Role of electrolyte composition on the acid stability of mixed-metal oxygen evolution catalysts

Abstract

Acid stability in catalysts that promote the oxygen evolution reaction (OER) may be achieved through either the introduction of electrolyte-modulated self-healing processes, or the fixation of OER active sites within stable, conductive oxide matricies. By varying the nature of the electrolyte, pH, buffer concentration, and ionic strength, the contributions of these two effects may be deconvoluted. Furthermore, we find that the nature of the buffer is capable of engendering OER with exceptional acid stability at moderate overpotentials.

---

The use of sunlight to produce liquid fuels directly from CO₂ is an important strategy to diversifying the global fuel supply in a renewable and sustainable manner.^1–4 Viable solar-generated liquid fuel processes must proceed at appreciable rates, at high energy efficiency (low overpotential) and with high carbon atom economy. While CO₂ reduction to fuels has been demonstrated at synthetically relevant rates,⁵ the process confronts the limitation of low carbon efficiency owing to the non-faradaic reaction of HO⁻ with CO₂ to produce super-stoichiometric amounts of carbonate, severely compromising the atom efficiency of CO₂-to-fuels catalysis.⁶ As has recently been demonstrated, a cell equilibrium is established with the complete consumption of HO⁻ by CO₂ to produce a purely bicarbonate electrolyte.⁷ This acid–base chemistry results in a loss of cell voltage and, more importantly, in extremely low CO₂-to-fuel carbon yields. The time for such conversion to carbonate depends on a cell's reservoir of hydroxide, giving rise to a characteristic S-curve where this hydroxide reservoir absorbs CO₂ until base is depleted, after which there is a rapid rise and then levelling upon concentration of a pure carbonate electrolyte. Cells reporting high current density are typically operated for shorter times than this transition to steady-state and consequently do not reflect the actual equilibrium conditions confronting a viable solar to liquid fuels process. Thus, the CO₃²⁻ problem is currently the greatest single obstacle to achieving a viable photoelectrochemical fuels synthesis.

Carbonate formation may be mitigated by performing the CO₂ reduction reaction (CO₂RR) in acidic solutions, thus providing an imperative for the design of acid stable catalysts for the anodic reaction (i.e., the oxygen evolution reaction, OER), which provides reducing equivalents and protons for CO₂RR. We have developed OER catalysts based on cobalt⁸ and nickel⁹ that can drive OER in water (pH > 5) with rapid kinetics and good energy efficiency. In more acidic conditions (pH < 5), we have further shown that, by tuning the chemistry of self-healing,¹⁰ metallate oxygen evolving catalysts (M-OECs) may be stabilized in acidic solution.¹¹ This stability may be enhanced further by embedding M-OECs in acid-stable metal oxide frameworks, such as MnO_x or PbO_x.¹² Using this approach, a series of acid-stable mixed-metal OECs, including CoFePbO_x, are able to maintain activity at pH 2.5 for >30 h. In such systems, the electrolyte plays an active role as a proton mediator for OER as well as a structural element in establishing the equilibrium needed for the self-healing of embedded active sites. We report herein studies seeking to explicitly define the role of electrolyte composition on film stability by comparing the OER activity of CoFePbO_x in phosphate (H₃PO₄/KH₂PO₄, KP_i) and sulphate (KHSO₄/K₂SO₄, KS_i), as these electrolytes possess similar chemical forms and pK_as (pK_a1(H₃PO₄) = 2.15 while pK_a2(H₂SO₄) = 2.00).¹³ We find that modulating the nature, concentration and pH of the electrolyte allows us to discern the contributions of self-healing and matrix-stabilization to catalyst longevity. Moreover, as an outgrowth of these studies, we find that when operated in sulphate buffer (in the absence of phosphate), the PbO_x framework can facilitate OER with exceptional acid stability at moderate overpotentials.

CoPbO_x and CoFePbO_x catalysts exhibit similar morphologies, as observed by SEM images of electrodeposited films, which are distinct from that of the filamentous PbO_x matrix (Fig. 1). EDS mapping indicates that the component metals are distributed homogeneously throughout the catalyst films (Fig. S1, ESI†). Both films in P_i and S_i perform OER with faradaic efficiencies of 100%.

Fig. 1 SEM images of (a) PbO_x, (b) CoPbO_x, (c) CoFePbO_x, and (d) FePbO_x for comparison. All samples were prepared on FTO substrate, and scale bar indicates 300 nm.

Chronopotentiometry measurements were performed on CoFePbO_x films at a constant current of 1 mA cm⁻². The films were immersed in stirred solutions containing either 100 mM of KP_i, 100 mM of KS_i, or 100 mM each of KP_i and KS_i. All solutions were buffered to pH 2.15 and contained enough KNO₃ supporting electrolyte to ensure a total ionic strength of 1.1 M. Representative chronopotentiometric traces are shown in Fig. 2. After a brief initial capacitive period, the operating potential for CoFePbO_x films settled at ∼1.6 V vs. NHE, at which OER was sustained (0.5 V overpotential for water splitting at pH = 2.15). After a period of operation, the film potentials underwent an inflection, rising rapidly before plateauing at an overpotential of 0.7 V for KS_i or 1.6 V for KP_i and KP_i/KS_i electrolytes, indicative of OER at the PbO_x matrix in KS_i and FTO substrate in KP_i, respectively. Direct O₂ measurements for CoFePbO_x films operated in KP_i and KS_i electrolyte solutions before and after inflection show 100% faradaic efficiency for OER, ruling out the possibility of unproductive oxidation reactions. Hence, the inflection time was used as a metric for film stability. On this basis, films operated in solutions buffered with 100 mM KP_i display extended stability at lowest overpotential as compared to those buffered with 100 mM KS_i.

Fig. 2 Stability of CoFePbO_x films performing OER at 1 mA cm⁻² in buffer solutions (pH 2.15) of 1 M KNO₃ + 0.1 M KP_i (red, P_i), 1 M KNO₃ + 0.1 M KS_i (green, S_i), or 0.8 M KNO₃ + 0.1 M KP_i + 0.1 M KS_i (orange, P_i/S_i): (a) chronopotentiometry curves, and (b) bar graphs indicating average stability (time required to reach the plateau current). η_pH2.15 = E_appl − E_therm where E_therm is the thermodynamic potential for water splitting at pH = 2.15 and E_appl = applied potential.

Film stability was examined for KP_i and KS_i electrolyte concentrations of 100 mM and 250 mM and at different pHs (Fig. S2, ESI†). For different concentrations of electrolyte, the overall buffer capacity was conserved in each electrolyte (see ESI†). At a concentration of 100 mM at pH 2.15, KP_i and KS_i buffers have a buffer capacity, β, of 0.074 and 0.072, respectively. At a concentration of 250 mM at pH 2.15, KP_i and KS_i buffers have a buffer capacity of 0.160 and 0.156, respectively. For both KP_i and KS_i, increasing buffer concentration led to a marginal decrease in film stability. This change cannot be attributed to ionic strength differences, as experiments performed in the presence of 1 M KNO₃ showed no difference in the stability of the film in the presence of KP_i whereas KS_i exhibited a slight decrease as compared to experiments run in the absence of KNO₃. More profound was the effect of pH on film stability. Representative chronopotentiometry curves and average film stability times at pH 1.03, 2.15 and 4.03 are shown in Fig. 3. For both electrolyte systems, film stability during OER operation decreased with decreasing pH; films operating in S_i exhibited significantly more instability than films operated in P_i. For these films, a precipitous drop in stability is observed between pH 2.15 and pH 1.03, consistent with the essential role of monobasic phosphate in promoting self-healing (vide infra).

Fig. 3 Stability of CoFePbO_x films performing OER at 1 mA cm⁻² in solutions containing (a) 1 M KNO₃ + 0.1 M KP_i and (b) 1 M KNO₃ + 0.1 M KS_i at different pH: 1.03 (light red and light green), pH 2.15 (red and green), and pH 4.03 (dark red and dark green). Representative chronopotentiometry curves (left), and bar graphs indicating average film stabilities (right) at each pH are shown.

We compared the elemental composition of films before and after OER in different electrolyte solutions using ICP-MS (see ESI† for experimental details). The results of these experiments are summarized in Fig. 4 (Table S1, ESI†). For both KP_i and KS_i electrolytes, film composition does not change significantly from the as-deposited film within the first 10 min of OER. Following 40 hours of electrolysis (post-inflection for all buffer conditions) all films exhibited significant metal loss. Films operated in KP_i electrolyte lose approximately 50% of their Fe content, 75% of their Co content, and 45% of their Pb content (vs. as-deposited films). This metal loss does not appear to change with addition of KNO₃ or at higher concentrations of KP_i. Similarly, Co loss is also observed and in even greater mangitude (88%) for CoFePbO_x films operated in KS_i. Interestingly, however, no Fe or Pb loss is observed for CoFePbO_x films operated in KS_i. The increased loss of Pb from films operated in KP_i relative to KS_i may reflect the formation of soluble Pb(IV) phosphate complexes, which have been reported to form electrochemically via oxidation of PbO₂ in the presence of phosphoric acid.¹⁴ Formation of the corresponding sulphate complexes has not been observed.

Fig. 4 Total metal content, Pb (grey), Fe (brown), and Co (magenta) of 1 cm² CoFePbO_x films after operation in specified buffers at pH 2.15 for 0 min, 10 min, or 40 h at 1 mA cm⁻² for OER.

The disparate behaviour exhibited by CoFePbO_x films operated in acidic KP_i and KS_i solutions is consistent with stabilization of the catalyst active site by (1) electrolyte-driven self-healing of active sites and (2) matrix stabilization of active sites within the PbO_x framework. The work described herein now enables us to disentangle the role of self-healing and matrix stabilization in acidic media.

With regard to (1) active site stabilization via self-healing, pair distribution functional (PDF) analysis of CoP_i films shows the catalyst to be composed of oxidic metallate clusters of 10–14 Co atoms.^15–17 The active sites have been shown to self-heal^10,18 where their self-assembly is balanced with OER such that the driving force required for self-assembly (i.e., catalyst regeneration) lies energetically within that for OER catalysis. For Co oxidic metallate clusters, when operating in acidic environments, the self-assembly process cannot be maintained owing to an inverse third-order dependence on proton activity,¹⁹ thus moving the catalyst system out of a regime of stability and imposing a limited operating lifetime on the catalyst. At pH 2.15, the dominant bases in P_i and S_i are H₂PO₄⁻ and SO₄²⁻, respectively, both of which are weak proton acceptors with similar basicity, and consequently neither is able to facilitate the deposition of CoP_i at the same rate as HPO₄²⁻. However, PDF analysis and reaction kinetics have shown that phosphate acts as a structural element that caps the edges of the metallate clusters²⁰ and hence provides added protection to the edge sites of the metallate clusters against protonation, thus shifting the equilibrium back towards self-assembly of the metallate active sites by impeding dissolution of the active sites during catalyst operation. This structural role of the buffer is further emphasized by the differing stability of CoFePbO_x in KP_i and KS_i. The P_i electrolyte is known to bind oxide surfaces with higher affinity via an inner-sphere (i.e., hydrogen bonding) mechanism.^21,22 Previous work has also shown that adsorbed phosphate is capable of stabilizing various metal oxide surfaces against corrosion.^23,24 In contrast, S_i has a significantly lower affinity for oxides, and studies have found evidence for an outer-sphere (i.e., ionic) mode of adsorption by sulphate onto these surfaces.^25–27

With regard to (2) matrix stabilization, the comparison of CoP_i and CoFePbO_x in Fig. 5 demonstrates the effect of the lead oxide matrix in enhancing acid stability. Elemental composition data summarized in Fig. 4 reveal sustained resistance of the PbO_x matrix to corrosion, further reinforcing the role of PbO_x as a stable structural component.

Fig. 5 Stability ascribed to (left) self-healing by comparison of CoFePbO_x films operating at 1 mA cm⁻² in 1 M KNO₃ and either 0.25 M KP_i (red) or 0.25 M KS_i (green) solution, buffered to pH 2.15, and (right) matrix stabilization by comparison of CoP_i (blue) and CoPbO_x (orange) films operating at 1 mA cm⁻² in a 1 M KNO₃, 0.25 M KP_i, solution, buffered to pH 2.5 (ref. 12).

The chronopotentiometric trace in Fig. 2a for CoFePO_x in S_i is intriguing, and highlights the important role that electrolyte can have on OER film stability and activity. The trace begins at an overpotential similar to that observed for CoFePbO_x in P_i. After approximately 1 h, the operating potential rises, consistent with the loss of Co from the film (Fig. 4) and the trace stabilizes at an overpotential of 0.7 V, which is almost 1 V lower than that observed for films operated in KP_i following similar Co loss. Control experiments indicate that this potential is representative of OER on a PbO_x film when operated in KS_i buffer (Fig. S3, ESI†). When PbO_x is operated at a similar pH, but in the presence of KP_i, a 1.7 V overpotential is required to achieve the same current density.¹² Though PbO_x in KS_i requires a higher overpotential for OER than CoFePbO_x, it is exceedingly stable in acid, maintaining activity for over 150 h with no sign of degradation (Fig. S4, ESI†). The stark contrast in operating potentials for PbO_x in KP_iversus KS_i may result from differences in oxide surface-binding between phosphate and sulphate, as described above. Notably, a similar explanation has been proposed in the lead-acid battery literature where phosphoric acid is known to be an effective additive for suppressing OER.²⁸ These results demonstrate the impact that buffer identity can have on the operation and activity of OER catalysts in acid and forecasts the need to consider the role of electrolyte in the development of active, stable, and earth-abundant OER catalysts.

Material is based upon work supported under the Solar Photochemistry Program of the Chemical Sciences, Geosciences and Biosciences Division, Office of Basic Energy Sciences of the U. S. Department of Energy DE-SC0017619. T. P. K. acknowledges support from a Graduate Research Fellowship from the National Science Foundation. We thank Dr Matthew Nava for helpful discussions.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 15729–15735.
2. S. Chu and A. Majumdar, Nature, 2012, 488, 294–303.
3. T. R. Cook, D. K. Dogutan, S. Y. Reece, Y. Surendranath, T. S. Teets and D. G. Nocera, Chem. Rev., 2010, 110, 6474–6502.
4. D. Abbott, Proc. IEEE, 2010, 98, 42–66.
5. F. P. G. de Arquer, C.-T. Dinh, A. Ozden, J. Wicks, C. McCallum, A. R. Kirmani, D.-H. Nam, C. Gabardo, A. Seifitokaldani, X. Wang, Y. C. Li, F. Li, J. Edwards, L. J. Richter, S. J. Thorpe, D. Sinton and E. H. Sargent, Science, 2020, 367, 661–666.
6. B. A. Zhang, C. Costentin and D. G. Nocera, Joule, 2019, 3, 1565–1568.
7. M. Ma, E. L. Clark, K. T. Therkildsen, S. Dalsgaard, I. Chorkendorff and B. Seger, Energy Environ. Sci., 2020, 13, 977–985.
8. M. W. Kanan and D. G. Nocera, Science, 2008, 321, 1072–1075.
9. M. Dinca, Y. Surendranath and D. G. Nocera, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 10337–10341.
10. C. Costentin and D. G. Nocera, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 13380–13384.
11. M. Huynh, D. K. Bediako and D. G. Nocera, J. Am. Chem. Soc., 2014, 136, 6002–6010.
12. M. Huynh, T. Ozel, C. Liu, E. C. Lau and D. G. Nocera, Chem. Sci., 2017, 8, 4779–4794.
13. M. Blasco-Ahicart, J. Soriano-López, J. J. Carbó, J. M. Poblet and J. R. Galan-Mascaros, Nat. Chem., 2018, 10, 24–30.
14. R. F. Amlie and T. A. Berger, J. Electroanal. Chem. Interfacial Electrochem., 1972, 36, 427–433.
15. C. L. Farrow, D. K. Bediako, Y. Surendranath, D. G. Nocera and S. J. L. Billinge, J. Am. Chem. Soc., 2013, 135, 6403–6406.
16. C. N. Brodsky, D. K. Bediako, C. Shi, T. P. Keane, C. Costentin, S. J. L. Billinge and D. G. Nocera, ACS Appl. Energy Mater., 2019, 2, 3–12.
17. P. Du, O. Kokhan, K. W. Chapman, P. J. Chupas and D. M. Tiede, J. Am. Chem. Soc., 2012, 134, 11096–11099.
18. D. K. Bediako, A. M. Ullman and D. G. Nocera, Catalytic Oxygen Evolution by Cobalt Oxido Thin Films. in Topics in Current Chemistry, ed. H. Tuysuz and C. K. Chan, Springer Publishing AG, Switzerland, 2016, 173–213.
19. Y. Surendranath, D. A. Lutterman, Y. Liu and D. G. Nocera, J. Am. Chem. Soc., 2012, 134, 6326–6336.
20. G. Kwon, H. Jang, J. S. Lee, A. Mane, D. J. Mandia, S. R. Soltau, L. M. Utschig, A. B. F. Martinson, D. M. Tiede, H. Kim and J. Kim, J. Am. Chem. Soc., 2018, 140, 10710–10720.
21. Y. Arai and D. L. Sparks, J. Colloid Interface Sci., 2001, 241, 317–326.
22. E. Bae and W. Choi, J. Phys. Chem. B, 2006, 110, 14792–14799.
23. E. D. Laudadio, J. W. Bennett, C. M. Green, S. E. Mason and R. J. Hamers, Environ. Sci. Technol., 2018, 52, 10186–10195.
24. W. Stumm and R. Wollast, Rev. Geophys., 1990, 28, 53.
25. J. S. Geelhoed, T. Hiemstra and W. H. Van Riemsdijk, Geochim. Cosmochim. Acta, 1997, 61, 2389–2396.
26. D. D. Hansmann and M. A. Anderson, Environ. Sci. Technol., 1985, 19, 544–551.
27. L. Charlet, N. Dise and W. Stumm, Agric., Ecosyst. Environ., 1993, 47, 87–102.
28. M. P. Vinod, K. Vijayamohanan and S. N. Joshi, J. Power Sources, 1998, 70, 103–105.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and additional tables and figures. See DOI: 10.1039/d0cc03863f

---