Catalysis of the methanolysis of acetylimidazole by lanthanum triflate

Article abstract of DOI:10.1139/v00-128

Methanolysis of acetylimidazole (1) and N-acetylimidazolepentamine-Co(III) (2) was found to be markedly accelerated in the presence of La(OTf)₃. Potentiometric titration of a solution of La³⁺(OTf^-)₃ gave a pK_a for the metal bound CH₃OH of 7.22. The kinetics of methanolysis of 1 and 2 were measured at 25°C at various pH under buffered conditions as a function of increasing La³⁺. Analysis of both the kinetic and potentiometric data indicates that the catalytically active species is a La³⁺-dimer, bridged by two methoxides, (CH₃OH)_nLa³⁺(CH₃O^-) ₂La³⁺(CH₃OH)ⁿ. The maximum second-order rate constants for attack of the dimer on 1 and 2 are 1.50 × 10³ M^-1 s^-1 and 1.42 × 10² M^-1 s^-1 respectively and both processes adhere to titration of a La³⁺(CH₃OH) to generate the active form. The results are explained in terms of a mechanism where the methoxy-bridged La³⁺ dimer transiently breaks a La³⁺-OCH₃ bond to expose both a CH₃O^- nucleophile and a La³⁺ which can act as a Lewis acid. Unlike the situation in water, the methanol results indicate that the medium greatly stabilizes and solubilizes the active dimer without the necessity of creating specially designed ligands to stabilize the dinuclear core.

Full text of DOI:10.1139/v00-128

1
247
COMMUNICATION
Catalysis of the methanolysis of acetylimidazole
by lanthanum triflate
Alexei A. Neverov and R.S. Brown
Abstract: Methanolysis of acetylimidazole (1) and N-acetylimidazolepentamine-Co(III) (2) was found to be markedly
3
+
–
accelerated in the presence of La(OTf) . Potentiometric titration of a solution of La (OTf ) gave a pK for the metal
3
3
a
bound CH OH of 7.22. The kinetics of methanolysis of 1 and 2 were measured at 25°C at various pH under buffered
3
3
+
conditions as a function of increasing La . Analysis of both the kinetic and potentiometric data indicates that the cata-
lytically active species is a La -dimer, bridged by two methoxides, (CH OH) La (CH O ) La (CH OH) . The maxi-
mum second-order rate constants for attack of the dimer on 1 and 2 are 1.50 × 10 M
respectively and both processes adhere to titration of a La (CH OH) to generate the active form. The results are ex-
plained in terms of a mechanism where the methoxy-bridged La dimer transiently breaks a La -OCH bond to ex-
pose both a CH O nucleophile and a La which can act as a Lewis acid. Unlike the situation in water, the methanol
3
+
3+
–
3+
3
n
3
2
3
n
3
–1 –1
2
–1 –1
s
and 1.42 × 10 M
s
3
+
3
3
+
3+
3
–
3+
3
results indicate that the medium greatly stabilizes and solubilizes the active dimer without the necessity of creating
specially designed ligands to stabilize the dinuclear core.
Key words: lanthanum ion, catalysis, methanolysis, activated amide.
Résumé : On a observé que la méthanolyse de l’acétylimidazole (1) et du N-acétylimidazopepentamine-Co(III) (2) est
3
+
–
fortement accéléré par la présence de La(OTf) . Un titrage potentiométrique d’une solution de La (OTf ) permet
3
3
d’établir que le pK du métal lié au CH OH est de 7,22. Opérant à divers pH, dans des conditions tamponnées, on a
a
3
3
+
mesuré les cinétiques de méthanolyse des composés 1 et 2 en fonction d’une concentration croissante de La .
L’analyse des données tant cinétiques que potentiométriques indique que l’espèce catalytiquement active est un dimère
de La3 comportant un pont par deux molécules de méthylates, (CH O) La (CH O ) La (CH OH) . Les constantes de
+
3+
–
3+
3
n
3
2
3
n
3
–1 –1
2
–1
vitesse maximales pour les attaques des composés 1 et 2 sont respectivement de 1,50 × 10 M
s
et 1,42 × 10 M
–
1
3+
s
et les deux processus adhèrent au titrage d’un La (CH OH) pour générer la forme active. On explique les résultats
3
3
+
en fonction d’un mécanisme dans lequel il se produit une dissociation transitoire d’une liaison La -OCH du dimère
3
de La3 relié par un pont méthoxy qui permet d’exposer un nucléophile CH O et d’un La qui peut agir comme acide
+
–
3+
3
de Lewis. Par opposition à la situation dans l’eau, les résultats obtenus avec le méthanol indiquent que le milieu stabi-
lise grandement et solubilise le dimère actif sans avoir à créer des ligands appropriés pour stabiliser le coeur
dinucléaire.
Mots clés : ion lanthane, catalyse, méthanolyse, amide activé. Communication 1250
X+
–
Much recent attention was focussed on the remarkable ac-
celeration provided by lanthanide cations (Ln ) for the hy-
of Ln (OH ) by encapsulation into cyclodextrins (1h, 2a),
n
X+
macrocycles (1g), or micelles (1d) but most research has
drolysis of phosphate esters (1) and, to a lesser degree,
amides and peptides (2). However mechanistic study of
these Ln -promoted hydrolyses is hindered by the fact that
in water, formation of insoluble hydroxides and gels of poor
definition is a problem, particularly under basic conditions.
A few attempts have been made to stabilize the active forms
been done at neutral or acidic conditions which limits the
catalytic ability of those ions. Only a few examples of Ln -
X+
X+
catalysis of amide bond hydrolysis are known (2), probably
X+
due to the poorer M -coordinating ability of the neutral >
N-C=O moiety compared to that of negatively charged di-
and monophosphate esters. Indeed, the reported cases of
X+
strong M -promoted hydrolysis of amide (peptide) sub-
strates are limited to those having additional binding centers
that promote preassociation of the metal (3).
Herein we focus on the La -catalyzed methanolysis of
Received June 21, 2000. Published on the NRC Research
Press website on August 22, 2000.
3
+
1
A.A. Neverov and R.S. Brown. Department of Chemistry,
the simple activated amide, acetylimidazole, (1), and its
Queen’s University, Kingston, ON K7L 3N6, Canada.
3+
ligand exchange inert Co -complex, N-acetylimidazolepen-
3
+
1Author to whom correspondence may be addressed.
Telephone and Fax: (613) 533-2624.
e-mail: rsbrown@chem.queensu.ca
tamine-Co(III), (AcImCo(NH3)5 , (2)) (4). These were cho-
sen because their aqueous hydrolyses are well understood,
and because it has yet to be shown that the parent compound
Can. J. Chem. 78: 1247–1250 (2000)
© 2000 NRC Canada

1
248
Can. J. Chem. Vol. 78, 2000
(
1) is subject to metal ion catalysis in water. This combina-
Fig. 1. (a) Plots of pseudo-first-order rate constants for the
methanolysis of AcIm (pHMeOH = 6.7, ᭿, left axis)) and
AcImCo(NH3)53 (pHMeOH = 7.6, µ = 0.2 (NaClO4) ᭹, right
axis) as a function of concentration of [La(OTf)3] at 25°C.
(b) Plots of the pseudo-first-order rate constants for methanolysis
tion allows us to investigate the role of electrostatic effects
and the possibility of catalysis arising from the binding of
the metal ion to the distal nitrogen of 1. We report here that
in methanol, homogeneous solutions can be obtained
throughout the entire pH range surrounding ionization of
+
3
+
of AcIm (pHMeOH = 7.33 ᭿g, left axis) and AcImCo(NH3)5
(pHMeOH = 7.6 ᭹, right axis) vs. [La(OTf)3], at 25°C.
3
+
La (CH OH). Moreover, through the use of buffered media
3
to control the pH, which can be easily measured using a re-
cently reported glass electrode technique (5), we can study
the reaction kinetics in methanol as simply as in water with-
out the annoying problems of metal–alkoxide precipitation.
To the best of our knowledge, this is the first time such an
approach has been used for kinetic investigations in metha-
2
nol solutions.
Our preliminary results show that the methanolysis rates3
of 1 and 2 markedly increase in the presence of lanthanum
trifluoromethanesulphonate (La(OTf) ), and this process is
3
very sensitive to pHMeOH. Pseudo-first-order rate constants,
kobs, obtained for the methanolysis of 1 and 2 (Fig. 1a, b)
show strong catalysis by La(OTf) , with the effect being lin-
3
early proportional to [La(OTf) ] at higher concentrations of
3
metal salt but with significant curvature at concentrations
–
4
lower than ~2 × 10 M. This indicates the involvement of
second- or possibly higher order terms in [La(OTf) ]. Unfor-
3
tunately, direct analysis of all the experimental k
vs.
obs
3
+
[
La ] data does not provide us with reliable kinetic parame-
t
ters, particularly in the case of 2 due to the high rate of the
background reaction. However the fact that the kobs vs.
[
La(OTf) ] dependencies look very similar for both sub-
3 t
strates eliminates the possibility that the higher reaction or-
der in [La ] is due to complexation of a La ion by the
3
+
3+
distal nitrogen of 1, and rather indicates the catalysis is due
3
+
to a concentration dependent change in speciation of La in
methanol such as formation of active dimers. Dimeric struc-
tures were previously observed in methanol solutions of
1
39
35
LaCl using La and Cl NMR (6).
3
3
–4
Analysis of the linear parts of the plots ([La ] > 2 × 10 M,
t
Fig. 1b) gives the second-order rate constants (k ) which are
2
then plotted vs. pHMeOH (as in Fig. 2) to reveal similar pro-
this indicates that the kinetically active species is a dimeric
complex consisting of two solvated La ions and two
methoxide anions as shown in Scheme 1.
files for both substrates with apparent kinetic pK s of
a
3
+
7
–7.5. In a separate experiment (not shown), potentiometric
titration of La(OTf) (1.33 mmol) between pH 6 and
3
MeOH
9
shows ionization of one proton per metal ion with
The equations for the equilibrium constants and the con-
servation of mass pertaining to the process in Scheme 1
MeOH
pKa
= 7.22. In combination with the pH-rate profiles
(
Fig. 2) and kobs vs. [La(OTf) ] dependencies (Fig. 1a, b),
provide the expression given in eq. [1] where k is the
3
o
2
N-methylimidazole (pKaMeOH = 7.60) and N-ethylmorpholine (pK
MeOH
= 8.28) were used to buffer methanol solutions. Total buffer concen-
a
trations were varied between 10–3 M and 10 M, and no buffer catalysis was observed. The [CH OH ] was determined using a combina-
–2
+
3
2
tion (glass/calomel) electrode, calibrated with Fisher Certified standard aqueous buffers. Values of pHMeOH were calculated by adding a
correction constant (2.24) to the experimental meter reading. (R.G. Bates. In Determination of pH. Theory and practice. Wiley, New York.
(
1973); E. Bosch, private communication.)
3
UV kinetics were monitored by observing the rate of loss of starting material at 240 nm, under pseudo-first-order conditions of excess metal
[amide] = 0.5 – 1.0 × 10–4 M) at 25°C. Pseudo-first-order rate constants (kobs) were evaluated by fitting the Abs. vs. time traces to a stan-
dard exponential model.
(
©
2000 NRC Canada

Communication
1249
Scheme 1.
3
–1 –1
(–14.8 ± 0.1)
2
2
Fig. 2. pH–Second-order rate constant profiles for La(OTf) -cata-
k = (1.50 ± 0.1) × 10 M s , K = 10
M (ki-
3
a
lyzed methanolysis of AcIm (᭿) and AcImCo(NH3)53 (᭹).
Curves are the result of NLLSQ fitting of experimental data to
eq. [3].
+
–1 –1
netic pK = 7.40) for 1, and k = (1.42 ± 0.02) × 10 M s ,
M (kinetic pK = 7.15) for 2. The lines
through the data in Fig. 2 are those generated by the fits to
a
(
–14.30 ± 0.02)
2
K = 10
a
a
eq. [3]. That the computed kinetic pK s are in good agree-
a
MeOH
ment with the pKa
obtained from the potentiometric ti-
tration lends strong credence to the idea that both processes
relate to the ionization of metal-coordinated methanol. Fur-
ther, the fact that only one inflection point is present in the
potentiometric titration curve indicates strong cooperation
between dissociation of both protons from a dimeric species.
While the exact structure of this dimer is uncertain, most
3
+
reasonable is one having two La ions bridged by two
methoxide anions as in 3.
Similar double-bridge structures were suggested for
La Cl (CH OH) (6) (with bridging chloride ions) and a
2
6
3
10
spontaneous rate constant for methanolysis in the absence of
metal ion.
3+
La -peroxide complex (with bridging peroxide dianions)
(
7). Hurst et al. (8), from a study of the hydrolysis of an
+
2
[
1]
kobs = k + (kK [H ] /K )((1 + 8(K +
0
a
d
a
3+
RNA model dimer promoted by La in water, proposed a
similar catalytically active metal dimer structure having two
lanthanide ions bridged by five HO .
+
2
3+
+ 2 0.5
2
+ 2
2
[
H ] )K [La ] /[H ] ) – 1) /(4(K + [H ] ))
d t a
–
3
+
The linear part of the plots of kobs vs. [La ] , (Fig. 1b),
used to calculate the second-order rate constant k , describe
the conditions where δkobs/δ[La ] is constant (eq. [2]),
t
3+
Comparison of the second-order rate constants for La -
2
3
–1 –1
3
+
promoted methanolysis of 1 and 2 ((1.50 ± 0.1) × 10 M s
and (1.42 ± 0.02) × 10 M s ) clearly indicates that coor-
dination of the (NH ) Co to the distal nitrogen does not
provide any enhancement of the catalytic effect of La .
t
2
–1 –1
3
+
k2 = δkobs/δ[La] = kαβ(1 – 1/(1 – β[La ] )0.5)
3
+
[
2]
t
t
3 5
3
+
+
2
+ 2 2
where α = K [H ] /(16K (K + [H ] ) ), and β = 8K (K +
H ] )/[H ] .
To satisfy the condition that δk /δ[La ] is constant, the
range of [La ] for analysis must be chosen such that
1 –β [La ] ) must be much bigger than 1, thereby provid-
a
d
d
d
a
Even though 2 has a much better leaving group than does 1
+
2
+ 2
[
–
3+
–
(
pK = 10.0 for ImCo(NH ) (4) vs. 14 for Im (9)), the
a 3 5
3
+
3+
obs
t
rate of its La -promoted reaction is 10-fold slower. This is
3
+
–
t
in contrast to our results for CH O -catalyzed methanolysis
of 2 where the second-order rate constant (k
10 M s ) is four orders of magnitude higher than that for
CH O attack on 1 (k
3
3
+
0.5
(
t
= 4.69 ×
CH3O
ing an expression relating k to K , the ionization constant of
7
–1 –1
2
a
3
+
La (HOCH ), eq. [3].
–
3
–1 –1 4
3
= 7.9 × 10 M s ). Strong
3
CH3O
+
2
acceleration of methoxide attack on 2 relative to 1 can be
explained by a combination of leaving group effects and
electrostatic interaction between the positively charged sub-
strate (2) and a negatively charged nucleophile. The latter
[
3]
k = kK /(2(K + [H ] ))
2 a a
NLLSQ fitting of pH–k data to eq. [3] gives the follow-
2
ing parameters:
4
A.A. Neverov, P. Montoya-Paelez, and R.S. Brown. Submitted for publication.
©
2000 NRC Canada

1
250
Can. J. Chem. Vol. 78, 2000
3
+
result thus shows that the La -promoted methanolysis
mechanism cannot involve attack of free methoxide ion
since that should result in much higher rate constants for 2
than for 1, contrary to what is observed. The slower reaction of
the lanthanide reaction with 2 can be accommodated by a pro-
References
1. (a) A. Blasko and T.C. Bruice. Acc. Chem. Res. 32, 475 (1999)
and refs. therein; (b) N.H. Williams, B. Tasaki, M. Wall, and J.
Chin. Acc. Chem. Res. 32, 485 (1999) and refs. therein;
(c) R.A. Moss, B.D. Park, P. Scrimin, and G. Ghirlanda. J.
Chem. Soc. Chem. Comm. 1627 (1995); (d) R.A. Moss, J.
Zhang, and K. Bracken. J. Chem. Soc. Chem. Comm. 1639
3+
-
cess involving the dimeric complex (La (CH O ) (CH OH) )
3), having a total positive charge of +4, wherein electro-
static repulsion causes a decrease in the rate for the posi-
tively charged substrate 2 relative to neutral 1.
It has been shown in the case of a structurally similar
2
3
2
3
m
(
(1997); (e) J. Sumaoka, S. Miyama, and M. Komiyama. J. Chem.
Soc. Chem. Comm. 1755 (1994); (f) J.R. Morrow, L.A. Buttrey,
and K.A. Berback. Inorg. Chem. 31, 16 (1992); (g) J.R. Mor-
row, L.A. Buttrey, V.M. Shelton, and K.A. Berback. J. Am.
Chem. Soc. 114,1903 (1992); (h) R. Breslow and B. Zhang. J.
Am. Chem. Soc. 116, 7893 (1994); (i) N. Takeda, M. Irisawa,
and M. Komiyama. J. Chem. Soc. Chem. Comm. 2773 (1994);
3
+
Co -hydroxide dimeric complex, that bridging hydroxides
1
0
do not possess nucleophilic activity, being at least 10 times
less active than the deprotonated oxo-bridged forms (10).
For the present methoxy-bridged dimer (3) deprotonation is
impossible so the only way to generate an active nucleophile
within the dimeric complex requires temporary breakage of
a lanthanum—methoxy bond. This creates a scissors-like
(j) P. Gómez-Tagle and A.K. Yatsimirsky. J. Chem. Soc. Dalton
Trans. 2957 (1998).
2
.
(a) T. Takarada, R. Takahashi, M. Yashiro, and M. Komiyama.
J. Phys. Org. Chem. 11, 41 (1998); (b) M. Yashiro, T. Takarada,
S. Miyama, and M. Komiyama. J. Chem. Soc. Chem.
Commun. 1757 (1994).
3
+
–
structure where the La -bound CH O is delivered to the
3
3
+
C=O carbon with the second La functioning as a Lewis
acid to stabilize the developing negative charge on the car-
bonyl oxygen as in 4.
3. (a) T.J. Przystas and T.H. Fife. J. Chem. Soc. Perkin Trans. 2,
3
2
93 (1990) and refs. therein; (b) J. Suh. Acc. Chem. Res. 25,
73 (1992) and refs. therein; (c) P. Tecilla, U. Tonellato, F.
Veronese, F. Fellugg, and P. Scrimin. J. Org. Chem. 62, 7621
1997); (d) J. Chin, V. Jubian, and K. Mrejen. J. Chem. Soc.
(
Chem. Commun. 1326 (1990); (e) L. Meriwether and F.
Westheimer. J. Am. Chem. Soc. 78, 5119 (1956); (f) T.J. Grant
and R.W. Hay. Aust. J. Chem. 18, 1189 (1965); (g) J.T. Groves
and R.M. Dias. J. Am. Chem. Soc. 101, 1033 (1979); (h) J.T.
Groves and R.R. Chambers. J. Am. Chem. Soc. 106, 630
(
(
1984); (i) J.T. Groves and J.R. Olson. Inorg. Chem. 24, 2715
1985); (j) L. Sayre, K.V. Reddy, A.R. Jacobson, and W. Tang.
Inorg. Chem. 31, 935 (1992); (k) T.H. Fife. Acc. Chem. Res.
2
1
6, 325 (1993); (l) T.H. Fife and R. Bembi. J. Am. Chem. Soc.
15, 11358 (1993); (m) T.N. Parac and N.M. Kostic′. J. Am.
The above study indicates several beneficial features stem-
3
+
Chem. Soc. 118, 51 (1996); (n) L. Singh and R.N. Ram. J.
Org. Chem. 59, 710 (1994); (o) P.A. Sutton and D.A.
Buckingham. Acc. Chem. Res. 20, 357 (1987); (p) E. Kimura,
I. Nakamura, T. Koike, M. Shionoya, Y. Kodama, T. Ikeda, and
M. Shiro. J. Am. Chem. Soc. 116, 4764 (1994); (q) T.N. Parac,
G.M. Ullman, and N.M. Kostic′. J. Am. Chem. Soc. 121, 3127
ming from studying La -promoted methanolysis relative to
3
+
La in water. Methanolyses have not been studied as exten-
sively as hydrolyses, undoubtedly due to the problems of
controlling and measuring the pH which now have been
overcome (5). By a combination of potentiometric titration
and kinetic data, we have shown that catalytically active
(
1999), and refs. therein; (r) A.M. Ridder and R.M. Kellogg.
3+
La dimers (3) are formed above the pK of the metal bound
a
In Comprehensive supramolecular chemistry. Edited by Y. Murakami.
Vol. 4. Supramolecular reactivity and transport: bioorganic
systems. Chap. 11. Elsevier Science Ltd., Oxford. 1996; (s) J.
Chin. Acc. Chem. Res. 24, 145 (1991), and refs. therein.
. J. MacB. Harrowfield, V. Norris, and A.M. Sargeson. J. Am.
Chem. Soc. 7282 (1976).
. (a) E. Bosch, F. Rived, M. Rosés, and J. Sales. J. Chem. Soc.
Perkin Trans. 2, 1953 (1999); (b) E. Bosch, P. Bou, H. Allemann,
and M. Rosés. Anal. Chem. 3651 (1996); (c) F. Rived, M.
Rosés, and E. Bosch. Anal. Chim. Acta, 374, 309 (1998).
CH OH. The fact that these dimers are far more soluble in
3
3+
methanol than the corresponding La -hydroxides are in aque-
ous solution simplifies kinetic studies and obviates the ne-
cessity of creating complex ligands to stabilize the dinuclear
catalytically active core, as is the case in aqueous studies (1,
4
5
8
, 10). The success observed here with the simple amides
3
+
and the fact that the La -dimer acts on the C=O unit sug-
3
+
gests that La -catalyzed methanolyses of more biologically
relevant amides, peptides, and oligonucleotides would be in-
teresting, and we will report on these studies in due course.
6. L.S. Smith, D.C. McCain, and D.L. Wertz. J. Am. Chem. Soc.
8, 5125 (1976).
9
7
8
. B.K. Takasaki and J. Chin. J. Am. Chem. Soc. 117, 8582 (1995).
. P. Hurst, B.K. Takasaki, and J. Chin. J. Am. Chem. Soc. 118,
9982 (1996).
Acknowledgment
The authors gratefully acknowledge financial support of
this research by the Natural Sciences and Engineering Re-
search Council of Canada (NSERC). We thank Dr. Elisabeth
Bosch for personal communications on determining the pH
of buffered methanol solutions.
9. R.J. Sundberg and R.B. Martin. Chem. Rev. 74, 471 (1974).
10. (a) D. Wahnon, A.-M. Lebuis, and J. Chin. Angew. Chem. Int.
Engl. 34, 2412 (1995); (b) N.H. Williams, A.-M. Lebuis, and
J. Chin. J. Am. Chem. Soc. 121, 3341 (1999).
©
2000 NRC Canada