Enols of carboxylic acid amides with β-electron-withdrawing substituents

Article abstract of DOI:10.1021/ja992059f

The effect of stabilizing enols of carboxamides by several two β- electron-withdrawing substituents was studied with the R¹R²CHCONHPh systems. When R¹R²CH₂ = Meldrum's acid (MA), the solid-state structure is that of the enol R¹R²C=C(OH)NHPh (7). In CDCl₃ solution the structure is 7, but there may be some exchange on the NMR time scale with a tautomer. B3LYP/6-31G** calculations show a significant preference for the enol R¹R²C=C(OH)NH₂ (12a) (R¹R²C = MA moiety) and a small preference for (MeO₂C)₂C=C(OH)NHPh (11b) over the amide structures. However, solid 11 has the amide structure (MeO₂C)₂CHCONHPh (11a). NMR spectra in CDCl₃ show >90% of 11a, but a minor species, probably 11b, is also present. In DMSO this species is not observed. The analogous dimedone-substituted anilide 10 exists both in the solid state and in solution as an enol of a ring carbonyl. Calculations show that HC(CO₂Me)₃ has a lower energy than its tautomeric enol. The effects of the push-pull structures of the enols on structural and spectrometric parameters, of the β-substituents, of the planarity of the system, of the acid derivative group (ester or anilide), and of the solvent as enol-stabilizing factors are discussed. Destabilization of the acid form contributes to the increased relative stability of the enols.

Full text of DOI:10.1021/ja992059f

J. Am. Chem. Soc. 2000, 122, 1325-1336
1325
Enols of Carboxylic Acid Amides with â-Electron-Withdrawing
Substituents
Jayanta Kumar Mukhopadhyaya, Stepan Sklena´k, and Zvi Rappoport*
Contribution from the Department of Organic Chemistry and the MinerVa Center for Computation
Quantum Chemistry, The Hebrew UniVersity, Jerusalem 91904, Israel
ReceiVed June 17, 1999. ReVised Manuscript ReceiVed NoVember 18, 1999
Abstract: The effect of stabilizing enols of carboxamides by several two â-electron-withdrawing substituents
was studied with the R1R2CHCONHPh systems. When R1R2CH2 ) Meldrum’s acid (MA), the solid-state
structure is that of the enol R1R2CdC(OH)NHPh (7). In CDCl3 solution the structure is 7, but there may be
some exchange on the NMR time scale with a tautomer. B3LYP/6-31G** calculations show a significant
preference for the enol R1R2CdC(OH)NH2 (12a) (R1R2C ) MA moiety) and a small preference for
(MeO2C)2CdC(OH)NHPh (11b) over the amide structures. However, solid 11 has the amide structure (MeO2C)2-
CHCONHPh (11a). NMR spectra in CDCl3 show >90% of 11a, but a minor species, probably 11b, is also
present. In DMSO this species is not observed. The analogous dimedone-substituted anilide 10 exists both in
the solid state and in solution as an enol of a ring carbonyl. Calculations show that HC(CO2Me)3 has a lower
energy than its tautomeric enol. The effects of the push-pull structures of the enols on structural and
spectrometric parameters, of the â-substituents, of the planarity of the system, of the acid derivative group
(ester or anilide), and of the solvent as enol-stabilizing factors are discussed. Destabilization of the acid form
contributes to the increased relative stability of the enols.
Introduction
Similar nucleophilic addition to ketenes substituted by two bulky
aryl groups (e.g., Ar ) 2,4,6-R3C6H2; R ) Me, i-Pr) generated
sufficiently long-lived enols which were observed by NMR
before tautomerizing to the acid derivatives 2.6
The relatively low Kenol values were ascribed to stabilization
of the acid form by resonative electron donation from the
heteroatom X (X ) OH, OR, NRR′, OCOR′′, halogen; cf. hybrid
2b, eq 1).2a A similar resonative electron donation from X should
also stabilize enols 1. However, since oxygen is more elec-
tronegative than carbon, the contribution of 1b to 1 when R1
and R2 are not strongly electron-withdrawing is lower than that
of 2b to 2, and Kenol is much lower than when X ) H.
In contrast with the extensive data on stable and short-lived
enols of aldehydes and ketones,1 enols of carboxylic acids and
their derivatives (1a) are regarded as very unstable compared
with their carbonyl analogues (2a) (eq 1).2 Except for a few
(4) (a) Graham, L.; Williams, D. L. H. J. Chem. Soc., Chem. Commun.
1991, 407. (b) March, J. AdVanced Organic Chemistry, 4th ed.; Wiley: New
York, 1992; p 590. (c) Williams, D. L. H.; Xia, L. J. Chem. Soc., Perkin
Trans. 2 1992, 747. (d) Nguyen, M. T.; Hegarty, A. F. J. Am. Chem. Soc.
1983, 105, 381. Hegarty, A. F.; Kelly, J. G.; Relihan, C. M. J. Chem. Soc.,
Perkin Trans. 2 1997, 1175. (e) Williams, D. L. H.; Xia, L. J. Chem. Soc.,
Perkin Trans. 2 1993, 1429. (f) Quinkert, G.; Kleiner, E.; Freitag, B.;
Glenneberg, J.; Billhardt, U.; Cech, F.; Schmieder, K. R.; Schudok, C.;
Steinmetzer, H.; Bats, J. W.; Zimmerman, G.; Du¨rner, G.; Rehm, D.; Paulus,
E. F. HelV. Chim. Acta 1986, 69, 469. (g) Vilsmaier, E.; Joerg, K.; Mass,
G. Chem. Ber. 1984, 117, 2947. (h) Bruce, M. I.; Walton, J. K.; Williams,
M. L.; Skelton, B. W.; White, A. H. J. Organomet. Chem. 1981, 212, C35.
(i) Galanski, M.; Keppler, B. K.; Nuber, B. Angew. Chem., Int. Ed. Engl.
1995, 34, 1103.
(5) (a) Andraos, J.; Chiang, Y.; Kresge, A. J.; Pojarlieff, I. G.; Schepp,
N. P.; Wirz, J. J. Am. Chem. Soc. 1994, 116, 73 and references therein to
Kresge’s and Wirz’s work. (b) Chiang, Y.; Jefferson, E. A.; Kresge, A. J.;
Popik, V. V.; Xie, R.-Q. J. Phys. Org. Chem. 1998, 11, 610. (c) Wagner,
B. D.; Arnold, B. R.; Brown, G. S.; Lusztyk, J. J. Am. Chem. Soc. 1998,
120, 1827.
(6) (a) O’Neill, P. J.; Hegarty, A. F. J. Chem. Soc., Chem. Commun.
1987, 744. Allen, B. M.; Hegarty, A. F.; O’Neill, P.; Nguyen, M. T. J.
Chem. Soc., Perkin Trans. 2 1992, 927. Allen, B. M.; Hegarty, A. F.;
O’Neill, P. J. Chem. Soc., Perkin Trans. 2 1997, 2733. (b) Frey, J.;
Rappoport, Z. J. Am. Chem. Soc. 1995, 117, 1161; 1996, 118, 5169, 5182.
(c) Frey, J.; Rappoport, Z. J. Am. Chem. Soc. 1996, 118, 3994. (d)
Rappoport, Z.; Frey, J.; Sigalov, M.; Rochlin, E. Pure Appl. Chem. 1997,
69, 1933.
earlier reports,3 only recently have some of these enols been
suggested, detected, or prepared as short-lived intermediates,4-6
and X-ray data of three derivatives were reported.4g-i Simple
systems, i.e., neither highly activated electronically nor highly
sterically crowded, were generated by nucleophilic addition to
ketenes or by enolization of an R-hydrogen to the acid derivative
function. They were observed by flash photolysis5 or suggested
as intermediates on the basis of kinetic and other evidence.3,4
(1) The Chemistry of Enols; Rappoport, Z., Ed.; Wiley: Chichester, 1990.
(2) For reviews on carboxylic acid enols, see: (a) Hegarty, A. F.; O’Neill,
P. In ref 1, Chapter 10, p 639. (b) Kresge, A. J., Chem. Soc. ReV. 1996, 25,
275.
(3) (a) Zimmerman, H. E.; Cutshall, T. W. J. Am. Chem. Soc. 1958, 80,
2893. (b) Zimmerman, H. E.; Cutshall, T. W. J. Am. Chem. Soc. 1959, 81,
4305. (c) Ried, W.; Junker, P. Justus Liebigs Ann. Chem. 1967, 709, 85.
(d) Lillford, P. J.; Satchell, D. P. N. J. Chem. Soc. B 1968, 897. Poon, N.
L.; Satchell, D. P. N. J. Chem. Res., Synop. 1983, 182. However, see: Sikaly,
H. R.; Tidwell, T. T. Tetrahedron 1986, 42, 2587. (e) Deno, W. C.; Billups,
W. E.; Distefano, R. E.; McDonald, K. M.; Schneider, S. J. Org. Chem.
1970, 35, 278. (f) Quinkert, G. Pure Appl. Chem. 1973, 33, 285. (g)
Montoneri, E.; Tempesti, E.; Giuffre, L.; Castold, A. J. Chem. Soc., Perkin
Trans. 2 1980, 662. (h) Hendon, J. E.; Gordon, A. W.; Gordon, M. J. Org.
Chem. 1972, 37, 3184. (i) Kagan, J.; Tolentino, L.; Ettlinger, M. G. J. Org.
Chem. 1975, 40, 3085.
10.1021/ja992059f CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/05/2000

1326 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Mukhopadhyaya et al.
A number of recent theoretical calculations7 and estimations8
had confirmed that Kenol values for simple acid derivatives will
be very low. The order of Kenol values for the parent H2Cd
C(OH)X (i.e., 1, R1, R2 ) H) as a function of X is H > alkyl
> OCHO > Br ∼ Cl > F > NH2 > NMe2 > OH, OMe.7a
Hence, enols of anhydrides and amides have a better chance of
being observed than enols of acids and esters.
Hybrid 1b should be more important and Kenol higher when
Câ carries negative charge delocalizing electron-withdrawing
groups (EWGs) R1 and R2, as found for Kenol of simple enols
1, X ) H.1 The effect of EWG on the stability of enols of acid
derivatives was not directly investigated, although occasional
reports had appeared (vide infra).
HCl,16b and (6c)16c appear as enols of amides 6 with an enolic
OH hydrogen bonded to a neighboring carbonyl. Finally, solid
pentakis(methoxycarbonyl)cyclopentadiene, HC5(CO2Me)5, ap-
pears as an enol of ester.17
Consequently, two â-EWGs may stabilize enols of amides
or esters in the solid. However, the structure in solution may
be different and should be determined, and support by calcula-
tions is desirable. We describe below such a combined approach
and an unequivocal example of an enol of amide.
Results
Synthesis and Structural Assignment of an Enol of Amide
Stabilized by a Meldrum Acid Moiety. Pak et al.18 reacted
alkyl and aryl isocyanates with Meldrum’s acid and wrote the
products as the enols, but their spectra could also fit the amide
form.18 In a similar reaction with phenyl isocyanate (eq 2), we
A possible complication is that the EWG may be the
enolization site. For example, in a competitive enolization of
an acid derivative (COX) with an aldehyde/ketone (COR), the
enolization will be exclusively on the COR group, since the
calculated Kenol value is 10 orders of magnitude higher for a
COR than for a COX group.7a Indeed, RCOCH(CO2Et)2, ArCH2-
COCH2COX, X ) OEt, NHPh, and â,â′-diketoesters enolize
exclusively on the keto group.9 Even with powerful EWGs,
enolization to give 1 is not necessarily observed: (O2N)2-
11
CHCOOEt10 and HC(CO2Me)3 are fully “ketonic”.
A search of the Cambridge Structural Database (CSDB) for
possible stable solid-state structures having the CdC(OH)NRR′
moiety reveal several such structures substituted by EWGs, such
as nitromalonamide,12 3a-d,13,14a and 4.14b X-ray structures of
obtained 5-(R-phenylamino-R′-hydroxy)methylene Meldrum
acid 7 in 74% yield. Its structural assignment in solution as 7
rather than the amide 8 is based on the room-temperature H
NMR (CDCl3) spectrum: δ 1.78 (Me), 7.26-7.47 (Ph, m),
11.16 (1H, br), and 15.70 (<1H, br, rapidly disappearing on
shaking the solution with D2O). The data are consistent with
assignment of the latter signals as NH and OH, respectively,
i.e., with structures 7 and 9 but not with 8. The δ 15.66 signal
1
(1H) is broad at 240 K and shifts to δ 15.62 and becomes sharper
at 223 K (Figure 1a). At 325 K, it is broad with intensity <1H.
The 1H NMR spectrum in CCl4/CDCl3 is similar. In CD3CN or
DMSO-d6 the spectrum is similar, but no signal was observed
at δ 15-16 (Figure 1b).
In the room-temperature 13C NMR spectra in CDCl3 and
DMSO-d6, most signals are at nearly the same δ values, but in
DMSO-d6 there are two signals of approximately the same
intensity at 166.48 (m) and 168.37 (s) ppm, whereas three
signals appear in CDCl3 at 164.32 (br, low intensity), 169.15,
and 170.84 (br, low intensity) ppm (Figure 2), and they become
sharp at 223 K (Figure 2a, inset). At 325 K, the signals are at
δ 169.26 and 170.82 with nearly identical intensity and δ 164.21
tetracyclines usually display a OdCsCH(CONHR)sCdO
moiety as part of ring A.15 However, solid (6a)‚HBr,16a (6b)‚
(7) (a) Sklenak, S.; Apeloig, Y.; Rappoport, Z. J. Am. Chem. Soc. 1998,
120, 10359 and refs 9 and 11 therein for work before 1996. (b) Gao, J. J.
Mol. Struct. (THEOCHEM) 1996, 370, 203. (c) Rosenberg, R. E. J. Org.
Chem. 1998, 63, 5562. (d) Rapoet, G.; Nguyen, M. T.; Kelly, S.; Hegarty,
A. F. J. Org. Chem. 1998, 63, 9669. (e) Sung, K.; Tidwell, T. T. J. Am.
Chem. Soc. 1998, 120, 3043. (f) Yamataka, H.; Rappoport, Z. Unpublished
results.
(8) (a) Bok, L. D. K.; Geib, K. H. Z. Phys. Chem. Abt. A. 1939, 183,
353. (b) Guthrie, J. P. Can. J. Chem. 1993, 71, 2123. (c) Guthrie, J. P.;
Liu, Z. Can. J. Chem. 1995, 73, 1395. (d) Amyes, T. L.; Richard, J. P. J.
Am. Chem. Soc. 1996, 118, 3129.
(9) Toullec, J. In ref 1, Chapter 6, (a) p 323; (b) p 357; (c) p 361.
(10) Kissinger, L. W.; Ungnade, H. E. J. Org. Chem. 1958, 23, 1340.
(11) Guthrie, J. P. Can. J. Chem. 1979, 57, 1177.
(12) Simonsen, O.; Thorup, N. Acta Crystallogr. Sect. B 1979, 57, 1177.
(13) 3a: de Mester, P.; Jovanovic, M. V.; Chu, S. S. C.; Biehl, E. R. J.
Heterocycl. Chem. 1986, 23, 337. 3b: Gavuzzo, E.; Mazza, F.; Carotti, A.
Casini, G. Acta Crystallogr. Sect. C 1984, 40, 1231. 3c: Bolton, W. Acta
Crystallogr. 1963, 16, 950.
(14) (a) Bideau, J.-P.; Bravic, G.; Filhol, A. Acta Crystallogr. Sect. B
1977, 33, 3847. Bideau, J. P.; Huong, P.; Toure, S. Acta Crystallogr. Sect.
B 1976, 32, 481. (b) Tranqui, D.; Vicat, J.; Thomas, M.; Pera, M. N.; Fillion,
H.; Duc, C. L. Acta Crystallogr, Sect. B 1976, 32, 1724.
(15) For a discussion, see: Gilli, G.; Bertolasi, V. In ref 1, Chapter 13,
pp 713-764, especially pp 736-740.
(16) 6a: de C. T. Carrondo, M. A. A. F.; Matias, P. M.; Heggie, W.;
Page, P. R. Struct. Chem. 1994, 5, 73. 6b: Bordner, J. Acta Crystallogr.
Sect. B 1979, 35, 219. 6c: Glatz, B.; Helmchen, G.; Muxfeldt, H.; Porcher,
H.; Prewo, R.; Senn, J.; Stezowski, J. J.; Stojda, R. J.; White, D. R. J. Am.
Chem. Soc. 1971, 101, 2171.
(17) For a review, see: Bruce, M. I.; White, A. H. Aust. J. Chem. 1990,
43, 949.
(18) Lee, H. K.; Lee, J. P.; Lee, G. H.; Pak, C. S. Synlett 1996, 1209.

Enols of Carboxylic Acid Amides
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1327
Figure 3. ORTEP drawing of 7.
a small new signal appears at δ 8.80, and δ 15.58 (0.55H) is
broad. On cooling a solution which stood 16 h at 345 K to room
temperature, the signal at δ 2.11 still remains, and the aromatic
region is more complex than that in the original room-
temperature spectrum. The room-temperature 13C NMR spec-
trum resembles that in CDCl3. At 360 K, δ 164.56 disappears,
and new signals appear at δ 30.77 and 129.18 and remain on
cooling a sample which stood for 30 min at 360 K to room
temperature. Hence, decomposition takes place at 345 and 360
K.
1
Figure 1. (a) H NMR spectrum of 7 in CDCl3 at 223 K: A, Me
1
signal; B,C, Ph signal; D, N-H signal; E, O-H signal. (b) H NMR
spectrum of 7 in DMSO-d6 at room temperature: A, Me signal; B,
DMSO signal; C,D, Ph signal; E, ?; F, N-H signal.
1
The H NMR spectrum in CD3CN at 220 K displayed OH,
NH, Ph, and Me signals. The aromatic region showed several
main signals at δ 7.35-7.57 and a small seven-line multiplet
at δ 7.24-7.30 with 20% of the aromatic intensity. These and
additional very small signals, e.g., at δ 11.4, indicate that more
than one species may be present. In the 13C NMR spectrum,
CR(m) is at δ 168.83, Câ(m) at δ 73.29, and δ(CO) at 163.36
(s) and 170.39 (s).
The δ values in the solid-state vacptppm and vacptppmnqs
13C spectra are similar to those in CDCl3 and DMSO and the
approximate values predicted by Chemdraw 4.5 (cf. 7′), sug-
gesting a similar structure in solution and in the solid state.
Figure 2. 13C NMR spectrum of 7 at room temperature and in (a)
CDCl3 (inset at 223 K) and (b) DMSO: A, Me signal; B, Câ signal; C,
DMSO signal; D, CMe2 signal; E-H, Ph signals; I,J, CO, CR signals.
is weaker. In 2,2-ditipyl-1-aminoethenols, CR is at δ 156-157,6d
and hence we ascribe the signal at δ 168.27 to CR. The other
two signals are due to the ester carbonyls.
In the coupled spectra Câ is at δ 75.31 (d, J ) 7.5 Hz) (in
ditipylacetic acid enol derivatives it is at δ 84-85.56c,d), thus
excluding structure 8 with a C-H coupled Câ. The δ difference
between the vinylic carbons, ∆CRâ ) 93 ppm, is a higher value
than that (74 ppm) found for the enols of ditipylacetamides.6d
The aromatic signals are identified by their multiplicites, the
lower intensity of Cp than Co/Cm, and the NOESY spectrum:
the higher field 1H triplet shows a cross-peak with δ 126.88;
thus it is assigned as Cp. The 2H triplet at δ 7.39 shows a cross-
peak with the 13C signal at δ 122.59 and is assigned as Co. The
remaining 2H Cm signal correlates with δ 129.73. These
assignments are used for all analogous cases. In the HMBC
correlated spectrum, the N-H signal shows a cross-peak with
Co, and the O-H signal shows no cross-peak.
In CDCl3 or DMSO-d6 to which a 1:1 H2O/D2O mixture was
added, the 13C signals of 7 show deuterium-induced shifts. In
CDCl3, the ortho and ipso proton signals appear as two lines
with spacings of 7 and 6 Hz, respectively.
X-ray Data of 7. The X-ray diffraction of solid 7 was
determined. The ORTEP picture is given in Figure 3, selected
bond lengths and angles are given in Table 1, and other bond
lengths, angles, positional and thermal parameters and a
stereoview are given in Tables S1-S3 and Figure S1 of the
Supporting Information. The main features are the following:
(a) The C1-C2 bond of 1.426(5) Å is longer than a CdC bond,
shorter than a single bond, and in the range for “push-pull”
enamines. (b) The C-N bond length is 1.333 Å. In a plot of
C1-C2 vs C-N bond lengths in enamines, the point for 7 lies
very close to the correlation line.19 (c) The C5-O5 bond length
of 1.221 Å is a normal value for a not strongly hydrogen bonded
The 1H NMR in Cl2CDCDCl2 at 245 K showed signals at δ
1.67 (6H), 7.09 (1H), 7.21-7.52 (4H), 10.99 (0.9H), 15.54-
(0.09H). At 345-360 K, a new δ 2.11 signal (1.7H, br) appears,

1328 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Table 1. Selected Crystallographic Data for 7
Mukhopadhyaya et al.
Table 2. Selected Crystallographic Data for 11
bond length, Å
angle, deg
bond length, Å
angle, deg
C3-C2-C5
C1-C2
C1-O1
C1-N
C3-O2
C5-O5
C2-C3
C2-C5
C3-O3
C5-O4
1.426(5)
1.300(4)
1.333(4)
1.240(4)
1.221(4)
1.420(5)
1.414(5)
1.341(4)
1.361(4)
C3-C2-C5
C1-C2-C3
C1-C2-C5
N-C1-C2
C2-C3-O2
C2-C5-O5
C2-C5-O4
C2-C3-O3
O1-C1-N
120.1(3)
117.2(3)
122.4(3)
121.5(3)
116.9(3)
126.1(3)
116.7(3)
118.6(3)
118.4(3)
120.1(3)
C1-C2
C1-O1
C1-N
C10-N
C3-O2
C7-O3
C2-C3
C2-C7
C3-C4
C6-C7
1.470(5)
1.264(4)
1.328(4)
1.407(4)
1.310(4)
1.246(4)
1.374(4)
1.444(4)
1.490(5)
1.495(5)
118.8(3)
118.4(3)
122.8(3)
118.7(3)
121.9(3)
118.1(3)
122.5(3)
114.4(3)
121.8(3)
119.4(3)
C1-C2-C3
C1-C2-C7
N-C1-C2
C2-C3-O2
C2-C7-C6
C2-C7-O3
C2-C3-C4
O1-C1-N
O1-C1-C2
6 C-C in Ph 1.355(6)-1.385(6) O1-C1-C2
O1-H(1O1)
O2-N(1O1)
N-H(1N)
1.125
1.385
0.942
O1-H(1O1)-O2 158.24
N-H(1N)-O5 138.54
6 C-C in Ph 1.371(5)-1.385(4) O1-H(O2)-O2 162.0 (154)a
O2-H
O1-H
O3-H
N-H
1.07(5) [0.85(5)]a N-H(1N)-O3 143.4 (142.0)a
1.62(5) [1.43]a
1.76 [1.85]a
O(5)-H(1N) 1.939
1.01(4) [(0.94(3)]a
a Data for a second crystallographic form.
following: (a) A normal Csp2sCsp2 C1sC2 single bond of 1.470
Å is found. (b) The C7sO3 bond is a short normal CdO bond,
and the C3sO2 length of 1.310 Å is closer to a single bond
length. The amide carbonyl C1-O1 bond length at 1.264 Å is
normal. (c) C3-C4 and C6-C7 have normal sp2-sp3 C-C
bond lengths. (d) The bond angles around C1 and C2 are
118.4°-122.8°, as expected. (e) The O1-O2 (O1′-O2′) dis-
tances are 2.44 (2.43) Å, and in the O-H-O moiety form
O2-H (O2′-H′) ) 1.07 (0.85) Å and the O1-H (O1′-H′)
H-bond is 1.62 (1.43) Å; i.e., the structure is unequivocally 10c.
(f) The N1-O3 and N1′-N3′ distances are 2.65 Å. The O3-
H(N1) distances are 1.76 and 1.85 Å in both forms. (g) The
dihedral O1C1N/C3C2C7 angle is 4.7°.
1
(b) In Solution. The H NMR spectrum of 10 in CDCl3
Figure 4. ORTEP drawing of 10c.
displays signals for Me, two CH2, Ph, NH, and a broad enolic
singlet (rapidly exchanging with D2O) at δ 1.15, 2.41 and 2.53,
7.17-7.56, 11.80, and 17.80. In DMSO-d6, the spectrum is
similar, but δ(OH) ) 17.46. The spectrum is consistent with
enolic forms 10b-d but not with the “amide” 10a. The 13C
NMR signals in CDCl3, in DMSO-d6, and in the solid state are
at approximately the same δ values (cf. 10′).
CdO. The C3-O2 bond length of 1.240 Å is close to the normal
value. The “amide carbonyl” C1-O1 bond is much longer
(1.300 Å) and closer to a single C-O bond length, as expected
for 7. (d) The two C2sCdO bond lengths (1.414, 1.420 Å)
are consistent with structure 7 having a Csp2-Csp2 bond. (e)
The bond angles around C1 and C2 are 118.4°-122.4°, as
expected for 7. However, C1 and C2 are also sp2-hybridized in
9, where C1-C2 is a single bond. (f) The C2-C3-O3 angle
of 118.6° fits either 7 or 9. (g) The O1-O2 distance is 2.46 Å,
the O1-H-O2 angle is 158.2°, and the O1-H distance is 1.125
Å, whereas the O2-H distance is significantly longer at 1.385
Å. This asymmetric nonlinear H-bond is consistent with form
7, rather than with 9. The N-O5 distance is 2.72 Å. The N-H
bond length of 0.942 Å, the O5-H bond of 1.939 Å, and the
N-H-O5 angle of 138.5° indicate a less symmetric, less linear,
and weaker H-bond than the O2-H-O1 H-bond. (h) The
system is nearly planar. H(N1) and O2 are e0.023 Å and C5 is
0.055 Å below the O1C1N plane, C2 and C3 are ca. 0.045 Å
above it, and only O5 is significantly (0.22 Å) below it. The
dihedral angle between planes O1C1N/C3C2C5 is 4.5°. Atoms
C1, N, O2, and O5 are 0.14-0.15 Å, H(N1) is 0.10 Å, O1 is
0.23 Å, and H(O1) is 0.27 Å above the C3C2C5 plane.
On shaking of a sample of 10 with D2O in THF, the OH
signal exchanges first, and the NH proton exchanges after a
longer time, but the CH2 groups do not exchange. After a
relatively short exposure to D2O, two deuterated samples, D1
with ca. 80% OD/30% ND, and D2 with ca. 90% exchanged
OH/78% ND, were obtained. The residual OH signal is doubled
to δ 17.79/17.81 singlets with 1:2 (D2) and 1:3 (D1) intensity
ratios. The NH appears as two signals at δ 11.73/11.75 with
ca. 7:3 (D1) and ca. 8:2 (D2) intensities.
The Dimedone Derivative, 10. (a) X-ray Data. The X-ray
data of 10 show two crystallographically different molecules,
mostly with similar bond lengths and angles. The structure is
that of enol 10c. The ORTEP drawing of one is given in Figure
4, selected bond lengths and angles are given in Table 2, and
other bond lengths, angles, positional and thermal parameters
and the stereoview are given in Tables S4-S6 and Figure S2
of the Supporting Information. The main features are the
For D1 in CDCl3, the coupled 13C NMR spectrum at 223 K
displays the expected δ values and couplings expected for 10c:
δ 168.83 (br s, CONHPh), 101.74 (d, coupled with OH, Câ),
193.41 (t, 3J ) 6.8 Hz, coupled with CH2, CO), and 194.17 (t,
(19) Oleksyn, B. J.; Stadnicka, K.; Sliwinski, J. In The Chemistry of
Enamines; Rappoport, Z., Ed.; Wiley: Chichester, 1994; Chapter 2, p 87.

Enols of Carboxylic Acid Amides
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1329
Figure 5. 1H NMR spectrum of 11 in CDCl3 at room temperature.
Signals A-F are those of 11a. Bottom, full spectrum: A, Me signal;
B, C-H signal; C-E, Ph signals; F, N-H signal. Signals G and H are
the N-H and O-H signals of an isomer (11b, 11c, or 11d). Top:
Expansion of the range of the F, G, and H signals.
3J ) 6.8 Hz, coupled with CH2 and OH, CR). At room
temperature. the D1 sample shows low-field signals at δ 197.41
and 194.17 and a new one at δ 193.09. In D2, δ 194.17 nearly
disappears, and δ 193.09, which is presumably an isotopically
induced shifted signal, grows. The amide signal is also split to
δ 168.90/169.29. At room temperature and at 223 K, the D1
spectrum shows splitting mostly of <0.12 ppm of many signals
to pairs with ca. 3:1 intensities.
The IR spectrum in Nujol displays broad concentration-
independent peaks at 3384, ca. 3187 (w) (OH and NH), 2820-
2480 (H-bonded OH), 1720 (s), and 1654 (CdO) cm-1. Sample
D1 displays a broad peak at ca. 2320 cm-1 and three much less
intense absorptions >2400 cm-1. In CHCl3, the peaks are at
3420-3400 (br), 1646-1640, and 1720 (weaker) cm-1, with
no signals at 2850-2400 cm-1. Since for dimedone δCO(CHCl3)
) 1712 cm-1, the data are consistent with structures 10b-d.
The Malonic Ester Derivative, 11. The malonate 11 shows
in CDCl3 at room temperature signals at δ 3.86 (6H, Me), 4.49
(1H, CH), 7.12-7.58 (Ph, m), and 9.25 (small NH). The CH
and NH signals exchange rapidly with D2O. Significantly, two
signals are observed at δ 17.05 (OH, 0.05H) and 3.81 (CH,
0.30H), as well as other aromatic and aliphatic signals (Figure
5). The two lower 13C NMR signals at room temperature are at
δ 159.60 and 166.10 (Figure S3a, Supporting Information). The
Figure 6. Aliphatic range of the 1H NMR spectrum of 11 in CDCl3 at
220 K. Signals F and I are those for the MeO and CH groups of 11a.
Signals A-E, G, and H are due to MeO groups of two or more of the
minor isomers (11b, 11c, or 11d).
IR spectrum in Nujol shows CdO peaks at 1751 and 1697 cm-1
.
In DMSO-d6, δ(CH) ) 4.77, and δ(NH) ) 10.33, but no OH
signal (δ >11) is observed (Figure S3b).
(MeO2C)2CHCONHPh
(MeO2C)2CdC(OH)NHPh
11a
11b
(MeO2C)2CHC(OH)dNPh
11c
MeO2CC(CONHPh)dC(OH)OMe
11d
Figure 7. ORTEP drawing of 11a.
The data are consistent with the presence of ca. 95% of the
amide 11a in CDCl3 solution. The minor isomer (ca. 5%) is an
enol, either 11b, 11c, or 11d. The 11a h 11b (or 11c or 11d)
equilibrium is rapid, judging by the rapid exchange of the CH
signal with D2O. The exchange was slower at 220 K, where a
signal at δ 9.61 (N-H of 11a) and two 0.05-0.02H signals at
δ 11.9 and 17.2 (presumably the N-H and O-H of 11b, 11c,
or 11d) are observed. These and the appearance of several small
signals at δ 3.78-3.94, close to the major MeO signal (Figure
6), suggest that more than one of the species 11b, 11c, or 11d
is present in small percentages.
positional parameters, and the stereoview are in Tables S7-S9
and Figure S4 in the Supporting Information. There are three
independent molecules, and in one (designated with ′′) C6 is
disordered. One molecule displays a H-bonded homopolymeric
network where O1 of one molecule is H-bonded to the NH of
another, whose O1 is H-bonded to the NH of a third one (Figure
S5, Supporting Information). The two other types of molecules
form a similar heteropolymer array, with a repeating second
molecule-third molecule H-bonded unit (Figure S6). The C1-
O1 (and C1′-O1′, C1′′-O1′′) bond length of 1.216(4) Å (1.25-
(4), 1.216(4) Å) is definitely a CdO bond, and C1-C2 (and
C1′-C2′, C1′′-C2′′) of 1.528(5) Å (1.521(3), 1.520(5) Å) is a
single bond.
Crystal Structure of 11. The crystal structure of 11 was
determined by X-ray diffraction as that of the amide 11a. The
ORTEP drawing is in Figure 7, selected bond lengths and angles
are in Table 3, and other bond lengths and angles, thermal and

1330 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Chart 1. Relative Energies of 12a-e
Mukhopadhyaya et al.
Table 3. Bond Lengths and Angles in Solid 12a
was performed, and vibrational frequencies, zero point energies
(ZPE), and ∆H and ∆G values (E ) electronic energy; H ) E
+ EZPE; G ) Gibbs free energy) were calculated.
bond length, Å
angle, deg
C1-C2
C1-O1
C2-C3
C2-C5
C3-O2
C3-O3
C4-O3
C5-O4
C5-O5
C6-O5
C1-N1
C7-N1
1.528(5)
1.216(4)
1.507(5)
1.499(5)
1.175(5)
1.318(5)
1.446(6)
1.200(4)
1.315(4)
1.467(6)
1.348(5)
1.423(5)
C1-C2-C3
110.6(3)
109.3(3)
113.5(3)
109.3(3)
121.3(3)
127.9(3)
116.4(3)
Compound 12. The calculations show five different distinct
species of 12 (12a-e, Chart 1, where “Cyc” ) O-CMe2-O
part of the Meldrum’s acid (MA) moiety), each having several
stable conformers. Geometries of 12a-c are in Chart 2, and
selected optimized dihedral angles and the structures of 12d
and 12e are in Table S10 and Figure S7 in the Supporting
Information. The most stable isomer is the enol of amide 12a,
which has only two stable conformers: the intramolecularly
H-bonded syn form is 19.9 kcal/mol more stable than the gauche
form (CdCsOsH dihedral angle ) -165.5°). Isomer 12b(syn),
with an enol on the ring ester carbonyl, is less stable by 1.2-
1.3 kcal/mol than 12a. 12b has two conformers: the intramo-
lecularly H-bonded syn is favored by 20.6 kcal/mol over the
anti form (CdCsOsH angle ) 177.6°). The 12a/12b inter-
conversion requires a shift by a fraction of an angstrom of the
enolic hydrogen between the two oxygens.
C1-C2-C5
N1-C1-C2
C3-C2-C5
O1-C1-C2
C1-N1-C7
N1-C7-C12
6 inter-ring CCC 120.0(4) ( 0.8
6 C-C in Ph 1.376(6) ( 0.009
a Data for one independent molecule are given. The values for the
other two molecules are very similar.
r-Aryl and r-Alkyl Malonate Esters. Since diethyl ben-
zylmalonate was reported to be 13% enolic,20 we searched for
enols in systems RCH(CO2Et)2, R ) Ph, PhCH2 or Ph3CCH-
(CO2Me)2. Their 1H NMR spectra in CDCl3 are consistent with
the diester structures.
Third in stability is the amide form 12c, whose most stable
(20) Skarzewski, J. Tetrahedron 1989, 45, 4593. However, it seems that
the compound meant is diethyl benzoylmalonate. Tarbell, D. S.; Price, J.
A. J. Org. Chem. 1957, 22, 245.
Theoretical Calculations. Relative stabilities of various
isomers were calculated for the Meldrum acid derivative 12,
(21) (a) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T. A.; Petersson,
G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andress, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R.
L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A. GAUSSIAN 94, Revision C.2;
Gaussian Inc.: Pittsburgh, PA, 1995. (b) Hehre, W. J.; Radom, L.; Schleyer,
P. v. R.; Pople, J. A. Ab Initio Molecular Orbital Theory; Wiley: New
York, 1986. (c) Becke, A. D. J. Phys. Chem. 1993, 98, 5648. (d) Lee, C.;
Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
the diester 11, and the triester 13 by using the density functional
theory hybrid method (B3LYP/6-31G**).21 Full optimization

Enols of Carboxylic Acid Amides
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1331
Chart 2. Calculated Structures of 12a-e
Chart 3. Relative Energies of 11a-c
conformer is ca. 15 kcal/mol less stable than 12a. The imino
form 12d is less stable than 12c and 12a by ca. 13 and 29 kcal/
mol, respectively. The least stable isomer is the enamine 12e;
its most stable calculated conformer is ca. 16 kcal/mol less stable
than 12c and ca. 3 kcal/mol than 12d.
12a, 12b, and 12d have an approximately planar N(O)Cd
C(CO2)2 moiety with a C(Od)sCsC(dO)sO dihedral angle
of 5-8°. 12c and 12e have a boat structure with Me2C and
CHCONH2 (or CHC(OH)dNH) groups above the central plane
containing the two ester groups. The chair conformers have
higher energies. The OO distances relevant to the hydrogen
bonds in 12a, 12b, and 12d are 2.50, 2.41, and 2.55 Å,
respectively.
The solvent effects on ∆E [(12a) - (12b)] were briefly
investigated by applying SCRF ) SCIPCM calculations,22
including geometry reoptimization, taking the ZPE from the gas-
phase calculations. The geometry changes and ∆E[(12a) -
(12b)] are small. 12a is more stable than 12b by 1.5 and 2.1
kcal/mol in heptane (ꢀ ) 1.92) and in water (ꢀ ) 78.39),
respectively.
An attempt to locate the transition state for the proton transfer
between 12a and 12b ended in either structure 12a or 12b. When
equal O-H distances were used, we obtained a “transition state”
with one negative frequency of 847 cm-1 corresponding to the
hydrogen migration between the oxygens, whose energy was
that of 12a. Consequently, the barrier to the process is close to
0 kcal/mol.
Compound 11. There are many conformers of the isomers
of 11. For enol 11b we calculated four conformers with syn
CdCsOsH and anti CdCsNsPh moieties which differ in
the relative orientation of the C(dO)O and the CdC groups.
Four conformers of 11a were calculated. The relative energies
(in kcal/mol) (Chart 3) of the most stable structures (Chart 4)
are 11b (0) > 11a (5.6) > iminoenol (11c) (20.2). We failed to
find a minimum corresponding to the ester enol 11d.
Compound 13. The energies of 13a and its enol 13b were
also calculated. The geometries are given in Chart 5. 13a is
favored over 13b by 3.9 kcal/mol (Chart 6).
The effects of â-dicarbomethoxy substitution on the enol and
the ester for both 11 and 13 and the imine in system 11 are
deduced from the isodesmic reactions 3a-7a and 3b-7b.
∆H,
kcal/mol
(MeO2C)2CdC(OH)NHPh + CH4 h
H2CdC(OH)NHPh + H2C(CO2Me)2
(MeO2C)2CHC(NHPh)dO + CH4 h
MeC(NHPh)dO + H2C(CO2Me)2
+37.0
+3.1
+0.6
+24.4
-1.4
+29.1
-4.8
-7.3
+16.5
-9.3
(3a)
(4a)
(5a)
(6a)
(7a)
(3b)
(4b)
(5b)
(6b)
(7b)
(MeO2C)2CHC(OH)dNPh + CH4 h
MeC(OH)dNPh + H2C(CO2Me)2
(MeO2C)2CdC(OH)OMe + CH4 h
H2CdC(OH)OMe + H2C(CO2Me)2
(MeO2C)2CHC(OMe)dO + CH4 h
MeC(OMe)dO + H2C(CO2Me)2
(MeO2C)2CdC(OH)NHPh + H2CdCH2 h
H2CdC(OH)NHPh + H2CdC(CO2Me)2
(MeO2C)2CH(NHPh)dO + H2CdCH2 h
MeC(NHPh)dO + H2CdC(CO2Me)2
(MeO2C)2CHC(OH)dNPh+H2CdCH2 h
MeC(OH)dNPh + H2CdC(CO2Me)2
(MeO2C)2CdC(OH)OMe + H2CdCH2 h
H2CdC(OH)OMe + H2CdC(CO2Me)2
(MeO2C)2CHC(OMe)dO + H2CdCH2 h
MeC(OMe)dO + H2CdC(CO2Me)2
Discussion
The calculated very large energy preference of the parent acid
derivative structure over its enol form7 indicates the need for a
significant driving force to overcome this difference. Both
calculations and NMR studies of â,â-di(bulky)aryl enols of acid
derivatives6,7f indicate that an increased steric bulk is still an
CH(CO2Me)3
(MeO2C)2CdC(OH)OMe
13a
13b
(22) Foresman, J. B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J.; Frisch,
M. J. J. Phys. Chem. 1996, 100, 16098.

1332 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Chart 4. Calculated Structures of 11a-c
Mukhopadhyaya et al.
Chart 6. Relative Energies of 13a and 13b
types, demonstrate potential problems, and give a glimpse of
the substituent effects. They also demonstrate unique features
of our enols, derived from the presence of the push-pull
interaction of the R- and â-substituents.
(a) The Meldrum Acid Systems 7 and 12. The O(1)-H
and O(2)-H distances indicate an enol structure for the solid
MA derivative 7. Planarity of the (OdC)2CdC moiety increases
the resonance stabilization of the enol (cf. 1b). Although
Kenol(CH3CONH2) > Kenol(CH3CO2CH3),7a 7 is not necessarily
more stable than the enol of ester isomer 9, since NH2 is a better
resonative electron-donating group than OMe and a â-CO2R
group delocalizes negative charge better than â-CONHPh.
The calculations enable comparison of the observed structure
of the N-Ph 7 (Figure 3) with that calculated (Chart 1) for the
N-H 12a and with experimentally nonavailable isomers. The
data for 7 and 12a show only small differences: <0.01 Å for
the C1-C2, C1-N, C1-O1, C5-O5, C3-C2, and C5-O4
bonds, and g0.03 Å for the C2-C3, C2-C5, and C3-O3
bonds. Higher deviations (obsd vs calcd in Å) for the O1-
H(1O) (1.125, 1.022), O3-H(1O) (1.385, 1.556), N-H(1N)
(0.942, 1.019), and O(5)-H(1N) (1.939, 1.912) bonds reflect
the higher errors associated with bonds to hydrogen. The angles
differ by <3°. The differences between 7 and 12b are larger,
except for the similar C5-O5, C2-C5, and N-H(1N) bond
lengths in both structures.
Chart 5. Calculated Structures of 13a and 13b
The calculated structures and energies of 12a and 12b do
not differ much. Both have a six-membered enone-containing
ring with one intramolecular H-bond. A proton shift by 0.46 Å
between two oxygens converts 12a to 12b. The differences of
the formal single and double bonds lengths in this ring are
smaller than expected: 0.036 and 0.037 Å for the CdC/CsC
bonds and 0.046 and 0.042 Å for the CdO/CsO bonds. The
energy difference of slightly more than 1 kcal/mol means that
both could coexist. The difference is remarkably smaller than
the 19.9 kcal/mol preference of the syn over the gauche
conformer of 12a. Indeed, a syn arrangement is observed for 7.
The 15 and 16 kcal/mol higher stability, respectively, of both
12a and 12b over the amide form 12c contrasts the ∆G
preferences of the “keto” forms CH3CONH2 and CH3CO2Me
over their enols by 28.9 and 30.3 kcal/mol at B3LYP/6-31G**//
B3LYP/6-31G**.7k Hence, the two cyclic ester groups enor-
mously stabilize the enol relative to the amide form by ca. 44
kcal/mol. The CH3CONH2/CH3C(OH)dNH difference of 14
kcal/mol7k resembles the 12c/12e difference of 16.4 kcal/mol,
indicating a small effect of the â-ester groups on the keto/imine
equilibria.
(b) The “Amide” Derivative 11. When the E conformation
of the ester groups of 7 disappears, a large part of the relative
stabilization of the enol is lost. The solid open-chain â-diester
with a Z conformation of the ester moieties exists as the amide
form 11a. However, the calculation that 11b is 5.6 kcal/mol
more stable than 11a predicts that solid 11 will exist as the
enol. The discrepancy may be due to the fact that not all the
conformers of both 11a and 11b were calculated. For 11b, all
the highly stable syn CdCsOsH conformers were calculated,
and we believe that we did not miss the most stable conformer.
insufficient driving force. However, from the scattered X-ray
data, a sufficient stabilization can be achieved by substitution
with EWGs and via intramolecular H-bonding. The evidence
for short-lived unobservable enols indicate that two strongly
â-EWGs are required but not always sufficient for observing
an enol of an acid derivative. This approach is used in this work.
Enols in the Solid State: Comparison with Calculations.
We will discuss separately the situation in the solid state, which
is not ambiguous, and that in solution. The X-ray structures for
three (EWG)2CHCONHPh systems display different structural

Enols of Carboxylic Acid Amides
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1333
However, if the most stable conformer of 11a was missed, the
calculated Kenol would be higher than the “real” one. This is
unlikely since the use of the experimental data for solid 11a
led to a conformer identical to a previously calculated one.
Packing forces in solid 11a could also be responsible for the
discrepancy.
(c) Tricarbomethoxymethane, 13. Of the two isomers of
13, the amide form 13a is 4.5 kcal/mol more stable than the
enol 13b. In 13b, the resonative system O4O6C3C2C1(dO9)
is nearly planar, with a 28.6° twist of the MeO2C group cis to
OMe from this plane. Hence, even with the extensive conjuga-
tion, reflected by the long CdC bond of 1.406 Å and the short
CsC(dO) bonds of 1.446 Å to the planar CO2Me group and
1.481 Å to the twisted one, two â-ester groups are insufficient
for a significant enolization of a third one.
Chemical Shifts. Due to the dipolar character, δCR should
be at a much lower field and δCâ at a much higher field than
δ13C ) 105-145 in simple ethylenic carbons.25b In 11, δCR )
ca. 168 and δCâ ) ca. 74. This raises assignment problems
since CR of the enol can be in the CO range of esters and amides.
The δ1H(OH) at δ > 15.5 in 7, 11b, and 10c indicates the
presence of a strong H-bond to the enolic hydrogen26 in a six-
membered ring which includes this bond. This resembles the
situation for the stable enols of â-diketones.
Observable Enols in Solution. CDCl3 and DMSO-d6 were
used as low-polarity and dipolar aprotic solvent, respectively.
When low solubility, signals overlap, and inaccessible temper-
atures were encountered, Cl2CDCDCl2 and CD3CN were used
instead. Significant differences in the behavior in the two
solvents were observed.
(a) The MA Derivative. The calculated 12a/12b energy
difference suggests that 7 and 9 may coexist in solution.
Whereas the Me, Ph, and NH signals in CDCl3 or Cl2CDCDCl2
are at the expected δ1H values for either isomer 7-9, the rapidly
exchanging δ 15.70 signal in CDCl3 and the lack of an aliphatic
C-H signal exclude structure 8. The uncoupled Câ, the two
δ13C ester signals, the H-C correlation spectra, and the
similarity of the solution and solid NMR spectra exclude
structure 9.
The large ∆CRâ of 93 ppm is consistent with the push-pull
structure 7. For comparison, we measured the 13C NMR spectra
of the analogues 14a27 and 14b,28 having two resonatively
electron-donating R-substituents. For 14a, δ13C(CDCl3, rt) )
(d) The Dimedone-Substituted Anilide. The diketo-activated
cyclic analogue of 7 enolizes on a ring carbonyl rather than on
the amide group. Since the calculated Kenol values of CH3COR
are g107 larger than that of CH3CONH2,7a the order of
preference should be 10c or 10d . 10b > 10a. Indeed, the
X-ray structure is that of 10c. The enol to enol interconversion
10c h 10a requires a proton shift from O2 to O1.
Enols in Solution. The isomeric composition in solution can
differ from the solid-state structure due to crystal-packing forces
and differential solvation. Several isomers can coexist in solution
and may undergo a rapid exchange. The strongly resonative
electron-donating NHPh and OH groups on C1 and the strong
CO2R or CO EWGs on C2 stabilize the enol and give C1-C2
a partial single bond character (cf. 1b). Structural and spectro-
metric parameters may then resemble those expected for the
amide form.
Structural Parameters. The C1-C2 bond in solid 7 is longer
and the C1-N bond shorter than their values in enamines, which
in frequency vs bond length histograms give maximum values
of 1.35 and 1.39 Å, respectively.19 In the CSDB, in 17 of the
206 enaminoesters RO2CC(R′)dC(R′′)NR′′′RIV, the C1-C2
bonds are longer than 1.400 Å and the group R′ is able to
resonatively delocalize the negative charge on C2 in the dipolar
structure. The model closer to 7 is PhC(NRR′)dC(CHO)CO2-
Et (C1-C2 1.409 Å, C-N 1.356 Å),23 and the extreme case is
RPhNC(R)dC(CO2Me)COCO2Me (C1-C2 1.470 Å, C-N
1.335 Å).24 Hence, the data for 7 are normal, when the additional
resonative electron donation by the OH group on C1 is
considered.
Similar changes are observed in the calculated structures 11b
and 13b, where the CdC and dCsC(dO) bond lengths are
1.44, 1.448, and 1.459 Å and 1.40, 1.446, and 1.480 Å,
respectively. In the amide analogues these bonds have normal
lengths. The observed C1-C2 and C1-N bond lengths are
1.528 and 1.348 Å in 11a, and the calculated C-C bond length
in 13a is 1.5290 Å. In contrast, in enol 10, C1-C2 is 1.470 Å,
C1-O1 is 1.264 Å, and C1-N is 1.328 Å, and these values
differ significantly from those for enol 7.
102.74 (s, Câ), 159.68 (s, COOR), and 192.36 (heptet, J ) 5
Hz, CR), ∆CRâ ) 89.4 ppm. For 14b, δ13C 79.11 (Câ),161.64
(CO2R), and 185.88 (CR), ∆CRâ ) 106.8 ppm. Hence, Câ of
14a is close to Câ in 7, and the high ∆CRâ values fit structure
7 and indicate that δ13CR can be in the range of a carbonyl
group or even lower.
The low intensity and broadening of the room temperature
CO signals and the intensity increase at 223 K may indicate an
exchange with an isomeric species such as 9 (cf. Chart 1). Since
the 13C-H correlation does not show cross-peaks corresponding
to 9, if it is present in the mixture its concentration must be
low.
The spectra of 7 in CD3CN resemble those in CDCl3. Due to
the unobserved OH and CR signals at room temperature in
DMSO-d6, exchange with 9 is more probable than in CDCl3,
but the similarity of the δ13C values of the solid and in DMSO-
d6 support structure 7.
(b) The Dimedone Derivative. The similar spectra of 10 in
solution and in the solid state fit structure 10c. We attribute
one of the low-field signals in the 13C NMR spectra to CR.There
is no evidence for the presence of other enols in solution.29
IR Frequencies. The longer C1-C2 bond length and the
partial double bond character of the â-EWG affect the IR
frequencies. Compared with νmax ) 1715-1730 cm-1 for R,â-
unsaturated CO or CO2Et groups lacking R-electron-donating
1
(c) The Malonic Ester Derivative 11. The H spectrum of
groups,25a
ν
max should appear at lower wavelengths. Indeed, for
11 in CDCl3 is that of the amide form 11a. However, the
additional small signal at δ 17.05 (OH), several MeO signals
at ca. δ 3.81 (Figure 6), and N-H signals (Figure 5) indicate
7 and 11, νmax(CHCl3) ) 1697 cm-1
.
(23) Freeman, J. P.; Duchamp, D. J.; Chidester, C. D.; Slomp, G.;
Szmuszkovicz, J.; Raban, M. J. Am. Chem. Soc. 1982, 104, 1380.
(24) Butler, R. N.; Gillan, A. M.; Lysaght, F. A.; McArdle, P.;
Cunningham, D. J. Chem. Soc., Perkin Trans. 1 1990, 555.
(25) Field, L. D.; Sternhell, C.; Kalman, J. R. Organic Structures from
Spectra, 2nd ed.; Wiley: Chichester, 1995; (a) p 15; (b) p 63.
(26) Perrin, C. L.; Nielson, J. B. Annu. ReV. Phys. Chem. 1997, 48, 511.
(27) Huang, X.; Chen, B. C. Synthesis 1986, 967.
(28) Salim, H., unpublished results.
(29) Mukhopadhyaya, J.; Rappoport, Z., unpublished results.

1334 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Mukhopadhyaya et al.
also the presence of ca. 5% of at least two enolic species,
presumably those of an amide (11b) and of an ester (11d), since
11c has a relatively high energy. Due to E/Z isomerism, 11d
can show four MeO signals, whereas 11b should display two
(R1 ) Ph, R2 ) OH), 9.7, 1.3 (R1R2Cd ) cyclopentadi-
enylidene), and 8.2, 1.0 (R1 ) Ph, R2 ) CN), respectively.30
MA (pKa ) 4.83 (H2O), 7.32 (DMSO))31 is a much stronger
acid than CH2(CO2Me)2 (pKa(DMSO) ) 15.87)31b or dimedone
(pKa(DMSO) ) 11.24).31 MA enolizes <1% in solution or in
the solid.32 We assume that CONHPh substitution does not
change enormously the pKa ratio of the two parent carbon acids,
unless steric hindrance to planarity is important, and that similar
MeO groups which are in different environments. Since the 13
C
NMR spectrum displays only signals of 11a and the IR peaks
at 1751 and 1697 cm-1 are consistent with the ester and amido
carbonyls of 11a, the other isomers are, indeed, present to a
low extent.
factors affect pKa (R1R2CH2) and pKa (R1R2CHCONHPh).
Hence, the ∆pKenol of two acid derivatives 2 will be as large as
K
K
The rapid disappearance of the methine proton of 11a in
CDCl3/D2O at room temperature suggests that a rapid H-D
exchange, presumably via the enol intermediate, takes place.
K
the ∆pKa of R1R2CH2. It is encouraging that the 11.7 kcal/
mol difference in pKa(DMSO) values between MA and CH2(CO2-
31
Me)2 and in the calculated ∆G values corresponding to the
1
The room-temperature H NMR spectrum in CCl4 is that of
Kenol values are identical.
11a (major) and an enol (δ 18.00) + CO2Me signals (minor).
The pKa difference of MA and CH2(CO2Me)2 was ascribed
to a stereoelectronic effect. The high acidity of MA is due to
the ring-enforced high ground-state E conformation of the ester
1
In DMSO, the H NMR spectrum is exclusively that of 11a.
Substituent and Structural Effects. (a) Dissection to the
Contributions of the Enol and the Carbonyl Form. Dissection
to the effects of the â-substituents on the enols and the ketones
is obtained from the calculated ∆H values of isodesmic transfer
reactions of the two â-substituents in 1 and 2 to CH4 (eqs 3a-
7a) or ethylene (eqs 3b-7b). Both series lead to similar
conclusions, and only those with ethylene are discussed. Positive
∆H values mean that the substituted species is more stable than
the parent system. Enol 11b is stabilized by ca. 29 kcal/mol
(eq 3b), and amide 11a and imine 11c are destabilized by ca. 5
and 7 kcal/mol, respectively (eqs 4b and 5b). Hence, the relative
stabilization of 11b arises from a mild amide destabilization
(14% of the effect) and a major (86% of the effect) enol
stabilization. The destabilization of 11a and 11c probably
reflects repulsion between the dipoles of the ester carbonyls
and the amido carbonyl of 11a or the imino nitrogen of 11c.
11a and 11c display similar electronic effects, since NHPh
replaces OH and simultaneously CdO replaces CdNPh.
groups.31 MO calculations on simple model carboxylic groups
33,34
for MA and CH2(CO2Me)2
indicate a higher acidity of the
E conformer, mainly ascribed to the greater MesO/CdO
dipole-dipole interaction than in the Z conformer.33 This
interaction in MeOAc is attractive in the Z conformer and absent
in the E conformer, and vanishes in their anions.34 From similar
considerations on 7 and 11, the large ∆pKenol value is mainly
K
due to a large ∆pKa difference.
(c) The Importance of Hydrogen Bonding. Two features
characterize strong H-bonds:26 (a) a short OO distance of the
OHO system (the H-bond energy rises sharply as the distance
decreases and is appreciable at 2.4 Å) and (b) a high δ(OH)
value. For the observable enols CH2(COR)2, R ) Me, Ph δ-
(OH) > 16. For solid 7, the O1-O2 distance is 2.46 Å and
δ(OH) ) 15.70 in CDCl3. Calculated OO distances in 11b, 13b
and 12a, 12b, and 12d are 2.42, 2.42, and 2.41-2.55 Å, and
δ(OH) of 11b in CDCl3 is 17.05. Although the H-bonds are
not entirely linear, both criteria for a strong H-bond are fulfilled.
Such bonds may stabilize the enols by >10 kcal/mol.26
Stabilization by the longer and weaker NHO bond is probably
compensated by a similar contribution in the amide. Hence, an
appreciable fraction (>0.3) of the increase in Kenol over the
parent system arises from this effect.
Solvent Effect on the Amide h Enol Equilibria. The
solvent effect on the calculated 12a/12b energy difference is
minor: 12a is more stable than 12b by 1.3 (gas phase, ꢀ )
1.0), 1.5 (heptane, ꢀ ) 1.92), and 2.1 (water, ꢀ ) 78.4) kcal/
mol; i.e., Kenol slightly increases on decreasing the polarity. This
is consistent with the observed spectral changes. In the low-
polar CDCl3, Cl2CDCDCl2, and CCl4, the spectra are those of
enol 7. In the more polar DMSO, the spectra cannot exclude a
small amount of an amide species. For 11, both the amide (95%)
and enol(s) are observed in CDCl3 or CCl4, but the enol is not
observed in DMSO-d6; i.e., Kenol seems higher in the chlorinated
solvents. A similar effect was observed for 1,3-dicarbonyl
compounds; e.g., for CH2(COMe)2, the enol comprises 97.6,
94.6, 82.6, 52.9, and 12.9% in the gas phase, in CCl4, in CHCl3,
in MeCN, and in water, respectively,9 and the trend for
acetoactanilide is similar.9 The explanation is that the H-bonded
A smaller enol stabilization (16.5 kcal/mol, 64% of the effect)
superimposed on a larger ester destabilization (ca. 9 kcal/mol,
32% of the effect) is calculated for ester 13. The 9.6 kcal/mol
higher stabilization of the enol of anilide 11a (5.7 kcal/mol)
than that of the ester 13b (-3.9 kcal/mol) agrees with the higher
Kenol values of enols H2CdC(OH)X. The ∆H values of 29.6
(OMe) and 27.6 (NH2) kcal/mol are due to enol stabilizations
of 2.5 (NH2) and 3.6 (OMe) kcal/mol and ketone stabilizations
of 17.0 and 20.1 kcal/mol, respectively.7a Hence, two â-ester
groups affect strongly the contribution of both species. The
MeCONH2 h MeC(OH)dNH equilibrium favors the imine by
12.8 kcal/mol, whereas 11c is disfavored by 20.2 kcal/mol
compared with the amide isomer.
(b) Relative Stabilities and Kenol Values. Enol 7 is highly
stable compared with isomer 8. A lower limit for Kenol is ca.
50, assuming that 2% of 8 could have been observed, but the
value is much higher since from Chart 1 Kenol (gas phase) ≈
1010.8. Even if the value will be reduced somewhat in CDCl3
(vide infra), it will still be the highest obtained so far. Enol 7
is relatively much more stable than enols of 11 and 13 and those
of other anilides 1, X ) NHPh with R1, R2 ) CN, NO2, CO2R.29
The Kenol value for 11 in CDCl3 is much lower: <0.05. The
calculated gas-phase Kenol value is 100 for 11 and 5 × 10-4 for
13.
(30) Andraos, J.; Chiang, Y.; Kresge, A. J.; Popik, V. V. J. Am. Chem.
Soc. 1997, 119, 8417.
(31) (a) Arnett, E. M.; Harrelson, J. A., Jr. J. Am. Chem. Soc. 1987,
109, 809. (b) Arnett, E. M.; Maroldo, S. L.; Schilling, S. L.; Harrelson, J.
A. J. Am. Chem. Soc. 1984, 106, 6759.
(32) (a) Eigen, M.; Ilgenfritz, G.; Kruse, W. Chem. Ber. 1965, 98, 1623.
(b)Arnett, E. M.; Harrelson, J. A. Gazz. Chim. Ital. 1987, 117, 237. (c)
Pluger, C. E.; Boyle, P. D. J. Chem. Soc., Perkin Trans. 2 1985, 1547.
(33) Wang, X.; Houk, K. N. J. Am. Chem. Soc. 1988, 110, 1870.
(34) Wiberg, K. B.; Laidig, K. E. J. Am. Chem. Soc. 1988, 110, 1872.
K
E
E
K
Kenol ) Ka /Ka , where Ka and Ka are the acidities of the
enol and keto forms, respectively. The higher the Câ-substitutent
negative charge delocalizing ability, the higher is KaK. The effect
E
K
of â-substituents on Ka is smaller, as reflected in the pKa ,
pKa values of R1R2CdC(OH)2: 27, 7 (R1 ) R2 ) H), 22, 6.6
E

Enols of Carboxylic Acid Amides
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1335
and stirring was continued for 30 min. The bright orange solution was
poured into 2 N HCl ice-cooled aqueous solution (100 mL). The white
solid precipitate was filtered, washed with cold water, and dried, giving
2 g (74%) of crude 7. Crystallization (EtOAc-petroleum ether, 60-
enol is less polar than the amide form, and hence the latter is
9,35
more solvated and stabilized, thus reducing Kenol
.
Deuterium-Induced 13C Chemical Shifts. The observed
deuterium-induced 13C chemical shifts on the Câ-D/Câ-H of the
amide form or the CR-OD/CR-OH of the enol form could help
to distinguish these structures. However, this tool, which was
applied to 2,2-ditipylethene-1,1-diols6c and for intramolecularly
H-bonded enols and related systems,36 was not an unequivocal
probe in our systems. In the mixture of mono- and dideuterated
species (O-D, N-D, O,N-d2), many deuterium-induced shifts
of mostly ca. 0.1 ppm were observed for both CR and more
remote carbons of 7 and 10c.
1
80 °C) gave pure 7, mp 109-110 °C. H NMR (CDCl3, 240 K): δ
1.78 (6H, s, Me), 7.27 (1H, t, p-Ph-H), 7.35-7.40 (2H, t, Cm), 7.45
(2H, d, Co), 11.14 (1H, br s, NH), 15.63 (<1H, br s, OH). The signal
at δ 15.63 disappeared immediately on shaking the solution with D2O.
1
The compound is stable in MeOH for 4 h at room temperature. H
NMR (CCl4/<10% CDCl3): δ 1.78 (6H, s, Me), 7.20-7.48 (5H, m,
1
Ph), 11.20 (1H, br s, NH), 15.94 (1H, br s, OH). H NMR (rt, CD3-
CN): δ 1.74 (6H, s, Me), 7.27-7.49 (5H, m, Ph), 11.03 (1H, br s,
NH). 1H NMR (220 K, CD3CN): δ 1.68 (6H, s, Me), 7.24-7.40 (5H,
m, Ph), 7.50-7.57 (4 small s), 10.95 (1H, NH), 11.4 (small s), 15.70
(1H, s, OH). 1H NMR (DMSO): δ 1.69 (6H, s, Me), 2.50 (small, m),
7.26-7.49 (5H, m, Ph), 10.99 (1H, br s, NH). 13C NMR (CDCl3): δ
26.02 (Me), 73.50 (Câ), 104.80 (CMe2), 121.89 (d of m, J ) 163 Hz,
Cm), 126.33 (d of t, J ) 163 Hz, Cp), 129.26 (d of d, J ) 163 Hz, Co),
134.67 (Cipso), 164.07 (small, CdO), 168.83 (small, CdO), 170.49 (CR).
13C NMR (CCl4, external C6D6 reference): δ 28.17, 75.16, 106.04,
123.47, 127.63, 130.93, 137.27, 171.03. 13C NMR (DMSO, coupled):
δ 25.59 (q of m), 73.67 (s), 104.61 (complex m), 122.77 (d of m),
126.19 (d of t), 129.04 (d of d), 134.74 (t of m), 166.48 (s), 168.37 (s).
13C NMR (CDCl3, shaken with 1:1 D2O/H2O): δ 26.18, 73.49, 104.92,
122.00/122.10, 126.38, 129.21, 134.61 + 134.69, 164.15, 168.82 +
168.89, 170.44. 13C NMR (rt, DMSO-d6, shaken with 1:1 D2O/H2O):
δ 26.22, 75.0, 104.71, 119.20+119.30 122.79, 123.03, 126.39, 129.56
+ 129.86, 135.36, 139.86, 167.26. The + sign indicates that pairs of
Conclusions. Several conclusions are drawn: (a) Two
strongly EWGs can effectively increase Kenol by the combination
of a stabilizing push-pull interaction of the enol and a
simultaneous destabilization of the “keto” form. (b) In favorable
cases, a solid enol of an amide is obtained. (c) In CDCl3 solution,
enol 7 is the major or exclusive species, and in 11 the enol 11b
is a minor but observable species. In DMSO-d6 or CD3CN, the
enolic hydrogen is not observed at room temperature, the major
species resembles that observed in CDCl3, but the percentage
of the enol is lower and a rapid amide h enol exchange may
occur. (d) Enolization at an RCO site may predominate over
that at a COX carbonyl. (e) CH(EWG)3 systems such as 7 or
11 decompose on raising the temperature. (f) The push-pull
character of the enols is reflected in their spectral parameters.
(g) Calculations show a much higher stability of the syn than
of the gauche conformer of enols. These conclusions indicate
that a systematic study with more systems is worthwhile. We
are involved now with such a study.
signals around it are due to isotope-induced shift. IR (CHCl3, νmax
,
cm-1): 3179 (w, br, OH), 1697 (CdO, s), 1651 (CdO, s). Microanaly-
sis. Calcd for C13H13NO5: C, 59.31; H, 4.98; N, 5.32. Found: C, 59.16;
H, 4.92; N, 4.85. HRMS: calcd for C13H13NO5 263.0794, found
263.0786. m/z (abundance relative to m/z 205, assignment): 263 (44,
M), 205 (100%, M - Me2CO), 159 (13%, M - CNHPh), 133 (35%,
C2(OH)NHPh). MS(EI): m/z 93 (100%, PhNH2+).
Experimental Section
Crystallographic Data for 7. C13H13NO5; space group P21/n; a )
18.058(3) Å, b ) 12.320(2) Å, c ) 5.644(1) Å; â ) 90.24(1)°; V )
1255.6 Å3; Z ) 4; Fcalcd ) 1.39 g cm-3; µ(Mo KR) ) 1.01 cm-1; no.
of unique reflections, 2336; no. of reflections with I g 3σI, 1522; R )
0.055; Rw ) 0.082.
General Methods. Melting points, FT IR spectra, NMR spectra,
and low- and high-resolution mass spectra were recorded as described
previously.6b The solid-state CPMAS 13C NMR measurements at 125.76
MHz were performed on a Bruker DMX-500 digital FT NMR
spectrometer equipped with a BL-4 CPMAS probehead and a high-
resolution/high-performance (HPHP) 1H preamplifier for solids. A solid
glycine standard (CdO at δ 176.03) was used for calibration. A
vacptppm variable-amplitude cross-polarization with a two-pulse phase
modulation broad-band proton decoupling) pulse program was used
for the spectrum. A vacptppmnqs (nqs ) nonquaternary and nonmethyl
signal supression) pulse program was utilized to afford spectral editing,
giving “quaternary-/methyl-only” spectra. Samples were placed in 4-mm
zirconia rotors and spun at a rate of 11.0 MHz.
Solvents and Materials. Meldrum’s acid, dimedone, phenyl isocy-
anate, diethyl benzylmalonate, and diethyl phenylmalonate were
purchased from Aldrich. Commerical deuterated solvents for NMR
spectroscopy (Aldrich) and solvents for chromatography were used
without further purification.
Diethyl Benzylmalonate. 1H NMR (CDCl3): δ 1.18 (6H, 2t, 2Me),
3.20 (2H, d, CH2), 3.64 (1H, m, CH), 4.12 (4H, q, 2CH2), 7.16-7.28
(5H, m, Ph).
Diethyl Phenylmalonate. 1H NMR (CDCl3): δ 1.24 (6H, t, 2Me),
4.20 (4H, 2q, CH2), 4.61 (1H, s, CH), 7.32-7.40 (5H, m, Ph).
Dimethyl triphenylmethylmalonate was available from a previous
work. 1H NMR (CDCl3): δ 3.47 (6H, s, 2Me), 5.38 (1H, s, CH), 7.18-
7.36 (15H, m, Ph).
5-(r-Phenylamino-r′-hydroxy)methylene Meldrum’s Acid (7). To
a solution of Meldrum’s acid (1.5 g, 10.4 mmol) in dry DMF (10 mL)
was added Et3N (2.8 g, 20.8 mmol) and the mixture was stirred for 5
min at room temperature. PhNCO (1.13 mL, 10.4 mmol) was added,
2-Hydroxy-4,4-dimethyl-5-oxo-N-phenylcyclohexenecarbox-
amide (10c). To a stirred solution of dimedone (1 g, 7.14 mmol) in
dry DMF (7 mL) is added Et3N (1.98 g, 14.29 mmol), followed by
PhNCO (0.77 mL, 7.14 mmol). The mixture is stirred for 40 min at
-10 to -12 °C and then poured into 2 N aqueous HCl solution (80
mL), and the precipitated solid is filtered, washed with cold water,
and crystallized (EtOAc-petroleum ether, 60-80 °C) to a white solid
(1.2 g, 48%), mp 84-5 °C. 1H NMR (CDCl3, rt): δ 1.15 (6H, s, 2Me),
2.41 (2H, s, CH2), 2.53 (2H, s, CH2), 7.17 (1H, t, p-Ph-H), 7.35 (2H,
t, m-Ph-H), 7.56 (2H, d, o-Ph-H), 11.77 (1H, br s, NH), 17.82 (1H, br
s, OH). The δ 17.82 signal rapidly exchange in D2O. At 223 K, δ(OH)
) 18.0 (<1H). δ(DMSO-d6): 1.03 (6H, s, Me), 2.45 (2H, s, CH2),
2.51 (2H, s, CH2), 7.17 (1H, t, p-Ph-H), 7.38 (2H, m, m-Ph-H), 7.56
(2H, d, o-Ph-H), 11.76 (1H, br s, NH), 17.46 (1H, br s, OH). 13C NMR
(CDCl3, 223 K): δ 27.74 (q, Me), 30.65 (s, CMe2), 44.65 (t, CH2),
50.44 (t, CH2), 101.72 (s, Câ), 120.72 (d of m, Cm), 124.62 (d of t, Cp),
3
128.73 (d of d, Co), 136.00 (t, J ) 9 Hz, Cipso), 168.83 (CR), 194.17
(q, J ) 8 Hz, CdO), 197.41 (t, 3J ) 6 Hz, CdO). The room-
temperature spectrum is similar. 13C NMR (DMSO-d6): δ 27.13, 30.48,
44.22, 50.30, 101.70, 120.86, 124.82, 128.97, 136.21, 168.90, 194.04,
197.34. IR (νmax, cm-1, Nujol): (br m, OH), 3186 (vw, br), 1720 (s,
CdO), 1654 (CdO, w). IR (νmax, cm-1, CHCl3): 3318, 3281 (m, OH),
1712 (w), 1654 (CdO, s). Microanalysis. Calcd for C15H17NO3: C,
69.48; H, 6.61; N, 5.40. Found: C, 69.27; H, 6.59; N, 5.49.
Crystallographic Data for 10c. C15H17NO3; space group P21/c; a
) 12.061(2) Å, b ) 12.752(2) Å, c ) 18.370(4) Å; â ) 104.45°; V )
2736(1) Å3; Z ) 4; Fcalcd ) 1.26 g cm-3; µ(Mo KR) ) 0.82 cm-1; no.
of unique reflections, 5065; no. of reflections with I g 3σI, 2911; R )
0.055; Rw ) 0.074.
(35) Floris, B. In ref 1, Chapter 4, p 147.
(36) (a) Hansen, P. E.; Bolvig, S.; Duus, F.; Petrova, M. V.; Kawecki,
R.; Kozerski, L. Magn. Reson. Chem. 1995, 33, 621. (b) Kozerski, L.;
Kawercki, R.; Krajewski, P.; Kwiecien, B.; Boykin, D. W.; Bolvig, S.;
Hansen, P. E. Magn. Reson. Chem. 1988, 36, 921.
Reaction of Dimethyl Malonate with Phenyl Isocyanate. A
solution of dimethyl malonate (1 g, 7.6 mmol), PhNCO (0.82 mL, 7.6

1336 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Mukhopadhyaya et al.
mmol), and Et3N (2.1 mL, 15.14 mmol) in dry DMF (7 mL) was heated
for 2.5 h at 55 °C and then stirred at room temperature for 14 h. The
mixture was poured into a cooled 2 N aqueous HCl solution (90 mL)
and extracted with EtOAc (125 mL), which was washed with water (3
× 40 mL), dried (Na2SO4), and evaporated. Chromatography of the
gummy residue (1.95 g) over silica gel with 25% EtOAc/petroleum
ether eluent gave a solid (1.03 g, 54%) which was recrystallized from
EtOAc/petroleum ether. The first precipitated crystals (20 mg), mp
238-240 °C, were identified by X-ray crystallography as N,N′-
diphenylurea, mp 240 °C. The main fraction, mp 98-100 °C, is 11.
1H NMR (CDCl3): δ 3.81 (0.33H, s, CO2Me of 11b), 3.86 (6H, s,
CO2Me), 4.49 (1H, s, CH), 7.14 (1H, t, p-Ph-H), 7.37 (2H, t, m-Ph-H),
7.56 (2H, d, o-Ph-H), 9.26 (ca 1H, br s, NH), 17.05 (0.05H, s, OH of
11b). The signals at 4.49, 9.25, and 17.05 ppm disappeared rapidly
appear at 163.31, 169.60, 169.95, and 174.73 with ca. 10% of the
intensity of the neighboring main signals. 13C NMR (uncoupled, DMSO-
d6, rt): δ 52.78 (t, Me), 60.46 (d, CH), 118.08 (small, overlapping),
119.02 (d of m, Cm), 123.92 (overlapping, d of t, Cp), 128.71 (d of d,
Co), 128.84 (d of d, small), 138.26 (t, J ) 9.4 Hz, Cipso), 160.94 (d of
d, J ) 9; 2.7 Hz, CONHPh), 165.16 (sextet, J ) 4 Hz, COOMe). IR
(νmax, cm-1, CHCl3): 3332 (br vw, OH) 1751 (m, CdO), 1697 (m,
CdO), 1613. IR (Nujol) 3326 (br, m), 1747 (s), 1691 (s), 1607 (s).
Microanalysis. Calcd for C12H13NO5: C, 57.37; H, 5.22; N, 5.58.
Found: C, 57.67; H, 5.21; N, 5.36.
Crystallographic Data. C12H13NO5; space group P21/c; a ) 9.908(4)
Å, b ) 44.385(8) Å, c ) 9.248(4) Å; â ) 107.97(3)°; V ) 3868(2)
Å3; Z ) 12, Fcalcd ) 1.29 g cm-3, µ (Cu KR) ) 8.65 cm-1, No. of
unique reflections 5716, No. of reflections with I g 3σI 2911. R )
0.058, Rw ) 0.086.
1
with D2O. H NMR (CCl4, external C6D6): δ 4.64 (s, Me), 4.74 (s,
Me), 5.21 (s, CH), 7.97 (m), 8.17 (t), 8.33 (d), 8.45 (d) [all Ph-H],
10.02 (s, NH), 12.93 (br s, NH), 18.00 (s, OH). 1H NMR (DMSO-d6):
δ 3.71 (6H, s, CO2Me), 4.77 (1H, s, CH), 7.08 (1H, t, p-Ph-H), 7.32
Acknowledgment. We are indebted to Mr. Niv Papo for
preliminary experiments, to Prof. M. Kaftory for his help with
the CSDB, to Dr. Shmuel Cohen for the crystallographic
determination, to Prof. Robert Glaser for the solid 13C NMR
spectra, and to the Israel Science Foundation for support.
1
(2H, t, m-Ph-H), 7.52(2H, d, o-Ph-H), 10.33 (ca. 1H, br s, NH). H
NMR (CD3CN, rt): δ 3.83 (3H, s, Me), 4.58 (0.45H, br, CH), 7.16-
7.55 (5H, m, Ph), 7.28 (0.3H, br s, Ph-H), 8.30 (0.16H, br s, Ph-H),
8.85 (0.4H, br s, Ph-H). 13C NMR (uncoupled CDCl3): δ 53.60 (Me,
q), 59.18 (CH, d), 120.15 (Cp, d of m), 124.93 (Cm, d of t), 128.98 (Co,
d of d), 137.05 (Cipso), 159.64 (CR, d of d), 166.08 (CO, hept, J ) 4.2
Hz).
At 220 K in the coupled spectrum in CDCl3, new signals (δ 41.17,
t) and (δ 53.83, q) (7% of the intensities) accompany the corresponding
major signals. The aromatic signals at 119.69, 124.81, 128.86, and
136.57 ppm were accompanied by a new overlapping multiplet at ca.
119.86, 124.52, 128.90 and 136.84 ppm, with 10% of their intensity.
The CH signal at δ 58.83 displays no satellite signal, but new signals
Supporting Information Available: Tables of bond lengths
and angles, position and thermal parameters, stereoviews of 7,
10c, and 11a, solid-state structures and 1H and 13C NMR plots
of 11a, B3LYP/6-31G** energies of various species, and
dihedral angles of 12a-e (PDF). This material is available free
of charge via the Internet at http://pubs.acs.org.
JA992059F