ELECTROCHEMICAL CONVERSION

Abstract

The present disclosure provides methods, compositions, devices, systems and uses that pertain to the electrochemical reduction of CO.sub.2 to CO. The application presents a class of electrodes, incorporating molecular catalysts in nanostructures, for robust and efficient electrochemical systems, specifically, selective and robust hybrid electrodes, by incorporating a rhenium (Re) catalyst into the structure of highly porous heterogeneous materials. These electrodes can be scaled up to desired manufacturing dimensions due to their robust nature and methods of preparation.

Claims

1. A composition comprising a rhenium catalyst and a carbon support wherein: the rhenium catalyst has a formula of Re(4,4′-R-2,2′-bipyridine)(CO).sub.3X; R is an electron donating group or an electron withdrawing group; X is a halogen, acetonitrile, CH.sub.3CN(OTf), or Py(OTf); and wherein the rhenium catalyst is dispersed on a surface of the carbon support.

2. The composition of claim 1, wherein the carbon support is multi-walled carbon nanotubes.

3. The composition of claim 1, wherein X is a halogen.

4. The composition of claim 1, wherein X is chloro.

5. The composition of claim 1, wherein R is an electron donating group.

6. The composition of claim 1, wherein the rhenium catalyst is Re(tBu-bpy)(CO).sub.3Cl.

7. The composition of claim 1, wherein R is an electron withdrawing group.

8. The composition of claim 1, wherein the composition is characterized by a current density of at least about 4 mA/cm.sup.2.

9. The composition of claim 1, wherein the composition is characterized by a current density of about 4 mA/cm.sup.2.

10. The composition of claim 1, wherein the composition is characterized by a turnover number (TON) greater than about 5600 and a turnover frequency (TOF) greater than about 1.6 s.sup.−1.

11. A method for electrocatalytically reducing CO.sub.2 to CO, comprising: contacting an electrode with CO.sub.2; wherein the electrode is in an aqueous solution having a pH of at least 4, comprising an electrolyte; wherein the electrode comprises the composition of claim 1; and wherein the method is performed at a temperature of at least about 5° C.

12-18. (canceled)

19. The method of claim 11, wherein the selectivity for CO over H.sub.2 is at least about 99%.

20. The method of claim 11, wherein the selectivity for CO over H.sub.2 is from about 30% to about 100%.

21. The method of claim 11, wherein the method is characterized by a Faradaic efficiency of at least about 99%.

22. The method of claim 11, wherein the electrolyte comprises KHCO.sub.3.

23. The method of claim 11, wherein the method is performed at a temperature of from about 5° C. to about 35° C.

24. (canceled)

25. The method of claim 11, wherein the pH of the aqueous solution is from about 6 to about 8.

26-30. (canceled)

31. A process for preparing an electrode, comprising: suspending a rhenium catalyst, a carbon support, and carbon nanofiber in ethanol to form a suspension; sonicating the suspension; drop-casting the suspension onto a glassy carbon plate to form a drop-casted glassy carbon plate; drying the drop-casted glassy carbon plate at a temperature from about 100° C. to about 180° C. for about 0.5 to about 24 hours; and wherein the rhenium catalyst is the rhenium catalyst according to claim 1.

32-39. (canceled)

40. The process of claim 31, wherein the suspension is drop-casted at a temperature from about 40° C. to about 80° C.

41. (canceled)

42. The process of claim 31, wherein the drop-casted glassy carbon plate is dried at a temperature of about 150° C. for about 1 hour.

43. (canceled)

Description

DESCRIPTION OF DRAWINGS

[0072] FIG. 1 is a schematic of a three-electrode cell configuration.

[0073] FIG. 2 is an image of an exemplary electrochemical cell in operation.

[0074] FIG. 3 is an image of the electrode surface.

[0075] FIG. 4A is an X-ray photoelectron spectrum (XPS) of Re 4f peak of freshly prepared Re-tBu/MWCNT and Re-tBu/MWCNT after 1 h controlled potential electrolysis (CPE) experiments. FIG. 4B is an XPS spectrum of N 1s peak of freshly prepared Re-tBu/MWCNT and Re-tBu/MWCNT after 1 h CPE experiments. FIG. 4C is an XPS spectrum of Cl 2p peak of freshly prepared Re-tBu/MWCNT and Re-tBu/MWCNT after 1 h CPE experiments.

[0076] FIG. 5 depicts transmission electron microscopy (TEM) images of Re(tBu-bpy)/MWCNT at different magnifications.

[0077] FIG. 6 depicts a cyclic voltammetry (CV) spectrum of Re(tBu-bpy)/MWCNT under N.sub.2 (lower curve) and CO.sub.2 (upper curve) in 0.5 M KHCO.sub.3 taken prior to the CPE experiment, using a scan rate of 100 mV/sA.

[0078] FIG. 7 depicts CPE experiments of Re(tBu-bpy)/MWCNT electrodes in CO.sub.2 saturated 0.5 M KHCO.sub.3 at −0.56 V vs. RHE for electrodes 1-5.

[0079] FIG. 8 is a plot showing the product distribution between CO and H.sub.2 measured as a function of time for a 7 h CPE experiment with a Re(tBu-bpy)/MWCNT (electrode 4) at −0.56 V vs. RHE in 0.5 M KHCO.sub.3.

[0080] FIG. 9 is depicts a CV plot of Re(tBu-bpy)/MWCNT at 25 mV/s in 0.5 M KHCO.sub.3 under CO.sub.2 atmosphere using electrode 4 as a working electrode, Ag/AgCl as a reference, and Pt as a counter electrode.

[0081] FIG. 10 depicts catalytic Tafel plots for Re(tBu-bpy)/MWCNT, CoPc-CN, CoPc-P4VP, COF-367-Co, and Mn—MeCN.

[0082] FIG. 11A depicts a high resolution spectrum of the Re 4f peak on bare GCE (carbon fabric). FIG. 11B depicts a high resolution spectrum of the N 1s peak on bare GCE. FIG. 11C depicts a high resolution spectrum of the Cl 2p peak on bare GCE.

[0083] FIG. 12 depicts a survey XPS for Re (tBu-bpy)/MWCNT. The top line on the left is before CPE, and the bottom line on the left is after CPE.

[0084] FIG. 13A depicts a scanning transmission electron microscopy (STEM) map of a Re-tBu/MWCNT material. FIG. 13B depicts energy-dispersive X-ray spectroscopy (EDS) maps of rhenium (top left), carbon (top right), chlorine (bottom left), and nitrogen (bottom right).

[0085] FIG. 14 depicts a CV spectrum of Re(tBu-bpy)/MWCNT electrodes 4 and 5 in 0.5 M KHCO.sub.3 under CO.sub.2.

[0086] FIG. 15A depicts a CV of control Re(tBu-bpy)/GCE under N.sub.2 and under CO.sub.2.

[0087] FIG. 15B depicts a CPE of control Re(tBu-bpy)/GCE under CO.sub.2 at −0.56 V vs. RHE.

[0088] FIG. 16 depicts a CPE of blank MWCNT/GCE under CO.sub.2 at −0.56 V vs. RHE.

[0089] FIG. 17A depicts a CV of Re(tBu-bpy)/CNF (carbon nanofiber) under N.sub.2 and under CO.sub.2. FIG. 17B depicts a CPE of Re(tBu-bpy)/CNF at −0.56 V vs. RHE under CO.sub.2 atmosphere.

[0090] FIG. 18 depicts a plot of CPE current vs. electroactive rhenium.

[0091] FIG. 19 depicts a CPE of 0.1 mM Re(tBu-bpy)(CO).sub.3Cl at −2.1 V vs. Fc.sup.+/0 in CO.sub.2 saturated MeCN/5% H.sub.2O solution.

[0092] FIG. 20 depicts a CV of Re(tBu-bpy)/MWCNT in CO.sub.2 saturated 0.1 M KHCO.sub.3.

[0093] FIG. 21 depicts a CPE of electrode 4 at −0.46, −0.56, −0.61 V vs. RHE in CO.sub.2 saturated 0.5 M KHCO.sub.3.

[0094] FIG. 22A depicts an N.sub.2 adsorption of a Re-loaded MWCNT/CNF composite in linear scale.

[0095] FIG. 22B depicts an N.sub.2 adsorption of a Re-loaded MWCNT/CNF composite in log scale.

[0096] FIG. 22C depicts density functional theory (DFT) pore size distributions of samples calculated from N.sub.2 isotherms.

[0097] FIG. 23 depicts the .sup.1H NMR spectrum of Re(tBu-bpy)MWCNT electrode soaked in CD.sub.3CN.

[0098] FIG. 24 depicts an infrared (IR) spectrum of Re(tBu-bpy)(CO).sub.3Cl and Re(tBu-bpy)/MWCNT.

[0099] FIG. 25A depicts the calibration IR spectra of Re(tBu-bpy)(CO).sub.3Cl in a KBr pellet.

[0100] FIG. 25B depicts IR spectra of Re-(tBu-bpy)/MWCNT in KBr pellet collected from electrode 2, electrode 3, and electrode 4.

[0101] FIG. 26 depicts CPE of Electrode 4 under N.sub.2 vs. CO.sub.2. at −0.56 V vs. RHE in 0.5 M KHCO.sub.3.

DETAILED DESCRIPTION

[0102] The present application provides a rhenium catalyst dispersed on multi-walled carbon nanotubes (MWCNTs) that can be used as an electrocatalyst for converting CO.sub.2 to CO. The rhenium catalyst can be, for example, Re(4,4′-tBu-2,2′-bpy)(CO).sub.3Cl. FIG. 1 is a schematic of a three-electrode cell configuration that the electrocatalyst can be used in. When used as an electrocatalyst, Re(tBu-bpy)/MWCNT electrodes can exhibit one or more of a Faradaic efficiency for CO FECO=99% and less than 1% H.sub.2, low operating potential (−0.58 V vs. RHE), the ability to sustain constant CO production for more than 7 h, and can work in water with a supporting electrolye (e.g., 0.5 M KHCO.sub.3).

[0103] The electrodes provided herein (e.g., Re(tBu-bpy)/MWCNT) can display excellent activity and selectivity in the electrochemical reduction of CO.sub.2 to CO. These electrodes operated at −0.58 V vs. RHE in 0.5 M KHCO.sub.3 with 99% selectivity for CO and only trace quantities of H.sub.2.

[0104] Some embodiments provide a composition comprising a rhenium catalyst and multi-walled carbon nanotubes; wherein: the rhenium catalyst has the formula Re(4,4′-R-2,2′-bipyridine)(CO).sub.3X; R is an electron donating group or an electron withdrawing group; X is a halogen, acetonitrile, CH.sub.3CN(OTf), or Py(OTf); and wherein the rhenium catalyst is dispersed on the surface of the multi-walled carbon nanotubes.

[0105] In some embodiments, the diameter of the nanotubes is about 5 nm to about 25 nm. For example, the diameter of the nanotubes is about 5 nm to about 10 nm, about 10 nm to about 15 nm, about 15 nm to about 20 nm, about 20 nm to about 25 nm, about 10 nm to about 20 nm, about 12 nm to about 18 nm, or about 15 nm.

[0106] In some embodiments, the loading of the transition metal catalyst (e.g., the rhenium catalyst), is from about 0.2 wt % to about 50 wt %. For example, from about 0.4 wt % to about 25 wt %, from about 2.5 wt % to about 30 wt %, from about 5 wt % to about 25 wt %, from about 10 wt % to about 23 wt %, from about 14 wt % to about 20 wt %, about 0.42 wt %, about 2.5 wt %, about 13.9 wt %, about 23.1 wt %, or about 25.2 wt %.

[0107] In some embodiments, X is a halogen, such as fluorine, chlorine, bromine, or iodine. In some embodiments, X is X is chloro.

[0108] In some embodiments, R is an electron donating group. In some embodiments, R is a C.sub.1-C.sub.10 alkyl group, for example, an isopropyl, t-butyl, or neopentyl group. In other embodiments, R is an electron withdrawing group. For example, R is F, Cl, CF.sub.3, —C(O)C.sub.1-C.sub.4 alkyl (e.g., acetyl), or —C(O)OC.sub.1-C.sub.4 alkyl (e.g., —C(O)OMe or —C(O)OEt).

[0109] In some embodiments, the rhenium catalyst is Re(tBu-bpy)(CO).sub.3Cl.

[0110] In some embodiments, the concentration of rhenium on the surface of the carbon support as measured by survey X-ray photoelectron spectroscopy is about 0.6 wt % to about 1.4 wt %. For example, about 0.8 wt % to about 1.2 wt %, about 0.9 wt % to about 1.1 wt %, about 1 wt %, or about 0.98 wt %.

[0111] In some embodiments, the concentration of nitrogen on the surface of the carbon support as measured by survey X-ray photoelectron spectroscopy is about 0.5 wt % to about 3.5 wt %. For example, about 1 wt % to about 3 wt %, about 1.5 wt % to about 2.5 wt %, about 1.7 wt % to about 2.3 wt %, about 2 wt %, or about 1.93 wt %.

[0112] In some embodiments, the concentration of chlorine on the surface of the carbon support as measured by survey X-ray photoelectron spectroscopy is about 0.6 wt % to about 1.4 wt %. For example, about 0.8 wt % to about 1.2 wt %, about 0.9 wt % to about 1.1 wt %, about 1 wt %, or about 0.98 wt %.

[0113] In some embodiments, the composition is characterized by a current density of at least about 0.1 mA/cm.sup.2. In some embodiments, the composition is characterized by a current density of at least about 1 mA/cm.sup.2. In some embodiments, the composition is characterized by a current density of from about 1 mA/cm.sup.2 to about 4 mA/cm.sup.2, from about 4 mA/cm.sup.2 to about 10 mA/cm.sup.2, from about 4 mA/cm.sup.2 to about 6 mA/cm.sup.2, from about 6 mA/cm.sup.2 to about 8 mA/cm.sup.2, or from about 8 mA/cm.sup.2 to about 10 mA/cm.sup.2, about 1 mA/cm.sup.2, about 1.3 mA/cm.sup.2, about 3.1 mA/cm.sup.2, or about 4 mA/cm.sup.2. In some embodiments, the composition is characterized by a current density of at least about 4 mA/cm.sup.2. In some embodiments, the composition is characterized by a current density of about 4 mA/cm.sup.2, about 5 mA/cm.sup.2, about 6 mA/cm.sup.2, about 7 mA/cm.sup.2, about 8 mA/cm.sup.2, about 9 mA/cm.sup.2, or about 10 mA/cm.sup.2. For example, composition is characterized by a current density of about 4 mA/cm.sup.2.

[0114] In some embodiments, the composition is characterized by a TON greater than about 4000. For example, the composition is characterized by a TON greater than about 4500, greater than about 4800, greater than about 5000, greater than about 5200, greater than about 5400, greater than about 5500, greater than about 5600, greater than about 5650, greater than about 5700, or about 5620, For example, the composition is characterized by a TON greater than about 5600.

[0115] In some embodiments, the composition is characterized by a TOF greater than about 1.0 s.sup.−1. For example, the composition is characterized by a TOF greater than about 1.0 s.sup.−1, greater than about 1.1 s.sup.−1, greater than about 1.3 s.sup.−1, greater than about 1.6 s.sup.−1, greater than about 2.0 s.sup.−1, greater than about 3.0 s.sup.−1, greater than about 4.0 s.sup.−1, greater than about 5.0 s.sup.−1, greater than about 10.0 s.sup.−1, greater than about 1 m.sup.−1, greater than about 10 m.sup.−1, greater than about 1 h.sup.−1, greater than about 6 h.sup.−1, greater than about 12 h.sup.−1, greater than about 50 h.sup.−1, greater than about 100 h.sup.−1, greater than about 200 h.sup.−1, greater than about 300 h.sup.−1, about 297 h.sup.−1, about 12 h.sup.−1, or about 1.6 s.sup.−1. For example, the composition is characterized by a TOF greater than about 1.6 s.sup.−1.

[0116] In some embodiments, the composition is characterized by a TON greater than about 5600 and a TOF greater than about 1.6 s.sup.−1.

[0117] Some embodiments provide a method for electrocatalytically reducing CO.sub.2 to CO, comprising: contacting an electrode as provided herein with CO.sub.2; wherein the electrode is in an aqueous solution having a pH of at least about 4; and wherein the method is performed at a temperature of at least about 5° C.

[0118] Some embodiments provide a method for electrocatalytically reducing CO.sub.2 to CO, comprising: contacting an electrode with CO.sub.2; wherein the electrode is in an aqueous solution having a pH from about 6 to about 8, comprising an electrolyte; wherein the electrode comprises multi-walled carbon nanotubes and the rhenium catalyst has the formula Re(4,4′-R-2,2′-bipyridine)(CO).sub.3X; R is an electron donating group or an electron withdrawing group; X is a halogen, acetonitrile, CH.sub.3CN(OTf), or Py(OTf); wherein the rhenium catalyst is dispersed on the surface of the multi-walled carbon nanotubes; and wherein the method is performed at a temperature of from about 5° C. to about 35° C.

[0119] In some embodiments, the selectivity for CO over H.sub.2 is from about 30% to about 100%. In some embodiments, the selectivity for CO over H.sub.2 is at least about 99%. In some embodiments, the Faradaic efficiency is at least about 99%. For example, the Faradaic efficiency is 100%.

[0120] In some embodiments, the electrolyte comprises KHCO.sub.3, HCl, NaOH, K.sub.2SO.sub.4 CH.sub.3COOH, H.sub.2CO.sub.3, NH.sub.4OH, and H.sub.2S. For example, the electrolyte comprises KHCO.sub.3.

[0121] In some embodiments, the concentration of the electrolyte is from about 0.01M to about 0.5M. For example, the concentration of the electrolyte is from about 0.05M to about 0.2M, 0.08M to about 0.13M, or about 0.1M.

[0122] For example, in some embodiments, the method is performed at a temperature of about 5° C. to about 30° C., about 5° C. to about 20° C., about 10° C. to about 25° C., about 15° C. to about 30° C., about 20° C. to about 35° C., or any value in between. In some embodiments, the method is performed at a temperature of from about 15° C. to about 25° C.

[0123] In some embodiments, the pH of the aqueous solution is from about 4 to about 10. In certain embodiments, the pH of the aqueous solution is from about 4 to about 6, from about 5.5 to about 9, from about 6 to about 8, from about 7.3 to about 10, or from about 6.5 to about 7.5. For example, the pH of the aqueous solution is about 7.3.

[0124] In some embodiments, the applied potential is from about −0.3 V to about −0.8 V. For example, the applied potential is from about −0.4 V to about −0.7 V, from about −0.5 V to about −0.6 V, or about −0.56 V.

[0125] In some embodiments, the method is performed in a low gravity environment (e.g., an environment where the force of gravity is less than that found on Earth). For example, the method is performed in Earth's thermosphere, Earth's exosphere, interplanetary space, on Earth's moon, or Mars. For example, the method is performed on Mars.

[0126] In some embodiments, the method further comprises producing O.sub.2 from the CO.sub.2.

[0127] Some embodiments provide a process for preparing an electrode, comprising:

[0128] suspending a rhenium catalyst, a carbon support (e.g., multi-walled carbon nanotubes), and carbon nanofiber in ethanol; sonicating the suspension; drop-casting the suspension onto a glassy carbon plate; drying the drop-casted glassy carbon plate at a temperature from about 100° C. to about 180° C. for about 0.5 to about 24 hours; and wherein the rhenium catalyst has the formula Re(4,4′-R-2,2′-bipyridine)(CO).sub.3X; R is an electron donating group or an electron withdrawing group; and X is a halogen, acetonitrile, CH.sub.3CN(OTf), or Py(OTf).

[0129] Some embodiments provide a process for preparing an electrode, comprising: suspending a transition metal catalyst, multi-walled carbon nanotubes, and carbon nanofiber in ethanol; sonicating the suspension; drop-casting the suspension onto a glassy carbon plate at a temperature from about 40° C. to about 80° C.; drying the drop-casted glassy carbon plate at a temperature from about 100° C. to about 180° C. for about 0.5 to about 24 hours; and wherein the transition metal catalyst has the formula M(4,4′-R-2,2′-bipyridine)(CO).sub.3X; R is an electron donating group or an electron withdrawing group; M is a transition metal; and X is a halogen, acetonitrile, CH.sub.3CN(OTf), or Py(OTf).

[0130] In some embodiments, X is a halogen, such as fluorine, chlorine, bromine, or iodine. In some embodiments, X is X is chloro.

[0131] In some embodiments, R is an electron donating group. In some embodiments, R is a C1-C10 alkyl group, for example, an isopropyl, t-butyl, or neopentyl group. In other embodiments, R is an electron withdrawing group. For example, R is F, Cl, CF.sub.3, —C(O)C.sub.1-C.sub.4 alkyl (e.g., acetyl), or —C(O)OC.sub.1-C.sub.4 alkyl (e.g., —C(O)OMe or —C(O)OEt).

[0132] In some embodiments, the transition metal is a Group III transition metal. In other embodiments, the transition metal is a Group IV transition metal. In still other embodiments, the transition metal is a Group V transition metal. In some embodiments, the transition metal is a Group VI transition metal. In other embodiments, the transition metal is a Group VII transition metal. In still other embodiments, the transition metal is a Group VIII transition metal. In some embodiments, the transition metal is a Group IX transition metal. In other embodiments, the transition metal is a Group X transition metal. In still other embodiments, the transition metal is a Group XI transition metal. In some embodiments, the transition metal is a Group XII transition metal.

[0133] In some embodiments, the transition metal is a Period 4 transition metal. In other embodiments, the transition metal is a Period 5 transition metal. In still other embodiments, the transition metal is a Period 6 transition metal.

[0134] In some embodiments, the transition metal is selected from iron, nickel, copper, palladium, and platinum.

[0135] Some embodiments provide a process for preparing an electrode, comprising:

[0136] suspending a rhenium catalyst, multi-walled carbon nanotubes, and carbon nanofiber in ethanol; sonicating the suspension; drop-casting the suspension onto a glassy carbon plate at a temperature from about 40° C. to about 80° C.; drying the drop-casted glassy carbon plate at a temperature from about 100° C. to about 180° C. for about 0.5 to about 24 hours; and wherein the rhenium catalyst has the formula Re(4,4′-R-2,2′-bipyridine)(CO).sub.3X; R is an electron donating group or an electron withdrawing group; and X is a halogen, acetonitrile, CH.sub.3CN(Otf), or Py(OTf).

[0137] In some embodiments, after sonicating the suspension, water is added. In some of these embodiments, after adding water, the suspension is sonicated further.

[0138] In some embodiments, the suspension is drop-casted at a temperature from about 30° C. to about 120° C. For example, the suspension is drop-casted at a temperature from about 30° C. to about 60° C., from about 40° C. to about 80° C., from about 60° C. to about 120° C., from about 50° C. to about 70° C., from about 55° C. to about 65° C., or from about 57° C. to about 63° C. In some embodiments, the suspension is drop-casted at a temperature of about 60° C.

[0139] In some embodiments, the suspension is drop-casted glassy carbon plate is dried at a temperature of about 150° C. for about 1 hour.

[0140] In some embodiments, the process is performed in a low gravity environment (e.g., an environment where the force of gravity is less than that found on Earth).

[0141] Some embodiments provide a method for electrocatalytically reducing CO.sub.2 to CO, comprising: contacting an electrode with CO.sub.2; wherein the electrode is in an aqueous solution having a pH from about 6 to about 8, comprising an electrolyte;

[0142] wherein the electrode comprises multi-walled carbon nanotubes and a transition metal catalyst having formula M(4,4′-R-2,2′-bipyridine)(CO).sub.3X; R is an electron donating group or an electron withdrawing group; X is a halogen, acetonitrile, CH.sub.3CN(OTf), or Py(OTf); M is a transition metal; and wherein the transition metal catalyst is dispersed on the surface of the multi-walled carbon nanotubes; and wherein the method is performed at a temperature of from about 5° C. to about 35° C.

[0143] A number of embodiments of the invention have been described.

Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention. Accordingly, other embodiments are within the scope of the following claims.

EXAMPLES

Materials and Methods

[0144] Materials: Chemicals were purchased from Sigma Aldrich and Fisher Scientific and used without further purifications. Multi-walled carbon nanotubes (MWCNT) (>98% carbon basis, O.D.×L6-13 nm×2.5-20 μm) and graphitized carbon nanofiber (CNF) (iron-free, composed of conical platelets, D×L 100 nm×20-200 μm) were purchased from Sigma Aldrich. Glassy carbon plates (SA-3, 100×100×t3 mm) were purchased from Tokai Carbon and pre-cut to the area 1×2 cm.sup.2.

[0145] Preparation of glassy carbon electrodes. Tokai glassy carbon plates were polished with an alumina slurry (0.05 μm) with subsequent sonication in water, methanol and acetone. Polished glassy carbon plates were connected to a copper wire using an alligator clip and were covered with an epoxy so that the exposed working area of glassy carbon surface was equal to 1×1 cm.sup.2.

Preparation and Characterization of Re(tBu-bpy)/MWCNT Electrodes.

[0146] Re(4,4′-tBu-2,2′-bpy)(CO).sub.3Cl was synthesized according to a previously published procedure. See Smieja J. M.; Kubiak C. P., Re(bipy-tBu)(CO).sub.3Cl—Improved Catalytic Activity for Reduction of Carbon Dioxide: IR-Spectroelectrochemical and Mechanistic Studies. Inorg. Chem. 2010, 49 (20), 9283-9289. A solution of Re(tBu-bpy)(CO).sub.3Cl (2.6 mM, 0.7 ml, 1 mg) in ethanol was added to 5 mg MWCNT and 2 mg CNF and sonicated for 5 minutes to disperse MWCNTs. DI water (0.3-0.5 ml) was added to the mixture to promote saturation and faster adsorption of Re(tBu-bpy)(CO).sub.3Cl by carbon nanotubes. The resulting yellow suspension was sonicated for 15 minutes until solution became colorless, which indicated of a complete adsorption of the catalyst by MWCNTs. This suspension was drop-casted onto a polished glassy carbon plate (1×1 cm.sup.2) at 60° C. using a hot plate. Electrodes were dried in the oven for 1 hour at 150° C. prior to CV and CPE experiments. Electrodes made of drop-casted Re(tBu-bpy)-MWCNTs dissolved in a water/ethanol mixture tend to form more even surfaces without visible precipitate at the edges, these electrodes crack less frequently in comparison to the electrodes drop-casted from pure ethanol solutions. Sonication in pure ethanol solution without addition of water resulted only in partial adsorption of the catalyst by MWCNTs and therefore precipitation of catalyst agglomerates occurred at the edges of the electrode. When Re(tBu-bpy)(CO).sub.3Cl catalyst was directly drop-casted onto MWCNTs, yellow agglomerates were seen at the surface of MWCNTs. These electrodes performed poorly in comparison to the electrodes prepared from sonicated Re(tBu-bpy)-MWCNTs ethanol/water solutions.

[0147] Electrochemistry. All cyclic voltammetry and controlled potential electrolysis (CPE) experiments were conducted using a Gamry Reference 600 potentiostat. A single compartment cell was used in all experiments with Re(tBu-bpy)/MWCNT/GCE as the working electrode (area 1×1 cm.sup.2), A Pt wire spool was used as a counter electrode and Ag/AgCl as a reference electrode. Electrochemical CO.sub.2 reduction experiments were performed on a Schlenk line using N.sub.2 or CO.sub.2 gas. Typically, a 0.5 M KHCO.sub.3 solution in DI water was used as the supporting electrolyte. All experiments were conducted with the compensation for iR drop. CPE experiments were performed in a single compartment Gamry cell, with modified carbon electrodes as working, Ag/AgCl as reference and Pt wire spool separated from the working electrode with a glass frit as the counter electrode. Products were monitored using gas chromatograph (Hewlett-Packard 7890A) with a molecular sieves column and helium as carrier gas. Gas analysis was performed using 1 ml gas-tight syringe with the gas sample taken from the headspace of the CPE cell. All potentials were measured versus Ag/AgCl electrode and converted to RHE according to the Nernst equation. Thus, the optimal operating potential of −0.56 V vs. RHE was determined as follows from cells employing Ag/AgCl reference electrodes:

E.sub.RHE=E.sub.Ag/AgCl+0.059 pH+E°.sub.Ag/AgCl

E.sub.RHE=(−1.20)+0.059×7.23+0.1976=−0.56 V

where E.sub.RHE is the converted potential vs RHE, E.sub.Ag/AgCl is the experimental potential measured against Ag/AgCl reference electrode, and E.sup.0.sub.Ag/AgCl vs NHE is 209 mV (3 M NaCl) at 25° C.

Example 1

X-Ray Photoelectron Spectroscopy (XPS)

[0148] High-resolution XPS analysis performed before and after 1 h Controlled Potential Electrolysis (CPE) experiments showed Re 4f (rhenium fourth shell f orbital), N is (nitrogen first shell s orbital), and Cl 2p (chlorine second shell p orbital) peak of freshly prepared Re-tBu/MWCNT and Re-tBu/MWCNT after 1 h CPE experiments (see FIG. 4). The Re 4f spectrum consisted of sharp peaks Re 4f.sub.5/2 and 4f.sub.7/2 at 44.3 and 41.8 eV respectively, with full widths at half-maximum (fwhm) of 1.09, 1.09 eV respectively, indicating a homogeneous environment around the rhenium center (FIG. 4A). No peaks were observed at 40.3, 42.7 eV expected for Re.sup.0 nanoparticles. The N is spectra consist of a single sharp peak at 400.29 eV (fwhm 0.98 eV) for the sample before CPE and slightly shifted N 1s peak at 400.35 eV (fwhm 0.98 eV) for the sample after 1 h CPE (FIG. 4B). The Cl 2p peaks were detected at 198.0 eV for Cl 2p.sub.1/2 and 199.7 eV for Cl 2p.sub.3/2 (FIG. 4C) for the sample before CPE and broader Cl 2 p peaks at 198.2 and 199.7 eV after CPE. XPS of blank Re(tBu-bpy)(CO).sub.3Cl on GCE without MWCNT presence revealed Re 4 f peaks at 44.0 and 41.6 eV. The Cl 2p peaks were detected at 198.7 eV for Cl 2p.sub.1/2 and 200.3 eV Cl 2p.sub.3/2). The N is spectra consist of a single sharp peak at 400.19 eV (fwhm 0.98 eV).

[0149] A series of XPS experiments were performed using a Kratos AXIS-SUPRA instrument equipped with a AL K-alpha monochromatic X-ray source operating at 225W. A pass energy of 160 eV was used for survey spectrum with 1 eV step size and a pass energy of 20 eV was used for details spectra, averaged over 5 scans, with 0.1 eV step size. Data were analyzed with CASA XPS software. All peaks were referenced to the is graphitic carbon peak (284.4 eV) in MWCNT. Peak fittings were performed with a Shirley-type background and Gaussian/Lorentzian line-shapes with 30% Gaussian shape. Survey XPS analysis supports the structural composition of Re(tBu-bpy)(CO).sub.3Cl for freshly prepared samples and for samples measured after CPE experiments. FIGS. 11A-11C depict XPS spectra of freshly prepared samples, wherein the peak integrations reveal atomic surface concentrations of 0.98%, 1.93% and 0.90% for Re (FIG. 11A), N (FIG. 11B), and Cl (FIG. 11C) respectively, which integrates to 1:1.97:0.92, consistent with expected Re/N/Cl ratio of 1:2:1 of Re(tBu-bpy)(CO).sub.3Cl. For samples measured after CPE the atomic surface concentrations are 0.98%, 1.96% and 0.83% for Re, N, and Cl respectively, with Re/N/Cl ratio of 1:2:0.85, indicating a diminished Cl concentration that can be explained by chloride dissociation. Only ca. 7.6% of the Cl content of the samples is lost after CPE. The diminished Cl peak can be explained by chloride dissociation. The dissociation of chloride is consistent with the mechanism for the homogeneous electrochemical reduction of CO.sub.2 to CO by Re(t-Bu)(CO).sub.3Cl, where the dissociation of chloride is the initiation step in the catalytic cycle, providing a vacant site for CO.sub.2 binding. Table 1 summarizes the survey XPS peak integration. FIG. 12 depicts the survey XPS for Re (tBu-bpy)/MWCNT. The top plot line in FIG. 12 is before CPE, and the bottom plot is after CPE.

TABLE-US-00001 TABLE 1 Survey XPS peaks integration. Re(tBu-bpy)(CO).sub.3Cl/MWCNT Re % N % C % O % Cl % before CPE 0.98 1.93 93.76 2.45 0.90 after CPE 0.98 1.96 93.63 2.64 0.83 Re(tBu-bpy)/GCE 1.08 2.07 74.49 21.38 0.99

Example 2

.SUP.1.H NMR Analysis

[0150] To analyze the composition of electrodes after CPE samples were soaked in deuterated acetonitrile and analyzed via .sup.1H NMR spectroscopy. Chemical shifts corresponding to bipyridine ligand were detected and matched with the spectra of Re(tBu-bpy)(CO).sub.3Cl. .sup.1H NMR (400 MHz, CD.sub.3CN): δ 1.45 (s, 18H, tBu), 7.64 (dd, 2H), 8.41 (d, 2H), 8.88 (d, 2H). FIG. 23 depicts the .sup.1H NMR spectrum of Re(tBu-bpy)MWCNT electrode soaked in CD.sub.3CN.

Example 3

IR Analysis

[0151] Solution IR. The IR experiments of the electrodes soaked in MeCN showed IR peaks u(CO) 2023, 1916, 1898 cm.sup.−1 and matched with the IR spectra of Re(tBu-bpy)(CO).sub.3Cl. FIG. 24 depicts an IR spectrum of Re(tBu-bpy)(CO).sub.3Cl (upper plot) dissolved in MeCN and Re(tBu-bpy)/MWCNT dissolved in MeCN (lower plot).

[0152] KBr pellet IR. A sample of Re(tBu-bpy)(CO).sub.3Cl of a known mass was pressed into a pellet with KBr salt and was analyzed using infrared spectroscopy. The IR spectra showed CO peaks at ν(CO) 2017, 1903, 1884 cm.sup.−1. These pellets with four known concentrations were used as a standard to plot a calibration line for the future measurements. FIG. 25A depicts the calibration IR spectra of Re(tBu-bpy)(CO).sub.3Cl in a KBr pellet; first (uppermost) plot at 1900 cm.sup.−1 corresponds to 0.350 mg Re-tBu, second plot at 1900 cm.sup.−1 corresponds to 0.175 mg Re-tBu, third plot at 1900 cm.sup.−1 corresponds to 0.066 mg Re-tBu, and fourth (lowest) plot at 1900 cm.sup.−1 corresponds to 0.044 mg Re-tBu. The solid samples of Re(tBu-bpy)/MWCNT with various Re(tBu-bpy)(CO).sub.3Cl loadings were pressed into KBr pellets to perform IR experiments of Re(tBu-bpy)(CO).sub.3Cl adsorbed on MWCNTs. These samples were prepared by mixing electrode materials scraped from the electrode surface with a surgical blade with dry KBr salt. The IR spectra of these samples matched with a standard and showed three CO stretches at 2019, 1900, 1884 cm.sup.−1. FIG. 25B depicts IR spectra of Re-(tBu-bpy)/MWCNT in KBr pellet collected from electrode 2 (lowest plot at 1900 cm.sup.−1), electrode 3 (middle plot at 1900 cm.sup.−1) and electrode 4 (highest plot at 1900 cm.sup.−1). Catalyst concentrations for three different Re(tBu-bpy)/MWCNT samples were calculated and are listed in Tables 2 and 3.

TABLE-US-00002 TABLE 2 IR data of Re(tBu-bpy)(CO).sub.3Cl in KBr pellet. Adsorption Extinction coefficient 1884 1903 2019 E 1884 E 1898 E 2019 Mass (mg) 1.556 1.478 1.467 19.142 18.182 18.047 0.350 0.837 0.811 0.821 20.593 19.954 20.200 0.175 0.429 0.420 0.458 21.110 20.667 22.537 0.088

TABLE-US-00003 TABLE 3 IR data of Re(tBu-bpy)(CO).sub.3Cl/MWCNT in KBr pellet. Adsorption Extinction coefficient 1884 1903 2019 E 1884 E 1898 E 2019 Mass (mg) 0.008 0.007 0.008 20.68 21.25 21.10 0.06 0.11 0.098 0.098 20.68 21.25 21.10 0.84 0.19 0.17 0.18 20.68 21.25 21.10 1.77

Example 4

Transmission Electron Microscopy (TEM)

[0153] Transmission electron microscopy (TEM) was conducted to study the structural morphology of the electrodes. FIG. 5 depicts TEM images of Re(tBu-bpy)/MWCNT at 25 kX magnification (electrode 3, first row left), 100 kX magnification (electrode 2, first row middle and electrode 3, first row right), 150 kX magnification (electrode 3, second row left), 280 kX magnification (electrode 3, second row middle), and 690 kX magnification (electrode 3, second row right). TEM revealed that Re(tBu-bpy)/MWCNT electrodes consist of hollow tubular nanotubes and conical nanofibers with the average diameter of 15 nm for nanotubes and 35 nm for nanofibers. The slightly wrinkled sidewalls of both carbon nanotubes and nanofibers suggest that the tubes are loaded with the catalyst that is homogeneously distributed over the surface of the nanotubes. The TEM at magnification of 100k×showed MWCNTs aggregated around CNFs which appear to add stability to the electrode material.

Example 5

Scanning Transmission Electron Microscopy (STEM)/Energy-Dispersive X-Ray Spectroscopy (EDS)

[0154] FIG. 13A depicts a scanning transmission electron microscopy (STEM) of a Re-tBu/MWCNT material and FIG. 13B depicts energy-dispersive X-ray spectra (EDS) of rhenium (upper left), carbon (upper right), chlorine (lower left), and nitrogen (lower right). FIGS. 13A-B revealed a homogeneous distribution of Re throughout the nanotube structures. The distribution of C and N elements matches that of the nanotube structures and overlaps due to close X-ray energies of C (0.277) and N (0.392). The EDS maps of Re and Cl show some response outside of the carbon nanotubes structures due to the instrumental noise that is typical at this magnification.

Example 6

Cyclic Voltammetry (CV)

[0155] The cyclic voltammetry (CV) experiments were performed using a three electrodes cell configuration, with platinum as a counterelectrode, Ag/AgCl as a reference and modified Re(tBu-bpy)/MWCNT as the working electrode at 100 mV/s. FIG. 2 is an image of an exemplary electrochemical cell in operation. FIG. 3 is an image of the electrode surface upon which CO bubbles have formed. CV experiments were performed in 0.5 M KHCO.sub.3 under N.sub.2 or CO.sub.2 atmosphere, unless stated otherwise. Electrodes with various Re(tBu-bpy) loadings were investigated through CV and CPE experiments and the results are summarized in Table 4. FIG. 6 depicts a CV spectrum of Re(tBu-bpy)/MWCNT under N.sub.2 (lower curve) and CO.sub.2 (upper curve) in 0.5 M KHCO.sub.3 taken prior to the CPE experiment, using a scan rate of 100 mV/sA. A significant increase in current is observed when the applied potential reached −0.56 V vs. RHE under CO.sub.2 in comparison to N.sub.2 atmosphere. Current densities of 30 mA/cm.sup.2 were achieved with electrodes containing 23.008 wt % of catalyst. The current density was found to decrease with decreasing catalyst loading and was less than 10 mA/cm.sup.2 for electrode 1. FIG. 14 depicts a CV spectrum of Re(tBu-bpy)/MWCNT electrodes 4 and 5 in 0.5 M KHCO.sub.3 under CO.sub.2; Re(tBu-bpy)/MWCNT electrodes were used as working electrode, Ag/AgCl as a reference electrode, and Pt as a counter electrode. The scan rate was 100 mV/s. The upper curve is 23.1 wt % Re-tBu and the lower curve is 25.2 wt % Re-tBu. Oversaturation of the carbon nanotube surface was found to occur when catalyst loading exceeded the concentration of 25 wt %, and this resulted in diminished current during CV and CPE experiments.

TABLE-US-00004 TABLE 4 Summary of electrocatalytic CO.sub.2 reduction using Re(tBu-bpy)/MWCNT electrodes. Re(tBu-bpy) charge [CO] FE FE TOF TON TOF electrode no. (wt %) (C) (μmol) (% CO) (% H.sub.2) (h.sup.−1) (CV) (CV) (s ) (mA/cm.sup.2) 1 0.42 3.7 16 84 15 297 2280 0.6 3.0 2 2.50 4.6 21 94 5 124 2703 0.8 3.3 3 13.9 11.1 57 99 1 38 16 1.4 3.1 4 23.1 14.4 73 99 1 27 619 1.6 4.0 5 25.2 7.6 33 99 1 12 4013 1.1 2.1 indicates data missing or illegible when filed

Example 7

Inductively Coupled Plasma—Mass Spectrometry (ICP-MS)

[0156] ICP-MS was performed on electrodes 1-5. The results of the analysis are shown in Table 5.

TABLE-US-00005 TABLE 5 Results of ICP-MS analysis of Re(tBu-bpy)/MWCNT. Re(tBu-bpy)/ Electrode MWCNT Re Re Re(tBu-bpy) Re(tBu-bpy) # mg mg wt % mg wt % 1 7.1 0.01 0.14 0.03 0.42 2 4.8 0.04 0.83 0.10 2.50 3 6.2 0.28 4.52 0.86 13.87 4 6.8 0.51 7.50 1.57 23.08 5 6.6 0.54 8.18 1.66 25.15

Example 8

Controlled Potential Electrolysis (CPE)

[0157] The composition, quantity, and rate of the product formation was investigated via controlled potential electrolysis (CPE) experiments. CPE experiments were performed in a three-neck cell (V=89 ml) at −0.56 V vs. RHE in 0.5 M KHCO.sub.3 solution (35-40 ml) saturated with CO.sub.2 gas. Under these conditions in the absence of the catalyst bare MWCNTs on a glassy carbon electrode produced H.sub.2 with 100% Faradaic efficiency and very low current densities (0.18 mA/cm.sup.2). When loaded with Re(tBu-bpy), these electrodes displayed high current densities which were found to be dependent on the catalyst loading. FIG. 7 depicts CPE experiments of Re(tBu-bpy)/MWCNT electrodes in CO.sub.2 saturated 0.5 M KHCO.sub.3 at −0.56 V vs. RHE. A current density of 1.0 mA/cm.sup.2 was achieved with electrode 1 (0.42 wt % Re-tBu, bottom plot), 1.3 mA/cm.sup.2 was achieved with electrode 2 (2.5 wt % Re-tBu, fourth plot from top), 3.1 mA/cm.sup.2 for electrode 3 (13.9 wt % Re-tBu, second plot), and reached the maximum of 3.8 mA/cm.sup.2 for electrode 4 (23.1 wt % Re-tBu, first (highest) plot). A decline in current density was observed for electrode 5, when catalyst loadings exceeded 25 wt % (third plot) suggesting that these electrodes are susceptible to a blocking effect at higher catalyst loadings. An electrode including 0% Re-tBu produced the fifth (lowest) plot. This was somewhat surprising since a Re catalyst loading of 25 wt % corresponds approximately to one Re atom per 170 carbon atoms in the hybrid material. The headspace was analyzed via gas chromatography after 1 h of electrolysis and revealed Faradaic efficiency (FE) of 99% for CO and less than 1% H.sub.2 for the electrodes 3, 4, and 5. The selectivity for CO decreased when the catalyst loadings were decreased and corresponded to FE(CO) of 97% for electrodes 7 (4.1 wt %) and 8 (6.7 wt %) and reached FE(CO) of 94% for electrode 2 and 6 (2.8 wt %) (see Tables 4 and 6). When the loading of the catalyst was decreased to 0.42% (electrode 1) the selectivity for CO decreased to 84%. The performance of electrode 4 was further investigated and CPE experiments were performed that proceeded for 7 hours at −0.56 V vs. RHE in 0.5 M KHCO.sub.3 solution saturated with CO.sub.2 gas. The headspace analysis via gas chromatography was performed every half hour. FIG. 8 shows the product distribution between CO (circular data points) and H.sub.2 (square data points) measured as a function of time for a 7 h CPE experiment with a Re(tBu-bpy)/MWCNT (electrode 4) at −0.56 V vs. RHE in 0.5 M KHCO.sub.3, revealing that the Faradaic efficiency of 99% for CO remained constant throughout the course of the experiment without deactivation of the catalyst. Table 7 shows CPE data for the 7 hour experiment in CO.sub.2-saturated 0.5 M KHCO.sub.3. FIG. 26 depicts CPE of Electrode 4 under N.sub.2 vs. CO.sub.2. at −0.56 V vs. RHE in 0.5 M KHCO.sub.3. Due to rapid CO formation and its low solubility in water, a profusion of CO bubbles was observed on the surface of the electrodes during CPE experiments.

TABLE-US-00006 TABLE 6 Summary of electrocatalytic CO.sub.2 reduction using Re(tBu-bpy)/MWCNT electrodes 6-8. Re(tBu-bpy) Charge [CO] FE FE TOF I Electrode # (wt %) (C) (μmol) (% CO) (% H.sub.2) (h.sup.−1) (mA/cm.sup.2) 6 (0.20 mg)* 2.8 7.3 35 94 6 101 2.0 7 (0.30 mg)* 4.1 8.7 44 97 3 83 2.4 8 (0.50 mg)* 6.7 10.2 51 97 3 60 2.9 *TOF calculated based on the total mass of the sample.

TABLE-US-00007 TABLE 7 CPE data for the 7-hour experiment in CO.sub.2 saturated 0.5M KHCO.sub.3. Gc area mol CO current efficiency % TON TOF (h−1) time (h) Coulombs H.sub.2 CO H.sub.2 CO H.sub.2 CO total Electrons mol Co/molCat 0.5 9.00 1.88 220.00 5.46E−07 4.60E−05 1.2% 99.4% 100.5% 9.33E−05 16.94 33.89 1.0 14.00 2.14 356.00 6.68E−07 7.20E−05 1.0% 99.2% 100.2% 1.45E−04 26.32 28.32 1.5 18.00 3.11 481.00 9.30E−07 9.30E−05 1.0% 99.9% 100.9% 1.87E−04 34.08 22.72 2.0 23.00 3.76 624.00 1.11E−06 1.18E−04 0.9% 99.3% 100.2% 2.38E−04 43.25 21.63 2.5 28.00 4.53 716.00 1.38E−06 1.42E−04 1.0% 97.7% 98.7% 2.90E−04 51.84 20.73 3.0 33.00 5.55 826.00 1.73E−06 1.67E−04 1.0% 97.7% 98.7% 3.42E−04 61.07 20.36 3.5 36.00 6.48 929.00 1.98E−06 1.84E−04 1.1% 98.6% 99.7% 3.73E−04 67.26 19.22 4.0 39.70 7.24 1025.00 2.21E−06 2.03E−04 1.1% 98.7% 99.8% 4.11E−04 74.21 18.55 4.5 43.30 8.12 1104.00 2.54E−06 2.23E−04 1.1% 99.5% 100.6% 4.49E−04 81.63 18.14 5.0 46.60 8.82 1218.00 2.70E−06 2.41E−04 1.1% 99.9% 101.0% 4.83E−04 88.18 17.64 5.5 49.70 9.75 1298.00 2.98E−06 2.57E−04 1.2% 99.8% 101.0% 5.15E−04 93.97 17.09 6.0 52.80 10.44 1365.00 3.19E−06 2.70E−04 1.2% 98.8% 100.0% 5.47E−04 98.82 16.47 6.5 55.70 11.48 1445.00 3.51E−06 2.86E−04 1.2% 99.1% 100.3% 5.77E−04 104.61 16.09 7.0 58.40 12.50 1523.00 3.82E−06 3.02E−04 1.2% 99.0% 100.2% 6.05E−04 110.26 15.75

Further Discussion of CV and CPE Results

[0158] The observed high CO:H.sub.2 selectivity of Re(tBu-bpy)/MWCNT electrodes is attributed to the high loadings of the Re(tBu-bpy) catalyst used in these experiments. At high concentration and surface coverage, selectivity for CO was almost 100% and remained constant throughout 7 hours of CPE. High surface coverage by the Re catalyst appears to promote CO.sub.2 reduction over native proton reduction that can occur at exposed carbon sites. When catalyst loadings were low (electrode 1, 0.7 wt %) hydrogen production was observed to occur with Faradaic efficiency for H.sub.2=15% and 84% for CO. When catalyst loadings were increased to 13.9 wt %, selectivity for H.sub.2 decreased to <1% with FE=99% for CO. To obtain maximum CO selectivity the Re(tBu-bpy) catalyst should be evenly dispersed onto the structure of MWCNTs to minimize carbon sites exposure. It is understood that minimization of carbon sites exposure suppresses hydrogen formation.

[0159] FIG. 15A depicts a CV of control Re(tBu-bpy)/GCE in 0.5 M KHCO.sub.3 under N.sub.2 (upper and lower plot at 0.0 potential) and under CO.sub.2 (middle two plots at 0.0 potential). FIG. 15B depicts a CPE of control Re(tBu-bpy)/GCE in 0.5 M KHCO.sub.3 under CO.sub.2 at −0.56 V vs. RHE. The FE (H.sub.2) 100%. The control experiments under a nitrogen atmosphere showed production of H.sub.2, and no CO formation during CPE experiments. Electrodes without MWCNTs with only Re(tBu-bpy) directly drop casted on glassy carbon surface showed almost no electrochemical activity toward CO.sub.2 reduction at the same conditions. FIG. 16 depicts a CPE of blank MWCNT/GCE in 0.5 M KHCO.sub.3 under CO.sub.2 at −0.56 V vs. RHE. The FE (H.sub.2)=100%. This bare MWCNT/GCE without Re(tBu-bpy) showed exclusive production of hydrogen with FE (H.sub.2)=100%. FIG. 17A depicts a CV of Re(tBu-bpy)/CNF in 0.5 M KHCO.sub.3 under N.sub.2 (two middle plot lines at −0.7 potential) and under CO.sub.2 (highest and lowest plot lines at −0.7 potential). FIG. 17B depicts a CPE of Re(tBu-bpy)/CNF at −0.56 V vs. RHE in 0.5 M KHCO.sub.3 under CO.sub.2 atmosphere. The electrodes without MWCNTs with only Re(tBu-bpy) directly drop casted on glassy carbon surface showed almost no electrochemical activity towards CO.sub.2 reduction under the same conditions. Control CV experiments with CNF demonstrated no change in current under CO.sub.2 in comparison to N.sub.2 with very low CPE current and hydrogen production with 100% Faradaic efficiency.

[0160] The TOF measured during CPE experiments and calculated per total concentration of catalyst were found to range from 178 h.sup.−1 to 12 h.sup.−1, depending on the catalyst loadings. It is important to note that the TOF values were calculated based on the total amount of Re(tBu-bpy) in the bulk material and are therefore the lower limit of the actual catalyst TOF values that should be calculated based on the amount of electroactive catalyst. The amount of electroactive catalyst can be obtained through integration of the area of a Re(tBu-bpy) CV. In acetonitrile Re(tBu-bpy)(CO).sub.3Cl displays two one electron reductions that were previously assigned to a bipyridine-based reduction followed by a metal based Re.sup.I/0 reduction. FIG. 9 depicts a CV plot of Re(tBu-bpy)/MWCNT at 25 mV/s in 0.5 M KHCO.sub.3 under CO.sub.2 atmosphere and used electrode 4 as a working electrode, Ag/AgCl as a reference, and Pt as a counter electrode. The scan rate was 25 mV/s. The first scan is the uppermost line at −0.4 V, the second scan is the line immediately below the first scan at −0.4V, and the third scan is the line immediately below the second scan at −0.4V. Re(tBu-bpy) redox peak was detected in 0.5 M KHCO.sub.3 at a scan rate of 25 mV/s at −0.38 V vs. RHE, prior to the catalytic wave, for freshly prepared electrodes, when the capacitive current is low. This feature was tentatively assigned to the bipyridine ligand-based redox process (first scan in FIG. 9). The CV peak shifts after the second scan and gets smaller after a third scan and completely diminishes after a fourth scan. The fact that the intensity of the peak decreases and completely vanishes with consecutive scans may indicate strong electronic coupling between MWCNTs under bias and the Re catalyst. The lack of Faradaic current in this case may be due to charge build up at the interface establishing an electric field gradient that changes the Fermi level of the electrode relative to the donor/acceptor states regardless of the applied potential. This eliminates the driving force for the outer-sphere electron transfer consistent with the double-layer theory for a molecule electronically coupled to the electrode. Jackson M. N.; Oh S.; Kaminsky C. J.; Chu S. B.; Zhang G.; Miller J. T.; Surendranath Y., Strong Electronic Coupling of Molecular Sites to Graphitic Electrodes via Pyrazine Conjugation. J. Am. Chem. Soc. 2018, 140 (3), 1004-1010. The quantity of electroactive Re(tBu-bpy) for electrode 4 obtained by integration of the first peak area was calculated to be 1.3×10.sup.−8 mol. This compares to a value of 1.8×10.sup.−8 for CoPc-CN, and 2×10.sup.−9 for COF-367-Co. Zhang X., Wu Z., Zhang X., Li L., Li Y., Xu H., Li X., Yu X., Zhang Z., Liang Y., Wang H.,, Highly Selective and Active CO.sub.2 Reduction Electro-Catalysts Based on Cobalt Phthalocyanine/carbon Nanotube Hybrid Structures. Nat. Commun. 2017, 8, 14675; Lin S., Diercks C. S., Zhang Y. B., Kornienko N., Nichols E. M., Zhao Y., Paris A. R., Kim D., Yang P., Yaghi O. M., Chang C. J., Covalent Organic Frameworks Comprising Cobalt Porphyrins for Catalytic CO.sub.2 Reduction in Water. Science 2015, 349 (6253), 1208-1213. Using an estimated value of electroactive Re in the Re(tBu-bpy)/MWCNT electrodes, a TON.sub.EA=5619 and TOF.sub.EA=1.6 s.sup.−1 were obtained. The amount of electroactive catalyst was determined for all electrodes in the manner described here and was found to range from 1-8% of the total catalyst loaded into the bulk electrode material. Table 8 provides the CPE current of electroactive species obtained from CV. Charge Q was calculated using the area under the curve per scan rate and moles of electroactive catalyst was calculated according to Equation: Γ=Q/nFA, where n is the number electrons (n=1), F is the Faraday constant, and A is the area of the electrode.

TABLE-US-00008 TABLE 8 Electroactive species obtained from CV and corresponding CPE current. Current Density Electrode # Mol.sub.EA (×10.sup.−8) (mA/cm.sup.2) 1 0.68 1.0 2 0.79 1.3 3 1.1 3.1 4 1.3 4.0 5 0.82 2.1

[0161] FIG. 18 depicts a plot of CPE current vs. electroactive rhenium, which shows that the amount of electroactive rhenium increased with increasing catalyst loadings until it reached a saturation level (electrode 5) which is attributed to the formation of agglomerates of inactive catalyst and physical blocking of the electrode surface. Thus, the amount of electroactive Re for electrode 5 decreased, which was also reflected in lower current densities for this electrode. The TOF.sub.EA values for all electrodes are summarized in Table 1. Although high surface coverage resulted in lower amounts of electroactive catalyst, the total coverage is necessary to suppress the competing hydrogen evolution reaction that occurs on the exposed sites of the carbon nanotubes.

[0162] The TOF for homogeneous CO.sub.2 reduction by 1 mM Re(tBu-bpy)(CO).sub.3Cl in acetonitrile in the presence of 3 M H.sub.2O corresponds to TOF=5.7 s.sup.−1 (average), with a maximum reported TOF of 2601 s.sup.−1 for 10 M H.sub.2O. Smieja J. M.; Sampson M. D.; Grice K. A.; Benson E. E.; Froehlich J. D.; Kubiak C. P., Manganese as a Substitute for Rhenium in CO.sub.2 Reduction Catalysts: The Importance of Acids. Inorg. Chem. 2013, 52, 2484-2491. The maximum effective loading for heterogeneous Re(tBu-bpy)/MWCNT was 23 wt %, which at the operating conditions corresponds to 0.1 mM Re(tBu-bpy), loadings 10 times lower than that for homogeneous catalysis. It is worth noting that these electrodes operate in aqueous media where the solubility of CO.sub.2 is almost 5 times lower than in acetonitrile. FIG. 19 depicts a CPE of 0.1 mM Re(tBu-bpy)(CO).sub.3Cl at −2.1 V vs. Fc.sup.+/0 in CO.sub.2 saturated MeCN/5% H.sub.2O solution; 0.1 M TBAPF.sub.6 was used as electrolyte; bare MWCNT/CNF/GCE was used as a working electrode; Ag/AgCl was used as a reference; and Pt was used as a counter electrode. The control CPE experiments depicted in FIG. 19 exhibited hydrogen production with FE of 87% and FE of 12% for CO. This was attributed to the limited diffusion and mass transport of the catalyst to the highly porous surface of the MWCNT. When incorporated into the structure of the MWCNT catalyst material no such limitations are observed because of the direct contact between the catalyst and the surface of the electrode.

[0163] Without addition of Bronsted acids, but in the presence of advantageous water, the determined overpotential (η) of Re(4,4′-tBu-bpy)(CO).sub.3Cl is equal to 0.856 V, considering CO.sub.2/CO equilibrium at −1.344 V vs. Fc.sup.+/0..sup.36 Since in aqueous solutions CO.sub.2/CO equilibrium corresponds to −0.11 V vs. RHE, (Chen Y.; Li C. W.; Kanan M. W., Aqueous CO.sub.2 Reduction at Very Low Overpotential on Oxide-Derived Au Nanoparticles. J. Am. Chem. Soc. 2012, 134, 19969-19972) the reduction potential of −0.56 V vs. RHE is equal to the overpotential of 0.45 V for CO evolution, which is 0.41 V lower than η for the homogeneous Re catalyst in acetonitrile.

[0164] Previous studies showed that reactivity and selectivity of Re-bpy in organic solvents depends on the acid strength (pKa). The acidity of 0.1 M KHCO.sub.3 solution saturated with CO.sub.2 corresponds to pH=6.81, and pH=7.23 for 0.5 M KHCO.sub.3. Chen C.; Zhang B.; Zhong Z.; Cheng Z., Selective Electrochemical CO.sub.2 Reduction Over Highly Porous Gold Films. J. Mater. Chem. A 2017, 5, 21955-21964. However, for the Re(tBu-bpy)/MWCNT system a trade-off exists between acidity of the solution and mass transport of CO.sub.2 to the active sites of the embedded catalyst. FIG. 20 depicts a CV of Re(tBu-bpy)/MWCNT in CO.sub.2 saturated 0.1 M KHCO.sub.3; electrodes 2, 3, 4 were used as working electrodes, Ag/AgCl was used as a reference, Pt was used as a counter electrode, and a scan rate of 100 mV/s was used. The uppermost curve at a potential of −0.7 is 23.1% Re-Bu, the middle curve at a potential of −0.7 is 13.9% Re-tBu, and the lowest curve at a potential of −0.7 is 25.2% Re-tBu. Thus, electrodes operated at −0.56 V vs. RHE in 0.1 M KHCO.sub.3 performed poorly in comparison to 0.5 M KHCO.sub.3 solution.

[0165] FIG. 21 depicts a CPE of electrode 4 at −0.46 V (bottom plot), −0.56 V (middle plot), and −0.61 V (top plot) vs. RHE in CO.sub.2 saturated 0.5 M KHCO.sub.3. The selectivity of the Re(tBu-bpy)/MWCNT is related to the applied potential. At −0.62 V vs. RHE the system produces CO with Faradaic efficiency of 97% vs. 3% for H.sub.2, but at −0.67 V vs. RHE, Faradaic efficiency decreased to 84% for CO and 15% for Hz. Thus the Re(tBu-bpy)(CO).sub.3Cl/MWCNT electrodes display overpotentials for CO and Hz evolution that favor CO evolution at less negative potentials than Hz evolution, the reverse of normal thermodynamic expectations for these reactions.

[0166] The comparison of the Re(tBu-bpy)/MWCNT catalyst to the best hybrid catalysts based on molecular catalysts on solid carbon supports are summarized in Table 9. Re(tBu-bpy)/MWCNT has a lower operating potential and highest Faradaic efficiency for CO in comparison to other catalysts.

TABLE-US-00009 TABLE 9 Comparison of existing molecular electrocatalysts on various carbon supports. Current Ecat. Density Charge TOF FE Catalyst (V vs. RHE) (mA/cm.sup.2) (C) (s.sup.−1) (CO) Re(tBu-bpy) −0.56 4.0 14.4 1.6 99 Mn-pyrrolyl −0.39 2.4 8.6 0.009 87 CoPc-CN −0.63 15 54 4.1 98 CoPc-P4VP −0.75 2.0 7.2 4.8 90 COF-367-Co −0.67 2.9 10.4 2.6 90

[0167] Catalytic Tafel plots were created to benchmark the catalytic system with the existing molecular catalysts and they represent the relationship between thermodynamic and kinetic parameters: η and TOF. Tafel plots were calculated based on log TOF multiplied by FE(CO) and plotted against the corresponding overpotentials. FIG. 10 depicts catalytic Tafel plots for Re(tBu-bpy)/MWCNT (electrode 4; second line from top at η=0.0), CoPc-CN (cobalt-2,3,7,8,12,13,17, 18-octacyano-phthalocyanine; third line from top at η=0.0), CoPc-P4VP (cobalt phthalocyanine poly-4-vinylpyridine at glassy carbon electrode at pH 4; lowest line at η=0.0), COF-367-Co (cobalt porphyrin in piphenyl-4,4′-dicarboxaldehyde covalent organic framework on carbon fabric; fourth line from top at η=0.0), and Mn—MeCN (Mn(4,4′-di(1H-pyrrolyl-3-propyl carbonate)-2,2′-bipyridine)(CO).sub.3MeCN].sup.+ (PF6).sup.−; uppermost line at η=0.0). The Tafel plot for this catalyst was calculated based on the reported TON and bulk concentration of the catalyst, under the assumption that TON and TOF should be higher if calculated per electroactive species. Re(tBu-bpy)/MWCNT displayed the best optimal characteristics of η and TOF in comparison to the other catalysts with overpotential of 0.45 and log TOF×FE(CO)=0.2.

Example 9

Gas Adsorption

[0168] Materials were characterized by N.sub.2 physisorption at 77 K using a Micromeritics ASAP 2020 after a 12 h activation under vacuum at 423 K. Brunauer-Emmet-Teller (BET) surface areas were calculated following established consistency criterion. Increasing loading of the rhenium complex results in surface area loss in excess of the added non-porous mass (Table 10), indicating surface adsorption.

TABLE-US-00010 TABLE 10 Results of gas adsorption measurements. Sample BET Surface Area (m.sup.2/g) Blank 175 ± 0.6 1.4% Re 120 ± 0.4 12.5% Re 107 ± 0.9 22.2%  87 ± 0.9 23.4%  83 ± 0.8

[0169] FIG. 22A depicts an N.sub.2 adsorption (77 K) of a Re-loaded MWCNT/CNF composite in linear scale; 0% Re (square data points, no corners on plot), 1.4% Re (circular data points), 12.5% Re (triangular data points, apex of triangles pointing up), 22.2% Re (triangular data points, apex of triangles pointing down), 23.9% Re (square data points, 2 corners on plot). FIG. 22B depicts an N.sub.2 adsorption (77 K) of a Re-loaded MWCNT/CNF composite in log scale; 0% Re (square data points, no corners on plot), 1.4% Re (circular data points), 12.5% Re (triangular data points, apex of triangles pointing up), 22.2% Re (triangular data points, apex of triangles pointing down), 23.9% Re (square data points, 2 corners on plot). FIG. 22C depicts density functional theory (DFT) pore size distributions of samples calculated from N.sub.2 isotherms (77 K); 0% Re (square data points, no corners on plot), 1.4% Re (circular data points), 12.5% Re (triangular data points, apex of triangles pointing up), 22.2% Re (triangular data points, apex of triangles pointing down), 23.9% Re (square data points, 2 corners on plot). Sample pore size distributions (PSD) were calculated using the Microactive software

package (Micromeritics) as multiwalled nanotubes under the DFT Pore Size model. The PSD of the blank sample shows a bimodal pore size distribution up to 10 nm—pore volume beyond this point is poorly defined and likely a result of interparticle condensation rather than intrinsic material porosity—which is closely mimicked by samples with low (1.4%) rhenium loading. Intermediate rhenium loadings (12.5%) result in noticeable changes in the PSD, where new pore distributions appear near 2.5 nm and 7 nm, likely from the disruption of the larger parent distributions. Moreover, the previously well-defined distribution at 5 nm is noticeably broadened. Further increasing (>20%) rhenium content continues to disrupt the pristine PSD, with a pronounced loss of cumulative porosity throughout the distribution. The changes in

[0170] PSD in conjunction with the decreases in BET surface area suggest that the rhenium complex is located within the pores of the substrate, though it is not possible to rule out the presence of external complex adsorption.

[0171] A number of embodiments of the invention have been described.

Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention. Accordingly, other embodiments are within the scope of the following claims.