Sonogashira reactions with propyne: Facile synthesis of 4-Hydroxy-2-methylbenzofurans from Iodoresorcinols

Article abstract of DOI:10.1021/op900263b

The Sonogashira reaction of terminal alkynes and ortho-halophe- nols with subsequent cyclization is a well-precedented method for the synthesis of substituted benzofurans. Here we describe the extension of this method to the coupling of 2-iodoresorcinols and terminal alkynes, including propyne, to give 4-hydroxy-2-substi- tuted benzofurans. In particular, we describe the screening, method development, and scaleup of the reaction with propyne using standard hydrogenation equipment.

Full text of DOI:10.1021/op900263b

Organic Process Research & Development 2010, 14, 180–187
Sonogashira Reactions with Propyne: Facile Synthesis of
4-Hydroxy-2-methylbenzofurans from Iodoresorcinols
Martin A. Berliner,†,* Eric M. Cordi,‡ Joshua R. Dunetz,† and Kristin E. Price‡,*
Research-API and Technology-API, Pharmaceutical Sciences, Pfizer Global Research and DeVelopment, 558 Eastern Point
Road, Groton, Connecticut 06340
Abstract:
The Sonogashira reaction of terminal alkynes and ortho-halophe-
nols with subsequent cyclization is a well-precedented method for
the synthesis of substituted benzofurans. Here we describe the
Figure 1. 4-Hydroxy-6-methoxycarbonyl-2-methylbenzofuran.
extension of this method to the coupling of 2-iodoresorcinols and
terminal alkynes, including propyne, to give 4-hydroxy-2-substi-
tuted benzofurans. In particular, we describe the screening,
method development, and scaleup of the reaction with propyne
using standard hydrogenation equipment.
zofuranones4 and phenol generation via Baeyer-Villiger oxida-
tion of a 4-carboethoxybenzofuran.5
Among general approaches to benzofurans, we were initially
most interested in well-documented furan annulation methods
involving oxidative cyclization of allylphenols.6 However, we
quickly determined that synthesis of the requisite allylresorcinol
5 via several different methods7 (Scheme 2) was not amenable
to scaleup. Suzuki coupling of the resorcinol bromide and
allylpinacol boronate proved to be low yielding, a direct
alkylation approach was ineffective and required protection of
the phenols, and Claisen rearrangement has previously been
established to proceed with the undesired regioselectivity.7c
These results prompted us to reconsider our synthetic strategy
and evaluate a Sonogashira-based process to access 1.
The coupling of terminal alkynes with aryl halides in the
presence of catalytic transition metal salts (Sonogashira reaction
and variants) has proven to be a powerful method for the
construction of carbon-carbon bonds. The scope and applica-
tion of the Sonogashira reaction has been the subject of several
recent reviews8 and its versatility has been demonstrated in
numerous large-scale applications.9 The use of Sonogashira
reactions in the synthesis of benzofurans and their precursors
has been demonstrated,10 but there are few examples of
Sonogashira reactions with resorcinol derivatives, and these
employ fully protected substrates and do not form benzofurans
Introduction
For a recent project, we had need to prepare kilogram
quantities of the 4-hydroxy-2-methylbenzofuran-6-carboxylic
acid ester 1 (Figure 1) to support drug discovery efforts. Initially,
our medicinal chemistry colleagues employed the existing three
step literature synthesis of 1-Et (the related ethyl ester) to
generate material (Scheme 1).1 Stobbe condensation of 5-me-
thylfurfural with diethyl succinate provides acid 2, which is
cyclized under Friedel-Crafts conditions to give protected
benzofuran 3. Deprotection of 3 provides the target benzofuran.
This procedure was sufficiently robust to enable delivery of lots
of up to 100 g of final product, but variable and modest yields
of the Stobbe reaction (40-60%) and complicated workup and
isolation2 led us to investigate other routes. The patent literature3
offers an alternative synthesis of benzofuran 1 that differs only
in the synthesis of ester acid 2, which is obtained in two steps
by reaction of 5-methylfurfural with phosphorane 4 followed
by tert-butyl ester deprotection. This process offers no practical
improvement over the medicinal chemistry synthesis due to an
increase in number of steps, the need to synthesize 4, and poor
atom economy from triphenylphosphine-derived residues. Other
reported syntheses of 4-hydroxybenzofurans rely upon arene
oxidation; examples include the aromatization of tetrahydroben-
(4) (a) Yamaguchi, S.; Yamamoto, K.; Ueda, T.; Morikawa, T.; Kawase,
Y. Bull. Chem. Soc. Jpn. 1989, 62, 4066–8. (b) Mphahlele, M. J.;
Moekwa, T. B. Org. Biomol. Chem. 2005, 3, 2469–2475.
(5) Sun, L.; Takaki, K.; Chen, J.; Iben, L.; Knipe, J. O.; Pajor, L.; Mahle,
C. D.; Ryan, E.; Xu, C. Bioorg. Med. Chem. Lett. 2004, 14, 5157–
5160.
(6) For example, see: (a) Mitsudome, T.; Umetani, T.; Nosaka, N.; Mori,
K.; Mizugaki, M.; Ebitani, K.; Kaneda, K. Angew. Chem., Int. Ed.
2006, 45, 481–485. (b) Yadav, A. K.; Singh, B. K.; Singh, N.; Tripathi,
R. P. Tetrahedron Lett. 2007, 48, 6628–6632.
* To whom correspondence should be addressed. E-mail: (M.A.B.) martin.
a.berliner@pfizer.com; (K.E.P.) kristin.e.price@pfizer.com.
† Research-API.
(7) For representative Suzuki reactions of allylboronic acids and boronate
esters, see: (a) Kotha, S.; Behera, M.; Shah, V. R. Synlett 2005, 1877.
(b) Fu¨rstner, A.; Seidel, G. Synlett 1998, 161. Claisen rearrangement: (c)
Occielli, E.; DePaoli, A; Nathansohn, G. Gazz. Chim. Ital. 1981, 111,
383.
‡ Technology-API.
(1) Simoni, D.; Romagnoli, R.; Baruchello, R.; Rondanin, R.; Rizzi, M.;
Pavani, M. G.; Alloatti, D.; Giannini, G.; Marcellini, M.; Riccioni,
T.; Castorina, M.; Guglielmi, M. B.; Bucci, F.; Carminati, P.; Pisano,
C. J. Med. Chem. 2006, 49, 3143–3152.
(8) Recent reviews: (a) Chinchilla, R.; Na´jera, C. Chem. ReV. 2007, 107,
874–922. (b) Pleino, N. Angew. Chem., Int. Ed. 2008, 47, 6954–6956.
(c) Doucet, H.; Hierso, J.-C. Angew. Chem., Int. Ed. 2007, 46, 834–
871. (d) Bunz, U. H. F. Chem. ReV. 2000, 100, 1605–1644. (e)
Sonogashira, K. J. Organomet. Chem. 2002, 653, 46–49.
(2) Complications were caused by the need to purge unreacted and
polymerized succinate residues.
(3) McKerrecher, D.; Rayner, J. W. WO Patent 046139 A1, 2004.
180
•
Vol. 14, No. 1, 2010 / Organic Process Research & Development
10.1021/op900263b  2010 American Chemical Society
Published on Web 12/10/2009

Scheme 1. Previous syntheses of benzofuran 1 Me and Et esters
Scheme 2. Attempted synthesis of 1 from allylresorcinols
Scheme 3. Synthesis of a 4-hydroxy-2-methylbenzofuran via
a sonogashira reaction
symmetric 2-haloresorcinol-5-carboxylic esters and propyne gas
(Scheme 3).12 Only a single example of benzofuran synthesis
using propyne has been published,13 which is surprising given
the prevalence of 2-methylbenzofurans in the medicinal chem-
istry literature.14 In this paper, we describe the development
and scaleup of a Sonogashira reaction with propyne to
synthesize a 4-hydroxy-2-methylbenzofuran using hydrogena-
tion screening equipment typically found in process develop-
ment laboratories.
Results and Discussion
as terminal products.11 A Sonogashira approach to 4-hydroxy-
2-methylbenzofuran 1 would offer the advantage of brevity as
the synthesis would be a single step from readily available,
When we started this work there were no reports in the
literature describing Sonogashira reactions between alkynes and
unprotected 2-halo-1,3-resorcinols. Therefore, our initial experi-
ments were designed to evaluate this reactivity. We initially
prepared several esters of 2-bromo- and 2-iodoresorcinols15 and
evaluated these in the Sonogashira reaction with methyl
propargyl ether and propargyl alcohol as test substrates (Scheme
4). Under standard reaction conditions,16 we obtained benzo-
furan products in good yields in reactions with iodoresorcinols;
the bromoresorcinols were recovered unchanged.
The first reactions with propyne were conducted as single
experiments in sealed glass pressure vessels. Excess propyne
(3-5 equiv) was measured by condensing the gas in a precooled
(-78 °C) graduated cylinder and using a cannula to transfer
(9) (a) Houpis, I. N.; Shilds, D.; Nettekoven, U.; Schnyder, A.; Bappert,
E.; Weerts, K.; Canters, M.; Vermuelen, W. Org. Process Res. DeV.
2009, 13, 598–606. (b) Li, H.; Xia, Z.; Chen, S.; Koya, K.; Ono, M.;
Sun, L. Org. Process Res. DeV. 2007, 11, 246–250. (c) Dorow, R. L.;
Herrinton, P. M.; Hohler, R. A.; Maloney, M. T.; Mauragis, M. A.;
McGhee, W. E.; Moeslein, J. A.; Strohbach, J. W.; Veley, M. F. Org.
Process Res. DeV. 2006, 10, 493–499. (d) Ripin, D. H. B.; Bourassa,
D. E.; Brandt, T.; Castaldi, M. J.; Frost, H. N.; Hawkins, J.; Johnson,
P. J.; Massett, S. S.; Neumann, K.; Phillips, J.; Raggon, J. W.; Rose,
P. R.; Rutherford, J. L.; Sitter, B.; Stewart, A. M., III; Vetelino, M. G.;
Wei, L. Org. Process Res. DeV. 2005, 9, 440–450.
(10) (a) Kundu, N. G.; Pal, M.; Mahanty, J. S.; De, M. J. Chem. Soc.,
Perkin Trans 1 1997, 2815–2820. (b) Bates, C. G.; Saejueng, P.;
Murphy, J. M.; Venkataraman, D. Org. Lett. 2004, 6, 1441–1444. (c)
Bang, H. B.; Han, S. Y.; Choi, D. H.; Yang, D. M.; Hwang, J. W.;
Lee, H. S.; Jun, J.-G. Synth. Commun. 2009, 39, 506–515. (d) Grigg,
R.; Sridharan, V.; Sykes, D. A. Tetrahedron 2008, 64, 8952–8962.
(e) Khan, M. W.; Alam, M. J.; Rashid, M. A.; Chowdhury, R. Bioorg.
Med. Chem. Lett. 2005, 13, 4796–4805.
(13) (a) Hong, Y.; Kania, R. S. U.S. Patent 0137395 A1, 2005. One use of
propyne in the synthesis of indoles via a Sonogashira-cyclization
reaction has been reported: (b) Xu, L.; Lewis, I. R.; Davidsen, S. K.;
Summers, J. B. Tetrahedron Lett. 1998, 39, 5159–5162.
(14) A SciFinder search revealed over 2000 unique 2-methylbenzofuran-
containing compounds prepared over the past 30 years and one
compound (trioxysalen) in clinical use.
(15) Bromoresorcinol derivatives were prepared by esterification (ROH,
H2SO4) of commercial 4-bromo-3,5-dihydroxybenzoic acid (Sigma-
Aldrich). Iodoresorcinols were prepared by iodination of commercial
methyl 3,5-dihydroxybenzoic acid with I2. See the experimental section
for details.
(11) (a) Gericke, K. M.; Chai, D. I.; Lautens, M. Tetrahedron 2008, 64,
6002–6014. (b) Hennrich, G.; Omenat, A.; Asselberghs, I.; Foerier,
S.; Clays, K.; Verbiest, T.; Serrano, J. L. Angew. Chem., Int. Ed. 2006,
45, 4203–4206. (c) Hauss, F.; Liu, J.; Michelucci, A.; Coowar, D.;
Morga, E.; Heuschling, P.; Luu, B. Bioorg. Med. Chem. Lett. 2007,
17, 4218–4222.
(12) When investigating a Sonogashira approach to 1, our initial concerns
centered on the cost and handling of propyne. While somewhat costly
on research scale, propyne (which is isolated in pure form by
distillation of MAPP gas) is readily available on scale from several
suppliers at low cost (less than $150/kg or $6/mol). MAPP gas is
composed of propyne, allene, propane, and propene.
(16) 5% Pd(Ph3P)2Cl2 and 5% CuI with i-Pr2NEt as base in DME was
chosen as a reference starting point for reaction exploration based on
literature precedent.
Vol. 14, No. 1, 2010 / Organic Process Research & Development
•
181

Scheme 4. Initial synthesis of 4-hydroxy-2-alkylbenzofurans
benzofuran formation. Solvents were chosen with a range of
polarities and to substitute for DME. A series of secondary and
tertiary amines were chosen to cover a range of steric bulk,
basicity, and nucleophilicity. A representative selection of the
bases and solvents screened can be found in Figure 2. Each
reaction was analyzed based on the relative amount of starting
iodide, desiodide, product, and related impurities20 as measured
by HPLC. The reaction run with dimethylacetamide (DMAc)
and diisopropylamine (i-Pr2NH) provided the highest yield and
purity.
the liquid to a cooled reaction vessel, which was then sealed
and heated. While labor intensive and not readily scalable, these
first experiments demonstrated that the reaction worked well
with propyne. The reaction proceeded at elevated temperature
(60 °C), reaching completion in less than four hours, and at
ambient temperature, reaching completion within 72 h. Yields
in these reactions were similar although the reaction at lower
temperature provided a higher purity product. On the basis of
these proof-of-concept experiments, we decided to initiate a
more extensive screen of reaction conditions. Since downstream
ester hydrolysis rendered identity of the ester unimportant, we
chose the 4-iodoresorcinol methyl ester (6) as the substrate for
additional optimization due to the ease of its synthesis and
crystallization.
Reaction Screening. Sonogashira reactions with propyne
have been reported using subsurface gaseous delivery at ambient
pressure17 (bubbling propyne through the reaction mixture),
liquid delivery18 (condensing propyne at cryogenic temperatures
and transferring it as a liquid) or in situ propyne formation.19
From a process development perspective, we elected to pursue
a different approach and employ hydrogenation screening
equipment to evaluate this chemistry. The use of pressure
reactors has several advantages, including convenient gas
metering and delivery on systems designed for low-volume
parallel screens (Biotage Endeavor) and lab scale preparative
runs (Biotage Atlantis and Parr reactors). A further advantage
is that the effects of propyne pressure and reaction temperature
can be explored while avoiding the handling challenges and
reproducibility problems associated with sparging and cryo-
loading techniques.
Given the high conversion and purity of the DMAc/i-Pr2NH
conditions, further small scale screening was carried out to
determine the most appropriate temperature and pressure of
propyne for this reaction. The proof-of-concept experiments
illustrated that temperatures above 50 °C increased byproduct
formation. For this reason, the temperature range in the screen
was kept between ambient and 45 °C. Pressure was varied
between 1 and 15 psig.21 The best product distribution was
observed at 35 °C and 10 psig (Figure 3).
Propyne Solubility Calculations. On the basis of the results
from these screening experiments, we sought further insight into
the relative solubility of propyne as a contributing factor in
reaction outcome. Accurate prediction of the vapor-liquid
equilibrium (VLE) of nonideal solutions (and therefore solubil-
ity) requires composition-dependent correlation for activity
coefficients (gammas) for each component of a mixture. To
estimate activity coefficients for propyne and each organic
solvent of interest in binary mixtures,22 we employed
COSMOtherm23 quantum chemical calculations of chemical
potential based on screening charge density of components.24
Pure component vapor pressure data and mixture activity
coefficients were paired in Aspen Plus25 to calculate propyne
solubility in each solvent of interest as a function of pressure
and temperature. The calculated equilibrium weight percent
concentration of propyne in each of seven solvents from 15 to
(20) Primary resorcinol-based impurities are enyne products derived from
alkynylation of the intermediate alkynylresorcinol prior to cyclization.
These include adduct i and lesser amounts of higher order enynes ii.
The first reaction parameters examined were solvent and
base. Bis(triphenylphosphine)palladium(II) dichloride and cop-
per iodide were used in all reactions, as they are competent
catalysts, readily available, and relatively inexpensive. As the
total iodoresorcinol loading was 200 mg for each of these
reactions, catalyst loading was not evaluated on this scale but
held at 5 mol % for both Pd and Cu. The reactions with varied
solvent and bases were carried out for 24 h at 25 °C and 5 psig
propyne. Under these conditions, it was expected that the proof-
of-concept conditions should show partial conversion, providing
an opportunity to find faster and/or cleaner conditions for
(21) When running on the Endeavor system, 100 g tanks of propyne were
attached to the secondary gas port. Each reaction was purged with N2
and then placed under a propyne blanket. No propyne purges were
carried out.
(22) Temperature-dependent activity coefficient data are frequently obtained
from binary VLE measurements. Measurement of VLE data in the
laboratory can be time consuming; therefore, computational methods
accounting for functional group contributions or molecular interaction
potential are used to estimate composition and temperature dependent
activities of components in a mixture of interest.
(17) (a) Fu¨rstner, A.; Mamane, V.; Seidel, G.; Laurich, D. Org. Synth. 2006,
83, 103–110. (b) Ritchey, R. N.; Yu, H. Org. Process Res. DeV. 2009,
13, 315–320.
(23) COSMOthermX, Version C21_0108, software for technical computa-
tion; COSMOlogic GmbH & Co. KG: Leverkusen, Germany, 2008.
(24) The non-random two-liquid (NRTL) solution model was chosen as a
descriptor for mixture activity coefficients. COSMOtherm data was
regressed into NRTL binary interaction parameters for each solvent
pairing with propyne, which were then transferred into an Aspen Plus
databank. Reference for NRTL: Seader, J. D.; Henley, E. J. Separation
Process Principles; Wiley: New York, 2006; pp 55-56.
(18) (a) Kim, J. T.; Butt, J.; Gevorgyan, V. J. Org. Chem. 2004, 69, 5638–
5645. (b) Nelson, M. L.; Ismail, M. Y.; McIntyre, L.; Bhatia, B.; Viski,
P.; Rennie, G.; Andorsky, D.; Messersmith, D.; Stapleton, K.;
Dumornay, J.; Sheahan, P.; Verma, A. K.; Warchol, T.; Levy, S. B.
J. Org. Chem. 2003, 68, 5838–5851.
(19) (a) Abraham, E.; Suffert, J. Synlett 2002, 328. (b) Rudyanto, M.;
Tomizawa, Y.; Morita, H.; Honda, T. Org. Lett. 2008, 10, 1921–1922.
(25) Aspen Plus, version 7.1 software for technical computation; Aspen
Technology: Burlington, MA, 2008.
182
•
Vol. 14, No. 1, 2010 / Organic Process Research & Development

Figure 2. Product distribution as a function of solvent and base. All reactions run with 200 mg of iodoresorcinol 6, 3 equiv base,
10 mL/g solvent, and 5 mol % of both catalysts at 25 °C under 5 psig propyne for 24 h.
50 °C at 10 psig is shown in Figure 4.26 Although calculated
solubility is higher than measured solubility for our system (vide
infra), the modeled data is consistent with the measured product
distribution in screening reactions (Figure 2), with the highest
calculated concentrations of propyne correlating to the highest
yielding reactions.
Scaleup of Reaction and Propyne Delivery. Once condi-
tions had been identified on small scale, we scaled up to
multigram runs. It became apparent immediately, through
variable reaction outcomes, that the development of a robust
method for propyne purging and pressurization was essential
to maximize propyne content in the reaction vessel. Conse-
quently, it was necessary to determine the number of propyne
purges needed to saturate the reaction mixture. Propyne was
charged to ∼400 mL of DMAc in a 600 mL Parr reactor.27
The reactor was weighed after each propyne purge cycle and
the solution was found to be ∼13% by weight propyne at
30 °C at ambient pressure (0 psig) after 4 cycles. Degassing is
audible (and visible on the pressure gauge) when the solution
is saturated with propyne. Given this result, the finalized purging
protocol was as follows: each reaction mixture was purged three
times with nitrogen without stirring at 50 psig (to remove O2
from the headspace), three times with nitrogen with stirring at
50 psig (to remove O2 from solution), and then four times with
propyne with stirring at 10 psig. Each propyne pressurization
takes a significant amount of time as the propyne largely
dissolves into the DMAc solution. The dissolution of propyne
is exothermic and the reactor temperature was controlled to
maintain an internal temperature between 25 and 35 °C. Because
(26) Complete descriptions and data on solubility calculations can be found
in the supporting information.
(27) The 600 mL Parr reactor used for these experiments has a dip tube
for subsurface gas delivery.
Vol. 14, No. 1, 2010 / Organic Process Research & Development
•
183

With the propyne purge protocol, it was possible to conduct
further small scale reactions (5-10 g) to examine the effect of
concentration and catalyst loading. The amount of solvent was
found to be crucial to high conversion, presumably since
lowering the amount of DMAc also decreases the amount of
propyne that can be dissolved for the reaction. As a result, we
continued to run the reaction with 10 volumes of DMAc. The
catalyst loading could be reduced to 3 mol % of both the Pd
and Cu catalysts when sufficiently pure 6 was used. However,
our bulk lot contained small amounts of iodine and necessitated
the use of 5 mol % catalyst to ensure adequate reaction
performance. Under these conditions, we observed typical
conversions of greater than 90% with formation of minimal
des-iodide byproduct on scales ranging from 5 g to 1.67 kg.
Workup Development. During initial development, stan-
dard aqueous workups were employed, requiring high volumes
and numerous washes to fully extract benzofuran 1 from the
aqueous mixtures. The crude product, isolated as a brown-black
semisolid of modest purity, was chromatographed on silica gel
and treated with thiol-modified silica gel to reduce Pd below
300 ppm.29 While serviceable for small-scale experiments, this
workup and isolation procedure was not suitable for larger-
scale work. Several factors complicated efforts to develop an
efficient isolation for 1, most notably the presence of propyne
gas30 and related polymers and oligomers, excess diisopropy-
lamine, as well as metal salts and aggregates.
Figure 3. Product distribution as a function of temperature
and pressure. All reactions run with 200 mg of 6, 3 equiv
diisopropylamine, 10 mL/g DMAc, and 5 mol % of both
catalysts for 24 h. Legend: black, 1; white, 6; light grey, des-
iodide; dark grey, byproduct.
To remove the high concentrations of i-Pr2NH and any
related salts, aqueous ammonium chloride31 was introduced as
the initial diluent following the reaction. Fortuitously, the
addition of water and aqueous NH4Cl expedited the degassing
of the remaining propyne in the reaction mixture.32 The removal
of the residual propyne decreased the overall volume signifi-
cantly and caused precipitation of dark-colored low-density
materials, presumably propyne polymers and oligomers.
To remove these polymeric materials and lower molecular
weight oligomers, the aqueous solution was treated with two
nonpolar washes (1% ethyl acetate in heptane).33 These washes
were performed at slightly elevated temperatures (40-50 °C)
to encourage dissolution of the nonpolar materials and prevent
crystallization of crude product from the aqueous phase.34
A
thick, dark rag layer of nonpolar polymeric material formed as
the phases separated and was discarded with the heptane. Any
residual polymer adhering to the tank walls or impeller
following the wash was removed with warm acetone. The
(28) The mass flow device installed in our kilo facility pressure reactor is
calibrated for hydrogen delivery and provided incorrect information
during propyne charging operations.
Figure 4. Calculated propyne solubility in organic solvents at
10 psig (weight percent propyne vs temperature).
(29) Palladium levels dropped somewhat with silica gel chromatography
(3% after aqueous work up to 0.5-1.0% after chromatography).
(30) On the basis of our dissolution studies, we anticipated that propyne
concentrations could be as high as 13 wt%.
(31) Water and saturated ammonium chloride were introduced in a 1:1
volume ratio. On scale, water was added first followed by the
ammonium chloride solution.
(32) As a result of the rapid degassing, when running on larger scale, the
water and ammonium chloride solution should be added slowly with
a N2 purge. Both the degassing and the heat increase related to the
dissolution of DMAc in water were roughly dose controlled.
(33) This solvent system was chosen to maximize extraction of propyne
related impurities and to minimize loss of benzofuran 1 from the
aqueous phase.
of the low pressure of the propyne source tank (60 psig at
20 °C), certain accommodations need to be made. In reactors
without automatic valve closures, the tank should be closed off
after purging so that the reaction mixture does not flow back
into the propyne regulator and tank as the pressure drops. The
low delivery pressure and high solubility of propyne in many
organic solvents make the traditional uptake data provided by
hydrogenation equipment and linear flow rates for delivery less
valuable. We found on kilogram scale that the easiest way to
measure propyne usage was to position the tank on a balance
to monitor the change in mass over time.28
(34) If the aqueous solution is stored at or below 20 °C, the product will
crystallize and hold on to a large amount of dark nonpolar impurities.
184
•
Vol. 14, No. 1, 2010 / Organic Process Research & Development

Table 1. Reaction of alkynes with iodoresorcinol 6
hydrogenation screening equipment. A combination of experi-
mental and computational results allowed us to develop an initial
understanding of key solubility properties of propyne in
common organic solvents and identified N,N-dimethylacetamide
as a particularly good solvent for reactions with propyne. We
developed a workup process that efficiently purged the metal-
rich propyne-derived polymers and oligomers, obviating the
need for chromatography and scavenging agents to meet purity
targets. This work supported the successful scaleup of a
Sonogashira reaction with propyne on kilo scale.
Experimental Section
General Experimental. Research samples of propyne were
purchased from Sigma-Aldrich and Advanced Gas Technolo-
gies, Inc. at 98% purity and used as received. Other reagents
and solvents were obtained from reputable suppliers and used
without further purification. 1H NMR were obtained at 400 MHz
and 13C NMR spectra were acquired at 100 MHz in DMSO-d6
unless otherwise noted. HPLC data were collected on an Agilent
1100 Series with a diode array detector, using an Agilent Zorbax
SB-CN column (4.6 × 150 mm) with a 10 min method (10:90
CH3CN:0.2% HClO4/water, 2 min isocratic, 6 min ramp to 90:
10 CH3CN:aqueous, then 2 min isocratic at 90:10) at 30 °C
and 2 mL/min flow rate. Propyne, like all liquefied petroleum
gases, is extremely flammable and forms explosive mixtures
in air. All users should thoroughly familiarize themselves with
available safety literature and validate their experimental setup
and environmental controls before initiating any experiments.
Methyl 4-iodo-3,5-dihydroxybenzoate (6).35 Methyl 3,5-
dihydroxybenzoate (4.00 kg, 97 wt/wt % pure, 23.1 mol),
sodium bicarbonate (5.82 kg, 69.2 mol, 3 equiv), water (40 L),
and THF (12 L) were charged to an inerted 200 L glass-lined
reactor, and the contents were cooled to 5 °C. Iodine (11.7 kg,
46.2 mol, 2 equiv) was dissolved in THF (28 L) and added
over 30 min while maintaining the internal temperature between
2-8 °C. After stirring 1 h, MTBE (20 L) was added in one
portion, followed by a solution of sodium sulfite (3.22 kg, 25.4
mol, 2 equiv) in water (12 L) over 30 min (gas evolution)
keeping the internal temperature between 2-10 °C. The
resulting biphasic mixture was warmed to ambient temperature
and the lower aqueous phase was withdrawn. The light yellow
product-rich organic phase was concentrated at atmospheric
pressure to 20 L and then displaced into methanol via constant-
volume azeotropic distillation, stopping when GC headspace
analysis indicated <2% THF. The dark red solution was cooled
to 45 °C and adjusted to 28 L with methanol. Water (28 L)
was then added over 15 min, during which time the product
crystallized as off-white to orange needles. The slurry was
cooled to ambient temperature, granulated for 2 h, and the
product was isolated by filtration, washed with 12 L of 1:1
methanol:water, and dried in the vacuum oven. A total of 5.68
kg (84%) of the title product was isolated as cinnamon-colored
needles. 1H NMR (400 MHz, DMSO-d6) δ: 10.50 (s, 2H), 6.93
(s, 2H), 3.79 (s, 3H). 13C NMR (100 MHz, DMSO-d6) δ: 166.0,
158.2, 130.5, 105.8, 81.8, 52.2. HRMS-ESI m/z: [M-H]- calcd
a Isolated yields. Products purified by crystallization or column chro-
matography. b Reactions with propyne conducted in hydrogenation equipment, not
standard glassware. c Isolated yield on a 5 kg batch of 6. Yields ranged from 64 to
75% on smaller batches.
extracted oligomers and precipitated polymers retained both Pd
and Cu, eliminating the need for further remediation.
Although benzofuran 1 could be crystallized directly from
the aqueous solution with cooling, the yields were comparatively
low (42%). As a result, the product was extracted into ethyl
acetate. Following a solvent displacement to methanol, 1 was
crystallized by slow addition of water. This scavenger-free
sequence was conducted successfully on 400 g scale (62% yield,
49 ppm Pd) and three times on 1.67 kg scale (68% combined
yield) as described in the experimental section.
Evaluation of Scope. We briefly evaluated the scope of the
Sonogashira reaction between iodoresorcinol 6 and various
alkynes (Table 1). This reaction appears to be general for this
substrate, providing the desired benzofuran products in good
yield after isolation and purification.
Conclusions
In conclusion, we have demonstrated an extension of the
Sonogashira-cyclization strategy for the synthesis of benzofurans
from terminal alkynes and 2-iodoresorcinols in an efficient two-
step synthesis of 4-hydroxy-2-methylbenzofuran 1 from 4-io-
doresorcinol 6 and propyne gas. In doing so, we developed a
method for screening reactions with propyne using existing
(35) Procedure reference: Denieul, M.-P.; Laursen, B.; Hazell, R.; Skryd-
strup, T. J. Org. Chem. 2000, 65, 6052–6060. The procedure was
modified to accommodate solvent restrictions in our scaleup facilities
and to improve robustness.
Vol. 14, No. 1, 2010 / Organic Process Research & Development
•
185

for C8H6O4127I, 292.93054; found, 292.93128. HPLC: rt 5.30
min (product), 4.03 min (starting material).
All ethyl acetate extracts from the three workups were
combined and concentrated at atmospheric pressure to 10 L
final volume. Methanol (25 L) was added and the remainder
of the ethyl acetate was displaced by constant-volume azeotropic
distillation until GC headspace analysis indicated <2% ethyl
acetate was present. The solution was then concentrated to 10
L and cooled to 30 °C. Water (12 L) was added slowly to
crystallize the product. The slurry was cooled to 15 °C,
granulated for 8 h, isolated by filtration, and washed with water
(28 L). After drying (vacuum, 50 °C) a total of 2.36 kg (11.5
mol, 68%) of the title product was obtained as an off-white
solid. 1H NMR (400 MHz, DMSO) δ: 10.19 (s, 1H), 7.52 (dd,
J ) 1.0, 1.1 Hz, 1H), 7.21 (d, J ) 1.1 Hz, 1H), 6.66 (dd, J )
0.9, 1.0 Hz, 1H), 3.83 (s, 3H), 2.44 (d, J ) 1.0 Hz, 3H). 13C
NMR (100 MHz, DMSO) δ: 166.5, 156.9, 155.0, 150.1, 125.4,
122.5, 108.3, 103.4, 100.7, 52.1, 13.9. HRMS-ESI m/z: [M-H]-
calcd for C11H10O4, 205.0495; found, 205.0507. HPLC: rt 6.04
min (product), 5.30 min (starting material).
General Procedure for Example Sonogashira Reactions.
Iodoresorcinol 6 (1 equiv), bis(triphenylphosphine)palladium
dichloride (3 mol %), copper(I) iodide (3 mol %), and
dimethylacetamide (10 volumes) were combined and degassed
with nitrogen. Alkyne (3 equiv) and diisopropylamine (3 equiv)
were charged and the mixture heated to 40-50 °C and held
until the reaction was complete. The crude reaction mixture
was partitioned between EtOAc and half-saturated ammonium
chloride solution, and the organic phase was removed, dried,
and concentrated to a brown oil. Chromatography on silica gel
was utilized to isolate pure benzofuran product.
General Procedure for Screening Sonogashira Reactions
between Propyne and Methyl 4-Iodo-3,5-dihydroxybenzoate
(6). Each glass Endeavor tube36 was charged with iodoresorcinol
6 (200 mg, 0.68 mmol), bis(triphenylphosphine)palladium
dichloride (24 mg, 0.034 mmol, 0.05 equiv), copper iodide (7
mg, 0.03 mmol, 0.05 equiv), solvent (10 mL/g), and amine base
(3 equiv). The tubes were placed into an Endeavor system and
sealed. The reactions were then purged with N2 twice with
stirring (250 rpm). The reactors were then warmed to the
appropriate temperature, stirred at 1000 rpm, and pressurized
with propyne to the target pressure. After 24 h, the reactions
were purged with N2 and removed for analysis.
4-Hydroxy-6-methoxycarbonyl-2-methylbenzofuran (1).
A 30 L hastelloy pressure reactor was charged with methyl
3,5-dihydroxy-4-iodobenzoate (1.66 kg, 5.64 mol, 1 equiv),
copper(I) iodide (0.050 kg, 0.05 equiv), bis(triphenylphosphine)-
palladium dichloride (0.20 kg, 0.05 equiv), N,N-dime-
thylacetamide (16.6 L, 10 volumes), and diisopropylethylamine
(1.71 kg, 3 equiv). The reactor was sealed and purged three
times with nitrogen (25 psig) with stirring. The reactor was then
purged four times with propyne with agitation. Dissolution of
propyne was exothermic; during this process the internal
temperature was controlled between 25-35 °C. Purges were
continued until the internal pressure reached 6 psig (first purge),
8 psig (second purge), and 10 psig (third and fourth purge), at
which time the reactor was vented. After four purges, the
solution was saturated (as demonstrated by continuous degassing
of the solution while the vent valve was open), so propyne was
added to adjust the internal pressure to 10 psig, the reaction
was heated to 38 °C and maintained under a 10 psig propyne
atmosphere for 20 h. Approximately 1.1-1.2 kg of propyne
was used for the purges and reaction pressurization. The dark,
homogeneous reaction mixture was cooled to 15 °C, sparged
with nitrogen for 15 min, and transferred to another reactor
where it was placed under vacuum for 30 min. The reaction
was repeated two more times on identical scale and the crude
reaction mixtures were combined.
For workup, the crude reaction mixture was split in three
equal portions and each portion was processed identically as
described below. The crude reaction mixture was transferred
to a 75 L reactor, heated to 40 °C, and water (16.4 L, 10
volumes) was added over 30 min, maintaining the internal
temperature between 40-50 °C. Saturated aqueous ammonium
chloride (16.4 L, 10 volumes) was then added, during which
time dark semisolid polymer precipitated from the reaction
mixture. While maintaining the internal temperature between
40-50 °C, the product-rich aqueous phase was washed twice
with 8.2 L (5 volumes each wash) of 100:1 heptane:ethyl acetate
to remove the insoluble polymer from the aqueous phase.
Between washes the reactor was rinsed with acetone to remove
any polymer that remained in the tank after the organic phase
was drained. The aqueous phase was returned to the cleaned
reactor, cooled to ambient temperature, and extracted three times
with ethyl acetate (8.2 L, 5 volumes each time).
4-Hydroxy-6-methoxycarbonyl-2-phenylbenzofuran (12).
Prepared from 6 (1.017 g) and phenylacetylene (1.2 mL) at 40
°C for 23 h. Chromatography (5-25% EtOAc/hexanes) pro-
vided 671 mg 12 (72% yield) as a beige solid. An additional
1
236 mg of uncyclized alkyne was obtained (25% yield). H
NMR (400 MHz, DMSO) δ: 10.48 (s, 1H), 7.95 (m, 2H), 7.65
(dd, J ) 1.0, 1.0 Hz, 1H), 7.52 (m, 2H), 7.51 (m, 1H), 7.43
(m, 1H), 7.28 (d, J ) 1.1 Hz, 1H), 3.86 (s, 3H). 13C NMR
(100 MHz, DMSO) δ: 166.3, 156.3, 155.1, 151.0, 129.3, 129.2,
129.1, 126.6, 124.8, 122.7, 108.5, 103.7, 99.8, 52.2. HRMS-
ESI m/z: [M + H]+ calcd for C16H12O4, 269.0808; found,
269.0810. HPLC: rt 7.79 min.
2-(Cyclohex-1-enyl)-4-hydroxy-6-(methoxycarbonyl)ben-
zofuran (13). Prepared from 6 (1.0 g) and 1-ethynylcyclohexene
(1.2 mL) at 40 °C for 24 h. Chromatography (5-10% EtOAc/
hexanes) provided 801 mg 12 (86% yield) as a beige solid. 1H
NMR (400 MHz, DMSO) δ: 10.35 (s, 1H), 7.52 (dd, J ) 1.1,
1.2 Hz, 1H), 7.22 (d, J ) 1.3 Hz, 1H), 6.81 (m, 1H), 6.56 (m,
1H), 3.83 (s, 3H), 2.34 (m, 2H), 2.23 (m, 2H), 1.71 (m, 2H),
1.62 (m, 2H). 13C NMR (100 MHz, DMSO) δ: 166.4, 157.7,
154.6, 150.7, 127.0, 126.6, 126.2, 122.5, 108.4, 103.3, 98.3,
52.1, 24.9, 24.2, 21.8, 21.5. HRMS-ESI m/z: [M + H]+ calcd
for C16H16O4, 273.1121; found, 273.1122. HPLC: rt 8.03 min.
4-Hydroxy-2-(2-hydroxypropan-2-yl)-6-(methoxycarbo-
nyl)benzofuran (14). Prepared from 6 (1.07 g) and 2-methyl-
3-butyn-2-ol (1.1 mL) at 40 °C for 24 h. Chromatography
(15-30% EtOAc/hexanes) provided 850 mg 12 (93% yield)
as a beige solid. 1H NMR (400 MHz, DMSO) δ: 10.24 (s, 1H),
7.55 (dd, J ) 1.0, 1.2 Hz, 1H), 7.22 (d, J ) 1.2 Hz, 1H), 6.75
(36) An ∼7 mL disposable glass tube purchased from Biotage for use on
the Endeavor system.
186
•
Vol. 14, No. 1, 2010 / Organic Process Research & Development

(d, J ) 0.9 Hz, 1H), 5.45 (s, 1H), 3.84 (s, 3H), 1.51 (s, 6H).
13C NMR (100 MHz, DMSO) δ: 166.4, 166.1, 154.9, 150.7,
125.9, 121.9, 108.3, 103.7, 97.9, 67.6, 52.1, 28.9. HRMS-ESI
m/z: [M + H]+ calcd for C13H14O5, 251.0914; found, 251.0914.
HPLC: rt 5.15 min.
Acknowledgment
The authors thank the following Pfizer colleagues: Teresa
Makowski, Mark Shaffer, Tim Houck, Jim Hallissey, Billy
Buzon, Dave Hall, and Supply Chain-API for assistance with
the development and scaleup of the Sonogashira reaction. The
authors thank Andrew Jensen, Brian Jones, Yanqiao Xiang, and
Zhongli Zhang for analytical support, and Dan Bowles, David
Damon, Alan Grubbs and Rob Singer for helpful discussions
of this chemistry.
2-[3-(tert-Butyldiphenylsiloxy)propyl]-4-hydroxy-6-(meth-
oxycarbonyl)benzofuran (15). Prepared from 6 (1.01 g) and
1-(tert-butyldiphenylsiloxy)pentyne37 (3.3 mL) at 40 °C for 24 h.
Chromatography (3-15% EtOAc/hexanes) provided 1.382 g
1
12 (82% yield) as an off-white solid. H NMR (400 MHz,
DMSO) δ: 10.20 (s, 1H), 7.60 (m, 4H), 7.52 (dd, J ) 1.0, 1.1
Hz, 1H), 7.43 (m, 2H), 7.41 (m, 4H), 7.22 (d, J ) 1.1 Hz, 1H),
6.66 (m, 1H), 3.84 (s, 3H), 3.72 (t, J ) 6.1 Hz, 2H), 2.90 (t, J
) 7.1 Hz, 2H), 1.96 (m, 2H), 1.00 (s, 9H). 13C NMR (100
MHz, DMSO) δ: 166.5, 160.2, 154.9, 150.2, 135.0, 133.1,
129.8, 127.8, 125.5, 122.2, 108.3, 103.5, 100.1, 62.4, 52.1, 29.8,
26.6, 24.2, 18.8. HRMS-ESI m/z: [M + H]+ calcd for
C29H32O5Si, 489.2092; found, 489.2096. HPLC: rt 9.30 min.
Supporting Information Available
Detailed information and discussion of the methods used to
1
calculated propyne solubility in reaction solvents and H and
13C NMR spectra of compounds 1, 6, 12, 13, 14, and 15. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Received for review October 15, 2009.
OP900263B
(37) Clark, R. C.; Lee, S. Y.; Boger, D. L. J. Am. Chem. Soc. 2008, 130,
12355–12369.
Vol. 14, No. 1, 2010 / Organic Process Research & Development
•
187