Formation and decomposition of N-alkylnaphthalimides: Experimental evidences and ab initio description of the reaction pathways

Article abstract of DOI:10.1002/poc.1768

The kinetics of hydrolysis of 1,8-N-butyl-naphthalimide (1,8-NBN) to 1,8-N-butyl-naphthalamide (1,8-NBAmide) and of 2,3-N-butyl-naphthalimide (2,3-NBN) to 2,3-N-butyl-naphthalamide (2,3-NBAmide), as well as the formation of the respective anhydrides from the amides were investigated in a wide acidity range. 1,8-NBN equilibrates with 1,8-NBAmide in mild alkali. Under the same conditions 2,3-NBN quantitatively yields 2,3-NBAmide. Over a wide range of acidities the reactions of the 1,8- and 2,3-N-butyl-naphthalamides (or imides) yield similar products but with widely different rates and at distinct pH's. Anhydride formation in acid was demonstrated for 1,8-NBAmide. The reactions mechanisms were rationalized in the manifold pathways of ab initio calculations. The differences in rates and pH ranges in the reactions of the 1,8- and 2,3-N-butyl-naphthalamides were attributed to differences in the stability of the tetrahedral intermediates in alkali as well as the relative stabilities of the five and six-membered ring intermediates. The rate of carboxylic acid assisted 1,8-N-Butyl-naphthalamide hydrolysis is one of the largest described for amide hydrolysis models. Copyright

Full text of DOI:10.1002/poc.1768

Research Article
Received: 21 December 2009,
Revised: 7 May 2010,
Accepted: 8 June 2010,
Published online in Wiley Online Library: 3 September 2010
(wileyonlinelibrary.com) DOI 10.1002/poc.1768
Formation and decomposition of
N-alkylnaphthalimides: experimental
evidences and ab initio description of the
reaction pathways
Teresa Cristina de Barrosay, Pedro Berci Filhob, Michel Loosc,d, Mario
a
a
a
*
´
Jose Politi , Hernan Chaimovich and Iolanda Midea Cuccovia
The kinetics of hydrolysis of 1,8-N-butyl-naphthalimide (1,8-NBN) to 1,8-N-butyl-naphthalamide (1,8-NBAmide) and of
2,3-N-butyl-naphthalimide (2,3-NBN) to 2,3-N-butyl-naphthalamide (2,3-NBAmide), as well as the formation of the
respective anhydrides from the amides were investigated in a wide acidity range. 1,8-NBN equilibrates with
1,8-NBAmide in mild alkali. Under the same conditions 2,3-NBN quantitatively yields 2,3-NBAmide. Over a wide
range of acidities the reactions of the 1,8- and 2,3-N-butyl-naphthalamides (or imides) yield similar products but with
widely different rates and at distinct pH’s. Anhydride formation in acid was demonstrated for 1,8-NBAmide. The
reactions mechanisms were rationalized in the manifold pathways of ab initio calculations. The differences in rates
and pH ranges in the reactions of the 1,8- and 2,3-N-butyl-naphthalamides were attributed to differences in the
stability of the tetrahedral intermediates in alkali as well as the relative stabilities of the five and six-membered ring
intermediates. The rate of carboxylic acid assisted 1,8-N-Butyl-naphthalamide hydrolysis is one of the largest
described for amide hydrolysis models. Copyright ß 2010 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: amide imide equilibria; anhydride from amides; kinetics imide hydrolysis; mechanism amide hydrolysis;
naphthalimide hydrolysis
Hydrolysis of 40-methoxysuccinanilic acid produces succinic
INTRODUCTION
anhydride and anisidine but an unstable equilibrium with
40-methoxy-N-phenylsuccinimide also occurs. In the case of
maleanilic acids, the low kinetic barrier for final product
formation prevents the observation of an intermediate imide.[13]
Hydrolyses of aromatic amides have been studied.[14–16] The
observed rate constant, kc, for acid hydrolysis of phthalamic acid
increases with the concentration of non-ionized carboxylic
Acid and alkaline amide hydrolysis, fundamental reactions in
biological systems, have been reviewed.[1–9] Acid catalysed
hydrolysis involves a water-assisted attack on the protonated
oxygen of the amide yielding a tetrahedral intermediate and
H3Oþ.[3–5,7] In the alkaline amide hydrolysis an anionic intermedi-
ate can either revert to reagents or proceed to products.[3–5]
Alkaline hydrolysis of amides may be first or second order in
OHꢀ.[4] Second order OHꢀ dependence can be related to the
formation of a doubly charged species, [3,4] although the existence
of a dianion has been disputed based on theoretical calculations.[5]
In some enzymatic reactions, the mobility restriction of the
substrate in the active site and the proximity of reactants in that
environment are the relevant catalytic factors.[8,9] This proximity
effect may lead to changes in mechanism or rate determinant
step.[10] Considering non-enzymatic systems, the presence of
double bond or a bulky group in a substrate containing a
carboxylic group near an amide decreases the degrees of freedom
of the reacting and catalytic groups and, as the in the case of
substituted N-alkyl maleamic acids, increases the reaction rate.[11]
Alkyl substituent that restrict the mobility of the relevant groups in
highly substituted amides may produce large rate accelerations.[12]
Hydrolyses of a series of dialkyl maleamic acids were studied in
a large range of conditions.[11] The rates of the acid catalysed
hydrolysis depend on the pKa of the carboxylic acid group and
are pH-independent from pH 0 to 2. [11–13] Anhydride formation
was proposed, and in some cases observed, below pH 5.[11]
˜
Correspondence to: I. M. Cuccovia, Instituto de Qu´ımica, Universidade de Sao
*
˜
Paulo, Av. Prof. Lineu Prestes 748, Sao Paulo, SP 05508-900, Brazil.
E-mail: imcuccov@quim.iq.usp.br
a
b
c
d
T. C. de Barros, M. J. Politi, H. Chaimovich, I. M. Cuccovia
Departamento de Bioqu´ımica, Instituto de Qu´ımica, Universidade de Sa˜o
Paulo, Brazil
P. B. Filho
Departamento de Qu´ımica e F´ısica, Instituto de Qu´ımica, Universidade de Sa˜o
Paulo, Campus de Sa˜o Carlos, Brazil
M. Loos
Departamento de Qu´ımica, Instituto de Qu´ımica, Universidade de Sao Paulo,
Brazil
˜
M. Loos
MLoos Consulting, Sao Paulo, Brazil
˜
y
ˆ
Present address: Instituto de Ciencias da Sau´de, Universidade Paulista, UNIP,
˜
Bauru, Sao Paulo, Brazil.
J. Phys. Org. Chem. 2011, 24 385–397
Copyright ß 2010 John Wiley & Sons, Ltd.

T. C. DE BARROS ET AL.
group.[14,15] The existence of the phthalic anhydride intermediate
was not demonstrated.[14]
were from Aldrich, 4-morpholineethanesulfonic acid, MES, N-
[tris(hydroxymethyl)methyl]-2-amino-ethanesulfonic acid, TES,
3-(cyclohexylamino)-1-propanesulfonic acid, CAPS and fluores-
camine, obtained from Sigma, were used as received.
n-Butylamine (Merck) was distilled.
1,8-NBN and 2,3-NBN were synthesized as described.[29–31]
1,8-NBAmide was prepared mixing 1,8-NBAnhydride (0.5 g in 1 ml
methanol) with 1.3 ml of n-butylamine and maintaining the
mixture 15 min at room temperature. The white precipitate,
formed upon cold water addition (2 ml) was filtered, washed with
cold water and dried in vacuum. Elemental analysis for the
butylammonium salt of 1,8-NBAmide was: Calcd. for C20H28O3N2:
C, 69.76%; H, 8.14%; N, 8.14%. Found: C, 69.51%; H, 8.18%; N,
8.43%.
2,3-NBAmide was prepared as described for 1,8-NBAmide.
After complete reaction, concentrated HCl was added and
2,3-NBAmide precipitate. The precipitate was thoroughly washed
with cold water and dried under vacuum over P2O5. Anal.: Calcd.
For C16H17O3N: C, 70.8%; H, 6.3%; N, 5.2%. Found: C, 70.1%; H,
6.3%; N, 5,7%.
All other reagents were P.A. grade or better and the solvents
were distilled. Aqueous solutions and buffers were prepared in
freshly glass bi-distilled water. The following buffers were used at
a final concentration of 0.01 M: sodium borate (pH 8.2–10.2), MES
(5.2–6.5) TES (pH 7.0–8.0), NaH2PO4 (pH 6–7), CAPS (10–11.3) and
sodium acetate (pH 3.4–5.4). HCl was used below pH 3 and NaOH
above pH 11.
The decomposition of N-methylphthalamic acid, NMA, at low pH,
can follow two different routes.[16] NMA cyclizes to N-methyl-
phthalimide, NMP, in an equilibrium reaction, at pH 7, the ratio NMP/
NMA increases reaching a maximum at pH 4.2 decreasing at lower
pHs. NMA can also form phthalic anhydride, which is rapidly
hydrolysed to phthalic acid.[16] The acid hydrolysis of NMA to phthalic
acid is dependent of the protonated carboxylic acid (pKa 3.7).[16]
Menger and Ladika described an interesting ‘synthetic
peptidase’, an aliphatic diacid amide showing intramolecular-
catalysed cleavage of the amide at pH 7 with participation of a
single carboxylic group with a pKa of 6.9.[17] Anhydride formation
and further hydrolysis to diacid was detected at neutral pH.[17]
The substrate geometry in this case is critical for the remarkable
rate of this amide cleavage.[17]
In amino acid substituted maleamic acids two carboxylic groups
can participate, over a limited pH range, in the catalysis of the
amide hydrolysis.[18] The protonation of the carboxylic group also
determines the rate of hydrolysis. The rate constant for amide
hydrolysis of carboxymethyl-di-isopropyl-maleamic acid, contain-
ing an aspartic acid at the amide nitrogen, is comparable to the
catalytic rate constant, kcat, for the hydrolysis of good substrates by
pepsin.[18] No anhydride was detected in the pH dependent
hydrolysis of N-(o-carbobenzoyl)-L-leucine, however this com-
pound ciclyses to N-phthaloylleucine in concentrated acid.[19]
At neutral pH limited imide formation from amides bearing a
proximal carboxylic acid has been observed.[16] At low pH’s,
cyclization to imide is competitive with the hydrolysis of the
amide, with anhydride formation. Only in a limited number of
cases anhydride formation can be experimentally verified.[16]
Recently the hydrolysis of several water-soluble, positively
charged, 1,4,5,8-naphthalenediimides (NDIs) were studied.[20] The
NDIs are completely hydrolysed at high pH (>12) to diamides,
while, from pH 10 to 12, one of the imides hydrolyses yielding
only the monoamide/monoimide. Below pH 10, the monoamide
concentration decreases and the diimide yield is almost 100%
at pH 8. The existence of mono and diimide is unusual and occurs
only in structural restricted compounds.[20]
Theoretical analysis of intramolecular catalysis by neighbouring
carboxylic acids of amide hydrolysis has stimulated considerable
interest.[21–28] In particular DFT calculations of benzamide
hydrolysis have suggests that the reaction, initiated by protonation
at the amidic oxygen, may be facilitated by two water molecules in
relatively concentrated acid, as reviewed recently.[7],[28]
Here, we describe the reactions of 1,8- and 2,3-N-butyl-
naphthalimide (1,8-NBN and 2,3-NBN) and 1,8- and 2,3-N-
butylnaphthalamide (2,3-NBAmide and 2,3-NBAmide) analyse
the formation and decomposition of those compounds over a
large range of pH, demonstrate that the reaction pathways are
comparable and only 1,8-NBAmide leads to stable anhydride. Both
experiment and theory refer to a system where amide hydrolysis
and/or imide formation involve the participation of a neighbouring
carboxylic group. The mechanisms are described in detail and can
be found in the manifold pathways of ab initio calculations.
Product isolation
1,8-NBN was isolated from aqueous solution after reacting
1,8-NBAmide (0.01 g) with NaOH 0.010 M (10.0 ml) during 24 h at
30 8C. The solid formed was filtered and washed with water. An
aqueous suspension of the product was extracted with CHCl3.
The organic phase was evaporated and the solid dried under
vacuum. The I.R. spectrum of the isolated product exhibited the
characteristics bands at n ¼ 1710 cmꢀ1 and n ¼ 1667 cmꢀ1 (C ¼O,
imide stretching) of 1,8-NBN.
1,8-NAnhydride was isolated from the reaction of 1,8-NBAmide
(0.01 g) with HCl 0.060 M (10.0 ml). The suspension was sonicated
until the crystals disappeared and then the solution became,
slowly, turbid and the product precipitated. The product was
extracted with CHCl3, the solvent evaporated and the solid was
dried under vacuum The i.r. spectrum of the isolated product has
the characteristic bands at: n ¼ 1786 cmꢀ1 and n ¼ 1749 cmꢀ1
(C ¼O of anhydride vibration and stretching) and n ¼ 1308 cmꢀ1
(C—O stretching of anhydride) corresponding to 1,8-NAnhydride.
Methods
n-Butylamine concentration was determined by measuring the
fluorescence of the reaction product with fluorescamine.[32]
A
standard curve was obtained by adding aliquots of a solution of
n-butylamine (5 ꢁ 10ꢀ4 M) in methanol to 2 ml of borate buffer
0.2 M, pH 10. Fluorescamine (5 mg/10 ml of acetone) was added
(0.5 ml) to the mixture and the fluorescence emission was
measured at l ¼ 475 nm (excitation l ¼ 390 nm) in a Hitachi
F-2000 Fluorescence Spectrophotometer.
EXPERIMENTAL SECTION
Kinetics
Materials
Determination of n-butylamine concentration in the reaction
mixture: Typically, 20 ml solutions containing the 1,8-NBAmide or
2,3-NBAmide (2 ꢁ 10ꢀ5 M) in a tightly closed reaction tube, at
1,8-Naphthalic anhydride, 1,8-NAnhydride, 2,3-Naphthalic
anhydride, 2,3-NAnhydride, boric acid, acetic acid, NaH2PO4
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 385–397

FORMATION AND DECOMPOSITION OF N-ALKYLNAPHTHALIMIDES
different pHs, were maintained at 30 or 50 8C. Aliquots were taken
at different times and maintained at ꢀ20 8C. At the end of the
reactions, the n-butylamine concentrations were determined
using 0.4 ml of the reaction sample, 1.6 ml of borate buffer
0.2 M, pH 10 and 0.5 ml fluorescamine (5 mg/10 ml acetone). For
reactions at pH < 1 aliquots of NaOH 5 M were added to the
borate buffer to reach pH 10. The observed rate constants, kc,
were determined from first order plots of fluorescence vs. time.
For imide hydrolysis and/or amide cyclization, kc were also
determined in a Beckman DU-7 or a Cary 50 Probe spectropho-
tometer using a quartz cell of 1.0 cm optical path. The hydrolysis
of 1,8-NBN (or cyclization of 1,8-NBAmide) was followed at
343 nm. The hydrolysis of 2,3-NBN (or cyclization of 2,3-NBAmide)
was followed at 364 nm or 262 nm. 1,8-NBN and 2,3-NBN stock
solutions (0.001 M or 5 ꢁ 10ꢀ4 M) were in DMSO. 1,8-NBAmide
and 2,3-NBAmide solutions were prepared in NaOH 0.1 M and
0.01 M, respectively, containing 15% DMSO. The final substrate
concentrations were between 1 and 2 ꢁ 10ꢀ5 M. All reactions
were followed for at least 5 half-lives and the observed rate
constants were calculated from first order plots. Rate constants
for 2,3-NBN hydrolysis above 0.020 M NaOH were obtained in a
Stopped Flow Spectrophotometer (Applied Photophysics).
Reactions were studied at 30 or 50 8C as described in the figure
legends. For each of the methods used (see above), the error in kc
was no more than 5%. For slow reactions (t0.5 > 12 h) the kc
obtained by either fluorescence or absorption, were equal and
the error reached 20%. All experiments were repeated, at least,
three times.
Computational procedures: ab initio calculations
The energetic of the chemical species involved in the reactions
were explored in the gas phase by ab initio quantum
chemical methods. The gas-phase geometries of all species
and transition states were fully optimized at the Restricted
Figure 1. Alkaline hydrolysis of 2,3-NBN at 30 8C. (A) Spectra of 2,3-NBN
(1.23 ꢁ 10ꢀ5 M) solution in NaOH (0.002 M) at different times: a – Initial
spectrum of 2, 3-NBN in water (t ¼ 0). b – Final spectrum after complete
reaction with NaOH (t ¼ 1). Intermediate spectra were taken each 3 min.
The final spectrum is identical to that of pure 2,3-NBAmide at the same
concentration. (B) pH effect on kc for 2,3-NBN alkaline hydrolysis. In the
inset is the plot of log kc vs. pH
*
Hartree Fock (RHF) level of theory using the standard 6-31G
basis-set.[33] Since we looked only for a qualitative insight
on the energetic of the process, the electron correlation energy
was estimated through single-point MP2 calculations at
the RHF geometries. For clarity, throughout the text, the
*
energy differences obtained using MP2/6-31G //HF/6-31G
*
calculations are presented and the RHF energy difference are
given between parentheses. The character of all stationary
Scheme 1. Reaction pathway of alkaline and water hydrolysis of 2,3-NBN
J. Phys. Org. Chem. 2011, 24 385–397
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

T. C. DE BARROS ET AL.
Figure 3. (A) Effect of pH on the kc of 2,3-NBAmide cyclization, (*), (D),
(^), (5), (~) or 2,3-NBN hydrolysis (*) at 50 8C (2.4 ꢁ 10ꢀ5 M). The
different symbols represent independent series of experiments and rate
constants measured by spectrophotometry ((*), (D), (^),(~),(*)) or
fluorescence (5). (B) Products fractions of 2,3-NBAmide reaction vs. pH
after 140 h at 50 8C. (*) %2,3-NBN; (D) %2,3-NBAmide; (*)
%n-butylamine
*
method, also at the HF/6-31G level. All calculations were
done using Gaussian 03.[34]
The energies of species of different stoichiometry were
always compared considering the same number of atoms and
electrons using H2O/OHꢀ/H3Oþ species (see Tables 1–4). For
anions the small basis set used only allowed rough reactivity
estimates.
In some cases solvation was estimated using the PCM
formalism, with the vacuum optimal molecular geometry at
the RHF or B3LYP level of theory.[35]
Figure 2. (A) Spectra of a solution of 2,3-NBN (2 ꢁ 10ꢀ5 M) at pH 6.5, at the
following times (hours): (a) 0; (b) 25; and (c) 58; (d) 145. Spectrum of
2,3-NBAmide solution (2 ꢁ 10ꢀ5 M), pH 6.5, at the following times (hours):
(A) 0; (B) 25; (C) 58; and (D) 145. (B) Reaction of 2,3-NBAmide followed by
absorbance change at 262 nm, at following pHs: (*) 3.0; (*) 4.0; (~) 4.5; (D)
5.0; (&) 5.5; (&) 6.0; (5) 6.5; (!) 7; and (^) 7.5. [2,3-NBAmide]¼ 2ꢁ 10ꢀ5 M.
(C) Reaction of 2,3-NBAmide followed by n-butylamine formation through
fluorescence measurements (see methods), at several pHs: (*) 0.0; (*) 1.0;
(D) 2.0; (~) 3.0; (5) 4.0; (&) 4.5; (&) 5.0; (^) 5.5; (^) 6.0; and (!) 6.5.
2,3-NBAmide¼ 1.85ꢁ 10ꢀ5 M. All experiments were done at 508C
In order to evaluate pKas of some intermediates, calculations
*
were done in the B3LYP formalism using the 6-31G base. The
pKas of the tetrahedral intermediates 24 and 124 (see Results,
Schemes 1 and 4), proposed in the mechanism of alkaline
hydrolysis of 1,8 and 2,3-NBN, were calculated using Eqns (1)
and (2), where AH and Aꢀ are the concentrations of the acid and
basic form of a generic weak acid.
1
DG0deprot;aq
(1)
(2)
points was determined calculating the second derivatives of
the potential energy with respect to atomic coordinates, at
the RHF level. The reaction paths from transition states were
determined using the Intrinsic Reaction Coordinate (IRC)
pKa ¼
2:303RT
DG0deprot;aq ¼ DGa0qðAHÞ ꢀ DGa0qðHþÞ ꢀ DGa0qðAꢀÞ
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 385–397

FORMATION AND DECOMPOSITION OF N-ALKYLNAPHTHALIMIDES
Scheme 2. Reaction pathway of acid hydrolysis of 2,3-NBAmide
In order to avoid both the lack of precision of the calculation of
DG0aqðHþÞ as well as the need of extremely large basis sets to
calculate the exact values of pKa, we calculated (pKa (24) ꢀ pKa
(124))[36,37] which can be obtained with smaller basis sets (Eqns 3
and 4).[37]where
Reactions of 2,3-NBAmide and 2,3-NBN below pH 9 were
studied only at 50 8C. At pH 6.5, using either 2,3-NBN or
2,3-NBAmide as substrates the same final spectrum was obtained
(Fig. 2A). At pH 6.5 the spectrum, after 170 h, corresponded to an
almost equal mixture of 2,3-NBN and 2,3-NBAmide starting from
either compound. As expected the rate constants were identical
using either 2,3-NBAmide or 2,3-NBN as substrates, between pH
6.5 and 8.5 (see below). A small amount of n-butylamine, less than
5% – indicative of diacid formation – were found in the samples
when 2,3-NBAmide was used as substrate. 2,3-NBN formation
from 2,3-NBAmide was followed for no more than 10 days
because in this pH range 2,3-NBAmide hydrolysis to 2,3-NDiacid
also occurs, at a slower rate. As the formation of 2,3-NDiacid
and n-butylamine is irreversible, at infinite time the products
should be only 2,3-NDiacid and n-butylamine.
pKað24Þ ꢀ pKað124Þ
1
¼
ðDG0deprot;aqð24Þ ꢀ DGd0eprot;aqð124ÞÞ
(3)
2:3003RT
DG0deprot;aqð24Þ ꢀ DGd0eprot;aqð124Þ
¼ DG0aqð24Þ þ DG0aqð122Þ ꢀ DGa0qð22Þ ꢀ DGa0qð124Þ (4)
DG0aq ¼ E0K þ ZPE þ DDG0!298K þ DGsolv
(5)
Below pH 6.5, 2,3-NBAmide produces 2,3-NBN and also
2,3-NBAnhydride which is rapidly hydrolysed to 2,3-NDiacid.
The decomposition reaction of 2,3-NBAmide, below pH 8.5, as
a function of time, was followed both by measuring 2,3-NBN
(at 262 nm) (Fig. 2B) and n-butylamine (see Methods), (Fig. 2C).
Both absorbance at 262 nm (Fig. 2B) and fluorescence
(see Methods) (Fig. 2C) increased to a plateau with time but
the final values of absorbance and fluorescence were pH
dependent.
To demonstrate that the product absorbing at 262 nm
was 2,3-NBN and not 2,3-NAnhydride (spectra of both
compounds are similar), samples of the reaction mixtures were
added to borate buffer 0.2 M, pH 10 after the absorbance’s
reached a constant value (10 days, for the slowest reactions).
The kc’s for the observed reactions were identical to that of
the hydrolysis of 2,3-NBN at pH 10. This confirmed that the
absorption peak at 262 nm was due only to 2,3-NBN.
2,3-NAnhydride is completely and rapidly hydrolysed to
2,3-Diacid in this pH range.[39,40]
The basicity in the gas phase (GB) was also calculated with
Eqn (6) using DG0gðHþÞ ¼ 6; 28 kcal=mol.[38]
GB ¼ DG0gðHþÞ þ DGg0ðAꢀÞ ꢀ DG0gðAHÞ
(6)
RESULTS
The UV spectrum of 2,3-NBN in water exhibited absorptions
maxima at 364 and 262 nm with molar absorption coefficients, e,
of 3.57 ꢁ 103 and 5.09 ꢁ 104 Mꢀ1 cmꢀ1, respectively (Fig. 1A, line
a). The absorption of 2,3-NBAmide at 364 nm was negligible and,
at 262 nm, the e is 4.9 ꢁ 103 Mꢀ1 cmꢀ1 (Fig. 1A, line b).
2,3-NBN hydrolysed in NaOH 0.002 M yielding a final spectrum
identical to that of pure 2,3-NBAmide (Fig. 1A, line b). The
isosbestic point at 246 nm (Fig. 1A) suggest the formation of a
single product. Between pH’s 9 and 12 (30 and 50 8C) the product
of 2,3-NBN hydrolysis was exclusively 2,3-NBAmide, no
n-butylamine was detected and the final absorbance’s, A , were
1
identical.
The product concentrations, at the end of 2,3-NBAmide
reaction, from pHs 0 to 10, were calculated using Eqns (7) and (8),
from the absorbance after 170 h, AT, and the [n-butylamine] (see
Methods), equivalent to the 2,3-NDiacid concentration.
kc for alkaline hydrolysis of 2, 3-NBN increased with pH, the
slope of the plot of log kc vs. pH was 1.0 at 30 8C (inset Fig. 1B) and
at 50 8C (not shown). The calculated second-order rate constants,
k2OH, at 30 and 50 8C, were 5.8 and 9.0 Mꢀ1 sꢀ1 respectively.
The mechanism of 2,3-NBN alkaline hydrolysis can be
described, as for other similar imides,[3–5] as a rate limiting
OHꢀ attack on a carbonyl centre of 2,3-NBN (123 in Scheme 1),
formation of a negatively charged tetrahedral intermediate 122
leading to 2,3-NBAmide 11 in an essentially irreversible step
(Scheme 1).
CT ¼ ½2; 3 ꢀ NBNꢂ þ ½2; 3 ꢀ Diacidꢂ þ ½2; 3 ꢀ NBAmideꢂ
(7)
AT ¼ "1 ꢁ ½2; 3 ꢀ NBNꢂ þ "2 ꢁ ½2; 3 ꢀ Diacidꢂ þ "3
ꢁ ½2; 3 ꢀ NBAmideꢂ
(8)
J. Phys. Org. Chem. 2011, 24 385–397
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

T. C. DE BARROS ET AL.
2,3-NBAmide as a function of reaction pH are in Fig. 3(A). The % 2,
3-NBN was small at pH 8.5, increased with decreasing pH
reaching a maximum between pHs 6.5 and 5.5 and decreased to
almost zero below pH 3.0. 2,3-NBAmide, the only product
above pH 9, decreased with acidity reaching zero below pH 5.0.
The % 2,3-Diacid increased from pH 6 to 0 reaching almost 90% of
the initial 2,3-NBAmide concentration below pH 3 (Fig. 3A). As
independently demonstrated above (Fig. 2A), at pH 6.5 the ratio
[2,3-NBN]/[2,3-NBAmide] is nearly 1.0.
The change in product ratios of 2,3-NBAmide reactions
from pH 8.5 to 5 (see Scheme 1) can be attributed to the
formation of a tetrahedral intermediate (122), negatively charged
in alkali but increasingly protonated (124) below pH 8.5. The
increase in the concentration of 122 favours the formation of
2,3-NBAmide (11) while that of 124 lead to 2,3-NBN (123) as
the pH decreases. From the data of Fig. 3(A), we can infer that
intermediate 122 has a pKa close to 6.5. As the pKa of the
carboxylic acid in 2,3-NBAmide was 3.8 (determined here from
absorbance change with pH at 334 nm, 30 8C (data not shown))
the changes in the observed product ratios (Fig. 3A) cannot be
correlated to protonation of the carboxylate group of
2,3-NBAmide.
The kc’s for 2,3-NBAmide reactions are in Fig. 3(B). kc’s
decreased from pH 8.5 to 5 reaching a pH independent value
(5.4 ꢁ 10ꢀ6 sꢀ1) between pH 7 and 5.0 (water reaction) and
increasing again below pH 5 (Fig. 3B). Since almost no
n-butylamine was detected from pH 8.5 to pH 6.5 the kc
at pH 6.5 can be attributed to the rate constant for 2,3-NBN and
_
2,3-NBAmide equilibrium attainment, kcCyc, due to the water
reaction (Scheme 1).
At pHs lower than 6.5, protonation of the carboxylate group of
2,3-NBAmide led to an increase of the rate of 2,3-NAnhydride
formation, rapidly hydrolysed to 2,3-Diacid (Scheme 2).[41] At pHs
lower than 2.5, kc for 2,3-NBAmide hydrolysis to 2,3-Diacid was
almost constant, twelve times higher than kCcyc, and the
predominant reaction product was 2,3-NDiacid (Fig. 3).
Below pH 1 the % 2,3-NBN increased again (Fig. 3B) implying
that an acid catalysis is operating at very low pH, increasing the
cyclization rate to 2,3-NBN, as noted with other amides.[19] We did
not extend this study further in this region.
The pH dependence of kc for 2,3-NBAmide decomposition
below pH 6 can be described by Eqn (9):[42]
kDiacid
1 þ Ka=½Hþꢂ
kc ¼ kCyc
þ
(9)
where Ka is the dissociation constant of the carboxylic group of
2,3-NBAmide and kDiacid is the rate constant for 2,3-NBAmide
hydrolysis to 2,3-NDiacid and butylamine. The values used
for the simulation of data of Fig. 3(B) with Eqn (9)
where: kcyc ¼ 5.4 ꢁ 10ꢀ6 sꢀ1
,
kDiacid ¼ 6.8 ꢁ 10ꢀ5 sꢀ1 and Ka ¼
Figure 4. (A) Spectra of reaction products of 1,8-NBAmide in HCl 0.01 M,
at several times (in minute): (a) t ¼ 0; b ¼ 0.5; c ¼ 1.0; d ¼ 1.5; e ¼ 2.5;
f ¼ 3.5; g ¼ 6.5; h ¼ 40, [1,8-NBN] ¼ 8.1 ꢁ 10ꢀ6 M. (B) Time effect on the
absorbance, at 343 nm, of 1,8-NBAmide cyclization at pHs: (D) 0.4; (~) 5.1;
(*) 5.8; (*) 6.1. 1,8-NBAmide (1.8 ꢁ 10ꢀ6 M). (C) pH effect on the kc of:
1,8-NBAmide cyclization (*), 1,8-NAnhydride hydrolysis (~) and
1,8-NAcid cyclization (D). Inset is the expanded region of the graph
between pH 1 and 6. All experiments were done at 30 8C
1.2 ꢁ 10ꢀ4 Mꢀ1 (pKa ¼ 3.92). The pKa value obtained by simulation
was similar to that experimentally obtained for 2,3-NBAmide
carboxylic acid (i.e., 3.80, see above).
1,8-NBAmide has an absorption maximum at 293 nm
(e293 ¼ 7700 Mꢀ1 cmꢀ1) (Fig. 4A, line a) and 1,8-NBN at 343 nm
(e343 ¼ 12 900 Mꢀ1 cmꢀ1).[21] The UV spectrum of 1,8-NAnhydride
is almost identical to that of 1,8-NBN with the same maximum at
343 nm and similar molar absorption coefficient.[39,40] At pH 2,
the spectra of a 1,8-NBAmide solution changed with time
becoming indistinguishable from that of 1,8-NAnhydride or
1,8-NBN (Fig. 4A).
In Eqns (7) and (8), CT is the initial 2,3-NBAmide concentration,
e1, e2 and e3 are the molar extinction coefficients of 2,3-NBN,
2,3-NBAmide and 2,3-NDiacid at 262 nm (50 900; 4900 and
4150 Mꢀ1 cmꢀ1, respectively). The % of 2,3-NBN, 2,3-NDiacid and
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 385–397

FORMATION AND DECOMPOSITION OF N-ALKYLNAPHTHALIMIDES
Scheme 3. Possible Reaction paths of compound 2 in acid conditions. In all structures, R ¼ CH3
Below pH 7, 343 nm absorbance of 1,8-NBAmide solution
increased to a maximum, Absmax, and then decreased, at slower
rate, to a minimum, Absmin (Fig. 4B). Absmax and Absmin were pH
dependent (Fig. 4B). The kinetics in Fig. 4(B) can be attributed to
an initial formation of 1,8-NAnhydride from 1,8-NBAmide, with
elimination of n-butylamine and a slower 1,8-NAnhydride/
1,8-NDiacid equilibration. A pH dependent equilibrium between
1,8-NAnhydride and 1,8-NDiacid is known.[39,40] We confirmed
that below pH 6 the absorption at 343 nm when Absmax was
reached was due to 1,8-NAnhydride and not to 1,8-NBN by
demonstrating that [n-butylamine] was identical to the expected
value and proving that the rate of decomposition of the product
in 0.01 M NaOH was that of 1,8-NAnhydride.
The rates of 1,8-NAnhydride formation from 1,8-NBAmide
from pH 7.5 to H0 ¼ ꢀ4 are in Fig. 4(C). The kc’s below pH 7.5 were
obtained from the first five half lives of the reactions to minimize
errors due to 1,8-NAnhydride hydrolysis (Fig. 4B).
The kc/pH dependence of 1,8-NAnhydride formation from
1,8-NBAmide was complex (Fig. 4C). Below pH 7, the kc’s
increased seemingly due to protonation of the carboxylic acid of
1,8-NBAmide (pKa1 near 4.0). A plateau is reached when the
carboxylic acid is fully protonated (Fig. 4C). Further protonation of
1,8-NBAmide increases the reaction rate (Fig. 4C).
Scheme 3 can be used to rationalize the data of Fig. 4(C).
Undissociated 1,8-NBAmide (2) yields 1,8-NBAnhydride (4).
Futher protonation of 2 yields 10, a unstable species in the
gas phase, producing 5. In solution, however, and consistent
with previous studies on amide reactions, 10 may be
stabilized by water. Consequently we have considered the
2 þ Hþ , 10 equilibrium.[7] The effective equilibrium constants
Ka1 and Ka2 are
½1ꢂ ꢃ Hþ
Ka1
¼
(10)
½2ꢂ
½10ꢂ ꢃ Hþ
½2ꢂ
Ka2
¼
(11)
J. Phys. Org. Chem. 2011, 24 385–397
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

T. C. DE BARROS ET AL.
where k1, k2 and kHþ are, respectively, the rate constants of the
reactions going from 2 and 10 and the specific acid catalysed
pathways from 10 to 1,8-Anhydride. The parameters values that
fitted the experimental data to Eqn (12) were: K1 ¼ 1 ꢁ 10ꢀ4
(pK1 ¼ 4.0), K2 ¼ 0.5 (pK2 ¼ 0.3), k1 ¼ 0.007 sꢀ1, k2 ¼ 0.042 sꢀ1 and
kHþ ¼ 0.0018 sꢀ1 Mꢀ1
. The pKa of the carboxylic acid of
1,8-N-acetyl-Naphthalamide (1,8-NAN)[43] is 3.47, and that
obtained using Eqn (12) for 1,8-NBAmide was 4.0, similar to
that of 1,8-NAN.
The spectrum of 1,8-NBN in 0.20 M NaOH (Fig. 5A, line a)
changes with time and, at the end of the reaction, it was identical
to that of 1,8-NBAmide (Fig. 5A, line k) and n-butylamine was not
detected at the end of the reaction. Above 0.2 M, 1,8-NBN reacted
with OHꢀ producing exclusively 1,8-NBAmide and the final
spectra were [NaOH] independent.
Between pHs 10 and 13, with either 1,8-NBAmide or 1,8-NBN
the absorbances at equilibrium were pH dependent. The
reversibility of the cyclization process was evidenced by reacting
1,8-NBN in NaOH at pH 13 to completion and then observing the
re-appearance of 1,8-NBAmide upon lowering the pH to 10.0 (not
shown). The kc’s for attaining equilibrium were identical starting
with either 1,8-NBAmide or 1,8-NBN. The pH dependence of the
calculated percentages of products starting the reaction with
1,8-NBAmide or 1,8-NBN are in Fig. 5(B) and (C).
1,8-NBN is stable below pH 9 (Fig. 5C), while 1,8-NBAmide
undergoes two concurrent reactions: cyclization to 1,8-NBN and
hydrolysis to 1,8-NDiacid and butylamine (Fig. 5B). The % of
1,8-NBN formed from 1,8-NBAmide decreased from pH 8.5 to 6
and the formation of 1,8-NDiacid (and n-butylamine) increases
reaching almost 100% below pH 6.0.
The equilibrium between 1,8-NBN (23) and 1,8-NBAmide in
alkali can be rationalized by Scheme 4. The minimum scheme
used to describe the kinetic process going from 23 to 25, can be
rewritten as
K3
23 þ ꢀOH ! 22
K4
22 þ ꢀOH ! 25 þ H2O
K5
25 þ H2O ! 1 þ OHꢀ
The equilibrium constants K3, K4 and K5 are
½22ꢂ
K3 ¼
(13)
(14)
(15)
½23ꢂ½OHꢂ
K4
½25ꢂ
¼
H2O ½22ꢂ½OHꢂ
½1ꢂ½OHꢂ
K5 ꢃ H2O ¼
½25ꢂ
Figure 5. (A) Spectra of reaction products mixture of 1,8-NBN in NaOH
0.2 M, 30 8C. Line f, 1,8-NBN, t ¼ 0 and line a, 1,8-NBAmide, t ¼ 1.
Intermediate spectra were taken at intervals of ca 2 min,
[1,8-NBN] ¼ 8.1 ꢁ 10ꢀ6 M. (B) Effect of pH in the % of products, after
equilibrium attainment, starting the reaction with 1,8-NBAmide: (*)
n-butylamine; (*) 1,8-NBN; and (&) 1,8-NBAmide. Line was calculated
using Eqn (17). (C) Effect of pH on the % of products after equilibrium
attainment starting the reaction with 1,8-NBN: (*) 1,8-NBN;
(D)1,8-NBAmide
The total concentration of species in equilibrium, Ct, is
½Ctꢂ ¼ ½1ꢂ þ ½22ꢂ þ ½23ꢂ þ ½25ꢂ
(16)
The concentration of 1,8-NBN, [23], can be calculated from the
final absorbance, at 343 nm, after equilibrium attainment, Abseq
1,8-NBAmide (1), as well as all other species (i.e., 24, 22, 21, 25
Scheme 4) do not absorb at this wavelength because the lack of
extended conjugation. The Abseq vs. pH data were fitted with
kc is given by Eqn (12), for pHs between 7 and H0 ¼ ꢀ2:
.
þ
þ
k1
k2
kH ½H ꢂ
kc
¼
þ
þ
(12)
1 þ K1=½Hþꢂ 1 þ K2=½Hþꢂ ð1 þ K2=½HþꢂÞ
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 385–397

FORMATION AND DECOMPOSITION OF N-ALKYLNAPHTHALIMIDES
Table 1. Energetic of the stable structures in acid in Scheme 3
RHF relative energy
(kcal/mol)
MP2 relative energy
(kcal/mol)
Product
RHF energy (u.a.)
MP2 energy (u.a.)
1 R 2H3OR
ꢀ929.760568
ꢀ930.092943
ꢀ930.076630
ꢀ930.233649
ꢀ930.190402
ꢀ930.162616
ꢀ930.179650
ꢀ930.184746
ꢀ930.179137
ꢀ930.120538
ꢀ930.197375
277.2
88.3
98.5
0.0
ꢀ932.503566
ꢀ932.799739
ꢀ932.781559
ꢀ932.924991
ꢀ932.882633
ꢀ932.847103
ꢀ932.881307
ꢀ932.878490
ꢀ932.876442
ꢀ932.834528
ꢀ932.894426
264.4
78.6
90.0
0.0
(2 R H2O) R H3OR
(3 R H2O) R H3OR
(4 R CH3NHR3 R H2O) R H2O
(5 R 2H2O) R H2O
(6 R CH3NH2 R H2O) R H2O
(7 R H2O) R H2O
27.1
44.6
33.9
30.7
34.2
71.0
22.8
26.6
48.9
27.4
29.2
30.5
56.8
19.2
(8 R H2O) R H2O
*
*
(8 R H2O) R H2O
R
(2 R H3O ) R H2O
(9 R H2O) R H2O
Table 2. Energetic of the stable structures in moderate alkaline condition of species in Scheme 4
RHF relative energy
(kcal/mol)
MP2 relative energy
(kcal/mol)
Product
RHF energy (a.u.)
MP2 energy (a.u.)
1 R H2O R OHS
21 R H2O R OHS
22 R H2O R OHS
23 R H2O R 2OHS
24 R 2OHS
ꢀ928.551
ꢀ928.515
ꢀ928.527
ꢀ928.438
ꢀ928.417
ꢀ928.501
ꢀ70.5
ꢀ47.9
ꢀ55.7
0.0
13.3
ꢀ39.3
ꢀ931.265
ꢀ931.231
ꢀ931.249
ꢀ931.136
ꢀ931.125
ꢀ931.222
ꢀ80.8
ꢀ59.6
ꢀ71.0
0.0
6.9
ꢀ54.0
25 R 2H2O
Eqn (17) (from Eqns 13–16):
been observed in the alkaline hydrolysis of N-acetylpirrole and
N-benzoylpyrrole.[44]
Abseq
1;8ꢀNBN
½Ctꢂ
½1; 8 ꢀ NBNꢂ ¼
"
Ab initio calculations
¼
(17)
ð1 þ K3½OHꢂ ꢃ ð1 þ K4K5 þ K4½OHꢂÞÞ
In these calculations, we used 1,8-N-Methyl-Naphthalimide
(1,8-NMN) and 1,8-N-Methyl-Naphthalimide (1,8-NAmide) as an
equivalent models for 1,8-NBN and 1,8-NBAmide, but throughout
we will use, for clarity, 1,8-NBN and 1,8-NBAmide. Summaries of
the gas phase computed possible reaction pathways for the
reactions of 1,8-NBN with water, and added Hþ or OHꢀ, and are
presented in Schemes 3 and 4, respectively and the species
labelling will be used throughout.
where Ct, the initial concentration of 2,3-NBAmide was
1.25 ꢁ 10ꢀ5 M and e1,8-NBN is 12 900 Mꢀ1 cmꢀ1
. The ratios
*
([1,8-NBN]/Ct) 100 were plotted in Fig. 5(B) and (C) as a function
of pH (between pHs 10 and 14). A good fit to the experimental
data of Fig. 5(B) and (C) with Eqn (17) was obtained using
K3 ¼ 5 Mꢀ1, K4/H2O ¼ 0.5 Mꢀ1 and K5 H2O ¼ 4 M.
*
From pHs 8.5 to 11.5, 1,8-NBAmide cyclizes almost quantitat-
ively to 1,8-NBN after 14 days (30 8C). The kc for 1,8-NBAmide
cyclization had a minimum (3.5 ꢁ 10ꢀ6 sꢀ1) between pHs 8.5 and
9.5, attributed to water reaction (Fig. 6A). As the 1,8-NBN/
1,8-NBAmide equilibrium is totally displaced to the 1,8-NBN in
this pH range, reactions were started with 1,8-NBAmide and the
value of the equilibrium constant was not determined.
All structures were optimized and the energies are in Tables 1
and 2.
Acid hydrolysis of 1,8-NBAmide, Scheme 3
Anhydride 4 was the most stable species, 19.2 kcal/mol (22.8 kcal/
mol) below amide 9. The protonation steps, (1 ! [2, 3, 4]),
(2 ! [5, 7, 9]) and (3 ! [5, 6, 9]) were exo-energetic and generally
proceeded without energetic barriers.
The total energy of species 4 and the deprotonated product
(CH3NH2) was 6.3 kcal/mol (0.2 kcal/mol) above 2, while 3 was
destabilized by 9.9 kcal/mol (9.8 kcal/mol) with respect to 2.
Entropic effects should greatly stabilize 4 over all other species
since an extra molecule is formed by this reaction.
Between pH 11.5 and 12.5 the plot of kc/[OH] vs. [OH] was
linear with a slope of 1 and, above pH 12.5, the slope decreases
to zero (Fig. 6B). kc for 1,8-NBAmide cyclization (or 1,8-NBN
hydrolysis) has a second-order dependence with [OH] below pH
12.5 and, above this pH, the rate limiting step of the reaction
changed leading to a first order dependence of kc with [OH]. The
changes in reaction order with [OH] in amide hydrolysis have also
J. Phys. Org. Chem. 2011, 24 385–397
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

T. C. DE BARROS ET AL.
Scheme 4. Possible Reaction paths of compound 23 in alkaline conditions
The only step where a transition state structure was found was
the elimination of amine and the formation of the corresponding
anhydride 5 ) 6 (Scheme 3, Table 1 and Fig. 7).
mol (þ2.4 kcal/mol) in an exoenergetic reaction (DE ¼ ꢀ11.4 kcal/
mol (ꢀ7.8 kcal/mol)) (Table 3). The activation energy is negative
at the MP2//RHF level and the TS only existed on the RHF
hypersurface. Optimization of 21 at the MP2 level of theory
directly led to 22. Deprotonation of 22 led spontaneously to
the opening of the imide cycle without any transition state,
yielding 25.
For completion, we calculated the other pathways from 22. The
transformation of 22 to 23, seen as the elimination of a hydroxyl,
(22 ! 23 R OHS), is highly endoenergetic, DE ¼ þ71.0 kcal/mol
(þ55.7 kcal/mol), but if the same reaction is seen as the
protonation of 22 at the –OH, (22 R H3OR ! 23 R 2H2O), it is
highly exoenergetic, DE ¼ ꢀ183 kcal/mol (ꢀ199 kcal/mol), and
proceeds without TS. These calculations suggested that this
reaction pathway could be favoured in mildly alkaline conditions.
In order to compare the reactivity of the imines 121 and 21
(Schemes 1 and 4) the reactions of the ring opening from
122 , 121 and 22 , 21were studied. The reaction of 21 ! 22
was displaced to 22, without the presence of TS. The inclusion of
solvent does not change the reaction pathway. 21 is not stable in
the more correct level of the calculation (MP2). For the reaction
121 ! 122, the species 121 is stable in all levels of the theory and
is a true reaction intermediate.
The energy analysis at this level of theory suggested that the
sequence going from 1 to 2, 5, 6 and finally to 4 requiring two
successive protonations was a favoured pathway (Scheme 3). An
alternative pathway with a single protonation going directly
from 1 to 4 was also possible proceeding without an energetic
barrier.
Alkaline hydrolysis, of 1,8-NBAmide, Scheme 4
The calculations of the reactions with OHꢀ were performed to
investigate, in a preliminary form, possible pathways yielding
the species that we observed experimentally in solution. For
the calculations we have considered the initial species for the
reactions of 1,8-NBAmide with OHꢀ to be 1 (Scheme 4). No direct
route for the reaction 1 ) 22 was found. However species 1 is in
equilibrium with 21, an isoprotonic form. As expected, the
stability of 1 was higher than that of 21 (ca 20 kcal/mol, Table 2)
but, on the other hand, the reactivity of 21 should be much
higher (Scheme 4).
The reaction from 21 (or 1) is critical for our comparison
because this step leads to the formation of the imide, by
cyclization of the open amide, a reaction that was readily
demonstrated experimentally only in alkali (Fig. 5). In fact 21
The energetic descriptions of 24/22 and 124/122 pairs were
calculated and the results are given in Table 3. The basicity (GB),
using Eqn (6) at the B3LYP level of the theory, of the pairs 24/22
and 124/122 in the gas phase are, respectively, 337.15 and
¼
yielded 22 through a virtually absent TS with DE ¼ ꢀ3.0 kcal/
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 385–397

FORMATION AND DECOMPOSITION OF N-ALKYLNAPHTHALIMIDES
Figure 7. Geometry of the transition state between 5 and 6
337.85 kcal/mol. The difference between the pKas of the
intermediates 24 and 124 (pKa(24)–pKa(124))was 0.82. The
difference in basicity between 24 and 124 was small and, in
the gas phase, 124 was more basic than 24 (0.7 kcal/mol),
while in aqueous phase the opposite occurs.
The absolute energies for C—N bond breakage in 122 and 22,
their respective TS and the relative energies in Table 4. The
reaction 21 ! 22 was displaced to 22 without the presence of TS.
The inclusion of a solvent effect (Table 4) did not modify this
conclusion. 21 was unstable at the more correct level of the
calculations (MP2). For the reaction 121 ! 122, 121 was stable at
all levels of theory, i.e., a true reaction intermediate. 121 can
suffer a transprotonation in order to be transformed in the
correspondent amine.
Figure 6. (A) pH effect on the kc of 1,8-NBN (&) and 1,8-NBAmide (*)
equilibrium reactions at 30 8C. (B) Effect of [OH] in the ratio of kc/[OH] in
the equilibrium reactions of 1,8-NBN (or 1,8-NBAmide). Inset is the
expansion of the linear region of Fig. 6(B)
Table 3. Comparison of the acidity of the tetrahedrical intermediates of 1,8 and 2,3-NBN species
Molecule
E0K
ZPE
DDG0_ > 298 K
DGsolv
DG0
aq
24
22
124
122
ꢀ782.453188
ꢀ781.890771
ꢀ782.453982
ꢀ781.891039
0.219578
0.204046
0.217651
0.203341
ꢀ0.037214
ꢀ0.039468
ꢀ0.040299
ꢀ0.040337
ꢀ0.015165
ꢀ0.094681
ꢀ0.013971
ꢀ0.096428
ꢀ782.288827
ꢀ781.820874
ꢀ782.290599
ꢀ781.824463
Energies in u.a.
Table 4. Comparison of the ring opening for 1,8 and 2,3-NBN
MP2
DE
DGSOLV
(kcal/mol)
DERHF
(kcal/mol)
DEMP2
(kcal/mol)
aq
RHF
0K
MP2
0K
Molecule
E
(u.a.)
E
(u.a.)
(kcal/mol)
122
ꢀ777.191879
ꢀ777.180197
ꢀ777.189699
ꢀ777.189636
ꢀ777.174574
ꢀ777.177269
ꢀ779.536328
ꢀ779.526723
ꢀ779.527441
ꢀ779.540183
ꢀ779.526633
ꢀ779.519064
ꢀ68.2
ꢀ55.8
ꢀ61.8
ꢀ65.9
ꢀ54.8
ꢀ55.7
0
0
0
TS 122-121
121
22
TS 22-21
21
7.3
1.4
0
9.4
7.8
6.0
5.6
0
8.5
13.2
18.4
12.0
0
19.7
23.4
Reactions Pathways in Schemes 1 and 4.
J. Phys. Org. Chem. 2011, 24 385–397
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

T. C. DE BARROS ET AL.
hydrolysis 1,8-NBAmide is one of the most efficient models yet
described.[11–20]
DISCUSSION
The mechanism of amide bond hydrolysis has been extensively
investigated. The stability of the amide bond of simple
peptides in water has led to focus much of the interest on
reactive amides instead of non-activated model molecules.
With non-activated amides containing nearby nucleophiles
hydrolysis can be efficiently catalysed. The accelerations
observed in chemical models are still modest.[45] The largest
catalytic factor, compared with the bimolecular reaction, is of
the order of 107 for acid catalysis and 106 for hydrolysis of
carboxymethyl-di-isopropyl-maleamic acid at pH near the
neutrality.[18] For enzymes, similarly calculated catalytic factors
Detailed DFT calculations have modelled the reaction
mechanisms of the amide hydrolysis in N-(o-carboxybenzoyl)-
L-aminoacid in acid.[28] In general terms our calculations agree
with those previous results. We note, however, that some of the
reaction pathways described here were not investigated.[28] In
particular, in most of the acid range we have not considered as
one of the most important products the deprotonated amine, a
highly unlikely species at this pH. Also here we have focused on
the species present between pH 0 and 5, where cyclization was
not considered to be significant.[28] In addition solvents may have
little influence on the reaction barriers.[28]
can be 1014
.
In moderate acid our calculations demonstrate that an
energetically possible pathway is the formation of anhydride
from the amide with amine loss (Scheme 3). The most favourable
route was 1, 2, 5, 6 to 4 (Scheme 3, Table 1). It is important to
note that this route is supported experimentally (see Fig. 5).
Our calculations using OHꢀ are, at the same time, in good
agreement with our experimental data and with some of the
calculations and analysis obtained in previous DFT analysis of
imide formation.[28] In particular the route 21, 22 to 23
(Scheme 4, Table 2) is the only one possible, suggesting that
in alkali the kinetic favoured product is the imide. Note that
Boyd[28] suggested that deprotonation of the NH of the amide
would be the favoured reaction route. In our case, the basic
conditions suggest that precisely this route should be the
preferred in the case of both 1,8-NBAmide and 2,3-NBAmide. We
considered the existence of a (Nꢀ, Oꢀ) dianionic species (25,
Scheme 4) and do not believe in the existence of a species
bearing two negatively charged oxygens bound to the same
carbon.[5]
We have compared the cyclization from the dissociated amides
1,8-NBAmide and 2,3-NBAmide. Both our data and previous
calculations[47] suggest that the activation barrier for cyclization
from this species is low when the cycle is a five-membered ring.
Our data show, in addition, that this barrier essentially disappears
when the resulting cycle is a stable six-membered ring (Table 4).
These calculations are again in excellent agreement with
experimental data (See Figs 3B and 5B, C).
Both experimentally and through calculations 1,8-NBAmide
yields essentially only imide at pH near 10. Calculations for
2,3-NBAmide also suggest that imide formation is kinetically
possible for this isomer and that the rate for 2,3-NBAmide should
be lower than that for 1,8-NBAmide, as experimentally found
here.
The strain in the molecule and proximity of reactive groups
can modulate the rates and products at almost all pH scale.
1,8-NBAmide and 2,3-NBAmide can both act as a peptides or/
and as peptidyl transferase at different pH range. So 1,8-NBAmide
is a better ‘peptidase’ than 2,3-NBN at acid pH, and it can
work, at alkaline pH, like a ‘peptidyl transferase’, with great
efficiency.
The search for catalytic efficiency has provided a clearer
picture of mechanistic changes of substrate structure dependent
amide hydrolysis. Some amides are hydrolysed both in alkali
and acid while others may produce imides. The stability of
the resulting imide is strongly affected by the strain imposed
by the substituents near the amide and carboxylic acid
groups.
The kinetic behaviours of 2,3-NBN and 1,8-NBN in alkali were
very different. First-order dependence of rate constants in alkali
are common in amide hydrolysis but changes in the order
dependence with [OH] has been described for amide hydrolysis
where the substituent amine has low pKa.[1] Here we have shown
that the putative tetrahedral intermediate does not go further
when the substrate is 1,8-NBN, and a second OHꢀ is needed to
yield products. The reactivity difference between, 2,3-NBN and
1,8-NBN can be attributed to the relative stability of the
mono-anionic tetrahedral intermediates (Figs 1B and 6B,
Schemes 1 (122) and 4 (22)). For 2,3-NBN 22 is unstable while,
for 1,8-NBN, 122 is stable and a second OHꢀ was necessary form
the corresponding amide.
It is clear, therefore, that 1,8-NBN is stable at moderately
alkaline pH while 2,3-NBN only stable near neutrality.
As with 1,8-NBAmide, 2,3-NBAmide yields 2,3-NBN but only
near neutrality. The pH range where 2,3-NBN/2,3-NBAmide is
highest is similar to that where the ratio between
N-methylphthalimide/N-methylphthalamic acid is maximum
and can be attributed to the similarity between the five member
ring of both imides.[16]
1,8- and 2,3-NBAmide produce imides with different equi-
librium constants and pH ranges. Imide formation is equivalent to
peptide bond synthesis at alkaline pH, in water. It is clear that in
carboxylate assisted amide reactions, differently than in the case
of simple formamides,[46] purely water reactions can have a role in
the pathway manifold, although at a slow rate. The pH range were
the cyclized form is predominant in 2,3-NBN is displaced two pHs
units when compared with 1,8-NBN. Protonation of the carboxylic
acid of 2,3-NBAmide and 1,8-NBAmide lead to the competitive
cleavage of the amide group to diacid or ciclysation to imide. In
acid only 1,8-NBAmide yields a stable anhydride, few examples of
stable anhydrides have been described.[17]
The stability of 1,8-NBN can be attributed to the low imide
cycle strain when compared to the 2,3-NBN. Furthermore, the
position of the reactive groups in 1,8-NBAmide is in the correct
configuration for ring closure.
The rate of amide hydrolysis in compounds containing
neighbouring carboxylic acids is extremely sensitive to the
positioning of the carboxyl group. In comparing the rate
constants for the reaction of carboxylic acid assisted amide
Acknowledgements
The authors acknowledge FAPESP, CNPq and CAPES for financial
support. This work was part of Master’s Thesis of T. C. B. at
Departamento de Bioquimica do Instituto de Qu´ımica da USP
(1991). T.C.B. had a fellowship from CAPES. The authors thank
Professor C. A. Bunton for his suggestions and help.
wileyonlinelibrary.com/journal/poc
Copyright ß 2010 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2011, 24 385–397

FORMATION AND DECOMPOSITION OF N-ALKYLNAPHTHALIMIDES
REFERENCES
[30] A. D. Campbell, M. R. Grimett, Aust. J. Chem. 1963, 16, 854–857.
[31] E. R. Triboni, P. Berci Filho, R. G. S. Berlinck, M. J. Politi, Synth. Commun.
2004, 34, 1989–1999.
[32] S. Udenfriend, S. Steins, P. Bohlen, W. Dairman, W. Leimgruber, M.
Weigele, Science 1972, 178, 871–872.
[1] D. Kahne, W. C. Still, J. Am. Chem. Soc. 1988, 110, 7529–7534.
[2] L. Meriwether, F. H. Westheimer, J. Am. Chem. Soc. 1956, 78,
5119–5123.
[3] R. S. Brown, A. J. Bennet, H. Slebocka-Tilk, Acc. Chem. Res. 1992, 25,
481–488.
[33] V. A. Rassolov, J. A. Pople, M. A. Ratner, T. L. Windus, J. Chem. Phys.
1998, 109, 1223–1229.
[34] Gaussian 03, Revision, C.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G.
E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T.
Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M.
Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E.
Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P.
Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels,
M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B.
Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J.
Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challa-
combe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, J.
A. Pople, Gaussian, Inc., Wallingford CT, 2004.
[35] V. Barone, M. Cossiand, J. Tomasi, J. Comp. Chem. 1998, 19, 404–417.
[36] J. R. Pliego, J. M. Riveros, Chem. Phys. Lett. 2000, 332, 597–602.
[37] S. Hwang, Y. H. Jang, D. S. Chung, Bull. Korean Chem. Soc. 2005, 26,
585–588.
[38] I. A. Topol, G. J. Tawa, S. K. Burt, A. A. Rashin, J. Phys. Chem. A 1997,
101, 10075–10081.
[39] T. C. Barros, S. Yunes, G. Menegon, F. Nome, H. Chaimovich, M. J. Politi,
L. G. Dias, I. M. Cuccovia, J. Chem. Soc. Perkin Trans. II 2001, 12,
2342–2350.
[40] K. Yates, J. C. Riordan, Can. J. Chem. 1965, 43, 2328–2334.
[41] S. F. Yunes, J. Gesser, H. Chaimovich, F. Nome, J. Phys. Org. Chem.
1997, 10, 461–465.
[4] S. O. Eriksson, Acta Chem. Scand. 1968, 22, 892–906.
[5] D. Cheshmedzhieva, S. Ilieva, B. Galabov, J. Mol. Struct. (Theochem.)
2004, 681, 105–112.
[6] J. P. Gutrie, J. Am. Chem. Soc. 1974, 96, 3608–3615.
[7] R. A. Cox, Can. J. Chem. 2008, 86, 290–297.
[8] A. J. Kirby, Advances in Physical Organic Chemistry (Eds: G. Gold, D.
Bethell), Academic Press, 1980, pp. 183–278.
[9] M. L. Bender, Chem. Rev. 1960, 60, 53–113.
[10] T. C. Bruice, S. J. Benkovic, J. Am. Chem. Soc. 1963, 85, 1–8.
[11] A. J. Kirby, P. W. Lancaster, J. Chem. Soc. Perkin II 1972, 1206–1214.
[12] M. F. Aldersley, A. J. Kirby, P. W. Lancaster, J. Chem. Soc. Perkin II 1974,
1487–1495.
[13] R. Kluger, J. C. Hunt, J. Am. Chem. Soc. 1989, 111, 5921–5925.
[14] M. L. Bender, Y. L. Chow, F. Chloupeck, J. Am. Chem. Soc. 1958, 80,
5380–5384.
[15] M. L. Bender, J. Am. Chem. Soc. 1957, 79, 1258–1259.
[16] J. Brown, S. C. K. Su, J. A. Shafer, J. Am. Chem. Soc. 1966, 88,
4468–4474.
[17] F. M. Menger, M. Ladika, J. Am. Chem. Soc. 1988, 10, 6794–6796.
[18] A. J. Kirby, R. S. McDonald, C. R. Smith, J. Chem. Soc. Perkin II 1974,
1495–1504.
[19] A. B. Onofrio, J. C. Gesser, A. C. Joussef, F. Nome, J. Chem. Soc. Perkin II
2001, 1863–1868.
[20] M. B. Kim, D. W. Dixon, J. Phys. Org. Chem. 2008, 21, 731–737.
[21] H. Park, J. Suh, S. Lee, Theochemistry 1999, 490, 47–54.
[22] K. Hori, A. Kamimura, K. Ando, M. Mizumura, Y. Ihara, Tetrahedron
1997, 53, 4317–4330.
[23] S. Antonczak, M. F. Ruiz-Lopez, J. L. Rivail, J. Am. Chem. Soc. 1994, 116,
3912–3921.
[24] S. Antonczak, M. F. Ruiz-Lopez, J. L. Rivail, J. Mol. Model. 1997, 3,
434–442.
[42] L. Eberson, Acta Chim. Scan. 1964, 18, 1276–1282.
[43] Calculated using Advanced Chemistry Development (ACD/Labs) Soft-
ware V8.14 for Solaris (ß 1994–2008 ACD/Labs).
[44] F. M. Menger, J. A. Donohue, J. Am. Chem. Soc. 1973, 95, 432–
437.
[45] W. P. Jencks, Catalysis in Chemistry and Enzymology, Dover, Inc., New
York, 1987.
[25] D. Bakowies, P. A. Kollman, J. Am. Chem. Soc. 1999, 121, 5712–5713.
[26] J. X. Guo, J. J. Ho, J. Phys. Chem. A 1999, 103, 6433–6441.
[27] H. Naundorf, G. A. Worth, H. D. Meyer, O. Kuhn, J. Phys. Chem. A 2002,
106, 719–724.
[46] H. Slebocka-Tilk, F. Sauriol, M. Monette, R. S. Brown, 2002, Can. J.
Chem. 80, 1343–1350.
[47] Z. Wu, F. Ban, R. J. Boyd, J. Am. Chem. Soc. 2003, 125, 3642–3648.
[28] Z. Wu, F. Ban, R. J. Boyd, J. Am. Chem. Soc. 2003, 125, 6994–7000.
[29] T. C. Barros, G. R. Molinari, P. Berci Filho, V. G. Toscano, M. J. Politi, J.
Photochem. Photobiol. 1993, 76, 55–60.
J. Phys. Org. Chem. 2011, 24 385–397
Copyright ß 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc