A regioselective synthesis of 2,4-dialkyl resorcinols

Article abstract of DOI:10.1016/j.tetlet.2005.01.036

A general regioselective synthesis of 2,4-dialkyl resorcinols from ethoxymethylenemalonate ethyl ester and dialkyl ketones is reported.

Full text of DOI:10.1016/j.tetlet.2005.01.036

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 1731–1734
A regioselective synthesis of 2,4-dialkyl resorcinols
Scott S. Ceglia,* Michael H. Kress, Todd D. Nelson and James M. McNamara
Merck Research Laboratories, Department of Process Research, 466 Devon Park Drive, Wayne, PA 19087, USA
Received 23 November 2004; revised 6 January 2005; accepted 7 January 2005
Available online 28 January 2005
Abstract—A general regioselective synthesis of 2,4-dialkyl resorcinols from ethoxymethylenemalonate ethyl ester and dialkyl
ketones is reported.
Ó 2005 Elsevier Ltd. All rights reserved.
As part of an ongoing effort to identify agents that are
OH
O
OH
agonists of PPAR, we required an efficient synthesis of
substrate 1.1 The core of this molecule can be perceived
as 2,4-dipropyl resorcinol 2. The syntheses of complex
molecules containing substituted phenols and/or resor-
cinols often times start with aromatic precursors that re-
quire protecting group manipulation and/or additional
synthetic steps to establish regiochemical control.2 Our
initial approaches toward 2 followed some of these tra-
ditional methodologies and suffered similar draw-
backs.3–5 We desired a more efficient synthesis of
resorcinol 2. As a result, this communication discloses
a concise synthesis of 2,4-dipropyl resorcinol 2, and a
general approach to other 2,4-dialkyl resorcinols that
is regioselective, convergent, and practical.
CO2Et
CO2Et
O
hydrolysis
decarbox.
HO
HO
3
2
4
O
O
EtO
OEt
OEt
+
O
5
6
Scheme 1.
compounds6 (i.e., 4 to 3). We envisioned that a rear-
rangement of the appropriately substituted a-pyrone 4
would afford the framework of our target dialkylated
resorcinol. Subsequent saponification and decarboxyl-
ation would give 2,4-dipropyl resorcinol 2 (Scheme 1).
Intermediate 4 was envisioned to be accessible based
on an extension of a literature precedent, by condensa-
tion of 5-nonanone 5 with 2-ethoxymethylenemalonate
ethyl ester (EMME) 6.7
N
O
OH
CF3
HO2C
O
HO
1
2
The keystone of our approach had its origins in limited
references to the isomerization of a-pyrones to aromatic
Our initial attempts at this approach toward 4 utilized
ethoxide as the base and generated pyrone 4 in a modest
HPLC assay yield (33%). The yield of pyrone 4 im-
proved slightly (42%) using KOtBu as the base. Interest-
ingly, with KOtBu as the base, a near equimolar amount
of arene 3 was observed. Because alkoxide bases have
been shown to induce the isomerization of a-pyrones,
we initially speculated that the observed product distri-
bution was under thermodynamic control. As such, we
Keywords: Resorcinol; Dianion; 2-Ethoxymethylenemalonate ethyl
ester.
*
Corresponding author. Tel.: +1 215 652 3361; fax: +1 215 993
3409; e-mail: scott_ceglia@merck.com
Merck Research Laboratories, Department of Process Research, 126
East Lincoln Ave., Rahway, NJ 07065, USA
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.01.036

1732
S. S. Ceglia et al. / Tetrahedron Letters 46 (2005) 1731–1734
Table 1. Isomerization attempt of pyrone 4
O
O
Base/solvent
Time
Temperature
% Pyrone 4
(recovery)a
EtO
OEt
OEt
+
O
5
LHMDS/THF
W/1 equiv EtOH
No age
1 h
1 h
5 °C
25 °C
40 °C
40 °C
27
38
90
93
6
1eq LHMDS
16 h
O
a Determined by HPLC assay yield wt % analysis versus standard.
O
CO2Et
EtO
CO2Et
O
- EtO
were surprised when we observed no further change in
product distribution upon prolonged age (>5 days) at
elevated temperature. Subsequent attempts to isomerize
isolated pyrone 4 to arene 3 using various bases (KOH/
EtOH, NaOEt/EtOH, LHMDS/THF), temperatures,
and reaction times were also unsuccessful and led mainly
to degradation of 4à. Intriguingly, attempted isomeriza-
tion of pyrone 4 with LHMDS and 1 equiv of EtOH,
(Table 1) resulted in the immediate disappearance of
pyrone 4; however; conversion back to pyrone 4 was
nearly quantitative after a more prolonged age time.
+ EtO
OLi
4
7
2eq LHMDS
O
OLi
OEt
OH
O
EtO
C-acylation
OEt
HO
OLi
These results indicate that pyrone 4 was not efficiently
undergoing isomerization to arene 3, but that arene 3
was being generated directly.8 We rationalized the direct
formation of arene 3 was occurring via an open chain,
dianion manifold (Scheme 2).9 This pathway would
require a minimum of 3 equivalents of base. The first
equivalent of base would generate the enolate of 5-non-
anone 5 that would add via Michael-addition into
EMME 6. The second equivalent would deprotonate
the ethanol liberated from the 1:1 5-nonanone/EMME
adduct 7. Finally, the last equivalent would generate
dianion 8. Intramolecular C-acylation of this dianion
intermediate results in arene 3. Formation of dianion 8
does not however insure the formation of arene 3. If
intramolecular O-acylation of dianion 8 occurs, pyrone
4 would result. Thus, the task at hand was to increase
the rate of C-acylation relative to the rate of O-
acylation.
3
8
NaOH
OH
HO
2
Scheme 2.
material consisted of minor amounts of pyrone 4,
open-chain isomers of 7, and a cyclobutanone interme-
diate 9 (determined via H NMR spectroscopy), that
1
presumably could form via the double-Michael addition
of one molecule of ketone 5 to EMME. Interconversion
of cyclobutanone adduct 9 to arene 3 under the reaction
conditions can be envisioned to occur via retro-Michael
to dianion 8, followed by ring closure to the more stable
tetrasubstituted arene 3 (Scheme 3). In fact, incorpora-
ting an extended age (>4 h) at 40 °C, led to the disap-
pearance of cyclobutanone adduct 9, and a higher
yield of arene 3.
We began our investigations into the direct formation of
arene 3 by screening a variety of base (>3 equiv) and sol-
vent systems. Due to incompatibilities of EMME 6 with
base, we established a reagent addition protocol in
which one equivalent of base was added to a solution
of ketone 5 in THF. EMME was then added to the
resulting enolate solution. Analysis of the reaction pro-
file via HPLC and H NMR spectroscopy at this point
showed a complex mixture of open-chained tautomers/
rotomers of intermediate 7.
1
Following this protocol our screening studies yielded the
following conclusions. The use of a strong base such as
LHMDS or NaHMDS (Table 2, entries 1, 2, and 3) re-
sulted in a high assay yield of arene 3, with only minor
amounts of pyrone 4 being formed. The nature of the
counter ion also played a role in the reaction. The use
of NaHMDS resulted in a lower yield of arene 3 (Table
2, entry 3 vs 1), and a significantly larger amount of pyr-
one 4. In addition, solvent choice was shown to affect
the course of the reaction. Use of a nonpolar solvent
such as hexane (entry 2) resulted in increased levels of
undesired pyrone 4, while a more polar solvent such as
The remaining 2.3 equiv of LHMDS were then charged
to the reaction mixture. After 30min at ambient temper-
ature, the reaction profile had coalesced to predomi-
nantly one peak in the HPLC chromatogram, but
contained only 50% of arene 3. The remainder of the
à We also monitored the formation of resorcinol 2, which could
potentially be formed under the basic reaction conditions by
saponification and decarboxylation of arene 3, but none was detected
by HPLC analysis.

S. S. Ceglia et al. / Tetrahedron Letters 46 (2005) 1731–1734
1733
OLi
EtO2C CO2Et
OH
EtO2C
OEt
CO2Et
HO
OLi
O
9
8
3
Scheme 3.
Table 3. LHMDS Base equivalent study
THF (entry 1) provided a better product distribution.
We hypothesized that the nonpolar solvent de-stabilizes
dianion 8 and causes rapid O-acylation (Table 2).
Entry
LHMDS charge
% Yield (3)
% Yield (4)
1
2
3
4
5
LHMDS (1 equiv)
LHMDS (2 equiv)
LHMDS (3 equiv)
LHMDS (3.3 equiv)
LHMDS (4 equiv)
2
37
75
85
88
61
25
3
Concomitantly, in order to test our original hypothesis,
we were exploring the effect that the number of equiva-
lents of base had on the formation of arene 3, since the
reaction could still progress through the mono or dian-
ion pathways under the screening protocol. (Scheme 2).
A base equivalency study (Table 3), showed that the
presence of excess base minimizes the formation of pyr-
one 4, resulting in an even higher yield of arene 3.
<1
<1
ester 3 at reflux in 5 N NaOH for an extended age
(>10h), followed by acidic work up. Our streamlined
procedure for the synthesis of resorcinol 2 resulted in
an overall 74% isolated yield starting from 5-nonanone
5.
On the basis of these equivalency studies, we switched to
an inverse addition protocol (B) to finish the base
screening study. The results of these experiments indi-
cated that the use of a strong coordinating base
(LHMDS) in a polar solvent using an inverse addition
protocol was optimal for maximizing the formation of
arene 3 and minimizing the formation of pyrone 4. This
optimized procedure for the formation of ester 3 re-
sulted in an 85% isolated yield starting from 5-nonanone
5.
Substituted resorcinols represent a ubiquitous structural
motif present in many natural products and pharmaceu-
tical agents.11 Resorcinols have also found utility as
building blocks for the development of a wide range of
macrocycles and molecules that exhibit interesting rec-
ognition properties.12 As such, we were quite interested
in whether this new approach to 2,4-dipropylresorcinol
2 could be extended to other enolizable ketones.
To this end, other symmetric ketones were condensed
with EMME under our optimized conditions (Scheme
4). The straight-chained alkyl ketones (5a, 5b, and 5c)
performed well, resulting in excellent yields of the 2,4-
dialkyl-5-carboethoxy resorcinols (10a, 10b, and 10c).
Reaction of EMME with more sterically demanding
symmetric ketones, such as R = isopropyl (5d) led to a
lower yield (47%) of 2,4-dialkyl-5-carboethoxy resor-
cinol. When utilizing diphenyl acetone (5e), selective for-
mation (73%) of pyrone was observed. This lower yield
With an optimized procedure in hand for the formation
of 3, we then turned our attention toward the hydrolysis/
decarboxylation of ester 3 to 2,4-dipropyl resorcinol 2.
This transformation was believed to be a relatively facile
conversion based on work done on similar resorcinillic
esters.10 Formation of 2 was accomplished by heating
Table 2. Reaction condition screening experiments
Entry Base (3.3 equiv) Solvent
Method (3) %a 3:4b
1
2
3
4
5
LHMDS
LHMDS
NaHMDS
LHMDS
LDA
THF
A
A
A
B
B
83
65
42:1
9:1
Hexanes
THF
THF
CO2Et
607:1
85
76
EtO
R
OH
R
O
R
>80:1
13:1
CO2Et
R
CO2Et
or
CO2Et
6
Hexanes/
ethylbenzene
THF
O
LHMDS
+
6
7
8
KOtBu
NaOEt
BuLi
B
B
B
42
4
1:1
0.1:1
2:1
HO
R
Ethanol
THF
R
O
16
(a) Dropwise addition of remaining 2.3 equiv of LHMDS to a pre-
cooled solution of 7.
11a, ---
11b, ---
11c, ---
11d, ---
11e, 72%
5a-e, R= a, Me
b, Et
10a, 88%
10b, 97%
10c, 88%
10d, 47%
10e, ---
(b) Dropwise addition of a solution of 7 to remaining 2.3 equiv of
LHMDS.
c, n-Pr
d, i-Pr
a HPLC assay yield.
e, Ph
b Ratio determined using HPLC assay yields calculated from LC
weight % standards of 3 and 4.
Scheme 4.

1734
S. S. Ceglia et al. / Tetrahedron Letters 46 (2005) 1731–1734
selectivity (a) Amorati, R.; Attanasi, O. A.; Ali, B. E.;
Filippone, P.; Mele, G.; Spadavecchia, J.; Vasapollo, G.
Synthesis 2002, 2749; (b) Graybill, T. L.; Casillas, E. G.;
Pal, K.; Townsend, C. A. J. Am. Chem. Soc. 1999, 121,
7730.
for the isopropyl and selectivity to form the pyrone for
the phenyl derivative is presumably due to the greater
steric demand imparted on the more rigid arene frame-
work versus that of the pyrone, as well as the potential
anion stabilization that can occur in the conjugated
phenyl derivative.
5. (a) Thermal rearrangement of resorcinol bis-allyl ether in
o-dichlorobenzene resulted in a 2:1 mixture of regioisom-
ers favoring 2,4-diallylresorcinol and resulted in only a
37% isolated yield from resorcinol; (b) Addition of
ethylmagnesium bromide to 2,4-dimethoxybenzaldehyde,
followed by catalytic hydrogenation under acidic condi-
tions, provided a near quantitative yield of 4-propylresor-
cinol dimethyl ether. Directed ortho-metallation (with n-
propyl iodide quench) did not completely alkylate; we
routinely observed 88–90% yield of the 2,4-dipropyl
resorcinol dimethyl ether that still required deprotection
to afford 2.
6. (a) Schmidt, Hans-Georg. EP 0074497, 1983; (b) Schmidt,
Hans-Georg. DE 32 35 019 A1, 1983; (c) Money, T.;
Comer, F. W.; Webster, G. R. B.; Wright, I. G.; Scott, A.
I. Tetrahedron 1967, 23, 3435; (d) Money, T. Chem. Rev.
1970, 70, 553.
In summary, a novel and convenient procedure was
developed for the synthesis of 2,4-dipropyl resorcinol 2
via condensation of 5-nonanone 5 with EMME 6 fol-
lowed by hydrolysis/decarboxylation. This methodology
was extended to other symmetrical ketones, which re-
sulted in the high yielding regioselective synthesis of
2,4-dialkyl resorcinols. Evaluation of the scope and
selectivity of this reaction with other unsymmetrical
and functionalized enolizable ketones will be part of
future studies.
Acknowledgements
7. (a) Milata, V. Aldrichim. Acta 2001, 34, 20; (b) Boger, D.
L.; Mullican, M. M. J. Org. Chem. 1984, 49, 4033.
8. Direct formation of functionalized aromatics has recently
been reported via [3+3] cyclization of (3-siloxymethylid-
ene)acetylacetone or 1,1,-diacylcyclopropanes with 1,3-
bis-silyl enol ethers in a similar fashion (a) Dede, R.;
Langer, P. Tetrahedron Lett. 2004, 45, 9177; (b) Bose, G.;
Nguyen, V. T. H.; Ullah, E.; Lahiri, S.; Gorls, H.; Langer,
P. J. Org. Chem. 2004, 69, 9128.
We would like to thank David M. Weingarten and Carl
LeBlond for preliminary experiments and helpful discus-
sion. We would also like to thank Pete Dormer and Lisa
Dimichele for their NMR support.
References and notes
9. For dianion reviews see: (a) Thompson, C. M.; Green, D.
Tetrahedron 1991, 47, 4223; (b) Petragani, N.; Yonashiro,
M. Synthesis 1982, 521; For dianion reviews see: (c)
Kaiser, E. M.; Petty, J. D.; Knutson, P. L. Synthesis 1977,
509.
10. (a) Anker, R. M.; Cook, A. H. J. Chem. Soc. 1945, 311; (b)
Marmor, R. S. J. Org. Chem. 1972, 2901; (c) Schmidt, H.-
W. Synthesis 1985, 778.
11. (a) Nicolau, K. C.; Vassilikogiannakis, G.; Simonsen, K.
B.; Baran, P. S.; Zhong, Y.-L.; Vidali, V. P.; Pitsinos, E.
N.; Couladouros, E. A. J. Am. Chem. Soc. 2000, 122,
3071; (b) Furstner, A.; Seidel, G. J. Org. Chem. 1997, 62,
2332; (c) Trifonov, L. S.; Bieri, J. H.; Prewo, R.; Dreiding,
A. S. Tetrahedron 1983, 39, 4243.
12. (a) Davis, C. J.; Lewis, P. T.; Billodeaux, D. R.; Fronczek,
F. R.; Escobedo, J. O.; Strongin, R. M. Org. Lett. 2001, 3,
2443; (b) Papefstathiou, G. S.; MacGillivray, L. R. Org.
Lett. 2001, 3, 3835; (c) Saito, S.; Rudkevich, D. M.;
Rebek, J., Jr. Org. Lett. 1999, 1, 1241.
1. (a) Adams, A. D.; Hu, Z.; von Langen, D.; Dadiz, A.;
Elbrecht, A.; MacNaul, K. L.; Berger, J. P.; Zhou, G.;
Doebber, T. W.; Meurer, R.; Forrest, M. J.; Moller, D. E.;
Jones, A. B. Bioorg. Med. Chem. Lett. 2003, 13, 3185; (b)
Kun Liu, et al. J. Med. Chem., in press.
2. Jones, R. J.; Van De Water, R. W.; Lindsey, C. C.;
Hoarau, C.; Ung, T.; Pettis, T. R. R. J. Org. Chem. 2001,
66, 3435.
3. Modest to poor regioselectivity of the Claisen rearrange-
ment of meta substituted allyl phenyl ethers has recently
been demonstrated to be sensitive to the electronic nature
of the substituents (a) Gozzo, F. C.; Fernandes, S. A.;
Rodrigues, D. C.; Eberlin, M. N.; Marsaioli, A. J. J. Org.
Chem. 2003, 68, 5493; (b) Bhide, B. H.; Akolkar, V. D.;
Prabhu, V. G. Indian J. Chem. 1992, 31, 143.
4. The rearrangement of 5-n-pentylresorcinol bis-allyl ether,
has been recently reported to afford 2,4-bisallyl-5-n-
pentylresorcinol in 70% yield with no mention of regio-