Nitrosation of dialkylhydroxylamines, and computational and NMR investigations of the interconversion of conformational isomers of N-nitroso-dimethylhydroxylamine

Article abstract of DOI:10.1002/poc.1723

The effect of acidity upon the rate of nitrosation of N-benzyl,O- methylhydroxylamine (3) in 1:1 (v/v) H₂O/MeOH at 25°C has been investigated. The pseudo-first-order rate constant (k_obs) for loss of HNO₂ as the limiting re

Full text of DOI:10.1002/poc.1723

Research Article
Received: 1 February 2010,
Revised: 26 March 2010,
Accepted: 31 March 2010,
Published online in Wiley Online Library: 15 June 2010
(wileyonlinelibrary.com) DOI 10.1002/poc.1723
Nitrosation of dialkylhydroxylamines, and
computational and NMR investigations of the
interconversion of conformational isomers of
N-nitroso-dimethylhydroxylamine
a
b
b
b
*
Juan Crugeiras , Howard Maskill , William MacFarlane , Iain D. Menneer ,
Ana Riosa and Miguel A. Riosa
The effect of acidity upon the rate of nitrosation of N-benzyl,O-methylhydroxylamine (3) in 1:1 (v/v) H2O/MeOH at 25 -
C has been investigated. The pseudo-first-order rate constant (kobs) for loss of HNO2 as the limiting reagent decreases
as [H3OR] increases. This is compatible with two parallel reaction channels (Scheme 2). One involves the direct reaction
of the free hydroxylamine with HNO2 (k1 ¼ 1.4 T 102 dm3 molS1 sS1, 25 -C) and the other involves the reaction of the
free hydroxylamine with NOR (k2 ¼ 5.9 T 109 dm3 molS1 sS1). In contrast, there is only a very slight increase in kobs
with increasing [H3OR] for nitrosation of N,O-dimethylhydroxylamine (4) in dilute aqueous solution at 25 -C to give
N-nitroso-dimethylhydroxylamine, 5. This also fits a two-channel mechanism (Scheme 3). Again, one involves the
nitrosation of the free base by NOR (k2 ¼ 8 T 109 dm3 molS1 sS1, 25 -C) but the other channel now involves catalysis by
chloride (k3 ¼ 1.3 T 108 dm3 molS1 sS1). Arising from these results, we propose an estimate of pKa ꢀ S5 for protonated
nitrous acid, (O ¼ N—OHR2 ), which is appreciably different from the literature value of R1.7.
The interconversion of cis and trans conformational isomers of 5 has been investigated by temperature-dependent
NMR spectroscopy in CDCl3, methanol-d4, toluene-d8 and dimethyl sulfoxide-d6. Enthalpies and entropies of reaction
*
and of activation have been determined and compared with computational values obtained at the B3LYP/6-31G level
of theory. The cis form is slightly more stable at normal temperatures and no solvent effects upon the thermodyn-
amics or kinetics of the conformational equilibrium were predicted computationally or detected experimentally. In
addition, key geometric parameters and dipole moments have been calculated for the cis and trans forms, and for the
lowest energy transition structure for their interconversion, in the gas phase and in chloroform. These results indicate
electronic delocalisation in the ground states of 5 which is lost in the transition structure for their interconversion.
Copyright ß 2010 John Wiley & Sons, Ltd.
Keywords: computational chemistry; conformational isomers; dialkylhydroxylamines; nitrosation; nitrosohydroxylamines;
NMR kinetics
(N2O3)[13–17] and the nitrosonium ion (NOþ),[18] all of which are
INTRODUCTION
potential nitrosating agents; in addition, it undergoes dec-
omposition to nitric oxide and nitric acid. There have been
many wide-ranging discussions of nitrosation of amino
compounds (including hydroxylamine) notably by Ridd and
Stedman and their colleagues,[18–22] and the field has been
comprehensively reviewed by Williams.[23] Particularly relevant
to our current investigation is the effect of acidity upon the rate
There are many known nitrosating agents, e.g. nitrosyl com-
pounds (halides, thiocyanate and acetate), nitrogen oxides (nitric
oxide, dinitrogen trioxide and dinitrogen tetroxide), the nitro-
sonium ion (NOþ) and nitroso sulphur compounds, but by far the
most common reagent is the undissociated weak acid, nitrous
acid (HNO2). Nitrosation of organic compounds has been studied
since the earliest days of systematic organic chemistry,[1–8]
¨
the first report being by Behrend and Konig who prepared
N-nitroso-dibenzylhydroxylamine by treating N,O-dibenzylhyd-
roxylamine (1 in Fig. 1) with nitrous acid.[1,2] Catalysis of nitro-
sation by nucleophiles has also been investigated.[9,10]
*
Correspondence to: H. Maskill, School of Chemistry, University of Newcastle,
Newcastle upon Tyne, NE1 7RU, UK.
E-mail: h.maskill@ncl.ac.uk
Nitrous acid (pKa ¼ 3.3) is readily formed in aqueous solution
by adding a nitrite salt (typically sodium nitrite) to an aqueous
solution of a strong acid.[11] Under these conditions, it exchanges
oxygen with the water, a reaction which can be exploited when
oxygen isotope labelling is desired in the nitroso group.[12] Under
acidic conditions, nitrous acid has been reported to be in
equilibrium with its protonated form, (H2NO2)þ,[9,10] its anhydride
a
b
J. Crugeiras, A. Rios, M. A. Rios
Department of Physical Chemistry, Faculty of Chemistry, University of
Santiago de Compostela, Santiago de Compostela, Galicia, Spain
H. Maskill, W. MacFarlane, I. D. Menneer
School of Chemistry, University of Newcastle, Newcastle upon Tyne, NE1 7RU,
UK
J. Phys. Org. Chem. 2011, 24 162–171
Copyright ß 2010 John Wiley & Sons, Ltd.

NITROSATION OF DIALKYLHYDROXYLAMINES
þ
henceforth as [RNH2OR0 ]0. Since [RNH2OR0þ]0 » [NaNO2]0, it will
(like [H3Oþ]0) remain approximately constant throughout the
reaction. However, a high level of methanol in water (1:1, v/v) was
needed to achieve the concentrations of hydroxylammonium
chlorides that were required to ensure they were in large excess
over the initial concentrations of nitrous acid. Similarly, as
[H3Oþ]0 » [NaNO2]0 and pKa ¼ 3.3 for HNO2,[11] the initial con-
centration of nitrous acid, [HNO2]0, was taken to be equal to the
initial concentration of the sodium nitrite, [NaNO2]0 (and it
decreased to zero as the reaction went to completion).
The UV range of 300–400 nm was monitored and the
crown-shaped absorbance attributed to nitrous acid (the limiting
reactant) was seen to decrease by a clean first-order rate law
(even though the overall changes in absorbances were small)
as broad absorbances due to the nitroso products increased. As
seen in Table 1, the range in reactivity for compounds 1–3 is
modest, as anticipated from the earlier work by Stedman.[21]
Compounds 1 and 2 were then shown to be first order in
hydroxylammonium by using different initial excess concen-
trations of RNH2OR0þ Clꢂ in the usual way (results not shown),[27]
and 3 was studied at different initial excess perchloric acid
concentrations (Table 2). Results in Tables 1 and 2 indicate that
precision and reproducibility are satisfactory. Although nitrosa-
tion is first order in both hydroxylammonium and nitrous acid,
Figure 1. Hydroxylamines 1–4 and N-nitroso,dimethylhydroxylamine, 5
of nitrosation of O-alkylhydroxylamines by nitrous acid rep-
orted by Stedman.[19–21]
We recently reported an investigation of the mechanism of the
acid-catalysed decomposition of a series of N-nitroso-dialkylhy-
droxylamines.[24–26] We now report the kinetics and mechanism
of nitrosation of N,O-dialkylhydroxylamines (1–3) in acidic 1:1
(v/v) aqueous methanol and of N,O-dimethylhydroxylamine (4) in
aqueous acid without the methanol cosolvent.
We also report an investigation of the conformational pro-
perties of N-nitroso-dimethylhydroxylamine (5) by NMR spec-
troscopy and theoretical methods; dynamic, thermodynamic and
structural parameters have been determined.
Table 2. Effect of initial hydronium ion concentration on
kinetics of nitrosation of N-benzyl,O-methylhydroxylamine (3)
under pseudo-first-order conditions, 1:1 (v/v) aqueous
methanol, 25.0 8Ca
METHODS AND RESULTS
Kinetics of nitrosation of N,O-dibenzylhydroxylamine (1),
tetrahydro-1,2-oxazine (2) and
N-benzyl,O-methylhydroxylamine (3)
105 Standard
error
[HClO4]0 /mol dmꢂ3
103 kobs/sꢂ1
Of the range of N-nitroso-dialkylhydroxylamines we had pre-
pared for solvolytic studies,[26] the formations of three by
nitrosation of the parent dialkylhydroxylamines (N,O-dibenzylhy-
droxylamine (1), tetrahydro-1,2-oxazine (2) and N-benzyl,O-
methylhydroxylamine (3)) were selected as representative for
mechanistic investigation and comparison with literature results.
All were available as the substituted hydroxylammonium chlori-
des which, given their pKa values (typically ꢁ5),[11,21] remain very
largely in their protonated forms in aqueous acidic solution. In
each case, therefore, the total initial concentration of all hydro-
xylamine species was taken to be equal to the initial con-
centration of the hydroxylammonium chloride and represented
0.10b,c
0.25b
0.50b
1.00d
8.25
6.42
5.59
5.28
7.52
5.06
20.9
4.92
a For all cells, [RNH2OR0þ]0 ¼ 0.075 mol dmꢂ3
.
b [NaNO2]0 ¼ 4.0 ꢃ 10ꢂ3 mol dmꢂ3
.
c Repeat of experiment 3 in Table 1 using different solutions
and spectrophotometer, etc.
d [NaNO2]0 ¼ 3.87 ꢃ 10ꢂ3 mol dmꢂ3
.
Table 1. Kinetics of nitrosation of N,O-dibenzylhydroxylamine (1), tetrahydro-1,2-oxazine (2) and N-benzyl,O-methylhydroxylamine
(3) under pseudo-first-order conditions in 1:1 (v/v) aqueous methanol containing perchloric acid, 25.0 8Ca
104 Standard
error
Wavelengthb
/nm
Hydroxylamine
103 kobs/sꢂ1
N,O-Dibenzylhydroxylamine (1)
Tetrahydro-1,2-oxazine (2)
N-Benzyl,O-methylhydroxylamine (3)
22.5
1.48
8.01
11.1
0.764
2.43
345
322
333
a In all cases, [HClO4]0 ¼ 0.1 mol dmꢂ3; [RNH2OR0þ]0 ¼ 0.075 mol dmꢂ3; [NaNO2]0 ¼ 4.0 ꢃ 10ꢂ3 mol dmꢂ3
.
b In each case, the wavelength at which the decreasing absorbance change was greatest was selected.
J. Phys. Org. Chem. 2011, 24 162–171
Copyright ß 2010 John Wiley & Sons, Ltd.
View this article online at wileyonlinelibrary.com

J. CRUGEIRAS ET AL.
Figure 3. The effect of [4Hþ]0 upon kobs for nitrosation of 4 in aqueous
Figure 2. Plot of pseudo-first-order rate constants against initial hydro-
niumionconcentrationsfornitrosationofN-benzyl-O-methylhydroxylamine
(3) by sodium nitrite in 1:1 (v/v) aqueous methanolic perchloric acid at
solution at [H3Oþ]0 ¼ 0.05 mol dmꢂ3
25 8C, I ¼ 1.0 (NaClO4).
,
[NaNO2]0 ¼ 5 ꢃ 10ꢂ5 mol dmꢂ3
,
25.08C ([NaNO2]0 ¼ 4 ꢃ 10ꢂ3 mol dmꢂ3; [3]0 ¼ 0.075 mol dmꢂ3
)
Table 2 and Fig. 2 indicate that increasing initial hydronium ion
concentrations lead to lower pseudo-first-order rate constants for
3, so a complex rate law is implicated.
higher values of [NaNO2]0, especially at the higher acidity. This
effect had been noted earlier by Stedman for nitrosation of the
parent hydroxylamine and attributed to competing nitrosation
by N2O3.[22] Consequently, our further experiments were all
carried out at [NaNO2]0 ¼ 5 ꢃ 10ꢂ5 mol dmꢂ3
.
Kinetics of nitrosation of N,O-dimethylhydroxylamine (4) in
aqueous acid
Figure 3 shows the effect of [4Hþ]0 (the initial concentration of
the hydroxylammonium ion) upon kobs at [H3Oþ]0 ¼ 0.05 mol dmꢂ3
;
Our results for the effect of acidity upon the rate of nitrosation of
3 in aqueous methanol (Table 2 and Fig. 2) were not as anti-
cipated from results reported earlier by Stedman who detected
minimal dependence of kobs upon [H3Oþ]0 for a series of O-alkyl-
hydroxylamines in water.[21] Consequently, we reinvestigated
the nitrosation of the simpler N,O-dialkylhydroxylamine (4) in
aqueous acid without methanol as a cosolvent. The reaction was
monitored by following the increase in absorbance due to the
formation of the N-nitroso product (5), and results for different
starting concentrations of sodium nitrite in the range of (0.05–
5.0) ꢃ 10ꢂ3 mol dmꢂ3 are shown in Table 3. As seen, the value of
kobs increases slightly with increasing [NaNO2]0; however, the
absorbance–time data deviated from first-order behaviour at
the results appear to be in accord with a rate law comprising first-
and second-order terms in [4Hþ]0, i.e. rate ¼ kobs [HNO2] with kobs
given by Eqn (1), and a plot of kobs/[4Hþ]0 versus [4Hþ]0 was
indeed linear with a ¼ (2.75 ꢄ 0.06) ꢃ 10ꢂ2 dm3 molꢂ1 sꢂ1 and
b ¼ (2.80 ꢄ 0.02) dm6 molꢂ2 sꢂ1 (Fig. 4).
2
kobs ¼ a½4Hþꢅ0 þ b½4Hþꢅ0
(1)
However, since the initial concentration þof the hydþroxylam-
monium is equal to that of chloride, b [4H ]20 ¼ b [4H ]0 [Clꢂ],
the second term in Eqn (1) could be due to nucleophile catalysis
by Clꢂ as previously reported by Stedman for halides and SCNꢂ in
Table 3. Dependence of kobs for nitrosation of 4 on [NaNO2]0
in aqueous acidic solution at 25 8C, I ¼ 1.0 mol dmꢂ3 (NaClO4)a
[H3Oþ]0 ¼ 0.10 mol dmꢂ3
[H3Oþ]0 ¼ 0.010 mol dmꢂ3
103[NaNO2]0/
mol dmꢂ3
103[NaNO2]0/
mol dmꢂ3
103 kobs/sꢂ1
103 kobs/sꢂ1
0.05b
0.5c
5.0d
8.58
9.35
14.0
0.05b
0.5c
5.0d
8.35
12.7
24.0
a [MeNH2OMeþ Clꢂ]0 ¼ 5 ꢃ 10ꢂ2 mol dmꢂ3
.
b Monitored at 240 nm.
c Monitored at 250 nm.
d Monitored at 330 nm.
Figure 4. Plot of kobs/[4Hþ]0 versus [4Hþ]0 for nitrosation of 4 in aqueous
solution at [H3Oþ]0 ¼ 0.05 mol dmꢂ3
25 8C, I ¼ 1.0 (NaClO4)
,
[NaNO2]0 ¼ 5 ꢃ 10ꢂ5 mol dmꢂ3
,
View this article online at wileyonlinelibrary.com
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 162–171

NITROSATION OF DIALKYLHYDROXYLAMINES
compounds,[21] and quite different from what we had found for 3
in aqueous methanol.
Determination of thermodynamic and kinetic parameters
for the conformational isomerisation of
N-nitroso-dimethylhydroxylamine (5) by NMR spectroscopy
and related computational methods
N-Nitrosamines and N-nitroso-hydroxylamines have been inves-
1
tigated in the past by H and 15N NMR spectroscopy,[28,29] and
conformational isomerism of several N-nitroso-dialkylhydroxy-
lamines was reported by Axenrod et al.[30] We have carried out
new 1H measurements on 5 at 500 MHz and ꢂ50 8C, and isotropic
NMR parameters were calculated using B3LYP/6-311 þ þG(d,p) in
Gaussian03 with the Gauge-Invariant Atomic Orbital (GIAO)
scheme and geometries optimised at the B3LYP/6-311 þ þG(d,p)
level.[31,32] Identical calculations were also carried out on
tetramethylsilane and the results used as references to calculate
proton and carbon chemical shifts. As seen in Table 4, the com-
puted and new experimental 1H chemical shifts for cis- and
trans-5 are in good agreement with each other and with earlier
experimental values. We have also measured 13C NMR spectra of
5 at 125.77 MHz and 23 8C in CDCl3 (assignments were confirmed
by two-dimensional HSQC experiments) for comparison with our
computed results in the gas (Table 4).
In order to determine the thermodynamic and kinetic para-
meters for the conformational isomerisation of compound 5,
Scheme 1, 1H NMR spectra were recorded at 500.16 MHz at
different temperatures in deuteriochloroform (263–328 K), meth-
anol-d4 (263–328 K), toluene-d8 (238–373 K) and dimethyl
sulfoxide-d6 (295–336 K; this relatively high lower temperature
is enforced by the melting point of DMSO). At the lowest
temperatures in each solvent, we observe two sharp singlets
(>NCH3 and —OCH3) for each of the two conformational isomers
Figure 5. The effect of increasing concentrations of NaCl upon kobs for
the nitrosation of 4 at [H3Oþ]0 ¼ 0.05 mol dmꢂ3, [4Hþ]0 ¼ 0.01 mol dmꢂ3
,
[NaNO2]0 ¼ 5 ꢃ 10ꢂ5 mol dmꢂ3, 25 8C, I ¼ 1.0 (NaClO4)
the nitrosation of several hydroxylamines by nitrous acid.[21]
To test this, Clꢂ was replaced by perchlorate using Amberlyte
IRA- 410 anionic resin. Although 100% exchange was not
achieved, the value of b in Eqn (1) was much reduced whereas a
was unchanged. In addition, Fig. 5 shows the effect of increasing
concentrations of NaCl upon kobs for nitrosation of 4 at [H3Oþ]0 ¼
0.05moldmꢂ3 and [4Hþ]0 ¼ 0.01 moldmꢂ3. The intercept divided
by [4Hþ]0 gave the result a ¼ 4.5 (ꢄ2.1) ꢃ 10ꢂ2 dm3 molꢂ1 sꢂ1
which, although of low precision, compares favourably with
the result given above, (2.75 ꢄ 0.02) ꢃ 10ꢂ2 dm3 molꢂ1 sꢂ1. The
slope divided by [4Hþ]0 gave the result b ¼ 2.56 (ꢄ0.07) dm6
molꢂ2 sꢂ1 in fair agreement with the value given above, 2.80
(ꢄ0.02) dm6 molꢂ2 sꢂ1
.
Finally in this context, Fig. 6 shows the effect of [H3Oþ]0
upon kobs for nitrosation of 4 at [4HþClꢂ]0 ¼ 0.05 mol dmꢂ3 in
water; clearly, the effect is only small up to 0.9 mol dmꢂ3 in
agreement with previous observations by Stedman for related
1
Table 4. Experimental and computed H and 13C NMR
chemical shifts (ppm relative to TMS) for cis- and trans-
a
N-nitroso-dimethylhydroxylamines (5) in CDCl3
cis-5
—OCH3
trans-5
>NCH3
>NCH3
—OCH3
dH observedb
dH observedc
dH calculatedd
dC observede
dC calculatedd
3.91
3.80
3.77
38.60
39.5
3.87
3.75
3.75
61.76
63.3
3.46
3.36
3.13
33.80
30.2
4.13
4.02
4.05
64.43
65.6
a The cis and trans short-hand descriptors here indicate the
relationship between the oxygens bonded to the respective N
atoms in the approximately planar forms (refer Schemes 1 and
4 and text); earlier authors have used cis and trans to describe
the relationships between the >N—CH3 and the O of the
nitroso.[30]
b
ꢂ30 8C, 60 MHz.[30]
c Present work, ꢂ50 8C, 500 MHz.
Figure 6. The effect of [H3Oþ]0 upon kobs for nitrosation of 4 in water at
d Present work, gas phase.
[4Hþ]0 ¼ 0.05 mol dmꢂ3
,
[NaNO2]0 ¼ 5 ꢃ 10ꢂ5 mol dmꢂ3
25 8C, I ¼ 1.0
,
e Present work, 23 8C, 125.77 MHz.
(NaClO4)
J. Phys. Org. Chem. 2011, 24 162–171
Copyright ß 2010 John Wiley & Sons, Ltd.
View this article online at wileyonlinelibrary.com

J. CRUGEIRAS ET AL.
exchange rate constants, are shown in Fig. 7. The 13C NMR spectra
at 125 MHz showed similar line broadening and coalescence
1
behaviour fully consistent with the results from the H spectra,
although the signal-to-noise ratios attainable in the times
available were insufficient to achieve comparable precision in
spectral simulations.
Scheme 1. Conformational isomerisation of N-nitroso-dimethylhydroxy-
lamine (5) (refer footnote in Table 4 regarding cis and trans nomen-
clature)
[a]
Each exchange rate constant corresponds to a rate of con-
formational equilibration and is the sum of the rate constants for
forward and reverse processes (kex ¼ kf þ kr) shown in Sch-
eme 1.[27] From these and the ratios of the rate constants from the
equilibrium constants, kf and kr can be separately determined at
each temperature. The enthalpies and entropies of activation for
kf and kr processes in the various solvents were then obtained
from plots using the Eyring equation: ln(k/T) ¼ 23.73 ꢂ (DHz/
8.31 T) þ (DSz/8.31); in all cases, the least squares fits to straight
lines had R2 > 0.994 and results for the kf process are inclzuded in
Table 5. Results for DHo (¼ DHz ꢂ DHrz) and DSo (¼ DS ꢂ DSzr )
of CH3N(NO)OCH3 (refer Fig. 7 for the spectrum in CDCl3 at 253 K).
Integration of the signals gives the relative proportions of the
conformational isomers and hence the equilibrium constant at
the measured temperature, which is the ratio of forward and
reverse rate constants for the conformational equilibration
(K ¼ kf/kr).
As the temperature is increased, the signals broaden as they
begin to coalesce due to interconversion of the conformational
isomers becoming faster compared with the time scale of the
NMR process; spectra in all solvents at the highest temperatures
comprise only two broad signals (one for —O—CH3 and one for
>N—CH3). At each temperature, a first-order exchange rate
constant, kex, was determined by line–shape analysis and sim-
f
f
were the same as results obtained by the more usual van’t Hoff
method from values of K at different temperatures shown (with
experimental values for DGo298) in Table 6.
Computational structural investigation on
N-nitroso-dimethylhydroxylamine (5)
1
ulation of the spectrum. The simulations of the H NMR spectra
were performed using Program gNMR,[33] allowance being made
for the variation of the proportions of cis- and trans-5 with
temperature, and also for the temperature dependence of the
various chemical shifts which was particularly marked in the case
of solutions in toluene. These latter effects can be attributed to
reversible association of the solvent and solute and the large
anisotropic diamagnetic susceptibility of toluene. In CDCl3 and
methanol, the chemical shifts of the >N—Me and —O—Me
resonances of the more abundant cis-5 were quite close, and their
assignments were confirmed by two-dimensional EXSY exper-
iments at 5 8C which related them to the corresponding widely
separated resonances in the less abundant trans isomer.
Representative experimental and simulated spectra in deuterio-
chloroform at specified temperatures, along with calculated
Structures were optimised using the B3LYP method with a
6-31G(d) basis set using the Gaussian03 program.[31,32] For
geometry optimisations in chloroform, the Polarised Continuum
Model (PCM) approximation was used (TABS ¼ 300.0, ALPHA ¼
1.21, TSNUM ¼ 70, RADII ¼ PAULING). Vibrational frequency cal-
culations were performed on all stationary points at the B3LYP
level to characterise them as minima (all real frequencies) or as
transition structures (one imaginary frequency). Stable cis and
trans conformational isomers, and two transition structures (TS1
and TS2) for their interconversion, were identified (Fig. 8). TS2
was found to be less stable than TS1, which was confirmed by
results of higher-level calculations (QCISD/6-311 þ G(d,p) and
B3LYP/6-311 þ þG(d,p)), and is not considered further. Key
Figure 7. Comparison of experimental and simulated NMR spectra for N-nitroso-dimethylhydroxylamine (5) in CDCl3 (temperatures are given below
experimental spectra and corresponding calculated exchange rate constants are given below the simulated spectra)
View this article online at wileyonlinelibrary.com
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 162–171

NITROSATION OF DIALKYLHYDROXYLAMINES
Table 5. Computed and NMR experimental activation parameters for the conformational isomerisation cis-5 ! trans-5
B3LYP
(gas)
B3LYP
(CHCl3)
NMR
(CDCl3)
NMR
(toluene)
NMR
(DMSO-d6)a
NMR
(CD3OD)
DHz/kJ molꢂ1
62.4
ꢂ9
68.2
ꢂ13
65.9
5
63.6
2
66.5
ꢂ5
64.1
ꢂ4
DSz/J Kꢂ1 molꢂ1
a Limited temperature range.
Table 6. Computed and NMR experimental thermodynamic parameters for the conformational isomerisation cis-5 $; trans-5
B3LYP
(gas)
B3LYP
(CHCl3)
NMR
(CDCl3)
NMR
(toluene)
NMR
(DMSO-d6)a
NMR
(CD3OD)
DHo/kJ molꢂ1
3.9
8.1
1.5 (0.55)
4.3
8.3
1.8 (0.48)
4.1
2.8
3.3 (0.26)
4.0
2.4
3.3 (0.26)
5.7
7.6
3.4 (0.25)
5.1
5.4
3.5 (0.24)
DSo/J Kꢂ1 molꢂ1
DGo298/kJ molꢂ1 (equil. constant)
a Limited temperature range.
geometric parameters (B3LYP/6-31G(d)) and dipole moments for
cis-5, trans-5 and TS1 in the gas phase are given in Table 7.
Zero-point vibrational energies, entropies and thermal corr-
ections to enthalpy and Gibbs free-energy values were obtained
from the unscaled harmonic frequencies and moments of
inertia by standard methods.[34] These led to the computed
thermodynamic parameters for the conformational isomerisation
cis-5 $; trans-5 included in Table 6 (along with NMR experimen-
tal values), and computed activation parameters are compared
with NMR experimental results in Table 5.
DISCUSSION
Mechanism of nitrosation of N,O-dibenzylhydroxylamine (1),
tetrahydro-1,2-oxazine (2) and
N-benzyl,O-methylhydroxylamine (3) in aqueous methanol
In aqueous acidic solution, nitrite anions will be essentially fully
protonated (pKa ¼ 3.3 for HNO2)[11] and it has been previously
assumed that the HNO2 so formed will be further protonated on
the hydroxyl but to only a small degree, Eqn (2).[9,10] On this
assumption, a kinetic scheme involving ONOHþ2 as a reactive
intermediate nitrosating agent may be derived (refer later).
However, the pKa of ONOHþ2 has recently been reported to be
1.7,[35,36] in which case its concentration would be appreciable in
aqueous acidic solutions of nitrites (even though it had eluded
previous attempts at detection), and the steady-state approxi-
mation may not be applied to it. This pKa value requires that
replacement of a hydrogen in H2O by a nitroso group is both
strongly acid-enhancing (pKa ¼ 15.7 for H2O and 3.3 for HONO)
and appreciably base-enhancing (pKa ¼ ꢂ1.74 for H3Oþ and þ1.7
for ONOHþ2 ). In contrast, our recent computational studies
Table 7. Key geometric parameters (B3LYP/6-31G(d)) and
dipole moments for cis-5, trans-5, and the transition structure
TS1 for their interconversion in the gas phasea
Parameter
cis-5
TS1
trans-5
˚
d(N—N)/A
—
1.351
1.220
1.571
1.183
1.347
1.225
165.8
3.6
˚
d(N O)/A
—
F(O—N—N—O)/o
ꢂ15.3
ꢂ119.2
m/D
3.4
350.3
1.3
306.7
P
u/o at N1
351.8
Figure 8. Gas phase optimised structures (B3LYP/6-31G(d)) for the cis
and trans conformational isomers of CH3N(NO)OCH3 (5) and two tran-
sition structures (TS1 and TS2) for their interconversion
a Calculated results in chloroform are only slightly different.
J. Phys. Org. Chem. 2011, 24 162–171
Copyright ß 2010 John Wiley & Sons, Ltd.
View this article online at wileyonlinelibrary.com

J. CRUGEIRAS ET AL.
indicate that ONOHþ2 is not only a stronger acid than H3Oþ, but
that it is also very unstable with respect to dissociation into NOþ
and water, Eqn (3).[37] Consequently, we combine Eqns (2) and (3)
[18]
into Eqn (4) and, following Reference
propose NOþ as the
nitrosating agent additional to nitrous acid in dilute aqueous
solutions of nitrite salts (refer later).
þ
HONO þ H3Oþ$;H2 O NO þ H2O
(2)
þ
H2 O NO$;NOþ þ H2O
(3)
(4)
HONO þ H3Oþ$;NOþ þ 2H2O; KNO
Hydroxylammonium ions (pKa ꢁ 5) are the conjugate acids of
moderately strong bases,[11,21] so will be deprotonated in aqu-
eous acids to only a very small degree; however, the free base will
be the most reactive nucleophile present even though its con-
centration is very low. We propose, therefore, the mechanism
shown in Scheme 2 with two parallel reaction channels. A third
possible channel involving catalysis by chloride to give ClNO can
be ruled out by the finþding that there is no kinetic term
second-order in [RNH2OR0 Clꢂ]0 for the nitrosation of 3 as was
observed for the nitrosation of 4 (refer later). The rate law derived
from the mechanism in Scheme 2 for compounds 1–3 in aqueous
methanol is included in Scheme 2.
Figure 9. Plot of pseudo-first-order rate constants against the reciprocal
of the initial hydronium ion concentrations for nitrosation of N-benzyl-O-
methylhydroxyꢂla3mine (3) in 1:1 (v/v) aqueous methanol at 25.0 8C. ([Na
NO2]0 ¼ 4 ꢃ 10 mol dmꢂ3; [RNH2OR0þ]0 ¼ 0.075 mol dmꢂ3; gradient ¼
3.30 ꢃ 10ꢂ4 mol dmꢂ3 sꢂ1 (standard error ¼ 1.4 ꢃ 10ꢂ5); intercept ¼ 4.98 ꢃ
10ꢂ3 sꢂ1 (standard error ¼ 7.6 ꢃ 10ꢂ5); correlation coefficient > 0.99)
The experimental rate law for the loss of nitrous acid is
pseudo-first order, i.e. rate ¼ kobs [HNO2], so, if the mechanism of
Scheme 2 is correct and assuming [HNO2] ꢁ [HNO2]T, kobs is
related to the parameters in Scheme 2 and Eqn (4) by Eqn (5).
Equation (6) predicts a linear relationship between kobs and the
reciprocal of [H3Oþ] with a finite intercept. The plot is shown in
Fig. 9, which indicates that the experimental results obtained for
3 are indeed compatible with the mechanism in Scheme 2.
Moreover, the two second-order rate constants k1 and k2 may be
determined from the gradient and intercept of the plot,
½RNHOR0ꢅ ꢆ Ka1 ꢆ ðk1 þ k2 ꢆ KNO ꢆ ½H3OþꢅÞ
ðKa1 þ 0½H3OþꢅÞ ꢆ ð1 þ KNO ꢆ ½H3OþꢅÞ
kobs
¼
(5)
respectively, with values for Ka1 and KNO
.
As ½H3Oþꢅ ꢇ Ka1 and KNO ꢆ ½H3Oþꢅ ꢈ 1 ðsee belowÞ;
Using [RNHOR0]0 ¼ 0.075 mol dmꢂ3, Ka1 ¼ 3.2 ꢃ 10ꢂ5 mol dmꢂ3
ꢀ
ꢁ
and KNO ¼ 3.5 ꢃ 10ꢂ7 dm3 molꢂ1 [23]
,
we obtain k1 ¼ 1.4 ꢃ
[34]
k1
kobs ¼ ½RNHOR0ꢅ0 ꢆ Ka1
þ k2 ꢆ KNO
(6)
102 dm3 molꢂ1 sꢂ1 and k2 ¼ 5.9 ꢃ 109 dm3 molꢂ1 sꢂ1. We conclu-
½H3OþꢅÞ
Scheme 2. Proposed mechanism and derived rate law for nitrosation of 1–3 by nitrous acid and nitrosonium ion in aqueous methanol
View this article online at wileyonlinelibrary.com
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 162–171

NITROSATION OF DIALKYLHYDROXYLAMINES
de that reaction between dialkylhydroxylamines and nitrosonium
ion is diffusion controlled, and the lower overall reactivity of 3 at
higher values of [H3Oþ] is because the increasing concentrations
of NOþ do not compensate for the decreasing concentrations of
the free hydroxylamine.
at 25 8C which compares well with the result of 1.5 ꢃ 107 dm3
molꢂ1 sꢂ1 at 0 8C reported by Stedman.[21] From the intercept in
Fig. 5 (5 ꢃ 10ꢂ4 sꢂ1) and the values for KNO and Ka1 given
above,[23,38,39] we obtain k2 ¼ 8 ꢃ 109 dm3 molꢂ1 sꢂ1, i.e. diffusion
control, in excellent agreement with the result obtained for 3
(refer above). The mechanism of Scheme 3, therefore, accounts
for the experimental results for nitrosation of compound 4 under
the experimental conditions we used.
Mechanism of nitrosation of N,O-dimethylhydroxylamine (4)
in aqueous acid
We saw above in the nitrosation of 3Hþ Clꢂ in aqueous methanol
that kobs decreased with increasing [H3Oþ] (Fig. 2) and that this is
due to the k1 term in Eqn (6) corresponding to the reaction of the
decreasing concentrations of the free base with HNO2 (the k1
route in Scheme 2). This was not observed in the reaction of 4Hþ
Clꢂ in water; in fact kobs increased slightly with increasing [H3Oþ]
(Fig. 6), the anticipated result on the basis of earlier findings by
Stedman for related compounds. Consequently, we rule out any
appreciable reaction between the free base 4 and HNO2 under
these reaction conditions. However, like Stedman, we observed
catalysis by chloride so propose the mechanism in Scheme 3 for 4
with two principal reaction channels. If the mechanism of
Scheme 3 is correct, kobs of the experimental rate law (rate ¼ kobs
[HNO2]) is related to the parameters in Scheme 3 by Eqn (7).
Comparison of the nitrosation of compounds 1–3 and 4
Nitrosation of compounds 1–3 was carried out in 1:1 (v/v)
aqueous methanol; compound 4 was nitrosated in aqueous
solution without the methanol; all substrates were used as their
þ
hydrochloride salts (RNH2OR0 Clꢂ). For 3 in aqueous methanol,
we have shown that a mechanism involving HNO2 and NOþ as
the effective nitrosating agents satisfactorily accommodates the
experimental findings, and HNO2 is more than 107 times less
reactive than NOþ which reacts under diffusion control. Signi-
ficantly, there was no evidence of catalysis by the counter-ion of
the substrate, Clꢂ. In contrast, a second-order term in [RNH2OR0
þ
Clꢂ] implicates Clꢂ as a nucleophile catalyst for nitrosation of 4 in
water. Under these conditions, both ClNO (formed in the catalytic
route) and NOþ react with 4 close to or at the diffusion limit, and
the much less reactive HNO2 cannot compete as a nitrosating
agent. Apart from the relatively minor differences in structure
between 3 and 4, the only obvious difference between the two
reactions is the presence of methanol as a cosolvent for 3.
Consequently, we have to conclude that the difference in solvents
affects the relative magnitudes of the various equilibrium and
rate constants sufficiently to determine the subtle balance bet-
ween the competing nitrosation routes for 1–3 on the one hand
and 4 on the other. There are insufficient data, however, either in
the literature or from our own work, to allow extrapolation and
accurate prediction of likely results in aqueous methanol
solutions of other compositions.
ðk2 ꢆ KNO þ k3 ꢆ KClNO:½ClꢂꢅÞ
kobs ¼ ½RNHOR0ꢅ0 ꢆ Ka1 ꢆ ½H3Oþꢅ ꢆ
(7)
(8)
ðKa1 þ ½H3OþꢅÞ
kobs ¼ ½RNHOR0ꢅ0 ꢆ Ka1 ꢆ ðk2 ꢆ KNO þ k3 ꢆ KClNO:½ClꢂꢅÞ
This complex rate law may be approximated to Eqn (8) since
K
a1 << [H3Oþ] (refer below). Equation (8) predicts a linear plot of
kobs versus total chloride concentration, [Clꢂ]T, as is observed in
Fig. 5. From the gradient of the plot (2.56 ꢃ 10ꢂ2 dm3 molꢂ1 sꢂ1),
[4Hþ]0 ¼ 0.01 mol dmꢂ3, and literature values for Ka1 (1.78 ꢃ
10ꢂ5 mol dmꢂ3),[38,39] KNO (3.5 ꢃ 10ꢂ7 dm3 molꢂ1 [23]
)
and KClNO
(1.1 ꢃ 10ꢂ3 dm6 molꢂ2),[23] we calculate k3 ¼ 1.3 ꢃ 108 dm3 molꢂ1 sꢂ1
Scheme 3. Proposed mechanism and derived rate law for nitrosation of 4 in aqueous perchloric acid containing sodium nitrite with catalysis by chloride
J. Phys. Org. Chem. 2011, 24 162–171
Copyright ß 2010 John Wiley & Sons, Ltd.
View this article online at wileyonlinelibrary.com

J. CRUGEIRAS ET AL.
Base strength of nitrous acid
In Scheme 2 and the associated text, we invoked NOþ as the
reactive nitrosating agent (in addition to HNO2) rather than its
precursor, ONOHþ2 , Eqns (2)–(4). This was based upon a com-
putational investigation of the protonation of HONO.[37] In the
event that ONOHþ2 is trapped by RNHOR0 faster than it fragments
to NOþ and H2O, Eqn (9), and ONOHþ2 rather than NOþ is the
reactive nitrosating agent, a mechanism analogous to that in
Scheme 2 can be readily devised. With the usual approximations,
the experimental rate constant is then given by Eqn (10), rather
than Eqn (6), where Ka2 is the equilibrium constant for the reverse
of Eqn (2), i.e. the acid dissociation constant of protonated nitrous
acid. If we assume that ONOHþ2 is comparable with ClNO as a
nitrosating agent,[41] i.e. k02 ꢀ 108 dm3 molꢂ1 sꢂ1, we calculate
pKa2 ꢀ ꢂ5. Whilst this is an approximate result, we suggest it is
more credible than the value of 1.7 recently reported.[35,36]
Scheme 4. Electronic structures and interconversion of conformational
[a]
isomers of N-nitroso-dimethylhydroxylamine (5) (refer footnote
Table 4 regarding cis and trans nomenclature)
in
resonance structures. We note that the degree of planarity at the
central N and the dihedral angles in cis and trans forms will reflect
the contribution of the alternative resonance canonicals, i.e. the
degree of delocalisation. We also note that internal rotation about
the central N—N bond which interconverts cis and trans forms of
5 reduces the degree of delocalisation, which will be close to zero
in the transition structures when the dihedral angle is about 908.
The N—N bonds in the two stable conformers are very similar
in length (Table 7), and shorter than a normal N—N bond,[42] i.e.
as expected for partial double bond character indicated in
Scheme 4. In contrast, the N—N bond in TS1 is appreciably
longer and corresponds to no discernable resonance interaction
in the transition structure. Correspondingly, the nitroso N—O
bond in TS1 is normal for a double bond between O and N
whereas the very similar N—O bonds in cis-5 and trans-5 are
longer than normal indicating partial single bond character.[42] In
accordance with these findings, the sums of angles at N1 shown
in Table 7 indicate (i) a high degree of pyramidality in TS1 but
only little in cis-5 and trans-5, (ii) dihedral angles in cis-5 and
trans-5 only about 158 out of planarity and (iii) appreciably
greater dipolarity in cis-5 and trans-5 than in TS1.
ꢀ
ꢁ
k1
½H3Oþꢅ
k20
kobs ¼ ½RNHOR0ꢅ0 ꢆ Ka1
þ
(10)
Ka2
Conformational isomerisation of
N-nitroso-dimethylhydroxylamine (5) by
1
temperature-dependent H NMR spectroscopy and
computational methods
Computed and experimental thermodynamic parameters for
cis-5 $; trans-5 are given in Table 6. Agreement amongst the
experimental results in the several solvents and the computed
results is within the probable experimental error of the NMR
results. The computed results indicate no major effect of sol-
vation by chloroform upon the equilibrium and none is found
experimentally in non-polar, H-bonding, and dipolar aprotic
solvents employed. Equilibrium favours the cis isomer but there is
still about 20% of the less stable trans form present at 25 8C.
Resolution of the free energy of reaction into enthalpy and
entropy terms indicates no huge compensating effects or diff-
erences amongst the solvents: the entropy change for the cis to
trans conversion is positive but very small so compensates to only
a small degree for the small positive enthalpy change associated
with the conversion.
The agreement between the experimental and simulated NMR
spectra shown in Fig. 7 is satisfactory indicating that the derived
rate constants are reliable. Computed and experimental acti-
vation parameters for the cis-5 ! trans-5 conversion are comp-
ared in Table 5. As with the enthalpies and entropies of reaction,
agreement amongst the experimental results in the several
solvents, and with the computed values, is within the probable
experimental error of the NMR results. Again, we observe no
computational prediction of a solvent effect upon the confor-
mational interconversion and none is found. This is perhaps
surprising in view of the computed greater dipolarity of the
ground state forms compared with the transition structure for
their interconversion (refer above and Table 7).
EXPERIMENTAL
Representative kinetics measurement of the nitrosation of
N-benzyl-O-methylhydroxylamine (3)
Aqueous perchloric acid (5.0 cm3, 0.4 mol dmꢂ3) was mixed with
methanolic N-benzyl-O-methylhydroxylammonium chloride (5.0 cm3,
0.30 mol dmꢂ3) to give solution A (10 cm3) which is 0.2 mol dmꢂ3
in perchloric acid and 0.15 mol dmꢂ3 in N-benzyl-O-methylhy-
droxylammonium chloride in 1:1 aqueous methanol.
Sodium nitrite (13.8 mg) was dissolved in 1:1 methanol/water
(25 cm3) to give solution B which is 8.0 ꢃ 10ꢂ3 mol dmꢂ3 in Na
NO2. Typically, 1 cm3 of each of the solutions A and B was injected
into a 1 cm pathlength quartz UV cell whereupon the data acq-
uisition was started at a wavelength in the range of 320–330 nm
(corresponding to the wavelength of greatest change in absorb-
ance as the reaction proceeds) in the usual manner.[27] (The initial
concentrations in the 1:1 methanol/water in the cell were [HCl
O4]¼ 0.1moldmꢂ3; [hydroxylammonium salt]0 ¼ 0.075 moldmꢂ3
and [NaNO2]0 ¼ 4.0 ꢃ 10ꢂ3 mol dmꢂ3).
The absorbance–time data were fitted to the exponential form
of the first-order rate law equation to determine the pseudo first
order rate constant.
NMR spectroscopic measurements of the conformational
equilibration of N-Nitroso-N,O-dimethylhydroxylamine (5)
Structure of N-nitroso-dimethylhydroxylamine (5)
Pulsed FT 1H (500.16 MHz) NMR spectra were obtained from
samples containing ca. 5% of 5 in 5 mm O.D. tubes on a JEOL
Lambda spectrometer using 708 flip angles and a digital
The interconversion of cis and trans conformational isomers of
compound 5 is shown in Scheme 4 along with their alternative
View this article online at wileyonlinelibrary.com
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 162–171

NITROSATION OF DIALKYLHYDROXYLAMINES
resolution in the transformed spectra of 0.15 Hz per point.
Temperature calibration was done using the 1H spectra of
samples of neat MeOH and 1,2-ethanediol. The number of
transients recorded depended upon the amount of line broad-
ening at the different temperatures and was adjusted to give a
satisfactory signal-to-noise ratio in each case. Any exponential
window applied to the FID prior to transformation was adjusted
to ensure the resulting additional line broadening was insignif-
icant in each case. 13C spectra were measured at 125.77 MHz, and
the EXSY and HSQC experiments used standard pulse sequences.
[21] T. B. D. Morgan, G. Stedman, M. N. Hughes, J. Chem. Soc. (B) 1968, 344.
[22] M. N. Hughes, G. Stedman, J. Chem. Soc. (B) 1963, 2824.
[23] D. L. H. Williams, Nitrosation Reactions and the Chemistry of Nitric
Oxide, Elsevier, Amsterdam, 2004.
[24] H. Maskill, I. D. Menneer, D. I. Smith, J. Chem. Soc., Chem. Commun.
1995, 1855.
[25] J. Haider, M. N. S. Hill, I. D. Menneer, H. Maskill, J. G. Smith, J. Chem.
Soc., Chem. Commun. 1997, 1571.
[26] J. I. Bhat, W. Clegg, M. R. J. Elsegood, H. Maskill, I. D. Menneer, P. C.
Miatt, J. Chem. Soc., Perkin Trans. 2 2000, 1435.
[27] H. Maskill, The physical basis of organic chemistry, Oxford University
Press, Oxford, 1985.
[28] G. J. Karabatsos, R. A. Taller, J. Amer. Chem. Soc. 1964, 86, 4373.
[29] T. Axenrod, P. S. Pregosin, G. W. A. Milne, J. Chem. Soc., Chem.
Commun. 1968, 702.
Acknowledgements
[30] T. Axenrod, M. J. Wieder, G. W. A. Milne, Tetrahedron 1969, 10, 401.
[31] J. B. Foresman, A. Frisch, Exploring Chemistry with Electronic Structure
Methods, (2nd edn). Gaussian Inc., Pittsburgh, PA, USA, 1996.
[32] Gaussian 03, Revision B.05. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G.
E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T.
Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M.
Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E.
Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P.
Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G.
Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G.
Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Pittsburgh PA, (2003).
[33] P. H. M. Budzelaar, Program gNMR. Cherwell Scientific Publishing,
Oxford, 1995.
The authors thank the Centro de Supercomputacion de Galicia
(CESGA) for computing facilities, and the EPSRC (UK) for partial
funding of NMR facilities at Newcastle.
REFERENCES
¨
[1] R. Behrend, E. Konig, Justus Liebigs Annalen de Chemie. 1891, 263, 218.
[2] R. Behrend, R. Lindner, Justus Liebigs Annalen de Chemie. 1893, 275,
136.
[3] V. Meyer, Justus Liebigs Annalen de Chemie. 1875, 175, 141.
[4] W. Wislicenus, Berichte der Deutschen Chemischen Gesellschaft. 1893,
26, 772.
[5] J. E. Leffler, Chem. Rev. 1950, 45, 385.
[6] J. E. Leffler, A. A. Bothner-By, J. Am. Chem. Soc. 1951, 73, 5473.
[7] A. Bothner-By, L. Friedman, J. Chem. Phys. 1952, 20, 459.
[8] A. B. Boese, Jr., L. W. Jones, R. T. Major, J. Am. Chem. Soc. 1931, 53,
3530.
[9] M. Hutchinson, G. Stedman, J. C. S. Perkin 2 1973, 93.
[10] T. A. Meyer, D. L. H. Williams, J. C. S. Perkin 2 1981, 361.
[11] D. D. Perrin, Dissociation Constants of Inorganic Acids and Bases in
Aqueous Solution, Butterworths, London, 1969. also published in Pure
Appl. Chem., 1969, 20 133.
[12] C. A. Bunton, G. Stedman, J. Chem. Soc. 1959, 3466.
[13] G. Y. Markovits, S. E. Schwartz, L. Newman, Inorg. Chem. 1981, 20, 445
[14] L. P. Hammett, Physical Organic Chemistry, McGraw-Hill Inc., New York,
1940. p. 294
[34] D. A. McQuarrie, Statistical Mechanics. Harper and Row, New York, 1976.
[35] E. Riordan, N. Minogue, D. Healy, P. O’Driscoll, J. R. Sodeau, J. Phys.
Chem. (A) 2005, 109, 779–786.
[36] C. Anastasio, L. Chu, Env. Sci. Technol. 2009, 43, 1108.
[37] J. Crugeiras, unpublished results.
[38] This value for PhNH2OMeþ is derived from the estimate of pKa1 ¼ 4.5
based upon literature values for analogues,[21] including NH3OHþ
(6.0)[11] and MeNH2OMeþ (4.75)[40] in water.
[39] M. Canle, LW. Clegg, I. Demirtas, M. R. J. Elsegood, J. Haider, H. Maskill,
P. C. Miatt, J. Chem. Soc., Perkin Trans. 2 2001, 1742 (for the effect of
acetonitrile as cosolvent).
¨
[15] C. Doring, H. Gehlen, Z. Anorg. Allg. Chem. 1961. 312, 32.
[40] T. C. Bissot, R. W. Parry, D. H. Campbell, J. Am. Chem. Soc. 1957, 79, 796.
[41] We thank Drs T. W. Bentley and R. A. McClelland for helpful discus-
sions on this issue.
[16] E. D. Hughes, C. K. Ingold, J. H. Ridd, J. Chem. Soc. 1958, 77.
[17] E. D. Hughes, C. K. Ingold, J. H. Ridd, J. Chem. Soc. 1958, 88.
[18] J. H. Ridd, Quarterly Reviews, 1961, 15, 418.
[42] D. R. Lide (Ed.). CRC Handbook of Chemistry and Physics, (89th edn).
CRC Press/Taylor and Francis, Boca Raton, FL, 2009.
[19] G. C. M. Bourke, G. Stedman, J. Chem. Soc., Perkin Trans. 2 1992, 161.
[20] M. N. Hughes, T. B. D. Morgan, G. Stedman, Chem. Commun. 1966, 241.
J. Phys. Org. Chem. 2011, 24 162–171
Copyright ß 2010 John Wiley & Sons, Ltd.
View this article online at wileyonlinelibrary.com