Mohamed M. Ibrahima,
G. A. M. Mersala,
Ahmed M. Fallataha,
Khaled Althubeitia,
Hamdy S. El-Sheshtawy
b,
Manal F. Abou Talebcd,
Manash R. Dasef,
Rabah Boukherroubg,
Mohamed S. Attia
*h and
Mohammed A. Amin*a
aDepartment of Chemistry, College of Science, Taif University, P.O. Box 11099, Taif 21944, Saudi Arabia. E-mail: mohamed@tu.edu.sa
bChemistry Department, Faculty of Science, Kafrelsheikh University, Kafr El Sheikh 33516, Egypt
cDepartment of Chemistry, College of Science and Humanities in Al-Kharj, Prince Sattam Bin Abdulaziz University, Al-Kharj, Saudi Arabia
dPolymer Chemistry Department, National Center for Radiation Research and Technology (NCRRT), Egyptian Atomic Energy Authority, Cairo, Egypt
eAdvanced Materials Group, Materials Sciences and Technology Division, CSIR-North East Institute of Science and Technology, Jorhat 785006, Assam, India
fAcademy of Scientific and Innovative Research (AcSIR), Ghaziabad 201002, India
gUniv. Lille, CNRS, Centrale Lille, Univ. Polytechnique Hauts-de-France, UMR 8520 – IEMN, F59000 Lille, France
hChemistry Department, Faculty of Science, Ain Shams University, Abbassia 11566, Cairo, Egypt. E-mail: Mohd_mostafa@sci.asu.edu.eg
First published on 11th March 2022
Three transition metal complexes (MC) namely, [TpMeMeCuCl(H2O)] (CuC), [TpMeMeNiCl] (NiC), and [TpMeMeFeCl2(H2O)] (FeC) {TpMeMe = tris(3,5-dimethylpyrazolyl)borate} were synthesized and structurally characterized. The three complexes CuC, NiC, and FeC-modified glassy carbon (GC) were examined as molecular electrocatalysts for the hydrogen evolution reaction (HER) in alkaline solution (0.1 M KOH). Various GC-MC electrodes were prepared by loading different amounts (ca. 0.2–0.8 mg cm−2) of each metal complex on GC electrodes. These electrodes were used as cathodes in aqueous alkaline solutions (0.1 M KOH) to efficiently generate H2 employing various electrochemical techniques. The three metal complexes' HER catalytic activity was assessed using cathodic polarization studies. The charge-transfer kinetics of the HER at the (GC-MC)/OH− interface at a given overpotential were also studied using the electrochemical impedance spectroscopy (EIS) technique. The electrocatalyst's stability and long-term durability tests were performed employing cyclic voltammetry (repetitive cycling up to 5000 cycles) and 48 h of chronoamperometry measurements. The catalytic evolution of hydrogen on the three studied MC surfaces was further assessed using density functional theory (DFT) simulations. The GC-CuC catalysts revealed the highest HER electrocatalytic activity, which increased with the catalyst loading density. With a low HER onset potential (EHER) of −25 mV vs. RHE and a high exchange current density of 0.7 mA cm−2, the best performing electrocatalyst, GC-CuC (0.8 mg cm−2), showed significant HER catalytic performance. Furthermore, the best performing electrocatalyst required an overpotential value of 120 mV to generate a current density of 10 mA cm−2 and featured a Tafel slope value of −112 mV dec−1. These HER electrochemical kinetic parameters were comparable to those measured here for the commercial Pt/C under the same operating conditions (−10 mV vs. RHE, 0.88 mA cm−2, 108 mV dec−1, and 110 mV to yield a current density of 10 mA cm−2), as well as the most active molecular electrocatalysts for H2 generation from aqueous alkaline electrolytes. Density functional theory (DFT) simulations were used to investigate the nature of metal complex activities in relation to hydrogen adsorption. The molecular electrostatic surface potential (MESP) of the metal complexes was determined to assess the putative binding sites of the H atoms to the metal complex.
In terms of sustainability and environmental impact, water electrolysis is the most promising technique for creating high purity H2 from renewable energy sources, as it solely produces oxygen as a by-product and emits no greenhouse gases.1–3 The rate of the cathodic half-reaction of water splitting, viz. the hydrogen evolution reaction (HER), determines the efficiency of the water electrolysis.9 A highly efficient and robust heterogeneous electrocatalyst is required to accelerate the HER kinetics.10 The efficient electrocatalyst should be able to reduce the HER overpotential below 100 mV. The greater the activity of a catalyst, the lower its overpotential.11
The efficient electrocatalyst should also have the highest possible electrochemically active surface area (ECSA), which is the active region on the catalyst's surface that is accessible for the charge transfer process, as well as a high faradaic efficiency, which is calculated as the ratio of the experimentally determined quantity of H2 to the calculated (theoretical) quantity of H2 assuming 100% faradaic efficiency.11 In addition to the electrochemical parameters mentioned above that can be used to evaluate the electrocatalyst's performance, the coveted electrocatalyst should have a high turnover frequency (TOF), which is defined as the amount of reactants that a catalyst can convert per catalytic site per unit time into the required product, as well as good stability.11
All of the above-mentioned conditions for a desirable electrocatalyst were met using platinum (Pt) and Pt-based materials. As a result, among active electrocatalysts reported in the literature, such materials remain the most efficacious electrocatalysts for the HER.11,12 However, because of their high cost and rarity, their usage in industry has been severely limited.11,12 To meet the industry's need for H2 generation, substantial scientific research has been performed to develop and manufacture non-Pt-based cost-effective and efficient electrocatalysts.12,13 For example, according to Jaramillo et al.,14 MoS2 demonstrated exceptional HER catalytic performance. There has recently been a lot of interest in studies on carbon-based catalysts with low overpotentials for the HER, such as N, B, or S-doped carbons.15,16
H2 may also be produced efficiently using molecular electrocatalysts, such as transition-metal complexes, which are the most stable materials and are used to store energy in chemical bonds as well as generate electricity through hydrogen and oxidation production.17 For instance, recently, two complexes with bipyridine ligand, namely [Co(bpy)2(SCN)2] 1 and [Ni(bpy)2(SCN)2] 2 were synthesized and characterized by C.-L. Wang et al.18 Both showed promising catalytic activity for the hydrogen evolution reaction (HER). In an aqueous solution of neutral buffer (pH 7.0), at an overpotential of 837.6 mV, complexes 1 and 2 electro-catalyzed the HER with turnover frequency (TOF) values of 699.6 and 1469 mol H2 per mol catalyst per h, respectively.
In another study, Z. Niu et al.19 prepared three molecular catalysts with square planar geometries based on mononuclear nickel(II) complexes, and investigated their catalytic activity for H2 generation in aqueous and non-aqueous solutions using cyclic voltammetry. The hydrogen production in both solutions was confirmed by controlled-potential electrolysis using gas chromatography. R. Li et al.20 synthesized two Co(II)/Ni(II) complexes based on nitrogenous heterocyclic ligand and explored them as stable electrocatalysts for the HER.
Incorporating transition metal (TM) complexes into cluster-based chalcogenide frameworks is a useful synthetic method for inducing structural diversity and controlling optoelectronic characteristics, which may boost electrocatalytic performance even further.21 For instance, J. Li et al.22 solvothermally synthesized and structurally characterized three novel isostructural neutral inorganic–organic open frameworks of gallium thioantimonate that are covalently bound with certain TM complexes. Thermal and chemical stability were good in the synthesized materials. The title compounds have tunable photocatalytic H2 evolution activity thanks to an adjustable TM center.
The hydrogenase enzymes23,24 are depicted as naturally occurring microorganisms that efficiently catalyze the reversible interconversion of protons to hydrogen using earth-abundant metals (Ni and Fe) with high turnover rates 6000–9000 ([Fe–Fe]) and 700 ([Ni–Fe]) moles of H2 per mole of enzyme per second for the HER.25 However, enzymes are difficult to produce in sufficient quantities for commercial uses, and their stability outside of their native system is generally limited.24,26 For these reasons, the structural and functional mimicking of hydrogenase's active site has recently acquired attention.27–29 As a result, the development of highly efficient, stable, and robust HER molecular electrocatalysts for the production of hydrogen, which are inexpensive and made of earth-abundant transition metal elements and different ligand combinations, has attracted the community's interest.27–29
Herein, three mononuclear metal complexes (MC) namely, [TpMeMeCuCl(H2O)] (CuC), [TpMeMeNiCl] (NiC), and [TpMeMeFeCl2(H2O)] (FeC) {TpMeMe = tris(3,5-dimethylpyrazolyl)borate} were reported as molecular catalysts supported on glassy carbon (GC) for the HER in an alkaline electrolyte (0.1 M KOH). The stability and catalytic mechanism of these molecular catalysts were investigated by chemical and electrochemical methods. Based on the achieved experimental data, we established a relationship between the functional pyrazolyl nitrogen of the pyrazolylborate ligand and the electrocatalytic water splitting stability of the corresponding CuC, NiC, and FeC complexes. The catalytic evolution of hydrogen on the three supported complexes was further studied using density functional theory (DFT) simulations.
DFT calculations using B3LYP/lanl2dz level of theory were used to explore the structures of metal complexes MC. Fig. 1 shows the optimized structures of the metal complexes in alkaline medium, where the chloride anions (Cl−) were substituted by hydroxyl groups (OH−). The structures of the metal complexes showed that each metal ion is coordinated to three nitrogen atoms from the anionic ligand TpMeMe−, forming symmetrical structures. The stable copper(II) complex CuC binds with one more water molecule (Cu–O48H2 = 2.16 Å), besides the OH− groups (Cu–O46 = 1.94 Å). While the iron(III) complex FeC binds with one water molecule (Fe–O50H2 = 2.05 Å) and two hydroxyl groups (Fe–O47 = 1.78 Å and Fe–O46 = 1.78 Å). These results were established by the TGA data and confirmed by the stable DFT structures. On the other hand, the neutral nickel(II) complex NiC binds with only one OH− group (Ni–O46 = 1.80 Å).
![]() | ||
Fig. 1 The optimized structures of (a) CuC, (b) NiC, and (c) FeC at B3LYP/lanl2dz level of theory in the gas phase. |
The formation of metal complexes (MC) was further evidenced from the high energy shift of ν(B–H) stretching vibration near 2500 cm−1 in the IR spectra of the metal complexes compared to that of the free ligand KTpMeMe, which appeared at 2436 cm−1 (Fig. 2). Similar shifts were reported for other Tp complexes.32 The IR spectra also showed other characteristic stretching and bending vibration bands at ∼2920, 1680, and 1150 cm−1, assigned to ν(C–H), ν(CN), and δ(C–H), respectively. Raman spectra of the ligand and its metal complexes (Fig. 3) revealed almost identical vibration bands with those found in their IR spectra. The non-electrolytic chloride ion in all complexes exhibits vibration bands, characteristics of the terminally-coordinated chloride ions.33 Additionally, the three metal complexes all featured weak bands at around 442, 449, and 452 cm−1, suggesting the existence of M–N bonds.34
The thermogram TGA-DTG-DSC of FeC (Fig. 3A) shows that the complex dehydrates one water molecule with an exothermic peak at 98 °C. It decomposes in the range of 120–900 °C (endothermic peak, DSC: 410 °C) to eliminate the organic part with a mass loss of 63.86% (calc. 63.27%), forming a mixture of ferrous(II) oxide and potassium oxide. The thermal analysis of the copper(II) complex CuC indicates that its decomposition takes place in several stages (Fig. 3B). The thermogram shows that the complex decomposes to eliminate the organic part in the range of 105–820 °C (endothermic peak, DSC: 461 K) with a mass loss of 81.98% (calc. 81.52%), generating a mixture of copper(II) and potassium oxide (found 19.64%; calc. 18.73%).
The electronic absorption spectrum of CuC in methanol (Fig. 4)35 exhibited weak bands in the lowest energy visible region attributable to d–d transitions of the Cu(II) ion in a square-pyramidal environment.36 The absorption spectrum of NiC displays three absorption bands at 330, 560, and 635 nm. An examination of these bands indicates that the complex has a tetrahedral geometry with ground state 3T1.37,38 In the regular tetrahedral nickel(II) complexes, only one d–d transition is observed in the visible region. The observation of the three transitions in the spectrum of NiC complex may be attributed to the distortion in the tetrahedral symmetry around the nickel(II) center. This spectral behavior is comparable with analogous tetrahedral nickel(II) complexes.39,40 Hence the nickel(II) ion has an sp3 hybridization with tetrahedral geometry. The intensity of the UV band at around 330 nm is consistent with a ligand-centered π → π* transition or a charge-transfer transition. The electronic spectrum of FeC exhibits four d–d transitions at 325, 374, 490, and 588 nm, assignable to 6A1g → 4T1g (G), 6A1g → 4T2g (G), 6A1g → 4Eg (G) and 6A1g → 4Eg (D), respectively in the octahedral stereochemistry.41 The magnetic moment value of CuC was determined to be 1.98 BM, confirming the presence of magnetically dilute Cu(II) complex,42 whereas the magnetic moment for FeC was close to the spin-only value of 5.63 and 5.44 μB, which is characteristic of high spin of d5 configuration.43,44
The quantitative elemental analysis of the three complexes CuC, NiC, and FeC were also determined using energy dispersive X-ray fluorescence (EDXRF) spectroscopy at atmospheric pressure and the obtained spectra are illustrated in Fig. 5. The Ar-Kα was found in all the spectra at 2.957 keV and it has no influence on the characteristic lines of Cu, Ni and Fe. Additional impurities of Ca and Zn were also found, which were originated form the used Mylar foil. In the case of complex NiC, the characteristic radiation Ni-Kα and Ni-Kβ was measured at 7.473 keV and 8.265 keV respectively. Addition escape peak of Ni-Kα was found at 5.733. The spectrum belongs to complex CuC, the Cu-Kα and Kβ were detected successfully at 8.042 keV and 8.906 keV respectively. Whereas in the case of complex FeC, the characteristic radiation of Fe–K lines could be easily detected free from interference at 5.895 keV and 6.491 keV for Kα and Kβ of Fe respectively. Escape peak of Fe-Kα could be recognized also at 4.155 keV. The % of the coordinated TpMe,Me in all complexes were calculated and omitted from the fundamental parameter program. This is due to the limitation of EDXRF setup whereas there is not possibility to detected elements of atomic number less than 11 (Z < 11). The weight% of complexes CuC, NiC, and FeC were found to be 15.27 ± 0.18, 15.22 ± 0.18, and 12.55 ± 0.18, respectively.
![]() | ||
Fig. 5 EDXRF spectra of complexes CuC, NiC, and FeC with the background (sample holder with Mylar window) using direct excitation of X-ray tube. |
Current densities in this work are obtained by dividing currents by the electrochemical active surface area (EASA), calculated from cyclic voltammetry measurements, see later. The various HER electrochemical kinetic parameters (Table 1) associated with such polarization measurements were derived from Tafel plots, Fig. 6(b).
Tested cathode | EHER/mV (RHE) | −βc/mV dec−1 | j0/mA cm−2 | η10/mV |
---|---|---|---|---|
Bare GC electrode | — | 254(6) | 6.3(0.15) × 10−4 | — |
GC-FeC (0.2 mg cm−2) | 305(5) | 162(3.3) | 1.1(0.05) × 10−2 | 467(8) |
GC-FeC (0.4 mg cm−2) | 255(3.4) | 160(3.2) | 2.5(0.07) × 10−2 | 423(7.6) |
GC-FeC (0.8 mg cm−2) | 230(3.2) | 158(3) | 3.7(0.1) × 10−2 | 372(5) |
GC-NiC (0.2 mg cm−2) | 180(2.6) | 157(2.8) | 4.3(0.3) × 10−2 | 278(4.2) |
GC-NiC (0.4 mg cm−2) | 107(2.2) | 160(2.6) | 10(0.12) × 10−2 | 250(4.6) |
GC-NiC (0.8 mg cm−2) | 80(1.7) | 155(2.2) | 14(0.35) × 10−2 | 213(4) |
GC-CuC (0.2 mg cm−2) | 45(1.5) | 118(1.8) | 29(0.6) × 10−2 | 176(4.4) |
GC-CuC (0.4 mg cm−2) | 30(1.1) | 114(2.1) | 60(0.9) × 10−2 | 140(2.6) |
GC-CuC (0.8 mg cm−2) | 25(0.5) | 112(1.6) | 70(1.1) × 10−2 | 120(2.2) |
Pt/C | 10(0.2) | 108(1.2) | 88(1.2) × 10−2 | 110(1.4) |
Such parameters, which include exchange current density (j0), cathodic Tafel slope (βc), and the overpotential required to deliver a current density of 10 mA cm−2 (η10), are essential to evaluate and compare the HER catalytic activity of the studied metal complexes 1–3.
Included also in Table 1 is another important HER electrochemical kinetic parameter, namely the HER onset potential (EHER), the potential at which the reduction of water molecules commences to liberate H2, followed by a rapid growth in the cathodic current. This parameter is directly estimated from the linear polarization plot, Fig. 6(a). It is one of the key electrochemical parameters employed to evaluate and compare electrocatalysts' performance towards the HER, since any alteration in the EHER's location on the linear polarization results in considerable changes in the exchange current density value.45 The literature has demonstrated that the anodic shift in EHER prompts hydrogen production with higher exchange currents at lower overpotentials, referring to enhanced electrocatalytic activity.45 Table 1 reveals, at any loading density, a considerable anodic shift in EHER, corresponding to improved catalytic performance, ongoing from GC-FeC to GC-CuC with the GC-NiC's EHER value locates in between. This anodic drift in EHER, which increases upon increasing the loading density, enhances following the sequence: GC-FeC (−255 mV vs. RHE @ 0.4 mg cm−2) < GC-NiC (−107 mV vs. RHE @ 0.4 mg cm−2) < GC-CuC (−30 mV vs. RHE @ 0.4 mg cm−2) < GC-CuC (−25 mV vs. RHE @ 0.8 mg cm−2).
The overpotential the electrocatalyst must acquire to attain a current density of 10 mA cm−2 (η10), another important HER electrochemical kinetic parameter employed for comparing and evaluating the catalytic performance of electrocatalysts,46 exhibits the same trend as for EHER values anodic drift. Here again, at any tested loading density, the lowest η10 value is estimated for the GC-CuC catalyst. These findings demonstrate the outstanding HER catalytic performance of the GC-CuC catalyst as it recorded the lowest EHER (25 mV vs. RHE) and η10 (120 mV) values among the other studied catalysts.
The GC-CuC catalyst's low EHER value, together with its steep polarization curve beyond its EHER, led to a j0 value of 0.29 mA cm−2 at 0.2 mg cm−2 loading density calculated by the Tafel extrapolation method, Table 1. This GC-CuC catalyst's j0 value increased with increase in loading density attaining a j0 value of 0.7 mA cm−2 at a loading density of 0.8 mg cm−2, approaching that of the commercial Pt/C catalyst performed here under the same operating conditions (0.88 mA cm−2).
These findings highlight the catalyst's loading density role in catalyzing the HER. The population of the HER active catalytic sites is expected to increase with increase in the loading density of the catalyst, as evidenced from the catalyst surface's active sites (n, mol) calculations (see later). This, in turn, expedites efficacious transfer of charge during the water molecules reduction process thus, boosts the HER kinetics.
Two distinct depressed semicircles can be seen in the Nyquist plots of all materials investigated. The electrocatalytic HER on Ni,47 Co,48 and some of their alloys13,49–51 in alkaline solutions exhibited such impedance behavior. Similar impedance results were reported earlier in our lab for HER investigations on Co50Ni23Ga26Al1.0 magnetic shape memory alloy,52 reduced graphene oxide nanosheets, some metallic nanoparticles-decorated titanium substrates, and cathodically activated carbon felt samples.53–56
At high frequency values, a first small semicircle (with diameter R1) was visible, whereas a large second semicircle (with diameter R2) appeared at low frequencies. Each of them can be assigned to the resistance–capacitance (RC) network. Each RC network at the catalyst/KOH interface corresponds to the charge-transfer resistance (R1 for the first semicircle and R2 for the second, Rtotal or Rct = R1 + R2) and associated double-layer capacitance (C1 for the first semicircle and C2 for the second, Ctotal = C1 + C2).56 The two semicircles appear as conspicuous humps in the related Bode graphs, Fig. S10 (ESI†). It is obvious from Fig. 7 that the size of the two semicircles varies according to the type the tested catalyst and its loading density. Also, referring to Fig. S10,† at a fixed loading density, both semicircles are clearly potential-dependent. These findings indicate that the two semicircles' size is connected to the HER kinetics.57
For comparison, EIS responses of bare GC and Pt/C electrodes with a single capacitive loop (Fig. S11, ESI†) were also recorded. The RC networks' elements were estimated using the equivalent circuit illustrated in Fig. S12 (ESI†), allowing us to gain a better grasp of the HER kinetics. This electrical equivalent circuit was described in more detail elsewhere.54
Based on the impedance data fitting, when the loading density of the catalyst was raised, the values of both R1 and R2, and hence the Rct value (Rct = R1 + R2), reduced considerably, indicating higher HER catalytic activity. For example, the Rct values recorded for the NiC catalyst decreased from 315 Ω cm2 at a loading density value of 0.2 mg cm−2 to 248 and 145 Ω cm2 at loading densities 0.4 and 0.8 mg cm−2, respectively. The CuC electrocatalyst showed the lowest Rct value among the other studied samples at any studied loading density thus, confirming the excellent HER catalytic activity of the CuC electrocatalyst. For instance, at a loading density value of 0.8 mg cm−2, the CuC electrocatalyst recorded an Rct value of 51 Ω cm2. This value is very close to that of the Pt/C (47 Ω cm2), and lower than those measured for GC-NiC (145 Ω cm2) and GC-FeC (450 Ω cm2) at the same loading density.
Another critical electrochemical parameter, namely the Tafel slope (βc), is also utilized to further assess the studied electrocatalysts' HER catalytic leverage towards the HER. The βc value recorded here for the commercial Pt/C catalyst is found to be 108 mV dec−1, close to that reported in the literature under the same operating conditions, 111 mV dec−1,58 meaning that the electrochemical measurements employed here are reliable.
The alkaline HER step mechanism can be represented by eqn (1)–(3):59
H2O + e− → Hads + HO− (Volmer step) | (1) |
The Volmer step is followed by either a combination of adsorbed species to form molecular H2:
Hads + Hads → Hads (Tafel step) | (2) |
H2O + e− + Hads → H2 + HO− (Heyrovsky) | (3) |
The Pt/C catalyst's βc value obtained here (108 mV dec−1), which is comparable with that reported in the literature (111 mV dec−1) under the same operating conditions,58 is close to the standard one (120 mV dec−1) reported in eqn (1), meaning that the alkaline HER mechanism comprises the Volmer step as the rate-limiting.60
The Tafel lines of FeC and NiC catalysts are parallel to each other, irrespective of the value of the catalyst's loading density, with βc values close to 160 mV dec−1. Also, the three studied CuC catalysts, namely CuC (0.2 mg cm−2), CuC (0.4 mg cm−2), and CuC (0.8 mg cm−2) exhibited parallel Tafel lines, but with reduced βc values (118, 114, and 112 mV dec−1, respectively). These findings indicate that the kinetics of the HER on the CuC catalyst's surface is faster than on the surfaces of FeC and NiC catalysts. This is evidenced from the marked drop in the Tafel slope values of the CuC catalyst since lower Tafel slopes often allude to a large number of catalytic active sites being available.60,61
The CuC (0.8 mg cm−2) catalyst's βc value (112 mV dec−1) is close to that of the Pt/C (108 mV dec−1) thus advocating a mechanism for the HER on the surface of that catalyst controlled by the Volmer step.60 The results show that the CuC catalyst can efficiently catalyze HER in 0.1 M KOH solution, which is similar to the most active HER electrocatalysts reported in the literature, Table S1 (ESI†).
The electrochemical active surface area (EASA) is another essential parameter used to compare different electrocatalysts' activity, as the kinetics of the HER over electrocatalysts is directly linked to their ECSA. Being directly related to ECSA, double layer capacitance (Cdl), determined from cyclic voltammetry measurements in non-faradic region, Fig. S1 (ESI†), is used to calculate ECSA.62–64 From the slopes of the current density versus scan rate (v) plots, Fig. S1 (ESI†), for any studied meal complex, Cdl values were found to increase with increase in the loading density. At any investigated loading density, this increase in Cdl enhances following the sequence: FeC < NiC < CuC. These findings confirm the high catalytic performance of the CuC electrocatalyst since electrocatalysts with higher Cdl values can provide higher adsorption density of reactive sites, and therefore promote efficient charge transfer, resulting in enhanced catalytic activity.
Eqn (4) is used to calculate the ECSA from Cdl.65
EASA = Cdl/Cs | (4) |
The catalyst surface's active sites (n, mol) are estimated using eqn (5):66
n = Qnet/2F | (5) |
For any studied catalyst, the value of n is found to enhance with the catalyst's loading density. At any tested loading density, the value of n increases obeying the order: FeC < NiC < CuC. These findings confirm the CuC's highest HER catalytic performance.
To further evaluate the catalytic performance of the investigated catalysts at a given overpotential, the turnover frequency (TOF) was calculated, which is the number of hydrogen molecules created per second per active site.66 The TOF values for the tested catalysts were computed using the EASA values in Table 2 at an overpotential of 500 mV, as discussed previously.52 Table 2 depicts the obtained TOF values, which feature a dramatic increase in the value of TOF as the loading density of the examined catalyst increases. The TOF value is always higher for the CuC at any studied loading density. To our knowledge, the TOF value recorded for the best catalyst, namely CuC, is comparable to the most efficient molecular catalysts and greater than many others described in the literature for electrochemical H2 production. Table S1 (ESI†) summarizes these comparisons.
Tested cathode | Cdl/mF cm−2 | ECSA/cm2 | Q × 103/C | n × 108/mol | TOF @ 0.5 V |
---|---|---|---|---|---|
GC-FeC (0.2 mg cm−2) | 11.26 | 375.3 | 4.2 | 2.2 | 872 |
GC-FeC (0.4 mg cm−2) | 20.27 | 675.7 | 6.8 | 3.5 | 988 |
GC-FeC (0.8 mg cm−2) | 30.55 | 1018.3 | 11.2 | 5.8 | 1124 |
GC-NiC (0.2 mg cm−2) | 15.55 | 518.3 | 7.4 | 3.8 | 927 |
GC-NiC (0.4 mg cm−2) | 28.32 | 944.0 | 12.4 | 6.4 | 1192 |
GC-NiC (0.8 mg cm−2) | 44.43 | 1481.0 | 19.8 | 10.3 | 1338 |
GC-CuC (0.2 mg cm−2) | 21.2 | 706.7 | 11.6 | 6.0 | 1154 |
GC-CuC (0.4 mg cm−2) | 36.71 | 1223.7 | 17.8 | 9.2 | 1510 |
GC-CuC (0.8 mg cm−2) | 53.02 | 1767.3 | 26.9 | 13.9 | 1820 |
![]() | ||
Fig. 8 Calculated MESP of the (a) CuC, (b) NiC, and (c) FeC metal complexes at B3LYP/lanl2dz in the gas phase. |
![]() | ||
Fig. 9 Optimized structures of the H atom adsorbed on the (a) CuC, (b) NiC, and (c) FeC metal complexes at B3LYP/lanl2dz in the gas phase. |
To retain the molecular nature of the metal complex MC catalysts under the alkaline conditions, it is necessary for the ligand to be tightly bound to the metal center. These catalysts were generated by treating metal(II) chloride (MCln) with one equivalent of ligand in absolute methanol, in which one tris(pyrazolyl)borate anion (TpMeMe−) is bound to the metal center with the fourth coordination site, occupied by a chloride in the case of NiC, whereas in CuC and FeC, the ligand is bound to the metal center with one and two more water molecules, respectively. Upon the addition of one equivalent of strong base, the chloride ligand in all complexes is substituted by hydroxide (OH−) anion, and the resulting Cu(II)-bound hydroxide complex, for example, adopts a square pyramidal structure with the OH-donor. Upon the addition of one equivalent of hydrochloric acid, the OH ligand can be protonated to regenerate the aqua complex. This reversible protonation of the ligand is an attractive feature for a water-oxidation catalyst, as it may allow for proton-coupled electron transfer (PCET), a common reaction step in water oxidation.
The actual catalyst for the current catalytic systems is generated in an alkaline medium, where the hydroxide anion must bind to the central metal ion of the catalyst. Thus, the initiation step of the present catalytic reaction involves binding of OH− to the central metal ion of the catalyst. In this context, the octahedral Fe(III) complex is coordinately saturated, and thus has no vacant coordination sites, which can be utilized to bind any reacting species during the catalytic process. The ability of such coordinately saturated complexes to dissociate (ML6 ⇌ ML5 + L) and consequently the energy of dissociation is one of the key factors in imparting catalytic properties to these metal complexes. On the other hand, the five coordinated square pyramidal Cu(II) complex and the four coordinated tetrahedral Ni(II) complex already have a vacant coordination site for the substrate binding and thus need no energy for ligand dissociation as in the case of the octahedral Fe(III) complex.
Although Cu(II) and Ni(II) complexes are both coordinatively unsaturated, Cu(II) complex exhibits higher catalytic activity. The superiority of Cu(II) can be explained by the possible mechanistic implication of the six coordinate intermediate [H2–ML–OH]. Based on Jahn–Teller theorem, this hexacoordinated species is highly unstable in the case of Cu(II) complex as compared to Ni(II) complex. According to the Jahn–Teller theorem, the hexa-coordinated Cu(II) complexes are usually associated with a tetragonal distortion reducing the symmetry from Oh to Dh.67,68 Thus, we propose that rapid structural reorganization of this active intermediate leads to rapid product formation and release. These structural changes will reduce the coordination number from six to five and return the catalyst to its original, stable form. As for the Ni(II) complex does not suffer from the Jahn–Teller effect, its catalytic reaction proceeds more slowly than it does in the case of the Cu(II) complex.
Copper and nickel are more labile than iron. This sequence of the lability degree could also be due to the electronic configuration of the metal center in these metal complexes and can be illustrated from the following discussion. The 3d-electronic configuration (electron density) of the metal ion is a key factor for determining the degree of lability or inertness of transition metal complexes.69 According to valence bond theory (VBT), the electron density of MC complexes is increasing in the following order CuC (4 lone pairs) > NiC (three lone pairs) > FeC (no lone pair). This is in accordance with the HER catalytic performance, in which the three tested metal complexes' activity follows the sequence: CuC > NiC > FeC.
In these solutions, the catalyst is particularly stable, as indicated by the humble loss in current after 6000 continuous potential cycles. The Tafel slope obtained after the last cycle (115 mV dec−1) is close to that recorded after the first cycle (112 mV dec−1), suggesting that catalyst potential cycling has a negligible influence on the mechanism of the HER. Because the current remains nearly constant throughout the operation, the chronoamperometry measurements (inset of Fig. 10) give an additional corroboration for the catalyst stability.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1ra08530a |
This journal is © The Royal Society of Chemistry 2022 |