Hydroxylamine Oxidation on Polycrystalline Gold Electrodes in Aqueous Electrolytes: Quantitative On-Line Mass Spectrometry under Forced Convection

Article abstract of DOI:10.1002/cphc.201900367

Herein, a method is presented that allows quantitative determination of faradaic efficiencies for dinitrogen (N₂) generation during the electrochemical oxidation of hydroxylamine (NH₂OH), (Formula presented.), on a polycrystalline gold Au(poly) disk electrode in aqueous electrolytes over a wide pH range. This tactic involves the use of an impinging jet electrolyte configuration incorporating a gas porous ring connected in turn to a mass spectrometer. The actual amount of N₂ generated at the Au(poly) disk was assayed using the oxidation of hydrazine (N₂H₄) in aqueous phosphate buffer (pH 7). This redox process yields N₂ as the only product, allowing a direct correlation to be established between the changes in the partial pressures of N₂ and the current flowing through the disk electrode. An analysis of the data collected revealed a strong dependence of (Formula presented.) both on pH and the applied potential. Although values of (Formula presented.) as high as 20 to 30 % were found in acid and neutral media over a narrow potential region, those in alkaline solution were far smaller in the entire potential range examined.

Full text of DOI:10.1002/cphc.201900367

Accepted Article
Title: HYDROXYLAMINE OXIDATION ON POLYCRYSTALLINE
GOLD ELECTRODES IN AQUEOUS ELECTROLYTES:
QUANTITATIVE ON-LINE MASS SPECTROMETRY UNDER
FORCED CONVECTION
Authors: Boguslaw Pozniak, Imre Treufeld, and Daniel Scherson
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemPhysChem 10.1002/cphc.201900367
Link to VoR: http://dx.doi.org/10.1002/cphc.201900367
A Journal of
www.chemphyschem.org

ChemPhysChem
10.1002/cphc.201900367
HYDROXYLAMINE OXIDATION ON POLYCRYSTALLINE GOLD ELECTRODES IN
AQUEOUS ELECTROLYTES: QUANTITATIVE ON-LINE MASS SPECTROMETRY UNDER
FORCED CONVECTION
+
Boguslaw Pozniak , Imre Treufeld, and Daniel Scherson
Department of Chemistry
Case Western Reserve University
Cleveland, OH 44106-7078
+
Institute of Heavy Organic Synthesis "Blachownia", ul. Energetykow 9,
4
7-255 Kedzierzyn-Kozle, Poland.
ABSTRACT
A method is herein presented that allows quantitative determination of faradaic efficiencies for
NH OH
2
dinitrogen, N
2
, generation, during the electrochemical oxidation of hydroxylamine, NH
2
OH, fN2
,
on a polycrystalline, Au(poly), gold disk electrode in aqueous electrolytes over a wide pH range. This
tactic involves the use of an impinging jet electrolyte configuration incorporating a gas porous ring
connected in turn to a mass spectrometer. The actual amount of N
assayed using the oxidation of hydrazine, N , in aqueous phosphate buffer (pH 7). This redox
process yields N as the only product, allowing a direct correlation to be established between the
changes in the partial pressures of N
2
generated at the Au(poly) disk was
2
H
4
2
2
and the current flowing through the disk electrode. An analysis of
NH OH
2
the data collected revealed a strong dependence of fN2
both on pH and the applied potential.
NH OH
2
Although values of fN2
as high as 20 to 30% were found in acid and neutral media over a narrow
potential region, those in alkaline solution were far smaller in the entire potential range examined.
INTRODUCTION
The conversion of nitrate, NO
3
-, derived primarily from fertilizers, into dinitrogen, N
, in aqueous elec-
2
trolytes has been regarded as among the most desirable tactics toward reducing deleterious effects asso-
1
ciated with algal blooms on often fragile ecosystems. As has been pointed out in the literature, this
goal could potentially be accomplished using electrochemical means and, indeed, several groups are
searching for electrocatalysts capable of carrying out this complex multielectron transfer process with
2
optimum efficacy and high specificity. Among the promising strategies being considered is the use of
two or more electrocatalysts acting in series, as illustrated more than a decade ago in our research
3
group. The specific electrode employed in that study consisted of Au nanoparticles dispersed on a
glassy carbon surface modified by a monolayer of adsorbed hemin, Hm. Polarization of such composite
-
2+
electrode at sufficiently negative potentials in mildly acidic NO
3
(aq) solutions containing Cd (aq) led
-
-
to the reduction of NO
subsequently reduced to hydroxylamine, NH
forts have focused on the search of electrocatalysts that might oxidize NH
3
on underpotential deposited cadmium on Au to yield nitrite, NO
2
, which was
2
OH, by the adsorbed macrocycle. More recently, our ef-
OH to N . In particular, in
2
2
situ surface-enhanced Raman scattering (SERS) data collected with roughened Au electrodes in mildly
acidic (pH 3) buffered aqueous electrolytes provided evidence for the presence of adsorbed nitrite,
-
4
NO
similar experiments performed in solutions containing NO
identical conditions displayed only the spectral feature attributed to NO
2
(ads), and NO
2
(ads) within the potential range in which NH
2
OH was found to oxidize. Moreover,
(aq) instead of NH OH under otherwise
(ads) over the range of 0.1 < E
-
2
2
-
2
1
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
<
0.8 V versus SCE with no additional features at higher potentials. These observations strongly sug-
gest that under the conditions selected for these studies the oxidation of NH OH on Au proceeds at
2
least in part through a pathway that does not involve nitrite as a solution-phase intermediate. Comple-
mentary experiments involving in situ normalized IRAS spectra yielded a prominent peak centered at
−
1
2
230 cm , characteristic of nitrous oxide, N
2
O, at a potential very close to the onset of NH OH oxida-
2
tion, the intensity of which reached a maximum at about the same potential as that found the dynamic
5
polarization recorded with a rotating disk electrode in separate measurements in the same electrolyte.
Based on the electrochemical and spectroscopic information, and theoretical simulations, the observed
-
behavior could be accounted for by an EEECE mechanism in which NO
NH OH oxidation, reacts with NH
2
(aq), one of the products of
2
2
OH yielding an electrochemically inert species. In related studies,
-
the groups of Koper and Feliu have shown that nitrite, NO
2
, can be reduced to N on quasi perfect
2
6
Pt(100) electrodes in alkaline media.
This communication describes a method that relies on on-line differential electrochemical mass
7
, 8
9
spectrometry (DEMS) under forced convection developed in our laboratory, to determine absolute
NH OH
2
faradaic efficiencies for generation of N
2
via oxidation of NH
2
OH, fN2
, on a solid polycrystalline
Au, Au(poly), disk electrode as a function of potential in aqueous electrolytes over a wide pH range. As
NH OH
2
will be shown, values of fN2
as high as 20 – 30% were found for acidic and neutral solutions, but
far smaller for alkaline media. Additional gaseous reaction products observed included N
the relative amounts of which were found to depend strongly on the pH of the solution.
2
O and NO,
Experimental
a.
the phosphate salt (3:1) (NH
Na HPO4, >99.0 %, Fisher Scientific), sodium phosphate monobasic anhydrous (NaH
Fisher BioReagents) and o-phosphoric acid (H PO , 85 %, Fisher Scientific) involved in the
Chemicals – Hydrazine monohydrate (N
OH, 98%, Acros Organics), sodium phosphate dibasic anhydrous
PO >99%,
2
H
4
·H O, 100.4%, Sigma-Aldrich), hydroxylamine, as
2
2
(
2
2
4
3
4
preparation of buffers, were all used as received. All solutions were made using ultrapure water (18.2
MΩ) generated by a Barnstead EASY-pureUV system and their pH measured by a pH-meter (Accumet
9
15, Fisher Scientific).
b. On-line Mass Spectrometry in the Presence of Convective Flow - The specific strategy
developed in our laboratory for acquiring on-line mass spectra under forced convection described in
detail in an earlier publication, relies on the use of a submerged laminar-flow jet of circular cross
section (0.5 mm in diameter) placed along the normal to the surface of a disk electrode (5 mm
diameter) surrounded by a concentric porous Teflon ring covered by a porous Teflon membrane
9
permeable to gases (see Insert, Fig. 1). A similar approach was later reported by Venkatachalam et
1
0
al., who used a wall-jet electrolyte impinging onto Pt film working electrode ca. 100 nm thick
sputtered onto a thin, hydrophobic membrane, which served as the inlet to the mass spectrometry
chamber. For the present experiments, the tip of the nozzle was placed ca. 5 mm below the disk
electrode surface. As before, the jet was produced by an external pump (FMI Labpump model RHV),
which was calibrated by pumping water from one graduated cylinder to another. After this step was
completed, no further adjustments to the relative positions of the jet and the cell were made for all
subsequent experiments.
c.
Electrochemical Cell and Instrumentation Either a reversible hydrogen electrode (RHE) or a
Ag/AgCl electrode were used as reference, and a graphite rod as the counter electrode for all solutions.
The total volume of electrolyte in the cell was ca. 360 mL. Prior to the actual experiments, all solutions
were deaerated by bubbling Ar (Airgas, prepurified). The electrode potential, E, was controlled with a
commercial potentiostat (RDE 4, Pine Instruments). Cyclic voltammetric and dynamic polarization
2
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
curves were acquired at a rate of  = 5 mV/s and recorded with a National Instruments Data
Acquisition Card (USB-6008) using custom programs written in LabView.
d.
Mass Spectrometry - Equipment and Calibration. The vacuum system consisted of a
manifold with connectors to the electrode assembly, a pressure gauge (Granville-Phillips 350), a turbo
pump, and a mass spectrometer (Residual Gas Analyzer, RGA 200, Stanford Research Systems)
-5
operating at a base pressure of ca. 1.6  10 Torr (mostly water). In order to reduce the overall
background pressure, and particularly that of N , the electrochemical cell, including the pump, were
placed in a small controlled atmosphere box, which was purged with Ar for an extended period of time
to remove atmospheric gases, at which point either N or NH OH were added to the solution through
a glass tube to avoid exposure to the atmosphere. The RGA 200 was operated in the "pressure versus
time" mode and set to monitor six gases: N (m/z 28), O (m/z 32), NO (m/z 30), N O (m/z 44) and
NO (m/z 46) and Ar (m/z 40), and N (m/z 14). No significant ion peaks were observed above m/z 46.
2
2
H
4
2
2
2
2
2
Ions were produced by electron impact ionization (70 eV). All experimental data reported herein was
collected at the fastest rate available with the instrument, i.e. one point per s. Although the instrument
was operated under ambient conditions, the temperature within the mass spectrometer chamber was
higher due to heat produced by the filaments. Before each set of measurements, the instrument was run
for several hours with the electrolyte jet impinging on the electrode to stabilize the pressure readings.
As is well known, N
correct their measured partial pressures in order to determine quantitatively the actual amounts of these
three species produced by the electrode entering the MS chamber. To this end, high purity N O gas (8g)
2
O undergoes fragmentation to yield both NO and N
2
, making it necessary to
2
contained in a small commercial cartridge (Ultra-Purewhip Cream Chargers WCC-MN-PW-UL-24,
Creamright, FL), was first transferred into an empty "whipping cream maker" (Mosa, WCD-MO-
LC25W) and then to a balloon connected in turn to a T-joint attached to the same line that supplies Ar
to the electrochemical cell. The 0.1 M phosphate buffer pH = 7 solution was first purged with Ar and
then with N
8, N ; 44, N
electrochemical inertness, no differences could be observed between data collected with the electrode
polarized at 0.0 V, or while cycling over the range -0.6 to 1.6 in N O saturated solutions. The relative
abundances of NO and N2, the two largest N O fragments, using the peak amplitude of N O as a
2
O, while monitoring seven different gases with the mass spectrometer, i.e. m/z = 18, H
2
O;
2
2
2
O; 32, O ; 30, NO; 46, NO ; and 14, N. As may be expected, in view of its
2
2
2
2
2
normalizing factor, are shown in the third column, Table I, where the numbers in parenthesis are those
reported by NIST. Although the values for m/z 28 were similar, that for m/z 30 found in this work was
only half. Such discrepancy is due, in all likelihood, to the specific response of our gas analyzer/mass
spectrometer.
Table I
Fragmentation Pattern of N
2
O
†
%
abundance
Fragment m/z
NO
30
28
16.4(31.13)*
11.1(10.81)
N
2
†
Values normalized by the amplitude of the N
2
O peak
*
http://chemdata.nist.gov/dokuwiki/doku.php?id=chemdata:start
On this basis, the changes in the partial pressures associated with both electrochemical and chemical
int
processes at the Au(poly)|electrolyte interface, pX , for X = N
2
, and NO will be given by the
difference between the measured pressure changes for these gases, pX , and the contributions due to
N
2
O fragmentation according to the data recorded in our laboratory shown in Table I, i.e.
3
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
int

p p 0.111p
N
N
N O
2
2
2
(1)
int
NO
p
p 0.164p
NO
N O
2
9
where p in the measured change in the pressure of N
2
O. As in previous work, the instrument was
on Au(poly) in 0.1 M phosphate buffer (PB, pH 7) as a standard.
as the only nitrogen containing product, and also very well-defined diffusion
limited currents, ilim. Under the conditions of our measurements,
N O
2
calibrated using the oxidation N
This reaction yields N
2
H
4
2
o
2/3 5/12 3/4 1/2 3/4
i 1.60knFC D  v a R
(2)
lim
f
where k is a dimensionless empirical constant, the value of which was determined in previous
9
experiments to be 0.868 ± 0.032, and thus in excellent agreement with that reported by Yamada et
1
1
o
al., i.e. 0.86; n, the number of electrons transferred, i.e. 4; F, Faraday’s constant; C and D, the bulk
3
−5
2
9
concentration in mol/cm , and the diffusion coefficient of N
kinematic viscosity in cm /s; v , the flow rate in cm /s; a, the internal diameter of a nozzle in cm, and R
2
H
4
, i.e. 1.33 × 10 cm /s; v, the
2
3
f
3
/4
is the radius of a disk electrode in cm. In agreement with Eq.(2) above, a plot of ilim vs vf (see Fig. 1),
and of ilim vs [N ] (not shown here), yielded straight lines with a negligible intercept.
2
H
4
3
/4
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Figure 1. Plot of ilim, vs.  f for the
oxidation of 1 mM N on Au(poly) in 0.1
2
H
4
M PB, pH = 7 recorded with the jet
impinging electrolyte set up (see text). The
statistical analysis yielded for the linear fit, a
slope, S = 1.3312 ± 0.01743 mA s /cm ,
an intercept, I = 0.05278 ± 0.01245 mA, and
an adjusted R = 0.99846. Insert: Schematic
3
/4
9/4
2
diagram of the jet impinging arrangement.
See text for details.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
 3f /4, cm9/4/s3/4
Quantitative correlations between the current flowing through the jet-impinging Au(poly) disk, i, at
a fixed v , and the partial pressure of N (m/z = 28), as measured
f
2
by the mass spectrometer, p , were sought from experiments in
N
2
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
N2
1
.0
lim
N2
which the disk potential, E, was scanned linearly at a rate ν = 5
p
0.8
0.6
Figure 2. Dynamic polarization curve (ν = 5 mV/s)
0
.4
.2
recorded with the jet impinging electrolyte system (v
.394 mL/s) in a 1.2 mM N in 0.1 M PB (pH = 7)
f
=
0
0
2
H
4
o
N2
p
0.0
solution (right ordinate, black curve). The curve in blue is
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
E, V vs Ag/AgCl
a plot of pN2 (left ordinate) vs E collected
simultaneously.
4
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
mV/s. Shown in Fig. 2, as a means of illustration, are plots of i vs E for v
f
= 0.394 mL/s (right ordinate,
black curve), and pN2 vs E, recorded simultaneously (left ordinate, blue curve) for the oxidation of 1.2
mM N in 0.1 M PB (pH = 7).
Systematic experiments of this type were performed by adding small aliquots of a concentrated N
2
H
4
2
H
4
stock solution to the cell under otherwise identical conditions, and the results then used to construct a
lim
lim
lim
o
N
o
N2
plot of pN2 vs ilim, where p  p  p , and
p
is the value of background pN2 recorded at
N
N
2
2
2
potentials negative to the onset of N
2
H
4
oxidation (see Fig. 2), derived from residual (background) N in
2
the chamber. As evidenced from the data in Fig. 3, such plot was found to be remarkably linear, allowing
for the faradaic efficiency for the electrochemical generation of N
2
in the case of NH
2
OH oxidation to be
determined with great accuracy (vide infra).
c, m M
lim
Figure 3. Plot of pN2 vs
6
i
N
7
lim for the oxidation of
in 0.1 M PB (pH =
), in the jet impinging
electrolyte arrangement, (v
0.394 mL/s) (see text for
5
4
3
2
H
4
f
=
details). The statistical
analysis yielded for the
linear fit (see green line),

2
-7
1
0
S = 2.846  10 Torr/mA;
-9
I = - 1.944  10 Torr;
2
R = 0.99982
0.0
0.5
1.0
1.5
2.0
ilim, mA
Additional experiments were performed in the same buffered N
2
H
4
solutions in which the potential was
first scanned from 0.0 V, a value at which no current flows, up to 0.95 V, i.e. the limiting current, ilim
,
for N oxidation, followed by a potential step down to 0.0 V vs Ag/AgCl. Based on the results
2
H
4
obtained (see Fig. 4), the time response of the system was ca. 2 s, which for ν = 5 mV/s amounts to 10
mV.
2.0
1.5
1.0
0.5
0.0
Fig. 4. Plot of i vs. time (red line, left
ordinate), recorded while scanning the
potential of the Au(poly) at 5 mV/s
from 0.0 V, i.e. i = 0, to 0.95 V, a value
at which the oxidation of N
2
4
H proceeds
0.95 V
under diffusion limited conditions
(black line), followed by a potential step
down to 0.0 V. Also shown in blue
symbols (right ordinate) is the
Scan
Starts
100 s
corresponding pN2 vs E response
Time
recorded simultaneously.
0.0 V
5
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
Results and Discussion
Unless otherwise specified, all experiments described in this section were performed with a
Au(poly) electrode in 10 mM NH OH in 0.1 M PB solutions of three different pH values, i.e. 2.92,
.00, and 12.03, and v
= 0.39 mL/s, yielding pN2 values well within the range involved in the
2
7
f
calibration curve (see Fig. 3). Shown in the lower panel, Fig. 5 (blue curve, right ordinate), are four
consecutive dynamic polarization curves ( = 5 mV/s) recorded toward positive, E, and subsequent
negative directions, E, (see arrows) for a pH 2.92 solution, in the range -0.1 V < E < 1.4 V vs
int
Ag|AgCl. Plots of pX vs E for X = N
and NO (see Eq.(1)), and the corresponding background
2
corrected N
2
O curve recorded simultaneously are displayed in the upper panel in this figure. Careful
inspection of these data revealed a number of interesting features. Specifically,
i.
No increase in the pressure beyond their background values was detected for all the gases
int
being monitored, i.e. pX = 0, for E in the range -0.1 V up to onset of the current, i, at E
int
ca. 0.4 V. As E was further increased, all pX vs E curves displayed bell-shape character,
yielding negligible values for E > 0.86 V, i.e. positive to the onset of Au oxide formation on
Au(poly) in the neat electrolyte (see grey curve in the lower panel, Fig. 5), despite the fact
that the current continued to increase up to 1.15 V and later decreased to non-negligible val-
ues. This behavior is consistent with the further oxidation of species, such as NO and NO
2
produced via NH
2
OH oxidation, to yield ionic species, such as nitrite and nitrate.
int
ii.
No changes in pX were found during the subsequent E, down to 0.86 V, at which point,
int
all pX suddenly increased to reach values much higher than those found during E at this
potential, to yield, thereafter, rather distorted bell-shaped curves down to 0.4 V.
int
iii.
Clear p
signals could be observed in the E scans at potentials at which the current
N O
2
was negligible, i.e. E < 0.4 V, which strongly suggest that Au(poly) can promote the hetero-
geneous catalytic decomposition of NH OH. It should be noted in this regard, that aqueous
NH OH solutions undergo spontaneous decomposition both in acid and alkaline media
yielding, respectively, N O and NH , and N and NH , a process that is catalyzed by a num-
ber of metal cations. In particular, the presence of ferric or ferrous ions in 50%
NH OH/H O solutions was found to generate N and N O in close to equal ratios. Similar
experiments in which iron oxide particles were dispersed in pure water yielded a much
2
2
2
3
2
3
1
2
2
2
2
2
1
2
higher N
2
:N
2
O ratio, i.e. ca. 2:1. Some evidence that Pt can catalyze NH
2
OH decomposi-
tion in neutral aqueous solutions was obtained from in situ mass spectrometric data reported
1
3
by Karabinas et al., who detected N
2
O at potentials more negative than the onset of elec-
trochemical oxidation of NH OH.
2
iv.
Both the electrochemical and mass spectrometric data yielded virtually identical results in
all four consecutive cycles.
6
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
E, V vs Ag/AgCl
0.4 0.6 0.8
Figure 5. Lower Panel. Dynamic
polarization curve ( = 5 mV/s) for
NH OH oxidation on Au(poly) electrode
-0.2
0.0
0.2
1.0
1.2
1.4
2
2
N2
2
in a 10 mM NH OH in 0.1 M PB
2
NO
solution (pH 2.92) recorded at a flow rate
f
v = 0.39 mL/s, (blue curve). Cyclic
4
N2O
voltammetry ( = 5 mV/s ) of the
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
5
electrode in the base electrolyte (no
NH OH, grey trace). Upper Panel. Plots
2
0
5
int
of pX vs E for X = N
(blue), and NO
2
-
(red), and corresponding background
-
10
corrected N O curve (black), recorded
2
-15
0.0
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
while acquiring the dynamic polarization
curve. See text for details.
E, V vs Ag/AgCl
Insight into the rather large hysteresis observed in the potential range 0.4 < E < 1.4 V was
gained from identical measurements, for which Ewas reversed at E = 1.1 V, i.e. just positive to
the onset of Au oxide formation. As shown in Fig. 6, both the electrochemical and MS response
for E and E were very similar. In marked contrast to the behavior found for the wider cycles,
int
both i and pX monotonically decreased (the roman numerals in the panels refer to the cycle
number), as the number of cycles was increased. A possible explanation for this effect may be
found in the formation of an adsorbed species derived from the oxidation of NH OH, which
2
blocks the electrode surface from further reaction for E ≤ 1.1 V, and either desorbs and/or oxi-
dizes at higher potentials to yield one or more species that have no affinity for the oxidized Au
1
4
surface. rotating disk electrode techniques for the low index faces of Au single crystal. As
mentioned in the Introduction the most likely species based on the in situ SERS data reported in
4
a previous communication is NO
2
(ads). Further evidence of an adsorbed species was obtained
from differences in the rates constants for the reac-
tion for the low index faces of Au single crystal as
measured with rotating disk electrodes.14 As the
E, V vs Ag/AgCl
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
I
0.8
0.9
1.0
1.1
Au oxide is reduced, the presumably bare Au sur-
face acquires very high activity toward NH OH ox-
2
1
N2
V
idation yielding, within the time scale of the meas-
urements, currents which exceed those found dur-
ing the scan toward positive potentials, as clearly
seen in Fig. 5 above.
1
I
NO
I

N2O
2
Similar measurements were also performed for
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
10mM NH
2
OH in neutral (pH 7) and alkaline (pH
1
2.03) PB solutions, and the i vs. E (thick grey
I
int
curves, right ordinates), and pX vs E (vide supra)
II
V
Figure 6. Same as Fig. 5 for a narrower potential
range. The roman numerals refer to the cycle
number for numerous consecutive cycles (see text
for details). Note that only a few such cycles are
shown for clarity.
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
E, V vs Ag/AgCl
7
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
curves for N
2
(blue) and NO (red) and background corrected N
2
O (black) curves vs potential
(
left ordinates, upper panels, Fig. 7) recorded simultaneously are shown in the upper panels,
Fig. 7. Although relatively large amounts of N O were found for measurements performed in
the entire pH range, the corresponding data in alkaline media yielded, respectively, values of
NO and N much higher and much lower than those found for the other two media, signaling
2
2
marked changes in the overall reaction pathways.
Faradaic Efficiencies for Dinitrogen Generation
int
The pN2 vs E and i vs. E data collected (see blue and thick grey curves in the upper panel, Fig.
lim
7
, respectively), together with the pN2 vs ilim plot in Fig. 3, make it possible to correlate quantitatively
NH OH
2
the fraction of the total current, i, associated with the oxidation of NH
2
OH, iN2
/ i
or faradaic
,
NH OH
2
efficiency for N
2
generation, fN2
, as a function of the applied potential, i.e.
NH OH
NH OH
int
-9
-7
2
2
fN2
=
iN
/ i = [( pN2 + 1.944 x 10 )/(5.692 x 10 )]/i
(3)
2
NH OH
2
where the expression for iN2
is based on the statistical fit of the calibration plot in Fig. 3 (see
OH to N is a two electron
oxidation. Plots of
caption in that figure) corrected by the fact that the oxidation of NH
process, and thus the current would be half that associated with that for N
2
2
2
H
4
NH OH
2
fN2
vs E for E(black) and E (red) based on the data in the upper panels are shown in the lower
E, V vs Ag/AgCl
0.6
E, V vs Ag/AgCl
0.4 0.6
E, V vs Ag/AgCl
0.4
0.8
1.0
0.0
0.2
0.8
1.0
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
6
5
4
3
2
1
1.2
1.0
0.8
0.6
0.4
0.2
0.0
7
6
5
4
3
2
1
0
1
.5
.0
pH 7
pH 2.92
pH 12.03
10
8
N2
3
2
1
0
N2O
N2O
1
NO
6
N2O
4
0.5
0.0
N2
N2
2
NO
NO
0
0
6
5
4
3
2
1
0
6
5
4
3
2
1
0
5
0
0
0
0
0
0
0
0
0
0
0
0
E
4
E
E
3
2
1
E
E
E
0.4
0.5
0.6
0.7
E, V vs Ag/AgCl
0.8
0.9
1.0
0.0
0.2
0.4
0.6
0.8
1.0
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
E, V vs Ag/AgCl
E, V vs Ag/AgCl
int
Figure 7. Upper Panels. Dynamic polarization (right ordinates, thick grey curves) and pX vs E for X
(blue), and NO (red), and corresponding background corrected N O curve (black), recorded
OH in PB solutions at the pH values indicated. Lower
=
N
2
2
simultaneously, for the oxidation of 10 mM NH
2
NH OH
2
Panels. Plots of fN2
vs E for E(black) and E (red) based on the data in the upper panels.
8
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
panels in Fig. 7, where the potential was restricted to a range in which the currents were significant
enough to avoid large uncertainties. Cursory inspection of the data for pH 2.92 indicates that, despite
NH OH
2
the big disparity in the values of i observed for E and E in the range 0.6 < E < 0.9, fN2
was
practically independent of the direction of the scan in the range 0.6 < E < 0.8 V, reaching values as high
as ca. 30% for small overpotentials. This was not the case for pH 7 where marked differences were
NH OH
NH OH
2
2
observed for fN2
in the same overpotential region. Lastly, the values of fN2
in the alkaline
solution were about an order of magnitude smaller than in the other two media at very high and very
low overpotentials.
Some of the on line DEMS data involving gases other than N
agreement with earlier results obtained using other techniques. In particular, the onset of N
coincides with the onset for the oxidation of NH OH as reported earlier in our laboratory using in situ
2
in neutral media are in qualitative
2
O detection
2
5
FTIR, for which the cross sections for absorption are very high. However, at least two factors
complicate a more quantitative analysis of the results presented in this work. Specifically, we have not
been able to identify electrochemical reactions that generate solely either N
as well defined as those involved in the case of N . In addition, NH OH reacts with NO
O as one of the products, and thus contributes to its DEMS signal. Although
2
O or NO under conditions
-
2
2
2
(aq) in
1
5
solution phase to yield N
2
it is, in principle, possible to use numerical simulations to account for this process, as we reported in
5
earlier work for a rotating disk electrode, the wall jet characteristics of our present system would make
such calculations non trivial and, as such, any determination of faradaic efficiencies for gases other
than N would be subject to rather large uncertainties.
2
Conclusions
Additional insights into the mechanism of NH OH on a solid Au(poly) electrode in aqueous phosphate
2
buffered electrolytes were gained from on line mass spectrometry under forced convection in a jet im-
pinging configuration. As evident from the data collected, this process generates N and N O in acidic
and neutral media, and mixtures of NO and N O in alkaline solutions as the main gaseous products
2
2
2
over a rather wide potential region. In acidic media (pH 2.92) no gases were detected for potentials
positive to the onset of oxide formation and negative to the onset of oxide reduction. The amounts of
gas detected during scans toward positive potentials, E, were smaller than those observed during E
for potential excursions into the oxide formation region. This was not the case when E was reversed
prior to oxide formation, for which the signals decreased monotonically upon continuous cycling. Con-
sistent with this phenomenon is the formation of an adsorbed species derived from the oxidation of
NH OH, which blocks the electrode surface from further reaction for E ≤ 1.1 V, and either desorbs
2
and/or oxidizes at higher potentials to yield one or more species that display no affinity for the oxidized
NH OH
2
Au surface. Faradaic efficiencies for dinitrogen generation, fN2
, as high as 30 - 40% were found in
acidic solutions for relatively small overpotentials independent of the direction of the scan. In the case
NH OH
2
of alkaline media, however, the corresponding fN2
were an order of magnitude smaller for much
NH OH
2
higher overpotentials. Lastly, for neutral media, the values of fN2
were found to be smaller for E
compared to those observed during E for most of the potential range examined.
9
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201900367
Literature Cited
1
.
Callisto, M.; Molozzi, J.; Barbosa, J. L. E., Eutrophication of Lakes. In In Eutrophication:
Causes, Consequences and Control, Ansari, A. A.; Gill, S. S., Eds. Springer: Netherlands, 2014; pp 55-
7
2
1.
.
Duca, M.; Koper, M. T. M., Powering denitrification: the perspectives of electrocatalytic nitrate
reduction. Energy & Environmental Science 2012, 5 (12), 9726-9742.
Chen, Y. J.; Zhu, H. F.; Rasmussen, M.; Scherson, D., Rational Design of Electrocatalytic
3
.
Interfaces: The Multielectron Reduction of Nitrate in Aqueous Electrolytes. Journal of Physical
Chemistry Letters 2010, 1 (13), 1907-1911.
4
.
Godoi, D. R. M.; Chen, Y.; Zhu, H.; Scherson, D., Electrochemical Oxidation of
Hydroxylamine on Gold in Aqueous Acidic Electrolytes: An in Situ SERS Investigation. Langmuir
2
5
010, 26 (20), 15711-15713.
. Jebaraj, A. J. J.; Kumsa, D.; Scherson, D. A., Oxidation of Hydroxylamine on Gold Electrodes
in Aqueous Electrolytes: Rotating Ring-Disk and In Situ Infrared Reflection Absorption Spectroscopy
Studies. The Journal of Physical Chemistry C 2012, 116 (12), 6932-6942.
6
.
Duca, M.; Figueiredo, M. C.; Climent, V.; Rodriguez, P.; Feliu, J. M.; Koper, M. T. M.,
Selective Catalytic Reduction at Quasi-Perfect Pt(100) Domains: A Universal Low-Temperature
Pathway from Nitrite to N-2. Journal of the American Chemical Society 2011, 133 (28), 10928-10939.
7
.
Lu, J.; Hua, X.; Long, Y.-T., Recent advances in real-time and in situ analysis of an electrode–
electrolyte interface by mass spectrometry. Analyst 2017, 142 (5), 691-699.
8
2
9
.
Baltruschat, H., Differential electrochemical mass spectrometry. J. Am. Soc. Mass Spectrom.
004, 15 (Copyright (C) 2012 American Chemical Society (ACS). All Rights Reserved.), 1693-1706.
Treufeld, I.; Jebaraj, A. J. J.; Xu, J.; de Godoi, D. M.; Scherson, D., Porous Teflon Ring-Solid
.
Disk Electrode Arrangement for Differential Mass Spectrometry Measurements in the Presence of
Convective Flow Generated by a Jet Impinging Electrode in the Wall-Jet Configuration. Analytical
Chemistry 2012, 84 (12), 5175-5179.
1
0.
Venkatachalam, S.; Angelici, R. J.; Woo, L. K.; Hillier, A. C., High Rate Detection of Volatile
Products Using Differential Electrochemical Mass Spectrometry: Combining an Electrode-Coated
Membrane with Hydrodynamic Flow in a Wall-Tube Configuration. Analytical Chemistry 2013, 85
(
1
12), 6059-6065.
1. Yamada, J.; Matsuda, H., Limiting Diffusion Currents In Hydrodynamic Voltammetry. 3. Wall
Jet Electrodes. Journal of Electroanalytical Chemistry 1973, 44 (2), 189-198.
2. Cisneros, L. O.; Rogers, W. J.; Mannan, M. S.; Li, X. R.; Koseki, H., Effect of iron ion in the
1
thermal decomposition of 50 mass % hydroxylamine/water solutions. Journal of Chemical and
Engineering Data 2003, 48 (5), 1164-1169.
1
3.
Karabinas, P.; Wolter, O.; Heitbaum, J., Mechanistic Studies With Mass Spectroscopic Cyclic
Voltammetry - Anodic-Oxidation Of Hydroxylamine On Pt. Berichte Der Bunsen-Gesellschaft-
Physical Chemistry Chemical Physics 1984, 88 (12), 1191-1196.
1
4.
Jebaraj, A. J. J.; Martins de Godoi, D. R.; Scherson, D. A., Pronounced Surface Sensitivity of
Hydroxylamine Oxidation on Gold Single-Crystal Electrodes in Acidic and Neutral Aqueous Solutions.
ACS Catalysis 2012, 2 (5), 911-915.
1
5.
Doring, C.; Gehlen, H., Uber Die Kinetik Der Reaktion Zwischen Hydroxylamin Und
Salpetriger Saure. Zeitschrift Fur Anorganische Und Allgemeine Chemie 1961, 312 (1-2), 32-44.
1
0
This article is protected by copyright. All rights reserved.