Mechanism of decomposition of the human defense factor hypothiocyanite near physiological pH

Article abstract of DOI:10.1021/ja2083152

Relatively little is known about the reaction chemistry of the human defense factor hypothiocyanite (OSCN^-) and its conjugate acid hypothiocyanous acid (HOSCN), in part because of their instability in aqueous solutions. Herein we report that HOSCN/OSCN^- can engage in a cascade of pH- and concentration-dependent comproportionation, disproportionation, and hydrolysis reactions that control its stability in water. On the basis of reaction kinetic, spectroscopic, and chromatographic methods, a detailed mechanism is proposed for the decomposition of HOSCN/OSCN^- in the range of pH 4-7 to eventually give simple inorganic anions including CN ^-, OCN^-, SCN^-, SO₃^2-, and SO₄^2-. Thiocyanogen ((SCN)₂) is proposed to be a key intermediate in the hydrolysis; and the facile reaction of (SCN) ₂ with OSCN^- to give NCS(=O)SCN, a previously unknown reactive sulfur species, has been independently investigated. The mechanism of the aqueous decomposition of (SCN)₂ around pH 4 is also reported. The resulting mechanistic models for the decomposition of HOSCN and (SCN) ₂ address previous empirical observations, including the facts that the presence of SCN^- and/or (SCN)₂ decreases the stability of HOSCN/OSCN^-, that radioisotopic labeling provided evidence that under physiological conditions decomposing OSCN^- is not in equilibrium with (SCN)₂ and SCN^-, and that the hydrolysis of (SCN)₂ near neutral pH does not produce OSCN^-. Accordingly, we demonstrate that, during the human peroxidase-catalyzed oxidation of SCN^-, (SCN)₂ cannot be the precursor of the OSCN^- that is produced.

Full text of DOI:10.1021/ja2083152

ARTICLE
pubs.acs.org/JACS
Mechanism of Decomposition of the Human Defense Factor
Hypothiocyanite Near Physiological pH
J ꢀo zsef Kalm ꢀa r, Kelemu L. Woldegiorgis, Bernadett Biri, and Michael T. Ashby*
Department of Chemistry and Biochemistry, University of Oklahoma, Norman, Oklahoma 73019, United States
S Supporting Information
b
ABSTRACT: Relatively little is known about the reaction chemistry of the human defense
ꢀ
factor hypothiocyanite (OSCN ) and its conjugate acid hypothiocyanous acid (HOSCN),
in part because of their instability in aqueous solutions. Herein we report that HOSCN/
ꢀ
OSCN can engage in a cascade of pH- and concentration-dependent comproportiona-
tion, disproportionation, and hydrolysis reactions that control its stability in water. On the
basis of reaction kinetic, spectroscopic, and chromatographic methods, a detailed mech-
ꢀ
anism is proposed for the decomposition of HOSCN/OSCN in the range of pH 4ꢀ7 to
ꢀ
ꢀ
ꢀ
2ꢀ
3
eventually give simple inorganic anions including CN , OCN , SCN , SO , and
SO4 . Thiocyanogen ((SCN) ) is proposed to be a key intermediate in the hydrolysis;
and the facile reaction of (SCN) with OSCN to give NCS(dO)SCN, a previously
2
ꢀ
2
ꢀ
2
unknown reactive sulfur species, has been independently investigated. The mechanism of the aqueous decomposition of (SCN)2
around pH 4 is also reported. The resulting mechanistic models for the decomposition of HOSCN and (SCN) address previous
2
ꢀ
ꢀ
empirical observations, including the facts that the presence of SCN and/or (SCN) decreases the stability of HOSCN/OSCN ,
that radioisotopic labeling provided evidence that under physiological conditions decomposing OSCN is not in equilibrium with
2
ꢀ
ꢀ
ꢀ
(
SCN) and SCN , and that the hydrolysis of (SCN) near neutral pH does not produce OSCN . Accordingly, we demonstrate
2
2
ꢀ
ꢀ
that, during the human peroxidase-catalyzed oxidation of SCN , (SCN) cannot be the precursor of the OSCN that is produced.
2
ꢀ
’
INTRODUCTION
reports indicate that the decomposition of OSCN is faster at low
ꢀ
pH and with high concentrations of SCN . Considering the
observation that (SCN) does not yield OSCN when it is
one might conclude that the
comproportionation of HOSCN and SCN to give (SCN) is
involved in the decomposition mechanism of OSCN near neutral
pH. However, isotopic exchange of the C atoms of OSCN and
SCN is not observed near neutral pH, as might be expected for a
There are approximately 200 references in the primary
ꢀ
ꢀ
1,2
2
literature that explicitly mention HOSCN and/or OSCN .
The publications are roughly equally distributed in the dental,
food, medical, and chemical literature. A very large number of
additional publications certainly concern OSCN ; however, they
do not explicitly mention the compound. For example, nearly
15ꢀ18
hydrolyzed at neutral pH,
ꢀ
2
ꢀ
ꢀ
ꢀ
ꢀ
6
000 journal articles involve the enzyme lactoperoxidase, for
19
ꢀ
ꢀ
reversible comproportionation:
which SCN is known to be the substrate in vivo, and OSCN is
3,4
the only known product. Interestingly, many of the publica-
tions in dental
chemical studies. Indeed, much of the current interest in OSCN
is derived from its relevance in food science and medicine.
14
HOSCN þ S CNꢀ þ Hþ f NCSꢀS CN þ H2O
14
ð1Þ
ð2Þ
1
,5,6
7ꢀ11
and dairy
literature predate most of the
ꢀ
2,13
NCSꢀS CN þ H2O f HOS CN þ SCN þ Hþ
14
14
ꢀ
1
There appears to be a regrettable lack of communication between
these disciplines; consequently, many inconsistencies exist in the
literature, particularly concerning the chemistry. In addition, the
primary conclusions of many of these studies have been rendered
suspect due to relatively recent advances in our understanding of
It is furthermore noteworthy that (SCN) accelerates the decom-
2
ꢀ
ꢀ
position of OSCN (i.e., OSCN and (SCN) react even at
2
neutral pH before the latter species decomposes; vide infra).
3
5
ꢀ
2ꢀ
4
When HOSCN is reacted with SCN , the SO that is
14
35
obtained does not contain the label. While the C and S tracer
studies do not preclude the involvement of NCSCN, O SCN ,
O SCN , and other possible intermediates, they suggest that the
reversible equilibrium between OSCN and (SCN) in the
ꢀ
ꢀ
the chemistry which suggest that OSCN should not exist under
2
ꢀ
the conditions employed in some of the studies. Finally, while it is
3
ꢀ
ꢀ
generally assumed that OSCN is the only biologically active
2
ꢀ
ꢀ
species produced from the oxidation of SCN in vivo, it is clear
presence of SCN (eqs 1 and 2) is not involved in the decom-
that many potentially active species must be formed during the
position mechanism. The eventual decomposition products of
ꢀ
ꢀ
ꢀ
ꢀ
2ꢀ
2ꢀ
redox/hydrolysis cascade of OSCN that produces the eventual
OSCN are CN , OCN , SO , and SO4 , depending on the
3
ꢀ
ꢀ
2ꢀ
2ꢀ 14,15
products CN , OCN , SO , and SO4
.
3
specific reactionconditions. These products may be accounted for
Until now, there have been no methodical investigations of the
ꢀ
kinetics and mechanisms of the decomposition of OSCN near
Received: September 2, 2011
Published: October 27, 2011
physiologically relevant neutral pH conditions. However, anecdotal
r 2011 American Chemical Society
19911
dx.doi.org/10.1021/ja2083152
|J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
by the following atom/electron balance (eqs 3ꢀ5, which does not
necessarily reflect the intimate mechanism or the intermediates
that are involved):
2
HOSCN f HO SCN þ HSCN
ð3Þ
2
HO2SCN þ HOSCN f HO3SCN þ HSCN
ð4Þ
HO3SCN þ H2O f H2SO3 þ HOCN and=or
H2SO4 þ HCN
ð5Þ
ð6Þ
which yield the net reaction of eq 6:
3
HOSCN þ H2O f ðH2SO3 þ HOCN and=or
H2SO4 þ HCNÞ þ 2HSCN
It has been previously suggested that the mechanism involves
dismutation of HOSCN (eq 3) because second-order kinetics
ꢀ
have been observed under pH conditions where OSCN
+
18,20
HOSCN predominate.
However, others have reported first-
order kinetics for the decomposition of OSCN (and under
ꢀ
specific conditions we have observed first-order kinetics as well;
1
9
vide infra). The mechanism of eqs 3ꢀ5 has been proposed to
20
ꢀ
produce H SO + HCN, but CN is not produced quantita-
2
4
Figure 1. pH dependence of the initial absorbance (A , top) and the
0
tively, presumably because it reacts with HOSCN to produce
NCSCN, which subsequently hydrolyzes to produce a 1:1
mixture of SCN and OCN . Importantly, although there is
observed pseudo-first-order rate constants (kobs, bottom) of the decom-
ꢀ
position of HOSCN + OSCN in the range from pH 4.37 to pH 5.66.
ꢀ
ꢀ 21
Markers: measured data (top: dots, 306 nm; crosses, 376 nm). Lines:
no experimental evidence for the intermediates that are involved
ꢀ
results of global data fit. [HOSCN] = 3.5 mM, [SCN ] = 0.25 M,
[HOAc] + [OAc ] = 0.50 M, μ = 1.0 M, 25.0 °C.
T,0
ꢀ
ꢀ 22,23
ꢀ
in the decomposition of OSCN , cyanosulfite, O SCN ,
and
2
ꢀ
24
cyanosulfate, O SCN , have been observed in nonaqueous
3
media (but not in water). However, many other potential
intermediates are possible, including the unknown anhydride
(
i.e., no absorbing intermediate is present at detectable quanti-
ꢀ
ties during the decomposition over wide ranges of [SCN ]
and pH). The measured pseudo-first-order rate constants (kobs
were found to be invariable to the buffer concentration and
to the initial analytical HOSCN concentration ([HOSCN]
(
cf. thiosulfinate esters, the anhydrides of sulfenic acids):
HOSCN f NCSðdOÞSCN þ H2O
In the present study we investigate the decomposition of HOSCN
)
2
ð7Þ
=
T,0
ꢀ
[HOSCN]
+ [OSCN ]
). The values of kobs are directly
0
0
ꢀ
and (SCN) in detail around pH 4.0 in acetate buffer, as well as the
proportional to [SCN ] over 1 order of magnitude change in
[SCN ], with a negligible ordinate intercept (Figure S3 of the
Supporting Information). The [H ] dependency of the reaction is
complex. At constant [SCN ] the initial spectrum of the reactant
solution recorded 0.75 ms after the jump to the desired pH
(second mixing) ispH-dependent asseen inFigure1. The series of
the initial spectra recorded at different pH can be successfully
analyzed by taking into account only the protolytic equilibrium
2
ꢀ
ꢀ
decomposition of OSCN around pH 7.0 in phosphate buffer, and
+
show that the observed mechanisms are relevant to the stability of
ꢀ
ꢀ
ꢀ
OSCN at physiological pH. As SCN is considered to be a pseudo-
25ꢀ27
halide in several aspects of reactivity,
considerations on analogous hydrolysis chemistry of halogens.
we base some of our
28ꢀ30
’
RESULTS
ꢀ
between HOSCN (Amax at 306 nm) and OSCN (Amax at
ꢀ
Decomposition of HOSCN/OSCN around pH 4. The
3
76 nm) (Figure S4 of the Supporting Information). Figure 1
+
kinetics of the reaction were studied in the range from pH 4.3 to
illustrates that the values of k increase with [H ]. The k vs
obs
obs
3
1
+
pH 6.0. Because the pK of HOSCN is 4.85, a mixtureofHOSCN
a
[H ] curve is sigmoidal, and it can be fit to the following function:
ꢀ
and OSCN was present. The decomposition of HOSCN/
OSCN was monitored photometrically at multiple wavelengths
ꢀ
0
þ 2
ξ ½H ꢁ
1
kobs
¼
ð8Þ
from 290 to 705 nm; two spectral features were observed: at 306
þ
þ 2
ξ þ ξ ½H ꢁ þ ½H ꢁ
2
3
(
shoulder) and 376 nm (maximum) (Figure S2 of the Supporting
0
Information). According to the singular value decomposition
where ξ , ξ , andξ are pH-independentconstants. Ifwetakeinto
account that kobs is directly proportional to [SCN ], the following
rate law is deduced:
1
2
3
ꢀ
(
SVD) analysis of multiwavelength data (SpecFit32), the reaction
ꢀ
follows strict pseudo-first-order kinetics when SCN is in large
excess over HOSCN/OSCN . Second-order kinetics were never
observed during the experiments.
ꢀ
14,15
þ 2
ꢀ
The half-life of the decom-
ξ1½H ꢁ ½SCN ꢁ
ꢀ
v ¼
½HOSCNꢁT
ð9Þ
þ
þ
2
position of HOSCN/OSCN is ∼0.5 s, depending on the actual
ξ þ ξ ½H ꢁ þ ½H ꢁ
2
3
ꢀ
+
conditions (Figure S2). For constant [SCN ] and [H ], principal
component analysis (PCA) of the collected time-resolved spectra
showed the overdominating variance of the first component
where ξ , ξ , and ξ are constants (the values of ξ are given in
Table S2 of the Supporting Information) and [HOSCN] is the
1
2
3
T
1
9912
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
a
Table 1. Relative Concentrations of Decomposition Products As Determined by Ion Chromatography
relative concentration, %
ꢀ
ꢀ
32ꢀ
42ꢀ
ꢀ
ꢀ
OCN
CN
SO
SO
SCN
HOSCN + CN pathway, %
ꢀ
HOSCN/OSCN (pH = 4.33)
11 ( 2 (17)
53 ( 8 (27)
12 ( 2 (12)
59 ( 6 (28)
8 ( 2 (11)
0 (0)
2 ( 1 (3)
5 ( 2 (8)
4 ( 1 (4)
7 ( 2 (10)
20 ( 6 (25)
17 ( 4 (19)
22 ( 2 (27)
18 ( 1 (18)
69 ( 6 (71)
55 ( 8 (73)
14
19
8
ꢀ
OSCN (pH = 6.82)
(
SCN)
2
2
(pH = 3.89)
(pH = 6.90)
23 ( 5 (19)
0 (0)
159 ( 7 (169)
144 ( 13 (172)
(
SCN)
17
a
Parentheses: calculated values based on the proposed stoichiometry of the reactions (Schemes 1 and 2) and on the indicated contribution to the total
ꢀ
consumption of the reactants of the HOSCN + CN pathway (Scheme 3, R14).
ꢀ
time-dependent analytical concentration of HOSCN + OSCN in
the reaction mixture.
To study the reversibility of the decomposition of HOSCN/
ꢀ
OSCN , triple-mixing stopped-flow experiments with in situ
UVꢀvis detection were performed. When aging times of
τ < 6t were employed using an acidic quench to produce
2
1/2
ꢀ
14,15,32
(
SCN) and a basic quench to produce OSCN ,
the
2
reactions were complete within the mixing time. The amount
of (SCN) or OSCN formed was quantitatively the same as the
remaining amount of HOSCN/OSCN that was expected at the
ꢀ
2
ꢀ
time of quenching its decomposition mixture (Figure S5 of the
Supporting Information). At aging times τ > 7t1/2, no spectral
2
change was observed after quenching.
Electrospray ionization mass spectrometry (ESI-MS) experi-
ꢀ
ꢀ
ments provided evidence for the formation of CN , OCN ,
2ꢀ
2ꢀ
SO4 , and traces of SO3
during the decomposition of
HOSCN/OSCN around pH 4. No intermediates, only the
end products and OSCN from the basic quench, were identified
ꢀ
Figure 2. Effect of [(SCN)
2
]
T,0 (the initial concentration of (SCN)
; see Discussion for details) on the observed
pseudo-first-order rate constants for the hydrolysis of (SCN) in the
2
+
ꢀ
ꢀ
ꢀ
2
(SCN)
3
+ HO(SCN)
in the triple-mixing quench-flow/ESI-MS experiments. The
amount of products relative to [HOSCN]T,0 are given in Table 1.
The measured amounts of products were invariable to the
2
range from pH 3.66 to pH 4.50 (indicated to the right from the plots):
(dots) measured data; (lines) results of global data fit. The kinetics of the
ꢀ
ꢀ
presence of excess SCN during the decomposition.
reaction deviates from pseudo-first-order at low [(SCN)
2
]
T,0. [SCN ] =
ꢀ
ꢀ
0.25 M, [HOAc] + [OAc ] = 0.50 M, μ = 1.0 M, and 25.0 °C.
Decomposition of HOSCN/OSCN around pH 7. The
ꢀ
kinetics of the decomposition of OSCN were studied in the
ꢀ
pH range of 6.1ꢀ7.2. The reaction is strictly pseudo-first-order
analytical concentration of HOSCN + OSCN in the reaction
ꢀ
ꢀ
when SCN is in a large excess over OSCN , with a half-life of
mixture. Acidic and basic triple-mixing quench studies were
ꢀ
∼
20 s. PCA of time-resolved photometric data revealed only one
performed, and the amount of (SCN) or OSCN formed was
2
ꢀ
detectable absorbing species (A
at 376 nm) during the
max
the same as the remaining amount of OSCN that was expected
ꢀ
reaction. The value of kobs was independent of the initial OSCN
concentration and independent from the buffer concentration.
kobs is proportional to [SCN ] with a nonzero intercept (Figure
at the time of quenching.
The ions OCN , SO4 , and SO3 were identified in the
ꢀ
2ꢀ
2ꢀ
ꢀ
ESI-MS spectra of the sample produced during the decomposi-
ꢀ
S6 of the Supporting Information):
tion of OSCN around pH 7. Evidence was also found for the
+
ꢀ
formation of elementary sulfur as HS
3
at m/z = 96.9228 was
kobs ¼ ξ ½SCN ꢁ þ ξ
ð10Þ
4
5
present in the +ESI-MS spectrum of the product solution.
According to the ion chromatography (IC) measurements, the
proportions of the products are significantly different from that
observed in acetate buffer as seen in Table 1. The measured
ꢀ
where ξ and ξ are [SCN ]-independent constants. The initial
4
5
spectrum of the reactant solution recorded 0.75 ms after pH
+
jump is independent of [H ], although k is proportional to
[
obs
+
H ] with a nonzero ordinate intercept (Figure S6):
product distribution around pH = 7 was the same regardless of
ꢀ
whether OSCN was generated from (SCN) or by the LPO/
2
þ
ꢀ
ꢀ
kobs ¼ ξ ½H ꢁ þ ξ
ð11Þ
6
7
SCN /H O system. No effect of an initial SCN excess was
2 2
observed on the product distribution.
where ξ and ξ are pH-independent constants. From the fitted
6
7
Decomposition of (SCN) around pH 4. Kinetic studies were
2
lines (e.g., Figures S6 and Table S2), a simple calculation leads to
+
ꢀ
performed in the range of pH 3.8ꢀ4.7. Since both (SCN)
2
and
is formed in
the following: ξ /[H ] = ξ /[SCN ] and ξ = ξ . Thus, the rate
ꢀ
ꢀ
3
4
6
5
7
ꢀ
SCN were present in the solutions, (SCN)
law for the decomposition of OSCN at pH 7 is
14
equilibrium. The latter equilibration is established in the mixing
14,15,33
þ
ꢀ
time over the entire range of pH.
The majority of the
v ¼ ðξ ½H ꢁ½SCN ꢁ þ ξ Þ½HOSCNꢁ
ð12Þ
8
9
T
kinetic experiments were followed at 300 nm, which is the
+
ꢀ
ꢀ 14,15,33
where ξ and ξ are constants, ξ = ξ /[H ] = ξ /[SCN ]
absorbance maximum of (SCN)3
.
On the basis of the
8
9
8
4
6
ξ = ξ = ξ (Table S2), and [HOSCN] is the time-dependent
UVꢀvis and MS spectra (vide infra) of the resulting solutions,
9
5
7
T
1
9913
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
[
HOCl] by changing the aging time before the second mixing
0
cycle (τ ), thus allowing the in situ-generated (SCN) to decom-
1
2
pose at pH 0.7 to various extents. The independent sets of
experiments gave identical results (Figure S8 of the Supporting
Information). The kinetic curves recorded at low [(SCN)2]T,0
match the end of the kinetic curves recorded at high [(SCN)2]T,0
under the same conditions (Figure S8). However, when starting at
high[(SCN)2]T,0 the theoretical error oftreatingthe whole course
of decomposition as a pseudo-first-order process and fitting kobs
accordingly is negligible, as only the last 5ꢀ10% of the data
deviates significantly from the single-exponential function which
fits the majority of the curve.
The following results were obtained at high [(SCN)2]T,0
unless stated otherwise. PCA of time-resolved multiwavelength
photometric data revealed that only one absorbing species (Amax
at 300 nm) is detectable during the reaction from 290 to 705 nm
ꢀ
+
at a given [SCN ] and [H ]. The pseudo-first-order kinetic
curves measured at 300 nm were extrapolated using single-
exponential functions to ꢀ0.6 ms (time correction; see details
in Supporting Information). The calculated zero-time absor-
CALC
ꢀ
bance at 300 nm (A0
) increases sigmoidally with [SCN ] at
+
constant [H ] (Figure S9 of the Supporting Information). Two
CALC
ꢀ
effects are evident in the A0
vs [SCN ] curve: (1) shifting of
ꢀ
the (SCN) /(SCN)3 rapid equilibrium and (2) the rate of
2
decomposition of the in situ-generated (SCN) at pH = 0.7 in the
2
first mixing cycle shows an inverse second-order dependence
ꢀ
14,15
ꢀ
on [SCN ].
Thus, at lower [SCN ], the [(SCN) ] is less
2 T,0
after the τ first aging-time (Figure S9; see details in the
1
Supporting Information). Data recorded at high and low
[
(SCN)2] showed consistency; kobs is inversely dependent
T,0
ꢀ
on [SCN ] (as seen in Figure 3):
ꢀ
Figure 3. pH and [SCN ] dependence of kobs of the decomposition of
ꢀ
(SCN)
2
in the range from pH 3.77 to pH 4.74. Top, [SCN ]
dependence: the initial concentration of (SCN) varied with [SCN ]
ξ10
ꢀ
kobs ¼ 1
ð13Þ
ꢀ
2
þ ξ ½SCN ꢁ
11
at pH 3.87 and pH 4.42 (see explanation in text); (dots) measured data
ꢀ
at high [(SCN)
2
]
T,0 from 1.0 to 0.4 mM; (crosses) measured data at
where both ξ10 and ξ11 are [SCN ]-independent constants. ξ
10
lower [(SCN)2] from 0.4 to 0.1 mM; (line) result of global data fit.
T,0
increases with increasing pH, and ξ11 slightly decreases with
ꢀ
ꢀ
[
HOAc] + [OAc ] = 0.50 M, μ = 1.0 M, and T = 25.0 °C. Bottom, pH
increasing pH (Table S2). At constant [SCN ], independent
dependence: (dots) measured data; (line) result of global fit; (inset) kobs
CALC
from [(SCN) ] , A0
increases to saturation with increas-
CALC
2 T,0
proportional to the calculated initial absorbance at 300 nm (A
0
).
(SCN)2]T,0 = 1.0 mM, [SCN ] = 0.25 M, [HOAc] + [OAc ] = 0.50 M,
μ = 1.0 M, and 25.0 °C.
ꢀ
ꢀ
ing pH (Figure S10 of the Supporting Information). The initial
spectrum of the reaction mixture recorded at different pH values
[
(
t = 1.5 ms) displays only one peak at 300 nm; thus, the fast
ꢀ
equilibrium which yields HOSCN/OSCN is not relevant under
the conditions employed.
approaches saturation with increasing pH, and A0
ꢀ
3
the product of the hydrolysis of the (SCN) /(SCN) mixture is
14,15
2
As illustrated in Figure 3, k
obs
CALC
ꢀ
not HOSCN/OSCN . The reaction follows first-order kinetics
and kobs
ꢀ
in the presence of a large excess of SCN with a half-life of 2ꢀ4
are proportional to one another. The curves can be fit with the
ms, depending on the actual conditions (Figure S7 of the
Supporting Information). The kinetic curves recorded at high
following functions both at high and low [(SCN)2]T,0
:
ꢀ
ξ ½OH ꢁ
[(SCN)2]T,0 (vide infra) cannot be fit with second-order func-
12
kobs
¼
ð14Þ
ð15Þ
ꢀ
tions, and the least-squares residuals are the same for single-
exponential and biexponential fits.
1 þ ξ ½OH ꢁ
13
CALC
As expected, kobs is generally independent of [(SCN) ]
.
2 T,0
obs
kobs ¼ ξ14A0
þ ξ15
However, for [(SCN)2]T,0 lower than ∼200ꢀ300 μM, k
where ξ , ξ , ξ , and ξ are pH-independent constants
12
13
14
15
decreases with decreasing [(SCN) ] (Figure 2). The photo-
2
T,0
(
Table S2), although dependent on [(SCN)2]T,0. The rate law
metric data were exceptionally low quality under these conditions
because the measured absolute change was only about 0.03
absorbance unit, which is comparable to the level of the noise in
the first few milliseconds of the measurements as a consequence of
the shock of the stopped-flow mixing cycle. Nonetheless, least-
squares fit of the kinetic curves gives better results with first-order
than with second-order kinetic functions. The [(SCN)2]T,0 was
established by two different means: (1) by changing the concen-
tration of HOCl in the first mixing cycle and (2) at constant
at high [(SCN)2]T,0 is the following:
ꢀ
ξ ½OH ꢁ
1
6
v ¼ 1
½ðSCNÞ2ꢁT
ð16Þ
ꢀ
ꢀ
þ ξ ½SCN ꢁ þ ξ ½OH ꢁ
17
18
where ξ , ξ , and ξ are constants (Table S2) and [(SCN) ] is
1
6
17
18
2 T
ꢀ
3
the time-dependent analytical concentration of (SCN) + (SCN ) .
2
Quenching the decomposition of (SCN) with HClO in
2
4
triple-mixing experiments yielded (SCN) and quenching with
2
1
9914
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
Scheme 1. Mechanism of Decomposition of HOSCN/
ꢀ
a
OSCN in the Range from pH 4 to pH 7
a
Above the line, proposed kinetic model; below the line, proposed
stoichiometry with X e 1 (kꢀ2, kꢀ3, and kꢀ4 are the rate constants of the
backward reactions of R2, R3, and R4, respectively; italicized reactions
occur after the rate-determining R6 step).
ꢀ
ꢀ
(
SCN) at pH 7. The eventual products at pH 7 (OCN , SCN ,
2
2
ꢀ
2ꢀ
SO3 , and SO4 ) were identified by MS and quantified by IC
the results are given in Table 1).
Reaction of (SCN) with HOSCN/OSCN . Figure 4 illustrates
(
ꢀ
2
representative results for the reaction of (SCN) with HOSCN/
2
ꢀ
ꢀ
OSCN . When mixing the two reactants at pH = 0.70 and high
SCN ] with (SCN) in small excess over HOSCN, an elevation
[
2
of the absorbance was observed at 300 nm which corresponds to
ꢀ
the production of (SCN) /(SCN) during the comproportio-
Figure 4. Change in absorbance at 300 nm after directly mixing
2
3
ꢀ
14
ꢀ
nation of HOSCN and SCN . The production of (SCN) was
complete within 100 ms, and it decomposes with homogeneous
solutions of (SCN)
2
and HOSCN/OSCN at pH = 0.70 (top) and at
2
pH = 3.89 (bottom). At pH = 0.70, HOSCN comproportionates with
ꢀ
14,15
SCN to give (SCN) which decomposes in a homogeneous second-
second-order kinetics.
When (SCN)2 was mixed with
2
ꢀ
ꢀ
order reaction. [(SCN)
2
]
T,0 = 1.4 mM, [HOSCN]T,0 = 1.1 mM,
HOSCN/OSCN at pH = 3.89 and high [SCN ], the absor-
bance decreased at 300, 306, and 376 nm (compared to the
references) and the reaction was over within the mixing time. We
were not able to study the kinetics of the process in detail to verify
the direct reaction of (SCN) with HOSCN/OSCN at pH 4, as
ꢀ
[SCN ] = 0.25 M, [HClO
4
] = 0.20 M, μ = 1.0 M, and 25.0 °C:
(dashed line) reference ([HOSCN]T,0 = 0); (continuous line) experi-
ꢀ
ment. At pH = 3.89, HOSCN/OSCN quenches (SCN) . [(SCN) ]
2
2 T,0
ꢀ
=
1.4 mM, [HOSCN]T,0 = 1.2 mM, [SCN ] = 0.25 M, [HOAc] +
ꢀ
ꢀ
2
[OAc ] = 0.50 M, μ = 1.0 M, and 25.0 °C: (dashed line) reference
maintaining the reaction conditions for a series of experiments
proved to be technically challenging. However, a semiquantitative
observation was made that when [(SCN)2]T,0 was varied from 0.1
to 0.4 mM with [HOSCN]T,0 = 0.4 mM, the kinetics showed
biphasic characteristics (a first step of 1ꢀ2 ms half-life and a second
step with a 0.4ꢀ0.5 s half-life) with a decrease in absorbance at
([HOSCN] = 0); (continuous line) experiment.
T,0
ꢀ
18,34
NaOH yielded OSCN ,
as evident by in situ photometric
detection. A significantly larger molar amount of (SCN) was
2
ꢀ
produced in the acidic quench than OSCN was produced in the
basic, and slightly more than the remaining amount of (SCN)2
300 nm where both (SCN) and HOSCN absorb (Figure S12 of
2
(
vide infra). The data sets were measured with high relative
the Supporting Information). The measured initial absorbance (t =
0.8 ms) at 300 nm increased with increasing [(SCN) ] , and the
starting absorbance of the second step (t = 100 ms) decreased with
increasing [(SCN)2]T,0 (Figure S13 of the Supporting In-
formation); thus the first step is associated with the reaction of
(SCN) with HOSCN/OSCN . The rate of the second step, the
decomposition of the excess HOSCN/OSCN , also increased with
increasing[(SCN) ] . IC analysis of the product solutions was not
error due to the low absorbance values or insufficient sta-
bility of the OSCN solutions (Figure S11 of the Supporting
2
T,0
ꢀ
Information).
ꢀ
ꢀ
2ꢀ
2ꢀ
ESI-MS data evidence that CN , OCN , SO , and SO3
4
ꢀ
are produced during the decomposition of (SCN) at pH 4
2
2
ꢀ
(
Table 1). No intermediates were identified in the quench-flow/
ꢀ
ESI-MS experiments. An excess of SCN present in the solutions
during the decomposition has no observable effect on the relative
amounts of products.
2
T,0
successful because all of the experiments were carried out in the
presence of high concentrations of SCN .
ꢀ
Decomposition of (SCN) around pH 7. No kinetic informa-
2
tion could be obtained for the decomposition of (SCN) since
the reaction is complete in the mixing time of the stopped-flow
2
’
DISCUSSION
ꢀ
instruments. Thus, on the basis of UVꢀvis and MS measure-
Mechanism of Decomposition of HOSCN/OSCN in the
Range from pH 4 to pH 7. The proposed mechanism of
ꢀ
ments, OSCN is excluded as the product of the hydrolysis of
1
9915
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
a
Scheme 2. Mechanism of Decomposition of (SCN) at pH 4
Table 2. Constants Resulting from Global Fit of the Kinetic
2
ꢀ
Data on the Decomposition of (SCN) and HOSCN/OSCN
2
in the Region from pH 4 to pH 7
constant
value
ꢀ
5
a
K
1
(1.17 ( 0.4) ꢂ 10
M
ꢀ1
ꢀ1
k
2
50.0 ( 0.9 M
s
3
ꢀ1
k
ꢀ2
(2.07 ( 0.09) ꢂ 10 s
ꢀ
2
ꢀ1 b
K
2
2.4 ꢂ 10
137 ( 23 s
M
ꢀ
1
k
3
c
11 ꢀ1 ꢀ1 b
k
ꢀ3
3.6 ꢂ 10
M
s
9
ꢀ1
K
ꢀ3
(2.65 ( 0.3) ꢂ 10 M
8
ꢀ1 ꢀ1
k
4
(1.29 ( 0.07) ꢂ 10 M
s
3
ꢀ1 b
k
ꢀ4
3.4 ꢂ 10 s
ꢀ
5
b
K
K
ꢀ4
2.7 ꢂ 10
M
ꢀ1 a
5
0.54 ( 0.1 M
6
ꢀ1 ꢀ1
a
k
6
(4.01 ( 0.3) ꢂ 10 M
s
Above the line, proposed kinetic model; below the line, proposed
2ꢀ
300 nm
3
ꢀ1
ꢀ1 b
E
7
{HO(SCN)
}
1.6 ꢂ 10 M cm
ꢀ5
15,31,33 b
stoichiometry with X e 1 (to be consistent with the definitions in Scheme 1,
a
k
ꢀ2, kꢀ3, and kꢀ4 are the forward rate constants and k
3
, k
4
, and k
2
are the
May be compared with the values of K
1
= 1.41 ꢂ 10 M and K
= 0.43
5
ꢀ
1
backward rate constants of reactions ꢀR2, ꢀR3, and ꢀR4, respectively).
M
that have been previously reported.
Deduced from other
c
constants. The value of kꢀ3 (italic) is overestimated; the limiting rate of
1
0
ꢀ1 ꢀ1
diffusion is ca. 10
M
s
in water.
ꢀ
decomposition of HOSCN/OSCN is illustrated in Scheme 1.
+
+
Because no second-order behavior was observed, the reaction of
And, at pH 7 we assume that k [H ] , k and [H ] , K ;
4
3
1
ꢀ
two HOSCN and/or OSCN molecules cannot be a major
therefore:
14,15
pathway under the conditions that were employed.
The
þ
ꢀ
reactive species is HOSCN, both at higher and lower pH. We
k2k3½H ꢁ½SCN ꢁ
K ðk þ k Þ
ν ¼
½HOSCNꢁT
ð21Þ
assume the formation of a previously unknown intermediate
1
ꢀ2
3
ꢀ
ꢀ
HO(SCN) from HOSCN and SCN in a reversible step.
2
ꢀ
SCN is regarded as a pseudo-halide and (SCN) as a pseudo-
halogen on the basis of their chemical reactivity.
2
The experimentally determined and the deduced limiting rate
laws agree both at pH 4 (eq 9 and eq 20, respectively) and at pH 7
(eq 12 and eq 21, respectively). However, the proposed mechan-
ism does not account for the observed intercepts of the k vs
SCN ] and k vs [H ] curves at pH 7 (Figure S6 and
parameter ξ in Table S2), which clearly indicates the presence
of an alternative mechanism at pH g 7 and [SCN ] e
HOSCN]T,0. The disproportionation of HOSCN/OSCN in
a bimolecular reaction is a feasible explanation.
1
4,15,33,35ꢀ37
ꢀ
The role of I OH in the hydrolysis chemistry of I has been
2
2
2
8ꢀ30
previously described in detail.
Accordingly, we propose
obs
ꢀ
+
the existence of equilibria R2ꢀR4 (Schemes 1 and 2). The
[
obs
dissociation of HO(SCN)2ꢀ
2
to (SCN) (R3) is assumed,
9
ꢀ
which can be proton-catalyzed at lower pH (R4). The general
rate law is given by
0
ꢀ
[
14,15
ꢀ
However, at
v ¼ k6½ðSCNÞ ꢁ½OSCN ꢁ
ð17Þ
2
ꢀ
pH > 11, OSCN is known to hydrolyze to thiocarbamate
ꢀ
38
The steady-state approximation is valid for [HO(SCN) ] and
S-oxide.
2
[
(SCN) ], because both species are proposed to be produced
A global data fit was performed to validate our kinetic model:
the complete experimental kinetic data set ([SCN ] and pH
dependence at pH 4 and pH 7) was fit to the proposed model,
with kobs and the initial absorbances at 306 and 376 nm defined as
dependent parameters, [H ] and [SCN ] defined as indepen-
dent parameters, and the floating constants were k , k , k , k ,
ξ , K , and E , E , E , and E (which were defined as the molar
absorptivities of HOSCN at 306 and 376 nm and the molar
absorptivities of OSCN at 306 and 376 nm, respectively). The
2
ꢀ
slowly and consumed rapidly. Furthermore, their concentrations
never exceed low μM that would result in a detectable signal as
ꢀ
ꢀ
both (SCN)3 (R6) and HO(SCN)2 (Table 2) have orders-of-
ꢀ
+
ꢀ
magnitude higher molar absorbances than HOSCN/OSCN :
2
ꢀ2
3
4
þ
ꢀ
2
αð1 ꢀ αÞk2k6ðk3 þ k4½H ꢁÞ½SCN ꢁ½HOSCNꢁ
T
ν ¼
α ¼
9 1 1 2 3 4
ꢀ
þ
kꢀ2ðkꢀ3½OH ꢁ þ kꢀ4Þ þ ð1 ꢀ αÞk6ðkꢀ2 þ k3 þ k4½H ꢁÞ½HOSCNꢁ
T
ꢀ
þ
½
H ꢁ
and ½HOSCNꢁT ¼ ½HOSCNꢁ þ ½OSCNꢀꢁ
ð18Þ
þ
model provided an excellent fit of the experimental data. The
resulting calculated constants are given in Tables 2 and S3.
Mechanism of Decomposition of (SCN) at pH 4. Scheme 2
K1 þ ½H ꢁ
Considering that (see the Supporting Information)
2
illustrates the proposed mechanism of the decomposition of
þ
ꢀ
kꢀ2ðK1 þ ½H ꢁÞðkꢀ3½OH ꢁ þ kꢀ4Þ
(
SCN) . The pH dependence of the initial spectrum around pH
2
þ
,
K k ðk þ k þ k ½H ꢁÞ½HOSCNꢁ
leads to the rate law
ð19Þ
ð20Þ
4 is the result of the protolytic equilibria ꢀR3 and ꢀR4, which
1
6
ꢀ2
3
4
T
are considered fast under the applied conditions. The equilibria
ꢀ
ꢀ
R3 and ꢀR4 are competitive with the formation of (SCN) in
3
þ
þ
ꢀ
R5. Because limiting second-order behavior is expected at low
[(SCN) , and limiting first-order behavior was observed at
high [(SCN) ] , a reversible first-order step is proposed to
k2ðk3 þ k4½H ꢁÞ½H ꢁ½SCN ꢁ
v ¼
½HOSCNꢁT
þ
þ
]
2 T
ðkꢀ2 þ k3 þ k4½H ꢁÞðK1 þ ½H ꢁÞ
2
T
ꢀ
precede a second-order step. We suggest that HOSCN/OSCN
1
9916
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
ꢀ
is produced in the dissociation of HO(SCN)2 in ꢀR2 and that
numerous independently recorded data sets were analyzed
simultaneously without considering the error of the independent
parameters (i.e., initial concentrations and pH), the results are
it subsequently reacts with (SCN) in the rate-limiting step R6.
2
The protonation state of the reacting species in R6 cannot be
determined on the basis of our kinetic experiments as we were
unable to measure the pH dependence of the reaction under
second-order conditions, and as all measurements were carried
out at least 0.3 pH units below the pK of HOSCN (K ). The
satisfactory.
Reaction of (SCN)
hydrolysis of (SCN)
ꢀ
with HOSCN and/or OSCN . The
2
ꢀ
to HOSCN/OSCN is well-defined at
2
both high and at low pH, but not at pH 4ꢀ8 because (SCN)
a
1
2
ꢀ
properties of reaction R6 are discussed in detail later. In deriving
does not hydrolyze to HOSCN/OSCN from pH 4 to 8 (vide
the general rate law, equilibrium ꢀR4 was omitted because K
=
supra). Considering the lack of other reactants in the (SCN)
2
ꢀ
4
ꢀ
K
K and because OSCN is assumed to react with (SCN)
solution, the simplest explanation for the absence of HOSCN/
ꢀ
3
w
2
ꢀ
(
R6; vide infra). Using the steady-state approximation for
HOSCN] and starting with eq 17:
OSCN under the described conditions is that (SCN)
2
reacts
ꢀ
[
with HOSCN/OSCN faster than (SCN)
2
hydrolysis can take
T
place. Indirect evidence was found for the existence of the
ꢀ
2
ꢀ
ν ¼ ð1 ꢀ αÞk k K ½OH ꢁ½ðSCNÞ ꢁ
ꢀ
2
6
ꢀ3
2 T
reaction of (SCN) with HOSCN/OSCN at pH 4 during the
2
ꢀ
ꢁꢀ1
study of the decomposition of (SCN) at low [(SCN) ] , and
2
2 T,0
ð1 ꢀ αÞk ½ðSCNÞ ꢁ
ꢀ
6
ꢀ
2 T
ꢂ αk2½SCN ꢁ þ
we obtained direct evidence by mixing the two reactants in
independent experiments.
ꢀ
1
þ K ½OH ꢁ þ K ½SCN ꢁ
ꢀ
3
5
ꢀ
ꢀ
ꢀ2
The rate-limiting step in the decomposition of HOSCN and
ꢂ ð1 þ Kꢀ3½OH ꢁ þ K5½SCN ꢁÞ
ꢀ
(
SCN) involves the reaction of OSCN with (SCN) (R6). In
2
2
þ
½
H ꢁ
principle, the product of such a reaction could be an oxo
α ¼
ꢀ
þ
K1 þ ½H ꢁ
ðSCNÞ ꢁ ¼ ½ðSCNÞ ꢁ þ ½ðSCNÞ ꢁ þ ½OHðSCNÞ ꢁ
derivative of the previously characterized (SCN)
3
. However,
ꢀ
ꢀ
we have not found a stable ground-state structure for a molecule
½
ꢀ
2
T
2
3
3
formulated as O(SCN)3 using computational methods
ð22Þ
(Supporting Information). Accordingly, we have described the
ꢀ
reaction of OSCN with (SCN) (R6) as a substitution to give
2
For high [(SCN) ] (see the Supporting Information)
ꢀ
2
T
“
O(SCN) ” and SCN . The reaction is assumed to take place
as a formal SCN for OSCN substitution because OSCN
is not known to react as a direct O-donor.
2
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
þ
ꢀ
k2ð1 þ Kꢀ3½OH ꢁ þ K5½SCN ꢁÞ½H ꢁ½SCN ꢁ
k K ½ðSCNÞ ꢁ
21,31,39
There
,
ð23Þ
ð24Þ
6
1
2 T
are two reasonable ground-state structures for “O(SCN)2”:
NCSꢀOꢀSCN (cf. a thiolsulfinate) and NCS(dO)ꢀSCN
and the limiting-rate law is
(
cf. a thiosulfenate ester). Remarkably, ab initio quantum chemi-
ꢀ
kꢀ2Kꢀ3½OH ꢁ
cal calculations yield stable ground states for both structures with
very similar total energies (Table S4 of the Supporting In-
formation). However, the activation barriers at room tempera-
ture (58 kcal/mol using a MP2/6-311+G** basis set, Figure S14
of the Supporting Information) appears to preclude their inter-
conversion, especially in the time frame of their expected life-
times. While the calculations do not distinguish between the
possible structures, evidence for an asymmetrical intermediate
has been previously provided by radio-labeling studies which
v ¼ 1
þ Kꢀ3½OH ꢁ þ K5½SCN ꢁ
½ðSCNÞ2ꢁT
ꢀ
ꢀ
Thus, reaction ꢀR2 is rate-limiting when [(SCN) ] is high. The
2
T
deduced rate law (eq 24) is the same as the experimentally
obtained rate law at high [(SCN)2]T,0 (see Results and eq 16). A
global data fit was performed: the experimental data set recorded
at high [(SCN) ] was fit to the limiting first-order model with
2
T,0
CALC
ꢀ
kobs and A
[
as dependent parameters and [SCN ] and
OH ] as independent parameters. Floating constants were as
follows: k 2, K , K , and E , E , and E (which are the molar
absorbances of (SCN) , (SCN) , and HO(SCN)2 , respec-
0
ꢀ
2ꢀ
ꢀ
demonstrated that SO4 in a pH 7 solution of OSCN and
ꢀ
ꢀ 19
ꢀ
ꢀ3
5
5
6
7
SCN is derived exclusively from OSCN , indicating the
conservation of the OꢀS bond. We therefore propose the
intermediate is NCS(dO)SCN.
ꢀ
3
ꢀ
2
tively, at 300 nm). The parameter k was approximated by fitting
6
the following equation to the kobs vs [(SCN) ]
curves
2 T,0
We were not able to experimentally determine the protonation
recorded at different pH values (Figure 2):
state of the reactants in R6; and although both HOSCN and
ꢀ
OSCN could act as nucleophiles toward (SCN) , the latter is
ꢀ
2
kobs ¼ ð1 ꢀ αÞk k K ½OH ꢁ½ðSCNÞ ꢁ
ꢀ2
6
ꢀ3
2 T, 0
presumably a much better nucleophile. It is improbable that only
!ꢀ1
ꢀ
the OSCN is reactive toward (SCN) , and we exclude this
ð1 ꢀ αÞk ½ðSCNÞ ꢁ
2
6
2 T, 0
ꢀ
ꢂ αk ½SCN ꢁ þ
possibility for the explanation of the absence of reaction R6 at pH
2
ꢀ
ꢀ
1
þ Kꢀ3½OH ꢁ þ K5½SCN ꢁ
0
.7. Instead, we propose that R6 becomes reversible, with
ꢀ
ꢀ
ꢀ2
(
SCN) as the favored product at low pH. This proposal is
2
ꢂ ð1 þ Kꢀ3½OH ꢁ þ K5½SCN ꢁÞ
supported by the observations made during the triple-mixing
quenching experiments. Elevated concentrations of (SCN)2
were observed after the acidic quench of the decomposition of
ð25Þ
þ
½
H ꢁ
α ¼
þ
K þ ½H ꢁ
1
(
SCN) at pH 4 (Figure S11), which can be accounted for by the
2
ꢀ
The experimental data set fits well to the proposed kinetic model,
and the obtained constants are listed in Tables 2 and S3.
In spite of the high quality of the fit, the limiting rate laws and
therefore the obtained parameters can be regarded only as
conversion of the steady-state intermediates HOSCN/OSCN
and NCS(dO)SCN to (SCN) at low pH. We assume that the
2
ꢀ
excess amount of OSCN after the basic quench was slightly less
than the excess amount of (SCN) after the acidic quench,
2
2
ꢀ
approximations as seen for example in the value of k which
because NCS(dO)SCN reacts to eventually give SO
CN (reactions R7 ꢀR10 ) in the highly basic solution.
and
ꢀ
3
4
1
0
ꢀ1 ꢀ1
ꢀ
0
0
exceeds 10
M
s . Although taking into consideration that
1
9917
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
Scheme 3. Mechanism of the Reaction of HOSCN with HCN
below pH 7
Biologically Relevant Reactions. Several important conclu-
sions can be drawn with respect to the biological relevance of the
results described herein. First, it is difficult to imagine (SCN)2
playing a significant role as a biological reactive intermediate at
neutral pH in an aqueous environment. The usual pathway to
ꢀ
produce (SCN) , comproportionation of HOSCN and SCN , is
2
14
ꢀ
magnitudes slower than the reaction of (SCN) with OSCN .
2
Thus, as soon as (SCN) is formed, it is already consumed
2
ꢀ
through its reaction with OSCN , which is a process that also
depletes the concentration of OSCN . Furthermore, if other
ꢀ
pathways exist for producing (SCN) , it is unlikely that (SCN)
2
2
plays a role in aqueous environments where the concentration of
OSCN is significant (as a consequence of the facile reaction
ꢀ
The production of (SCN) from HOSCN/OSCN is slower by
ꢀ
2
2
orders of magnitude compared to the production of HOSCN/
OSCN from (SCN) ; thus, the steady-state concentrations of
NCS(dO)SCN and (SCN) are practically zero during the
decomposition of HOSCN/OSCN in the region from pH 4
R6), especially in the mucosae where the steady-state concentra-
ꢀ
ꢀ
2
tion of OSCN can achieve high micromolar concentrations.
2
Second, we have shown that no mechanistic proposals which
ꢀ
involve (SCN) as a precursor can be feasible for the formation of
2
ꢀ
18,19,40ꢀ44
to pH 7. We note these measurements were only semiquantita-
tive and obtained with large experimental error.
OSCN at neutral pH
because these species are not in
ꢀ
dynamic equilibrium. Furthermore, OSCN and (SCN) cannot
2
ꢀ
Stability of the Intermediates. HO(SCN)2 was not ob-
coexist longer than a few milliseconds at neutral pH. Third, on
the basis of the kinetic data, we have confirmed that relatively
served during the previous investigations of the hydrolysis of
SCN) , and the comproportionation of HOSCN and SCN in
ꢀ
ꢀ
(
short-lived derivatives of OSCN are produced during the
2
14,15
ꢀ
ꢀ
the range from pH 0 to pH 2,
exist at pH < 3 as K 4 = 2.65 ꢂ 10 (pKa
because HO(SCN) cannot
decomposition of OSCN , including the product of the reaction
of (SCN) and OSCN , which we have formulated as NCS-
2
ꢀ5
(SCN)
2
ꢀ
= 4.58). The
ꢀ
2
ꢀ
intermediate HO(SCN)2 can only be a minor transient species
in the hydrolysis of (SCN) to give OSCN at pH > 8 because
R3 is a fast equilibrium, and ꢀR2 is shifted to the left with the
(dO)SCN on the basis of the results of ab initio MO calculations
ꢀ
2
and isotope tracer studies that favor an unsymmetrical inter-
19
ꢀ
mediate (vide supra). It is unlikely that the intermediate
deprotonation of HOSCN.
contains a symmetrical -SꢀOꢀS- linkage, as has been suggested
45
We propose that NCS(dO)SCN is formed over a wide range
in some early literature in the field. Fourth, we found that a high
ꢀ
ꢀ
of pH (0ꢀ8) where (SCN) and HOSCN or OSCN coexist for
relative concentration of OCN is produced during the decom-
2
ꢀ
some milliseconds, long enough for their reaction. However at
position of OSCN at neutral pH. This raises the question of
+
ꢀ
high [H ] and [SCN ], reaction R6 becomes reversible and
whether the observed MPO-catalyzed carbamylation of proteins
+
ꢀ
shifts toward the initial compounds. At lower [H ], the rapid
in sites of inflammation is caused by OCN as a direct product of
ꢀ
ꢀ
ꢀ
hydrolysis of NCS(dO)SCN is assumed. At high [OH ],
the enzymatic oxidation of SCN , or whether the OCN
concentration is elevated because of the decomposition of the
ꢀ
(
SCN) is quantitatively converted to HO(SCN)2 . Thus, the
2
ꢀ
46
relative rate of formation of NCS(dO)SCN becomes negligible
compared to the relative rate of formation of OSCN . On the
excess OSCN present.
ꢀ
basis of these considerations, no intermediates are expected to
be detectable with quench-flow/ESI-MS experiments, and none
were found.
’
CONCLUSIONS
The mechanism of decomposition of the human defense
Fast Steps and Stoichiometry. The proposed mechanism for
the hydrolysis of NCS(dO)SCN as seen in Schemes 1 and 2
factor HOSCN is a complicated multistepped process near
physiological pH. The lifetime of HOSCN decreases with
decreasing pH and with increasing SCN concentration.
0
0
35ꢀ37
ꢀ
(
R7 ꢀR10 ) is based on analogy with the halogens,
as no
kinetic data are available for these fast reactions that take place
after the rate-limiting steps. The product ratios determined by
the IC experiments (Table 1) differ significantly from the
predicted stoichiometry of the reactions as seen in Schemes 1
Furthermore the presence of (SCN) dramatically decreases
2
the availability of HOSCN as a consequence of their rapid
reaction with one another.
Some intermediates of the decomposition of HOSCN have
ꢀ
and 2. We propose that as [CN ] increases during decomposi-
already been described in the literature (e.g., (SCN) and
2
ꢀ
tion, its reaction with HOSCN becomes competitive among the
fast reactions (Scheme 3, R14 and R15). Reaction R14 has
(SCN)3 ), and some intermediates are proposed exclusively
on the basis of the analogy of the pseudo-halide SCN with the
halides (e.g., HO(SCN)2 , O SCN , O SCN ); the formation
of NCS(dO)SCN in the reaction of (SCN) with OSCN is
ꢀ
2
1
ꢀ
ꢀ
ꢀ
been studied in detail, and its pH-dependent rate is maximal
2 3
ꢀ
around pH 7, which is in agreement with the observed increase
2
ꢀ
in the OCN ratio. Table 1 also contains calculated product
proposed for the first time in this study on the basis of modeling
of the kinetics (from which the stoichiometry is obtained), ab
initio molecular orbital calculations (Supporting Information),
ratios indicating the contribution of R16 to the overall
ꢀ
ꢀ
decomposition of (SCN) and HOSCN/OSCN . At pH 7,
2
ꢀ
19
the SCN concentration is smaller and the OCN concentra-
tion is larger than anticipated, which is evidence for an
alternative hydrolysis pathway under these conditions. The
stoichiometry of the reactions at pH 7 is balanced with
inclusion of the production of elementary sulfur (which was
detected by ESI-MS); however, the mechanism of its forma-
tion is not clear.
and previously published isotope tracer studies (that indicate
the intermediate is unsymmetrical). The eventual stable products
of the decomposition of HOSCN at neutral pH are mainly simple
ꢀ
ꢀ
2ꢀ
2ꢀ
inorganic anions (e.g., OCN , SCN , SO , and SO4 ), some
3
of which can participate further in biochemical reactions.
Several processes described in the present study are highly
relevant in a biological setting. Evidence is provided that HOSCN
1
9918
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
cannot coexist with (SCN) at physiological pH and that they are
not in dynamic equilibrium with each other. Furthermore, pre-
vious mechanistic proposals for the formation of HOSCN in
CCl
4
into a solution of 0.1 M NaOH, where (SCN)
2
promptly
2
ꢀ
ꢀ 34,40
hydrolyzes to OSCN and SCN .
obtained when the resulting OSCN concentration of the aqueous
solutions did not exceed 4 mM. NaOSCN solutions were standardized
Quantitative 1:1 mixtures were
ꢀ
peroxidase systems that invoke (SCN) as an intermediate during
2
ꢀ
32,34,38
photometrically.
the enzymatic oxidation of SCN with H O cannot be correct,
2
2
because the product of the hydrolysis of (SCN) at neutral pH is
not HOSCN. On the basis of these conclusions, the biochemical
General Stopped-Flow and Quenched-Flow Methods.
Double-mixing stopped-flow experiments were performed using a Hi-Tech
SF-61 DX2 instrument equipped with both a photomultiplier tube
2
role of (SCN) is questionable. However, the possible roles of the
2
(PMT) for single-wavelength detection and a diode array for multiple
other reactive sulfur species that are produced during the
wavelength detection. A reference PMT is employed in this instrument.
Absorbance traces were collected using a Xe arc lamp and an optical cell
with a 1.00 cm path length. Triple-mixing stopped-flow and quenched-
flow experiments were carried out using a Bio-Logic SFM-400 instru-
ment. A single PMT, a Xe arc lamp, and a 1.00 cm flow-through optical
cell were used for absorbance measurement in stopped-flow mode. The
instrument can be converted for sample collection in quench-flow mode.
The dead time of both stopped-flow instruments were measured by the
reaction of 2,6-dichlorophenolꢀindophenol (DCIP) with ascorbic acid
decomposition of HOSCN remain to be investigated.
’
EXPERIMENTAL METHODS
Chemicals and Solutions. ACS reagent grade or better chemicals
were used without further purification. Ion exchanged and Millipore
ultrafiltered (Milli-Q synthesis A10) water was used to prepare all
solutions. NaOH solutions were standardized with potassium hydrogen
phthalate, and HClO
Acetate buffers (pH = 3.7ꢀ5.7) were made from NaOAc and HClO
and phosphate buffers (pH = 5.8ꢀ7.8) from KH PO and NaOH. The
ionic strengths of the solutions were established using NaClO . NaSCN,
4
solutions with the standardized NaOH solutions.
53
(
AA) under pseudo-first-order conditions with AA in excess. By
4
plotting the experimentally obtained kobs values against [AA], deviation
2
4
ꢀ
1
from linearity was observed above kobs = 600 s for the Hi-Tech
instrument, which means that higher kobs values can be determined only
with the elevation of relative error (see details in the Supporting
Information). The dead time of the Hi-Tech instrument is 1.35(3) ms
in double-mixing mode. For the Bio-Logic instrument in stopped-flow
mode, dead-time calibration was performed whenever switching to a new
mixing sequence. Calibration of the aging time was also necessary for the
Bio-Logic instrument in triple-mixing stopped-flow and quench-flow
mode for every mixing sequence and was carried out according to the
4
NaOAc, KH
ethylenediamine tetraacetic acid (Na
2
PO
4
, and NaClO
4
stock solutions contained 1 mM sodium
EDTA) to mask trace metal
2
impurities. An Orion Sure-Flow Ross combined-pH electrode attached
to an Orion expandable ion analyzer E920 was employed for all pH
measurements and an Irving-factor correction was applied. The
electrode was calibrated following IUPAC recommendations. For
UVꢀvis absorbance measurements an HP8452A diode array photo-
meter was used with a cell with either a 1.0, 5.0, or 10.0 cm path length.
4
7
4
8
ꢀ
recommendation of the manufacturer using the OH catalyzed hydro-
Preparation of NaOCl, (SCN) , and NaOSCN Solutions.
NaOCl stock solutions were prepared by sparging Cl gas (99.5+%)
2
lysis of 2,4-dinitrophenyl acetate. The experimentally determined and the
precalculated dead times and aging times were in excellent agreement
within 9% RSD. The temperature during all the stopped-flow and
quenched-flow measurements was maintained at 25.0 ( 0.1 °C using
Lauda RC-20 circulators equipped with chilled water heat exchangers.
Double Mixing Stopped-Flow Measurements (Hi-Tech).
2
into a 0.3 M solution of NaOH. The process was stopped at a NaOCl
concentration less than 0.10 M. The basic NaOCl solutions were
49
50
standardized by UVꢀvis spectroscopy and iodometric titration.
For stopped flow measurements, (SCN) solutions were generated
2
in situ in the first mixing cycle of the stopped-flow equipment by mixing
an acidic (0.2ꢀ0.4 M HClO
slightly acidified (pH = 5ꢀ6) solution of HOCl (0.5ꢀ10 mM).
reactions were complete within the mixing time. At least a 25-fold excess
of NaSCN had to be used to avoid the overoxidation of (SCN) by
HOCl and get quantitative yields of (SCN) . The highest concentration
of (SCN) obtained by this method was 5 mM which decomposed with a
4
) solution of NaSCN (0.2ꢀ1.2 M) with a
To study the kinetics of the decomposition of (SCN)
2
or HOSCN/
1
4,15
ꢀ
The
OSCN at a given pH, the in situ generated acidic or basic reagent
solution was mixed with the appropriate buffer in the second mixing
cycle, and the reaction was followed photometrically. The concentration
of the buffer was chosen to be at least in a 100-fold excess over the
2
2
ꢀ
reactant. The resulting pH was always measured independently. SCN
2
was used in high excess and for concentration dependence studies; its
concentration was varied in the NaSCN solution used for the generation
of the reagents. A 2.0 s aging time before the second mixing (τ ) gave the
1
2
half-life of some seconds. The analytical concentration of (SCN) was
determined photometrically, taking into account the formation of
ꢀ
ꢀ 33
(SCN)
3
in the presence of SCN . For analytical measurements, to
ꢀ
avoid the high SCN background in the samples, (SCN) was synthe-
sized in CCl
tially extracted into a 0.1 M solution of NaOH. (SCN)
standardized photometrically.
highest yield from the in situ prepared reagents and ensured the excellent
reproducibility of the experiments. The reaction of (SCN)2 with
2
34,51,52
4
by oxidizing Pb(SCN)
2
with Br
2
and was sequen-
in CCl was
ꢀ
HOSCN/OSCN at pH 1 and pH 4 was also studied by double-mixing
2
4
34,52
2
stopped-flow. The in situ-generated (SCN) (pH 0) was reacted with
Solutions of NaOSCN were prepared using three different methods:
1) NaOSCN solutions were generated in situ for stopped-flow mea-
freshly prepared solutions of NaOSCN (pH 13) to produce a reaction
mixture with a pH between 0.58 and 0.70. For experiments around pH 4,
the solution of NaOSCN was prepared with a sufficient amount of
NaOAc for the resulting mixture to be buffered. NaOSCN was
synthesized either by the extraction method or by mixing NaSCN and
NaOCl solutions with a hand mixer. The NaOSCN solutions were used
within 2 min after synthesis, and a fresh solution was prepared for each
experiment. Reference experiments were done without using hypochlor-
(
surements in the first mixing cycle of the instruments by mixing a
solution of 0.2ꢀ1.5 M NaSCN in 0.2ꢀ0.4 M NaOH with a solution of
3
2,34,38
0
.5ꢀ15 mM NaOCl from pH 11 to pH 12.
The reaction was
complete within 2 s, and at least a 50-fold excess of NaSCN was required
for a quantitative yield of OSCN . The maximum concentration of
NaOSCN obtained was 7 mM, which was stable for several minutes.
ꢀ
3
8
ite or (SCN) in CCl during the preparation of one or the other
(
2) Solutions of NaOSCN were generated at pH 7 using the lactoper-
2
4
34,40
oxidase system (LPO/NaSCN/H
excess of SCN was undesirable. For quantitative yield, a 4-fold excess of
SCN over H O2 was required. The maximum concentration of
OSCN obtained by this method was 4 mM.
2
O
2
)
when the presence of a high
reactant.
ꢀ
Triple-Mixing Stopped-Flow and Quench-Flow Measure-
ments (Bio-Logic). Triple-mixing experiments were performed to
quench the decomposition of the studied sulfur species and to create an
ꢀ
2
ꢀ
34,40
The enzyme was
not removed from the system before further experiments. (3) Solutions
opportunity for quantitative analysis of the reaction mixtures. (SCN)
OSCN was generated in situ and allowed to decompose at the desired
2
or
ꢀ
ꢀ
ꢀ
with 1:1 SCN to OSCN ratio were prepared by extracting (SCN) in
2
1
9919
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
pH (after mixing with a buffer) for τ aging times; the reaction was
ARTICLE
2
solution to a final pH of 13 before analysis. Background samples were
prepared from the buffers.
quenched either with HClO (jump to pH 0) or with NaOH (jump to
4
pH 13). The samples were analyzed either in situ by UVꢀvis (stopped-
flow mode) or externally (quench-flow mode). Reference experiments
were performed in double-mixing mode with in situ UVꢀvis detection.
Electrospray Ionization Mass Spectrometry. ESI-MS mea-
surements were performed with an Agilent 6538 UHD Accurate-Mass
Q-TOF MS with a dual ESI ion source. A 5 μL aliquot of each sample
solution was introduced via loop injection with a 0.2 mL/min methanol
stream. The temperature of the drying gas was 180 °C. The voltage
applied on the ion source was 5000 V. Spectra were recorded at 2 GHz
and averaged by Masshunter B.03.01 software from Agilent. The m/z
scale was calibrated externally using Agilent ESI tune mix (6 points) and
corrected internally with Agilent ESI reference mass solution (2 points).
The error of the m/z measurement was less than 7.0 ppm in positive
mode and less than 6.0 ppm in negative mode. The aim of the MS
Data Treatment. The raw data sets generated in the various types
of measurements were treated with the instrument-controlling software.
Further manipulation, when necessary, was done by MS Excel 2003
(Microsoft). Specfit/32 (Spectrum Software Associates) was used for
the evaluation of polychromatic UVꢀvis data. LevenbergꢀMarquard
least-squares fits and kinetic data simulations were performed with
Scientist 2.0 (Micromath).
’
ASSOCIATED CONTENT
S
Supporting Information. Experimental and computa-
b
tional details, equations used for global fits, schemes showing
the Bio-Logic SFM-400 flow chart and stopped-flow parameters
of the Hi-Tech SF-61 DX2, figures displaying the concentration
and pH dependences of the measured values of kobs during the
experiments was to identify the intermediates and products of the
ꢀ
ꢀ
decomposition of (SCN) and HOSCN/OSCN in various pH values.
2
decomposition of HOSCN and (SCN) , UV-vis spectra of
2
For product identification (SCN)
2
and OSCN solutions were gener-
ꢀ
HOSCN and OSCN , results of triple-mixing experiments to
ated in situ in the Bio-Logic instrument, and after mixing with the
appropriate buffer they were collected in quenched-flow mode and
allowed to decompose for 5 min. Intermediate identification was
achieved using triple-mixing quench-flow experiments (acidic and basic)
quench decomposition of (SCN) , experiments for mixing
2
(
SCN) with HOSCN, details of the molecular orbital calcula-
2
tions and a reaction coordinate diagram for the interconversion
of NCSꢀOꢀSCN and NCS(dO)ꢀSCN, a table identifying ESI-
MS peaks, a table with the empirical rate constants for the
experimentally determined rate laws, a table of computed molar
absorptivities obtained from a global fit, and a table of computed
total energies for the two ground states and one transition state
2
that were performed with aging times τ = 0.5t1/2, 1t1/2, and 2t1/2, which
allowed the construction of reaction-time-resolved MS spectra. Approxi-
mately 5 min elapsed between sample preparation and the measure-
ments. All samples were diluted 1/10ꢀ1/50 before introduction to the
MS, to avoid complication from the high ionic strength. All samples were
analyzed both in negative and positive ion mode. The process of
compound identification and more details on the measurements can
be found in the Supporting Information.
structures of O(SCN) . This material is available free of charge
2
via the Internet at http://pubs.acs.org.
’
AUTHOR INFORMATION
Ion Chromatography. Ion chromatography (IC) was performed
on a Dionex ICS-3000 instrument. The sample holder tray, column,
detector, and cell were thermostatted to 20, 25, 30, and 35 °C,
Corresponding Author
MAshby@ou.edu
ꢀ
ꢀ
respectively. The injection loop was 10 μL. CN and SCN were
ꢀ
2ꢀ
measured using integrated amperometry detection, and OCN , SO
3
,
2ꢀ
and SO
4
were measured using conductivity detection. An Ionpac AS16
’
ACKNOWLEDGMENT
column was used for the anions measured, using integrated amperometry
detection, and the eluents were as follows: from ꢀ5 (preinjection
equilibrium) to +10 min, isocratic 6.25 mM NaOH; from 10 to 25
min, isocratic 60 mM NaOH. An Ionpac AS18 column and an ASRS
This research was supported by the National Science Founda-
tion (Grants CHE-0503984 and CHE-0911328) and the Oklahoma
Center for the Advancement of Science and Technology (Grant
HR08-003). The authors are grateful to Dr. G ꢀa bor Lente for
critically reading the manuscript.
2
mm suppressor were used for ions detected by conductivity, and
isocratic 12.5 mM NaOH was employed for 30 min. Each sample was
injected three times to ensure reproducibility. Calibration of the methods
was performed usinga series of standard solutions. The IC measurements
were performed to quantify the anions produced during the decomposi-
tion of the investigated sulfur species at different pH values. Because
’ REFERENCES
(
(
1) Ashby, M. T. J. Dent. Res. 2008, 87, 900.
2) Hawkins, C. L. Free Radical Res. 2009, 43, 1147.
ꢀ
SCN is one of the expected products, synthetic methods which require
(3) Reiter, B.; Harnulv, G. J. Food Prot. 1984, 47, 724.
(4) Pruitt, K. M.; Tenovuo, J. O., Eds. The Lactoperoxidase System:
Chemistry and Biological Significance, Immunology Series, Vol. 27;
Dekker: New York, 1985.
ꢀ
an initial large excess of SCN were not feasible for these experiments.
(SCN)
2 4
was synthesized in CCl and extracted directly to a 50 mM
acetate or phosphate buffer, to a concentration of lessthan 5 mM (SCN)
2
(
5) Ihalin, R.; Loimaranta, V.; Tenovuo, J. Arch. Biochem. Biophys.
006, 445, 261.
6) Pruitt, K. M. J. Oral Pathol. 1987, 16, 417.
(7) Boots, J. W.; Floris, R. Int. Dairy J. 2006, 16, 1272.
8) Seifu, E.; Buys, E. M.; Donkin, E. F. Trends Food Sci. Technol.
2005, 16, 137.
in the aqueous phase and was allowed to decompose for 15 min before
ꢀ
2
the analysis. OSCN was produced by extracting (SCN) synthesized in
2
(
CCl to a 0.1 M NaOH solution, which was immediately mixed with a
4
ꢀ
phosphate or acetate buffer during constant stirring. OSCN was
(
alternatively prepared by using the LPO system in the region from pH
ꢀ
6.5 to pH 7.3 in phosphate buffer. The resulting HOSCN/OSCN
(
(
(
(
9) Pruitt, K. Advanced Dairy Chemistry, 3rd ed.; 2003; Vol. 1, p 563.
10) Naidu, A. S. Nat. Food Antimicrob. Syst. 2000, 103.
11) Wolfson, L. M.; Sumner, S. S. J. Food Prot. 1993, 56, 887.
12) Wang, J.-G.; Mahmud, S. A.; Nguyen, J.; Slungaard, A. J. Immunol.
solutions, regardless of the method of synthesis, contained 50 mM buffer
ꢀ
and a maximum of 2 mM reactant. Solutions of HOSCN/OSCN were
allowed to completely decompose for 30 min prior to analysis. To study
ꢀ
the effect of initial SCN excess on the decomposition products which
2
006, 177, 8714.
was present during all the kinetic studies, additional experiments were
(13) Wang, J.-G.; Mahmud, S. A.; Thompson, J. A.; Geng, J.-G.; Key,
N. S.; Slungaard, A. Blood 2006, 107, 558.
ꢀ
performed with buffer solutions containing 8- to 11-fold SCN excess
over the reactants. All samples were diluted to 10-fold by a NaOH
(14) Nagy, P.; Lemma, K.; Ashby, M. T. Inorg. Chem. 2007, 46, 285.
1
9920
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921

Journal of the American Chemical Society
ARTICLE
(
15) Barnett, J. J.; McKee, M. L.; Stanbury, D. M. Inorg. Chem. 2004,
3, 5021.
16) Rayson, M. S.; Mackie, J. C.; Kennedy, E. M.; Dlugogorski, B. Z.
Inorg. Chem. 2011, 50, 7440.
4
(
(
(
(
(
(
(
17) Mamman, S.; Iyun, J. F. Int. J. Pure Appl. Chem. 2007, 2, 429.
18) Aune, T. M.; Thomas, E. L. Eur. J. Biochem. 1977, 80, 209.
19) Thomas, E. L. Biochemistry 1981, 20, 3273.
20) Walden, P.; Audrieth, L. F. Chem. Rev. 1928, 5, 339.
21) Lemma, K.; Ashby, M. T. Chem. Res. Toxicol. 2009, 22, 1622.
22) Kornath, A.; Blecher, O.; Ludwig, R. J. Am. Chem. Soc. 1999,
121, 4019.
(
(
(
23) Kornath, A.; Blecher, O. Z. Anorg. Allg. Chem. 2002, 628, 625.
24) Jander, G.; Gruttner, B.; Scholz, G. Chem. Ber. 1947, 80, 279.
25) Okulik, N. B.; Jubert, A. H.; Castro, E. A. Adv. Quantum Chem.
Bonding Struct. 2008, 119.
26) Okulik, N. B.; Jubert, A. H.; Castro, E. A. J. Struct. Chem. 2008,
9, 922.
27) Boehland, H.; Golub, A. M.; Koehler, H.; Lisko, T. P.; Samoilenko,
(
4
(
V. M.; Skopenko, V. V.; Cincadze, G. V. Chemistry of Pseudohalides; Dr.
Alfred Huethig Verlag: Heidelberg, Germany, 1979.
(
28) Lengyel, I.; Epstein, I. R.; Kustin, K. Inorg. Chem. 1993, 32,
5880.
(
(
29) Palmer, D. A.; Van Eldik, R. Inorg. Chem. 1986, 25, 928.
30) Truesdale, V. W.; Luther, G. W.; Greenwood, J. E. Phys. Chem.
Chem. Phys. 2003, 5, 3428.
31) Nagy, P.; Jameson, G. N. L.; Winterbourn, C. C. Chem. Res.
Toxicol. 2009, 22, 1833.
32) Ashby, M. T.; Carlson, A. C.; Scott, M. J. J. Am. Chem. Soc. 2004,
26, 15976.
(
(
1
(33) Barnett, J. J.; Stanbury, D. M. Inorg. Chem. 2002, 41, 164.
(34) Nagy, P.; Alguindigue, S. S.; Ashby, M. T. Biochemistry 2006,
45, 12610.
(
(
(
35) Figlar, J. N.; Stanbury, D. M. J. Phys. Chem. A 1999, 103, 5732.
36) Figlar, J. N.; Stanbury, D. M. Inorg. Chem. 2000, 39, 5089.
37) Nagy, P.; Beal, J. L.; Ashby, M. T. Chem. Res. Toxicol. 2006,
19, 587.
(
38) Nagy, P.; Wang, X.; Lemma, K.; Ashby, M. T. J. Am. Chem. Soc.
2
007, 129, 15756.
(
(
39) Skaff, O.; Pattison, D. I.; Davies, M. J. Biochem. J. 2009, 422, 111.
40) Aune, T. M.; Thomas, E. L.; Morrison, M. Biochemistry 1977,
1
6, 4611.
41) van Dalen, C. J.; Whitehouse, M. W.; Winterbourn, C. C.;
Kettle, A. J. Biochem. J. 1997, 327, 487.
(
(42) Van Dalen, C. J.; Kettle, A. J. Biochem. J. 2001, 358, 233.
(43) Slungaard, A.; Mahoney, J. R., Jr. J. Biol. Chem. 1991, 266, 4903.
(44) Das, D.; De, P. K.; Banerjee, R. K. Biochem. J. 1995, 305, 59.
(45) Pollock, J. R.; Goff, H. M. Biochim. Biophys. Acta 1992, 1159, 279.
(46) Wang, Z.; Nicholls, S. J.; Rodriguez, E. R.; Kummu, O.; H €o rkk €o ,
S.; Barnard, J.; Reynolds, W. F.; Topol, E. J.; DiDonato, J. A.; Hazen, S. L.
Nature Med. 2007, 13, 1176.
(47) Irving, H. M.; Miles, M. G.; Pettit, L. D. Anal. Chim. Acta 1967,
3
1
1
1
8, 475.
(
48) Covington, A. K.; Bates, R. G.; Durst, R. A. Pure Appl. Chem.
983, 55, 1467.
(
49) Adam, L. C.; Fabian, I.; Suzuki, K.; Gordon, G. Inorg. Chem.
992, 31, 3534.
50) Gordon, G.; Pacey, G. E.; Cooper, W. J. Water Chlorination
990, 6, 29.
(
(51) Wood, J. L. Org. React. (New York, NY, United States) 1946, 240.
(52) Lemma, K.; Ashby, M. T. J. Org. Chem. 2008, 73, 3017.
(53) Tonomura, B.; Nakatani, H.; Ohnishi, M.; Yamaguchi-Ito, J.;
Hiromi, K. Anal. Biochem. 1978, 84, 370.
1
9921
dx.doi.org/10.1021/ja2083152 |J. Am. Chem. Soc. 2011, 133, 19911–19921