Synthesis of β-keto esters in-flow and rapid access to substituted pyrimidines

Article abstract of DOI:10.1021/jo101783m

We have developed an in-flow process for the synthesis of β-keto esters via the BF₃·OEt₂-catalyzed formal C-H insertion of ethyl diazoacetate into aldehydes. The β-keto esters were then condensed with a range of amidines to give a variety of 2,6-substituted pyrimidin-4-ols.

Full text of DOI:10.1021/jo101783m

pubs.acs.org/joc
asymmetric version has recently been reported4l using Sc(OTf)3
and a bis-N-oxide chrial ligand.
Synthesis of β-Keto Esters In-Flow and Rapid Access
to Substituted Pyrimidines
Although this approach constitutes an elegant method for
the synthesis of β-keto esters, the use of diazo compounds is
potentially hazardous and this is particularly an issue when
working on a larger scale. We considered that if it was pos-
sible to adapt this reaction for use “in-flow” this might offer
several potential benefits over the “batch” process as exo-
thermic processes, the release of gases, and the use of toxic or
hazardous reagents are well tolerated.5 To illustrate the
synthetic utility of the in-flow process, we planned to use
the resulting β-keto ester products to synthesize a range of
substituted pyrimidines.
Our initial efforts focused upon selecting a suitable Lewis
acid that could be used for catalyzing the addition of ethyl
diazoacetate to aldehydes (the Roskamp reaction) under
flow conditions. Roskamp originally reported that SnCl2
was the catalyst of choice under batch conditions, but he also
Hannah E. Bartrum,† David C. Blakemore,‡ Christopher
J. Moody,*,† and Christopher J. Hayes*,†
†School of Chemistry, University of Nottingham, University
Park, Nottingham, NG7 2RD, United Kingdom, and
‡Pfizer Global Research and Development, Ramsgate Road,
Sandwich, Kent, CT13 9NJ, United Kingdom
chris.hayes@nottingham.ac.uk; c.j.moody@nottingham.ac.uk
Received September 10, 2010
reported that BF3 OEt2 could be used, albeit with slightly
3
diminished yields. As the BF3 OEt2-catalyzed reaction re-
3
mains homogeneous throughout, we chose to use this as our
basis for the development of an in-flow process. We began
our studies by comparing the effectiveness of BF3 OEt2 and
3
SnCl2 as catalysts for the batch-wise addition of ethyl dia-
zoacetate (2) to hydrocinnamaldehyde (1), in order to select
an optimum catalyst loading for initial use under flow con-
ditions (Table 1).
This preliminary study showed that BF3 OEt2 was an
3
effective catalyst for the batch-wise production of the desired
β-keto ester product 3(76% yield at 10 mol % loading, entry 5),
and this was comparable to the results obtained using either
We have developed an in-flow process for the synthesis of
β-keto esters via the BF3 OEt2-catalyzed formal C-H
3
SnCl2 (83% yield at 10 mol % loading, entry 1) or SnCl2
2H2O (81% yield at 10 mol % loading, entry 3). Further-
3
insertion of ethyl diazoacetate into aldehydes. The β-keto
esters were then condensed with a range of amidines to
give a variety of 2,6-substituted pyrimidin-4-ols.
more, we found that the loading of BF3 OEt2 could be
3
lowered to 1 mol % (entry 6) without affecting either yield
or conversion. Lowering the loading of either SnCl2 (entry 2)
or SnCl2 2H2O (entry 4) to 1 mol % resulted in significant
3
reductions in yield of the β-keto ester 3.
We next investigated adapting this batch reaction into a
β-Keto esters are versatile intermediates in organic synthe-
sis, and have been extensively used in the construction of
natural products and other biologically active heterocyclic
molecules.1 As the number of commercially available func-
tionalized β-keto esters is limited, a number of methods have
been developed for their preparation.2 One of the simplest
approaches involves an acid-catalyzed formal C-H inser-
tion of ethyl diazoacetate into an aldehyde3 and since the
original report by Roskamp,3a various Lewis acids have been
flow process (Scheme 1). We used a Vaportec R2þ/R4 reactor
(4) (a) Nomura, K.; Iida, T.; Hori, K.; Yoshii, E. J. Org. Chem. 1994, 59,
488. (b) Dudley, M. E.; Morshed, M. M.; Brennan, C. L.; Islam, M. S.;
Ahmad, M. S.; Atuu, M.-R.; Branstetter, B.; Hossain, M. M. J. Org. Chem.
2004, 69, 7599. (c) Yadav, J. S.; Subba Reddy, B. V.; Eeshwaraiah, B.; Reddy,
P. N. Tetrahedron 2005, 61, 875. (d) Yadav, J. S.; Subba Reddy, B. V.;
Purnima, K. V.; Jhansi, S.; Nagaiah, K.; Lingaiah, N. Catal. Commun. 2008,
ꢀ
9, 2361. (e) Fructos, M. R.; Dıaz-Requejo, M.; Perez, P. J. Chem. Commun.
´
2009, 5153. (f) Jeyakumar, K.; Chand, D. K. Synthesis 2008, 11, 1685.
(g) Dhavale, D. D.; Patil, P. N.; Mali, R. S. J. Chem. Res., Synop. 1994, 152.
(h) Murata, H.; Ishitani, H.; Iwamoto, M. Tetrahedron Lett. 2008, 49, 4788.
(i) Sudrik, S. G.; Balaji, B. S.; Singh, A. P.; Mitra, R. B.; Sonawane, H. R.
Synlett 1996, 369. (j) Balaji, B. S.; Chanda, B. M. Tetrahedron 1998, 54,
13237. (k) Bandgar, B. P.; Pandit, S. S.; Sadavarte, V. S. Green Chem. 2001, 3,
247. (l) Li, W.; Wang, J.; Hu, X.; Shen, K.; Wang, W.; Chu, Y.; Lin, L.; Liu,
X.; Feng, X. J. Am. Chem. Soc. 2010, 132, 8532.
(5) (a) Baxendale, I. R.; Hayward, J. J.; Ley, S. V.; Tranmer, G. K.
ChemMedChem. 2007, 2, 768. (b) Ley, S. V.; Baxendale, I. R. Chimia 2008,
62, 162. (c) Ley, S. V.; Baxendale, I. R. In Systems Chemistry, Proceedings of
the Beilstein Bozen Symposium, Bozen, Italy, May 26-30th, 2008; pp
65-85. (d) Baxendale, I. R.; Ley, S. V.; Mansfield, A. C.; Smith, C. D.
Angew. Chem., Int. Ed. 2009, 48, 4017. (e) Wheeler, R. C.; Benali, O.; Deal,
M.; Farrant, E.; MacDonald, S. J. F.; Warrington, B. H. Org. Process Res.
Dev. 2007, 11, 704.
used (SnCl2,3 TiCl4,4a ZrCl4,4a BF3 OEt2,4b NbCl5,4c
3
Ag(TPA),4d [IPrAu(NCMe)]BF4,4e MoO2Cl2,4f activated alu-
mina,4g mesoporous silica,4h zeolites,4i,j clay4k) and the first
(1) (a) Hill, M. D.; Movassaghi, M. Chem.;Eur. J. 2008, 14, 6836.
(b) Lagoja, I. M. Chem. Biodiversity 2005, 2, 1. (c) Jain, K. S.; Chitre,
T. S.; Miniyar, P. B.; Kathiravan, M. K.; Bendre, V. S.; Veer, V. S.; Shahane,
S. R.; Shishoo, C. J. Curr. Sci. India 2006, 90, 793.
(2) Benetti, S.; Romagnoli, R.; De Risi, C.; Spalluto, G.; Zanirato, V.
Chem. Rev. 1995, 95, 1065.
(3) (a) Holmquist, C. R.; Roskamp, E. J. J. Org. Chem. 1989, 54, 3258.
(b) The formal C-H insertion of R-diazoketones into aldehydes has also
been reported: Padwa, A.; Hornbuckle, S. F.; Zhang, Z.; Zhi, L. J. Org.
Chem. 1990, 55, 5297.
8674 J. Org. Chem. 2010, 75, 8674–8676
Published on Web 11/17/2010
DOI: 10.1021/jo101783m
r
2010 American Chemical Society

Bartrum et al.
JOCNote
TABLE 1. Batch Optimization of Catalyst Loadinga
TABLE 2. In-Flow β-Keto Ester Synthesis and Subsequent Pyrimidine
Formation
entry
Lewis acid
mol %
yield (%)
1
2
3
4
5
6
SnCl2
SnCl2
SnCl2 2H2O
10
1
10
1
10
1
83
32
81
35
76
83
3
SnCl2 2H2O
BF3 OEt2
3
3
3
BF3 OEt2
aReactions performed at 0.125 mM in CH2Cl2 at 23 °C.
SCHEME 1. In-Flow Preparation of β-Keto Ester 3
entry
R1
R2 product yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
CH2CH2Ph
Me
c-Pr
Ph
SH
H
Me
c-Pr
Me
c-Pr
Ph
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
5n
5o
5p
5q
5r
5s
70a
67a
33a,c
63a,d
30a
CH2Ph
54a
54a
(CH2)5CH3
62a
67b
51b,c
33b
H
78b
controlled by Vaportec Flow Commander software in order
to allow optimization of the process. Hydrocinnamaldehyde
was once again used as our test substrate, employing 1 mol %
C6H11
Me
c-Pr
Me
Ph
Me
Me
Me
60a
CH(CH3)CH2CH2CH3
62b
50b,c
70b
of BF3 OEt2 catalyst, and we systematically varied reaction
3
CH2OBn
CH2CH2OBn
(CH2)3CtCH
parameters including temperature, flow rate, and back pres-
sure in order to find optimum conditions.
52a,e
63b
During optimization, we found that a back pressure of
390 psi was required on the system in order to minimize out-
gassing of the N2 generated inside the coil reactor. Under
these conditions 30 °C was found to be the optimum reactor
temperature (with 39 °C being the maximum) to ensure ef-
ficient conversion without exceeding the system pressure
limit. A residence time of 7.2 min (equating to a flow rate
of 0.139 mL/min from each of the two pumps) was found to
be optimal (>99% conversion at steady state), with shorter
residence times (i.e., ca. 5 min) leading to poorer conversions,
and longer residence times (i.e., 10 min) offering no further
advantage in conversion or isolated yield. This optimized
procedure gave high conversion and produced the desired
product 3 in good yield (74%) (Scheme 1), and furthermore it
could be applied to a wider range of aldehydes (Table 2).
Analysis of the crude reaction mixture by 1H NMR spec-
troscopy indicated that the β-keto ester was sufficiently pure
to be used in subsequent transformations without purifica-
tion. To demonstrate this, we treated the crude reaction mix-
ture with acetamidine hydrochloride6 (DBU, EtOH), which
provided the corresponding pyrimidin-4-ol7 5a in excellent
yield (70%, 2 steps) (Scheme 2). The 70% yield obtained over
the two steps compares favorably with the individual yields
obtained when the two separate steps were carried out on
purified material (Table 1, entry 6 and Scheme 2).
CH2CH(CH3)CH2CH2CHC(CH3)2 Me
35b,f
a1.05 equiv of amidine HCl and 2.0 equiv of DBU. b2.1 equiv of
3
amidine HCl and 4.0 equiv of DBU. cReaction refluxed for 18 h.
3
dNaOEt and thiourea. e10 mol % of BF3 OEt2 used. fModified flow pro-
cedure, see main text.
3
SCHEME 2. Synthesis of Pyrimidin-4-ol 5a from β-Keto Ester 3
investigated the two-step formation of a wider variety of
pyrimidinols from in-flow generated β-keto esters. Thus, a
range of aldehydes were converted in-flow into the corre-
sponding β-keto esters. To the output from the flow was
added the desired amidine hydrochloride and DBU in EtOH
in a modification of a literature procedure.6 The resulting
solution was stirred under Ar at rt for 18 h (formamidine,
acetamidine, and cyclopropylcarbamidine hydrochlorides)
or heated to reflux for 18 h (benzamidine hydrochloride).
The solvent was then removed in vacuo and the crude mate-
rial purified by column chromatography to give a range of
novel pyrimidinols in moderate to good yields over two steps
from the starting aldehydes (Table 2, Figure 1).
As the pyrimidine heterocycle is an important core struc-
ture in a range of biologically active molecules,1 we next
(6) Alker, D.; Campbell, S. F.; Cross, P. E.; Burges, R. A.; Carter, A. J.;
Gardiner, D. G. J. Med. Chem. 1989, 32, 2381.
(7) For the in-flow synthesis of a pyrimidine from an yne-one see:
Baxendale, I. R.; Schou, S. C.; Sedelmeier, J.; Ley, S. V. Chem.;Eur. J.
2010, 16, 89.
Once again, the yields obtained for the two-step process
(using acetamidine HCl) were in general comparable to the
3
yields of the intermediate β-keto esters when isolated and
J. Org. Chem. Vol. 75, No. 24, 2010 8675

Bartrum et al.
JOCNote
over the two steps for the 2-methyl- and 2-cyclopropyl-derived
pyridimidinols (e.g., entries 8 and 9). Reaction with benza-
midine required harsher conditions (reflux, 18 h) but gave
reasonable yields of the phenyl-substituted pyrimidinols
(e.g., entry 10). In addition to forming C2-alkyl- and aryl-
substituted pyrimidines, it was also possible to synthesize a
sulfur analogue (entry 4) in good yield by reacting thiourea
(NaOEt in EtOH) with an in-flow-generated β-keto ester.
When β-keto ester formation was attempted with a range
of aromatic aldehydes (e.g., benzaldehyde, p-anisaldehyde,
p-NO2-benzaldehyde, 2-furaldehyde, 2-pyridine carboxal-
dehyde) a mixture of products was obtained. This is perhaps
not surprising given that previous reports in the literature
have found that use of BF3 OEt2 as a catalyst for this
3
transformation gives low yields of the desired β-keto esters
in addition to significant amounts of 2-aryl-3-hydroxy-
2-acrylic esters.4b
In conclusion, we have synthesized a variety of novel 2,6-
substituted pyrimidin-4-ols utilizing a two-step procedure
involving preparation of intermediate β-keto esters in flow,
followed by subsequent condensation reactions with a range
of amidines.
Experimental Section
Representative Procedure for the Synthesis of Pyrimidin-4-ols.
A solution of hydrocinnamaldehyde (1) (0.500 mmol, 0.25 M)
and BF3 OEt2 (1 mol %) in CH2Cl2 (2 mL) was injected into one
3
of the sample loops of the R2þ unit. The other sample loop was
loaded with a solution of ethyldiazoacetate (2) (0.505 mmol) in
CH2Cl2 (2 mL). The valves of the loop were set to load and the
reagents pumped through the system using CH2Cl2 as a system
solvent at a flow rate of 0.139 mL/min per pump. The reagents
combined in a T-piece before entering a 2 mL coil reactor (inside dia-
meter =1 mm, length =2.9 m) (PFA), which was maintained at
30 °C by the R4 unit. A back pressure regulator (390 Psi) was
added in line after the reactor to ensure the nitrogen produced
during the reaction stayed in solution. The output (8 mL total
volume) was collected directly into a round-bottomed flask. To the
output from the flow reaction (crude β-ketoester 3 in 8 mL of
CH2Cl2) was added EtOH (8 mL), acetamidine hydrochloride
(52 mg, 1.10 equiv), and DBU (150 μL, 2.00 equiv). The resulting
solution was stirred under Ar at rt for 48 h, after which time the
solvent was removed in vacuo and the crude material purified by
column chromatography (2% MeOH/CH2Cl2) to give the py-
rimidinol 5a (74 mg, 70%) as colorless needles, mp 126-127 °C
(lit.10 mp 125-127 °C); found C 72.7, H 6.6, N 13.1, C13H14N2O
requires C 72.9, H 6.6, N 13.1; Rf 0.6 (10% MeOH/CH2Cl2);
FIGURE 1. Structures of pyrimidin-4-ol products.
therefore the yield over two steps for these derivatives gives
an indication of the efficiency of the β-keto ester formation.
Primary and secondary aldehydes were well tolerated by
this two-step procedure and gave reasonable yields of a
variety of pyrimidinols (entries 1-15). Pleasingly the proce-
dure was compatible with the presence of oxygen function-
ality in the side chain of the aldehyde (entries 16 and 17). The
flow process also tolerated the presence of an alkyne moiety
(entry 18) or alkene (entry 19). Although the yield for the
reaction with citronellal (entry 19) is relatively low, it is an
encouraging result given the propensity for the aldehyde to
undergo a cyclization in the presence of acid.8 For this ex-
ample we utilized a modified flow procedure, which involved
delivery of the aldehyde and ethyl diazoacetate (as a solution
in CH2Cl2) from one R2þ injection loop, and delivery of
ν
max(CHCl3)/cm-1 3011, 2865, 2760, 1663, 1606, 1562, 1497, 1455,
1383, 1304, 1184, 967, and 856; δH (400 MHz; CDCl3) 13.13 (1H,
br s), 7.31-7.26 (2H, m), 7.22-7.17 (3H, m), 6.12 (1H, s),
3.02-2.96 (2H, m), 2.86-2.81 (2H, m), 2.48 (3H, s); δC (100
MHz; CDCl3) 169.5, 166.0, 158.7, 140.8, 128.5, 128.4, 126.3, 109.6,
39.4, 34.1, 21.7; m/z(ESI) found 237.1001 (M þ Na C13H14N2NaO
requires 237.0998).11
10 mol % BF3 OEt2 (in CH2Cl2) from the other.9 This pro-
3
cedure avoided exposing citronellal to the Lewis acid prior to
combination with ethyl diazoacetate, and led to isolation of
the pyrimidinol 5s in 35% yield (entry 19) after the hetero-
cycle formation step. In general, good yields were obtained
(8) Clark, B. C., Jr.; Chamblee, T. S.; Iacobucci, G. A. J. Org. Chem.
1984, 49, 4557.
Acknowledgment. We thank the EPSRC (EP/G027919/1)
and Pfizer for financial support of this work.
(9) This procedure could be used in any situation where the aldehyde is
sensitive to being in contact with a Lewis acid for prolonged periods of time,
but it offers no advantage to the standard protocol described in Scheme 1 and
Table 2 for examples where the aldehyde and Lewis acid are compatible.
(10) Murray, T. P.; Hay, J. V.; Portlock, D. E.; Wolfe, J. F. J. Org. Chem.
1974, 39, 595.
Supporting Information Available: Detailed experimental
procedures, spectroscopic data, and copies of 1H NMR spectra
for all new compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
(11) Sakamoto, T.; Yoshizawa, H.; Yamanaka, H. Chem. Pharm. Bull.
1984, 32, 2005.
8676 J. Org. Chem. Vol. 75, No. 24, 2010