Oxidation of ammonium on platinum in acidic solutions

Article abstract of DOI:10.1149/1.2405851

Oxidation of ammonium on polycrystalline Pt has been studied in perchloric and sulfuric acid solutions at room temperature. Ammonium was electrochemically active, and N2 and NO were detected using differential electrochemical mass spectroscopy. Cyclic voltammetry and electrochemical quartz crystal microbalance data showed that ammonium affects the formation and reduction of Pt oxides, probably due to formation of very stable adsorbed nitrogen or nitrogen-oxygen species stable down to the Hads region. An oxidation peak occurred at 0.8 VRHE in positively going scans, decreasing strongly with increasing sweep rate. A corresponding reduction shoulder at about 0.66 VRHE was seen, probably caused by formation and reduction of N Hx,ads or Nads species. Formation of Pt oxides was shifted to slightly higher potentials in the presence of ammonium, and the reduction charge of the Pt oxide reduction peak at 0.80 VRHE was independent of sweep rate, indicating that the amount of Pt oxides formed was limited by other adsorbates. In particular, the most strongly adsorbed hydrogen was shifted to lower potentials by adsorbed species formed at high potentials as well as adsorption of bisulfate mutually stabilized by ammonia. The voltammetric response in the hydrogen desorption region was not affected, showing that all adsorbed species were desorbed at low potentials in the negatively going scan.

Full text of DOI:10.1149/1.2405851

Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
B263
0
013-4651/2006/154͑2͒/B263/8/$20.00 © The Electrochemical Society
Oxidation of Ammonium on Platinum in Acidic Solutions
a,c, ,z
b,
b,
Rune Halseid, * Jesse S. Wainright, * Robert F. Savinell, ** and
a,
Reidar Tunold *
a
Department of Materials Science and Engineering, Group of Electrochemistry, Norwegian University of
Science and Technology, NO-7491 Trondheim, Norway
b
Department of Chemical Engineering, Case Western Reserve University, Cleveland Ohio 44106, USA
Oxidation of ammonium on polycrystalline Pt has been studied in perchloric and sulfuric acid solutions at room temperature.
Ammonium was electrochemically active, and N2 and NO were detected using differential electrochemical mass spectroscopy.
Cyclic voltammetry and electrochemical quartz crystal microbalance data showed that ammonium affects the formation and
reduction of Pt oxides, probably due to formation of very stable adsorbed nitrogen or nitrogen-oxygen species stable down to the
Hads region. An oxidation peak occurred at 0.8 VRHE in positively going scans, decreasing strongly with increasing sweep rate. A
corresponding reduction shoulder at about 0.66 VRHE was seen, probably caused by formation and reduction of NHx,ads or Nads
species. Formation of Pt oxides was shifted to slightly higher potentials in the presence of ammonium, and the reduction charge
of the Pt oxide reduction peak at 0.80 VRHE was independent of sweep rate, indicating that the amount of Pt oxides formed was
limited by other adsorbates. In particular, the most strongly adsorbed hydrogen was shifted to lower potentials by adsorbed species
formed at high potentials as well as adsorption of bisulfate mutually stabilized by ammonia. The voltammetric response in the
hydrogen desorption region was not affected, showing that all adsorbed species were desorbed at low potentials in the negatively
going scan.
©
2006 The Electrochemical Society. ͓DOI: 10.1149/1.2405851͔ All rights reserved.
Manuscript submitted September 26, 2005; revised manuscript received October 6, 2006.
Available electronically December 27, 2006.
−
͑1−␦͒−
+ ␦e−
Background for the study.— Our interest in oxidation of ammo-
nium in acidic solutions stems from the observation that ammonia
affects the performance of polymer electrolyte membrane fuel cells
OH ꢀ OH
͓1͔
͓2͔
͓3͔
͓4͔
͓5͔
͓6͔
͓7͔
͓8͔
ads
NH3,aq ꢀ NH3,ads
1
-3
͑
PEMFCs͒. Both partial and full recovery has been reported when
NH3,ads + OH͑
1−␦͒−
→ NH2,ads + H O + ͑1 − ␦͒e
−
operating the PEMFC on neat hydrogen after exposure to ammonia,
and it is of interest to determine if the recovery mechanism may be
related to electrochemical oxidation of ammonium. Further, ad-
sorbed species on the Pt surfaces in the fuel cell ͑FC͒ catalyst layer
may impede the rate and selectivity of the desired FC reactions, the
hydrogen oxidation reaction ͑HOR͒, and the oxygen reduction reac-
tion ͑ORR͒. However, this study will not address ammonia poison-
ing of fuel cells directly, but the data reported in this paper are
useful to better understand the poisoning and recovery mechanism
ads
2
͑1−␦͒−
−
NH2,ads + OHads → NHads + H O + ͑1 − ␦͒e
2
NHx,ads + NHy,ads → N H
2
n,ads
͑1−␦͒−
−
N H
2
+ nOHads → N2,ads + nH O + n͑1 − ␦͒e
n,ads
2
N2,ads → N2,g
3
of ammonium in PEMFC as we have reported elsewhere.
NHads + OH͑
1−␦͒−
→ Nads + H O + ͑1 − ␦͒e
where x and y is 1 or 2 and n = x + y.
−
ads
2
Objectives of this study.— The aim of the present study is to bet-
ter understand oxidation of ammonium in acidic solutions which has
not been studied in any detail before, but has been reported to be
Although the reaction mechanism is complex, one should note
4
extremely slow. However, this was stated in the context of bulk
the key role played by adsorbed NH or NH in forming the N–N
2
oxidation of ammonia in direct ammonia fuel cells ͑DAFC͒, where
very high oxidation currents are desirable. In this study volatile
oxidation products, and at which potentials these form, were as-
sessed using differential electrochemical mass spectrometry
bond before further oxidation takes place; see Eq. 5. Gerischer and
Mauerer found that the electrode deactivated quickly initially, prob-
5
ably due to partial blocking by OHads. This initial deactivation was
easily reversed. If the electrode was polarized for longer periods, a
more persistent deactivation was found, ascribed to formation of
Nads considered to be a self-poisoning reaction in the Gerischer-
͑
͑
DEMS͒. Classical electrochemical methods like cyclic voltammetry
CV͒ and rotating disk electrode ͑RDE͒ were also used. Formation
6
of adsorbed products was detected using electrochemical quartz
crystal microbalance ͑EQCM͒. An overview of the available litera-
ture on ammonia oxidation in alkaline solutions is given next.
Mauerer mechanism. de Vooys et al. found that there was a very
distinct onset of this latter deactivation process above 0.57 VRHE.
7
Gootzen et al. found, using DEMS, that the rate of nitrogen evolu-
tion was stable for at least 10 min at an electrode potential of 0.65
VRHE, but the total current fell quickly. Two species were thus
formed during oxidation of ammonia: nitrogen and an adsorbed spe-
cies which gradually saturated the electrode surface, which can be
explained by self-poisoning of the surface by Nads; see Eq. 8. Was-
Oxidation of ammonia in alkaline solutions.— Different reac-
tion mechanisms for oxidation of ammonia in alkaline solutions
4,5
4
have been suggested. Oswin and Salomon proposed a reaction
mechanism where ammonia is electrochemically dehydrogenated in
steps to form N , which then combine to form N . Gerischer and
8
ads
2
5
mus et al. found that the adsorption process of species originating
Mauerer have suggested a different reaction mechanism that is bet-
ter supported by available experimental data. The mechanism is
rather complex partly because the nature of the intermediate species
is not known
from NH at 0.40 V
was complete in about 5 min. The adsorp-
3
RHE
tion was associated with an oxidative current.
7
8
Both Gootzen et al. and Wasmus et al. report experimental
evidence for the Gerischer-Mauerer mechanism based on adsorbate
8
experiments. Wasmus et al. reported that when a positively going
sweep was applied after adsorption at 0.40 VRHE, nitrogen evolution
was detected by DEMS, and was associated with an oxidative cur-
rent. This showed that the source of nitrogen evolution was a species
with oxidation number less than 0, e.g., NH2,ads or NHads. If the
adsorbed intermediate species leading to nitrogen evolution had
*
Electrochemical Society Active Member.
*
* Electrochemical Society Fellow.
c
Present address: Department of Surface Chemistry and Catalysis, University of
Ulm, Albert-Einstein-Allee 47, D-89081 Ulm, Germany.
E-mail: rune.halseid@uni-ulm.de
z
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

B264
Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
been Nads, nitrogen evolution would occur with no net faradaic cur-
rent. Gootzen et al. found that the current changed from reductive
to oxidative with time when negatively going sweeps were halted at
EQCM.— An experimental setup from Elchema ͑EQCN-701 and
PS-305 potentiostat͒ was used in the EQCM measurements. Ohmic
compensation was used at sweep rates higher than 50 mV/s. The
cell resistance was measured using a Solartron 1280B. An EG&G
175 sweep generator was used, and the data were logged by an
oscilloscope ͑Tektronix TDS 420A͒ or an AD card ͑DaqBoard,
IOtech͒.
7
0.55 VRHE, which can be explained assuming that two reactions take
place at the same time. The first reaction is the reductive formation
of NHx,ads from Nads ͑see Eq. 8͒; the second reaction is the oxidative
formation of nitrogen from NHx,ads via N Hn,ads; see Eq. 5-7.
2
Strongly adsorbed nitrogen species can be reduced by operating
the electrode at potentials in or close to the hydrogen adsorption
region. In negatively going sweeps, Wasmus et al. reported sig-
The crystals used were Pt sputtered, AT-cut, unpolished 10 MHz
crystals ͑International Crystal Manufacturing͒ with no adhesion
layer between the quartz and the Pt layer. The mass sensitive area
was 5–8% smaller than the electrochemically active area. After the
WE had been platinized, see below, 20–30 cyclic sweeps were al-
lowed prior to any measurements to stabilize the Pt surface. The WE
was regularly cleaned by sweeping the potential between 0.025 and
7
,8
8
11
nificant formation of NH , detected as the fragment m/e = 15 at
3
potentials below 0.50 VRHE. The exact nature of the adsorbate,
which was the source of this NH formation, was not reported. They
3
also observed an m/e = 30 ͑NO͒ signal at potentials below 0.15
VRHE when the upper reversal limit in the CV was at least 1.20
VRHE, and the NO signal was much stronger when scanning to 1.60
VRHE. de Vooys et al. concluded that adsorbates formed from NO
dissolved in acidic electrolyte were reduced to ammonia at low
1
.25 VRHE at 50 mV/s. The standard sweep rate used was 50 mV/s,
with potential sweep limits 0.025 and 1.25 VRHE. The roughness
factor was typically 15–25 as determined from the hydrogen under-
2
potential deposited ͑UPD͒ charge assuming that 210 ␮C/cm corre-
9
sponds to a full monolayer coverage.
potentials.
A complication of the EQCM method is that the mass response
may change dramatically if the crystal is cycled for a long time,
Formation of NO and N O by oxidation of ammonium has been
2
8
reported by Wasmus et al., and commenced at approximately 0.80
1
2-14
even though the recorded CVs show very small changes.
The
VRHE with a plateau of maximum rate at potentials above 1.00 VRHE
.
change is believed to be caused by roughening of the electrode
However, the intensities of the NO mass signals were much smaller
x
1
4
surface on a mesoscopic scale, i.e., not the same roughness as
measured by electrochemical techniques like hydrogen UPD ob-
served with CV. To avoid this the crystals were regularly platinized
by pulse deposition in a solution of platinic acid in hydrochloric acid
depositing a total of 5 ␮g Pt each time to restore the electrode. The
Pt black deposit was rinsed well with purified water followed by
exposure to a 50/50 mixture of concentrated nitric and sulfuric acid
than those seen for nitrogen at the same potential. Formation of
nitrous oxides only at high electrode potentials suggests that the
presence of adsorbed oxygen species on the electrode is a prerequi-
site for these reactions.
Experimental
All glassware used was thoroughly washed and then placed in a
hot solution of 5–10 wt % H O in 0.5 M sulfuric acid. The equip-
1
5
for about 30 s, and then thoroughly rinsed with purified water.
The frequency response of the quartz crystal is proportional to
the mass change of the active area of the crystal according to the
2
2
ment was then rinsed with purified water ͑Barnstead NANOpure II
system followed by a MilliPore Milli-Q UV Plus system͒. The RDE
cell was cleaned in concentrated KOH overnight before it was
boiled in high-purity water and rinsed. The reference electrode ͑RE͒
used in all experiments, except in the RDE experiments, was a re-
versible hydrogen electrode ͑RHE͒ consisting of a platinized Pt
gauze in a glass tube sealed on one end, filled with the same elec-
1
6
Sauerbrey equation, which is applicable to small changes in
1
4
mass, but other factors than mass changes may also influence the
1
4
frequency response. The theoretical sensitivity of the crystal used
in this work, which has a fundamental resonance frequency of
1
0 MHz, is −0.8673 ng/Hz. The mass sensitivity was calibrated by
bulk deposition of Cu from a 5 mM solution of cupric sulfate in
.5 M H SO at 0.06 VRHE. The Sauerbrey coefficient was found to
trolyte as the cell, and electrochemically charged with H . A satu-
2
0
rated calomel electrode was used as RE in the RDE experiments, but
the potentials are quoted against the RHE. The counter electrode
2
4
be −1.24 ± 0.10 ng/Hz. The absolute calibrated value is signifi-
cantly higher than the theoretical value. The reason for the deviation
is most likely that the theoretical Sauerbrey coefficient is calculated
͑
CE͒ was a Pt wire in all experiments.
TraceMetal grade acids and NH OH ͑Fisher͒ as well as
4
17
assuming infinite dimension of the crystal. Less than 100% current
NaOH·H O ͑Alfa Aesar, 99.996% pure on metal basis͒ were used.
2
efficiency for copper deposition may also contribute to the
deviation.
EQCM data for perchloric acid were measured using Trace-Select
chemicals ͑Fluka͒, and for the RDE and adsorbate experiments Su-
praPur chemicals ͑Merck͒ were used. Continuous nitrogen purge
was used to remove oxygen from the solutions ͓Ultrahigh-purity
RDE and flow cell experiments.— An RDE rotator from Pine
Instruments ͑AFASRE͒ combined with a Pine AFRDE5 bipoten-
tiostat was used, and the analog data were logged with an AD card
͑
UHP͒ grade from Praxair and AGA͔. Ar ͑MTI Gase, N 6.0͒ was
used for purging in the RDE and adsorbate experiments. UHP he-
lium from Praxair was used in the DEMS experiments because N2
was one of the species of interest.
͑
6036E from National Instruments͒ controlled by LabVIEW ͑Na-
tional Instruments͒. Measurements were performed in a standard
electrochemical cell with three separate compartments for the RE,
CE, and WE. The WE substrate was a glassy carbon disk ͑Sigradur
G from Hochtemperatur Werkstoffe GmbH͒ mounted in a PFTE
DEMS.— A comprehensive description of the DEMS system
10
used is given by Wasmus et al. The major difference in our work
2
was how the working electrode ͑WE͒ was prepared. A solution of
holder, and had a geometric area of 0.164 cm . The active part of the
2
.2 wt % BN-18 ͑polyvinyl butyral resin͒ in 97.8 wt % diacetone
alcohol was thoroughly mixed 2:1 by weight with Pt black ͑HiSPEC
000, Alfa Aesar͒ and then painted by hand onto a microporous
WE was prepared by pipetting a 20 ␮L aliquot aqueous suspension
͑2 mg catalyst per mL being constantly sonicated to prevent settling͒
of carbon-supported Pt ͑20 wt % Pt on C, E-TEK͒ onto the glassy
carbon substrate. The droplet was dried in a N2 stream and then
another 20 ␮L aliquot of dilute, aqueous Nafion solution was added
1
PTFE substrate to form a thin, even layer. The ink was allowed to
dry for at least 1 day. The electrode was cleaned by repeatedly step-
ping the potential between 0.03 and 1.60 VRHE. The cleanliness of
1
8
on top to mechanically stabilize the catalyst.
1
9
the WE was checked by monitoring the m/e = 28 ͑CO͒ and 44
The flow cell has been described in more detail elsewhere. The
WE, a 9 mm diameter polycrystalline Pt cylinder ͑MaTecK, 99.95%
purity͒, was mounted into the thin-layer flow cell pressing against an
8
͑
CO ͒ signals. In a separate experiment CO was bubbled through
2
2
the solution, and the adsorbed species formed by CO reduction at
2
2
0.20 VRHE on the WE were stripped off in a 5 mV/s positively going
approximately 50 ␮m thick spacer exposing 0.283 cm Pt surface to
scan after the electrolyte had been bubbled with He. The mass signal
m/e = 44 closely followed the oxidation current, verifying that the
applied sweep rate was low enough.
the thin-layer electrolyte volume. The same potentiostat and data
aquisition as for the RDE was used. The electrolyte entered the
center of the WE compartment and flowed in a radial pattern over
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
B265
Figure 1. ͑Color online͒ ͑a͒ CVs at 5 mV/s in neat 0.1 M H SO ͑– –͒ as
Figure 2. ͑Color online͒ ͑a͒ CVs at 5 mV/s in neat 0.1 M HClO ͑– –͒ as
2
4
4
well as in 0.1 M H SO + 0.1 M NH OH ͑—͒ and DEMS data for ͑b͒
well as in 0.1 M HClO + 0.05 M NH OH ͑—͒ and DEMS data ͑b͒ m/e
2
4
4
4
4
m/e = 28, ͑c͒ 30, and ͑d͒ 44, measured using the electrolyte containing am-
monium. Thick lines in ͑b͒–͑d͒ are used for positively going sweep, thin
lines for negatively going sweep.
= 28, ͑c͒ 30, and ͑d͒ 14, measured using the electrolyte containing ammo-
nium. Thick lines in ͑b͒–͑d͒ are used for positively going sweep, thin lines
for negatively going sweep.
8
VRHE. The highest NO and N O signals were detected at potentials
2
the WE towards the edge at a flow rate of approximately 5 ␮L/s.
The design allows easy electrolyte exchange and flushing while
maintaining potential control, enabling adsorbate experiments.
higher than 1.20 VRHE in alkaline solutions compared to about 1.4
8
VRHE in the acidic solutions studied here. No m/e = 44 ͑N O͒ sig-
2
nals were detected in this study, probably due to the high back-
ground level.
Results
Coulometry.— The total charge passed as a function of potential
was determined. The net oxidation charge passed during one com-
plete CV cycle was calculated both for the neat 0.5 M sulfuric acid,
Volatile species formed by oxidation of ammonium.— Potentio-
dynamic measurements with simultaneous measurements of mass
spectra in both sulfuric and perchloric acid solutions were per-
formed. The results are shown in Fig. 1 and 2. Substantial back-
ground levels in the DEMS system were present for some of the
species of interest, in particular m/e = 14, 28, and 44. The signal-
to-noise ratio was enhanced using the MS software to average 8 to
+
acid
, and for the acid containing 0.5 M ammonium, Q oN xH
4
Q
, for dif-
ox
ferent sweep rates between the potential limits 0.05 and 1.25 VRHE
.
The net oxidation of ammonium in one complete cycle ⌬Q is
ox
+
4
NH
acid
ox
found by ⌬Qox = Qox − Q , and is shown in Fig. 3a. The influ-
1
6 individual data points. The DEMS data were corrected for back-
ence of the upper reversal potential on the net oxidation charge was
also investigated at 50 mV/s as shown in Fig. 3b. The net charge
passed in electrolytes containing ammonium was positive, i.e., a net
oxidation process was taking place, for upper reversal potentials
above 1.0 VRHE. The net oxidation increased with decreasing sweep
rate, i.e., longer times at high potentials. This is consistent with the
observation of NO evolution in the DEMS experiments. No signifi-
ground drift, and data from 5–8 different sweeps were averaged to
further minimize the noise.
As seen from Fig. 1 and 2, there was a stronger m/e = 28 signal
͑
N ͒ in the positively going sweep in the perchloric than in the
2
sulfuric acid solution. The maximum of the m/e = 28 signal ap-
peared at about 0.79 VRHE in both solutions. An N signal of similar
2
magnitude was measured in both solutions at about 0.68 VRHE in
negatively going sweeps. The m/e = 14 signal ͑N͒ was also moni-
tored in the perchloric solution ͑see Fig. 2d͒, and followed the
m/e = 28 signal closely. The m/e = 28 signal is thus due to nitrogen
evolution.
We also detected an m/e = 30 signal in both solutions, but some-
what stronger in the sulfuric than in the perchloric solution. This
signal is attributed to formation of NO at the WE at potentials higher
than 1.25 VRHE. In alkaline solutions it has been found that nitric
cant net oxidation current due to N evolution at intermediate po-
2
tentials was seen. The variation in upper reversal limit was per-
formed at a higher sweep rate than in the DEMS experiments ͑50 vs
5
mV/s͒. From Fig. 3a we would expect a higher net oxidation
charge, consistent with the DEMS data also in the range of N2
evolution, if a lower sweep rate had been used.
Effect of ammonium concentration.— DEMS measurements
were also performed with lower concentrations ͑data not given here,
oxides ͑NO and N O͒ were formed at potentials higher than 0.80
see Ref. 11͒. The N signal ͑m/e = 28͒ could also be detected in a
2
2
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

B266
Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
Figure 3. ͑Color online͒ ͑a͒ Difference in net oxidation charge, ⌬Q , be-
ox
tween acid with addition of NH OH and neat sulfuric acid during one com-
4
plete CV cycle at different sweep rates ␯ reversed at 1.25 VRHE. ͑b͒ ⌬Qox as
in ͑a͒, but here with varying upper reversal limits with a fixed sweep rate of
Figure 5. ͑Color online͒ CV recorded at 50 mV/s on platinized Pt in elec-
trolytes with different ammonium and acid concentration to assess the com-
bined effect of pH and ammonium concentration.
50 mV/s.
0
.1 M HClO solution containing 0.025 M NH OH. The NO signal
4
4
͑
m/e = 30͒ was detectable in the solution containing 0.01 M
was affected so that the onset of oxide formation was less sharp
when ammonium was present. There was also a slightly higher oxi-
dation current at potentials above 1.0 VRHE with ammonium present
in the solution. The Pt oxide reduction peak at 0.80 VRHE in the
negatively going scan decreased with increasing concentration of
ammonium, but there was a broad reduction wave following from
NH OH but not with 0.05 M NH OH. The observed CVs and cor-
4
4
responding EQCM responses for different concentrations of ammo-
nium in 0.5 M H SO are shown in Fig. 4a and b, respectively. The
mass responses of the electrodes in neat acidic solution seen in our
study were similar to data reported by other authors.
some researchers have reported larger mass changes,
the hydrogen adsorption region, whereas the mass gain in the oxide
region was similar in all literature sources; see Ref. 11 for further
details.
2
4
1
2,13
However,
mainly in
2
0,21
0
.70 down to 0.50 VRHE. The reduction peak for adsorption of
strongly bound hydrogen at 0.25 VRHE was affected, and a shoulder
appeared at about 0.18 VRHE. There was also a slightly higher re-
ductive current at potentials below 0.10 VRHE. The hydrogen desorp-
tion peaks were not affected by the presence of ammonium.
A shoulder appeared at approximately 0.80 VRHE in the CV in the
positively going scan; see Fig. 4a. The Pt oxide formation region
There was an increasing hysteresis in the EQCM response with
increasing ammonium concentration which was particularly notice-
able at potentials below 0.90 VRHE; see Fig. 4b. A crossover was
seen at approximately 0.92 VRHE in the positively going sweep
where the adsorbed mass in electrolytes containing ammonium be-
came larger than in the neat acidic solution. The adsorbed mass at
high potentials was higher the higher the ammonium concentration
was. Also, the adsorbed mass on the electrode was higher in the
negatively going sweep, showing that a stable adsorbate was formed
at high potentials. This is consistent with the findings in the CV in
the same potential range where a higher reductive current was ob-
served. Higher reductive currents have also been seen in the pres-
6
ence of ammonia in alkaline solutions.
Effect of pH on oxidation of ammonium.— Obviously the pH,
the ionic strength, as well as the concentration of anions will change
upon addition of ammonium hydroxide to the neat acidic solutions
which also can affect the behavior of Pt. It is for example likely that
ammonium adsorbs as ammonia on Pt through the lone electron pair
so that a proton must be released upon adsorption. This means that
the pH of the solution will affect the surface coverage of NH3,ads and
thus the ammonium oxidation rate; see Eq. 2 in the Gerischer-
Maurer mechanism. Two sets of experiments were made to evaluate
this. First, NaOH was added to neat sulfuric acid instead of NH OH.
4
The CV and corresponding EQCM response, which are not shown
here, were very similar to those seen in neat acidic solution. This
indicates that the effect of pH, ionic strength, and increased concen-
tration of ͑bi͒sulfate alone on the observable CVs and EQCM data is
small.
In the second measurement 0.5 M NH OH was added to the neat
.5 M sulfuric acid electrolyte. Then, an additional amount of sul-
4
Figure 4. ͑Color online͒ ͑a͒ CV and ͑b͒ corresponding EQCM response
0
measured on platinized Pt at 50 mV/s in electrolyte containing 0.5 M H SO
2
4
furic acid equivalent to 0.5 M was added; see Fig. 5. The effect of
ammonium was partially reversed by addition of more acid to the
0.5 M H SO + 0.5 M NH OH solution. However, there was still a
base electrolyte. The mass response for neat acidic solution is also shown
with an offset of 20 ng/cm2 for clarity. The legend applies to both plots.
Arrows indicate the effect of increasing ammonium concentration. Arrows in
the upper part of ͑b͒ indicate the path followed by the mass response through
one CV cycle.
2
4
4
significant influence of ammonium compared to the neat acidic so-
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
B267
Figure 6. ͑Color online͒ ͑a͒ CVs of Pt in 0.5 M HClO + 0.25 M NH OH
4
4
Figure 7. ͑Color online͒ ͑a͒ CVs of platinized Pt in 0.5 M H SO
2 4
at 50 mV/s with different upper reversal potentials. Refer to the text for
explanation of arrows and numbers. ͑b͒ Detail of the CVs in the hydrogen
adsorption region; the arrows indicate the effect of increasing reversal po-
tential. Upper reversal potentials from 0.6–1.1 VRHE are shown. Similar be-
havior was also seen in sulfuric acid solution ͑Ref. 11͒, but is not included
here.
+
0.5 M NH OH at 50 mV/s. ͑b͒ Details of the CVs in the hydrogen adsorp-
4
tion region; the arrows indicate the effect of increasing reversal potential.
Upper reversal potentials from 0.6–1.5 VRHE are shown.
EQCM data were also recorded with different upper reversal
potentials in sulfuric solutions as shown in Fig. 8. The mass changes
seen in the low potential region were not directly due to adsorption
of hydrogen because the mass change due to this process is only
lution, suggesting that there is a combined effect of ammonium con-
centration and pH. The change in pH and ammonium concentration
was accompanied by changes in ionic strength and concentration of
2
14
about 2.3 ng/cm , and also we find a decrease in mass detected by
EQCM upon hydrogen adsorption. Most likely the mass changes in
this region are due to desorption of water or specificaly adsorbed
͑
bi͒sulfate as well, so that the three factors cannot be separated by
1
4
these experiments.
anions like ͑bi͒sulfate. Further, there was a more noticeable hys-
teresis between the positively and negatively going sweeps in the
double-layer region when ammonium was present commencing at
Effect of the upper reversal potential.— The upper reversal po-
tential was also an important parameter; see Fig. 6. The CV in the
positively going scan was not affected by the reversal potential, i.e.,
the electrode surface was fully restored in the negatively going scan,
also independently on the applied scan rate as discussed later; see
Fig. 9. There was no effect of reversal potential on the hydrogen
0
.30 VRHE; see also Fig. 4b. These species were not formed electro-
chemically because there was no differences in current in the posi-
tively going CVs compared to neat acidic solutions. The mass data
measured by EQCM in neat acidic solution are shown for compari-
son in Fig. 8, and do not show any hysteresis for reversal potentials
below 0.60 VRHE. The adsorbed mass in positively going sweeps
was not dependent on the upper reversal potential used, which is
consistent with our observation that the CV in the positively going
scan is not affected either.
11
adsorption features in neat acidic solutions.
The peak potential for peak “1” occurred at a slightly higher
potential in the sulfuric acid system ͑0.81 VRHE͒ compared to per-
chloric acid ͑0.76–0.80 VRHE͒. The peak potential was slightly de-
pendent on sweep rate, especially in the perchloric acid system
Effect of sweep rate.— The effect of sweep rate on the CVs both
in sulfuric and perchloric acid was assessed, and the data for the
sulfuric acid are shown in Fig. 9. In neat acid the only difference is
in the oxide region where the amount of Pt oxides formed and re-
duced decreases with increasing sweep rate. In electrolytes contain-
ing ammonium, formation of Pt oxides at high potentials also de-
͑
+30 mV shift in the positive direction when increasing the sweep
11
rate one decade͒. Note that this peak occurred at the same potential
as where nitrogen evolution was observed by DEMS; see Fig. 2.
Two reduction peaks were visible in the reverse scan in perchlo-
ric solutions at high potentials ͑“2” and “3” in Fig. 6a͒. Peak 3 was
associated with the reduction of Pt oxides, occurring at 0.80–0.81
VRHE consistent with our observations in neat perchloric acid. The
Pt oxide reduction peak was more rounded and also smaller in the
presence of ammonium. We propose that peak 2 is associated with
peak 1 forming an adsorbed surface redox couple in the system as
further discussed below. Peak 2 occurred at 0.66–0.67 VRHE, at the
11
same potential where N formation was found by DEMS in nega-
2
tively going sweeps; see Fig. 2.
Different reversal potentials were also applied in sulfuric acid
containing ammonium as shown in Fig. 7. Peaks 1 and 2 were not
observable in sulfuric acid solutions at 50 mV/s sweep rate. There
was a clear effect of ammonium on the hydrogen adsorption region
as shown in Fig. 7b. More reduction charge was passed at potentials
below 0.2 VRHE as the reversal potential increased. This was also
seen in the perchloric acid system. However, increasing the reversal
potentials further than 1.25 VRHE did not further affect the CV at low
potentials in the hydrogen adsorption region. The influence of the
reversal potential indicates that at least some of the effects in the
hydrogen adsorption region stemmed from processes taking place at
high potentials. The hydrogen desorption peaks are all identically
independent upon the upper reversal potential because the species
causing the changes in the hydrogen adsorption region are reduced
and removed at low potentials.
Figure 8. ͑Color online͒ EQCM data for Pt in 0.5 M H SO
2
4
+
0.5 M NH OH at 50 mV/s at different upper reversal potentials. The
4
EQCM data measured with neat 0.5 M H SO are also shown for compari-
2
4
2
son with an offset of 5 ng/cm for clarity.
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

B268
Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
Figure 9. ͑Color online͒ The effect of the applied sweep rate on the differ-
ential capacitance ͑i.e., current divided by the sweep rate͒ in ͑a͒ neat 0.5 M
H SO and in ͑b͒ 0.5 M H SO + 0.5 M NH OH. The arrows indicate the
Figure 11. ͑Color online͒ CV performed at 5 mV/s on an RDE with carbon-
supported Pt ͑20 wt % Pt/C, E-TEK͒ in 0.50 M HClO + 0.25 M NH OH at
room temperature with no rotation ͑solid line͒ and with 1000 rpm ͑dotted
line͒.
2
4
2
4
4
4
4
effect of increasing sweep rate. The sweep rates used were 500, 200, 100, 50,
0, 10, and 5 mV/s.
2
creases with increasing sweep rate, but the Pt oxide reduction peak
is independent of sweep rate. The peak in front of the Pt oxide
formation decreases with increasing sweep rate, as does the reduc-
tion shoulder in the double-layer region in the negatively going
scans. The peak related to formation of HUPD at about 0.25 VRHE is
less suppressed at slower sweep rates, indicating that the removal of
adsorbates formed at higher potentials suppressing HUPD is a slow
process. Formation of the adsorbed species related to peak 1 is also
slow as the peak is not seen at high sweep rates, see Fig. 9.
loss in electrolyte containing ammonium was also higher in the
range 0.12–0.08 VRHE in the negatively going scan.
RDE and flow cell experiments.— RDE experiments were per-
formed to clarify if adsorbed or solution species were responsible
for features 1 and 2 observed in CVs measured in ammonium con-
taining electrolyte. If the species in question is a solution species,
peak 2 in the negatively going scan should be strongly minimized
when applying rotation. As is evident from Fig. 11 this is not the
case, and the species responsible for peaks 1 and 2 in Fig. 6 is thus
an adsorbed species.
Time derivative of the mass response.— The time derivative of
the mass changes for neat 0.5 M H SO4 and acid with 0.5 M
2
A flow cell was used to study peaks 1 and 2 shown in Fig. 6. The
WE was held at 0.60 VRHE for 4 min while 0.5 M HClO4
NH OH added is shown in Fig. 10. The curves shown are the aver-
4
age of five different measurements to lower the noise. The mass gain
rate in the double-layer region in the positively going sweep was
slightly lower for the electrolyte containing ammonium. Notice that
there was a slightly higher rate of mass increase in the region 0.8–
+
0.25 M NH OH flowed through the cell at a rate of 5 ␮L/s. After
4
this adsorption period, neat 0.5 M HClO was flushed through the
4
cell to rinse out ammonium, and a positively going scan was then
initiated. The scan limits were 0.55–0.82 VRHE in the first four
sweeps, thereafter sweeping up to 1.15 VRHE and down to 0.06
VRHE. This is compared to the same adsorption experiment, but
where the four intermediate cycles were not performed. From the
1
.0 VRHE in the positively going sweep in the electrolyte containing
ammonium. Also, there was a slightly lower rate of mass loss in the
oxide reduction peak in the negatively going scan. Further, the mass
loss rate associated with the adsorption of the most strongly ad-
sorbed hydrogen was lower in the electrolyte containing ammonium.
However, there was a significantly higher mass loss rate in the re-
gion between the two hydrogen adsorption peaks. The rate of mass
Figure 12. ͑Color online͒ Adsorption in 0.5 M HClO + 0.25 M NH OH at
4
4
0
.60 VRHE for 4 min followed by a positively going scan. The dotted line is
the CV in neat 0.5 M HClO electrolyte, the thin, black solid line is the first
4
scan between the intermediate scan limits ͑0.55–0.82 VRHE͒, the thick black
line is measured after the four intermediate scans shown in the insert. The
gray solid line was measured when not performing the intermediate scans.
The flow rate of both electrolytes was 5 ␮L/s and a polycrystalline Pt elec-
trode was used.
Figure 10. ͑Color online͒ Time derivative of the EQCM data for neat 0.5 M
H SO and with additions of 0.5 M NH OH recorded at 50 mV/s. The data
2
4
4
shown are the average of five data sets for each solution followed by a
numeric 9-point Savitzky-Golay filter ͑Ref. 22͒.
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
B269
insert in Fig. 12 it is clearly seen that the peaks 1 and 2 decrease
rapidly with each scan in the narrow range. Further, note that the
scans between the intermediate scan limits only affect peaks 1 and 2
in the following complete stripping scan as compared to the strip-
ping scan where no intermediate scans were performed.
mechanism, it seems natural to assume that the adsorbed species
could be Nads formed from NHx,ads in the positively going sweep.
Future studies, including techniques such as Fourier transform infra-
red, may be helpful to resolve this matter.
6
Formation of Nads was also reported by de Vooys et al. for
alkaline solutions but then in the potential range 0.4–0.6 VRHE. It is
not clear if Nads would be further oxidized at high potentials. In the
reverse sweep, Nads may be reduced, forming either nitrogen via
NHx,ads with no net faradaic current, or reductively formed ammo-
Discussion
Effects on hydrogen desorption and the double-layer region in
positively going sweeps.— The hydrogen desorption peaks are not
affected by the presence of ammonium. This shows that all adsorbed
species are removed in the negatively going scan so that the effect of
ammonium in the positively going scan is not initially noticeable.
7
nium as reported by Gootzen et al. in alkaline solution. Adsorbates
resulting from NO exposure is only reduced at potentials below 0.25
9
VRHE
.
6
The N signal in the DEMS in the negatively going sweep is very
This was also reported by de Vooys et al. in alkaline solution.
2
similar both in perchloric and sulfuric acid electrolytes; see Fig. 1
Further, there is no difference between CVs measured in neat acidic
solutions and acids containing ammonium in the double-layer region
in the positively going scan. Thus, electrochemical formation of
and 2. This indicates that the source of N formation in the nega-
2
tively going sweep is an adsorbed species with similar surface cov-
erage in both sulfuric and perchloric acid containing ammonium.
adsorbed species does not take place at potentials below 0.65 VRHE
,
and oxidation of ammonium commences only at about 0.65 VRHE. A
Effects on platinum oxides.— The Pt oxide formation and reduc-
tion is significantly affected by the presence of ammonium. The
onset of Pt oxide formation is less sharp compared to neat acidic
solution; see Fig. 4a. There is a significantly higher oxidation cur-
rent at potentials above 1.0 VRHE in the presence of ammonium.
Formation of NO was seen by DEMS only at potentials above 1.2
VRHE; see Fig. 1c and 2c.
significant oxidation current was observed at potentials as low as
7,8
0.45 VRHE in alkaline solutions.
EQCM data show that an adsorbed species is formed at poten-
tials above 0.25 VRHE because the hysteresis of the mass response
increased upon addition of ammonium; see Fig. 8. We propose that
this is adsorption of bisulfate stabilized by ammonia as also reported
23
for Pt͑111͒ by Shingaya et al.
The Pt oxide reduction peak appearing at about 0.8 VRHE is
shifted negatively in the presence of ammonium; see Fig. 4a. How-
ever, the most noticeable effect of ammonium is that the charge
associated with the Pt oxide reduction peak is almost independent of
sweep rate, in contrast to what is found in neat acid where the peak
decreases with increasing sweep rate, Fig. 9. The charge associated
with the Pt oxide reduction peak is also much smaller than in the
neat acidic solution. Similar observations were made by Wasmus et
al. and Gootzen et al. in alkaline electrolytes. This indicates that
the amount of “normal” Pt oxides formed is limited by another
adsorbed surface species. Figure 12 shows that the adsorbed species
responsible for peaks 1 and 2 is not responsible for the changes
observed in the oxide region.
Presence of a redox couple.— Peaks 1 and 2 in Fig. 6 coincide
with the N peaks seen in the DEMS measurements in positively
2
and negatively going scans, respectively; see Fig. 1b and 2b. It is
therefore likely that these peaks are associated with nitrogen evolu-
tion. The ratio between the MS signal and the faradaic current was
calibrated from measurements of ammonia oxidation in alkaline so-
lution ͑data not shown here͒. This calibration accounts for collection
efficiency, ionization probability in the MS, as well as other factors
influencing the signal strength in the DEMS. We estimate that the
8
7
N signal in the positively going scan accounts for less than 10% of
the current. This shows that the charge associated with peak 1 is
2
mainly due to a different reaction than N evolution, i.e., formation
2
of an adsorbed intermediate species in the ammonium oxidation
process as further discussed below. Further, it is likely that peak 2
represents reduction of the species formed in peak 1 in the positively
going sweep. A closer inspection of the EQCM mass response in
Fig. 4b and 10 reveals that the mass of the WE increases faster in
ammonium containing solution than in neat acid solution in the posi-
tively going sweep in the range 0.75–0.95 VRHE, possibly caused by
adsorption of the species involved in peak 1. Further, CVs recorded
with and without rotation shown in Fig. 11 in an RDE cell were
identical, showing that the species is adsorbed.
Adsorbed N-O species could also be present. Detection of NO by
DEMS at high potentials ͑see Fig. 1 and 2͒ suggests the presence of
adsorbed N-O species at high potentials. de Vooys et al. proposed
6
that metal oxynitrides are present on the Pt surface in alkaline solu-
tions containing ammonia at high potentials, and further proposed
that the oxynitride species were very stable and inactive in the for-
8
mation of NO and N2O observed. Wasmus et al. found, when
sweeping from high to low potentials, that nitric oxides were
evolved at low potentials and explained this by reduction of highly
oxidized nitrogen species formed at high potentials.
The results in Fig. 12 show that the adsorbate formed at 0.60
VRHE is actively taking part in the process related to peaks 1 and 2.
The peaks decreased when sweeping back and forth with the nar-
rower potential limits. This reduction can partly be due to evolution
Effects in the double-layer region in negatively going scans.—
A substantially higher reduction current was found in the double-
layer region in negatively going sweeps in the presence of ammo-
nium than in neat acidic solutions. The reductive current increases
with ammonium concentration ͑Fig. 4a͒ and upper reversal potential
͑Fig. 6 and 7͒, whereas it decreases with increasing sweep rate; see
Fig. 9. We propose that less adsorbed species are formed in the
positive scan at high sweep rate, whereas more adsorbed species are
formed at high ammonium concentration and also with higher rever-
sal potentials. The reduction charge in the double-layer region is
also much higher in alkaline solutions containing ammonia than in
of N thus consuming the adsorbate, or due to other processes where
2
other volatile or soluble species are formed. We see it as unlikely
that the peaks are due to formation-reduction of Pt oxides, as the Pt
oxide formation and reduction at higher potentials is not affected by
the species responsible for peaks 1 and 2. This is concluded from the
fact that the CVs are identical at higher potentials independent of if
the sweeps between the narrow potential limits are performed or not.
Peak 1 is more noticeable in perchloric than in sulfuric acid ͑see Fig.
7
,8
6
and 7͒ probably also because ͑bi͒sulfate adsorbes more efficiently
neat alkaline solutions,
suggesting that the same species are
on Pt, thus partially blocking the surface for adsorption of other
formed in alkaline solutions.
species which in turn is reflected in the lower N yield in sulfuric
2
compared to perchloric acid.
The nature of the species has not been identified, nor the chemi-
cal reactions that yield the species responsible for peaks 1 and 2.
Effects on hydrogen adsorption.— There is a significant effect
on the hydrogen adsorption peaks when ammonium is present in the
solution. The most marked effect is that the peak at 0.25 VRHE as-
sociated with the most strongly bound hydrogen is shifted to lower
potentials. This effect is more pronounced with increasing ammo-
nium concentration ͑see Fig. 4a͒, increasing upper reversal potential
The observation of N evolution at the same potentials as peaks 1
2
and 2, however, indicates that the species is related to the mecha-
nism leading to N evolution. Based on the Gerischer and Mauerer
2
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

B270
Journal of The Electrochemical Society, 154 ͑2͒ B263-B270 ͑2007͒
for the CVs in the ͑bi͒sulfate system ͑see Fig. 7͒, and higher sweep
rates ͑see Fig. 9͒. The effect on the most strongly adsorbed hydrogen
in the perchloric system is small; see Fig. 6.
The hydrogen adsorption region is strongly affected by ammo-
nium, in particular the most strongly adsorbed hydrogen, the forma-
tion of which is shifted to lower potentials. Both species formed at
very high potentials as well as coadsorption of ammonia and bisul-
fate on Pt are responsible for the shift of hydrogen adsorption to
lower potentials. The observation that the most strongly adsorbed
hydrogen is less negatively shifted in perchloric acid also shows that
there is an anion effect in the sulfuric acid system due to coadsorp-
tion of ammonia and ͑bi͒sulfate. The hydrogen desorption region is
not affected by ammonium in either acidic solution, showing that the
electrode surface is fully restored in the reductive scan.
23
Shingaya et al. found that ammonia and bisulfate coadsorbes
on Pt͑111͒, i.e., ammonia and bisulfate mutually stabilize each other
so that they desorb at lower potentials than in neat acidic solution,
thus shifting formation of strongly adsorbed hydrogen to lower po-
tentials. It is also known that the peak associated with the most
strongly adsorbed hydrogen is strongly affected by adsorption of
2
4
23
bisulfate. The findings of Shingaya et al. are very similar to our
findings; the effect on the most strongly bound hydrogen peak is
noticeable even at quite low upper reversal potentials ͑0.80 VRHE
;
see Fig. 7͒, and the effect is more pronounced at higher sweep rates;
see Fig. 9b. For 0.1 M HF containing 0.01 M ammonium, Shingaya
et al. reported very small effects on the hydrogen adsorption
Acknowledgments
Norsk Hydro ASA and The Research Council of Norway sup-
ported this work financially.
2
3
features. This is in line with our observations in perchloric acid
where the effect of ammonium on strongly adsorbed hydrogen is
much smaller than in sulfuric acid.
Norwegian University of Science and Technology assisted in meeting the
publication costs of this article.
At potentials below 0.15 VRHE the reduction current increases
with increasing reversal potential both in sulfuric and perchloric
acid solutions containing ammonium; see Fig. 6 and 7. The EQCM
data show that there is a mass loss on the EQCM which is not seen
References
1
2
. F. A. Uribe, T. A. Zawodzinski, Jr., and S. Gottesfeld, J. Electrochem. Soc., 149,
A293 ͑2002͒.
. H. J. Soto, W-K. Lee, J. W. Van Zee, and M. Murthy, Electrochem. Solid-State
Lett., 6, A133 ͑2003͒.
. R. Halseid, P. J. S. Vie, and R. Tunold, J. Power Sources, 154, 343 ͑2006͒.
. H. G. Oswin and M. Salomon, Can. J. Chem., 41, 1686 ͑1963͒.
. H. Gerischer and A. Mauerer, J. Electroanal. Chem. Interfacial Electrochem., 25,
421 ͑1970͒.
6. A. C. A. de Vooys, M. T. M. Koper, R A. van Santen, and J. A. R. van Veen, J.
Electroanal. Chem., 506, 127 ͑2001͒.
in neat acidic solution in the potential region 0.12–0.08 V
negatively going scan; see Fig. 10. Wasmus et al. reported forma-
in the
RHE
8
3
4
5
tion of N , NO, and N O at these low potentials in their alkaline
2
2
electrolyte. We suggest that a similar desorption reaction occurs in
our acidic system where a highly oxidized, very stable adsorbed
nitrogen species is reduced to form either nitrous oxides, N , or
2
7
. J. F. E. Gootzen, A. H. Wonders, W. Visscher, R. A. van Santen, and J. A. R. van
Veen, Electrochim. Acta, 43, 1851 ͑1998͒.
8. S. Wasmus, E. J. Vasini, M. Krausa, H. T. Mishima, and W. Vielstich, Electrochim.
ammonium, or possibly all these species. Removal of these species
is relatively slow, as seen by the fact that the effect on formation of
strongly bound Hads at about 0.25 VRHE is less pronounced when
applying lower sweep rates; Fig. 9b.
Acta, 39, 23 ͑1994͒.
9
. A. C. A. de Vooys, M. T. M. Koper, R. A. van Santen, and J. A. R. van Veen,
Electrochim. Acta, 46, 923 ͑2001͒.
1
0. S. Wasmus, S. R. Samms, and R. F. Savinell, J. Electrochem. Soc., 142, 1183
1995͒.
11. R. Halseid. Ph.D. Thesis, Norwegian University of Science and Technology ͑2004͒.
Conclusions
͑
We have found that ammonium is electrochemically active in
acidic solutions at room temperature, but the oxidation is slow. CV
and EQCM data show that presence of ammonium affects the for-
mation and reduction of Pt oxide at high potentials involving the
formation of very stable nitrogen or nitrogen-oxygen species. Some
of these species are removed from the electrode only at very low
potentials in the hydrogen adsorption region, and removal of these
species is slow. An oxidation peak occurs in positively going scans
at about 0.8 VRHE, which decreased strongly with increasing sweep
rate, and a corresponding reduction shoulder at about 0.66 VRHE is
seen which could be due to formation of NHx,ads species on the
surface. Formation of Pt oxides is shifted to slightly higher poten-
tials in the presence of ammonium, and the charge of the normal Pt
oxide reduction peak observed at about 0.80 VRHE is independent on
sweep rate, indicating the amount of normal Pt oxides formed is
limited by other adsorbates.
1
2. R. Raudonis, D. Plaušinaitis, and V. Daujotis, J. Electroanal. Chem., 358, 351
1993͒.
3. C. P. Wilde, S. V. De Cliff, K. C. Hui, and D. J. L. Brett, Electrochim. Acta, 45,
3649 ͑2000͒.
͑
1
14. V. Tsionsky, L. Daikhin, M. Urbakh, and E. Gileadi. in Electroanalytical
Chemistry—A Series of Advances, A. J. Bard and I. Rubinstein, Editors, Vol. 22, p.
1
, Marcel Dekker Inc., New York ͑2004͒.
1
5. Z. Jusys and S. Bruckenstein, Electrochem. Solid-State Lett., 1, 74 ͑1998͒.
16. G. Sauerbrey, Z. Phys., 155, 206 ͑1959͒.
17. C. Gabrielli, M. Keddam, and R. Torresi, J. Electrochem. Soc., 138, 2657 ͑1991͒.
1
8. T. J. Schmidt, U. A. Paulus, H. A. Gasteiger, and R. J. Behm, J. Electroanal.
Chem., 508, 41 ͑2001͒.
19. Z. Jusys, J. Kaiser, and R. J. Behm, Electrochim. Acta, 49, 1297 ͑2004͒.
20. Y. Lim and E. Hwang, Bull. Korean Chem. Soc., 18, 6 ͑1997͒.
21. Z. Jusys, H. Massong, and H. Baltruschat, J. Electrochem. Soc., 146, 1093 ͑1999͒.
22. A. Savitzky and M. J. E. Golay, Anal. Chem., 36, 1627 ͑1964͒.
23. Y. Shingaya, H. Kubo, and M. Ito, Surf. Sci., 428, 173 ͑1999͒.
24. J. C. Huang, W. E. O’Grady, and E. Yeager, J. Electrochem. Soc., 124, 1732
͑1977͒.
Downloaded on 2014-10-22 to IP 128.148.252.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).