Elimination mechanisms in the aminolysis of sulfamate esters of the type NH2SO2OC6H2X - Models of enzyme inhibitors

Article abstract of DOI:10.1002/ejoc.200800366

The kinetics of the reaction of 4-nitrophenyl sulfamate NH ₂SO₂OC₆H₄NO₂-4 (1a) in acetonitrile (ACN) with a series of pyridines (pK_a range ca. 8 units) and alicyclic amines (pK_a range ca. 3.6 units) has been studied in the presence of excess amine at various temperatures. The compounds 1a-1f are important as model substrates for the medicinally important sulfamate esters 667-coumate and emate and analogues. Pseudo-first-order rate constants (k _obsd.) have been obtained mainly by the release of 4-nitrophenol/4-nitrophenoxide. Slopes of plots of k_obsd. vs. [amine] gave second-order rate constants (k₂), and Broensted plots were biphasic for the aminolysis (with alicyclic amines) with an initial slope β₁ = 0.53 and a subsequent slope β₂ = 0.19. The change in slope occurs near the first pK_a of 1a (17.9) in ACN. Leaving-group effects were probed by using the same series of phenyl sulfamates, i.e. 1a-f and the alicyclic amines N-formylpiperazine and pyrrolidine. The reactions were considered to be dissociative in nature involving E2- and E1cB- type mechanisms with the phenyl sulfamate anion 2 being involved in pyridine and in the weaker alicyclic amines (β₁ segment) and a phenyl sulfamate dianion 3 being involved with the stronger alicyclic bases (β₂ segment). The calculation of Leffler indices (α) for bond-forming (base...H⁺) and bond-breaking (S-OAr) steps allows fuller interpretation of the mechanisms occurring, which are seen as having the N-sulfonylamines, HN=SO₂ and ^-N=SO₂ on the reaction pathways leading to products. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejoc.200800366

FULL PAPER
DOI: 10.1002/ejoc.200800366
Elimination Mechanisms in the Aminolysis of Sulfamate Esters of the Type
NH2SO2OC6H4X – Models of Enzyme Inhibitors
William J. Spillane,*[a] Andrew O’Byrne,[a] and Cheryl J. A. McCaw[a]
Keywords: Elimination / Enzyme models / Kinetics / Inhibitors
The kinetics of the reaction of 4-nitrophenyl sulfamate
NH2SO2OC6H4NO2-4 (1a) in acetonitrile (ACN) with a series
of pyridines (pKa range ca. 8 units) and alicyclic amines (pKa
range ca. 3.6 units) has been studied in the presence of ex-
cess amine at various temperatures. The compounds 1a–1f
are important as model substrates for the medicinally impor-
tant sulfamate esters 667-coumate and emate and analogues.
Pseudo-first-order rate constants (kobsd.) have been obtained
mainly by the release of 4-nitrophenol/4-nitrophenoxide.
Slopes of plots of kobsd. vs. [amine] gave second-order rate
constants (k2), and Brönsted plots were biphasic for the ami-
nolysis (with alicyclic amines) with an initial slope β1 = 0.53
and a subsequent slope β2 = 0.19. The change in slope occurs
near the first pKa of 1a (17.9) in ACN. Leaving-group effects
were probed by using the same series of phenyl sulfamates,
i.e. 1a–f and the alicyclic amines N-formylpiperazine and
pyrrolidine. The reactions were considered to be dissociative
in nature involving E2- and E1cB- type mechanisms with the
phenyl sulfamate anion 2 being involved in pyridine and in
the weaker alicyclic amines (β1 segment) and a phenyl sulf-
amate dianion 3 being involved with the stronger alicyclic
bases (β2 segment). The calculation of Leffler indices (α) for
bond-forming (base···H+) and bond-breaking (S–OAr) steps
allows fuller interpretation of the mechanisms occurring,
which are seen as having the N-sulfonylamines, HN=SO2
–
and N=SO2 on the reaction pathways leading to products.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
Generally, the active esters are esters of sulfamic acid,
i.e. they contain the NH2SO2O entity. The subject has been
reviewed over the years,[3,4] and several quite recent ac-
counts have appeared.[5–7] These esters act by inhibiting or
blocking enzymatic pathways in the body, and it is thought
that they sulfamoylate essential amino acid residues in the
enzyme, thereby rendering it inactive. Over much the same
period of time, we have been investigating the hydrolysis
and aminolysis mechanisms of various esters such as 4-ni-
trophenyl methylsulfamate, 4-nitrophenyl benzylsulfamate
and a series of 4-nitrophenyl phenylsulfamates.[8,9] Latter
studies have focussed more on 4-nitrophenyl sulfamate
NH2SO2OC6H4-NO2-4 (1a), because this compound could
act as a simple model for the more complex and therapeuti-
cally important esters.[10] The thrust of this current work
is to examine the kinetics and mechanism of reaction in
acetonitrile (ACN) of compound 1a and five other phenyl
sulfamate derivatives, 1b–f.
Introduction
There has been a renaissance of interest in sulfamate es-
ters RNHSO2ORЈ over approximately the last 12 years be-
cause of their potential in medicinal chemistry where cer-
tain types of esters, particularly those of the type
NH2SO2OR have been showing considerable promise in a
wide variety of applications. To take a few important exam-
ples, some sulfamates inhibit steroidal sulfatases (STS) and
carbonic anhydrase (CA), and thus the synthesis of those
steroids that encourage the growth of hormone-dependent
cancers like breast and prostate cancer and other diseases[1]
can be blocked. The results of phase I clinical trials on 667-
coumate, an STS inhibitor, which shows promise in the
treatment of breast cancer have appeared recently.[2]
Results and Discussion
pKa Studies
In order to understand more fully the mechanism of ami-
nolysis of these sulfamate esters, pKa measurements in ACN
at 25 °C for compound 1a for the equilibria shown in Equa-
tions (1) and (2) were first carried out (Scheme 1). Com-
[a] Chemistry Department, National University of Ireland,
Galway, Ireland
E-mail: william.spillane@nuigalway.ie
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
4200
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4200–4205

Aminolysis of Sulfamate Esters
pound 1a was the substrate mainly used in this study. The Table 1. log k2 values for pyridinolysis at 37 °C in ACN of 1a and
values of pKa and Hammett σ used.
pKa for the ionization of 1a was found to be 17.9, in excel-
[a]
[b]
lent agreement with a value of 17.8 previously determined
in this laboratory.[10] The pKa value for the second ioniza-
tion gave an average value of 21.08 in ACN on the basis of
three determinations by using two different analytical wave- 4-Dimethylamino
2-Amino-4-methyl
lengths.
Pyridine
log k2
pKa
σpyr
4-Pyrrolidino
4-Amino
–1.06
–1.75
–1.26
–2.32
–1.92
–2.17
–3.07
–2.87
–3.38
18.2
17.2
17.6
15.1
14.7
14.3
13.9
12.3
10.5
–0.90
–0.57
–0.83
–0.41
–0.57
–0.27
–0.33
0
3,4-Diamino
2-Amino
2,4,6-Trimethyl
Pyridine
3-Chloro
0.37
[a] Measured in ACN at 25 °C.[8,9] [b] Most of the σ values were
available in: D. D. Perrin, B. Dempsey, E. P. Sergeant, pKa Predic-
tion for Organic Acids and Bases, Chapman & Hall, London, 1981;
the σ value of –0.90 for 4-pyrrolidine and the σ0 values of –0.27
and –0.13 for the 2-NH2 and 2-Me compounds, respectively, were
obtained from: C. Hansch, A. Leo, D. Hoekman, Exploring
QSAR – Hydrophobic, Electronic and Steric Constants, American
Chemical Society, Washington, DC, 1995.
Scheme 1. Ionizations of phenyl sulfamate esters.
In the second ionization [Equation (2)], a higher pKa (less
acidic) would be expected because of the insulating effect
of the negative nitrogen centre now present and thus greater
difficulty in removing protons.
Brönsted (β), Hammett (ρ) and Thermodynamic Studies
The rate law followed for the aminolysis of the sulfamate
esters used in this work had the form:
d[XC6H4OH or XC6H4O–]/dt = –d[sulfamate ester]/dt =
kobsd.[sulfamate ester]
Figure 1. Brönsted-type plots for the pyridinolysis of 1a (black dia-
monds), the bases left to right being 3-chloropyridine, pyridine, 2-
aminopyridine, 3,4-diaminopyridine, 2-amino-4-methylpyridine, 4-
aminopyridine, 4-(dimethylamino)pyridine and 4-pyrrolidinopyrid-
ine, and for the aminolysis (alicyclic amines) of 1a (squares), the
bases left to right being thiomorpholine, morpholine, N-formyl-
piperazine, N-(2-hydroxyethyl)piperazine, N-(2-aminoethyl)pipera-
zine, piperazine, piperidine, and pyrrolidine in ACN at 37 °C. The
usual correction has been applied for piperazine by using p = 1 and
q = 2 and both N-(2-aminoethyl)piperazine and piperazine having
the same pKa values and virtually identical rates appear together
in the plot.
where kobsd. = k0 + k2·[amine]. The k0 terms for uncatalyzed
reactions were found to be negligible in this work carried
out under pseudo-first-order conditions, i.e., a large excess
of amine present relative to the amount of substrate. The
k2 values are the second-order rate constants for the ami-
nolysis reaction. Two types of bases were employed: firstly
a set of pyridines with a pKa range in ACN of ca. 8 units,
and secondly a set of alicyclic amines with a range spanning
ca. 3.6 pKa units. The second-order rate constants for the
pyridinolysis and aminolysis (alicyclic bases) of 1a are given
in Table 1 together with the pyridine pKa values and the
Hammett σ values used. Brönsted-type plots of log k2 steric effects. The data for the alicyclic amines has been
against amine pKa values are shown in Figure 1 for the pyr- published previously.[10]
idinolysis and for the aminolysis of 1a.
These β values indicate a moderate transfer of proton
The Brönsted β value for the pyridinolysis plot is 0.28 (r from substrate nitrogen atom to pyridine in the reaction.
= 0.970). This value is very close to those previously ob- From previous examination of the reactions of various sul-
tained when a series of 12 pyridines were treated with 1a in famate esters in ACN and other organic solvents the pyrid-
chloroform.[11] Because pyridine pKa values are not avail- inolysis is likely to involve an E2 mechanism with reasonable
able in CHCl3, sets of pKa values for water and ACN were N–H cleavage in the substrate, and the aminolysis (alicyclic
used instead. Plots of log k2 against pyridine pKa values (in amines) for the Brönsted β1 and β2 segments will follow
water) gave a slope of 0.23 (r = 0.930), and against pyridine E2-/E1cB-type mechanisms. The plot for the aminolysis
pKa values (in ACN) a slope of 0.28 (r = 0.920) was ob- with alicyclic bases could be regarded as presenting a steady
[11]
tained.
The point for 2,4,6-trimethypyridine is not plot- change (continuum) or two straight lines with slopes β1 and
ted in Figure 1, because it deviates well below the line. This β2 of 0.53 (r = 0.981, 5 points) and 0.32 (r = 0.982, 3
is not surprising, because being substituted in both ortho points), respectively. Opting for the smooth continuous
positions it may be expected to deviate somewhat due to curve also gives a good fit to a suitable quadratic equation
Eur. J. Org. Chem. 2008, 4200–4205
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4201

W. J. Spillane, A. O’Byrne, C. J. A. McCaw
FULL PAPER
Table 2. Activation parameters[a] for the aminolysis by alicyclic amines of 1a in ACN.
Alicyclic amine[b]
∆HϮ/kJmol–1 ∆SϮ/Jmol–1K–1 Alicyclic amine[b]
∆HϮ/kJmol–1
∆SϮ/Jmol–1K–1
Thiomorpholine (16.2)
Morpholine (16.6)
N-Formylpiperazine (17.0)
N-(2-Hydroxy)piperazine (17.6)
47.5Ϯ5
42.9Ϯ4
61.2Ϯ6
59.5Ϯ5
–128.2Ϯ9
–138.0Ϯ10
–75.8Ϯ6
N-(2-aminoethyl)piperazine (17.9)
piperazine (18.2)
piperidine (18.8)
49.2Ϯ5
53.4Ϯ5
49.9Ϯ5
60.1Ϯ6
–105.7Ϯ8
–87.3Ϯ7
–98.5Ϯ8
–59.5Ϯ6
–73.5Ϯ6
pyrrolidine (19.8)
[a] Eyring plots (correlation coefficients were Ͼ 0.99 in all cases) were generated for each amine by using rate constants at between four
and nine different temperatures and covering temperature ranges of usually 25+ °C. Thiomorpholine, which reacts very slowly, was studied
over a 16 °C temperature range. The errors shown are standard deviations.. [b] The numbers in parentheses are the pKa values in ACN
for the amines.
(y = –0.0668x2 + 2.8075x –29.815) with r = 0.995. However, tial proton removal from the substrates to generate dianion
because biphasic Brönsted plots have been found before 3a. By using the log k2 and σpyr data (Table 2) a Hammett
with some sulfamate esters,[8] and because the entropy ρ value of –1.82 (r = 0.97) was found, and this would indi-
changes measured (Table 2) do not display a continuous cate that the pyridines have removed the acidic NH hydro-
change, but show firstly an increase (become less negative) gen atom partially. Substantial removal of the proton would
with a change in base, and at approximately a pKa value of lead to a much larger negative ρ value of the order of –5 to
18 again become quite negative and start to increase for a –6 because the ρ value for the ionization of the pyridinium
second time, we regard the plot as being most likely bi- ions of the bases in Table 1 (pKa against σpyr plot) is –6.1 (r
phasic. With other biphasic plots involving, for example, 4- = 0.963). The Brönsted β value of 0.28 for the pyridinolysis
nitrophenyl benzylsulfamate in ACN with eight strong supplies the same information and corroborates this point.
bases whose pKa values cover a range of 4–5 units and in All the alicyclic bases used were of the required “same
particular straddle pKa ≈ 18, we have observed very similar structural type” possessing a secondary –NH– moiety
behaviour in entropy changes for the reaction, and for this within a saturated six- or five-membered (in the case of pyr-
substrate the entropy increases from –112 Jmol–1 K–1to rolidine) ring system, however, one of them (octahydro-
–55 Jmol–1 K–1, and then at a point corresponding approxi- indole surprisingly), like 2,4,6-trimethylpyridine, deviated
mately to the pKa value of the substrate the entropy de- considerably from the plot and is not included. Its slower
creases to –114 Jmol–1 K–1 and then increases again to reaction may be due to a substantial steric effect caused by
–15 Jmol–1 K–1.[8] When the Brönsted plot is a straight line, the attached six-membered ring, and it cannot be used with
for example, with 4-nitrophenyl sulfamate and a set of pyr- the other bases.
idines covering a pKa range of ca. 7 units, the entropy dis-
plays a steady increase moving to the weaker bases. The
Leaving Group Effects
activation study in this work was carried out by using eight
of the nine alicyclic amines shown in Figure 1. The ∆HϮ
and ∆SϮ values determined from Eyring and Arrhenius
plots are given in Table 2 for the aminolysis of 1a.
To shed further light on the mechanism occurring in
these aminolysis reactions, leaving group effects were
probed in the “β1 segment” and in the “β2 segment” of the
alicyclic amine plots in Figure 1. The substrates 1a–f were
treated with N-formylpiperazine, which falls about midway
on the β1 plot, and with pyrrolidine, which is the last point
on the β2 plot. By using the log k2 and pKa data (Table 3)
Brönsted βlg (N-formylpiperazine) and βlg (pyrrolidine) val-
ues of –0.43 (r = 0.992) and –0.66 (r = 0.984), respectively,
were determined and the plots are drawn (Figure 2). The
The value for β1 points to substantial proton removal
during the reaction, but the value for β2 indicates that con-
siderably less protons are now transferred to the stronger
bases. This arises, because it is more difficult to remove the
second proton from 2a than it was to remove the first pro-
ton from 1a. It may be noted that the changeover from one
plot to the other seems to be occurring at around the first
pKa value of 1a (17.9). Around pKa ≈ 18 the second proton
is being removed by the stronger bases, and the reactive
species then becomes the dianion 3a. Above the second pKa
value of 1a (21.1) corresponding to the ionization 2a i 3a,
eventually the dianion 3a is fully formed, and general catal-
ysis by bases should cease at that point, and a straight line
parallel to the x-axis would be expected. Then the plot for
the alicyclic bases would be triphasic (Figure 1). The
strongest alicyclic base available in this study was pyrroli-
dine (pKa = 19.8), and thus we were unable to demonstrate
this predicted levelling off of the Brönsted plot.
Table 3. log k2 values for the aminolysis (pyrrolidine and N-formyl-
piperazine) at 41.5 °C in ACN of a series of 4-substituted phenyl
sulfamate esters and values of pKa and Hammett σ used.
[a]
log k2
Leaving group X in pKa
σ
Pyrrolidine N-Formylpiperazine H2NSO2OC6H4X-4
–5.00
–4.52
–4.00
–3.00
–2.92
–0.324
–4.16
Me
H
F
Br
Cl
27.45 –0.17
27.2
26.57 0.062
25.53 0.227
25.44 0.232
20.7 0.778
[b]
0.0
–3.40
–3.05
–3.05
–1.11
The plot in Figure 1 for the pyridinolysis does not dis-
play biphasic behaviour and neither do the other plots show
such behaviour. This is due to the fact that even the strong-
NO2
[a] pKa value in ACN of the leaving phenol. [b] Substantial overlap
between the absorbances of the amine and the leaving phenol pre-
est pyridine bases are not basic enough to achieve substan- cluded rate measurements.
4202
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4200–4205

Aminolysis of Sulfamate Esters
same log k2 values and the appropriate σ values (Table 3) 3.66, respectively.[17] These values relate to the measurement
gave Hammett ρlg values for N-formylpiperazine of 3.2 (r = in water and are therefore lower than the values obtained
0.995) and for pyrrolidine of 5.1 (r = 0.996).
in ACN, but nevertheless they also point to substantial
negative charge buildup in the transition states for the Paths
A and B.
Figure 2. Brönsted leaving-group plot for aminolysis at 41.5 °C in
ACN with N-formylpiperazine (black diamonds) the substrates
from left to right being 1a, 1c, 1d, 1b and 1e, and with pyrrolidine
(squares) the substrates from left to right being 1a, 1c, 1d, 1b, 1e
and 1f. The pKa values in ACN of the corresponding phenolic leav-
ing groups[12] are used in the plot. The value for 4-fluorophenol in
ACN was estimated from a plot of the pKa values in water against
ten pKa values available in ACN.[12]
Figure 3. Hammett leaving-group plot for the aminolysis at 41.5 °C
in ACN with N-formylpiperazine (black diamonds) the substrates
from left to right being 1e, 1b, 1d, 1c and 1a, and with pyrrolidine
(squares) the substrates from left to right being 1f, 1e, 1b, 1d, 1c
and 1a.
A likely mechanistic scenario for the reactions outlined
above of 1a, its anion 2a, and the dianion 3a would be the
scheme proposed by Thea and Williams[13] some years ago
for the reaction in water (Scheme 2). In this two major
Mechanism of Aminolysis
A substitution mechanism where the amines act not as pathways involving an N-sulfonylamine 4 and a unique an-
bases but as nucleophiles attacking the sulfur atom leading ionic sulfonylamine 5 leading to identical products were
to a pentavalent intermediate can be ruled out under the proposed and supported with kinetic evidence. In ACN in
present circumstances. This is because the NH proton is the presence of moderate bases a mechanism (Path A) in-
quite acidic (ca. 8 in aqueous organic media[13]) and the volving the anions 2 and N-sulfonylamine (4) seems prob-
aryl oxides are good leaving groups. However, an SN-type able, and with the stronger bases (pKa Ͼ ca. 18) the mecha-
reaction is very likely in certain situations (i) for those esters nism is likely to involve the dianion 3, albeit in very low
that do not possess an acidic hydrogen atom, e.g. Me2N- concentration, because the pKa value of 2a is ca. 21.1, and
SO2Oar, because they have to choose an associative the anionic sulfonylamine 5 (Path B).
route;[9,14] and (ii) current work has shown that at pH = 2
the neutral form of the esters 1a–1f react by nucleophilic (α) for the bond-forming (base···H+) and bond-breaking
attack by water in a bimolecular TS.[15]
(S···OAr) steps is possible if β1 and βlg are added to give
For this present work calculation of the Leffler indices
The two values of βlg (–0.43 and –0.66) for the two βeq, and β2 and βlg are added to give βeq. These calculations
amines used (Figure 2) would indicate a reasonable degree give further insights into the reactions under study, namely,
of S–O bond cleavage and phenolate-ion character in the 1a Ǟ 2a Ǟ 4 Ǟ products, and 2a Ǟ 3a Ǟ 5 Ǟ products.
transition states, both in the “β1 and β2” segments of the The value of β1 and the associated βlg and βeq values are
reaction. Recent work[16] has suggested that in the scissile 0.53, –0.43 and 0.96, respectively, and β2 and its associated
S–O bond there is significant lengthening in the slow step, βlg and βeq values are 0.19, –0.66 and 0.85, respectively.
and that the dissociative TS is sulfur-trioxide-like. Thea and
For the reaction 1a Ǟ 2a Ǟ 4 Ǟ products, α (bond-
Williams interpreted the βlg values of –1.21 (intermediate forming step) will be 0.53/0.96 = 0.55, and α (bond-break-
pH) and –1.79 (high pH) in water as supporting E1cB dis- ing step) will be 0.43/0.96 = 0.44. These values of the Leffler
sociative processes for the hydrolysis/aminolysis of com- indices indicate that removal of the proton to the base is
pounds 1.[13] The Hammett ρlg values (Figure 3) are 3.2 slightly more advanced than the cleavage of the sulfur–oxy-
(using N-formylpiperazine) and 5.1 (using pyrrolidine). gen bond for this reaction, and there is a small imbalance
Both suggest that a substantial amount of negative charge between the two steps.
is being carried in the transition states and may support the
For the second reaction in stronger bases 2a Ǟ 3a Ǟ 5
breakup of an anionic substrate 2 via an N-sulfonylamine Ǟ products, α (bond-forming step) is 0.19/0.85 = 0.22 and
4 (path A) and of the dianion 3 to give the anionic sulfon- α (bond-breaking step) is 0.66/0.85 = 0.78. This indicates
ylamine 5 (Scheme 2). Hammett ρ values at intermediate that for this reaction there is a very substantial imbalance
and high pH levels can be calculated from the data by Thea between the progress of these two crucial steps, and the pro-
and Williams,[17] and the values obtained were 2.32 and ton is only about 20% removed from the substrate, while
Eur. J. Org. Chem. 2008, 4200–4205
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4203

W. J. Spillane, A. O’Byrne, C. J. A. McCaw
FULL PAPER
Scheme 2. Favoured mechanistic pathways for the hydrolysis of 1a and other derivatives 1b–1f.
the S–OAr bond is cleaved to almost 80%. The removal of
The TS for Path A should lie on or somewhat below line
the second proton would be expected to be much more diffi- C, and the TS for Path B should be on or below line A but
cult than the removal of the first proton from 1a because well above line B. The TSs for Path A and Path B should
of the negative charge on 2a, and thus it is not surprising resemble those shown in Scheme 4:
to find that in the TS for Path B this lags well behind the
sulfur–oxygen bond-breaking step.
A free energy diagram for these elimination processes
serves to illustrate more clearly the changes occurring dur-
Scheme 4. TSs for Path A and Path B in Scheme 2. In Path B the
ing the reaction of these sulfamate esters in moderate bases
and stronger bases. Such reaction maps have been used to
follow more easily the changes taking place during the
course of elimination reactions,[18,19] and the changes taking
place during the elimination reactions in this work have
been delineated now in Scheme 3.
larger δ symbols represent substantial bond cleavage, and in both
TSs the distances between the various atoms/groups attempt to
show the variable length of the bonds.
Finally, Path A should be seen as an E2-/E1cB-type
mechanism and Path B as an E2 mechanism in which little
bond formation but extensive bond breaking has occurred,
and both are likely to involve very transient N-sulfonyl-
amine intermediates as shown in Scheme 2.
Conclusions
In this work an eliminative route in the aminolysis of
compounds 1a–f is strongly supported and the likely de-
composition pathways (A and B) are shown above
(Scheme 2). The establishment and interpretation of several
structure–reactivity relationships (Hammett and Brönsted
plots) point to substantial scission of the S–O bonds in
compounds 1 in both pathways to products. The breakup
of these compounds in ACN (studied by Thea and Williams
et al. in water) appear to follow E2-/E1cB-type and E2-type
mechanisms with the involvement of N-sulfonylamine inter-
mediates as outlined in Scheme 2. These important findings
throw light on the likely mechanisms involved in the inhibi-
tion of various enzymes such as STS and CA which certain
sulfamates such as 667-coumate, emate etc. can bring
about.
Scheme 3. Free energy reaction map for the elimination of pro-
ton(s) and –OAr from sulfamate esters NH2SO2OAr (1). Line C
represents Path A, i.e. 1a Ǟ 2a Ǟ 4 Ǟ products (see Scheme 2),
and shows how bond-forming between base and H+ is somewhat
more advanced than S–OAr bond-breaking and the line is pulled
off a little off centre towards the bottom right hand corner of the
diagram. Line A represents Path B, i.e. 2a Ǟ 3a Ǟ 5 Ǟ products
(see Scheme 2), and this line is dragged substantially towards the
top left hand corner reflecting the facts that S–OAr bond breaking
is well ahead of base and proton bond forming. Line B shows a
“central” E2 mechanism where bond-breaking and bond-forming
both occur to the same extent in the TS.
Experimental Section
Substrates and Reagents: The amines used in this work were puri-
fied by distillation under reduced pressure or recrystallization from
an appropriate solvent. The six substrates used, i.e 1a–f, were syn-
thesised according to a literature method[20] and gave yields of
70+%. Briefly, the method involved cooling of a stirred mixture of
4204
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4200–4205

Aminolysis of Sulfamate Esters
the appropriate phenol in dimethylacetamide with sulfamoyl chlo-
ride and allowing the reaction to occur at room temperature (moni-
tored by TLC). After the reaction was deemed to be complete, the
mixture was poured into brine, extracted with three aliquots of
ethyl acetate, washed exhaustively with brine, dried with magne-
sium sulfate and concentrated under reduced pressure. Final purifi-
cation was performed by column chromatography and, if necessary,
recrystallization from toluene. The purified materials gave C, H, N
microanalyses within Ϯ0.5 of the calculated values except for 1b
(N 0.89), 1e (C 0.99) and 1f (C 0.55). Melting points were generally
close to the literature values,[16,21] and the 1H NMR and IR spectra
were in agreement with their structures. The ACN used in this work
was HPLC grade, and Karl Fischer analysis showed that the water
content (as a% by volume) in ACN for four samples from different
batches was 0.058, 0.048, 0.122 and 0.099. As a check on the effects
of these small amounts of water in these ACN samples on pKa
values we measured the pKa values of five amines (diisopropyl-
amine, piperidine, morpholine, 4-DMAP and 4-pyrrolopyridine) in
the four samples and found only small variations within the experi-
mental error of the pKa determination. The largest variation was
for 4-DMAP, which was 17.56Ϯ0.07, but most were much lower.
Acknowledgments
The Millennium and Corrib Funds of the National University of
Ireland, Galway are thanked for grants. C. McC. and A. O’B. thank
Enterprise Ireland and Galway Co. Co. for grants. Drs. A. Bruzzi
and M. Cuffe of the Public Analysts Laboratory, Galway are
thanked for Karl Fischer analyses of the ACN solvents.
[1] P. Nussbaumer, P. Lehr, A. Billich, J. Med. Chem. 2002, 45,
4310–4320.
[2] S. J. Stanway, A. Purohit, L. L. W. Woo, S. Sufi, D. Vigushin,
R. Ward, R. H. Wilson, F. Z. Stanczyk, N. Dobbs, E. Kulin-
skaya, M. Elliott, B. V. L. Potter, M. J. Reed, R. C. Coombes,
Clin. Cancer Res. 2006, 12, 1585–1592.
[3] L. W. Woo, N. M. Howarth, A. Purohit, H. A. M. Hejaz, M. J.
Reed, B. V. L. Potter, J. Med. Chem. 1998, 41, 1068–1083.
[4] D. Poirier, L. C. Ciobanu, R. Maltais, Expert Opin. Ther. Pat.
1999, 9, 1083–1099.
[5] a) J.-Y. Winum, A. Scozzafava, J.-L. Montero, C. T. Supuran,
Med. Res. Rev. 2004, 25, 186–228 ; b) C. T. Supuran, Nat. Rev.
Drug Discovery 2008, 7, 168–181.
[6] M. J. Reed, A. Purohit, L. W. L. Woo, S. P. Newman, B. V. L.
Potter, Endocr. Rev. 2005, 26, 171–202.
Determination of pKa Values in ACN: The ionizations shown
(Scheme 1) in Equations (1) and (2) were measured at 25 °C by
using the procedures previously described.[8] The pKa value of the
first ionization of 1a i 2a [Equation (1)] was measured by using
4-nitrophenol as indicator and λanalytical = 220 nm. A value of 17.9
was obtained. The pKa value of the second ionization of 2a i 3a
[Equation (2)] by using in each case 2,4-dinitrophenol as indicator
gave values of (λanalytical in nm in parentheses): 20.96 (325), 21.13
(325) and 21.16 (305). This gives an average value of 21.08. The
pKa values in ACN of the amines used have been reported pre-
viously (see Table 1, footnote [a]) except that of 3,4-diaminopyr-
idine which was measured in this work as 14.7.
[7] A. Horvath, A. Billich, Expert Opin. Ther. Pat. 2005, 15, 1541–
1553.
[8] W. J. Spillane, P. McGrath, C. Brack, A. B. O’Byrne, J. Org.
Chem. 2001, 66, 6313–6316.
[9] W. J. Spillane, G. Hogan, P. McGrath, J. King, C. Brack, J.
Chem. Soc., Perkin Trans. 2 1996, 2099–2104.
[10] C. J. A. McCaw, W. J. Spillane, J. Phys. Org. Chem. 2006, 19,
512–517.
[11] G. B. M. Hogan, Ph. D. Thesis, National University of Ireland,
Galway, 1993.
[12] K. Izutsu, Acid-Base Dissociation Constants in Dipolar Aprotic
Solvents, IUPAC Chemical Data Series no. 35, Blackwell, Ox-
ford, 1990.
Kinetics: Rates were measured by using Cary 1 or Cary 50 UV
spectrophotometers fitted with thermostatted cell holders. All reac-
tions were carried out under pseudo-first-order conditions, i.e. ex-
cess amine (typically 0.2 to 0.02 ) with an initial substrate concen-
tration from 1ϫ10–4 to 5ϫ10–5 . The initial concentration of sub-
strate used did not effect the rates. The reactions were generally
monitored by the detection of the appropriate X-phenol/X-phenox-
ide ion. A few kinetic runs were monitored by the disappearance
of sulfamate ester and the agreement between the rates obtained
this way and those obtained by monitoring the appearance of X-
phenol/X-phenoxide was good. From plots of kobsd. vs. [amine],
straight lines were obtained, and the slopes of these plots gave k2,
the second-order rate constants for the aminolysis reactions. The
small amounts of water in the ACN did not lead to a hydrolysis
term in the rate law for these reactions, because plots of kobsd. vs.
[amine] passed through the origin or very close to it.
[13] S. Thea, G. Cevasco, G. Guanti, A. Williams, J. Chem. Soc.,
Chem. Commun. 1986, 1582–1583.
[14] A. Williams, K. T. Douglas, J. Chem. Soc., Perkin Trans. 2
1974, 1727–1732.
[15] W. J. Spillane, C. J. A. McCaw, N. P. Maguire, Tetrahedron Lett.
2008, 49, 1049–1052.
[16] E. Denehy, J. M. White, S. J. Williams, Chem. Commun. 2006,
314–316.
[17] The liberated X-phenols in the study in ref.[11] shown in Fig-
ure 1 were (with pKa values in water in parentheses): 4-NO2
(7.15), 4-CN (7.95), 3,5-diCl (8.2), 3-NO2 (8.4), 3-CN (8.55)
and 4-Cl (9.38).
[18] S. Thea, A. Williams, J. Chem. Soc., Perkin Trans. 2 1981, 72–
77.
[19] H.Maskill, The Physical Basis of Organic Chemistry, Oxford
University Press, Oxford, 1985, pp. 414–415.
[20] M. Okada, S. Iwashita, N. Koizumi, Tetrahedron Lett. 2000,
41, 7047–7051.
[21]
C. K. Patel, A. Galisson, K. James, C. P. Owen, S. Ahmed, J.
Pharm. Pharmacol. 2003, 55, 211–218.
W. J. Spillane, G. Hogan, P. McGrath, J. Phys. Org. Chem.
1995, 8, 610–616.
Product Studies: In earlier product studies conducted in ACN,
CHCl3 and 50% aqueous ACN, the sulfamide product has been
monitored by reversed-phase HPLC,[9,22] and in ACN the ab-
sorbances of spent kinetic solutions of 4-nitrophenol in amine were
almost identical with those of spiked solutions.[8,9]
[22]
Received: April 10, 2008
Published Online: July 4, 2008
Eur. J. Org. Chem. 2008, 4200–4205
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4205