Probing persistent intramolecular C-H· · ·X(X = O, S, Br, Cl, and F) bonding in solution using benzyl meldrum's acid derivatives

Article abstract of DOI:10.1021/jo802311w

Persistent intramolecular interactions between acidic C-H hydrogens and a variety of acceptors (X) (X = O, S, Br, Cl, and F) in solution were probed by ¹H NMR experiments, using 5-benzyl Meldrum's acid derivatives. To bring about formation of i

Full text of DOI:10.1021/jo802311w

Probing Persistent Intramolecular C-H · · ·X (X ) O, S, Br, Cl, and
F) Bonding in Solution Using Benzyl Meldrum’s Acid Derivatives
Eric Fillion,* Ashraf Wilsily, and Dan Fishlock1
Department of Chemistry, UniVersity of Waterloo, Waterloo, Ontario N2L 3G1, Canada
efillion@uwaterloo.ca
ReceiVed October 18, 2008
Persistent intramolecular interactions between acidic C-H hydrogens and a variety of acceptors (X) (X
) O, S, Br, Cl, and F) in solution were probed by 1H NMR experiments, using 5-benzyl Meldrum’s acid
derivatives. To bring about formation of intramolecular C-H · · · X bonding, ortho-substituted benzyl
Meldrum’s acids were designed, for which hydrogen bonding occurred through a six-membered ring.
Introduction of substituents on the aromatic moiety and in the tether allowed variation of electronic and
steric factors. The superior acidity of Meldrum’s acids impacted the ability of the C-H hydrogen to
engage in nonclassical C-H · · · X bonds and drove the conformational properties of benzyl Meldrum’s
acid, in combination with steric factors, namely A1,3-allylic strain. Further understanding of intramolecular
C-H · · · X bonds was gained by characterization of benzyl Meldrum’s acids in the solid state, by X-ray
analysis, and by the conformation correlated to the observations made in solution.
Introduction
Nonclassical C-H· · ·X (where X ) O and N) hydrogen
bonds are prominent and well documented in the solid state.2
However, limited experimental evidence supports persistent
C-H· · ·X hydrogen bonds in solution, which have been
postulated to be of particular importance in a wide array of
biological phenomena, as well as in supramolecular chemistry.
Infrared and proton magnetic resonance spectroscopy have been
utilized to survey the propensities of haloforms and related
compounds to act as proton donors in intermolecular hydrogen
bond with DMSO, pyridine, acetone, and Et3N.3,4 Other reports,
using 1H NMR, concern intramolecular o-carboranyl C-H· · ·N
and C-H· · ·O interactions.5 Further evidence of C-H· · ·X
bonding in this system has been extended to acceptors such as
Br, Cl, and F.6 The common feature of the haloforms and the
o-carboranes is the relatively high acidity of the hydrogen
involved. Consequently, the formation of persistent C-H· · ·X
FIGURE 1. Meldrum’s acid (1) and 5-benzyl derivatives.
bonds in solution should be promoted by enhancing the acidity
of the C-H donor and thus its ability to hydrogen bond.7,8
Meldrum’s acid (1) and derivatives are valuable reagents in
organic synthesis (Figure 1), serving as powerful acylating
agents in C-C,9 C-O, and C-N bond forming reactions10 and
as building blocks in the synthesis of numerous natural
(5) (a) Cappelli, A.; Giorgi, G.; Anzini, M.; Vomero, S.; Ristori, S.; Rossi,
C.; Donati, A. Chem. Eur. J. 2004, 10, 3177–3183. (b) Alekseyeva, E. S.;
Batsanov, A. S.; Boyd, L. A.; Fox, M. A.; Hibbert, T. G.; Howard, J. A. K.;
MacBride, J. A. H.; Mackinnon, A.; Wade, K. J. Chem. Soc., Dalton Trans.
2003, 475–482. (c) Donati, A.; Ristori, S.; Bonechi, C.; Panza, L.; Martini, G.;
Rossi, C. J. Am. Chem. Soc. 2002, 124, 8778–8779. (d) Gibb, C. L. D.; Stevens,
E. D.; Gibb, B. C. J. Am. Chem. Soc. 2001, 123, 5849–5850. (e) Mizuno, K.;
Ochi, T.; Shindo, Y. J. Chem. Phys. 1998, 109, 9502–9507. (f) Afonin, A. V.
Russ. Chem. Bull. 1996, 45, 1145–1147.
(6) Ohta, K.; Yamazaki, H.; Endo, Y. Tetrahedron Lett. 2006, 47, 1937–
1940.
(7) A correlation between the mean C-H· · ·O distances in crystal structures
and the pKa values of carbon acids was reported, see: Pedireddi, V. R.; Desiraju,
G. R. J. Chem. Soc., Chem. Commun. 1992, 988–990.
(1) Current address: Roche Palo Alto LLC, Technical Sciences-Chemical
Synthesis, 3431 Hillview Ave, Palo Alto, CA, 94304.
(2) For reviews on the C-H · · · O bond in the solid state, see: (a) Castellano,
R. K. Curr. Org. Chem. 2004, 8, 845–865. (b) Steiner, T. Angew. Chem., Int.
Ed. 2002, 41, 48–76. (c) Desiraju, G. R; Steiner, T. The Weak Hydrogen Bond
in Structural Chemistry and Biology; Oxford University Press: Oxford, UK, 1999.
(3) Huggins, C. M.; Pimentel, G. C. J. Chem. Phys. 1955, 23, 1244–1247.
(4) Allerhand, A.; Schleyer, P. v. R. J. Am. Chem. Soc. 1963, 85, 1715–
1723.
(8) In the solid state, the acidity of the X-H bond affects the lengths of the
X-H · · · O (X ) C, N, O) hydrogen bonds rather than the oxygen atom basicity,
see: Taylor, R.; Kennard, O. J. Am. Chem. Soc. 1982, 104, 5063–5070.
10.1021/jo802311w CCC: $40.75
Published on Web 12/29/2008
2009 American Chemical Society
J. Org. Chem. 2009, 74, 1259–1267 1259

Fillion et al.
TABLE 1. 1H NMR Chemical Shifts of Benzyl Meldrum’s Acid and Benzyl Malonate Derivatives
H(5) chemical shifta b c
∆ ppme
,
,
entry
compd
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
9
10
11
12
13
14d
15d
16d
17d
18d
19d
4.13 (4.63) (t)
4.29 (4.45) (d)
5.38 (5.18) (s)
3.74 (4.77) (t)
3.67 (4.72) (d)
3.58 (4.56) (s)
3.47 (s)
3.59 (s)
3.50 (s)
5.29 (s)
5.05 (s)
5.21 (5.04) (s)
3.75 (t)
3.83 (d)
4.60 (s)
3.66 (t)
0.39 (-0.14) (2-5)
0.62 (-0.27) (3-6)
1.80 (0.62) (4-7)
-0.11 (8-7)
0.01 (9-7)
-0.08 (10-7)
1.71 (11-7)
1.47 (12-7)
1.63 (0.48) (13-7)
0.09 (14-17)
0.21 (15-18)
0.78 (16-19)
9
10
11
12
13
14
15
16
17
18
3.62 (d)
3.82 (s)
a
c
1Η ΝΜR chemical shifts in CDCl3. b H(5) multiplicity is in parentheses. 1Η ΝΜR chemical shifts in DMSO-d6 are in parentheses. d H(2) chemical
shift. e ∆ ppm in DMSO-d6 and the compared compounds are in parentheses.
products.11 Considerable attention has also been paid to the
exceptionally high C-H acidity of 1.12 The hydrogens located
R to the two carbonyl groups (5-position), with a remarkable
pKa of 4.83-4.9313 (7.3 in DMSO),14 confer unique chemical
properties to Meldrum’s acid (1), one of them being the ability
to participate in hydrogen bonding. X-ray structural determi-
nation of this atypical carbon-based Brønsted acid revealed the
presence of a nonclassical intermolecular C-H· · ·O bond (2.43
Å) between the oxygen atom of the carbonyl group and the
pseudoaxial hydrogen H(5) at the 5-position.15,16 Toward an
understanding of the interactions between acidic C-H hydro-
gens and a variety of acceptors (X), we set out to probe
persistent hydrogen bonding in solution using Meldrum’s acid
derivatives. To bring about formation of an unprecedented
intramolecular hydrogen bond, ortho-substituted benzyl Mel-
drum’s acids were designed for which hydrogen bonding would
occur through a six-membered ring. As illustrated in Figure 1,
introduction of substituents (R) on the aromatic moiety and in
the tether (Y) would allow variation of electronic and steric
factors.
Herein, we document benzyl Meldrum’s acid derivatives as
a class of compounds newly identified as displaying persistent
intramolecular C-H· · ·X bonds (X ) O, S, Br, Cl, and F)
(9) (a) Fillion, E.; Dumas, A. M. J. Org. Chem. 2008, 73, 2920–2923. (b)
Fillion, E.; Dumas, A. M.; Hogg, S. A. J. Org. Chem. 2006, 71, 9899–9902. (c)
Fillion, E.; Wilsily, A. J. Am. Chem. Soc. 2006, 128, 2774–2775. (d) Fillion, E.;
Dumas, A. M.; Kuropatwa, B. A.; Malhotra, N. R.; Sitler, T. C. J. Org. Chem.
2006, 71, 409–412. (e) Fillion, E.; Fishlock, D. J. Am. Chem. Soc. 2005, 127,
13144–13145. (f) Fillion, E.; Fishlock, D.; Wilsily, A.; Goll, J. M. J. Org. Chem.
2005, 70, 1316–1327. (g) Fillion, E.; Fishlock, D. Org. Lett. 2003, 5, 4653–
4656.
1
(Figure 1) in solution, as evidenced by H NMR experiments
and solid-state X-ray analysis.
(10) (a) Wilsily, A.; Fillion, E. Org. Lett. 2008, 10, 2801–2804. (b) Fillion,
Results and Discussion
´
E.; Carret, S.; Mercier, L. G.; Tre´panier, V. E. Org. Lett. 2008, 10, 437–440.
For reviews on Meldrum’s acid in synthesis, see: (c) Chen, B.-C. Heterocycles
1. Characterization of Persistent Intramolecular C-H· · ·X
(X ) O, S, Br, Cl, and F) Bonds in Solution with H NMR
´
1991, 32, 529–597. (d) Strozhev, M. F.; Lielbriedis, I. E.; Neiland, O. Ya. Khim.
1
Geterotsikl. Soedin. 1991, 57, 9–599. (e) McNab, H. Chem. Soc. ReV. 1978, 7,
Spectroscopy. (a) Identification of Anomalous Chemical
Shift Differences in Benzyl Meldrum’s Acid Derivatives. In
the course of our studies using Meldrum’s acids as acylating
agents in the catalytic intramolecular Friedel-Crafts reaction,
5-(2,3-dimethoxybenzyl) Meldrum’s acids 2-4 were prepared,
in which the degree of substitution at the benzylic position was
varied (Table 1).9f It was noted that the 1H NMR chemical shift
difference for H(5) spanned 1.25 ppm; as substitution increased
at the benzylic position, so did deshielding of H(5) (entries 1-3).
To determine if the origin of this unexpected chemical shift
variation was attributable to the gem-dimethyl group, unsub-
stituted counterparts 5-7 were synthesized. The latter showed
345–358.
(11) Ivanov, A. S. Chem. Soc. ReV. 2008, 37, 789–811.
(12) (a) Nakamura, S.; Hirao, H.; Ohwada, T. J. Org. Chem. 2004, 69, 4309–
4316. (b) Byun, K.; Mo, Y.; Gao, J. J. Am. Chem. Soc. 2001, 123, 3974–3979.
(c) Perez, G. V.; Perez, A. L. J. Chem. Ed. 2000, 77, 910–915.
(13) (a) Pihlaja, K.; Seilo, M. Acta Chem. Scand. 1969, 23, 3003–3010. (b)
Pihlaja, K.; Seilo, M. Acta Chem. Scand. 1968, 22, 3053–3062.
(14) Arnett, E. M.; Maroldo, S. G.; Schilling, S. L.; Harrelson, J. A. J. Am.
Chem. Soc. 1984, 106, 6759–6767.
(15) Pfluger, C. E.; Boyle, P. D. J. Chem. Soc., Perkin Trans. II 1985, 1547–
1549.
(16) A survey of the Cambridge Crystallographic Database revealed a very
limited number of Meldrum’s acids monosubstituted at the 5-position. From these,
a single example of intramolecular C-H · · · O bond, which occurs via a seven-
membered ring, was found, see: Snyder, C. A.; Selegue, J. P.; Dosunmu, E.;
Tice, N. C.; Parkin, S. J. Org. Chem. 2003, 68, 7455–7459.
1260 J. Org. Chem. Vol. 74, No. 3, 2009

Persistent Intramolecular C-H· · ·X Bonding in Solution
TABLE 2. Concentration Data for Meldrum’s Acid 13 in CDCl3
TABLE 3. Temperature Effect on Meldrum’s Acids 2-4 in CDCl3
1H NMR chemical
shift of H(5) (ppm)
H(5) chemical shift (ppm)
Meldrum’s
acid
∆ ppm
[-50 °C - 55 °C]
entry
concn (M)
entry
at -50 °C
at 55 °C
1
2
3
4
5
0.044
0.16
0.35
0.70
1.38
5.22
5.22
5.23
5.24
5.26
1
2
3
2
3
4
4.31
4.49
5.59
4.07
4.18
5.25
0.24
0.31
0.34
TABLE 4. Van der Waals Contacts Contribution of Ortho
Substituent to ∆ ppm in CDCl3 and DMSO-d6
an insignificant variation of 0.16 ppm (entries 4-6). A direct
comparison of compounds 2-4 with 5-7 (∆ ppm in Table 1)
revealed that the largest 1H NMR chemical shift difference, 1.80
ppm, was observed for gem-dimethyl-substituted Meldrum’s
acid 4 versus 7. Obviously, the presence of the gem-dimethyl
group had a large influence on the chemical shift of H(5) in
2-4. However, since increasing substitution at the benzylic
position of Meldrum’s acids 5-7 had a slight shielding effect
on H(5), an additional factor had to be involved. It was
hypothesized that the methoxy group present at the ortho
position of the aromatic moiety was key in deshielding H(5).
Indeed, 1H NMR chemical shifts of 3,4-dimethoxybenzyl, 3,5-
dimethoxybenzyl, and 4-methoxybenzyl Meldrum‘s acids 8, 9,
and 10 revealed no significant difference when related to 7
(entries 7-9). The influence of the o-methoxy group on the
chemical shift of H(5) was further confirmed with Meldrum’s
acids 11-13. The introduction (entries 10 and 11) or absence
(entry 12) of an additional electron-donating methoxy group
substituent at various positions of the aromatic had little
influence on the chemical shift of H(5), which remained
unusually deshielded for a CsH bond of this type, with a ∆
ppm between 1.47 and 1.71 when compared with 7.
These observations were unique to 5-benzyl Meldrum’s acids.
As shown in Table 1 (entries 13-18), the chemical shift
variations of H(2) for malonate analogues 14-16 and their
corresponding derivatives 17-19 lacking the methoxy substit-
uents revealed a similar dependence of benzylic substitution.
However, with a maximum ∆ ppm of 0.78 for 16, this effect
was much lower than that observed with 5-benzyl Meldrum’s
acids. Therefore, the greater C-H acidity of Meldrum’s acids
in comparison to malonates likely plays a role in deshielding
H(5) when the ortho position of the aromatic moiety is
substituted with a methoxy group.
From these results, the polarization of the C-H bond and
the ensuing deshielding of H(5) in Meldrum’s acids bearing a
methoxy group at the ortho position of the aromatic moiety
suggested a persistent intramolecular C-H· · ·O bond in solu-
tion. The ability of Meldrum’s acid to engage in nonclassical
C-H· · ·O hydrogen bonds as a result of its superior acidity, in
combination with steric factors, deserved thorough investigation.
(b) Distinguishing Between van der Waals Interactions
and Hydrogen Bonding as Causes of H(5) Chemical Shift
Changes. From these premises, we set out to probe persistent
intramolecular hydrogen bonding in solution by studying the
insensitivity of proton chemical shift to concentration and
temperature changes on benzyl Meldrum’s acid derivatives 2,
3, 4, and 13. As shown in Table 2, varying the concentration
of 13 from 0.044 to 1.38 M had no impact on the 1H chemical
shift of H(5), which remained deshielded.
H(5) chemical shift (ppm)
∆ ppm
entry
R
R′
in CDCl3
in DMSO in CDCl3 in DMSO
1
2
3
H
H (20)
3.69
3.63
4.13
4.66
4.47
4.45
-0.05a
-0.04b
0.55c
-0.11
-0.25
-0.11
Me H (21)
Me Me (22)
a Compared with 5. b Compared with 6. c Compared with 7.
stantial deshielding of H(5) compared to 4, with ∆ ppm values
of 0.39 and 0.62, respectively, versus 1.80 ppm (Table 1, entries
1 and 2 versus 3). It was then determined if the H(5) chemical
shift variations observed for acids 2 and 3 in CDCl3 were the
result of van der Waals contacts rather than persistent intramo-
lecular hydrogen bonding. The influence on the chemical shift
of H(5) by a non-hydrogen bond acceptor group located at the
ortho position of the aromatic was estimated by preparing
compounds 20-22 (Table 4). While slightly larger, the ethyl
group was rationally selected to mimic the hydrogen bond
acceptor methoxy group.
A substantial effect was only observed for 22, which bears a
benzylic gem-dimethyl group (Table 4, entry 3). In CDCl3, the
ethyl group deshielded H(5) in 22 when compared with 7 by
0.55 ppm. Once again, the gem-dimethyl compound displayed
a unique behavior, which likely indicates that the presence of
the benzylic substituents forces the group at the ortho position
of the aromatic to be in close proximity with H(5). Indeed, in
CDCl3, a nOe between the methylene of the ethyl groups and
H(5) was observed.
The fact that H(5) in Meldrum’s acids 20 and 21 relative to
5 and 6 is slightly shielded, while 2 and 3 relative to 5 and 6 is
deshielded, suggests that van der Waals interactions do not
account for the increased chemical shift of H(5) in 2 and 3.
Considering that the only difference between 2 and 20, and 3
and 21, respectively, is the presence of an oxygen atom versus
a methylene group, it is logical to conclude that this deviation
in chemical shift is caused by hydrogen bonding. In the same
vein, while H(5) of 22 is deshielded by 0.55 ppm relative to 7,
the much larger difference between H(5) of 4 compared to 7
(1.80 ppm) must be caused by a significant intramolecular
C-H· · ·O bond in solution.
In solution, it was then concluded that the chemical shift
variations of H(5) in 2-4, in comparison to their unsubstituted
analogues 5-7, were due to hydrogen bonding between the
o-methoxy group and H(5), in accordance with the temperature
studies. These data undoubtedly established that Meldrum’s
acids 2-4 form a persistent C-H· · ·O bond in solution, the
strength of this interaction being enhanced by increasing
substitution at the benzylic position.
For acids 2-4, the effect of temperature on H(5) chemical
shift was minimal as judged by the slight variation over a range
of 105 °C (Table 3). While the above results suggested the
formation of persistent intramolecular C-H· · ·O bonds in
Meldrum’s acids 2-4, compounds 2 and 3 showed nonsub-
Van der Waals interactions as the sole cause of the H(5)
chemical shift changes were further ruled out by studying the
J. Org. Chem. Vol. 74, No. 3, 2009 1261

Fillion et al.
TABLE 5. 1H NMR Spectroscopic Data for 5-(2-Substituted
Benzyl) Meldrum’s Acids in CDCl3 and DMSO-d6
TABLE 6. Spectroscopic Data for 28 and 29 in CDCl3
entry
n
H(5) chemical shift (ppm)
∆ ppma
,
entry
X
H(5) chemical shifta (ppm)
∆ ppmb c
1
2
1 (28)
2 (29)
4.69
5.06
1.11
1.48
1
2
3
4
5
6
7
8
H (7)
OMe (13)
Et (22)
Me (23)
F (24)
Cl (25)
3.58 (4.56)
5.21 (5.04)
4.13 (4.45)
4.18
1.63 (0.48)
0.55 (-0.11)
0.60
0.98 (0.01)
1.92
a The compounds were compared with 7.
4.56 (4.57)b
5.50
Conformationally constrained systems 28 and 29 displayed a
smaller 1H NMR chemical shift difference, compared with
dimethoxy substrate 4, likely a result of a nonoptimal orientation
of the oxygen lone pairs in the C-H· · ·O interaction (Table
6). This further rules out van der Waals contacts as the cause
of the increased chemical shift of H(5), as these should be
relatively unaffected by such a conformational change.
Br (26)
SMe (27)
5.73
2.15
6.50 (6.27)b
2.92 (1.71)
1Η ΝΜR chemical shifts in DMSO-d6 are in parentheses. b All the
a
c
compounds were compared with 7.
parentheses.
∆ ppm in DMSO-d6 are in
(c) Contribution of Geminal Benzylic Substituents to
the Presence of C-H· · ·X (X ) O, S, F) Bonding. The effect
of the nature of the geminal benzylic substituents on hydrogen
bond formation was then examined (Table 7). Sterically
congested diethyl ester and diallyl derivatives 30 and 31
displayed deshielded H(5) when compared with analogous
compounds 39 and 40, respectively (entries 1 and 2 versus
entries 10 and 11). Similar observations were made for
cyclopentene, cyclopentane, and cyclohexane derivatives 32, 33,
and 34 (entries 3-5), and dimethoxy cyclohexane substrate 45
(entry 16). In all these models, the chemical shift differences
(∆ ppm) in CDCl3 were not as sizable as that observed for the
dimethyl counterpart 13, indicating that the greater steric
hindrance afforded by the gem-diester, gem-diallyl groups, and
geminal cyclic substituents had a direct impact on the ability
of the ortho substituent to interact with H(5) in solution.
Nonetheless, the ∆ ppm values in CDCl3 were indicative of
persistent intramolecular C-H· · ·O bonding in solution for
30-34 and 45, but to a less extent than in gem-dimethyl systems
4 and 13. This hypothesis was corroborated by studying the
effect of an ethyl group located at the ortho position of the
aromatic on the chemical shift of H(5), in the cyclohexyl series,
with Meldrum’s acid 35 (entry 6). The ∆ ppm of 0.20 in CDCl3
for 35 was almost 3-fold lower than the one observed for 22
(0.55), despite detecting a NOE between the methylene of the
ethyl groups and H(5), and was inferior to the ∆ ppm values
for all Meldrum’s acids 30-34 and 45. Therefore, van der Waals
contacts arguments did not account for the observed chemical
shift variations for 30-34 and 45 in CDCl3; it was concluded
that intramolecular hydrogen bonding was key in the deshielding
of H(5). The smaller ∆ ppm values for this series of compounds
seem to indicate that adverse steric effects were introduced,
which counterbalanced the formation of a persistent C-H· · ·O
bond in solution. As a result, for compounds 30-34 and 45, an
equilibrium of intramolecular hydrogen bonded versus non-
bonded benzyl Meldrum’s acid species was formed to a variable
degree, similar to 2 and 3, thus furnishing less deshielded H(5)
chemical shifts. In contrast, compounds 4 and 13 which have
larger ∆ ppm values in CDCl3 suggest that a greater percentage
of the molecules were present in the hydrogen bonded form.21,22
effect of other potential H-bond acceptors. Thus, electronic
factors controlling the formation of persistent C-H· · ·X (X )
S, Br, Cl, and F) through a six-membered ring were then
explored in detail. As the chemical shift difference was most
pronounced between 4 and 7, this effect was studied using gem-
dimethylbenzyl-substituted Meldrum’s acid derivatives 22-27.
As depicted in Table 5, the chemical shift values of H(5) were
determined and compared with reference compound 5-benzyl
Meldrum’s acid 7. The ethyl-substituted 5-benzyl Meldrum’s
acid 22 discussed previously, and the methyl derivative 23,
which lack the ability to hydrogen bond but are also more
sterically demanding than 7, slightly affected the chemical shift
of H(5) (entries 3 and 4).17 Halogens, also considered strong
hydrogen bond acceptors, interacted effectively with H(5)
(entries 5-7), bromide and chloride leading to substantial
deshielding of H(5). Thiomethoxy-substituted Meldrum’s acid
27 induced the strongest effects on H(5) chemical shift (entry
8).18
5-(2-Alkylbenzyl) Meldrum’s acids 22 and 23 showed a ∆
ppm between 0.55 and 0.60 compared to control model 7. On
the other hand, Meldrum’s acids 24-27, which contain potential
hydrogen bond acceptors, show greater chemical shift differ-
ences, ranging from 0.98 to 2.92 ppm. Were this difference
caused solely by van der Waals interactions, the chemical shift
should increase with the van der Waals radius of the involved
atom. However, comparing 22 and 23 to 24, in which a smaller
fluoride atom (1.47 Å)19 causes a greater chemical shift
difference than a larger carbon atom (1.70 Å), suggests that
hydrogen bonding accounts for the deshielding of H(5). The
same is true of 26 versus 27, in which smaller sulfur (1.80 Å)
deshields more strongly than larger Br (1.85 Å). The fact that
∆ ppm correlates to polarizability of the involved atom rather
than van der Waals radius is further proof of intramolecular
hydrogen bonding in 5-(2-substituted benzyl) Meldrum’s acids.
Aside from the presence of a hydrogen bond accepting atom
as a requirement for their formation, hydrogen bonds have been
shown to depend on the directionality of the involved lone pair.20
(17) Similarly to the observations made with 22, strong nOe between the
methyl group located on the aromatic moiety and H(5) was measured for 23.
(18) For intramolecular C-H · · · X (X ) O, S) interactions in N-acyl-4-
isopropyl-1,3-thiazolidine-2-thiones and related 1,3-oxazolidin-2-ones, see: Cosp,
A.; Larrosa, I.; Anglada, J. M.; Bofill, J. M.; Romea, P.; Urp´ı, F. Org. Lett.
2003, 5, 2809–2812.
The effect of other potential H-bond acceptors was investi-
gated. A negligible ∆ ppm of 0.17, comparable to that of ethyl
derivative 35 (∆ ppm ) 0.20), was observed for fluoro
Meldrum’s acid derivative 36 and suggested the absence of a
(19) Bondi, A. J. Phys. Chem. 1964, 68, 441–451.
(20) Steiner, T.; Saenger, W. J. Am. Chem. Soc. 1992, 114, 10146–10154.
1262 J. Org. Chem. Vol. 74, No. 3, 2009

Persistent Intramolecular C-H· · ·X Bonding in Solution
1
TABLE 7. Effect of Benzylic Substituents on the H NMR Chemical Shifts of Benzyl Meldrum’s Acids
H(5) chemical shift (ppm)a b
∆ ppmc
,
entry
Meldrum’s acid
1
2
3
4
5
6
7
8
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
5.31
4.76 (4.93)
4.16
4.19 (4.28)
3.89 (4.04)
3.64 (3.85)
3.61 (3.79)
5.63 (5.31)
5.09 (5.06)
4.85
3.69 (4.24)
3.67
3.40 (4.13)
3.44 (3.79)
3.70 (4.60)
4.08
0.46 (30-39)
1.07 (0.69) (31-40)
0.49 (32-41)
0.79 (0.15) (33-42)
0.45 (0.25) (34-43)
0.20 (0.06) (35-43)
0.17 (0.00) (36-43)
2.19 (1.52) (37-43)
1.39 (0.46) (38-44)
9
10
11
12
13
14
15
16
0.64 (45-43)
b
a Chemical shifts in CDCl3. 1Η ΝΜR chemical shifts in DMSO-d6 are in parentheses. c The compared compounds are in parentheses.
persistent C-H· · ·F bond in solution.23 On the other hand, by
TABLE 8. Solvent Data for Meldrum’s Acids 7 and 13
increasing the polarizabitity of the acceptor, the additional steric
hindrance imposed by cyclic benzylic substituents was success-
fully overcome and C-H· · ·X bond formation favored. As
depicted in Table 7, thiomethoxy cyclohexyl substrate 37 led
to the largest chemical shift difference in this series of
Meldrum’s acids, with a ∆ ppm of 2.19 (entry 8), supporting
an intramolecular C-H· · ·S interaction in solution.
H(5) chemical
shift (ppm)
entry
solvent
for 13
for 7
∆ ppm (13 - 7)
1
2
3
4
5
6
C6D6
CDCl3
CD3NO2
CD3CN
acetone-d6
DMSO-d6
∆ ppm (high - low)
5.20
5.21
5.40
5.28
5.37
5.04
0.36
3.25
3.58
4.06
4.05
4.31
4.56
1.31
1.95
1.63
1.34
1.23
1.06
0.48
In addition, subtle variation in the structure of the geminal
disubstituent had a profound impact on the ∆ ppm as shown
with cyclohexane and cyclohexene derivatives 34 and 38 (entries
5 and 9). When one of the sp3-hybridized methylene substituents
on the benzylic carbon center was replaced with a less sterically
demanding sp2-hybridized methine group (Table 7, entry 9), the
chemical shift of H(5) substantially increased; Meldrum’s acid
38 revealed a H(5) chemical shift of 5.09, in comparison to
3.89 for 34. The large ∆ ppm of 1.39 clearly showed that the
deshielding of H(5) was due to C-H· · ·O hydrogen bonding,
and not an artifact caused by the electronics of the alkene
functional group.
drum’s acids. In CDCl3, the chemical shifts of H(5) and related
∆ ppm values likely correlate with the relative strength of the
intramolecular hydrogen bonds.
It was envisaged that further information on the relative
strength of the intramolecular C-H· · ·X (X ) O, S, F) hydro-
gen bonding would be revealed through solvent effects, which
were initially probed by using Meldrum’s acid 13 (Table 8). In
all solvents examined, including strong hydrogen bond accep-
tors, H(5) of Meldrum’s acid 13 exhibited a similar downfield
chemical shift, with a small difference of 0.36 ppm between
the highest and lowest values. On the other hand, the chemical
shift of H(5) for Meldrum’s acid 7, which does not possess the
ability to form an intramolecular hydrogen bond, spanned 1.31
ppm. In DMSO-d6, the smallest difference (0.48 ppm) of the
chemical shift values of H(5) between 7 and 13 was obtained.
These results suggest that 13 formed a persistent intramolecular
hydrogen bond in solution that was not disrupted by the nature
of the medium, except in DMSO-d6, in which an equilibria of
intramolecular and intermolecular hydrogen-bonded species was
observed, resulting in a decrease of the chemical shift.
(d) Relative Intramolecular C-H· · ·X (X ) O, S, F)
Bond Strength in CDCl3 versus DMSO-d6. At this stage, a
persistent intramolecular C-H· · ·X (X ) O, S, F) bond in
solution had been demonstrated for a variety of benzyl Mel-
(21) The origins of the gem-dialkyl effect were discussed in thermodynamic
terms, see: (a) Allinger, N. L.; Zalkow, V. J. Org. Chem. 1960, 25, 701–704.
(22) For a study on the gem-dialkyl effect on intramolecular bonding in
2-substituted propane-1,3-diols, see: Schleyer, P. v. R. J. Am. Chem. Soc. 1961,
83, 1368–1373.
(23) These two types of interactions involve different physicochemical factors
and the small ∆ ppm for 36 may have resulted from a combination of several
effects unlike the C-H · · ·C contact in 35.
J. Org. Chem. Vol. 74, No. 3, 2009 1263

Fillion et al.
TABLE 9. Solvent Data for Meldrum’s Acids 34 and 43
The relative strength of the C-H· · ·O bond for Meldrum’s
1
acids 2-4 was also evaluated. The H chemical shifts of H(5)
H(5) chemical
shift (ppm)
for 2 (4.63 ppm), 3 (4.45 ppm), and 4 (5.18 ppm) in DMSO-d6
were measured and spanned 0.73 ppm, whereas the 1H chemical
shifts of H(5) for 2 (4.13 ppm), 3 (4.29 ppm), and 4 (5.38 ppm)
in CDCl3 spanned 1.25 ppm (Table 1, entries 1-3). In DMSO-
d6, comparing 2-4 with analogues 5-7 and 20-22 revealed
negligible and negative ∆ ppm values for 2 (∆ ppm 2 vs 5 )
-0.14 and ∆ ppm 2 vs 20 ) -0.03) and 3 (∆ ppm 3 vs 6 )
-0.27 and ∆ ppm 3 vs 21 ) -0.02), but significant and positive
∆ ppm values for 4 (∆ ppm 4 vs 7 ) 0.62 and ∆ ppm 4 vs 22
) 0.73). That data suggest that in DMSO-d6, an equilibrium of
intramolecular and intermolecular hydrogen-bonded species was
established for 4, thus lowering the chemical shift of H(5). For
compounds 2 and 3, this equilibrium was entirely displaced
toward the intermolecular hydrogen-bonded species, providing
consistent increases in H(5) chemical shifts (Table 1, entries 1
and 2), equivalent to the chemical shift values observed for 20
and 21 (Table 4, entries 1 and 2). The H(5) chemical shift
increases for 2 and 3 in DMSO-d6 versus CDCl3 indicated that
the intramolecular C-H· · ·O interactions were weaker than the
intermolecular C-H· · ·OS(CH3)2, and accordingly, weaker than
the intramolecular hydrogen bond found in Meldrum’s acid 4.
A parallel was made with monomethoxy Meldrum’s acid 13
and 38, for which the intramolecular C-H· · ·O interactions were
of comparable strength as the one observed for 4, with ∆ ppm
values in CDCl3 of 1.63 and 1.39 and DMSO-d6 of 0.48 and
0.46, respectively (Table 1, entry 12, and Table 7, entry 9). Of
note, Meldrum’s acid 22 did not show a deshielded H(5) in
DMSO-d6, with relatively similar chemical shifts for 20-22
(Table 4). This observation may suggest an alternative confor-
mation for 22 in DMSO-d6 compared to that in CDCl3, a
consequence of intermolecular C-H· · ·O hydrogen bonding of
22 with DMSO-d6.
entry
solvent
for 34
for 43
∆ ppm (34 - 43)
1
2
3
4
5
6
C6D6
3.85
4.16
4.08
4.12
3.89
4.04
0.31
3.19
3.58
3.57
3.67
3.44
3.79
0.60
0.66
0.58
0.51
0.45
0.45
0.25
CD3NO2
CD3CN
acetone-d6
CDCl3
DMSO-d6
∆ ppm (high - low)
ppm) in the chemical shift values of H(5) between 34 and 43
was measured. The ∆ ppm values discussed above are indicative
of a persistent intramolecular hydrogen bond in solution for
Meldrum’s acid 34, which overall parallel the observations
reported in Table 8 with use of gem-dimethyl substrates 7 and
13. The trend was similar to the one observed with 7 and 13
(Table 8), the smallest ∆ ppm values were in DMSO-d6 and
the largest in C6D6 (Table 9, entries 6 and 1). Again, these results
suggested that DMSO-d6 displaced the equilibrium of intramo-
lecular hydrogen bonded versus intermolecular hydrogen bonded
benzyl Meldrum’s acid species toward the latter. The fact that
the chemical shift of H(5) increases in DMSO-d6 compared to
CDCl3, despite the disruption of the C-H· · ·O bond, suggests
a substantial conformational change, which places H(5) in an
entirely different chemical environment. As discussed previ-
ously, a ∆ ppm of 0.45 in CDCl3 for 34 was indicative of a
weak C-H · · · O bond and the results obtained in DMSO-d6
corroborated this deduction.
That trend was also noted for compounds 31 and 33 for which
the chemical shifts of H(5) in DMSO-d6 were slightly larger
than in CDCl3 (Table 7, entries 2 and 4). Moreover, while the
H(5) chemical shifts of 31 in CDCl3 were slightly lower than
those in 13, compound 31 provided ∆ ppm values in DMSO-
d6 superior to the one observed for the gem-dimethyl analogue
13 (0.69 versus 0.48 ppm) (Table 7, entry 1). The correlation
observed for the gem-dimethyl compounds regarding the relative
intramolecular C-H· · ·X (X ) O) bond strength in CDCl3
versus DMSO-d6, illustrated by the chemical shifts of H(5) and
∆ ppm values in both solvents, does not seem to hold for benzyl
Meldrum’s acids displaying greater steric hindrance at the
benzylic position. Therefore, in DMSO-d6, the chemical shift
of H(5) does not correlate exclusively with the strength of the
intramolecular C-H· · ·O bond, but also to conformational
changes.
Similar solvent effects were observed with other hydrogen
1
bond acceptors. The acquisition of the H NMR spectra of 27
and 37 in strong hydrogen bond acceptor solvent DMSO-d6
further substantiated intramolecular C-H· · ·S bonding in these
compounds (Table 5, entry 8 and Table 7, entry 8); the H(5)
chemical shifts were insensitive to solvent polarity and remained
practically unchanged at a distinctly downfield position com-
pared to 7 and 43, respectively.
In comparison to CDCl3, the chemical shift of H(5) for
Meldrum’s acid 24 was unaffected in DMSO-d6 at 4.57 ppm,
the same shift as 7 (Table 5, entries 1 and 5), suggesting that
the intramolecular C-H· · ·F bond in 24 is of equal strength to
the intermolecular bond C-H· · ·OS(CH3)2. The chemical shift
of H(5) for Meldrum’s acid 36 was equal in DMSO-d6 to the
chemical shift of H(5) for acid 43, with a ∆ ppm of zero, which
further supported the absence of C-H· · ·F in 36 (Table 7,
entries 7 and 14). These data suggested that 36, like 43,
exclusively formed an intermolecular C-H· · ·OS(CH3)2 bond
in DMSO-d6.
In the above sections, we delineated through a series of
experiments that 5-benzyl Meldrum’s acid derivatives bearing an
acceptor group at the ortho position of the arene moiety form
persistent intramolecular C-H· · ·X (X ) O, S, Br, Cl, and F)
1
bonds in solution as characterized by H NMR spectroscopy. It
was also established that the presence of a gem-dimethyl group at
the benzylic position is an essential structural requirement in
promoting hydrogen bonding, while a larger cyclic substituent
weakened the intramolecular interaction in CDCl3. Next, intramo-
lecular C-H· · ·X (X ) O, S, F) bonds in solid state will be
characterized and correlated with the results obtained in solution.
2. Characterization of Intramolecular C-H· · ·X (X )
O, S, F) Bonds in the Solid State by X-ray Crystal-
lography. Further information on the intramolecular hydrogen
bonding C-H· · ·X (X ) O, S, F) bond in Meldrum’s acid
derivatives was revealed through X-ray analysis of 2, 3, 4, 10,
22, 24, 27, 35, 36, 38, 43, 45, and 46. Two criteria are typically
Solvent effect studies were also carried out with 34 and 43.
As shown in Table 9, the chemical shift of H(5) for Meldrum’s
acid 43, which does not possess the ability to form an
intramolecular hydrogen bond, spanned 0.60 ppm, and a smaller
range of 0.31 ppm was observed for 34. It is noteworthy that
the chemical shift of H(5) for 43 was affected to a lesser extent
by the nature of the solvent then Meldrum’s acid 7. Conse-
quently, a smaller range of 0.60 ppm for 43 versus 1.31 ppm
for 7 was observed. In DMSO-d6, the smallest difference (0.25
1264 J. Org. Chem. Vol. 74, No. 3, 2009

Persistent Intramolecular C-H· · ·X Bonding in Solution
ates an intramolecular C-H· · ·O bond. Since the interatomic
distance H(5) · · ·O(20) in the solid-state structures of 4 arises
through attractive forces, this correlates with the NMR data in
solutionsthe downfield shift of the acidic proton. Chemical shift
variations were much less pronounced for compounds 2 and 3
(Table 1), and accordingly, the interatomic distance H(5) · · ·O(20)
in the solid state was significantly longer. A correlation between
the C-H · · · O distance and the tendency of the triad C(5)-
H(5)· · ·O(20) to assume a linear configuration (angle (θ)) was
also observed. The bond angle (θ) C(5)-H(5)· · ·O(20) (130.6°)
found in 4 was larger than the one found in 2 and 3 (Table 10).
Despite these structural differences in the solid state, compounds
2 and 3 met the criteria for a C-H· · ·O bond, which parallels
the results obtained in solution.2,24
1
In solution, H NMR studies supported the presence of a
persistent C-H· · ·F bond in Meldrum’s acid 24. In the solid
state, bond length and directionality criteria were met (Table
10), with a H · · ·F distance of 2.13 Å, which is much shorter
than the sum of the van der Waals radii for H (1.20 Å) and F
(1.47 Å) atoms.25 Similarly, Meldrum’s acid 27 containing a
thiomethoxy acceptor revealed a C-H· · ·S bond, with a H · · ·S
distance of 2.43 Å, which is substantially shorter than the sum
of the van der Waals radii for H (1.20 Å) and S (1.80 Å) atoms.
The conformation observed in the solid state is in relation with
the results obtained in solution, for which Meldrum’s acid 27
displayed a large chemical shift variation for H(5), likely a result
of the superior polarizability of the sulfur atom. Of note,
regardless of the acceptor atom in C-H· · ·X (X ) O, S, F),
the difference between the sum of the van der Waals radii and
the observed H · · ·X distance in the solid state was constant for
Meldrum‘s acids 4, 24, and 27, between 0.54 and 0.57 Å.
The assessment of the X-ray structure of compounds 2-4,
24, and 27 ascertained a number of conformational similarities.
First, in all the structures, the Meldrum’s acid moiety was in a
boat conformation in which the acidic H(5) hydrogen occupied
the pseudoaxial position and the benzylic group the pseudoequa-
torial one. Second, the substituents on the benzylic carbon and
C(5) Meldrum’s acid’s carbon were staggered. This feature
places one of the substituents on the benzylic carbon almost
perfectly antiperiplanar to the C(5)sH(5) bond of the Meldrum’s
acid moiety, and the aryl group gauche to H(5) and one of the
carbonyl groups. Third, the conformation adopted by the
aromatic group minimizes the A1,3-allylic strain; the hydrogen
at the unsubstituted ortho position (6-position) of the aromatic
eclipses the benzylic substituent, which is antiperiplanar to the
C(5)sH(5) bond.
FIGURE 2. X-ray structures of Meldrum’s acids 4 and 46. The dihedral
angle H(5)-C(5)-C(11)-C(22) for 4 is 176.1°. Only one molecule
of the data set is illustrated. The dihedral angle for the second molecule
is 177.9°. The dihedral angle H(5)-C(5)-C(11)-C(22) for 46 is 59.4°.
TABLE 10. Solid Phase Structures
∆ (sum of the
van der Waals
bond angle
(θ deg)
C(5)-H(5)· · ·X
temp bond length (Å) radii - obsd X-ray
acid
(K)
C(5)-H(5)· · ·X
distance) (Å)
2
3
4
4
180
180
2.55 (X ) O)
2.57 (X ) O)
0.17
0.15
0.57
0.56
(0.59)
115.9
119.7
130.6
297a 2.15 (X ) O)
180a 2.16 (X ) O)
(2.13)b (X ) O)
180
128.2
(128.0)b
10
22
24
27
35
36
38
43
45
46
180
180
180
150c
150
150
150d
180
180
2.42 (X ) C)
2.13 (X ) F)
2.43 (X ) S)
0.48
0.54
0.57
132.1
138.2
2.11 (X ) O)
2.12 (X ) O)
0.61
0.60
125.5
134.7
a A crystal phase change was observed at 291 K so data are reported
for both phases. b The unit cell contains two molecules, thus two sets of
data. c The unit cell contains four molecules, thus four sets of data.
d The unit cell contains eight molecules, thus eight sets of data.
The strong tendency of benzyl Meldrum’s acid to adopt that
conformation was confirmed by analyzing the X-ray structure
of 4-methoxy derivative 10. The features observed for com-
pounds 2, 3, 4, 24, and 27 were found in 10, even in the absence
of an intramolecular hydrogen bond.26 The presence of the
benzylic gem-dimethyl group then seems to reinforce the
positioning of the ortho substituent of the aromatic ring in close
proximity to H(5). This was further established by examining
the X-ray structure of 22. The conformation of 22 in the solid
state was comparable to that of structures 2, 3, 4, 10, 24, and
used to judge the presence of the C-H· · ·X hydrogen bond in
the solid state, namely bond length and directionality.2 First, a
H· · ·X (X ) O, S, F) distance equal to or shorter than the sum
of the van der Waals radii of H and X is required. Second, the
directionality is characterized by the C-H· · ·X (X ) O, S, F)
bond angle (θ), a linear bond angle being ideal, and 110° being
the minimum requirement.
The X-ray structures of Meldrum’s acids 2, 3, and 4 were
examined to detect the presence of an intramolecular C-H· · ·O
bond between H(5) of Meldrum’s acid and the oxygen of the
o-methyl ether of the aromatic moiety (Figure 2). In Meldrum’s
acid 4, an obvious H(5) · · ·O(20) contraction was observed (by
about 0.4 Å) when compared with analogous compounds 2 and
3 (Table 10). The H · · ·O distance was considerably shorter (2.13
to 2.16 Å) than 2.72 Å, the sum of the van der Waals radii for
oxygen (1.52 Å) and hydrogen atoms (1.20 Å), which substanti-
(24) Compounds 2 revealed, through analysis of the crystal packing diagram,
an intermolecular C(5)-H(5) · · · O(10) bond (2.24 Å). Close contacts C(5)-
H(5) · · · O(9) (2.90 Å) and C(5)-H(5) · · · O(10) (3.25 Å) were observed for
Meldrum’s acid 3.
(25) Analysis of the crystal packing diagram showed the formation of an
intermolecular C(5)-H(5) · · · O(10) bond (3.13 Å).
(26) An intermolecular C(5)-H(5) · · · O(9) bond (2.58 Å) was revealed by
analysis of the crystal packing diagram.
J. Org. Chem. Vol. 74, No. 3, 2009 1265

Fillion et al.
difference between the sum of the van der Waals radii and the
observed H· · ·O distance in the solid state was constant, at 0.61
and 0.60 Å, respectively, and paralleled the differences observed
for the dimethyl substrates (Table 10).
The conformation of Meldrum’s acid 36 in the solid state,
for which no experimental evidence clearly supported a
persistent C-H· · ·F bond in solution, was similar to the one
observed for compound 46 and revealed the absence of a
C-H· · ·F bond (Figure 2). Exceptionally, the Meldrum’s acid
moiety was in a chair conformation, placing H(5) at the
pseudoequatorial position and the large alkyl group at the
pseudoaxial position, respectively. The Meldrum’s acid moiety
was positioned at the equatorial position of the cyclohexane.29
The conformation of Meldrum’s acids 35 and 43 in the solid
state was closely related to that of 36. It was rationalized that
the conformation of 35 in solution was likely similar to the one
in the solid state, as the chemical shift of H(5) was slightly
affected by the presence of the ethyl group (Table 7, entry 6).
On the other hand, a substantial chemical shift variation for
H(5) in CDCl3 was observed for Meldrum’s acid 22, strongly
supporting that the ethyl group was in the close proximity of
H(5) in solution (Table 4, entry 3). Indeed, a different
conformation was observed for dimethyl analogue 22 in the solid
state.
FIGURE 3. Meldrum’s acid 46.
27. In solution, NOE studies suggested that the ethyl group was
in the vicinity of H(5). In addition, the deshielding of H(5)
observed by 1H NMR in solution indicated that van der Waals
contacts played a role (Table 4, entry 3). Indeed, in the solid
state, the C(5)-H(5)· · ·C(18) distance of 2.42 Å was shorter
than 2.90 Å, the sum of the van der Waals radii for the carbon
atom (CH2 group) in the ethyl group (1.70 Å) and hydrogen
atoms (1.20 Å).27 These data enabled us to draw a parallel
between the conformations of 22 in solution and in the solid
state, and seemed to be matching. The difference of 0.48 Å
between the sum of the van der Waals radii and the observed
H(5)· · ·C(18) distance in 22 was less important that the ones
determined for C-H· · ·X (X ) O, S, F), a consequence of the
attractive nature of the hydrogen bonding interaction.
The results obtained with 2-4, 10, 22, 24, and 27 suggested
that 5-[1-aryl-1-methylethyl] Meldrum’s acids have a well-
defined three-dimensional structure in solution and in the solid
state. In solution, this preferred conformation is constrained and
has a very limited degree of freedom as demonstrated by the
absence of temperature effect on Meldrum’s acids 2-4 (Table
3). Even when repulsive van der Waals contacts with H(5) are
present, as in 22, the conformation is preserved. A1,3-Allylic
strain is a major contributor to the conformation adopted by
5-[1-aryl-1-methylethyl] Meldrum’s acids, and when an acceptor
(X) group is introduced at the ortho position of the aromatic
ring, the formation of a persistent intramolecular C-H· · ·O bond
in solution is promoted.28 The importance of A1,3-allylic strain
was confirmed with Meldrum’s acid 46 (Figure 3). The presence
of two o-methoxy substituents substantially increased A1,3-allylic
strain at the expense of hydrogen bonding. As a result,
compound 46 adopted a different conformation, in which all
eclipsing interactions were minimized (Figure 2). Such confor-
mation clearly explains the chemical shift observed for H(5) in
CDCl3 at 3.76 ppm, resulting in a low ∆ ppm of 0.18 when
compared with 7.
In this section, sound evidence of the ability of the acidic
C-H bond to drive the conformational properties of benzyl
Meldrum’s acids in the solid state through the formation of
intramolecular C-H· · ·X (X ) O, S, F) bonds was presented.
The data obtained by X-ray crystallography were consistent with
1
the H NMR studies in solution. However, due to possible
packing effects, no direct correlations were found between
C-H· · ·X (X ) O, S, F) bond distances in the solid state and
the chemical shifts and/or the ∆ ppm values in solution.
Conclusion
In summary, we established the occurrence and persistence
1
of C-H· · ·X (O, S, Br, Cl, and F) bond in solution using H
NMR spectroscopy for a large number of benzyl Meldrum’s
acids. The latter are novel and reliable probes for the evaluation
of this type of nonclassical interaction in solution. The
persistence of the C-H· · ·X bond in solution was demonstrated
to be dependent upon structural features present on the aromatic
moiety and the benzylic position of the benzyl Meldrum’s acid
derivatives. The observations presented in this paper highlight
the large potential of Meldrum’s acid in developing an
understanding of the function and nature of C-H · · · X
interactions.
These data established that the conformational properties of
benzyl Meldrum’s acid in solution correlate to the conformation
in the solid state. The parallel between solution and solid-state
three-dimensional structure was further studied with cyclohexyl
derivatives 35, 36, 38, 43, and 45.
The X-ray structure of Meldrum’s acid 45 displayed a short
C-H· · ·O of 2.12 Å, and a large C(5)-H(5)-O(20) angle (θ;
Table 10). These observations are consistent with our analysis
of the presence of intramolecular hydrogen bonding in solution.
Similarly, the X-ray structure of Meldrum’s acid 38 revealed a
short C-H· · ·O distance of 2.11 Å (Table 10). In solution, no
nOe signal could be perceived between the hydrogens of the
alkene and H(5), suggesting an antiperiplanar orientation of these
groups as observed in the solid state. For 38 and 45, the
Experimental Section
General Procedure A: Preparation of Mono-Benzylic
Meldrum’s Acids via a One-Pot Reductive Alkylation. Mel-
drum’s acid (1.0 equiv) and the appropriate benzaldehyde (0.98
equiv) were dissolved in absolute EtOH (0.5-1.0 M), followed by
the addition of catalytic piperidinium acetate (0.1 equiv). The
resulting solution was stirred vigorously for 30 min, and then cooled
to 0 °C. NaBH3CN (1.5 equiv) was added portionwise over 1 h,
(29) A survey of the Cambridge Crystallographic Database revealed a single
structure of benzyl Meldrum’s acids, namely 5-(4-tert-butyl-1-phenylcyclohexyl)-
2,2-dimethyl-1,3-dioxane-4,5-dione. In the absence of a hydrogen bond acceptor,
5-(4-tert-butyl-1-phenylcyclohexyl)-2,2-dimethyl-1,3-dioxane-4,5-dione displayed
a conformation similar to that of compounds 35, 36, and 43, see: Davis, A. P.;
Egan, T. J.; Orchard, M. G. Tetrahedron 1992, 48, 8725–8738.
(27) Analysis of the crystal packing diagram showed an intermolecular C(5)-
H(5) · · · O(9) bond (2.84 Å).
(28) (a) Hoffmann, R. W. Chem. ReV. 1989, 89, 1841–1860. (b) Wender,
P. A.; Ternansky, R. J. Tetrahedron Lett. 1985, 26, 2625–2628.
1266 J. Org. Chem. Vol. 74, No. 3, 2009

Persistent Intramolecular C-H· · ·X Bonding in Solution
and the reaction was allowed to warm to room temperature. The
reaction was monitored by TLC and upon completion was carefully
quenched with 10% HCl (extreme caution should be exercised due
to the eVolution of HCN gas). Vigorous stirring was maintained
until gas evolution had ceased, after which time the reaction was
concentrated in vacuo to remove EtOH. The residue was resus-
pended in 10% HCl and extracted four times with CH2Cl2. The
combined organic layers were dried over MgSO4, filtered, and
concentrated. The resulting product was generally found to be of
General Procedure C: Preparation of Malonates via Acid-
Catalyzed Ring Opening of Meldrum’s Acid Derivatives. In a
typical reaction, Meldrum’s acid derivative was dissolved in a 1:1
mixture of Et2O:MeOH (0.2 M) and concentrated sulfuric acid (1-2
equiv). The reaction mixture was heated to 50 °C for 24 h. After
cooling to room temperature, the reaction mixture was diluted with
Et2O, washed with NaHCO3 (aq) and brine, then dried over MgSO4
and filtered. Concentration of the organic layer provided the crude
product that was purified by flash chromatography on silica gel.
Dimethyl 2-(2,3-Dimethoxybenzyl)malonate (14). Compound
14 was prepared by ring opening of 5-(2,3-dimethoxybenzyl)-2,2-
dimethyl-1,3-dioxane-4,6-dione (2), using procedure C to provide
a 73% yield of a clear oil after purification by flash chromatography
1
good purity by H NMR, but was recrystallized from MeOH or
purified by flash chromatography on silica gel.
5-Benzyl-2,2-dimethyl-1,3-dioxane-4,6-dione (5). Compound 5
was prepared by the reductive alkylation of Meldrum’s acid with
benzaldehyde according to procedure A. Flash chromatography (4:1
hexanes:EtOAc) of the crude reaction mixture provided a white
solid in 47% yield; 1H NMR (CDCl3, 300 MHz) δ 7.31-7.20 (m,
5H), 3.74 (t, J ) 4.8 Hz, 1H), 3.46 (d, J ) 4.5 Hz, 2H), 1.70 (s,
3H), 1.46 (s, 3H); 13C NMR (CDCl3, 75 MHz) δ 165.3, 137.2,
129.7, 128.6, 127.2, 105.2, 48.1, 32.1, 28.4, 27.2.
1
(9:1 hexanes:EtOAc); H NMR (CDCl3, 300 MHz) δ 6.91 (t, J )
7.9 Hz, 1H), 6.78 (d, J ) 8.1 Hz, 1H), 6.71 (d, J ) 7.6 Hz, 1H),
3.83 (s, 3H), 3.82 (s, 3H), 3.78 (t, J ) 7.8 Hz, 1H), 3.66 (s, 6H),
3.19 (d, J ) 7.8 Hz, 2H); 13C NMR (CDCl3, 75 MHz) δ 169.4,
152.5, 147.3, 131.3, 123.7, 122.3, 111.4, 60.4, 55.6, 52.4, 52.0,
29.8; HRMS(EI) m/z calcd for C14H18O6 (M+) 282.1103, found
282.1100.
General Procedure B: Conjugate Addition of Grignard
Reagents to Meldrum’s Alkylidenes. Meldrum’s alkylidene was
suspended in dry THF (0.1 M) and cooled to 0 °C with stirring
under inert atmosphere. A solution of the appropriate Grignard
reagent (2-3 equiv) in THF (or Et2O) (0.5 to 3 M) was added
dropwise, and then the reaction was allowed to warm to room
temperature and monitored by TLC. Upon completion, the reaction
was quenched with 5% HCl and extracted three times with EtOAc.
The combined organic layers were washed with H2O and brine,
then dried over MgSO4 and filtered. Concentration of the organic
layers provided the crude product that was purified by either
recrystallization or flash chromatography on silica gel.
5-[1-(2,5-Dimethoxyphenyl)-1-methylethyl]-2,2-dimethyl-1,3-
dioxane- 4,6-dione (12). Conjugate addition of methylmagnesium
bromide to 5-(1-(2,5-dimethoxyphenyl)ethylidene)-2,2-dimethyl-
1,3-dioxane-4,6-dione (prepared by condensation of Meldrum’s acid
and 2′,5′-dimethoxyacetophenone) according to procedure B pro-
vided a 50% yield of a light yellow solid after recrystallization.
Mp 117-118 °C (MeOH); 1H NMR (CDCl3, 300 MHz) 7.21 (d, J
) 8.5 Hz, 1H), 6.49-6.44 (m, 2H), 5.05 (s, 1H), 3.81 (s, 3H),
3.77 (s, 3H), 1.74 (s, 3H), 1.67 (s, 3H), 1.61 (s, 6H); 13C NMR
(CDCl3, 75 MHz) 164.1, 159.2, 157.4, 127.6, 127.3, 104.1, 103.7,
99.3, 55.1, 55.0, 53.0, 40.0, 28.4, 26.8, 25.5; HRMS(EI) m/z calcd
for C17H22O6 (M+) 322.1416, found 322.1423.
ACKNOWLEDGMENT. This work was supported by the
Natural Sciences and Engineering Research Council of Canada
(NSERC), the Canadian Foundation for Innovation, the Ontario
Innovation Trust, and the University of Waterloo. We gratefully
acknowledge Boehringer Ingelheim (Canada) Ltd., AstraZeneca
Canada Inc., and the Merck Frosst Centre for Therapeutic
Research for unrestricted support of this research. A.W. is
indebted to NSERC for CGS-M and CGS-D scholarships and
D.F. the Government of Ontario for an OGSST scholarship.
The late Dr. Nicholas J. Taylor (Meldrum’s acids 2, 3, 4, 10,
and 46), University of Waterloo, Dr. Jalil Assoud (Meldrum’s
acids 22, 24, 27, and 45), University of Waterloo, and Dr. Alan
Lough (Meldrum’s acid 35, 36, 38, and 43), University of
Toronto, are gratefully acknowledged for X-ray structure
determination. We thank Aaron M. Dumas, University of
Waterloo, for helpful suggestions and proofreading the manuscript.
Supporting Information Available: Experimental proce-
dures, NMR spectra, X-ray structures, and cif files. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO802311W
J. Org. Chem. Vol. 74, No. 3, 2009 1267