1,4-Addition of an aryllithium reagent to diethyl ketomalonate. Scalable synthesis of ethyl 1-(hydroxymethyl)-1,3-dihydroisobenzofuran-1-carboxylate

Article abstract of DOI:10.1016/j.tetlet.2012.05.052

While optimizing the synthesis of pharmaceutical building block 3 [ethyl 1-(hydroxymethyl)-1,3-dihydroisobenzofuran-1-carboxylate], we encountered an unusual addition of an aryllithium reagent to the ketone oxygen atom of diethyl ketomalonate. Compound 3 was ultimately prepared on a large scale by a two-step sequence involving (1) annulation of a functionalized Grignard reagent with diethyl ketomalonate and (2) selective mono-reduction of a geminal diester using lithium tri-tert-butoxyaluminum hydride. 2012 Elsevier Ltd. All rights reserved.

Full text of DOI:10.1016/j.tetlet.2012.05.052

Tetrahedron Letters 53 (2012) 5467–5470
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
1,4-Addition of an aryllithium reagent to diethyl ketomalonate. Scalable
synthesis of ethyl 1-(hydroxymethyl)-1,3-dihydroisobenzofuran-1-carboxylate
⇑
Benjamin N. Rocke , Edward L. Conn, Shane A. Eisenbeis, Roger B. Ruggeri
Pfizer Worldwide Research and Development, Eastern Point Rd., Groton, CT 06340, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 November 2011
Revised 9 May 2012
Accepted 10 May 2012
Available online 17 May 2012
While optimizing the synthesis of pharmaceutical building block 3 [ethyl 1-(hydroxymethyl)-1,3-
dihydroisobenzofuran-1-carboxylate], we encountered an unusual addition of an aryllithium reagent
to the ketone oxygen atom of diethyl ketomalonate. Compound 3 was ultimately prepared on a large
scale by a two-step sequence involving (1) annulation of a functionalized Grignard reagent with diethyl
ketomalonate and (2) selective mono-reduction of a geminal diester using lithium tri-tert-butoxyalumi-
num hydride.
Keywords:
Directed ortho metalation
Ó 2012 Elsevier Ltd. All rights reserved.
1
,4-Addition
Annulation
Mono-reduction
Diethyl ketomalonate
Introduction
of a pre-constructed ring system would minimize the length of the
synthetic route. Elaboration to alcohol 3 relied on the reactivity of a
Drug discovery programs depend on the rapid synthesis of
experimental medicines. To this end, research organizations main-
tain stores of small, multi-functional molecules that can be readily
incorporated or transformed into novel structures of pharmaceuti-
cal interest. Naturally, robust synthetic procedures to access these
building-block molecules are critical, as they allow rapid progres-
sion from milligram to multi-gram preparations.
benzylic methylene group through iterative deprotonation and
treatment with electrophilic reagents. Whereas CO was the only
2
5
,6
precedented electrophile giving a product in the desired (carbox-
ylic acid) oxidation state, we were able to access the ethyl ester
directly in higher yield via inverse addition to ethyl cyanoformate.
Theoretically, addition of phthalan anion to a solution of excess
electrophile should suppress subsequent reaction of the similarly
electrophilic product. We found compound 2 to be sufficiently
acidic to allow hydroxymethylation simply using paraformaldehyde
Historically, organometallic methods have been limited due to
incompatibility with pharmaceutically desirable polar functional-
1
ity, but recent developments have led to increased versatility. For
2
instance, the use of in situ protecting groups and, significantly, the
3
development of a myriad of functionalized Grignard and organo-
4
zinc reagents have produced highly attractive, scalable methods.
1. tBuLi, TMEDA, THF
2. add to EtO2CCN
O
In the course of our medicinal chemical research program, we
desired a preparative method for the chiral, conformationally
restricted alcohol intermediate 3 to support advanced studies.
The following account details our efforts culminating in a scalable
synthetic route, as well as the observation of an unusual reaction
defying the expected reactivity pattern of a ketone.
O
O
45 %
O
1
2
Results and discussion
O
O
(
CH O) , cat. DBU,
2 n
OH dioxane
Our milligram-scale synthesis of alcohol 3 is shown in Scheme 1.
We selected phthalan (1) as our starting point, reasoning that use
O
35 %
(
56 % b.r.s.m)
3
⇑
Corresponding author. Tel.: +1 860 715 0089; fax: +1 860 441 4734.
E-mail address: benjamin.n.rocke@pfizer.com (B.N. Rocke).
Scheme 1. Milligram-scale synthesis of pharmaceutical building block 3.
0
040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2012.05.052

5
468
B. N. Rocke et al. / Tetrahedron Letters 53 (2012) 5467–5470
annulation
steps (selective tosylation of the primary alcohol, intramolecular
nucleophilic displacement, and selective mono-reduction of the
geminal diester), we opted rather to explore methods that could
circumvent formation of the lactone side product.
O
O
"
"
O
O
OH
O
+
We reasoned that a suitable annulation precursor could be de-
rived from benzaldehyde (10) by in situ protection using lithiated
O
O
O
0
2a
N,N,N -trimethylethylenediamine and subsequent directed ortho
lithiation. After treatment of this carbanion (11) with diethyl keto-
malonate and an acidic workup, we isolated a product nearly con-
3
4
5
reduction
1
sistent with structure 12 in terms of H NMR and mass spectra.
Figure 1. Annulation approach to compound 3.
However, we were surprised to observe UV
MeCN/H O) for this compound having only an unconjugated ben-
zene ring. Reduction of this intermediate using triethylsilane under
kmax = 276 nm
(
2
in the presence of a catalytic amount of DBU. Use of paraformalde-
hyde (vs gaseous or aqueous formaldehyde) provided a suitable
balance between the need for anhydrous conditions and ease of
handling. These conditions proved to be superior to traditional
methods involving stoichiometric enolates, which we found to
be highly sensitive to both time and temperature.
Whereas this synthesis was succinct, and alcohol 3 was ob-
tained in reasonable yield, two safety considerations precluded
its use on larger scale: (1) tert-butyllithium solution is pyrophoric
and (2) cyanide, a byproduct of addition to ethyl cyanoformate, is
highly acutely toxic. Furthermore, chromatographic purification of
ester 2 was difficult, and we observed it to undergo slow air
oxidation.
1
1
the action of boron trifluoride afforded a compound clearly
1
inconsistent with structure 13 by H NMR spectroscopy. Rather,
the analytical data were consistent with the isomer 15. Apparently,
1,4-addition of the aryllithium reagent 11 to diethyl ketomalonate and
subsequent intramolecular aldol reaction occurred, completely revers-
ing the traditional reactivity of diethyl ketomalonate. Instead of
hemiacetal 12, we obtained the similarly-behaved vinylogous
7
1
2
hemiacetal 14.
To further prove the structure of compound 14, we treated it
with hot aq HCl in dioxane. Ester saponification and decarboxyla-
tive elimination of the resulting b-hydroxy carboxylic acid afforded
benzofuran-2-carboxylic acid (16), which was spectroscopically
identical to an authentic sample.
Changing strategies, we anticipated that the 1,3-dihydro-iso-
benzofuran ring system could be constructed via annulation of
The unexpected reactivity of aryllithium reagent 11 toward
diethyl ketomalonate can be rationalized in these terms: Whereas
diethyl ketomalonate does not typically show nucleophilic suscep-
tibility on the oxygen atom of its central carbonyl (1,2-addition is
8
an appropriate toluene zwitterion synthon (5) with diethyl keto-
9
malonate (4) (Fig. 1). Inspired by a report from Ayers, we expected
that the resulting geminal diester could be selectively mono-re-
duced to install the desired b-hydroxy ester.
1
3,14
usually favored),
1,4-addition is thermodynamically feasible
1
5,16
In our first attempt to reduce this strategy to practice, directed
ortho lithiation of benzyl alcohol (6) was accomplished using nBu-
when other factors intervene.
We postulate that in this case,
the transition state leading to compound 12 or 14 is sterically
encumbered due to the combined presence of the diamine direct-
ing group, the neighboring lithium alkoxide moiety, and coordi-
nated solvent molecules. Kinetically, this would favor addition to
the ketone oxygen atom, which is significantly more exposed than
1
0
Li/TMEDA (Scheme 2). Treatment of the resulting carbanion (7)
with diethyl ketomalonate afforded predominately the undesired
lactone 8 along with the desired diol 9. Although we hypothesized
that compound 9 could be converted into compound 3 in three
OH
1
. nBuLi, TMEDA,
hexane
OLi
Li
2. diethylketomalonate,
THF
OH
O
+
O
OH
EtO C OH
2
EtO2C CO2Et
6
7
8 (major product)
9
OH
O
O
O
OH
CO Et
2
EtO C
EtO C
1
2
. LiN
CO2Et
2
2
EtO C
2
N
,
OLi
THF
12
13
3
3
. diethyl
O
. nBuLi
ketomalonate
N
Li
N
δ 3.80
BF ·OEt ,
3
3
2
24-38 %
Et SiH
CO Et
2
δ 5.93
CO Et
3
1 %
O
2
10
11
OH
1
5
CO Et
2
CO Et
2
aq. HCl,
dioxane
O
14
δ 7.02
CO H
2
44 %
O
16
Scheme 2. Unexpected 1,4-addition of aryllithium reagent 11 to diethyl ketomalonate observed during our initial attempts to access compound 3 via an annulation.

B. N. Rocke et al. / Tetrahedron Letters 53 (2012) 5467–5470
5469
Table 1
Reactivity of selected metalated benzaldehydes toward diethyl ketomalonate
1
1
. LiN
N
O
,
OH
. THF
CO Et
CO Et
2
2
R
R
2. metalation conditions
. diethyl ketomalonate
O
3
a
Entry
1
Bezaldehyde
Metalation conditions
Product
Yield /%
O
OH
2
2
a
a
29c
nBuLi (3 equiv), 24 h, ꢀ20 °C
CO Et
2
CO Et
2
O
OH
O
2
3
nBuLi (3 equiv), 48 h, ꢀ20 °C
CO
2
Et
33b
11
CO Et
2
O
Cl
O
Cl
OH
2
a
nBuLi (3 equiv), 2.5 h, ꢀ20 °C
nBuLi (3 equiv), 2.8 h, ꢀ20 °C
CO Et
2
CO Et
2
O
O
OH
2
a
CO Et
2
4
5
22
CO Et
2
O
Cl
Cl
O
OH
CO
2
a
2
Et
53b
nBuLi (3 equiv), 24 h, ꢀ20 °C
CO Et
2
O
t-BuO
HO
t-BuO
O
O
CO Et
2
6
PhLi (3 equiv), 7 h, rt19
14c
CO Et
2
O
OMe
a
b
c
Refers to isolated yields. As noted, some products contained small amounts of impurities derived from diethyl ketomalonate and not containing aromatic rings.
Product isolated in ca. 90% purity.
Product isolated in ca. 85% purity.
Cl
I
Cl
2. diethyl ketomalonate,
O
O
1
. iPrMgCl, THF
Δ
O
LiAl(OtBu) H, THF
3
OH
57-66 %
42 %
CO Et
O
2
EtO C
MgCl
2
17
18
13
Scheme 3. Scalable synthesis of compound 3.
3
the fully-substituted carbon atom.17 We also note precedence for
the 1,4-addition of organomagnesium, -aluminum, and -zinc re-
agents to select iminomalonate diesters.18
To explore the generality of this result, we applied these condi-
tions to a series of substituted benzaldehydes (Table 1). Accord-
ingly, we found the reaction to tolerate simple substitution at
any position (entries 2–5), as well as 2,3-ring fusion (entry 1). A
particularly high yield in the reaction of 4-tert-butoxybenzalde-
hyde and lower yields in the case of 2- and 4-chlorobenzaldehyde
suggest that electron donating substituents facilitate the reaction
18.21 Using diethyl ketomalonate as the electrophile, this strategy
quickly yielded positive results in the laboratory. Treatment of 2-
iodobenzyl chloride (17) with iPrMgCl at ꢀ5 °C afforded the re-
agent 18, which reacted smoothly with diethyl ketomalonate at
temperatures below ꢀ50 °C. Finally, overnight heating of the reac-
tion mixture induced cyclization, reliably affording compound 13
in acceptable yield (Scheme 3).
Reduction of compound 13 according to published procedure9
afforded a surprisingly low yield of alcohol 3, due to (1) incomplete
conversion of compound 13 and (2) formation of significant
quantities of compound 2 via retro-Claisen or retro-aldol side
reactions. Optimization of these conditions (equiv of reducing agent,
temperature, and time) led to a modified procedure whereby com-
pound 13 is treated with only 2 equiv of lithium tri-tert-butoxyalu-
minum hydride for a short period at reflux. Reduction according to
this procedure afforded compound 3 in 42% yield (Scheme 3).
(
entries 3–5). Interestingly, substrates bearing an ortho directing
group in the 3-position either failed to produce product (3-fluoro-
2
0
benzaldehyde ) or afforded the initially expected 1,2-addition
product in low yield (entry 6).
Still having not succeeded in our original goal, we focused our
efforts on a transformation of the functionalized Grignard reagent

5
470
B. N. Rocke et al. / Tetrahedron Letters 53 (2012) 5467–5470
1
0. Meyer, N.; Seebach, D. Angew. Chem., Int. Ed. Engl. 1978, 17, 521–522. For a
review of directed ortho metalation, see: Snieckus, V. Chem. Rev. 1990, 90, 879–
33.
Conclusion
We have developed a scalable synthesis of the pharmaceutical
9
11. The low yield observed in this reaction was due to formation of oligomer side
products, presumably through Friedel–Crafts alkylations.
1
building block 3 [ethyl 1-(hydroxymethyl)-1,3-dihydroisobenzofu-
ran-1-carboxylate] via annulation of a functionalized Grignard
reagent with diethyl ketomalonate and subsequent selective
mono-reduction of the intermediate geminal diester 13. In the
process, we encountered an unusual electrophilic oxygenation of
protected and ortho-lithiated benzaldehydes. In the future, this
reaction might provide orthogonal access to 2,2-disubstituted
2. LCMS and/or 1H NMR data for the crude product and all chromatography peaks
from the reaction 10 ? 14 failed to detect isomeric products.
3. For recent examples of 1,2-additions to diethyl ketomalonate, see: (a)
Sunazuka, T.; Handa, M.; Hirose, T.; Matsumaru, T.; Togashi, Y.; Nakamura,
K.; Iwai, Y.; O¯ mura, S. Tetrahedron Lett. 2007, 48, 5297–5300; (b) Okrasa, K.;
1
Levy, C.; Wilding, M.; Goodall, M.; Baudendistel, N.; Hauer, B.; Leys, D.;
Micklefield, J. Angew. Chem., Int. Ed. Engl. 2009, 48, 7691–7694; (c) Frings, M.;
Atodiresei, I.; Wang, Y.; Runsink, J.; Raabe, G.; Bolm, C. Chem. Eur. J. 2010, 16,
4577–4587.
2
2
23
dihydrobenzofurans and 2-carboxybenzofurans. Significantly,
it might also lead to a method for electrophilic oxygenation,
currently a limited but important transformation.24
1
1
4. In our hands, lithium, magnesium, and zinc derivatives of benzene all reacted
with diethyl ketomalonate in the typical 1,2-fashion. Furthermore, the adduct
of
lithiated
2-{[2-(dimethylamino)ethyl]methylamino}ethanol
and
phenyllithium also underwent 1,2-addition to diethyl ketomalonate.
5. The two electron-withdrawing groups on the central carbonyl of diethyl
ketomalonate attenuate the polarity of the carbonyl, rendering its oxygen atom
susceptible to nucleophilic attack. Atomic partial charge calculations according
to extended Hückel theory (CambridgeSoft Chem 3D Pro 12.0) for the carbonyl
in acetone are ꢀ0.52 (O) and +0.52 (C). In contrast, values for the central
carbonyl in diethyl ketomalonate are ꢀ0.39 (O) and +0.35 (C).
Acknowledgments
We thank Dr. Yong Tao for helpful discussions in the prosecu-
tion of this project. We are also grateful for help from Drs. David
Piotrowski and Vincent Mascitti in the preparation of this
manuscript.
16. In quantitative terms, the thermodynamic course of this reaction is a function
of the pK
the pK of the observed 1,4-addition product (malonate diester, pK
expected to be similar to the pK of the desired 1,2-addition product (alcohol,
pK estimates were based on predictions by ACD Labs pK DB 12.01
a
’s of the products obtained directly following addition. Accordingly,
a
a
ꢁ 12) is
a
Supplementary data
a
ꢁ 10). (pK
a
a
software for diethyl 2-phenoxymalonate and diethyl 2-hydroxy-2-
phenylmalonate.)
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.tetlet.2012.05.
1
1
1
7. Use of 1.5 or 1.1 equiv of nBuLi in the metalation step (vs 3 as described in Refs.
2a and 19) resulted in a corresponding decreases in the yield of compound 14;
no effect on the qualitative outcome of the reaction was observed.
8. (a) Niwa, Y.; Takayama, K.; Shimizu, M. Tetrahedron Lett. 2001, 42, 5473–5476;
0
52.
(
b) Calí, P.; Begtrup, M. Synthesis 2002, 63–66; (c) Niwa, Y.; Takayama, K.;
References and notes
Shimizu, M. Bull. Chem. Soc. Jpn. 2002, 75, 1819–1825.
9. For an example of metalation of 3-methoxybenzaldehyde, see: Comins, D. L.;
Brown, J. D. J. Org. Chem. 1989, 54, 3730–3732.
1.
2.
3.
4.
5.
For example, the lithium analog of reagent 18 must be prepared and handled at
100 °C. See Parham, W. E.; Jones, L. D.; Sayed, Y. A. J. Org. Chem. 1976, 41,
184–1186.
(a) Comins, D. L.; Brown, J. D. J. Org. Chem. 1984, 49, 1078–1083; (b)
Roschangar, F.; Brown, J. C.; Cooley, B. E., Jr.; Sharp, M. J.; Matsuoka, R. T.
Tetrahedron 2002, 58, 1657–1666.
ꢀ
20. For an example of metalation of 3-fluorobenzaldehyde, see University of
Pittsburg. Methods for preparation of camptothecin analogs having antitumor
activity. U.S. Patent 6,150,343 A, Nov 21, 2000.
21. Delacroix, T.; Bérillon, L.; Cahiez, G.; Knochel, P. J. Org. Chem. 2000, 65, 8108–
8110.
1
Recent developments include: (a) Lin, W.; Baron, O.; Knochel, P. Org. Lett. 2006,
22. 2,2-Disubstituted dihydrobenzofurans are currently difficult to access except
under transition metal or strong acidb catalysis. Recent examples include: (a)
Wang, X.; Lu, Y.; Dai, H.-X.; Yu, J.-Q. J. Am. Chem. Soc. 2010, 132, 12203–12205;
(b) Garcia, O.; Bofill, J. M.; Nicolas, E.; Albericio, F. Eur. J. Org. Chem. 2010, 3631–
3640.
23. Salicylaldehydes (vs benzaldehydes) are typical precursors to 2-carboxy-
benzofurans. Examples include: (a) Witiak, D. T.; Newman, H. A. I.; Poochikian,
G. K.; Fogt, S. W.; Baldwin, J. R.; Sober, C. L.; Feller, D. R. J. Med. Chem. 1978, 21,
833–837; (b) Coterón, J. M.; Catterick, D.; Castro, J.; Chaparro, M. J.; Díaz, B.;
Fernández, E.; Ferrer, S.; Gamo, F. J.; Gordo, M.; Gut, J.; De las Heras, L.; Legac, J.;
Macro, M.; Miguel, J.; Muñoz, V.; Porras, E.; de la Rosa, J. C.; Ruiz, J. R.; Sandoval,
E.; Ventosa, P. .; Rosenthal, P. J.; Fiandor, J. M. J. Med. Chem. 2010, 53, 6129–
6152; (c) Li, Q.-B.; Zhou, F.-T.; Liu, Z.-G.; Li, X.-F.; Zhu, W.-D.; Xie, J.-W. J. Org.
Chem. 2011, 76, 7222–7228.
a
8, 5673–5676; (b) Clososki, G. C.; Rohbogner, C. J.; Knochel, P. Angew. Chem., Int.
Ed. Engl. 2007, 46, 7681–7684.
Recent developments include: (a) Wunderlich, S. H.; Knochel, P. Angew. Chem.,
Int. Ed. 2007, 46, 7685–7688; (b) Mosrin, M.; Knochel, P. Org. Lett. 2009, 11,
1
837–1840.
(a) Coote, S. J.; Davies, S. G.; Middlemiss, D.; Naylor, A. J. Organomet. Chem.
989, 379, 81–88; (b) Brown, D. S.; Elliott, M. C.; Moody, C. J.; Mowlem, T. J. J.
1
Chem. Soc., Perkin Trans. 1 1995, 1137–1144; (c) Azzena, U.; Demartis, S.;
Melloni, G. J. Org. Chem. 1996, 61, 4913–4919; (d) Tomooka, K.; Wang, L.-F.;
Okazaki, F.; Nakai, T. Tetrahedron Lett. 2000, 41, 6121–6125; (e) Fort, Y.; Gros,
P.; Rodrigues, A. L. Tetrahedron Lett. 2002, 43, 4045–4048.
6
.
.
A recent report described an alternative approach to phthalan-1-carboxilic acid
esters: See Zhu, S.-F.; Song, X.-G.; Li, Y.; Cai, Y.; Zhou, Q.-L. J. Am. Chem. Soc.
2
010, 132, 16374–16376.
For examples of related enolate hydroxymethylations using strong bases, see:
a) Faissner, R.; Huttner, G.; Kaifer, E.; Rutsch, P. Eur. J. Inorg. Chem. 2003, 1681–
693; (b) Lukeman, M.; Scaiano, J. C. J. Am. Chem. Soc. 2005, 127, 7698–7699.
24. Electrophilic oxygenation of organometallic reagents is currently accessed via
MoOPH or N-sulfonyl oxaziridines,b or indirectly via oxidation of a boronate
a
7
c
(
1
ester. See: (a) Vedejs, E.; Engler, D. A.; Telschow, J. E. J. Org. Chem. 1978, 43,
188–196; (b) Davis, F. A.; Chem, B.-C. Chem. Rev. 1992, 92, 919–934; (c)
Fujikawa, N.; Ohta, T.; Yamaguchi, T.; Fukuda, T.; Ishibashi, F.; Iwao, M.
Tetrahedron 2006, 62, 594–604.
8
9
.
.
Corey, E. J.; Cheng, X.-M. The Logic of Chemical Synthesis; John Wiley & Sons:
New York, 1989.
Ayers, T. A. Tetrahedron Lett. 1999, 40, 5467–5470.