Effect of Ammonia on Pt, Ru, Rh, and Ni Cathodes During theAlkaline Hydrogen Evolution Reaction
Ramasamy Palaniappan and Gerardine G. Botte*
Center for Electrochemical Engineering Research, Chemical and Biomolecular Engineering Department, 165 Stocker Center, OhioUniversity, Athens, Ohio 45701, United States
*S Supporting Information
ABSTRACT: The effects of ammonia on the hydrogen evolution reaction (HER) with Pt,Rh, Ru, and Ni have been investigated using electrochemical methods. The activity of thecatalysts for HER, in the presence of ammonia follows the trend Rh > Pt > Ru > Ni. Pt andRu exhibit a decrease in the activity toward HER. Cyclic voltammetric (CV) studies showthat the presence of ammonia results in weaker metal−hydrogen underpotentialdeposition (M−Hupd) bonds. In addition, the activity of HER on Pt and Ru is retardedin the presence of ammonia, supporting the theory that the activity for HER is higher onthe metal surface with Hupd than on the bare metal. On the contrary, the activity for HERon Rh surface in the presence of ammonia is enhanced, suggesting the M−Hupd bond is abarrier to the HER. CV for Ni in the presence of ammonia shows a Ni(OH)2 reductionpeak. However, the overpotentials for HER on Ni are very high such that no significantdifference is observed in the activity of HER.
1. INTRODUCTION
The history of ammonia oxidation dates back to 1836 andpossibly beyond that.1 Recent developments in electrocatalystsfor ammonia electro-oxidation have led to the recognition ofammonia as a safe and clean hydrogen carrier for distributedpower generation.2−7 It facilitates a robust and economicalmethod to produce hydrogen as opposed to the traditional hightemperature cracking of ammonia and high-energy consumingwater electrolysis processes. The reactions for an ammoniaelectrolytic cell comprise of ammonia electro-oxidation at theanode and alkaline water reduction at the cathode as shown ineqs 1−3.8 Pt and Pt−Ir alloys have been identified as the mostsuitable electrocatalysts for the oxidation of ammonia.9−12 Theanodic electro-oxidation of ammonia has been identified as thelimiting reaction for this process, as it reaches a maximum at 0.6to 0.8 V versus reversible hydrogen electrode (RHE) onpolycrystalline Pt.12−14 On the other hand, the cathodichydrogen evolution reaction (HER), shown in eq 2, is knownfor its facile kinetics.
+ → + +
= −
− −e
E
Anode: 2NH 6OH N 6H O 6
0.77 V vs SHE
3 2 2
0 (1)
+ → +
= −
− −e
E
Cathode: 6H O 6 3H 6OH
0.82 V vs SHE
2 2
0 (2)
→ + =EOverall: 2NH N 3H 0.059 V3 2 20
(3)
In order to realize higher efficiencies it is required to operatethe electrolytic cell at voltages, such that the potential at theanode is maintained below its limiting value. It is desired thatcathode electrocatalysts facilitate high hydrogen production
efficiencies at such voltages. Previous studies have reported theoperation of an ammonia electrolyzer stack in a divided cellmode, where the anolyte containing ammonia and ammonia-free catholyte are fed in separate compartments.15,16 However,during the electrolysis of ammonia, traces of ammonia candiffuse into the cathodic compartment at the rate of 6.33 ×
10−8 mol cm−2 s−1.17 It has also been noticed that the pH of theanode compartment drops while that at the cathode compart-ment is found to increase during a single pass in an ammoniaelectrolyzer.16
Operating the electrolytic cell in a continuous stirred tanktype mode would help alleviate the problems associated withthe divided-cell electrolyzer. Therefore, it is critical that thepresence of ammonia does not provide any detrimental effecton the stability of the catalyst and its activity toward HER. Anunderstanding of HER kinetics on the electrocatalyst is vital tocomprehend the effect of ammonia on the HER for thatmaterial. HER has been a subject of interest for over twocenturies, since the invention of water electrolysis. Ever sinceHoriuti published on the matter,18,19 numerous researchershave independently studied the reaction and have madesignificant contributions to the mechanism of HER.20−25
HER has played an important role in the development ofelectrochemical principles, including the famous Butler−Volmer equation in 1924.26
A wide range of materials has been tested for their efficacy aselectrocatalysts for the HER in alkaline media, of which thegroup IV metals in the periodic table and their alloys have
received predominant attention.27−41 They are preferred tonoble metal based cathodes for the HER due to their relativelow-cost. Nevertheless, they have drawbacks owing to their easeof deactivation, high overpotentials, and low activity towardHER.42,43 A wide variety of noble metals (including Pt, Ir, Rh,Ru, Pd) and their alloys,27,44−50 oxides25,51 and perovskiteshave been evaluated for activity toward HER. Studies rangefrom theoretical modeling, such as quasi steady statecalculations,52 density functional theory studies,24 and volcanoplot (exchange current density for hydrogen evolution versusenthalpy of formation of M-H bond) simulations,21 toexperiments utilizing cyclic voltammetry (CV), linear sweepvoltammetry (LSV),27,53 constant potential, and electro-chemical impedance spectroscopy (EIS).54,55 In addition, ithas been reported that the HER is sensitive to theelectrocatalyst crystal structures.56,57
The HER proceeds through Volmer−Heyrovsky−Tafel(VHT) mechanism in both acidic and alkaline environ-ments.26,27,55 The mechanism in the alkaline medium isshown in eqs 4−6. The Heyrovsky (or electrochemicaldesorption) step and the Tafel (or chemical recombination)step are parallel and competing reactions. It is important toknow which of these steps will be rate-determining, as itdetermines the stability of the electrode. The extent of thereaction proceeding through one of the two steps should bedirectly related to the resistance of the other (i.e., the reactionwill proceed via the step with a lower resistance or a higher rateconstant), which can be evaluated by potential decay and EIS.
+ + ⇌ +− −Volmer: M H O e MH OH2 (4)
+ + ⇌ + +− −Heyrovsky: MH H O e M H OH2 2
(5)
+ ⇌ +Tafel: 2MH 2M H2 (6)
EIS is one of the standard techniques to estimate theimpedance of various components of an electrochemicalreaction and, therefore, rate constants for the steps involvedin the mechanism.54,55,58−60 Armstrong and Henderson derivedthe equivalent circuits for a two-step electrochemical reactionwith one intermediate in the absence of diffusion control byevaluating the impedance from their corresponding rateequations.61 Harrington and Conway extended this procedureto derive the impedance equations for the three-step HER andsubsequently verified the validity of the equations usingcomplex plane plots obtained from EIS experiments using Ptelectrodes in 0.5 M sodium hydroxide solution.54,62
However, the effect of ammonia on HER has not beenreported. This can be better understood by evaluating thekinetic parameters such as Tafel slope (TS), exchange current
density (i0), and rate constant, in the presence and absence ofammonia. Based on an extensive theoretical study, Greeley etal., concluded that Pt, Ir, Rh, Ru, Pd, Ni, and Co displayed highactivities for HER.24 In addition, Pt, Rh, Ru, and Ni continuedto show high activities for HER, when alloyed with q widerange of transition materials.54,62
Within this context, the objective of this paper is todetermine the effect of ammonia on the alkaline HER kineticson Pt, Ru, Rh, and Ni electrocatalysts using voltammetric,alternating current (AC) impedance, and current transienttechniques.24 It has been shown that potassium hydroxide(KOH) solution exhibits maximum ammonia electro-oxidationrate at a concentration of 5 M, therefore it was chosen for ourstudies. Carbon fiber paper (CFP) substrates were used forelectrodeposition of the catalysts as they demonstrated a betterperformance for the electrolysis in alkaline medium.4,9 In orderto address the main objective, the following tasks wereperformed:
1. Pt, Rh, Ru, and Ni electrodes were prepared byelectrodeposition onto CFP substrates. The electrodeswere then characterized using scanning electronmicroscopy (SEM), energy dispersive X-ray (EDX)spectroscopy, and X-ray diffraction (XRD) while theelectrochemical active surface area (EASA) was deter-mined using CV.
2. The kinetic parameters of the electrodes for HER inaqueous 5 M KOH were determined in the presence andabsence of ammonia using LSV and EIS techniques.
3. The effects of ammonia on long-term stability of theelectrocatalysts were evaluated using constant potentialexperiments.
2. MATERIALS AND METHODS
All of the chemicals used in this study were of analytical gradeand were obtained from Alfa Aesar. All the experiments wererepeated in order to check for reproducibility, and the errorsreported in the paper were calculated through errorpropagation, taking into account the instrumental error andstandard deviations in the measurements. All electrochemicalexperiments were performed with a potentiostat (Solartron1287) coupled with a frequency response analyzer (Solartron1252A) with an electrochemical interface. Hg/HgO saturatedwith KOH (20% by weight), purchased from Koslow Scientific,was used as the reference electrode for measurements inalkaline solution, whereas an Ag/AgCl electrode saturated withKCl, purchased from Fisher Scientific, was used in acidicsolutions. A glass-body luggin capillary was used to house thereference electrode in order to reduce the solution resistanceand avoid contamination of the solution.
Table 1. Electroplating Solution Composition, Operating Conditions, and Results for Constant Potential Electrodeposition ofPt, Rh, Ru, and Ni on Untreated Carbon Fiber Paper (Toray TGP-H-30) Substratea
material electrodeposition solution composition temp (°C)potential (V)vs Ag/AgCl
average current density(mA cm−2) total deposition time (s)
aElectrodeposition was carried out for all catalysts until a loading of 5 mg was achieved, except iridium for which there was no appreciable mass gainafter 2 h.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXB
2.1. Electrode Preparation and Physical Character-ization. The electrocatalysts used in this study were preparedby electrodeposition onto two untreated Toray TGP-H-30CFP, purchased from Electrochem Inc., of size 2 cm ×2 cmsandwiched on either side of an 18-mesh titanium gauzesubstrate of purity 99.9% (Alfa Aesar) as reported by Boggs andBotte.9 The substrate was then degreased with acetone andrinsed with deionized water, dried in an oven at 80 °C andweighed using a microbalance with an accuracy of 0.1 mg, priorto and after the electrodeposition. The electrodepositions wereconducted at constant potentials that were chosen based onmetal ion reduction peaks in the CV as described in theSupporting Information available online. The conditions for Pt,Rh, Ru, and Ni electrodeposition are detailed in Table 1. Theelectrodeposition experiments were conducted in a 250 mLPyrex glass beaker on a stirring hot plate with a built-inthermocouple and a glass-ceramic surface from Corning. Thepotentials are provided in Table 1. The contents of the beakerwere stirred at 100 rpm using a magnetic stirrer bar to maintainuniform temperature and minimize mass transfer limitations;the temperature was maintained using the thermocouple built-in with the hot plate. A 5 cm × 5 cm Pt foil of thickness 0.01cm and purity 99.999% (ESPI Metals) and an Ag/AgClelectrode saturated with KCl were used as the auxiliary andreference electrodes, respectively. The electrodepositions werecarried out until a loading of 1.25 mg cm−2 was obtained. TheCFP on one side of the sandwich was cut for physicalcharacterization of the electrode, which was used for surfacemorphology by SEM and elemental analysis by EDX studiesusing a JEOL JSM 6390 scanning electron microscope. Thesignature crystallographic orientation of the synthesizedelectrodes were studied using XRD on a Rigaku Ultima IVusing Cu Kα (1.54 Å) radiation generated with 40 kV and 30mA, at a scan rate of 0.5° min−1.2.2. Pretreatment and EASA Measurements. All the
electrodes were pretreated, prior to the electrochemical
measurement. Choice of a pretreatment procedure is depend-ent upon the nature of the electrocatalyst and the substrate andis crucial to obtaining reproducible results. The pretreatmentwas performed in two stages: first by performing a CV in 1 MKOH solution at 50 mV s−1 in the region where there is nooxide formation, until a sustained periodic state for thevoltammogram was achieved, so that any impurity that mighthinder the electrochemical measurement in the potential rangewould be removed. Second, the electrode was subjected to apotential of −1.42 V versus Hg/HgO in 1 M KOH for 1 h inorder to remove any surface oxides that might have formed onthe electrode, during the previous step or while drying.35 Steps1 and 2 were repeated until a reproducible surface wasobtained.The electrodes were then subjected to EASA measurements
followed by the CV, LSV, and EIS studies. The EASA for Ptand Rh were evaluated by measuring the hydrogen adsorptioncharge in 0.5 M sulfuric acid,63−66 while the EASA for Ni wasevaluated by the charge required to produce α-Ni(OH)2 in 1 MKOH using CV.67 The solution was deaerated by bubblingargon gas (99.999% purity, from Praxair) at 15 mL min−1 for 15min prior to the measurements.Carbon monoxide (CO) stripping voltammetry was used to
measure the EASA for Ru and to verify the EASA for Ptobtained via hydrogen adsorption.68,6970The CO strippingvoltammetry was carried out under a fume hood. The testsolution (1 M KOH for Pt and 0.5 M H2SO4 for Ru) wassaturated with carbon monoxide, of purity 99.9% from Praxair,which is evident from the open circuit voltage reaching steadystate. The electrode was held at a constant potential of 25 mVversus RHE in the respective test solution. The solution wasthen purged with argon gas for 15 min, while continuing tohold the electrode at the same potential. Thereafter, CV testswere carried out on the working electrode in the CO electro-oxidation region. The CV for EASA measurement was carriedout with a 5 cm ×5 cm Pt foil auxiliary electrode at a scan rate
Figure 1. EDX for metals Pt, Rh, Ru, and Ni are labeled A, B, C, and D respectively. The SEM images of the corresponding metals at 2500× areplaced exactly below the EDX images. It can be observed that the plating occurs in a globule-type fashion occurring uniformly along the lengths ofthe carbon fibers for all but Ru, where the deposition occurred in a more layered structure.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXC
of 50 mV s−1. The sustained periodic state for the voltammo-grams was obtained after 5 cycles.2.3. Electrochemical Characterization. The CV, LSV,
EIS, and constant potential experiments were carried out with a3-electrode setup in 5 M KOH solution in the presence andabsence of 1 M ammonia. A 5 cm ×5 cm Ni gauze (10 meshwoven from 0.1 mm diameter wire, Alfa Aesar) was used as thecounter electrode for each experiment. The Ni gauze wassandblasted using a micro sand blaster (Swam Blaster, CrystalMark, Inc.) with alumina (size 27.5 μm, purchased from CrystalMark, Inc.) for 2 min, then degreased with acetone and rinsedwith deionized water.The CV tests were carried out by scanning the potential from
−1.1 to 0.1 V versus Hg/HgO and back at 10 mV s−1. Whilethe LSV experiments were carried out by scanning the potentialbetween −0.9 V to −1.3 V versus Hg/HgO at 1 mV s−1. Thelow scan rate of 1 mV s−1 was sufficient to maintain the reactionunder kinetic-control by allowing enough time for the diffusionof fresh reacting species to the electrode surface. Sustainedperiodic results were obtained for both CV and LSV after thefifth scan.The EIS experiments were performed by applying an AC
amplitude of 10 mA with a frequency range between 10 kHzand 0.1 Hz superimposed over a DC potential −1.1 V versusHg/HgO in the HER region. The electrodes were polarized atthe DC potential for 15 min prior to the EIS measurements.Constant potential experiments were carried out in a solution
of 5 M KOH in the presence and absence of ammonia at apotential of −1.1 V versus Hg/HgO for 3 h.
3. RESULTS AND DISCUSSION
3.1. Electrodeposition and Physical Characterization.Pt, Rh, Ru, and Ni were successfully deposited on the CFPelectrodes and verified by an increase in the electrode mass aswell as by SEM and EDX studies. The results for theelectrodeposition (average current and catalyst loading) arereported along with the electrodeposition solutions andconditions in Table 1. SEM and EDX results for the electrodesafter electrodeposition are reported in Figure 1. The SEMimages for Pt, Rh, and Ni show globule type deposition withthe size of each globule in the range 10 to 100 nm andpertaining to carbon fibers in the first two layers, consistentwith the literature. On the other hand, the SEM image for Rushows a layer type deposition. The EDX images show that theelectrodes are predominantly comprised of respective metalalong with carbon from the substrate and trace amounts ofoxygen that occurred as a result of exposure to atmosphere.The slightly higher oxygen content on the nickel electrodecould be attributed to the existence of hydroxides.In order to effectively compare the intrinsic electrochemical
activity of an electrocatalyst for a reaction, it is crucial to
normalize the current obtained for the material with its EASA(the area of the electrode material that is active to anelectrochemical reaction). The EASA for the catalysts wereevaluated as mentioned in the Supporting Information availableonline. Table 2 provides the results of the EASA, the specificarea of the catalysts, and the method adopted to evaluate theEASA. The EASA for Pt obtained using CO strippingvoltammetry agreed with that obtained using the hydrogenadsorption method, providing credence to the technique used.It can be observed that Ni exhibits an EASA similar to that ofPt. However, Ru displayed the highest EASA and roughness,consistent with the SEM images, which indicate a rough layertype deposition that is equally prevalent in the gaps betweenthe carbon fibers. The high EASA for Rh is attributed to thesmaller particle size of the deposits. The results reported in thisstudy, unless mentioned otherwise, have been normalized withrespect to the EASA.
3.2. Voltammetry Studies. Figure 2 illustrates the Tafelplots for the catalysts under investigation in 5 M KOH in the
presence and absence of 1 M ammonia. It can be observed fromFigure 2 that the noble metals exhibit two different TS’s in theregion below and above −200 mV. However, for Ni, twodifferent TS’s were measured in regions below and above −240mV. This is in agreement to that observed by Vilekar et al.,based on quasi-steady state theory.52 The TS in the low
Table 2. Electrochemical Active Surface Area for Pt, Rh, Ru, and Ni Electrodeposited on Carbon Fiber Papera
Ni Ni→Ni2+ 116833.5 514 227.30 56.83 47.35aErrors reported were calculated through error propagation using instrumental errors for current and voltage. Ru had the highest surface area amongthe electrodes prepared.
Figure 2. Tafel plots for steady state linear sweep voltammograms forPt, Rh, Ru, and Ni in 5 M KOH in the presence and absence of 1 Mammonia. Voltammograms without ammonia are shown by dashedlines (--) while solid lines (−) represent the presence of ammonia.Inset to figure represents the data in the region of interest −110 mV to−250 mV for enhanced clarity. Current densities (ireal) were evaluatedbased on the electrochemical active surface area. Voltammograms werecarried out at a potential scan rate of 1 mV s−1.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXD
overpotential (<200 mV) region has been attributed to thereaction proceeding through the Volmer−Tafel pathway, whilethe TS at high overpotential (>200 mV) is due to the Volmer−Heyrovsky path.52 The Tafel constants were estimated byfitting the LSV data to the Tafel equation after correcting forthe IR drop using linear regression with a 95% confidence level.The HER kinetics at the low overpotential region is of
particular interest to ammonia electrolysis as one tries toachieve maximum hydrogen production at low cell voltages andprevent the anode of the electrolytic cell from reaching itslimiting potential. In addition, it was observed that at highoverpotentials the difficulty for hydrogen to escape the wetsurface of the untreated CFP electrode was enhanced.Consequently, it is also observed that the high TS thataccompanies the high i0 at high overpotentials, implying thatthe HER currents in this region is masked by diffusion.Therefore, only the Tafel parameters (TS and i0) calculated atthe lower overpotential region (between −118 mV and −200mV) will be used in this study.The Tafel constants along with the overpotentials at current
density of 0.25 mA cm−2 (corresponding to a hydrogenevolution rate of 1.73 l h−1 for an electrode with EASA 1 m2)are reported in Table 3. The overpotential at 0.25 mA cm−2,
without ammonia, follows the order Pt < Ru < Rh < Ni,whereas the exchange current densities followed the trend Pt >Rh > Ru > Ni in this region. In the absence of ammonia, Ptdisplayed the least overpotential and highest i0 among thecatalysts studied, confirming the highest activity of Pt for HERas seen from the volcano plots.The TS obtained for the catalysts in 5 M KOH in the
absence of ammonia in the low overpotential region increasedin the order Ni < Ru < Pt < Rh. The TS for the noble metalcatalysts follow the trend observed by Fournier et al., for thesame catalysts embedded in a porous LaPO4 matrix in 1 MKOH.71 Lower TS indicate higher hydrogen evolution rate foran increase in overpotential. Although Ni has the lowest TS, italso has the highest overpotential at a current density of 0.25mA cm−2 as well as an i0 at least 100 times lower than the othercatalysts. The higher overpotentials exhibited by Ni electrodescould be due to the existence of hydroxides even at potentialsas low as −1.0 V versus Hg/HgO confirmed by Silverman inthe revised EMF-pH diagrams for the Nickel−Water system.72
The existence of hydroxides results in a high activity for theelectrochemical dissociation of water. However, it also exhibitsa considerably low activity for the desorption steps, resulting inthe formation of hydrides thereby leading to the deactivation ofNi.It can also be observed from Figure 2 that, in the presence of
ammonia, the overpotential-log i behavior for the catalystscontinue to exhibit a low TS in the low overpotential region,and a high TS in the high overpotential region. This behavior issimilar to that without ammonia, implying that the VHTmechanism still applies to the HER in the presence of 1 Mammonia. It can be noticed that the overpotentials for Pt, Ru(inset to Figure 2) and Ni (Figure 2) are higher due to a partialsurface deactivation reaction in the presence of ammonia.71
However, the overpotentials observed for Rh with 1 Mammonia is lower than that observed without ammonia; thiscould be attributed to surface activation thereby increasing theHER on Rh. From Table 3 it can be noted that theoverpotentials for HER at 0.25 mA cm−2 in 5 M KOH with1 M ammonia follows the trend: Rh < Pt < Ru < Ni.The cyclic voltammograms for Pt, Rh, Ru, and Ni in 5 M
KOH with 1 M ammonia and ammonia-free solutions in theregion preceding the HER, shown in Figures 3A through 3D,provide a deeper insight to the surface process. It is known thatcatalyst utilization and the kinetics for HER are dependentupon the surface coverage of the overpotential depositedhydrogen (Hopd) which is weakly bonded to the catalyst surfaceand is considered to be the actual intermediate for theHER.23,54,55 Hopd is in turn dependent upon the potential forthe hydrogen underpotential deposition (Hupd), i.e., thehydrogen that is adsorbed at potentials lower than equilibriumpotential (E0) as shown in eq 7. It is agreed upon that the Hupd
is strongly bound to the catalyst surface.23,54
+ + ⇌ + <− − E EM H O e MH OH2 upd
0(7)
where, E0 = 0 V versus SHE. The potential of occurrence forthe Hupd is indicative of the strength of the M−Hupd bond, i.e.,the more positive it is, relative to the HER, the stronger theM−Hupd bond will be.23,73,74 However, the role of the Hupd onHER is a controversial subject, with opposing views: Markovicet al.,23 reported that the strong Hupd hinders the HER, whileConway et al. claims that the presence of Hupd enhances therate of HER.20,73,74 Coincidentally, by evaluating the shift in theHupd peaks that arises due to the addition of ammonia andrelating the shifts in potential for Hupd formation to the changesin HER kinetics provides important information to understandthe role of Hupd.In the absence of ammonia, the CV for Pt (Figure 3A) shows
three reversible peaks A01, A02, and A03 associated with Hupd
in the Pt (111), Pt (100) planes and the reduction of thesurface oxide or OH− ion desorption for Pt, respectively.23,75
However the CV for Rh and Ru, in the absence of ammoniawere similar.45,69 The cathodic peaks B01, B02 (Figure 3B) forRh and peaks C01, C02 (Figure 3C) for Ru can be attributed tothe Hupd and the reduction of the metal-oxide that was formedduring the forward scan, thereby implying that the over-potential for Hupd follows the trend Pt > Rh > Ru. On the otherhand, the CV for Ni indicates a small plateau type feature atD01, attributed to the reduction of the Ni(OH)2.
76,77 However,the absence of a distinct reduction peak indicates thetransformation of Ni(OH)2 from the reversible α-phase to airreversible β-phase. The presence of oxides for Ni at such low
Table 3. Tafel Parameters Obtained from Linear SweepVoltammetric Studies for Pt, Rh, Ru, and Ni Electrodes in 5M KOH in the Presence and Absence of Ammoniaa
material
ammoniaconcentration
(M)
Tafel slope(±3 mVdecade−1)
exchange currentdensity (±0.002
mA cm−2)
overpotential,η, iesa = 0.25mA cm−2
(±1 mV)
Pt 0 −157 0.043 −125
1 −171 0.031 −152
Rh 0 −220 0.040 −162
1 −219 0.069 −125
Ru 0 −140 0.021 −147
1 −222 0.047 −159
Ni 0 −95 0.0005 −255
1 −108 0.001 −260aThe Tafel slope and exchange current density were calculated forcathodic HER at the low and high overpotential regions separately.The exchange current density and the overpotentials at 0.25 mA cm−2
were evaluated based on the EASA of the catalyst.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXE
potentials can be verified by the Pourbaix diagram and obviatesthe existence of Hupd.
72
The voltammograms in the presence of ammonia displaythree distinctive features: (1) the absence of the metal-oxidereduction, (2) the significantly larger cathodic currents in theHupd region between −0.4 V to −0.9 V, and (3) the presence ofan additional reduction peak. The absence of the metal oxidedesorption peak indicates the removal of oxides by taking partin the ammonia electro-oxidation reaction or its masking by thesignificantly larger ammonia electro-oxidation currents.For Pt, the potential scan in the positive direction shows a
peak A14 corresponding to the adsorption of ammonia andanother at −0.25 V corresponding to the onset of ammoniaelectro-oxidation. Upon evaluating the charges under thereversible peak A11−A11′, after accounting for the doublelayer and desorption of ammonia, this agrees with the chargerequired for Hupd in the absence of ammonia, implying that thepeak A11 is due to Hupd. The additional reduction peaks A12and A13 can be attributed to the reduction of the intermediatesof ammonia oxidation and ammonia desorption as verified bystripping studies by de Vooys et al.12 and differentialelectrochemical mass spectroscopy studies by Wasmus et al.78
In the presence of ammonia, Rh and Ru did not exhibit anydistinct ammonia electro-oxidation peaks during the potentialscan in the positive direction. This has been attributed to
surface blockage by intermediates from chemical or a minorelectrochemical oxidation of ammonia along with the OH− ionsthat were adsorbed at considerably lower potentials.12 This canbe verified by the presence of an additional reduction peaksB12 and C12 for Rh and Ru, respectively. The additionalreduction peaks B12 and C12 and larger Hupd peaks B11 andC11 could be attributed to either the ammonia desorption or ashift in the surface-oxide reduction peak in the negativedirection or both. In addition, the presence of ammoniaindicates no metal-oxide reduction peaks for Rh and Ru.In addition, it could be observed that the Hupd peaks for Pt
A01 and A02 appear to have slightly shifted by 120 mV to 180mV in the negative direction, implying a weaker Pt−Hupd bond.The shift in the Hupd to a more negative potential could be aresult of adsorbed ammonia or intermediates for ammoniaelectro-oxidation. Similarly, the presence of ammonia causesthe Hupd peaks B01 for Rh and C01 for Ru, to be shifted in thenegative direction by 40 mV and 30 mV, respectively, implyinga slightly weaker M−Hupd bond. The extent of the shift appearsto be caused by the sensitivity of electrocatalyst surface toammonia.The CV for Ni Figure 3D in the presence of ammonia
indicates peak D11 corresponding to the reduction of Ni(OH)2to Ni.76,79 This suggests a reversible nature of the hydroxideand could in turn imply that the presence of ammonia prevents
Figure 3. Cyclic voltammograms for (A) Pt, (B) Rh, (C) Ru, and (D) Ni in 5 M KOH with the presence and absence of 1 M ammonia. (−) Solidlines indicates presence of ammonia while the (--) dashed lines indicate the absence of it. The voltammograms shown above are representative of thesteady state cycle achieved at the 5th cycle and at a potential scan rate of 10 mVs−1. It can be seen that in the presence of ammonia all the Hupd peakis more pronounced and occurs at a slightly more negative potential for all the catalysts.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXF
the conversion of the hydroxide from the reversible α phase toits irreversible β phase. The Figure also shows that theammonia adsorption or electro-oxidation peaks are absent,since they do not occur until considerably higher potentialsthrough a direct electron transfer reaction with the NiOOHcatalyst.76
Upon adding ammonia and by correlating the shift in theHupd with the change in the overpotential of the catalyst, therole of Hupd on the HER could be understood. The addition ofammonia results in weaker M−Hupd bonds for Pt, Rh, and Ru.The weaker M−Hupd bond (Figures 3A and 3C) isaccompanied by a decrease in the HER activity on Pt and Ruin the presence of ammonia, thereby implying that the activityfor HER is higher on a surface with Hupd than on the baremetal.20,73,74 This hypothesis can also be confirmed from theshift in Hupd (i.e., an Hupd of −140 mV results in anoverpotential decrease of 27 mV for Pt, whereas a Hupd shiftof −40 mV results in an overpotential decrease of 11 mV).Thereby validating the hypothesis by Conway et al. that theactivity of HER is thermodynamically more favorable on thesurface with the stronger Hupd, rather than on the bare metalitself.74 However for Rh, an increase in activity toward HER isobserved with the weaker M−Hupd bonds. The addition ofammonia hinders the formation of the Hupd bonds on Rhthereby leaving room for the formation of Hopd, which isreported to enhance the HER on the Rh surface. This suggeststhat a HER on Rh could follow a different mechanism. Apossible mechanism for HER on Rh might consist of the asurface where Hupd bonds act as a barrier to HER and Hopd
being the actual intermediate for HER, as proposed byMarkovic et al.23
3.3. EIS Studies. The EIS technique can be used to identifythe rate constants for each step in the reaction mechanism.Figure 4 shows the various equivalent circuit models that wereevaluated to describe the HER.58,59,71 The procedure for
estimating the best equivalent circuit for the impedance data isexplained in the Supporting Information to this article (foundonline).
3.3.1. Effect of Ammonia on Equivalent Circuit Parame-ters for HER. A list of equivalent circuit models that wereevaluated for HER is described in Figure 4A−E. TheArmstrong−Henderson model61 (Figure 4A) describes atypical case for HER, the model is usually used in a scenariowhere the Nyquist plots indicate the presence of two timeconstants. The constant phase element (CPE) model (Figure4B) is an extension to the Randles circuit and a simplifiedversion of the Armstrong−Henderson model, where theimpedance of the parallel Rp−Cp branch is eliminated due tothe considerably low pseudocapacitance. The CPE model isdesigned to fit systems with one time constant and frequencydispersion.80 Additionally, in the derived circuits (Figure 4Cthrough 4E), a CPE element replaced the capacitor in theArmstrong−Henderson model, were evaluated.The EIS results for Pt, Rh, Ru, and Ni are presented in
Figures 5A through 5D, in the form of Nyquist (or) Cole−Cole(or) complex-plane plots along with its equivalent circuit fit andthe fit obtained from the rate constants derived for the VHTmechanism (Section 3.4). As a rule of thumb, the highfrequency intercept in the Nyquist plot represents the solutionresistance, while the diameter of the semicircle represents theFaradaic resistance. Harrington and Driessche81 give anexcellent description on how to decipher an impedance spectraand the significance of the equivalent circuit elements forspecific cases. The Nyquist plots indicate the presence of onlyone time constant and no Warburg-type feature at the lowerfrequency range, indicating that the HER is charge transfercontrolled. Nevertheless, due to its practical significance toHER, the Armstrong−Henderson model was evaluated to fitthe impedance spectra. Figures 5A to 5C show that the CPEmodel offered the best possible fit for HER on the noble metals.The CPE accounts for the capacitance dispersion withfrequency as a result of the inhomogeneity of the electrodesurface,58,71,80 while Figure 5D shows that the Armstrong−Henderson model offered the best fit for the Ni electrodes,confirming that found in the literature.58,61 The derived circuitsdid not offer any significant improvement in the fit whencompared with the CPE model, suggesting a faster recombi-nation reaction. The total impedance for all the models inFigure 4 obeys the equation:
= + +− − −Z R Z Z( )T s f
1CPE
1 1(8)
where Rs is solution resistance, Zf is faradaic impedance, andZCPE is impedance due to the CPE. Since, all of the models arebased on the Armstrong−Henderson equivalent circuit, theFaradaic impedance can be expressed as
ω= + +Z R j C R1/( (1/ ))f ct p p (9)
where Rct is charge transfer resistance, Rp is polarizationresistance, Cp is pseudocapacitance, and ω is angular frequency.The equation for the impedance of the CPE element is given by
ω=ϕZ T j1/[ ( ) ]CPE (10a)
= +ϕ ϕ− − −T C R R[ ]dl s
1ct
1 (1 )(10b)
where Cdl is the double layer capacitance, and Φ is the anglethat describes the extent of frequency dispersion of the doublelayer capacitance. The equivalent circuit fitting suggests that the
Figure 4. Equivalent circuit models evaluated for the representation ofHER impedance spectra at E = −1.161 V vs Hg/HgO. Figures Athrough E depict: Armstrong−Henderson model (A-H), ModifiedRandles circuit (CPE model), A-H model with capacitance dispersionat a high frequency, A-H model with capacitance dispersion at lowfrequency, and A-H model with capacitance dispersion at low and highfrequency domains, respectively. Rs − solution resistance, Cdl − doublelayer capacitance, Rf − Faradaic resistance, Rct − charge transferresistance, Rp − polarization resistance or pseudoresistance, Cp −
pseudocapacitance. Constant phase element (CPE), CPEdl and CPEp
accounts for frequency dispersion in double layer capacitance andpseudocapacitance, respectively.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXG
pseudocapacitance term is negligible for the noble metals and
obeys the CPE model, thereby implying that the H adsorption
formed as a result of the Volmer step quickly recombined to
form hydrogen, and in turn suggesting that the water
dissociation step is the slow step. However for Ni, the
pseudocapacitance offers significant impedance toward the
HER, thereby following the Armstrong Henderson model. The
procedure adopted for optimal equivalent circuit fitting is
described in the Supporting Information and supplemental
figure (Figure S4) to this paper that is found online.
Figure 5. Nyquist (or) complex-plane plot for the experimental impedance spectra for HER obtained with electrocatalysts: Pt (A), Rh (B), Ru (C),and Ni (D) at E = −1.102 V vs Hg/HgO (η = −166 mV) in the presence and absence of ammonia. In addition, the spectra for equivalent circuitmodels (described in inset to Supplemental Figure S4 to this paper found online) that best fit the experimental values are shown along with thespectra obtained by using the rate constants evaluated for the VHT mechanism. The experimental data for HER in the absence and presence ofammonia are represented by clear (red) and filled (blue) square markers, respectively.
Table 4. Circuit Parameters for the Equivalent Circuit Fit to the Impedance Spectra for the HER on Pt, Rh, Ru, and Ni in thePresence and Absence of Ammonia at an Overpotential of −166 mVa
Material Ammonia Concentration (M)
Goodness of Fit
Rs (Ω) CPEdl- T CPEdl- P Cdl (mF cm‑2) Rf (Ω cm2) CP (mF)Chi-Square Sum-Square
Pt 0 0.012 1.211 0.039 0.136 0.663 50 31.06 -
1 0.006 0.602 0.046 0.108 0.671 42 44.07 −
Rh 0 0.005 0.412 0.065 0.108 0.810 70 35.82 -
1 0.003 0.286 0.054 0.129 0.793 74 36.30 −
Ru 0 0.007 0.549 0.050 0.174 0.755 44 37.68 -
1 0.012 0.831 0.039 0.192 0.722 36 51.50 −
Ni 0 0.012 0.837 0.023 0.005 1.000 20 498.75 15.69
1 0.012 1.000 0.022 0.006 1.000 28 553.43 16.81aThe CPE model offered the best fit the impedance data for Pt, Rh, and Ru. However the Armstrong−Henderson model offered the best fit for theimpedance data for the Ni electrode. It can be observed that Pt offered the least Faradaic resistance in the absence of ammonia, while Rh offered theleast resistance in the presence of ammonia. The trend in the Faradaic resistance follows the trend in overpotentials in accordance with the LSVstudies in the presence and absence of ammonia.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXH
The values for the circuit elements for the equivalent circuitfit for the impedance spectra at −166 mV, along with thegoodness of fit (χ2 and sum squared values) are reported inTable 4. The maximum average error for the fit, estimated fromthe sum squared values for 45 data points, for the noble metal is0.016 Ω obtained using the CPE model, whereas the averageerror for Ni is 0.043 Ω obtained using the Armstrong−Henderson model. The results for other models exhibitedhigher errors and therefore were not reported. From Table 4, itcould be observed that there is no significant difference in thesolution resistance, whereas the double layer capacitance perunit area decreased in the order Rh > Pt > Ru > Ni in theabsence and presence of ammonia. In addition, it is noticed thatthe double layer capacitance is greater than the expected 20 mFcm−2, for all the noble metals. The higher Cdl values is specificto the alkaline medium and specifically to the coadsorption ofthe OH− anion in the double layer region.23 On the other hand,the Faradaic resistance per unit area increased in the order Pt <Ru < Rh < Ni, as observed for the trend in overpotentialsobtained from LSV results. The addition of ammonia did notresult in any change in the trend for the double layercapacitance per unit area. Meanwhile the Faradaic resistance(Rf) increased for Pt, Ru, and Ni, altering the trend in Rf to Rh< Pt < Ru < Ni. The results are consistent with the LSV studies,confirming their validity. The higher activity on Rh can beattributed to (1) the weakening of the surface hindering Rh−Hupd bonds,23 or (2) the high tendency to adsorb OH− ionsand therefore its reduced activity for Hupd formation45 or (3) itsselectivity of reducing nitrates and its intermediates, or acombination of the three. Further study is required to explainthe higher activity for Rh toward HER in the presence ofammonia. Evaluating the variation in the Hopd surface coverageand rate constants for the Volmer, Heyrovsky, and Tafel stepsin the presence of ammonia could offer a deeper understandingto the effect of ammonia on the HER.3.3.2. Effect of Ammonia on Kinetic Rate Constants for
HER and Hopd Surface Coverage. Upon assuming Langmurianadsorption, and neglecting the backward reactions for theHeyrovsky and Tafel steps, the final expression for the rateequations for reactions in eqs 4−6 is given by:
υ θ β η θ β η
θ θ
= − − − −
= − −
−
−
k f k f
k k
(1 ) exp( ) exp((1 ) )
(1 )
1 1 1 1 1
1 1 (11a)
υ θ β η θ= − = k f kexp( )2 2 2 2 (11b)
υ θ= k3 32
(11c)
where k1, k−1, k2, and k3 are the potential independent rateconstants; η is the overpotential; θ is the fractional surfacecoverage of Hopd species; β1 and β2 are the transfer coefficients.
where r0 is the rate of electron consumption, r1 is the rate ofsurface coverage, and q is the charge required for the formationof a monolayer of Hopd. The expression for fractional surfacecoverage at a steady state can be obtained by solving thequadratic equation, from eq 11b, for θ
θ = − + +
+ + + + +
−
−
k k k
k k k k k k
[( )
( ) 8 ]/4
1 1 2
1 1 22
1 3 3 (13)
Upon linearizing eqs 12a and 12b; applying the phasors tointroduce the AC component and eliminating θ, one obtains
ω= = + +Y Z A B j C1/ /( )f f (14)
where
η= − ∂ ∂ θA F r( / )0 (14a)
η η= − ∂ ∂ ∂ ∂θ θB F q r r( / )( / ) ( / )20 1 (14b)
η= − ∂ ∂ θC F q r( / )( / )1 (14c)
The expressions for the equivalent circuit elements from theEIS experiments in relation to the HER can be obtained bycomparing eqs 14 with 9,54
=R A1/ct (15a)
= − +R A C B A1/(( / ) )p2
(15B)
= −C A B/p2
(15c)
The HER kinetics is usually described by four independentkinetic parameters (k1, k−1, k2, k3) and the fractional surfacecoverage θ as described in eqs 11a−11c and eq 13 respectively.β1, and β2 was assumed to be 0.5. However, EIS experimentsyield only 3 values for Rct, Rp, and Cp for eqs 15a−15c. The
Table 5. Rate Constants and Surface Coverage Were Estimated for the Catalysts in the Presence and Absence of Ammonia at anOverpotential of −166 mVa
MaterialAmmonia Concentration
(M)
Rate Constant (mol cm‑2 s‑1)Surface Coverage
q(μC cm‑2 )
Fractional Coverage θ(monolayers)
k1(x10‑10)
k‑1
(x10‑6)k2
(x10‑8)k3
(x10‑6)kav
(x10‑10)
Pt 0 3.22 2.47 0.42 0.25 9.89 9.05 0.043
1 2.26 2.10 0.35 0.25 7.01 7.58 0.036
Rh 0 2.64 0.98 1.29 1.71 8.54 4.99 0.023
1 2.76 0.50 1.16 0.80 8.91 5.50 0.025
Ru 0 2.77 0.24 1.33 0.45 8.96 4.41 0.021
1 2.02 0.19 1.13 0.42 6.57 3.78 0.018
Ni 0 30.67 3.73 0.0002 0.0029 0.06 28.69 0.112
1 32.74 3.28 0.0001 0.0022 0.05 34.43 0.134aRate constants and surface coverage were derived for the catalysts using the equivalent circuit parameters shown in Table 4 by the Levenberg−Marquardt algorithm for the Volmer−Heyrovsky−Tafel mechanism. The addition of ammonia leads to a decrease in the rate constant for the sloweststep, k1 for Pt and Ru, and k2 for Ni. However, for Rh, the addition of ammonia results in an increase in the rate constant for the Volmer step (k1).
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXI
value for i0 obtained from LSV studies was used in theexpression for the average rate constant described in eq 16. Thesolutions to the rate constants were obtained by maintaining θconstant.
β η= + = = − + − − − −
k k k i F f k k1/( ) /2 exp( )/av 11
21
0 1
1
2
1
(16)
The potential independent rate constants for the reactions4−6 were estimated, by fitting the circuit elements fromimpedance measurements to eqs 15a−15c, derived from theimpedance equation for the VHT mechanism. The Levenberg−Marquardt algorithm was used for the nonlinear fitting with atolerance level of 1 × 10−14. The estimated rate constants werethen used in the impedance eq 14 to re-evaluate the impedancespectra in the interested frequency range. The surface coveragewas varied until the rate constants were able to closely predictthe impedance spectra. These potential independent rateconstants and surface coverage are reported in Table 5.From Table 5, at an overpotential of −166 mV, the trends in
the average rate constants (kav) in the absence and presence of1 M ammonia agreed with the results obtained using the LSVstudies. In addition, for all the noble metals the rate constantfor the Volmer or the water dissociation step (k1) is lower thanthat of the desorption steps. However, the rate constant for thewater dissociation step (k1) for the Ni electrode is higher thanthat of the desorption steps, as expected, thereby causing a lowsurface coverage for the noble metal electrodes, and a highersurface coverage for Ni as mentioned in Table 5. Amongdesorption steps, the rate constant for the nonelectrochemicalTafel step was higher than the electrochemical Heyrovsky stepfor all the catalysts at −166 mV. In addition, Pt presents thehighest rate constant for the slow Volmer step, explaining itssuperior activity.The addition of ammonia results in a decrease in the average
rate constant (kav) for HER, which is observed for Pt, Ru, andNi (a schematic representation of the mechanism is provided inFigure S5). Moreover, the presence of ammonia results in thereduction of the rate constant k1 for the slow Volmer step for Ptand Ru, accompanied by a drop in the rate constant for thedesorption steps thereby resulting in a lower Hopd surfacecoverage. This is evident for Pt and Ru, which in turn cause adecline in the average rate constant and activity for HER. Onthe other hand, an increase in k1 was observed for Rh, whichexplains the rise in the Hopd surface coverage and therefore theincrease in the average rate constant and activity for HER.For Ni, although an increase in k1 and Hopd surface coverage
is observed with the addition of ammonia, this results in areduction in the rate of the slow rate-determining Tafel step.This decrease in the rate constant of the rate determining Tafelstep explains the Hopd higher surface coverage and loweractivity toward HER for Ni.3.4. Constant Potential Studies. Figure 6 shows the
current density−time plots at a potential of −1.1 V versus Hg/HgO for the CFP based Pt, Rh, Ru, and Ni electrocatalysts in 5M KOH, in the presence and absence of ammonia. A highsurface area Ni gauze was used as the auxiliary in place of a Ptfoil counter to avoid the depletion of ammonia through electro-oxidation. The current densities shown were evaluated basedon the EASA. All four catalysts present sufficient stability at theapplied potential in the presence and absence of ammonia.From Figure 6, Ni has the lowest cathodic current densityirrespective of the presence of ammonia. In addition, it is alsoobserved that Pt continued to display the highest current
density in the absence of ammonia despite losing 0.38 mA cm−2
after 3 h, confirming to the LSV and EIS studies. Although, theaddition of ammonia resulted in a minimum current loss of 0.1mA cm−2 for Pt and Ru, the catalysts appear to be stable. Onthe contrary, an increase in the current density with time isobserved for Rh. This performance is very similar to theperformance of a deactivated Rh electrode in 1 M KOH.82
Wrona et al. attribute this deactivation to traces of unknownmetal impurities on the electrode. The authors also reportedthat the surface of Rh electrodes can be purified (removesurface blockage species) during HER in alkaline media bycycling the potential of the electrode between −1.0 and 0.3 V vsa Hg/HgO reference electrode. On the contrary, an increase inthe current density with time is observed for the Rh electrodein the presence of ammonia. This means that the presence ofammonia in the electrolyte causes a self-purification of thesurface of the Rh electrode. It is hypothesized that thedeactivation of the surface of the Rh electrode is due to theformation of a Hupd layer on Rh, which acts as a barrier toHER. In addition, the higher HER current in the presence ofammonia could be attributed to the weaker Rh−Hupd bondformation, evident from CV measurements (as seen in Figure3B), which tends to easily break off from the surface therebycreating fresh active sites for HER. The rather slow reactivationcould be due to the constant competition from Hupd formationat −1.1 V, specific only to HER sites, resulting in a delay inreaching the steady state. The studies confirm highest activitytoward HER on Rh in the presence of ammonia as expectedfrom the LSV and EIS results.
4. CONCLUSION
A study of the effect of ammonia on the HER on Pt, Rh, Ru,and Ni electrodes was performed. The following conclusionswere made based on the results obtained:
1. The LSV, EIS, and chronoamperometric results for HERin ammonia free alkaline solutions were similar to thatreported in the review on HER. The activity of thecatalysts for HER followed the order Pt > Ru > Rh > Ni.
2. The presence of ammonia results in a slight loss in theHER activity for Pt, Ru, and Ni. However Rh displayed a
Figure 6. Current density-time plot for constant potential studies forPt, Rh, Ru, and Ni in 5 M KOH in the presence and absence ofammonia at −1.1 V versus Hg/HgO for a period of 3 h. It can beobserved that the addition of ammonia resulted in lower reductioncurrent densities for all materials except Rh. Rh also displayed themost stable and highest activity toward HER in the presence ofammonia.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXJ
higher activity toward HER in the presence of ammonia,resulting in the trend in activity Rh > Pt > Ru > Ni.
3. EIS results confirm that the noble metal catalysts have aslow water dissociation step and a fast recombinationstep, while the vice versa was observed for Ni. Theaddition of ammonia resulted in a decrease in the averagerate constant of the slow step for Pt, Ru, and Ni.However, the opposite behavior was observed for Rh.
4. EIS results also indicate that the presence of ammoniacauses a decrease in the rate constant k1 of the rate-determining Volmer step for Pt and Ru and k3 for Ni,thereby explaining the decrease in their activity towardHER. Whereas, the addition of ammonia results in theincrease in k1 substantiating the increase in Hopd surfacecoverage and in-turn the HER activity on Rh.
5. The addition of ammonia resulted in the weakening ofthe metal Hupd bond accompanied by a decrease in theactivity of the catalysts toward HER for Pt and Ru.Thereby supporting the hypothesis that HER is moreactive on the surface with hydrogen than the surfacecontaining the bare metal. On the other hand, anincrease in the activity toward HER was observed for Rh,suggesting that the HER on Rh is higher on the baremetal surface.
In conclusion, Rh offers the highest activity toward HER inthe presence of ammonia among the catalysts evaluated.Further studies need to be performed to evaluate the effect ofammonia on the activity of HER for bimetallic alloys.
■ ASSOCIATED CONTENT
*S Supporting Information
A full description of the procedure used to determine theelectrodeposition potential, EASA, along with the estimation ofrate constant mechanism by evaluating the equivalent circuitparameters to represent the impedance spectra for the HER inthe presence and absence of ammonia is given. The CVs usedto determine the electrodeposition potential and EASA, XRDpatterns for the synthesized electrocatalysts, and flowchart usedfor estimating the rate constants are provided. Finally, a cartoondepicting the proposed mechanism of HER in the presence andabsence of ammonia on Pt, Ru, and Ni is given. Thisinformation is available free of charge via the Internet athttp://pubs.acs.org.
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
The authors would like to acknowledge the financial support ofthe Center for Electrochemical Engineering Research (CEER)at Ohio University and the Department of Defense through theU.S. Army Construction Engineering Research Laboratory(W9132T-09-10001). The content of the information does notreflect the position or the policy of the U.S. government. XRDmeasurements were performed at CEER. The authors wouldalso like to thank Dr. Howard Dewald and Dr. MadhivananMuthuvel for their feedback and valuable discussions.
■ REFERENCES
(1) Daniell, J. F. On the Electrolysis of Secondary Compounds.Philos. Trans. R. Soc. London 1839, 129, 97−112.(2) Botte, G. G. Electro-Catalyst for Oxidation of Ammonia inAlkaline Media. U.S. Patent 7,803,264, 2010.(3) Bonnin, E. P.; Biddinger, E. J.; Botte, G. G. Effect of Catalyst onElectrolysis of Ammonia Effluents. J. Power Sources 2008, 182, 284−290.(4) Boggs, B. K.; Botte, G. G. On-Board Hydrogen Storage andProduction: An Application of Ammonia Electrolysis. J. Power Sources2009, 192, 573−581.(5) Botte, G. G.; Vitse, F.; Cooper, M. Electrocatalysts for theOxidation of Ammonia in Alkaline Media. U.S. Patent 7,485,211, 2009.(6) Cooper, M.; Botte, G. G. Hydrogen Production from the Electro-oxidation of Ammonia Catalyzed by Platinum and Rhodium on RaneyNickel Substrate. J. Electrochem. Soc. 2006, 153, A1894−A1901.(7) Despic, A. R.; Drazic, D. M.; Rakin, P. M. Kinetics ofElectrochemical Oxidation of Ammonia in Alkaline Solution. Electro-chim. Acta 1966, 11, 997−1005.(8) Vitse, F.; Cooper, M.; Botte, G. G. On the Use of AmmoniaElectrolysis for Hydrogen Production. J. Power Sources 2005, 142, 18−22.(9) Boggs, B. K.; Botte, G. G. Optimization of Pt−Ir on Carbon FiberPaper for the Electro-oxidation of Ammonia in Alkaline Media.Electrochim. Acta 2010, 55, 5287−5293.(10) Endo, K.; Nakamura, K.; Katayama, Y.; Miura, T. Pt−Me (Me =Ir, Ru, Ni) Binary Alloys as an Ammonia Oxidation Anode.Electrochim. Acta 2004, 49, 2503−2509.(11) Vidal-Iglesias, F. J.; Solla-Gullon, J.; Montiel, V.; Feliu, J. M.;Aldaz, A. Screening of Electrocatalysts for Direct Ammonia Fuel Cell:Ammonia Oxidation on PtMe (Me: Ir, Rh, Pd, Ru) and PreferentiallyOriented Pt(100) Nanoparticles. J. Power Sources 2007, 171, 448−456.(12) De Vooys, A. C. A.; Koper, M. T. M.; Van Santen, R. A.; VanVeen, J. A. R. The role of Adsorbates in the Electrochemical Oxidationof Ammonia on Noble and Transition Metal Electrodes. J. Electroanal.Chem. 2001, 506, 127−137.(13) Gerischer, H.; Mauerer, A. Untersuchungen zur AnodischenOxidation von Ammoniak an Platin-Elektroden. Electrochim. Acta1970, 25, 421.(14) Endo, K.; Katayama, Y.; Miura, T. A Rotating Disk ElectrodeStudy on the Ammonia Oxidation. Electrochim. Acta 2005, 50, 2181−2185.(15) Diaz, L. A.; Botte, G. G. Electrochemical Deammonification ofSynthetic Swine Wastewater. Ind. Eng. Chem. Res. 2012, 51, 12167−12172.(16) Kim, K. W.; Kim, Y. J.; Kim, I. T.; Park, G. I.; Lee, E. H.Electrochemical Conversion Characteristics of Ammonia to Nitrogen.Water Res. 2006, 40, 1431−1441.(17) Palaniappan, R.; Botte, G. G. Efficacy of Potassium Poly-(acrylate) Gel Electrolyte as a Substitute to Aqueous Electrolytes forAlkaline Ammonia Electrolysis. Electrochim. Acta 2013, 88, 772−781.(18) Horiuti, I.; Polanyi, M. Exchange Reactions of Hydrogen onMetallic Catalysts. Trans. Faraday Soc. 1934, 30, 1164.(19) Horiuti, J.; Okamoto, G. Note on the Mechanism of CatalyticInterchange of Hydrogen with Water and Alcohol. Trans. Faraday Soc.1936, 32, 1492.(20) Conway, B. E.; Bai, L. Determination of the AdsorptionBehaviour of Overpotential-Deposited Hydrogen-Atom Species in theCathodic Hydrogen-Evolution Reaction by Analysis of Potential-Relaxation Transients. J. Chem. Soc., Faraday Trans. 1: Phys. Chem.Condens. Phases 1985, 81, 1841.(21) Conway, B. E.; Bockris, J. O. M. Electrolytic HydrogenEvolution Kinetics and Its Relation to the Electronic and AdsorptiveProperties of the Metal. J. Chem. Phys. 1957, 26 (3), 532−541.(22) Conway, B. E.; Lijun, B.; Tessier, D. F. Data Collection andProcessing of Open-Circuit Potential-Decay Measurements Using aDigital Oscilloscope. Derivation of the H-Capacitance Behaviour ofH2-Evolving, Ni-Based Cathodes. J. Electroanal. Chem. 1984, 161, 39−49.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXK
(23) Markovic, N. M.; Sarraf, S. T.; Gasteiger, H. A.; Ross, P. N.Hydrogen Electrochemistry on Platinum Low-Index Single-CrystalSurfaces in Alkaline Solution. J. Chem. Soc., Faraday Trans. 1996, 92,3719−3725.(24) Greeley, J.; Jaramillo, T. F.; Bonde, J.; Chorkendorff, I.;Norskov, J. K. Computational High-Throughput Screening ofElectrocatalytic Materials for Hydrogen Evolution. Nat. Mater. 2006,5, 909−913.(25) Subbaraman, R.; Tripkovic, D.; Chang, K.-C.; Strmcnik, D.;Paulikas, A. P.; Hirunsit, P.; Chan, M.; Greeley, J.; Stamenkovic, V.;Markovic, N. M. Trends in Activity for the Water ElectrolyserReactions on 3d M(Ni,Co,Fe,Mn) Hydr(oxy)oxide Catalysts. Nat.Mater. 2012, 11, 550−557.(26) Bagotsky, V. S., Fundamentals of Electrochemistry, 2nd ed.; JohnWiley & Sons: Hoboken, NJ, 2006; pp 263−272.(27) Tilak, B.; Ramamurthy, A.; Conway, B. High PerformanceElectrode Materials for the Hydrogen Evolution Reaction fromAlkaline Media. Proc. - Indian Acad. Sci, Chem. Sci. 1986, 97, 359.(28) Ahn, S. H.; Hwang, S. J.; Yoo, S. J.; Choi, I.; Kim, H. J.; Jang, J.H.; Nam, S. W.; Lim, T. H.; Lim, T.; Kim, S. K.; et al. ElectrodepositedNi Dendrites with High Activity and Durability for HydrogenEvolution Reaction in Alkaline Water electrolysis. J. Mater. Chem.2012, 22, 15153−15159.(29) Choquette, Y.; Brossard, L.; Lasia, A.; Menard, H. Study of theKinetics of Hydrogen Evolution Reaction on Raney-Nickel Compo-site-Coated Electrode by AC Impedance Technique. J. Electrochem.Soc. 1990, 137, 1723−1730.(30) Daftsis, E.; Pagalos, N.; Jannakoudakis, A.; Jannakoudakis, P.;Theodoridou, E.; Rashkov, R.; Loukaytsheva, M.; Atanassov, N.Preparation of a Carbon Fiber-Nickel-Type Material and Investigationof the Electrocatalytic Activity for the Hydrogen Evolution Reaction. J.Electrochem. Soc. 2003, 150, C787−C793.(31) El Deab, M. S.; El Shakre, M. E.; El Anadouli, B. E.; Ateya, B. G.Hydrogen Evolution on Stacked Copper Screen Electrodes fromFlowing Alkaline Solutions. Int. J. Hydrogen Energy 1996, 21, 273−280.(32) Fan, C. L.; Piron, D. L. Electrodeposition as a Means ofProducing Large-Surface Electrodes Required in Water Electrolysis.Surf. Coat. Technol. 1995, 73, 91−97.(33) Fan, C. L.; Piron, D. L.; Meilleur, M.; Marin, L. P. HydrogenEvolution in Alkaline Solution on Electrolytic Nickel Cobalt andNickel Iron Deposited with Different Bath Compositions and CurrentDensities. Can. J. Chem. Eng. 1993, 71, 570−574.(34) Fan, C. L.; Piron, D. L.; Sleb, A.; Paradis, P. Study ofElectrodeposited Nickel−Molybdenum, Nickel−Tungsten, Cobalt−Molybdenum, and Cobalt−Tungsten as Hydrogen ElectrodesinAlkaline Water Electrolysis. J. Electrochem. Soc. 1994, 141, 382−387.(35) Krstajic, N.; Popovic, M.; Grgur, B.; Vojnovic, M.; Sepa, D. Onthe Kinetics of the Hydrogen Evolution Reaction on Nickel in AlkalineSolution. Part I. The Mechanism. J. Electroanal. Chem. 2001, 512, 16−26.(36) Martinez, S.; Metikos-Hukovic, M.; Valek, L. ElectrocatalyticProperties of Electrodeposited Ni−15Mo Cathodes for the HER inAcid Solutions: Synergistic Electronic Effect. J. Mol. Catal. A 2006,245, 114−121.(37) Pierozynski, B. On the Hydrogen Evolution Reaction at Nickel-Coated Carbon Fibre in 30 wt % KOH Solution. Int. J. Electrochem. Sci.2011, 6, 63−77.(38) Raj, I. A.; Vasu, K. I. Transition Metal-Based Cathodes forHydrogen Evolution in Alkaline Solution - Electrocatalysis on Nickel-Based Ternary Electrolytic Codeposits. J. Appl. Electrochem. 1992, 22,471−477.(39) Ramesh, L.; Sheshadri, B. S.; Mayanna, S. M. ElectrolyticPreparation and Characterization of Ni−Fe−Mo Alloys: CathodeMaterials for Alkaline Water Electrolysis. Int. J. Energ. Res. 1999, 23,919−924.(40) Rausch, S.; Wendt, H. Morphology and Utilization of SmoothHydrogen-Evolving Raney Nickel Cathode Coatings and PorousSintered-Nickel Cathodes. J. Electrochem. Soc. 1996, 143, 2852−2862.
(41) Shervedani, R. K.; Madram, A. R. Kinetics of HydrogenEvolution Reaction on Nanocrystalline Electrodeposited Ni62Fe35C3
Cathode in Alkaline Solution by Electrochemical Impedance Spec-troscopy. Electrochim. Acta 2007, 53, 426−433.(42) Divisek, J.; Schmitz, H.; Steffen, B. Electrocatalyst Materials forHydrogen Evolution. Electrochim. Acta 1994, 39, 1723−1731.(43) Giz, M. J.; Silva, J. C. P.; Ferreira, M.; Machado, S. A. S.;Ticianelli, E. A.; Avaca, L. A.; Gonzalez, E. R. Progress on theDevelopment of Activated Cathodes for Water Electrolysis. Int. J.Hydrogen Energy 1992, 17, 725−729.(44) Jaksic, M. M.; Johansen, B.; Tunold, R. ElectrochemicalBehaviour of Platinum in Alkaline and Acidic Solutions of Heavy andRegular Water. Int. J. Hydrogen Energy 1993, 18, 817−837.(45) Jaksic, M. M.; Johansen, B.; Tunold, R. ElectrochemicalBehaviour of Rhodium in Alkaline and Acidic Solutions of Heavy andRegular Water. Int. J. Hydrogen Energy 1994, 19, 35−51.(46) Jaksic, M. M.; Johansen, B.; Tunold, R. ElectrochemicalBehaviour of Palladium in Alkaline and Acidic Solutions of Heavy andRegular Water. Int. J. Hydrogen Energy 1993, 18, 111−124.(47) Jaksic, M. M.; Johansen, B.; Tunold, R. ElectrochemicalBehaviour of Iridium in Alkaline and Acidic Solutions of Heavy andRegular Water. Int. J. Hydrogen Energy 1994, 19, 321−335.(48) Norskov, J. K.; Bligaard, T.; Logadottir, A.; Kitchin, J. R.; Chen,J. G.; Pandelov, S. Trends in the Exchange Current for HydrogenEvolution. J. Electrochem. Soc. 2005, 152, J23−J26.(49) Roue, L.; Blouin, M.; Guay, D.; Schulz, R. Comparative Study ofthe Electrochemical Behavior of Polycrystalline and NanocrystallineRu Powder in NaOH Solution. J. Electrochem. Soc. 1998, 145, 1624−1631.(50) Schmickler, W.; Trasatti, S. Comment on Trends in theExchange Current for Hydrogen Evolution. J. Electrochem. Soc. 2006,153, L31−L32.(51) Rajeswari, J.; Kishore, P.; Viswanathan, B.; Varadarajan, T. FacileHydrogen Evolution Reaction on WO3Nanorods. Nanoscale Res. Lett.2007, 2, 496−503.(52) Vilekar, S. A.; Fishtik, I.; Datta, R. Kinetics of the HydrogenElectrode Reaction. J. Electrochem. Soc. 2010, 157, B1040−B1050.(53) Subbaraman, R.; Tripkovic, D.; Strmcnik, D.; Chang, K.-C.;Uchimura, M.; Paulikas, A. P.; Stamenkovic, V.; Markovic, N. M.Enhancing Hydrogen Evolution Activity in Water Splitting byTailoring Li+−Ni(OH)2−Pt Interfaces. Science 2011, 334, 1256−1260.(54) Bai, L.; Harrington, D. A.; Conway, B. E. Behavior ofOverpotentialDeposited Species in Faradaic ReactionsII. ACImpedance Measurements on H2 Evolution Kinetics at Activated andUnactivated Pt Cathodes. Electrochim. Acta 1987, 32, 1713−1731.(55) Lasia, A., Applications of Electrochemical Impedance Spectros-copy to Hydrogen Adsorption, Evolution and Absorption into Metals.In Modern Aspects of Electrochemistry; Conway, B. E.; White, R. E.,Eds.; Springer: New York, 2002; Vol. 35, pp 1−50.(56) Barber, J. H.; Conway, B. E. Structural Specificity of the Kineticsof the Hydrogen Evolution Reaction on the Low-Index Surfaces of PtSingle-Crystal Electrodes in 0.5 M dm−3 NaOH. J. Electroanal. Chem.1999, 461, 80−89.(57) Barber, J.; Morin, S.; Conway, B. E. Specificity of the Kinetics ofH2, Evolution to the Structure of Single-Crystal Pt Surfaces, and theRelation Between OPD and UPD H. J. Electroanal. Chem. 1998, 446,125−138.(58) Lasia, A.; Rami, A. Kinetics of Hydrogen Evolution on NickelElectrodes. J. Electroanal. Chem. 1990, 294, 123−141.(59) Chen, L. L.; Lasia, A. Study of the Kinetics of HydrogenEvolution Reaction on Nickel-Zinc Alloy Electrodes. J. Electrochem.Soc. 1991, 138, 3321−3328.(60) Los, P.; Lasia, A. Electrocatalytic Properties of AmorphousNickel Boride Electrodes for Hydrogen Evolution Reaction in AlkalineSolution. J. Electroanal. Chem. 1992, 333, 115−125.(61) Armstrong, R. D.; Henderson, M. Impedance Plane Display of aReaction With an Adsorbed Intermediate. J. Electroanal. Chem. 1972,39, 81−90.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXL
(62) Harrington, D. A.; Conway, B. E. ac Impedance of FaradaicReactions Involving Electrosorbed IntermediatesI. Kinetic Theory.Electrochim. Acta 1987, 32, 1703−1712.(63) Pozio, A.; Francesco, M. D.; Cemmi, A.; Cardellini, F.; Giorgi, L.Comparison of High Surface Pt/C Catalysts by Cyclic Voltammetry. J.Power Sources 2002, 105, 13−19.(64) Baraka, A. M.; Shaarawy, H. H.; Hamed, H. A. Electrodepositionof Rhodium Metal on Titanium Substrates. Anti-Corros. Methos. Mater.2002, 49, 277−282.(65) Søgaard, M.; Odgaard, M.; Skou, E. M. An Improved Methodfor the Determination of the Electrochemical Active Area of PorousComposite Platinum Electrodes. Solid State Ionics 2001, 145, 31−35.(66) Tarasevich, E. I. K.; R, M. Electrochemical Determination ofSurface Area of Metals. Russ. Chem. Rev. 1978, 47, 416.(67) Grden, M.; Alsabet, M.; Jerkiewicz, G. Surface Science andElectrochemical Analysis of Nickel Foams. ACS Appl. Mater. Inter.2012, 4, 3012−3021.(68) Gasteiger, H. A.; Markovic, N.; Ross, P. N.; Cairns, E. J. COElectrooxidation on Well-Characterized Pt−Ru Alloys. J. Phys. Chem.1994, 98 (2), 617−625.(69) Spendelow, J. S.; Lu, G. Q.; Kenis, P. J. A.; Wieckowski, A.Electrooxidation of Adsorbed CO on Pt(111) and Pt(111)/Ru inAlkaline Media and Comparison with Results from Acidic Media. J.Electroanal. Chem. 2004, 568, 215−224.(70) Giorgi, L.; Pozio, A.; Bracchini, C.; Giorgi, R.; Turtu, S. H2 andH2/CO Oxidation Mechanism on Pt/C, Ru/C and Pt−Ru/CElectrocatalysts. J. Appl. Electrochem. 2001, 31, 325−334.(71) Fournier, J.; Wrona, P. K.; Lasia, A.; Lacasse, R.; Lalancette, J.M.; Menard, H. Catalytic Influence of Commercial Ru, Rh, Pt, and Pd(Similar-to-0.1 Atomic Percent) Intercalated in Graphite on theHydrogen Evolution Reaction. J. Electrochem. Soc. 1992, 139, 2372−2377.(72) Silverman, D. C. Technical Note: Revised EMF-pH Diagram forNickel. Corrosion 1981, 37, 546−548.(73) Conway, B. E.; Lijun, B. Determination of Adsorption of OPDH Species in the Cathodic Hydrogen Evolution Reaction at Pt inRelation to Electrocatalysis. J. Electroanal. Chem. 1986, 198, 149−175.(74) Conway, B. E.; Tilak, B. V. Interfacial Processes InvolvingElectrocatalytic Evolution and Oxidation of H2, and the Role ofChemisorbed H. Electrochim. Acta 2002, 47, 3571−3594.(75) Markovic, N. M.; Gasteiger, H. A.; Ross, P. N. OxygenReduction on Platinum Low-Index Single-Crystal Surfaces in AlkalineSolution: Rotating Ring DiskPt(hkl) Studies. The J. Phys. Chem. 1996,100, 6715−6721.(76) Visscher, W.; Barendrcht, E. The Anodic Oxidation of Nickel inAlkaline Solution. Electrochim. Acta 1980, 25, 651−655.(77) Jaksic, M. M.; Brun, J.; Johansen, B.; Tunold, R. Some SpecificPotentiodynamic Features of Nickel Electrodes in Alkaline AqueousMedia. Int. J. Hydrogen Energy 1995, 20, 265−273.(78) Wasmus, S.; Vasini, E. J.; Krausa, M.; Mishima, H. T.; Vielstich,W. DEMS-Cyclic Voltammetry Investigation of the Electrochemistryof Nitrogen Compounds in 0.5 M Potassium Hydroxide. Electrochim.Acta 1994, 39, 23−31.(79) Kapalka, A.; Cally, A.; Neodo, S.; Comninellis, C.; Wachter, M.;Udert, K. M. Electrochemical Behavior of Ammonia at Ni/Ni(OH)2Electrode. Electrochem. Commun. 2010, 12, 18−21.(80) Brug, G. J.; van den Eeden, A. L. G.; Sluyters-Rehbach, M.;Sluyters, J. H. The analysis of Electrode Impedances Complicated bythe Presence of a Constant Phase Element. J. Electroanal. Chem. 1984,176, 275−295.(81) Harrington, D. A.; van den Driessche, P. Mechanism andEquivalent Circuits in Electrochemical Impedance Spectroscopy.Electrochim. Acta 2011, 56, 8005−8013.(82) Wrona, P. K.; Lasia, A.; Lessard, M.; Menard, H. Kinetics of theHydrogen Evolution Reaction on a Rhodium Electrode. Electrochim.Acta 1992, 37, 1283−1294.
The Journal of Physical Chemistry C Article
dx.doi.org/10.1021/jp405191c | J. Phys. Chem. C XXXX, XXX, XXX−XXXM
