Multiple changes rate-determining, step the acid base catalyzed cyclizations of ethyl N-(p-nitrophenyl)hydantoates caused by methyl substitution

Article abstract of DOI:10.1139/v99-079

The slopes of the pH-rate profiles for the cyclization of 2-methyl- and 2,3-dimethyl hydantoates 1-NPU and 2-NPU between pH 1 and 7 change from 1 to 0 and then back to 1. A reaction first order in H⁺ was observed with the latter compound. The 2

Full text of DOI:10.1139/v99-079

849
Multiple changes in rate-determining step in the
acid and base catalyzed cyclizations of ethyl
N-(p-nitrophenyl)hydantoates caused by methyl
substitution
Iva B. Blagoeva, Anthony J. Kirby, Asen H. Koedjikov, and Ivan G. Pojarlieff
Abstract: The slopes of the pH–rate profiles for the cyclization of 2-methyl- and 2,3-dimethyl hydantoates 1-NPU and
2-NPU between pH 1 and 7 change from 1 to 0 and then back to 1. A reaction first order in H+ was observed with the
latter compound. The 2,2,3-trimethyl derivative 3-NPU showed only one reaction first order in OH–, but complex acid
catalysis is described by slopes 0, –1, 0, and finally –1 again. The cyclizations were general base catalyzed, with
Brønsted β values of 0.5–0.6. The OH– catalysis at higher pH for 1-NPU and 2-NPU showed inverse solvent kinetic
isotope effects and deviated from the Brønsted relationships, while that for 3-NPU showed a normal effect and
complied with the Brønsted relationship. The accelerations due to the gem-dimethyl effect were lost with the OH– and
general base-catalyzed reactions of 3-NPU. This behaviour is due to a change from the rate-determining formation of
the tetrahedral intermediate with 1-NPU and 2-NPU to the rate-determining breakdown with 3-NPU, due to steric
hindrance to protonation of the leaving ethoxy group. The OH– reaction at higher pH involves attack of the ureide
anion with 1-NPU and 2- NPU, becoming concerted with deprotonation when catalyzed by general bases and changing
to acid inhibition of the anion of the tetrahedral intermediate at low pH. With 3-NPU at higher pH, T– is in
equilibrium and the conjugate acids of the general bases accelerate its breakdown by protonating the ethoxy group.
Acid catalysis of the cyclization of 3-NPU at higher pH is also protonation of the leaving group from T0 changing to
the rate-determining formation of T at lower pH. The latter mechanism is preferred for the cyclization of 2-NPU.
Key words: gem-dimethyl effect, mechanism, general base catalysis, proton transfer, steric hindrance.
Résumé : Les pentes des profils pH–vitesse pour la cyclisation des 2-méthyl- et 2,3-diméthylhydantoates 1-NPU et
2-NPU entre les pH 1 et 7 passe de 1 à 0 avant de revenir à 1. Avec le dernier composé, on observe que la réaction
est du premier ordre en H+. Le dérivé 2,2,3-triméthylé 3-NPU ne donne qu’une réaction du premier ordre en OH–,
mais une catalyse acide complexe décrite par des pentes de 0, –1, 0 et finalement –1 encore une fois. Les cyclisations
sont soumises à une catalyse générale par les bases avec des valeurs β de Brønsted qui vont de 0,5 à 0,6. Les 1-NPU
et 2-NPU soumis à une catalyse par OH–, à des pH plus élevés, mettent en évidence l’existence d’effets isotopiques
cinétiques inverses du solvant et des déviations des relations de Brønsted. Avec OH–, les accélérations provoquées par
l’effet gem-diméthyle sont perdues ainsi que les réactions soumises à une catalyse générale par les bases. Ce
comportement est dû au fait que dans les cas des 1-NPU et 2-NPU l’étape cinétiquement déterminante est la formation
d’un intermédiaire tétraédrique alors que dans le cas du 3-NPU l’étape déterminante est son bris qui est dû à
l’empêchement stérique à la protonation du groupe éthoxy partant. La réaction OH–, à un pH plus élevé, implique une
attaque de l’anion uréide par les 1-NPU et 2-NPU qui devient concertée avec la déprotonation lorsqu’elle est soumise à
une catalyse générale par les bases et qui change à une inhibition acide de l’anion de l’intermédiaire tétraédrique à
pH plus bas. Avec le 3-NPU, à des pH plus élevés, T– est en équilibre et les acides conjugués des bases générales
accélère son bris en protonant le groupe éthoxy. La catalyse acide de la cyclisation du 3-NPU à des pH plus élevés est
aussi la protonation du groupe partant de T0 qui se change en formation cinétiquement limitante de T à des pH plus
faibles. Ce dernier mécanisme est préféré pour la cyclisation du 2-NPU.
Mots clés : effet gem-diméthyle, mécanisme, catalyse générale par les bases, transfert de proton, empêchement stérique.
[Traduit par la Rédaction] Blagoeva et al. 859
Received December 2, 1998.
This paper is dedicated to Jerry Kresge in recognition of his many achievements in chemistry.
I.B. Blagoeva, A.H. Koedjikov, and I.G. Pojarlieff.1 Institute of Organic Chemistry, Bulgarian Academy of Sciences, ul. Acad. G.
Bonchev block 9, Sofia 1113, Bulgaria.
A.J. Kirby. University Chemical Laboratory, Cambridge CB2 1EW, U.K.
1Author to whom correspondence may be addressed. Telephone: +359-2-7334-116. Fax: 359-2-700-225. e-mail:
ipojarli@orgchm.bas.bg
Can. J. Chem. 77: 849–859 (1999)
© 1999 NRC Canada

850
Can. J. Chem. Vol. 77, 1999
the workings of the gem-dimethyl effect in reactions through
tetrahedral intermediates.
Introduction
The hydroxide ion catalyzed cyclization of ethyl 2,2,3-
trimethyl-5-phenyl hydantoate, 3-PUE, is slower than that of
the dimethyl compound 2-PUE (1). This result is surprising
because the reaction of 3-PUE should be subject to the gem-
dimethyl effect (2): indeed, a normal gem-dimethyl effect
operates on the acid-catalyzed cyclizations of these com-
pounds (1-PUE–3-PUE) (3, 4). Data on the general acid-
base catalysis and solvent kinetic isotope effects (SKIE) on
PUE and their 5-methyl analogues (1-MUE–3-MUE) led us
to the conclusion that the loss of the gem-dimethyl effect
was due to steric hindrance to proton transfer of the leaving
ethoxide group, resulting in a change in the rate-determining
step (r.d.s.) from formation to breakdown of the tetrahedral
intermediate (3). The balance of evidence as to which r.d.s.
is preferred, with the less heavily substituted compounds
changing into an alternate one in the fully methylated deriv-
atives, was delicate and thus further evidence was sought for
these systems. The improved leaving group ability of the p-
nitrophenylureido group was intended to bias the system to-
wards rate-determining departure of the ethoxy group, thus
allowing a simpler evaluation of the various effects.
Grounds for such reasoning were the alternative cyclizations
of dicarbamoylglycine 4: the specific base-catalyzed (SBC)
ring closure with the N-isopropylcarbamoyl moiety was at-
tributed to r.d. attack of the anion while the GBC reaction
involved the end aryl group to r.d. departure of the hydroxy
group (5).
Experimental
Materials
Inorganic reagents and buffer components were of analyti-
cal grade and were used without further purification. Potas-
sium hydroxide and buffer solutions were prepared with
CO2-free distilled water. D2O and DCl (20 wt.% in D2O), 99
at.% were from Aldrich.
Ethyl esters of methyl substituted 5-(4-nitrophenyl)hydantoic
acids were prepared by reaction of the corresponding amino
acid esters with 4-nitrophenyl isocyanate. The amino acid
esters were converted into the free base form and distilled
prior to use. The typical experimental procedure consisted in
addition of 10% excess of 4-nitrophenyl isocyanate
(3.3 mmol in 5 mL dry benzene) to an ice-cooled solution of
the amino acid ester (3 mmol in 5 mL dry benzene). A heavy
precipitate was formed immediately and the reaction mixture
was left for 20 min at room temperature. The precipitated
product was filtered, washed with dry benzene, and dried in
a vacuum desiccator with P2O5. The products could not be
heated or recrystallized due to their fast cyclization at higher
temperatures. 3-NPU cyclized readily in the presence of
traces of moisture and was best stored in sealed ampoules at
low temperatures. The purity of the products was checked
by NMR and the yields of pure esters were 62–64%.
Ethyl 3-methyl-5-(4-nitrophenyl)hydantoate (1-NPU): mp
124–125°C. IR νmax(CHCl3)/cm–1: 1727 (CO ester), 1676
(CO ureido). H NMR (CDCl3) δ (ppm): 1.31 (t, J 7.1 Hz,
3H, CH3CH2), 4.15 (s, 2H, N-CH2), 3.16 (s, 3H, CH3N),
4.25 (q, J 7.1 Hz, 2H, CH2O), 7.08 (s, 1H, HN), 7.54–8.19
(m, 4H, Ar).
O
1
2
3
R
MUE Me
PUE Ph
2
R
H
R
H
1
R
1
2
3
OEt
1
R
Me H
Me Me
3
NHR
N
NPU 4-nitrophenyl
Me
O
Ethyl 2,3-dimethyl-5-(4-nitrophenyl)hydantoate (2-NPU):
mp 130–131°C. IR νmax(CHCl3)/cm–1: 1727 (CO ester), 1672
(CO ureido).1H NMR (CDCl) δ (ppm): 1.30 (t, J 7.2 Hz, 3H
CH3CH2), 1.49 (d, J 7.35 Hz,, 3H CH3CH), 3.04 (s, 3H,
CH3N), 4.23 (q, J 7.2 Hz, 2H, -CH2O), 5.04 (q, J 7.35 Hz
1H,-CHN), 6.99 (s, 1H, HN), 7.56–8.20 (m, 4H, Ar).
Cl
NHCO
CONH
N
CH2
Cl
Ethyl 2,2,3-trimethyl-5-(4-nitrophenyl)hydantoate (3-NPU):
mp 125–127°C. IR νmax(CHCl3)/cm–1: 1725 (CO ester), 1670
(CO ureido). H NMR (CDCl) δ (ppm): 1.29 (t, J 7.1 Hz,
CO2H
1
4
3H, CH3CH2), 1.51 (s, 6H, (CH3)2C), 3.07(s, 3H, CH3N),
4.21 (q, J 7.1 Hz, 2H, CH2O), 7.28 (s, 1H, HN), 7.49–8.34
(m, 4H, Ar).
Unexpectedly, the hydroxide-catalyzed reaction of the
p-nitrophenyl esters mirrored the behaviour of the ureido es-
ters with less negative ω-substituents, showing only a small
increase in rate with 3-NPU compared to a large one under
acid catalysis, and a normal SKIE opposed to the inverse
one with 1-NPU and 2-NPU. In contrast to MUE and PUE,
however, the p-nitrophenyl derivatives showed highly com-
plex rate profiles, varying extensively upon methyl substitu-
tion.
We now report how this interplay of the gem-dimethyl ef-
fect and reaction mechanism can be understood from an
analysis of the reaction profiles, general base catalysis, and
SKIE. The results obtained support our previous interpreta-
tion of the loss of the gem-dimethyl effect in the HO–-cata-
lyzed cyclization in esters 3 and allow further insight into
Product analysis
The cyclizations of the ethyl hydantoates studied in this
paper proceeded quantitatively to the corresponding hydan-
toins. Good isosbestic points were obtained for all three
compounds studied and the end-point absorbances for the ki-
netic runs were identical within experimental error with the
absorbances of model solutions of the respective hydantoins
described previously (6).
Kinetic measurements
Rate constants were determined at 25.0 ± 0.01°C under
pseudo-first-order conditions in the thermostatted cell com-
© 1999 NRC Canada

Blagoeva et al.
851
Fig. 1. Rate profiles for the cyclization of 1-NPU (᭺) and
2-NPU (᭝).
Fig 2. Rate profile for the cyclization of 3-NPU.
Scheme 1.
_
O
O
OEt
1
R
1
R
k
k
_
partment of a Unicam SP-800 spectrophotometer. The rates
of cyclization of the substituted esters were followed by
monitoring the decrease of the absorbance at 330 nm due to
the 4-nitrophenylureido group. Reactions were initiated by
injecting 20 mL of 1 × 10–2 M stock solutions of the sub-
strates in dry THF into 2.7 cm3 of preheated buffer solution.
Ionic strength was maintained constant (1.0 M) with KCl.
Measurements of pH values and calculations of the observed
pseudo-first-order rate constants (kobs) were as previously
described (7). Good linear plots (r > 0.999) of ln(At – A∞)
against time were obtained over three half-lives and kobs val-
ues were reproducible to within 3% with the exception of
the slow ring closure of 1-NPU in hydrochloric acid below
pH 2 where the hydrolysis of the ester group was significant.
Since the product acids cyclized more rapidly than the es-
ters, plots of ln(At – A∞) against time sloped downwards
with time. In this case the rate constants were obtained from
the initial part of the reaction where linearity was preserved:
initial rates in 0.1 M Hcl up to ca. 20% conversion in 0.01 M
HCl.
f1
OEt
3
NR
+ HA
2
+
A
R
2
R
N
3
NHR
r1
N
Me
Me
O
O
_
T
UH
O
1
R
k
_
+ EtOH + A
f2
3
NR
2
R
k
N
r2
Me
O
Below pH = 1, hydrolysis of the ester group was observed
and acid catalysis could not be detected. With the faster
cyclization of compound 2-NPU the reaction could be stud-
ied up to 1 M HCl and an acid-catalyzed region could be
registered in addition to base catalysis. The latter is also best
described by equations leading to the same changes of
slopes against pH, from 1 to 0 and back again to 1. The min-
imum between the acid-catalyzed and base-catalyzed regions
Solvent kinetic isotope effects
The solvent kinetic isotope effects (SKIE) for kH were de-
termined by comparison of kobs values measured in solutions
of HCl or DCl of the same concentration. The kOH values for
calculation of SKIE were determined by extrapolation to
zero buffer concentrations of kobs in phosphate or cacodylate
buffers both in H2O and D2O. For the hydroxide-catalyzed
reaction k0 values were divided by the activity of HO– or
DO– to obtain kOH or kOD. pD values were obtained by add-
ing 0.4 to the pH-meter readings and the aOD values were
calculated using pKw = 14.86 (8).
occurs around pH
= 1, illustrating the very poor
nucleophilicity of the neutral p-nitrophenylureido group.
The same minimum occurs around pH 3 with 2-PUE (3).
The behaviour of ester 3-NPU, presented separately in
Fig. 2 to avoid overlap in the base-catalyzed region, is dis-
tinctly different: only one segment, first order in [OH–], is
observed, while the acid catalysis region is complex.
The rate law for base catalysis can be derived from
Scheme 1, depicting the main steps in the reorganization of
bonds from reactants to products.
The steady state solution is
Results and discussion
kf1[A− ]kf2[HA]
Rate profiles
[1]
kobs =
(kr1 + kf2)[HA]
1. Base catalysis
With ureido ester 1-NPU the slope of the plot of log kobs
versus pH changes from 1 to 0 and then back again to 1 (Fig. 1).
When the bases are water and hydroxide ion this becomes
© 1999 NRC Canada

852
Can. J. Chem. Vol. 77, 1999
107 dm3 mol–1 s–1 are observed for the kOH of 1-NPU and
a
Table 1. Kinetic solvent isotope effects for ring closure of
methyl substituted ethyl 5-(4-nitrophenyl)hydantoates at 25°C
for the kOH of 3-NPU (Table 2). Participation of OH– in a
concerted process in which the N—C bond is formed while
the NH proton is pulled off (formula 5, B = OH–) is also un-
likely because, according to the libido rule of Jencks (12),
the base has to be of strength intermediate between that of
the proton-donating group in the ground and product states.
The alternative SB-GAC mechanism for N—C formation,
i.e., attack of the ureide anion concerted with proton transfer
Substrate
[LCl] mol dm–3 (kH / kD)obs
pH (H)
5.60a
kOH / kOD
1-NPU
0.002
0.005
0.01
2.9
2.4
1.8
3.0
2.9
0.8
2.5
2.6
1.1
0.78
2-NPU
3-NPU
0.005
0.01
5.57b
0.6
0.90
0.005
0.01
5.62b
5.82c
1.8
1.9
δ+
B
_
H
O
δ
1
0.90
O
OEt
NR
aCacodylate buffer 20% base.
bPhosphate 10% base.
cPhosphate 15% base.
R
OEt
NR
δ+
3
1
R
H
B
2
R
3
δ −
2
R
MeN
MeN
5
6
(kf1w + kf1OH[OH− ])(kf2w + kf2H[H+ ])
kr1w + kf2w + (kr1H + kf2H)[H+ ]
δ+
B
H
[2]
kobs =
_
δ
1
O
OEt
1(a). Ethyl hydantoates 1-NPU and 2-NPU: Various limit-
ing conditions lead to simplified forms of eq. [2], describing
rate profiles with slopes changing from 1 to 0 and then back
to 1. In the following discussion2 we designate the apparent
R
3
NR
2
R
MeN
a
second-order constant at lower pH as kOH and the one at
7
higher pH as kOHb. Two conditions have to be met (9) in or-
der that two different kOH are observed in situations de-
scribed by eq. [2]. The first is that the two forward fractions
for breakdown of T–, kf2/(kr1 + kf2), for H2O and H+ cataly-
sis, respectively, have to be different. The first one will op-
erate at higher pH and the second one at lower pH. Two
different reactions of the same order are usually associated
with a change in the rate-determining step; in the case under
discussion, when kr1w < kf2w then kr1H > kf2H or vice versa.
The second condition for two reactions of the same order to
be observed is that the intermediates involved should not
equilibrate. In such a case the reactions go simply by the
lowest energy path available to, for example, T–, which in
turn corresponds to single first order in the [OH–] region in
the rate profile. In our case two different kOH can be ob-
served when one or both processes are general acid–base
catalyzed.
to the ester carbonyl leading to T0 (formula 6, B = OH–) was
refuted (3) for the ω-methyl and ω-phenyl analogues MUE
and PUE as it contradicts Jencks’ requirement for T– to be
an unstable intermediate (13); this apparently applies also
for NPU to permit the SBC catalyzed reaction of 1-NPU and
2-NPU presented by kOHb, discussed below. We are thus left
with the last possible proton transfer as the cause of the nor-
mal SKIE, that is, water donating a proton to the ethoxy
group cleaving from T– (formula 7, B = OH–). As discussed
before (3), a concerted mechanism is possible with a late
C—O bond cleavage. In terms of Scheme 1 and eq. [2], this
a
mechanism requires kr1 > kf2. The process kOH is observed
at low pH when kr1w + kf2w < (kr1H + kf2H)[H+] so the in-
equality kr1H > kf2H will apply. When kf1w > kf1OH[OH–] and
if kf2w > kf2H[H+] in the region for base catalysis, the ob-
served rate constant becomes
As is generally the case for kinetically equivalent mecha-
nisms, the preference must be based on non-kinetic evi-
dence. Comparison of the rate profiles of Figs. 1 and 2 with
kf1wkf2w
[3]
kobs =
k
r1H[H+ ]
a
the SKIE data listed in Table 1 shows that the kOH values
for 1-NPU and 2-NPU, as well as the single kOH constant for
3-NPU, exhibit normal isotope effects (kH > kD) These allow
a choice between various possible mechanisms. The similar-
ity in the pK’s of the p-nitrophenylureido group and water,
the conjugate acid of OH–, could give rise to a normal SKIE
(10) for proton removal from nitrogen (a pK of 14 has been
reported for p-nitrophenylurea (11)). However, if this step
was rate determining, a Brønsted β value of 1 is expected,
contrary to the observed values of 0.5–0.6 (vide infra Ta-
ble 6). Such a deprotonation would be diffusion controlled
and would certainly be much faster than the overall rates in
at least two of the cases; second-order constants of ca. 106–
The rate constant first order in [OH–] at low pH can be then
be presented as
KT– kf2w
[4]
kOaH
=
Kw
where kT– is the equilibrium constant for formation of T–
when A– = H2O. Thus at low pH the first order in [OH–] be-
haviour is determined by the equilibrium concentration of T–,
and the rate-determining step is the water-catalyzed decom-
position of T–. Such a mechanism is referred to as acid inhi-
bition of T– (14).
2 For the sake of brevity exhaustive consideration of the various cases afforded by eq. [2] will be avoided.
© 1999 NRC Canada

Blagoeva et al.
853
Table 2. Rate constants for H+, OH–, and H2O catalyzed ring closure of ethyl 5-(4-nitrophenyl)hydantoates at 25°C and ionic strength 1.0 M.
kOH dm3 mol–1 s–1
kOH dm3 mol–1 s–1 b
kw s–1
k2H dm3 mol–1 s–1
k1H dm3 mol–1 s–1
b
a
Compound
1-NPUa
(4.23±0.22) × 104
(3.85±0.20) × 106
kOH
(6.94±0.69) × 106
(2.07±0.019) × 108
k1w s–1
(1.36±0.11) × 10–5
(2.18±0.16) × 10–4
k2w s–1
196±30
107±16
2-NPUc
(8.34±0.61) × 10–5
(1.21±0.36) × 10–3
3-NPUd
(5.20±0.37) × 106
(1.10±0.18) × 10–3
(2.46±0.37) × 10–1
35.2±16.9
aEquation [5].
bkw / k2HKw.
cEquation [6].
dEquation [12].
Scheme 2.
_
_
O
O
O
O
_
_
+
OEt
OEt
+ A
+ A
+ HOEt
OEt
(
+
NAr
H
H
(
)
)
NAr
NAr
NHAr
H
A
k
f2w
k
r1wb
k
r1wa
C
N (C O)
_
O
O
O
_
+ OEt
+ HA
OEt
+ HA
OEt
NAr
_
)
(
NAr
NAr
Upon increasing the pH, the plateau is reached with 1-
NPU and 2-NPU when the [H+] term in the denominator can
be neglected. Now kobs = kf1w because, as discussed above,
kr1w < kf2w in order for the second slope of 1 in the rate pro-
file to be observed at higher pH. Then kobs = kf1OH [OH–].
With the above limitations eq. [2] can be simplified to
The inequality kr1w < kf2w means that p-nitrophenylureido
leaves less readily than ethoxide at high alkalinity against
expected fugacities while the opposite is true at low
pH (kr1H > kf2H). The reason for the former inequality can be
sought in reactivities of stepwise vs. concerted processes
(15). The acid-catalyzed breakdown of T– can be described
by the alternate route diagram (16) shown in Scheme 2.
When HA = H2O, the leaving of the ureido group is a two-
step process along the edges of the diagram: splitting off of
the ureide anion, kr1wa, followed by proton transfer from
H2O to the nitrogen atom, kr1wb. This is not a concerted pro-
cess for reasons discussed above explaining why OH– can-
not act as a general base in the reverse reaction and, in terms
of the alternate route diagram, this happens because the en-
ergy of the bottom right-hand corner is considerably lower
than that of the top-left corner. On the other hand, as dis-
cussed above for 3-NPU, the ethoxy group cleaves along the
diagonal in a one-step process with both bonds changing si-
multaneously, kf2w. This occurs because the energy of the
bottom-right corner is raised, ethanol being a weaker acid by
more than two pK units, and that of the top-left corner is
lowered because of the greater basicity of the ethoxy group.
Thus a valley along the diagonal is created, with the energy
of the transition state apparently lower than that for the at-
tack of the ureide anion.
kw + kOH[OH− ]
1 + k2H[H+ ]
[5]
kobs =
where kOH = kObH = kf1OH, kw = kf1w, k2H = kr1H /kf2w, and kOaH
= kw/(k2HKw). The rate constants for 1-NPU in Table 2 were
obtained by fitting the rate data to eq. [5]. In the case of 2-
NPU the equation was expanded with a term for acid cataly-
sis.
kw + kOH[OH− ]
1 + k2H[H+ ]
[6]
kobs
=
+ k1H[H+ ]
The kOHb constants for 1-NPU and 2-NPU refer to an SBC
process as evidenced by the inverse isotope effect at pH =
5.6 where, according to Fig. 1, the second first order in
[OH–] process dominates. Because, as already discussed,
kr1w < kf2w, this is r.d. formation of T– taking place by pre-
liminary ionization of the ureido esters.
The crossover of reactivities for H2O and H3O+ catalysis
of partitioning of the intermediate is an interesting problem.
When HA = H3O+ the relationships in the diagram change
significantly. Considered in the reverse direction, the
© 1999 NRC Canada

854
Can. J. Chem. Vol. 77, 1999
T–, concerted with catalysis by water. When the ethyl group
is replaced by hydrogen, as is the case with the respective
hydantoic acids, the intermediate becomes less crowded and
slowing down is no longer observed.3 The GBC cyclization
of thioureido esters (17) and the above-mentioned (5) GBC
cyclization of the aryl moiety of the N,N-dicarbamoylglycine
4 are also believed to be limited by breakdown of T–.
Scheme 3.
O
HO
OEt
1
R
o
f1
o
r1
1
R
k
OEt
3
NR
+ HA
2
+
HA
R
2
R
N
3
k
NHR
N
Me
Me
O
O
o
T
2. Acid catalysis
UH
The complex course for acid catalysis of the cyclization of
3-NPU can be accommodated by an equation predicting four
regions of slopes 0, –1, 0, and –1 in the rate profile on de-
creasing pH. Such an equation can be derived from
Scheme 3, similar to Scheme 1, involving H+ and water-
catalyzed formation and decomposition of T0.4 This would
then yield an equation in the format of eq. [2], the super-
scripts 0 denoting that they are related to T0.
O
1
o
R
k
f2
3
NR
o
2
+ EtOH + HA
R
k
r2
N
Me
O
deprotonation by water along the right-hand edge is now
quite an uphill reaction, offering no easy bypassing of the
zwitterions at the top-left corner. These are highly unstable
and so will enforce general catalysis. This applies to both re-
actions, kr1H and kf2H and, when the mechanism is the same,
the intrinsically better fugacity of the p-nitrophenylureido
group determines reactivity.
(kf01w + kf01H[H+ ])(kf02w + kf02H[H+ ])
(kr01w + kf02w) + (kr01H + kf02H)[H+]
[7]
kobs =
To describe the rate profile for acid catalysis we chose the
combination kr01w > kf02w and kr01H < kf02H because when [T–]
reaches equilibrium (as assumed in the preferred mechanism
for kOH of 3-NPU) so should [T0] and the constants for water
catalysis are important in the overlapping region. So when
all the [H+] terms are small one gets
1(b). Ethyl hydantoate 3-NPU: In contrast to 1-NPU and
2-NPU, the rate profile of 3-NPU (Fig. 2) is characterized
by a single rate constant first order in [OH–]. This implies
that the partitioning ratios for water and H+ catalyzed break-
down of T– of 3-NPU are biased in the same way, i.e., no
change in the r.d.s. takes place. Why 3-NPU is different can
be understood by comparing with previous results on MUE
kf01wkf02w
kr01w + kf02w
[8]
kobs
=
≈ KT0 kf02w
b
Equation [8], the water-catalyzed decomposition of T0, is
kinetically equivalent to another pathway: the H+-catalyzed
decomposition of T–, kobs = KT– kf2H, and the latter is most
probably a more efficient one.
After the plateau described by eq. [8], most likely the ad-
joining slope of –1 is due to acid-catalyzed decomposition
of T0:
and PUE (3). The inverse SKIE observed with the kOH con-
stants for 1-NPU and 2-NPU and the normal effect with the
kOH constant for 3-NPU repeat the pattern of 1–3-MUE and
PUE. Significant also is the disappearance of the gem-dimethyl
effect with the kOH of 3-NPU: in contrast to acid catalysis,
where 3-NPU reacts 15 times faster than 2-NPU, the rates in
the higher pH region are almost the same (Table 2). The
change of mechanism brought about by the introduction of
the extra methyl group in 3-MUE or 3-PUE was explained
by steric hindrance to protonation of the oxygen atom of the
ethoxy group. The ester ethyl group will adopt the least
crowded conformation, blocking easy protonation of the ethoxy
oxygen because its lone pairs are effectively shielded in a
half-chair conformation by the methyl groups on C-5 and the
aryl ring on N-2 (see 8; for clarity, one of the methyl groups
at C-5 in the formula is represented by a single bond).
[9]
kobs = KT0 kf02H[H+ ]
followed by a change in the rate-determining step when
(kr01w + kf02w) < (kr01H + kf02H)[H+]:
[10] kobs = kf01w
This formally acid-catalyzed reaction according to Scheme 3
is also most likely the kinetically equivalent general base ca-
talysis by water.
Finally, when kf1w < kf1H[H+]
O-
CH
Ar
O
2
O
CH3
[11] kobs = kf01H[H+ ]
N
N
H
The rate profile can then be described by eq. [12]:
(k1w + k1H[H+ ])(k2w + k2H[H+])
Me
H
H
[12] kobs
=
+ kOH[OH− ]
O
1 + k2H[H+ ]
H
H
8
kT– k
kOH
=
f2w; k2w = KT– kf2H or = KT0 kf02w; k2H = KT0 kf02H
;
Kw
k1w = kf01w; k1H = kf01H
Thus there is good reason to believe that, for 3-NPU,
steric hindrance to protonation determines r.d. breakdown of
.
3 To be published.
4
Catalysis by H+ can be formally assumed to produce To as T+ loses proton to bulk water under diffusion control.
© 1999 NRC Canada

Blagoeva et al.
855
Fig. 3. Plot of kbuf against fraction base in acetate buffers for
2-NPU.
Fig. 4. Plot of kbuf against fraction base for 1-NPU in neutral
phosphate. Insert excluding data for 0.9 fraction base.
The above assignments of mechanism agree well with the
SKIE values listed in Table 1. In 0.90 M LCl it is 0.8 for 2-
NPU. This value applies for a “clean” reaction and is the
same as the values of 0.7–0.8 found for ω-phenylhydantoic
esters.(4) This and other evidence with the latter (4) sup-
ported rate-determining formation of T via the transition
state 9:
Plots of kbuf vs. fraction (of) base were convex to a vary-
ing degree. The “nonlinear” behaviour is illustrated by the
examples in Figs. 3 and 4.
This nonlinearity is appreciable when a wide range of
buffer ratios are studied. With narrow ranges it is indicated
by the negative intercepts in plots of kbuf vs. fraction base. A
solution was sought by means of equations derived from
Scheme 1. When a general base is included in Scheme 1, the
full steady state treatment yields a complex solution that can
reasonably be analyzed only under simplifying assumptions.
The constants for general base catalysis were obtained by
nonlinear regression curve fitting of the equations shown be-
low. The parameters describing the rate profile as a function
of pH were fed in as known constants. A series of equations
were tried: only those that gave good fits will be discussed.
Unless stated otherwise, all equations used have a 1 +
k2H[H+] term in the denominator; k2H from the rate profiles
(Table 2) varies from around 200 in 1 to 30 in 3, which
means that at pH > 3 its influence becomes negligible.
In cases when the general base is HPO4 2– we found previ-
ously (3) that upward trends at high buffer ratios are de-
scribed well by the addition of a kBOH × [B] × [OH–] term
δ−
A H
O
OEt
1
R
3
δ+
NHR
2
R
MeN
O
9
The somewhat larger SKIE value of 1.1 for 3-NPU does
not contradict the conclusion for kf1H as rate determining at
the highest acidity because it is most probably enlarged by
contribution from the water-catalyzed reaction. According to
Table 1, effects of about 2.5–3 are observed at lower [LCl]
with 2-NPU and 3-NPU.
3. General base catalysis
In contrast to the case of ω-phenylureido esters where GC
was clearly expressed only in the case of 3-PUE and diffi-
cult to establish with the remaining compounds, the effect of
buffer catalysis was significant and could be readily moni-
tored with all three p-nitrophenylureido derivatives.
Corrections for buffer failure were carried out as pH var-
ied strongly at low fractions of base in acid phosphate and
glycine buffers. Changes in pH were satisfactorily repro-
duced by solving the equation for the ionization
3–
interpreted as PO4 catalysis. An extensive series of data in
neutral phosphate for ester 1 was treated by means of
eq. [13]:
kw + kOH[OH− ] + kB[B] + kBOH[OH− ][B]
[13] kobs
=
1 + k2H[H+ ]
(The kH term is negligible in neutral phosphate.) This equa-
tion gave the very good fit as illustrated in Fig. 5, which
compares the experimental with the calculated points
(eq. [14] is a multivariable function and cannot be presented
as a two-dimensional graph).
(CA − aH)aH
KAH
=
(CAH + aH)
3–
The rate constant for catalysis for PO4 is obtained as
where KAH is the dissociation constant at I = 1 M KCl ac-
cording to ref. 7, CA and CAH are the concentrations as
weighed, and the actual concentration of the buffer compo-
nents may be approximated as [A–] = CA – aH, and [AH] =
CAH + aH.
kPO = kBOHKw/K
.
3 −
2 −
HPO4
4
Alternately the deviations from linearity could be de-
scribed by equations including a kBH term in the denomina-
tor. The following two equations proved most useful with
© 1999 NRC Canada

856
Can. J. Chem. Vol. 77, 1999
Table 3. Buffer catalysis data for the cyclization of ethyl 3-methyl-5-(4-nitrophenyl)-hydantoate, 1-NPU, at 25°C and ionic
strength 1.0 M.
Conc. range
mol dm–3
104 kB
104 k2BH(k1BH
)
a,b
dm3 mol–1 s–1b
dm3 mol–1 s–1
pKAH
Buffer acid
H3O+
Runs
% base
–1.74
(–1.26)
2.45
0.136 ± 0.011c
H3N+CH2CO2H
0.1–0.5
0.1–0.5
0.1–0.5
3
3
3
3
4
4
4
4
4
4
4
5
4
4
9
8
5
30
50
70
90
50
70
30
50
70
20
30
50
10
30
50
70
90
0.274 ± 0.067d
27 000 ± 18 000
(0.32 ± 0.17)
0.1–0.5
HCO2H
3.57
4.62
0.01–0.5
0.01–0.5
0.1–0.5
0.1–0.5
0.1–0.5
0.01–0.3
0.01–0.3
0.01–0.4
0.016–0.2
0.016–0.5
0.016–0.5
0.016–0.4
0.016–0.35
2.77 ± 0.13d
7.28 ± 0.19e
21 500 ± 2600
(0.432 ± 0.189)
5000 ± 2.9800
CH3CO2H
(CH3)AsO2H
6.19
6.48
56.7 ± 2.8f
–
H2PO4
42.7 ± 3.1g (21.3)
978 000h
(326 000)
(4.23 ± 0.22) × 108
2–
HPO4
12.32
(11.84)
15.74
H2O
(16.04)
apKAH values from ref. 7.
bStatistically corrected values in parentheses.
cFirst-order rate constant.
dEquation [15].
eEquation [14].
fEquation [5] + kB[B].
gEquation [13].
hCalculated from kBOH
.
compounds 1-NPU and 2-NPU. They derive from eq. [1] un-
der certain conditions.
Fig. 5. Logarithms of the first-order rate constants for
cyclization of 1-NPU in neutral phosphate: (᭺) experimental
data, (᭹) theoretical points obtained by means of eq. [14] and
parameters in Tables 2 and 3.
kw + kOH[OH− ] + kB[B]
[14] kobs
[15] kobs
=
=
1 + kH[H+ ] + k2BH[BH+ ]
kw + kOH[OH− ] + kB[B] + k1BH[BH+ ]
1 + kH[H+] + k2BH[BH+ ]
In the case of ester 2-NPU the term for H+ catalysis was
added as in eq. [6].
Equation [14] is an extension of eq. [5], derived by as-
suming that kBH behaves as k2H in the case of 1 and 2, that
is, kr1BH > kf2BH and kf2BH[BH+] < kf2w in the range studied.
Then kB = kf1B and k2BH = kr1BH/kf2w (kf2w > kr1w). The in-
crease of k2BH[BH+] with the decrease of fraction base will
cause at the same time kbuf to decrease faster than demanded
by a linear function. These assumptions mean that kB is in-
volved in the formation of T– and can be measured directly
as long as formation of T– is r.d. for the overall reaction.
When k2BH [BH+] becomes greater than 1 + kH[H+] there is a
change in the r.d.s.; actually eq. [14] predicts reaching the
situation determined by eqs. [3] and [4], the first step be-
Equation [15] improves the fit further in some cases. The
coming an equilibrium and T– giving products by kf2w
.
term k1BH [BH+] in the numerator could either be due to
© 1999 NRC Canada

Blagoeva et al.
Table 4. Buffer catalysis data for the cyclization of ethyl 2,3-dimethyl-5-(4-nitrophenyl)-hydantoate, 2-NPU, at 25°C and ionic
857
strength 1.0 M.
Conc. range
mol dm3
104 kB
104 k2BH
a,b
dm3 mol–1 s–1,b
dm3 mol–1 s–1
pKAH
Buffer acid
H3O+
Runs
% base
–1.74
(–1.26)
1.78
2.18 ± 0.16c
H3PO4
0.01–1
0.01–1
0.1–1
0.1–1
0.1–1
0.1–1
0.1–1
0.01–0.2
0.01–0.2
0.01–0.2
0.01–0.2
0.05–0.2
0.01–0.2
0.01–0.2
0.01–0.2
0.05–0.2
0.01–0.15
0.01–0.15
0.01–0.15
0.01–0.15
0.01–0.15
6
6
5
5
3
3
3
4
4
4
4
3
4
4
4
3
4
4
4
4
4
30
50
70
90
50
70
90
30
50
70
90
10*
30
50
70
90
10
20
30
50
70
16.1 ± 2.0d
120 000 ± 21 300
22 400 ± 5500
(2.26)
H3N+CH2CO2H
HCO2H
2.45
3.57
12.7 ± 1.3d
65.9 ± 5.6e
CH3CO2H
(CH3)AsO2H
H2O
4.62
6.19
533 ± 64d
78 800 ± 19 100
1930 ± 170e
15.74
(3.85 ± 0.20) × 1010
(16.04)
apKAH values from ref. 4.
bStatistically corrected values in parentheses.
cFirst-order rate constant.
dEquation [14].
eEquation [6] + kB[B].
kf2BH[BH+] not being negligible in eq. [1] and result from
kf1w × kf2BH. Independent GAC is another possibility.
However, within the experimental data available for com-
pounds 1 and 2, a definite choice cannot be made between
eqs. [14] and [15]. Equation [15] performed better with 1-
NPU but there is no considerable improvement over
eq. [14]. In contrast,with ureido ester 2-NPU eq. [14] is defi-
nitely the better choice as eq. [15] fails. With the stronger
buffer bases, acid catalysis becomes unimportant. In acetate
and cacodylate buffers, either a straightforward equation for
GBC (adding kB × [B] to eqs. [5] or [12]) best described the
results or the more complex equations gave insignificant im-
provement together with unrealistic additional constants.5
For compound 3-NPU the following modification of
eq. [12] was preferred in the case of the more acidic buffers
such as acid phosphate, glycine, and formate (the latter gave
large errors for the BH constants):
The data obtained in acetate were best described by an
equation assuming independent GBC. In cacodylate and
neutral phosphate only a few measurements at low fraction
base could be taken due to the fast reaction and the rate con-
stants were obtained by assuming independent GBC.
The rate constants for GBC listed in Tables 3–5 yield the
Brønsted plots shown in Fig. 6. The lines drawn are linear
fits obtained from the data for the general bases with pK’s
between 2 and 6.5. The data for ureido ester 3-NPU in acid
phosphate and glycine buffers deviate strongly from the
Brønsted plot and the β-value was determined from the re-
maining four data points. The slopes of the lines are pre-
sented in Table 6. As observed before for the cyclization of
ethyl 5-phenylhydantoates, the points for water catalysis fall
on the correlation lines. It is noteworthy that the data for
OH– catalysis repeat the pattern observed with the ω-methyl
and ω-phenyl analogues of compounds 1-NPU–3-NPU. Es-
ters 1-NPU and 2-NPU exhibit strong positive deviations
while the point for 3-NPU fits the line, a behaviour parallel
to the SKIE results.
[16] kobs
(k1w + k1H[H+ ])(k2w + k2H[H+]) + kB[B] + k1BH[BH+]
=
1 + k2H[H+ ] + k2BH[BH+ ]
The present series of Brønsted linear relationships for three
NPU hydantoates with increasing numbers of methyl groups
revealed an unexpected feature: the gem-dimethyl effect on
+ kOH[OH− ]
5 It should be noted also that eq. [13] on the one hand and eqs. [14] and [15] on the other give different kB constants in principle. With
eq. [13] kB is approached at low [OH–], while with eqs. [14] and [15] kB is approached at high [OH–] (low [BH+]).
© 1999 NRC Canada

858
Can. J. Chem. Vol. 77, 1999
Table 5. Buffer catalysis data for the cyclization of ethyl 2,2,3-trimethyl-5-(4-nitrophenyl)-hydantoate, 3-NPU, at 25 oC
and ionic strength 1.0 M.
Conc. range
mol dm3
104kB
104k2BH(k1BH
)
a,b
dm3 mol–1 s–1b
dm3 mol–1 s–1
pKAH
–1.74
Buffer acid
H3O+
Runs
% base
11.0 ± 1.8c
(–1.26)
1.78
(2.26)
H3PO4
0.01–1
0.01–1
0.1–1
0.1–1
0.1–1
0.1–1
0.1–1
0.01–1
0.1–0.5
0.1–0.5
0.1–0.5
0.01–0.5
0.01–0.5
0.01–0.2
0.01–0.3
0.008–0.2
0.008–0.048
6
30
50
70
90
30
50
70
90
30
50
70
30
50
70
20
10
15
492 ± 104d
73 700 ± 44 200
(992 ± 426)
6
5
5
3
3
3
4
4
4
4
9
9
4
4
8
4
H3N+CH2CO2H
2.45
759 ± 134d
60 100 ± 26 100
(518 ± 258)
HCO2H
3.57
4.62
165 ± 7.2d
172e
3350 ± 2880
(16.3 ± 6.2)
CH3CO2H
1190 ± 40e
(CH3)AsO2H
H2PO4
6.19
6.48
17 500 ± 1500 f
14 900 ± 5100 f
(7450)
–
H2O
15.74
(5.20 ± 0.37) × 1010
(16.04)
apKAH values from ref. 4.
bStatistically corrected values in parentheses.
cFirst-order rate constant
dEquation [16].
eEquation [12] + kB[B].
fSee text.
Fig. 6. Brønsted plots of the rates GBC for the cyclization of 1-
NPU (᭝), 2-NPU (᭹), and 3-NPU (ٗ). See text for linear
relationships.
Table 6. Brønsted β values for general base catalysis of the
cyclization of ethyl p-nitrophenylhydantoates.
Compound
β
No. of points
r
1-NPU
2-NPU
3-NPU
0.50 a
0.58
0.63
5
5
4
0.9559
0.9866
0.9638
aIf the data for neutral phosphate are discarded, a slope of 0.60 is
obtained in which case the point for alkaline phosphate falls on the
correlation line.
the rates for GBC showed the same saturation behaviour as
observed with the OH– catalysis. Figure 6 demonstrates
how, except for the deviating points acid phosphate and
glycine, the kB values for 3-NPU are insignificantly larger
than those for 2-NPU, though the increases for 2-NPU over
1-NPU amount to almost two orders of magnitude. This fact
indicates that our previous conclusion (3) on the mechanism
of GBC of the cyclization of ω-phenylhydantoic esters is in-
correct. No data could be obtained for 1-PUE. The similarity
in the β values of ca. 0.6 for 2-PUE and for 3-PUE was in-
terpreted as evidence of the same mechanism and the close
catalytic constants as due to an early transition state with re-
spect to ring formation, namely r.d. formation of T– con-
certed with deprotonation of the ureido group. If the same
mechanism applies to NPU, obviously it does not fit the
large gem-dimethyl effect observed between 1-NPU and 2-
NPU. The latter is compatible with two other possibilities.
Either GBC of 3-NPU involves a different mechanism than
that of 1- and 2-NPU and the similar β values are accidental
or all three involve GAC of the breakdown of T–, the loss of
© 1999 NRC Canada

Blagoeva et al.
859
the gem-dimethyl effect with 3-NPU being caused by steric
hindrance to protonation. Actually the question can be an-
swered within the framework of the mechanisms accepted
for the rate profiles. GBC must comply with the rate-
determining step of the reaction as a whole. Thus under con-
ditions where formation of T– is rate determining, the GBC
reaction cannot proceed with r.d. breakdown unless the r.d.s.
for the whole reaction is changed, a situation discussed
above in relation to the curvature of the kobs against fraction
base plots. In the particular case of NPU, the r.d.s. in the
various regions of the rate profiles are established above
with a fair degree of certainty. The rate profiles (Fig. 1) in-
dicate that the in whole region above pH 2.5 for esters
1-NPU and 2-NPU (the plateau and kOHb) where GBC was
actually measured the reaction proceeds with rate-
determining formation of T. Alternatively, under conditions
where an equilibrium between reactant and T– is established,
as assumed for 3-NPU above pH ca. 4, GBC formation of T–
cannot accelerate the reaction because the [T–]/[UH] ratio
becomes time independent. Thus we have to conclude the 3-
NPU on the one hand and 1-NPU–2-NPU on the other react
by different mechanisms and that the similarity in β values is
coincidental. Supporting evidence is the fact that the data for
of Chemistry for a Journals Grant for International Authors
(to I.G.P.).
References
1. I.B. Blagoeva, D.T. Tashev, and A.J. Kirby. J. Chem. Soc.
Perkin Trans. 1, 1157 (1989).
2. C.K. Ingold. J. Chem. Soc. 119, 305 (1921); C.K. Ingold, S.
Sako, and J.F.Thorpe. J. Chem. Soc. 1117 (1922); N.L.
Allinger and V. Zalkow. J. Org. Chem. 25, 701 (1960); I.B.
Blagoeva, B.J. Kurtev, and I.G. Pojarlieff. J. Chem. Soc.
Perkin Trans. 2, 1115 (1979); A.J. Kirby. Adv. Phys. Org.
Chem. 17, 183 (1980); L. Mandolini. Adv. Phys. Org. Chem.
22, 17 (1986).
3. E. Atay, I.B. Blagoeva, A.J. Kirby, and I.G. Pojarlieff. J.
Chem. Soc. Perkin Trans. 2, 2289 (1998).
4. E. Atay, I.B. Blagoeva, and I.G. Pojarlieff. Dokl. Bolg. Akad.
Nauk, 52, No. 2, (1999). In press.
5. O. Belafdal, M. Bergon, M. Calmon, and J.P. Calmon. J. Org.
Chem. 54, 4193 (1989).
6. A.H. Koedjikov, I.B. Blagoeva, I.G. Pojarlieff, and A.J. Kirby.
J. Chem. Soc. Perkin Trans. 2, 2479 (1996).
7. I.B. Blagoeva. J. Chem. Soc. Perkin Trans. 2, 127 (1987).
8. A.K. Covington, R.A. Robinson, and R.G. Bates. J. Phys.
Chem. 70, 3820 (1966); J.C. Fishbein and W.P. Jencks. J. Am.
Chem.Soc. 110, 5075 (1988).
9. D.F. Detar. J. Am. Chem. Soc. 104, 7205 (1982).
10. F. Hibbert. Adv. Phys. Org. Chem. 22, 113 (1986).
11. J. Kavalek, V. Sterba, and S. El Bahaie. Collect. Czech. Chem.
Commun. 48, 1753 (1983).
2–
GBC by H2PO4 and glycine with 3-NPU deviate from the
log kB / pK correlation; in the low pH region a change in the
r.d.s. is indicated by the rate profile.
Thus the best explanation for the loss of the gem-dimethyl
effect in the GBC cyclizations of 3 appears once again to be
a change to r.d. breakdown of the tetrahedral intermediate
caused by steric hindrance to proton transfer as assumed by
us for the OH– catalysis of other ureido esters. Thus GBC
with 1-NPU and 2-NPU has to be attributed to rate-
determining formation of T– concerted with deprotonation of
the ureido group, structure 5, while that with 3-NPU to the
general acid-catalyzed breakdown of T–, structure 7.6
12. W.P. Jencks. Chem. Rev. 72, 702 (1972).
13. W.P. Jencks. Acc. Chem. Res. 9, 425 (1976).
14. K.J. Davies and M.I. Page. J. Chem. Soc. Chem. Commun.
1448 (1990).
15. T.C. Bruice. Annu. Rev. Biochem. 45, 331 (1976).
16. R.A. More O’Ferrall. J. Chem. Soc. B: 274 (1970).
17. J. Kavalek, V. Machacek, G. Svobodova, and V. Sterba. Col-
lect. Chech. Chem. Commun. 51, 375 (1986).
Acknowledgments
We thank the National Foundation for Scientific Research
of Bulgaria for funding this research and the Royal Society
6 Tables of observed rate coefficients for 1-, 2-, and 3-NPU have been deposited as supplementary material (6 pages), which may be pur-
chased from: The Depository of Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, Canada, K1A
0S2.
© 1999 NRC Canada