-mesh served as the counter electrode, while a RHE (MiniRHE from Gaskatel) was used as the reference electrode. The anode compartment was filled with 12 mL of 1 M KOH, while the cathode compartment contained 12 mL of 1 M MBY in 0.3 M KOH/H2O. The two half-cells were separated with the help of a Nafion117 cation exchange membrane. Each measurement was repeated at least two times.

Product quantification for the ECH with pellet electrodes
Analysis of the electrolysis products in the PEEK H-type cell was performed via headspace GC-MS analysis. The quantification took place by directly analyzing 1 mL of the obtained catholyte utilizing a Shimadzu GC-MS QP2020 equipped with a HS20 headspace sampler.

Electrochemical investigations in a zero-gap electrolyzer
Electrochemical investigations in membrane electrode assemblies were performed in an inhouse built single cell electrolyzer with an electrode area of 12.57 cm 2 (40 mm diameter). A compressed Ni-foam served as the anode, a porous carbon electrode coated with 5 mg cm -2 of the respective catalysts served as the cathode, employing PTFE gaskets and a torque value of 5 N m over eight screws to ensure a leak-less operation. The used Fumasep membranes were conditioned overnight in 1 M KOH before use. During electrolysis, the anolyte and catholyte solutions were circulated through the half cells at flow rate of 12 ml min -1 , while a homogeneous liquid distribution was guaranteed through the use of Ti-based serpentine flow fields. A copper plate in direct contact with the flow field served as the current collector plate. Electrolysis was performed for 2 h, with samples being taken at the end of electrolysis from both the anolyte and catholyte. The respective chronopotentiometric curves up to a current density of 80 mA cm -2 (1 A) were recorded on a Gamry Interface 1010B potentiostate/galvanostate, while for current density values above 80 mA cm -2 , a Gamry Reference 3000 potentiostate/galvanostate equipped with a 30K booster module was employed.
For each experiment, a new MEA was constructed/investigated, with each experiment being investigated at least two times.

Electrode Fabrication
To prepare the catalyst PTFE-containing catalyst inks, 0.5 g of the mechanochemically synthesized Fe3Ni6S8 catalyst was mixed with 15 g 2-propanol, 4 mL H2O and 0.2 g Triton X 100. The mixture was placed into an ultrasonic bath for 5 min. Afterwards, the ink was dispersed at 13600 rpm using a T 25 digital Ultra-Turrax for 1 min. Afterwards, the respective amount of a 60 wt% PTFE dispersion was added while stirring. The suspension was then homogenously spray-coated with an Iwata SBS airbrush on a 8.5x8.5 cm carbon cloth or carbon felt that was heated on a hot plate to 100 °C. Afterwards, the PTFE-containing electrodes were heat-treated at 240 °C for 20 min to remove the surfactant in the ink.
For the PVDF containing ink, a 5 wt% PVDF solution in methanol/acetone was added in a solution consisting of 0.5 g Fe3Ni6S8 in 4 mL H2O and 15 g of methanol, to reach the desired amount of binder. The ink was homogenously spray-coated with an Iwata SBS airbrush on a 8.5x8.5 cm carbon cloth or carbon felt that was heated on a hot plate to 100 °C. No further treatment steps were performed in the case of PVDF-containing inks.
The resulting catalyst coated sheets were cut into circular electrodes of a diameter of 40 mm with the help of an iron punch. The catalytic loading was determined through the weight diffrence between the non-coated substrates and dried coated electrodes.

NMR Analysis
The

Quantification of generated hydrogen via GC-MS
Quantification of the generated hydrogen during ECH in zero-gap electrolyzers was performed with the help of an online Shimdazu QP2020 GC-MS equipped with a Supelco Carboxen 1010 Plot column. The catholyte reservoir was completely sealed with Ar flowing through at a flow rate of 10 ml min -1 . Gas-samples were taken every 20 min for a total of 2 h.

H2 Quantification during electrolysis in the zero-gap electrolyzer
Compared to our experiments under ambient conditions, during the H2 quantification the total FE for the ECH decreases from ca 80 % to 55 %. This decrease could be attributed to the operation of the cell under the slight differential pressure of 40 mbar, which could have affected the fluid dynamics of the flow during electrolysis.
Moreover, from our control experiment we were able to obtain a total faradaic efficiency of 90 % including the generated H2, and ECH products in the anolyte and catholyte. Since our NMR analysis show no signs of decomposition or side-products, we are confident that the remaining percentages could be attributed to small leaks within the electrolyzer or gas-sampling loop, since our ECH electrolyzer has not been yet optimized towards a completely leakfree operation for gaseous products during electrolysis.

Product quantification for the zero-gap experiments
The faradaic efficiency of MBE and MBA was calculated with equation 1 (with np as amount of product p in mol, z as number of transferred electrons (z = 2 for MBE, z = 4 for MBA), F as Faraday constant (96485 A s mol -1 ), i as applied absolute current in A and t as reaction time (7200 s): The amount of product np was calculated with equation 2 (with I as the integral of the product peak (6H) in the 1 H-NMR spectrum normalized to the peak integral of the internal standard (4H), and nSt as the amount of internal standard in the NMR sample (1.25 µmol)): The yield was determined with equation 3: FE and yield error bars were calculated from the results of duplicates.
The potential values have been calculated by determining the arithmetic mean of the respective potential data for all data points of the last 10 min of the experiment for both of the two measurements. The arithmetic mean and the standard deviation of the resulting values was determined as the final data point of the respective experiment.  Figure S10. SEM Analysis of the catalyst coated PTLs at magnifications of x100, x500, x1000. Figure S11. EDX analysis of a H23 coated PTL with 5 mg cm -2 of catalytic loading.     Regarding the comparison with the current state-of-the-art systems, it is important to notice that most thermocatalytic reactors show a sharp increase in MBA generation after reaching a peak for the MBE conversion. Despite this finding, we specifically selected these peak values for the comparison with our electrochemical route.

Hydrogenation of MBY: Conversion factors & Reaction rates
The reaction rate (in h -1 ) was calculated according to Eq. 4: where $%& is the concentration of MBY at the start of reaction, the conversion, is the reservoit volume, while and -./012"3 correspond to the total time of electrolysis and moles of catalyst used.
At catalytic loading of 5 mg cm -2 was used, amounting to 62.85 mg of catalyst. For the used Fe3Ni6S8 catalyst this corresponds to 8.108·10 -5 mol, with the time of electrolysis being 2 h.
in which n 89:,;<=>=?@ is the amount of MBY at the start of the reaction (0.05 mol) and $%&,A,7 the detected amount of MBY in the catholyte at the end of the electrolytic reaction. With being total time of electrolysis in seconds, and -./012"3 and moles of used catalyst.

Long-term ECH experiments with pentlandite-based electrodes
To assess the catalytic stability of our Ni6 catalyst, the ECH was performed for 100h at 160 mA cm -2 with the optimal cathode parameters. The electrolyte volume was set to 938 L to enable a theoretical conversion to MBE of 100% in 25h and ensure necessary laboratory safety precautions. The catholyte and anolyte were exchanged with fresh solutions after every 25h until the 100h mark.