Deprotonation Sites of Acetohydroxamic Acid Isomers.
A T heoretical and E xperimental Study
Marı ´ a L. Senent,†Alfonso Nin ˜o,‡Camelia Mun ˜oz Caro,‡Satunino Ibeas,§Begon ˜a Garcı ´ a,*,§
J ose ´M. Leal,§Fernando Secco,|and Marcella Venturini|
Departamento de Astrofı ´ sica Molecular e Infrarroja, Serrano 113b, Madrid 28006, Spain,
Grupo de Quı ´ mica Computacional, Escuela Superior de Informa ´tica, Universidad de Castilla la Mancha,
Paseo de la Universidad, 4, 13071 Ciudad Real, Spain, Departamento de Quı ´ mica, Universidad de
Burgos, Misael Ban ˜uelos s/ n, 09001 Burgos, Spain, and Dipartimento di Chimica e Chimica Industriale,
Universita ` di Pisa, Via Risorgimento 35, 56126 Pisa, Italy
Received February 4, 2003
Theoretical (ab initiocalculations) and experimental (NMR, spectrophotometric, and potentiometric
measurements) investigations of theisomers of acetohydroxamic acid (AHA) and their deprotonation
processes have been performed. Calculations with the Gaussian 98 package, refined at the MP2-
(FC)/AUG-cc-pVDZ level considering the molecule isolated, indicate that the Z(cis) amide is the
most stable form of the neutral molecule. This species and the less stable (Z)-imide form undergo
deprotonation, giving rise to two stable anions. Upon deprotonation, the E(trans) forms give three
stable anions. The ab initio calculations were performed in solution as well, regarding water as a
continuous dielectric; on the basis of the relative energies of the most stable anion and neutral
forms, calculated with MP2/PCM/AUG-cc-pVDZ, N-deprotonation of the amide (Z or E) structure
appeared to be the most likely process in solution. NMR measurements provided evidence for the
existence of (Z)- and (E)-isomers of both the neutral and anion forms in solution. Comparisons of
the dynamic NMR and NOESY (one-dimensional) results obtained for the neutral species and their
anions were consistent with N-deprotonation, which occurred preferentially to O-deprotonation.
The (microscopic) acid dissociation constants of the two isomers determined at 25 °C from the pH
dependence of the relevant chemical shifts, pKE ) 9.01 and pKZ) 9.35, were consistent with the
spectrophotometric and potentiometric evaluations (pKHA) 9.31).
Hydroxamic acids are very useful reagents with inter-
esting biological activity and medical applications. Due
to their ability to form stable chelates with a variety of
metal ions, hydroxamic acids play a key roleas bioligands
in the microbial transport of iron (siderophores); the ease
in forming metal chelates facilitates the function of
enzymes in oxygen and electron transport and in other
life-sustaining processes.1Despite their interesting prop-
erties, hydroxamic acids have remained as a poorly
characterized class of organic compounds; assignment of
the correct structure has been a major difficulty because
these compounds may adopt several forms.2Knowledge
of the preferred ionization sites and the protonation-
deprotonation mechanism is essential tolearning therole
played by hydroxamic acids in biological and complex-
Previously, we investigated acetohydroxamic acid us-
ing experimental and theoretical tools.3,4At high acidity
levels, the neutral AHA molecule can accept one proton.
The stable cations and neutral AHA forms have been
studied with ab initiomethods of good quality;3,4calcula-
tions predict bonding of the proton tothe hydroxylamine
oxygen group, which produces a third stable cation in the
presence of solvent, the CO oxygen being the most basic
site. Calculations in the gas phase suggest different
amidic and imidic structures that become stabilized by
The most stable AHA forms in the solid state were
found to be those prone to intermolecular H-bonding;5
Z-E isomerism was observed for AHA in DMSO6and
for monoalkylhydroxamic acids in several solvents.7On
thebasis of theoretical calculations and comparativedata
†Departamento de Astrofı ´ sica Molecular e Infrarroja.
‡Universidad de Castilla la Mancha.
§Universidad de Burgos.
|Universita `di Pisa.
(1) Brown, D. A.; Chidambaram, M. V. In Metal Ions in Biological
Systems; Marcel Dekker: New York, 1982; Vol. 14.
(2) Bauer, L.; Exner, O. Angew. Chem., Int. Ed. Engl. 1974, 13, 376.
Ventura, O. N.; Rama, J . B.; Turi, L.; Dannenberg, J . J . J . Am. Chem.
Soc. 1993, 115, 5, 5754. Ventura, O. N.; Rama, J . B.; Turi, L.;
Dannenberg, J . J . J . Phys. Chem. 1995, 99, 131.
(3) Mun ˜oz-Caro, C.; Nin ˜o, A.; Senent, M. L.; Ibeas, S.; Leal, J . M. J .
Org. Chem. 2000, 65, 405.
(4) Garcı ´ a, B.; Ibeas, S.; Leal, J . M.; Senent, M. L.; Nin ˜o, A.; Mun ˜oz-
Caro, C. Chem-Eur. J . 2000, 6, 2644.
(5) Lindberg, B.; Berndtsson, A.; Nilsson, R.; Nyholm, R.; Exner,
O. Acta Chem. Scand. 1978, A32, 353.
(6) Brown, D. A.; Glass, W. K.; Mageswaran, R.; Girmany, B. Magn.
Reson. Chem. 1988, 26, 970.
(7) (a) Brown, D. A.; Glass, W. K.; Mageswaran, R.; Mohammed, S.
A. Magn. Reson. Chem. 1991, 29, 40. (b) Brown D. A.; Cufle, L. P.;
Fitzpatrick, G. M.; Fitzpatrick, N. J .; Glass, W. K .; Herlihy, K . M.
Collect. Czech. Commun 2001, 66, 99.
10.1021/jo0341564 CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/18/2003
J . Org. Chem. 2003, 68, 6535-65426535
obtained with amides,8it has been reported that the
Z(cis) conformation of monohydroxamate group (CON-
HOH) is preferentially stabilized by intramolecular H-
bonding in nonpolar solvents9or by H-bonding in water.10
NMR and UV studies performed on the structure and
acid-base behavior, respectively, of N-phenylbenzohy-
droxamic acid have shown that the Z(cis):E(trans) ratio
strongly depends on the solvent used.11A property that
has been paid some attention but is the origin for certain
controversy is the acid-base behavior of hydroxamic
acids. O- versus N-deprotonation has received different
interpretations. First, it was accepted that hydroxamic
acids are O-acids,6,12-14but extensive IR and UV mea-
surements in dioxaneand aqueous alcohol15indicatethat
hydroxamic acids are N-acids, a conclusion supported by
17O NMR and FT-IRC studies of benzohydroxamate ion
in MeOH.12,16Potentiometric measurements compatible
with O- and N-deprotonation of hydroxamic acids have
been reported2,17-19assuming that the acid dissociation
constant Kexp ≈ KOMe + KNMe; however, irregularities
attributed to solvent effects have been observed.2,14b,19,20
The most probable deprotonation processes of AHA
were also investigated with ab initio methods. Previous
ab initio calculations of the isolated molecules confirm
the NH-acidity on the basis of the anions relative
energies.21-23Most of the ab initio studies on acid-base
properties of hydroxamic acids concern formoacetohy-
droxamic acid, which appears todissociatemost probably
by N-deprotonation.24-27Toget additional insight intothe
deprotonation sites of AHA, this work undertakes a
theoretical and experimental study of this acid at differ-
ent temperatures; the solvent effect was included con-
sidering water as a continuous dielectric. Relevant NMR
information on the deprotonation process and on confor-
mational isomerism is compared with thethermodynamic
results. The microscopic constants determined by NMR
are consistent with the macroscopic value obtained from
potentiometric and UV-vis measurements.
The ORIGIN 3800 IRIX 64 processor computer of the
CTI of C.S.I.C. (Madrid) was used for computation. All
ab initiocalculations were performed with the Gaussian
98 (Revision A.1x)28program at the MP2/AUG-cc-pVDZ
level. Original Fortran codes have been designed for the
treatment of the data. The zero-point vibrational energy
and thermodynamic properties of the species in the gas
phase have been calculated at the MP2/AUG-cc-pVDZ
level using the harmonic approximation and the rigid
rotor approximation for all the conformers.
In the calculations performed in solution, water was
regarded as a continuous dielectric, characterized by a
constant permittivity. The Polarized Continuum Model
(PCM) was used, implemented in the Gaussian 98
package29using the same basis at the MP2 level. The
cavity radii are those recommended for the UAHF
model.29The solvent dielectric constant has been set at
78.5 for T ) 298.15 K, 74.8 for T ) 308.15 K, and 71.5
for T ) 318.15 K.30The changes of free energy in solution
have been computed according tothe recipes given in ref
R esults and Discussion
T heoretical R esults: Neutral F orms. The starting
species for the quantum mechanical analysis were the
neutral stable structures already described (Figure 1),
which werefound with theMP2/cc-pVDZ method and full
geometry optimization.3The calculations performed with
the Gaussian 98 package were refined using the AUG-
cc-pVDZ basis set, which is more appropriate for the
treatment of anion species.28,32Searching for the stable
anions and simulation of the chemical dissociation pro-
cesses were performed starting from the reoptimized
(8) (a) Hupe, D. J .; Wu, D. J . Am. Chem. Soc 1977, 99, 7653. (b)
Yamaki, R. T.; Paniago, E. B.; Carvalho, S.; Howarth, O. W.; Kam, W.
J . Chem. Soc., Dalton Trans. 1977, 24, 4817. (c) Keeffe, J . R.; J encks,
W. P. J . Am. Chem. Soc. 1983, 105, 265.
(9) Liczynska-K ochany E.; Iwamura, H. J . Org. Chem. 1982, 47,
(10) Brown, D. A.; Coogan, R. A.; Fitzpatrick, N. J .; Glass, W. K.;
Abukshima, D. A.; Shiels, L.; Ahlgre ´n, M.; Smolander, K.; Pakkanen,
T. T.; Pera ¨kyla ¨, M. J . Chem. Soc., Perkin Trans. II 1996, 2673.
(11) Garcı ´ a, B.; Ibeas, S.; Mun ˜oz, A.; Leal, J . M.; Ghinami, C.; Secco,
F.; Venturini, M. Inorg. Chem. 2003, in press.
(12) Palm, V. A., Ed. Tables of Rate and Equilibrium Constants of
Heterocyclic Organic Reactions; VINITI: Moscow, 1975.
(13) Monzky, B.; Crumbliss, A. L. J . Org. Chem. 1980, 45, 4670.
(14) (a) Bagno, A.; Comuzzi, C.; Scorrano, G. J . Am. Chem. Soc.
1994, 116, 916. (b) Bagno, A.; Scorrrano, G. J . Phys. Chem. 1996, 100,
(15) Exner O.; Kaka ´c, B. Collect. Czech. Chem. Commun. 1963, 28,
(16) Exner, O.; Holubek, J . Collect. Czech. Chem. Commun. 1965,
(17) (a) Bordwell, F. G.; Fried, H. E.; Hughes, D. L.; Lynch, T. S.;
Satish, A. V.; Whang, Y. E. J . Org. Chem. 1990, 55, 3330. (b) Bordwell,
F. G.; Liu, W. Z. J . Am. Chem. Soc. 1996, 118, 8777.
(18) Decouzon M.; Exner, O.; Gal, J . F.; Maria P. C. J . Org. Chem.
1990, 55, 3980.
(19) Exner, O.; Hradil, M.; Mollin, J . Collect. Czech. Chem. Commun.
1993, 58, 1109.
(20) (a) Monzky, B.; Crumbliss, A. L. J . Org. Chem. 1980, 45, 4670.
(b) Brink, C. P.; Crumbliss, A. L. J . Org. Chem. 1982, 47, 1171. (c)
Brink, C. P.; Fish, L. L..; Crumbliss, A. L. J . Org. Chem. 1985, 50,
(21) Ventura, O. N.; Rama, J . B.; Turi, L.; Dannenberg, J . J . J . Am.
Chem. Soc. 1993, 115, 5754.
(22) Yamin, L. J .; Ponce, C. A.; Estrada, M. R.; Vert, F. Tomas J .
Mol. Struct. (THEOCHEM) 1996, 360, 109.
(23) Remko, M. J . Phys. Chem. A 2002, 106, 5005.
(24) Remko, M.; Mach, P.; Scheleyer, P. R.; Exner, O. J . Mol. Struct.
(THEOCHEM) 1993, 279, 139.
(25) Stinchcomb, D. M.; Pranata, J . J . Mol. Struct. (THEOCHEM)
1996, 370, 25.
(26) Remko, M. Phys. Chem. Chem. Phys. 2000, 2, 1113.
(27) Yen, S. J .; Lin, C. Y.; Ho, J . J . J . Phys. Chem. A 2000, 104,
(28) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J . A. Gaussian 98, revision A.1x; Gaussian, Inc.: Pittsburgh,
(29) Barone, M.; Cossi, B.; Mennucci, B.; Tomasi, J . J . Chem. Phys.
1997, 107, 3210.
(30) Riddik, J . A.; Bunger, W. B.; Sakano, T. K. Organic Solvents,
4th ed.; J ohn Wiley and Sons: New York, 1987; Vol. II, p 75.
(31) Tomasi J .; Persico M. Chem. Rev. 1994, 94, 2027.
(32) Kendall, R. A.; Dunning, T. H., J r.; Harrison, R. J . J . Chem.
Phys. 1992, 96, 6796.
Senent et al.
6536 J . Org. Chem., Vol. 68, No. 17, 2003
AHA presents different stable structures and shows
amide-imide tautomerism; conformational analysis pro-
vides stable acidic forms. The internal rotation of the
C-N bond that enables interconversion of the Z(cis) h
E(trans) forms is hindered by barriers higher than 25
kcal/mol;4hence, thetwoforms can betreated as different
species. The conformers generated by OH rotation and
CH3torsion can be classified as (Z)-amidic, (E)-amidic,
(Z)-imidic, and (E)-imidic forms. The(Z)- and (E)-isomers
are assumed to coexist, giving rise to different deproto-
nation processes leading toformation of independent (Z)-
and (E)-anions, respectively. Thus, the probability of the
process involving the (E)-form must be taken into con-
sideration for a proper interpretation of the mechanisms.
The minima for the MP2/cc-pVDZ and MP2/AUG-cc-
pVDZ potential energy surfaces agree well. As expected,
the diffuse functions modify slightly the relative energies
of the neutral species; the energy change ranged from
2.8% (in the most stable (E)-imide tautomer) to 15.3%
(in the most stable (Z)-imide tautomer). Table 1 lists the
relative energies of the most stable neutral forms with
respect tothe (Z)-amide tautomer (E ) -283.673447 au).
The amide forms display nonplanar geometry (Figure 1),
whereas the imide forms (not shown) adopt planar
structures with formation of a CdN double bond. The
intramolecular O3...H6-O5 H-bond and the steric hin-
drance of the methyl group cause the relative stabilities.4
Another significant feature is the difference in dipole
moment between the amide and imide forms, since the
polar solvent effect is more significant on the amide
structure; moreover, although the energy levels are
separated by relatively high barriers, the (Z)-amide, the
(E)-amide, and the (Z)-imide tautomers show similar
stability. The (E)-amide tautomer lies only 1.6 kcal/mol
abovethe(Z)-geometry; themost stable(Z)- and (E)-imide
tautomers (termed (Z)-II imide tautomer and (E)-III
imide tautomer) lie 1.1 and 5.4 kcal/mol above the (Z)-
amide form, respectively.
Totest the validity of the second-order Mo ¨ller-Plesset
approximation for determining relative energies, the
calculations of the isolated molecule have been repeated
with the MP4 method. With the exception of the (E)-
amide, whose MP4 relative energy drops from 1.6 to 0.8
kcal/mol, there are not significant changes with respect
to the data shown in Table 1.
Anion F orms. (Z)-AHA displays the three (Z)-Ia, (Z)-
Ib, and (Z)-II anions (Figure 2), the most stable of them
being (Z)-Ia. (E)-AHA gives rise to three less stable
anions. In the case of Ib-type anions, (E)-Ib represents a
conformer, whereas the geometry of (Z)-Ib is unstable,
although a rotamer of (Z)-Ib may exist. In these struc-
tures, the remaining H atom is bonded to O5, O3, and
N4, respectively; their MP2/AUG-cc-pVDZ relative ener-
gies are listed in Table 1. Extension of the basis set
causes an energy change of some 45%, this feature
indicating that calculations of the AHA anions require
the use of basis sets containing diffuse orbitals.
The (Z)-Ia and (Z)-II anions arise from geometry
optimization of the neutral (Z)-amide after elimination
of one proton, either H7+ or H6+ (see Figure 3, A ) (Z)-
amide; I ) (Z)-imide form II); in the amide tautomer, H6
is bound to O5, and H7 to N4. After elimination of H7+
(R1 process), optimization affords the most stable Ia
anion (E ) -283.121064 au). Loss of H6 (R2 process)
gives rise tothe (Z)-II anion (E ) -283.0936 au). On the
T ABL E 1.
AHA Considering the Molecule Isolated and in the Presence of Solventa
MP2/AUG-cc-pVDZ Calculated R elative E nergies, ER (kcal/mol), of the Stable Neutral and Anionic F orms of
isolated molecule PCM
aSolvation free energies,∆Gs, are given in kcal/mol. Dipole moments, µ, are given in Debyes.bEa) -283.673447 au.cEb) -283.688362
au.dEc) -283.121064 au.eEd) -283.221794 au.
F IGUR E 1. Neutral (Z)- and (E)-amidic forms of AHA.
F IGUR E 2. (Z)- and (E)-anions of AHA.
Deprotonation Sites of Acetohydroxamic Acid Isomers
J . Org. Chem, Vol. 68, No. 17, 2003 6537
other hand, H7 is bound toO3 in the (Z)-imide tautomer;
in this case, elimination of H6 (R3 process) produces the
(Z)-Ia anion, and loss of H7+ gives oneof the(Z)-Ia anion
rotamers. The conformers generated by the OH and CH3
internal rotations produce rotamers of the (Z)-Ia, (Z)-Ib,
and (Z)-II anions. Loss of H atoms from the neutral (E)-
forms gives risetothreedifferent anions (Figure4). After
elimination of H6+ or H7+ (RE2 and RE1 processes), the
(E)-amide tautomer can produce the most stable (E)-II
anion (E ) -283.101532 au) and (E)-Ia anion (E )
-283.100375 au). The(E)-imideform can producethe(E)-
Ia and (E)-Ib anions (E ) -283.075842 au) after elimina-
tion of H7+ or H6+ (RE3 or RE4 processes).
Theproduct of theR3 process is the(Z)-Ia anion rather
than the (Z)-Ib form, since this was found tobe unstable
at the MP2 level. Figure 2 shows two(Z)-structures, (Z)-
Ia and (Z)-Ib anions, connected by the interconversions
O3‚‚‚H6-O5 and O3-H6‚‚‚O5 (see Figure 1). The stable
(Z)-Ia anion has a planar structure, with the methyl
group antieclipsed with respect toO3, and H6 connected
toO5. The H-bond closes a cycle with a 2.5448 Å distance
between the two oxygens. The O3‚‚‚H6 equilibrium
intramolecular H-bond distance was evaluated to be
1.8170 Å and the MP2/AUG-cc-pVDZ dipole moment to
be 3.2959 D. However, (E)-Ia anion exhibits minimum
energy conformations. Some rotamers of the (Z)-Ib anion
shown in Figure 2 display some stability, although this
anion, denoted as 4 A2 by Yazal and Pang,33is reported
to be unstable with the Density Functional Theory.
There is no experimental evidence of the (Z)-Ib anion.
The (Z)-Ia anion T (Z)-Ib anion interconversion was
simulated searching for the minimum energy path.
Figure 5 plots the change in energy as a function of R (R
) O3H6); theremaining coordinates werefully optimized
for each selected structure. The (Z)-Ia anion (R ) 1.8170
Å) represents the unique minimum of the MP2/AUG-cc-
pVDZ path, although thegradient decreases in theregion
where the RHF curve (dashed line) shows a second
minimum (R ) 0.97 Å). RHF calculations led to two
equilibrium structures separated by a transition state
(RTS) 1.07 Å). The stability of the (Z)-Ib anion decreases
according to a correlation already observed.2
The two isomers of anion II, (E)- and Z, lie 12.3 and
17.2 kcal/mol above the (Z)-Ia anion (Table 1). The
interconversion (Z)-II anion T (E)-II anion requires that
a potential barrier of 50.3 kcal/mol be overcome. In the
(Z)-II anion, themethyl group is antieclipsed with respect
toO3. TheO3C2N4O5 frameis distorted tominimizethe
steric interactions of the two oxygens (O3O5 ) 2.9295
Å). For instance, the O5N4C2 angle changed from 107.6°
in the (Z)-Ia anion to 128.2° in the (Z)-II anion. In the
(E)-II anion, the methyl group eclipses O3; the two
oxygens are disengaged, but the charge distribution
blocks theintramolecular H-bond stabilization. TheGibbs
energy change of the interconversion (Z)-II anion T (E)-
II anion was evaluated tobe -5.7 kcal/mol, a value that
also decreases with temperature.
Solvent E ffect. In ab initiocalculations performed in
solution, water was regarded as a continuous dielectric,
characterized by a constant permittivity. The Polarizing
Continuum Model (PCM) was used, implemented in the
Gaussian 98 package.29Table1 lists therelativeenergies
(ER) of the most stable anions and neutral forms calcu-
lated with MP2/PCM/AUG-cc-pVDZ; also listed are the
solvation energies (∆GS(MP2)), defined as the difference
between the MP2 energies of the isolated species and the
corresponding values in solution and the solvation ener-
(33) El Yazal, J .; Pang, Y. P. J . Chem. Phys. 1999, 103, 8346.
F IGUR E 3. Deprotonation process of (Z)-AHA.
F IGUR E 4. Deprotonation of (E)-AHA.
F IGUR E 5.
interconversion as a function of the O3H6 bond distance.
Energy change of the O3H6‚‚‚O5O3‚‚‚H6O5
Senent et al.
6538 J . Org. Chem., Vol. 68, No. 17, 2003
gies calculated at the RHF level (∆Gs(RHF)). The two
parameters ∆GS(MP2) and ∆Gs(RHF) involve three non-
electrostatic contributions (cavitation, dispersion, and
repulsion energies). As expected, the solvation effect on
the neutral forms is negligible and somewhat larger for
amides compared to imides. This effect, however, is
important for the anions. The most stable (Z)-form is the
(Z)-Ia anion (E ) -283.221794 au), though the relative
energy of the (Z)-II anion drops from 17.2 to 6.5 kcal/
mol. For the (E)-forms, the solvent reverts the relative
order of stability of (E)-II and (E)-Ia anions and favors
(E)-N-deprotonation. The solvation free energies calcu-
lated at the Hartree-Fock level are shown in the last
column of Table 1.
The presence of solvent increases all dipole moments.
The dipole moments of the (Z)-Ia anion and the (Z)-II
anion produce strong N- and O-deprotonation, respec-
tively. A simple Mulliken analysis shows that the prox-
imity of the two oxygens in the (Z)-II anion raises both
the negative charge around the O5 and the electric
moment. As expected, thelargest solvation effect appears
in the polar species (Z)-II, which becomes stabilized by
74.0 kcal/mol upon solvent addition. Solvation augments
the Z-O-deprotonation probability.
Deprotonation Processes. The deprotonation site of
AHA still is a matter of controversy, and generally a
pronounced influence of the molecule environment is
accepted.34,35For an isolated molecule (ideal gas model),
the variation of the thermodynamic properties corre-
sponding tothe R1, R2, and R3 processes of the (Z)-forms
(Figure 3) and the four RE1, RE2, RE3, and RE4
processes (Figure 4) at 298.15, 308.15, and 318.15 K are
listed in Table 2 along with the properties calculated in
the presence of solvent. For an isolated molecule, we
consider the process AH h A-+ H+and treat the system
as an ideal gas. The entropy values of the neutral and
anionic species were taken from the Gaussian package
output. In the presence of water, the Gibbs energies can
be evaluated with the Born-Haber cycle:
where∆Gs(AH), ∆Gs(A-), and ∆Gs(H+) arethesolvation
Gibbs energies of the acid species, anion, and proton,
respectively (Table 1); ∆Gs(H+) was calculated as ∆Gs
(H3O+). Its value was calculated to be -107.5 kcal/mol
with MP2/AUG-cc-pVDZ, far from theexperimental value
(-264.61 kcal/mol) employed in ref 36; ∆Gisolated moleculewas
evaluated with the gas ideal model and from the De
dissociation energies (∆H ) De+ p∆V and ∆G ) ∆H -
T∆S). Thedetermination of Derequires calculation of the
total electronic energies and the zero-point energies
performed in a harmonic analysis (ZPVE ) 48.9 ((Z)-A),
49.0 ((E)-A), 48.9 ((Z)-I), 48.6 ((E)-I), 40.6 ((Z)-Ia), 40.0
((E)-Ia), 39.6 ((Z)-Ib), 40.7 ((Z)-II), and 40.6 ((E)-II) (kcal/
mol)). In the case of the solvent, we did not include the
solvent effect on thevibrational analysis. In addition, the
variation of the nonelectrostatic contributions to the
solvation energy with the temperature has been ne-
glected given the range of the temperature variation.
If the molecule is isolated, then the most likely process
is N-deprotonation of the (Z)-amide, which produces (Z)-
Ia anion (or formation of (Z)-Ia from the imide, in an
equilibrium mixture of amide and imide). In this case,
∆G298.15 has been calculated to be 332.0 kcal/mol, in
agreement with theexperimental data of ref 18 measured
in the gas phase (339.1 kcal/mol). This result supports
previous ab initiocalculations21,23and coincides with that
determined for formohydroxamic acid.2,24The next prob-
able processes are N and O deprotonation of the (E)-
amide; from thermodynamic considerations, (Z)-O-depro-
tonation appears tobeunlikely. Also, in solution themost
probable process is the formation of the (Z)-Ia anion from
the (Z)-amide and the (E)-Ia anion from the (E)-amide
(∆G298.15) 170.7 kcal/mol and ∆G298.15) 175.4 kcal/mol,
respectively). However, (Z)-N-deprotonation is favored.
The twoprocesses, (Z)-O-deprotonation (∆G298.15) 177.1
kcal/mol) and (E)-O-deprotonation (∆G298.15) 178.5 kcal/
mol), have a similar probability.
∆G298.15 obtained from the experimental pKHA value
(see below) was smaller than the theoretical value in
solution. The causes of the differences between experi-
mental and calculated pKHA values are discussed in
Spectrophotometric Measurements: Acid Dis-
sociation Constant. Within the ordinary pH range,
AHA behaves as a simple monoprotic acid whose dis-
sociation, represented by reaction 2,
(34) Eigen, M.; Kruse, W. Z. Naturforsch. B 1963, 18, 857.
(35) Diebler, H.; Secco, F.; Venturini, M. J . Phys. Chem. 1984, 88,
(36) Liptak, M. D.; Shields, G. S. J . Am. Chem. Soc. 2001, 123, 7314.
(37) Oliveira Silva, C.; Nascimento, M. A. C. Adv. Chem. Phys. 2002,
T ABL E 2.
T hermodynamic Properties (kcal/mol) Corresponding to the Most Probable deprotonation Processes of AHA
isolated molecule (AH h A-+ H+) PCM (AHsh A-s+ H+s)
R1: amide h (Z)-Ia anion
R2: amide h (Z)-II anion
R3: imide h (Z)-Ia anion
RE1: amide h (Z)-Ia anion
RE2: amide h (Z)-II anion
RE3: imide h (Z)-Ia anion
RE4: imide h (Z)-Ib anion
∆Gsolution) ∆Gisolated molecule- ∆Gs(AH) + ∆Gs(A-) +
HA h A-+ H+
Deprotonation Sites of Acetohydroxamic Acid Isomers
J . Org. Chem, Vol. 68, No. 17, 2003 6539
was evaluated by UV titrations (Figure 6). The absor-
bance-pH data pairs were analyzed according to the
In eq 3, A represents the absorbance measured during
titration. AA- and AHA stand for the absorbances of the
anion and the undissociated acid, respectively. KHA
represents the acid dissociation constant, and the pa-
rameter m is close to unity in aqueous solutions of pH
3-11.38An iterative minimization procedure allows the
AA-, AHA, and pKHAvalues tobe determined. Table 3 lists
the pKHAvalues at different temperatures and the ionic
strength I ) 0.2 M, which are in good agreement with
literature values,39along with the ∆H° ) 5.3 kcal/mol
and ∆S° ) -23.5 cal K-1mol-1values of reaction 2. Table
3 also lists the potentiometric results determined at the
Molar absorptivities, ?, of both the undissociated acid
and the anion were determined at several wavelengths
and at two different ionic strengths (Table 4); the
observed independence of the anion ? values of the ionic
strength reveals that dimerization does not ocurr in
aqueous solution, a result confirmed by the1H NMR
experiments described below.
1H NMR R esults. We are aware of only a single pK
value published in the literature, the discussion being
focused on which anion (O-anion or N-anion) is prefer-
entially formed. Occurrence of the (Z)- and (E)-isomers
of AHA in organic solvents has been reported, but not in
aqueous solution. Brown et al.6reported cis-trans isom-
erism in DMSO-d6by1H,13C, and15N NMR measure-
ments; NMR spectra have enabled clean assignment of
the OH and NH protons of the (Z)-amide and (E)-amide
isomers, yielding a 90:10 Z:E ratio in this solvent. On
the other hand, experiments performed by Bagnoet al.14
in aqueous solution at pH 1 and pH 12 showed a single
signal for AHA at each pH value. Theseauthors provided
three possible interpretations for this behavior: (a) the
two forms undergo fast exchange, (b) the two forms
display negligibledifferencein chemical shifts, or (c) only
a single form is present in solution. They concluded that
O-deprotonation is prevailing. However, in contrast with
these authors, at room temperature, we found two1H
NMR signals in water in the region characteristic of the
methyl resonance, this feature suggesting that the neu-
tral and anionic forms of AHA are present in aqueous
solution as (Z)- and (E)-structures.
Figure 7 shows that both the neutral AHA form (pH
) 5.20) and its anion (pH ) 11.52) exhibit two signals.
Sofar, there exists noevidence on which tautomer form,
Z or E (amide or imide, respectively), is prevailing;
however, for N-phenybenzohydroxamic acid, NMR mea-
surements allowed the signal appearing upfield to be
assigned tothe (Z)-form. Moreover, kinetic results showed
that the (Z)-tautomer is prevailing in water (92%).11
Finally, the theoretical calculations reported in this
work clearly indicate that the (Z)-form of AHA is prevail-
ing both in the gas phase and in solution. On these
grounds, it appears reasonable to assign the high-
intensity and the low-intensity signals to the (Z)- and
(E)-forms, respectively. Integration of the signals of
Figure 7 indicates that the neutral (Z)-isomer amounts
to 97% and the (Z)-anion amounts to 94% of the total
Figure 7 shows that the signals shifted tohigher fields
with increasing pH. Titration curves based on pH/
chemical shifts are displayed in Figure 8a for the higher
intensity signal ((Z)-AHA) and in Figure 8b for the lower
intensity signal ((E)-AHA). Introduction of the δAH, δA-,
and δ chemical shifts corresponding to the acidic, basic,
and intermediate forms, respectively, intoeq 3 (δ instead
of A) enables calculation of the KZ and KE microscopic
dissociation constants of the(Z)- and (E)-tautomer forms.
Since the NMR experiments were performed in deuter-
ated water, to reliably compare a reaction in water with
the same reaction in deuterium oxide, it was necessary
tousethefollowing conversion: actual pH of D2O solution
) apparent pH - γ. Given that the pKHAvalue is around
9.0 (Table 3a), application of the Burton and Shiner
treatment40leads tothe ratioKH2O/KD2O) 3.5 for an acid
species of similar strength and therefore to ∆pK ) γ )
0.54. Hence, the values referring to an aqueous solution
were pKE) 9.01 ( 0.01, and pKZ) 9.35 ( 0.03 at 25 °C.
The difference in acid strength between the twoisomers
can be ascribed to the difference in the strength of the
(38) Garcı ´ a, B.; Leal, J . M. Collect. Czech. Chem. Commun. 1987,
(39) (a) Wu, H. J . Am. Chem. Soc. 1977, 99, 1977. (b) Yamaki, R.
T.; Paniago, E. B.; Carvalho, S.; Howarth, O. W.; Kam, W. J . Chem.
Soc., Dalton Trans. 1997, 24, 4817. (c) Keeffe, J . R.; J encks, W. P. J .
Am. Chem. Soc. 1983, 105, 265. (d) Monzky, B.; Crumbliss, A. L. J .
Org. Chem. 1980, 45, 4670. (e) Ryaboi, V. I.; Shenderovich, V. A.;
Strizhev, E. F. Russ. J . Phys. Chem. 1980, 54, 730. (f) Ryaboi, V. I.
Zh. Fiz. Khim. 1980, 54, 1279.
(40) Burton, C. A.; Shiner, V. J ., J r. J . Am. Chem. Soc. 1961, 83,
F IGUR E 6. Set of spectral curves corresponding toionization
equilibrium of AHA. T ) 25 °C, I ) 0.2 M.
T ABL E 3.
pKHA Values at Different T emperatures (I )
9.451 ( 0.003
9.327 ( 0.003
9.236 ( 0.004
9.116 ( 0.002
8.975 ( 0.003
9.45 ( 0.04
9.30 ( 0.01
9.24 ( 0.05
9.11 ( 0.01
8.93 ( 0.03
8.84 ( 0.06
AA- - AHA
1 + 10-mpH+mpKHA+ AHA
Senent et al.
6540 J . Org. Chem., Vol. 68, No. 17, 2003
H-bond, either intramolecular or involving solvent; H-
bonding stabilizes the proton of the (Z)-isomer, thus
decreasing its acidity.
The distribution of AHA among different isomer spe-
cies makes it difficult to decide which of the OH or NH
forms preferentially dissociates. Figure 7 displays no
signal ascribable to the H of AHA. Although the fast
exchange with deuterium oxide prevents elucidation of
which H of each isomer is moreacidic and alsodiscussion
of whether the (Z)- and (E)-species behave as OH or NH
acids in aqueous solution, the theoretical results listed
in Table 2 show that in aqueous solution, the most likely
processes correspond toN-deprotonation of both (Z)- and
(E)-forms, namely, the R1, R3 and the RE1, RE3 pro-
cesses shown in Figures 3 and 4, respectively.
As outlined in the Experimental Section, the temper-
ature effect on the NMR signals in the 5-85 °C range
(not shown) indicates that the signal of the (E)-isomer of
the neutral species becomes broader starting from about
30 °C and overlaps with the baseline, whereas the two
signals of the anions remain unchanged within the same
temperature range; this feature points to a slow Z--
E-and a fast Z-E interconversion, relative tothe NMR
time scale. This result was confirmed by NOESY (one-
dimensional) experiments (not shown) that reveal that
the exchange between the anions is negligible at 25 °C
(even at the high mixture time of 1 s) and modest at 80
°C (mixture time 0.4 s). This feature bears relation to
the CdN double-bond character through which the
isomeric interconversion occurs and reveals that theZ-E
conversion barrier is much larger for the anions than for
the corresponding neutral molecules. Likewise, calcula-
tions in the gas phase showed that the Z-- E-inter-
conversion barrier amounts to about 10 kcal/mol above
T ABL E 4.
Molar Absorptivities, E, of Both the Protonated and Nonprotonated Species
I ) 1.00 M
I ) 0.25 M
1318 ( 47
1462 ( 87
997 ( 36
1046 ( 65
687 ( 26
730 ( 51
455 ( 22
499 ( 42
338 ( 21
383 ( 39
248 ( 20
292 ( 37
I ) 1.00 M
I ) 0.25 M
4139 ( 296
4873 ( 140
4470 ( 13
4769 ( 99
3924 ( 57
4058 ( 38
3017 ( 48
3074 ( 16
2105 ( 35
2142 ( 8
1378 ( 22
1398 ( 7
F IGUR E 7. AHA signals in D2O in the region characteristic of the methyl resonance. T ) 25°C; (a) neutral forms, pH ) 5.20; (b)
anionic forms, pH ) 11.52.
F IGUR E 8. Variation of the chemical shift, δ, in D2O as a function of pH. T ) 25 °C; (a) high-intensity signal ((Z)-AHA); (b)
low-intensity signal ((E)-AHA).
Deprotonation Sites of Acetohydroxamic Acid Isomers
J . Org. Chem, Vol. 68, No. 17, 2003 6541
that of the Z-E conversion. These results are consistent Download full-text
with a preferencefor N-deprotonation (R1 processes) over
O-deprotonation (R2 processes), since the C-N bond in
the N-anion exhibits a more extended double-bond char-
acter than in the O-anion (as inferred from the resonant
forms), which hinders the rotation and Z-- E-inter-
conversion. Comparison of the dynamic NMR results for
the neutral species with those of the anions supports the
assumption that theCdN double-bond character is much
less pronounced in the first case than in the second,
which results in theprevalenceof theneutral (CH3(O)C-
NH(OH)) amide forms ((Z)-A and (E)-A) relative to the
imide (CH3(OH)CdN(OH)) forms ((Z)-I and (E)-I). There-
fore, the reaction in Scheme 1 is suggested.
According to Scheme 1, the macroscopic dissociation
constant KHAis given by eq 4.
The NMR data of Figure 7 yield KIS) 3/97 and KIS-)
6/94. Introduction of the KIS, KZand KE values into eq 4
yields KHA ) 4.63 × 10-10, in good agreement with the
value determined at 25 °C by spectrophotometric and
potentiometric measurements (Table 3). Finally, NMR
experiments performed at different initial AHA concen-
trations demonstrated that chemical shifts were not
influenced by the substrate concentration. Consistently,
this feature supports the observation that in water, AHA
is present only in the monomer form. It should be noted,
for this purpose, that N-phenylbenzohydroxamic acid was
found to dimerize in acetone but not in water.11
Ab initio calculations indicate that the (Z)- and (E)-
isomers of acetohydroxamic acid present neutral and
anionic stable forms. NMR spectroscopy also shows the
existence of (Z)- and (E)-tautomers in aqueous solution,
which sofar had been only a theoretical hypothesis. Each
tautomer has a different acidic strength. NMR spectros-
copy allows the microscopic deprotonation constants of
the (Z)-AHA and (E)-AHA forms to be evaluated, the
combination of which provided an overall acid dissocia-
tion constant in good agreement with the value deter-
mined by spectrophotometric and potentiometric mea-
surements. Comparison of the dynamic NMR and NOESY
(one-dimensional) results for neutral species with those
for anions reveals the occurrence of N-deprotonation in
the (Z)-amide and (E)-amide in aqueous solution. The
theoretical calculations agree with this conclusion.
E xperimental Section
Chemicals. All chemicals used were analytical grade.
Twice-distilled water was used to prepare the solutions and
as a reaction medium as well. The ligand, acetohydroxamic
acid (AHA), was of the highest purity commercially available
(>99%). Stock solutions of AHA were kept in the dark at 4
°C. Perchloric acid and sodium perchlorate or potassium
chloride were used to attain the desired medium acidity and
ionic strength, respectively. NMR measurements were per-
formed in D2O (99.9%). ThepH variation in NMR was attained
with KOD (40% w/w, previously diluted with D2O).
Methods. Hydrogen ion concentration of solutions below
[H+] e 0.01 M were determined by pH readings with a pH
meter, calibrated with different buffer solutions within a pH
range of 1.6-12.5. A combined glass electrode was used after
replacing theusual KCl bridgewith 3 M NaCl in order toavoid
precipitation of KClO4. The electrode was calibrated against
sodium perchlorate-perchloric acid solutions of known con-
centration and ionic strength to directly give -log [H+].
1H NMR measurements were performed with a 400 MHz
instrument at 9.4 T (operating at 399.941 MHz). The spectra
were recorded at different temperatures between 5 and 85 °C
using a spectral window of 15 ppm; acquisition times were13.2
s. NOESY (one-dimensional) were determined with nonselec-
tive proton saturation under the following conditions: relax-
ation delay, 1.000 s; mixing between 0.2 and 1 s; acquisition
time, 0.225 s. A Gauss apodization of 0.040 s and an FT size
of 2048 × 2048 were used for data processing.
The absorption titrations were performed on a Diode Array
spectrophotometer with a Peltier accesory to control the
temperature. Experiments were performed at 15, 25, 35, 45,
55, and 65 °C. Increasing amounts of KOH were added by a
microsyringe toa solution of AHA previously thermostated in
the measuring cell. Fluctuations in temperature were within
(0.1 °C throughout. The medium acidity and ionic strength
were kept constant at the desired values during each titration.
The data were evaluated by nonlinear least-squares proce-
Acknowledgment. Financial support by Ministerio
deCiencia y Tecnologı ´ a, Spain, Projects AYA2002-02117
and BQU2002-01061, DGESIC PM98-0073, and J unta
de Castilla y Leo ´n BU26-02, Spain, are gratefully
acknowledged. The authors thank Prof. J . Tomasi for
useful suggestions and discussion.
Supporting Information Available: Theoretical results
of the neutral and anionic forms. This material is available
free of charge via the Internet at http://pubs.acs.org.
SCHE ME 1
1 + KIS
Senent et al.
6542 J . Org. Chem., Vol. 68, No. 17, 2003