Application of a batch microwave unit for scale-up of alkoxycarbonylation reactions using a near-stoichiometric loading of carbon monoxide

Article abstract of DOI:10.1021/op800296d

The ethoxycarbonylation of iodobenzene was performed on the 1 mol scale in batch mode using microwave heating. The reaction was performed using both an excess and a near stoichiometric loading of carbon monoxide, comparable yields being obtained. Six different alkoxycarbonylation reactions were then performed simultaneously on the 50 mmol scale using a near-stoichiometric loading of carbon monoxide with excellent conversions in each case.

Full text of DOI:10.1021/op800296d

Organic Process Research & Development 2009, 13, 634–637
Application of a Batch Microwave Unit for Scale-Up of Alkoxycarbonylation
Reactions Using a Near-Stoichiometric Loading of Carbon Monoxide
Mauro Iannelli,† Fabio Bergamelli,† Chad M. Kormos,‡ Stefano Paravisi,† and Nicholas E. Leadbeater*,‡
Milestone s.r.l., Via Fatebenefratelli, 1/5, Sorisole (BG), Italy, and Department of Chemistry, UniVersity of Connecticut, 55
North EagleVille Road, Storrs, Connecticut 06269-3060, U.S.A.
Abstract:
prepressurized with carbon monoxide.6 Reactions were per-
formed using 0.1 mol % palladium acetate as catalyst with no
additional ligand required. DBU proved to be the optimal base
for the reaction, and a range of aryl iodide substrates can be
converted to the corresponding esters using this methodology.
Reactions were complete within 20 min at 125 °C. We
subsequently developed a methodology for performing the
reaction using a near-stoichiometric loading of carbon monox-
ide.7 We loaded the reaction vessel with the requisite quantity
of carbon monoxide and then used nitrogen to reach a total
pressure of 10 bar. We had to increase the palladium loading
to 0.5 mol % to obtain good yields of the desired ester products.
Reactions were again complete within 20 min at 125 °C.
Recently we have become interested in developing scale-
up strategies for microwave-promoted reactions. This is essential
if the technique is going to continue to gain acceptance and
become practical at the process level. Possible approaches to
scale-up include both batch and continuous-flow processing.
Recent work in our laboratory and others has been focused on
exploring both possibilities.8-10 For part of this, we have probed
the scale-up of our alkoxycarbonylation chemistry. Our initial
efforts were directed at performing the reaction in batch mode
using a multimode microwave reactor equipped with eight
heavy-walled quartz reaction vessels.11 This gave us the
capability to load up to eight reaction vessels and run them
simultaneously by means of a reaction carousel. Using the
ethoxycarbonylation of iodobenzene as a test reaction and
starting with 0.1 mol of iodobenzene across the eight reaction
vessels, we obtained an overall conversion of 91% and an
isolated yield of 81% after chromatography. The disadvantage
of this approach is that it takes time to load each vessel with
reagents, pressurize each with carbon monoxide, and at the end
of the reaction, decant each product mixture. Our next approach
was to use a single reaction vessel of larger capacity. To achieve
this we used a scientific microwave unit equipped with a 300-
The ethoxycarbonylation of iodobenzene was performed on the 1
mol scale in batch mode using microwave heating. The reaction
was performed using both an excess and a near stoichiometric
loading of carbon monoxide, comparable yields being obtained.
Six different alkoxycarbonylation reactions were then performed
simultaneously on the 50 mmol scale using a near-stoichiometric
loading of carbon monoxide with excellent conversions in each
case.
1. Introduction
Palladium-catalyzed carbonylation of aryl halides offers a
one-step route to a range of products including carboxylic acids,
esters, and amides.1,2 Of these variants, alkoxycarbonylation
(synthesis of esters) accounts for the majority of the industrial
applications. In our laboratory we have been focusing attention
on the use of simple ligandless palladium complexes as catalysts
for carbonylation chemistry.3 We perform our reactions using
microwave heating, this being a valuable tool for synthetic
chemists because it is possible to enhance the rate of reactions
and, in many cases, improve product yields.4 Larhed and co-
workers have used Mo(CO)6 as a source of carbon monoxide
for the preparation of amides, esters, and carboxylic acids from
aryl halides using microwave heating.5 Advantages of using
Mo(CO)6 as a replacement for gaseous CO include the fact that
it is a solid and is easily used on a small scale with commercially
available microwave apparatus with no modification required.
However, Mo(CO)6 is expensive and toxic, and its use results
in metal waste-this being a particular problem if the reaction
is to be scaled up. We have found that it is possible to perform
alkoxycarbonylation reactions of aryl iodides in reaction vessels
* Author for correspondence. Email: nicholas.leadbeater@uconn.edu.
† Milestone s.r.l.
‡ University of Connecticut.
(1) For a recent review see: Barnard, C. F. J. Organometallics 2008, 27,
5402–5422.
(6) Kormos, C. M.; Leadbeater, N. E. Org. Biomol. Chem. 2007, 65–68.
(7) Kormos, C. M.; Leadbeater, N. E. Synlett 2007, 2006–2010.
(8) For reviews see: (a) Lehmann, H. In New AVenues to Efficient Chemical
Synthesis; Seeberger, P. H., Blume, T., Eds.; Springer-Verlag: Berlin,
2007. (b) Kremsner, J. M.; Stadler, A.; Kappe, C. O. Top. Curr. Chem.
2006, 266, 233–278. (c) Roberts, B. A.; Strauss, C. R. Acc. Chem.
Res. 2005, 38, 653–661.
(2) For background to the area see: (a) Kolla´ır, L., Ed. Modern Carbo-
nylation Methods; Wiley: Weinheim, Germany, 2008. (b) Zapf, A.;
¨
Beller, M. Chem. Commun. 2005, 431–440. (c) Skoda-Foldes, R.;
Kolla´ır, L. Curr. Org. Chem. 2002, 6, 1097–1119.
(3) Kormos, C. M.; Leadbeater, N. E. Synlett 2006, 1663–1666.
(4) For a recent overview of the field see: Loupy, A., Ed. MicrowaVes in
Organic Synthesis; Wiley: Weinheim, 2006.
(9) For an evaluation of microwave reactors for use in a kilolab
see: Lehmann, H.; LaVecchia, L. J. Assoc. Lab. Autom. 2005, 10, 412–
417.
(5) For examples see: (a) Appukkuttan, P.; Axelsson, L.; Van der Eycken,
E.; Larhed, M. Tetrahedron Lett. 2008, 49, 5625–5628. (b) Lagerlund,
O.; Larhed, M. J. Comb. Chem. 2006, 8, 4–6. (c) Wu, X. Y.; Larhed,
M. Org. Lett. 2005, 7, 3327–3329. (d) Wu, X. Y.; Nilsson, P.; Larhed,
M. J. Org. Chem. 2005, 70, 346–349. (e) Wannberg, J.; Larhed, M.
J. Org. Chem. 2003, 68, 5750–5753.
(10) Moseley, J. D.; Lenden, P.; Lockwood, M.; Ruda, K.; Sherlock, J.-P.;
Thomson, A. D.; Gilday, J. P. Org. Process Res. DeV. 2008, 12, 30–
40.
(11) Bowman, M. D.; Holcomb, J. L.; Kormos, C. M.; Leadbeater, N. E.;
Williams, V. A. Org. Process Res. DeV. 2008, 12, 41–57.
634
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development
10.1021/op800296d CCC: $40.75  2009 American Chemical Society
Published on Web 02/13/2009

Scheme 1. Alkoxycarbonylation of iodobenzene
mL Teflon vessel.12 We performed the reaction on scales up to
75 mmol scale using the optimized conditions from our first
small-scale protocol and obtained a quantitative conversion in
the ethoxycarbonylation of iodobenzene. Turning to our near-
stoichiometric methodology and deciding to perform the reac-
tion on the 100 mmol scale, we calculated that an initial loading
of 250 psi (17 bar) of CO would equate to 1.1 equivalents.
Running the reaction on this scale and using a catalyst loading
of 0.1 mol % resulted in >95% conversion to product and, after
column chromatography, an 86% isolated yield. We were not
able to scale up the reaction further since, at the 0.1 mol level,
we were approaching the pressure limits of the apparatus.
Knowing that further scaling of the reaction was going to be
important, we have recently turned to a microwave unit
comprising a single reaction vessel of 3.0 L capacity and a
working volume of approximately 2 L. A discussion of the
scale-up of our alkoxycarbonylation methodology to the 1 mol
level as well as running multiple 50 mmol reactions simulta-
neously in the reactor is the subject of this technical note.
2 L in volume in a Teflon vessel. In addition, it is possible to
perform multiple smaller reactions at the same time in either
Teflon, glass, or quartz vessels. This is achieved by placing all
the loaded smaller reaction vessels into a dedicated rack.
Uniform temperature conditions are achieved by positioning
all the individual sample vessels in a pool of microwave-
absorbing fluid. It is the effective temperature of this fluid that
is monitored and controlled throughout the heating process.
Since the entire microwave cavity is under pressure, in normal
operation boiling never occurs in individual sample containers.
Loose-fitting covers on each container prevent loss of sample
or cross-contamination. We did not use these since we wanted
to ensure a carbon monoxide environment was in each reaction
vessel. The unit is able to operate at temperatures up to 260 °C
and at pressures of 200 bar. All the reaction parameters (time,
temperature, pressure, microwave power, stirring speed) can
be monitored and controlled during the run through software
installed on a dedicated touch-screen terminal. In using this unit,
we envisaged simply using a mixture of nitrogen and carbon
monoxide in the prepressurization stage, thus allowing us to
perform our alkoxycarbonylation chemistry efficiently on a
larger scale in one vessel or in a number of smaller vessels.
2.2. Scaling the Alkoxycarbonylation Reaction. To initiate
our study we chose to focus on the ethoxycarbonylation of
iodobenzene on the 1 mol scale. Our results are summarized
in the context of our previous reports in Scheme 1. We decided
first to perform the reaction on the 1 mol scale using 1.4
stoichiometric equivalents of carbon monoxide. We placed the
3.0 L capacity Teflon vessel containing the iodobenzene, DBU
(1.1 equiv), palladium acetate (0.1 mol %), and ethanol (1.8 L)
into the UltraCLAVE microwave unit. After first purging the
reactor with nitrogen three times, we then loaded with nitrogen
to a pressure of 8 bar before loading with 34 bar of CO
(calculated to be 1.4 equiv) and then a further 15 bar of nitrogen
for a total pressure of 57 bar. This sequence was important since
the internal pressure forms the seal between the vessel and the
lid of the unit. For safety reasons we wanted to ensure a tight
seal before loading the carbon monoxide into the reactor and
thus prepressurized to greater than the sealing limit with nitrogen
first. We then heated the contents of the reactor to 125 °C over
the period of 15 min and held at this temperature for 30 min.
During the course of the reaction a drop in pressure to 40 bar
2. Results and Discussion
2.1. Microwave Apparatus Used. Our work was performed
using the Milestone UltraCLAVE.10,13 The unit has found most
application in the area of sample digestion.14 Its design is based
on a high-pressure autoclave. It has a single, large reaction
chamber of 3.0 L volume which is prepressurized with nitrogen
to a level of approximately 15 bar. The prepressurization step
also allows for easy introduction of reactive gases when
required. Microwave energy is introduced into the reactor
through a port located directly at the end of the waveguide.
This design allows a microwave field of uniform intensity to
be introduced directly into the reaction chamber overcoming
the issues related to nonhomogeneous microwave field in large
cavities. The pressurized chamber in the system serves simul-
taneously as the microwave cavity and the reaction vessel. By
using this unit it is possible to run one large reaction of up to
(12) Bowman, M. D.; Schmink, J. R.; McGowan, C. M.; Kormos, C. M.;
Leadbeater, N. E. Org. Process Res. DeV. 2008, 12, 1078–1088.
(13) For a description of the apparatus, see: Borowski, K. J.; Schoenfeld,
C. Am. Lab. 2006, 38, 26–29.
(14) (a) Krachler, M.; Le Roux, G.; Kober, B.; Shotyk, W. J. Anal. At.
Spectrom. 2004, 19, 354–361. (b) Krachler, M.; Mohl, C.; Emons,
H.; Shotyk, W. Spectrochim. Acta B 2002, 57, 1277–1289. (c) Brunner,
I.; Brodbeck, S. EnViron. Pollut. 2001, 114, 223–233.
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
635

Table 1. Alkoxycarbonylation of aryl iodides on the 50
was observed, this clearly indicating consumption of the carbon
monoxide. Upon completion of the heating stage, we then
allowed the reactor to cool to 55 °C before slowly releasing
the residual pressure (32 bar) into a well-ventilated area. The
cool-down phase was somewhat lengthy (approximately 45 min)
because of the significant size of the vessel and the thick cavity
walls. After removal of the reaction vessel from the unit and
workup of the product mixture, pure ethyl benzoate was
obtained in 79% yield, this being comparable to that obtained
in our previous smaller-scale studies.
mmol scalea
Our next objective was to perform the reaction using a near-
stoichiometric loading of carbon monoxide. This methodology
has significant advantages when working at larger scales, the
major improvement being in safety. When working with a near-
stoichiometric loading of carbon monoxide not only is the
overall inventory of gas smaller, but also almost all the CO
has been consumed at the end of the reaction, thus limiting the
quantity liberated to the atmosphere upon depressurizing the
reactor to access the vessel. By reducing the amount of CO
used, it may also be possible to prolong the lifetime of catalysts
and increase turnover numbers, and if using radiolabeled carbon
monoxide, this methodology would represent a significant cost
saving since isotopically labelled CO is expensive. Working
on the 1 mol scale we attempted the ethoxycarbonylation of
iodobenzene using a 1:1.08 stoichiometric ratio of aryl iodide
to carbon monoxide. We calculated that this equated to an initial
CO loading of 27 bar. Again, after purging with nitrogen three
times, we first ensured effective sealing of the vessel in the
unit by prepressurizing with nitrogen (11 bar) before loading
the carbon monoxide followed by a further 12 bar of nitrogen,
giving a total initial pressure of 50 bar. Using the same heating
program as before, we held the reaction mixture at 125 °C for
30 min. Upon cooling the reaction vessel, venting residual
pressure, and working up the product mixture, we obtained an
80% isolated yield of ethyl benzoate. This shows that the
reaction can be performed using a near-stoichiometric loading
of carbon monoxide on the 1 mol scale with a product yield
comparable to that obtained in our initial 1 mmol trials in
method optimization studies or to our efforts in scaling the
reaction to the 0.1 mol level.
2.3. Performing Multiple Alkoxycarbonylation Reactions
Simultaneously. In the knowledge that we could scale-up our
alkoxycarbonylation reaction in one vessel, we next wanted to
attempt a scale-up of six individual ethoxycarbonylation reac-
tions simultaneously using the microwave unit, each being
performed at the 50 mmol level. To achieve this we loaded six
substrates into individual 150- mL capacity glass reaction
vessels, choosing iodobenzene, 2-iodotoluene, 4-iodotoluene,
4-iodoanisole, 2-iodoanisole, and 4-iodoacetophenone. To each
we added DBU (1.1 equiv), palladium acetate (0.1 mol %),
and ethanol (90 mL). The loaded vessels were placed (without
covers) into a rack, and this in turn was attached to the lid of
the microwave unit. A volume of ethanol was placed inside
the large Teflon vessel, thus serving as the requisite pool of
microwave-absorbing fluid. After making sure that the smaller
vessels sat in the pool of ethanol in the larger vessel and
ensuring effective sealing in the unit by prepressurizing with
nitrogen, we loaded the reactor with 1.08 stoichiometric
a 50 mmol aryl halide, 90 mL of ethanol, 0.1 mol % Pd(OAc), 1.1 equiv of
DBU, 8 bar CO (1.08 equiv). Reaction mixture ramped from rt to 125 °C over the
period of 15 min and held at this temperature for 30 min.
equivalents of carbon monoxide (8 bar, based on six 50 mmol
reactions) followed by further nitrogen, to give a total initial
pressure of 50 bar. Using the same heating program as before,
we held the reaction mixture at 125 °C for 30 min. After cooling
the reaction vessel and venting residual pressure, the six vessels
were removed from the microwave unit and products isolated.
The results are shown in Table 1. In each case an excellent
conversion to the desired ester product was observed. Of note
is that no observed cross contamination of the reaction mixtures
was observed. This confirms that, under the conditions used,
the prepressurization of the microwave cavity ensures that the
partial pressures of the aryl iodide starting materials or ester
products never exceed that required to allow them to exit the
individual vessels. The use of a moderating fluid allows us to
negate problems arising from differential microwave absorp-
tivity of reaction mixtures when performing parallel syntheses.
Previous results from our group have shown that the heating
characteristics of the substrate can affect significantly the
outcome of a reaction, particularly in cases when reactions
involving poorly microwave-absorbing solvents are used.15
Another parameter for consideration when performing a syn-
thesis using a range of substrates in a multimode microwave
apparatus is into which vessel the temperature probe should be
placed and thus is used as the control. In our case this is not an
issue since the temperature probe is inserted into the moderating
fluid.
(15) Leadbeater, N. E.; Schmink, J. Tetrahedron 2007, 63, 6764.
636
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development

3. Summary
bar of carbon monoxide (1.4 equiv based on iodobenzene)
followed by 15 bar of nitrogen, this giving a total pressure of
57 bar. The reaction mixture was heated to 125 °C over the
period of 15 min. The contents of the vessel were then held at
this temperature for 30 min. The reaction mixture was stirred
throughout the run. Upon completion of the heating stage, the
reaction mixture was cooled to 55 °C, this taking 45 min; then
the reaction chamber was slowly vented and opened. The
contents of the reaction vessel were added into diethyl ether
(500 mL). The brown precipitate that formed (DBU ·HCl salt)
was filtered off, dissolved in water, and extracted with diethyl
ether (3 × 50 mL). All the organic phases were collected,
concentrated under reduced pressure, and washed with 1 N HCl
(2 × 150 mL) and saturated KCl solution (3 × 150 mL). After
adding petroleum ether (200 mL of bp 30-40 °C), the organic
phase was washed further with distilled water (2 × 50 mL)
and dried over MgSO4. It was then concentrated under reduced
pressure to afford ethyl benzoate (119 g, 79% yield).
In this technical note we have shown that alkoxycarbony-
lation reactions can be scaled up successfully in batch mode
using a sealed-vessel multimode microwave unit. The ethoxy-
carbonylation of iodobenzene was performed on the 1 mol scale
using both an excess and a near-stoichiometric loading of carbon
monoxide. Yields comparable to those previously reported for
the reaction on smaller scales were obtained. Six different
alkoxycarbonylation reactions were performed simultaneously
on the 50 mmol scale using a near stoichiometric loading of
carbon monoxide with excellent conversions in each case.
4. Experimental Section
4.1. General Experimental. All reagents were obtained
from commercial suppliers and used without further purification.
1H NMR spectra were recorded at 293 K on a 300 or 400 MHz
spectrometer.
4.2. Equipment. Reactions were performed in a Milestone
UltraCLAVE multimode microwave unit. The instrument
consists of a continuous microwave power delivery system with
power output from 0-1000 W. Operating pressures up to 200
bar and temperatures up to 260 °C can be achieved. Single
reactions were performed in a 3.0-L capacity Teflon vessel. The
temperature of the contents of the vessel was monitored using
a shielded thermocouple inserted directly into the reaction
mixture by means of a Teflon coated stainless steel thermowell.
The contents of the vessel were stirred by means of a rotating
magnetic plate located below the floor of the microwave cavity
and a Teflon-coated magnetic stir bar in the vessel. To perform
multiple reactions at the same time, 150-mL capacity Teflon
reaction vessels were used. Ethanol was employed as the
moderating fluid, and the temperature of this was recorded
during the course of a run. In all cases, the microwave cavity
was prepressurized to 50 bar. Pressure was measured using a
transducer. Product mixtures were cooled at the end of a run
by passing a flow of coolant around the outside of the stainless
steel microwave cavity. All operations (sealing, pressurizing,
heating, cooling, venting, and opening the vessel in the
microwave cavity) were undertaken using computer control.
During the course of a run, integrated sensors continuously
monitored and displayed internal pressure, temperature, and
microwave power applied. The software dynamically adjusted
the applied microwave power in real time to follow a predefined
temperature profile.
4.3. Experimental Procedures for the Reactions Per-
formed in This Study. Ethoxycarbonylation of Iodobenzene
Using an Excess of Carbon Monoxide. Palladium acetate (220
mg, 1 mmol) and ethanol (1.8 L) were combined in a 3.0 L
Teflon vessel placed into the microwave unit and stirred. DBU
(167.0 g, 163.9 mL, 1.1 mol) and iodobenzene (204.0 g, 112.1
mL, 1 mol) were added. The microwave cavity was closed and
the vessel loaded and purged three times with nitrogen. In cycle
1, the vessel was pressurized with nitrogen from atmospheric
to 30 bar, and then the pressure was released to 8 bar. In cycles
2 and 3 the pressure was raised to 30 bar and then released
back to 8 bar. Following this, the vessel was loaded with 34
Ethoxycarbonylation of Iodobenzene Using a Near-Stoichio-
metric Loading of Carbon Monoxide. An identical procedure
was used with the exception that in the final stage of the nitrogen
purge cycle the vessel was pressurized with nitrogen from
atmospheric to 30 bar, and then the pressure was released to
11 bar. Following this, the vessel was loaded with 27 bar of
carbon monoxide (1.08 equiv based on iodobenzene) followed
by 12 bar of nitrogen, this giving a total pressure of 50 bar.
Upon workup ethyl benzoate (120 g, 80% yield) was obtained.
Ethoxycarbonylation of Six Aryl Iodides Using a Near-
Stoichiometric Loading of Carbon Monoxide. To each of six
Teflon vessels (150 mL capacity) equipped with a magnetic
stir bar and placed into a six-position rack were added palladium
acetate (11 mg, 0.05 mmol) in ethanol (90 mL), DBU (8.37 g,
8.2 mL, 55 mmol), and aryl iodide (50 mmol). Ethanol (400
mL) was placed into the 3.0 L Teflon vessel as a moderating
fluid. The vessel was loaded and purged three times with
nitrogen. In cycle 1, the vessel was pressurized with nitrogen
from atmospheric to 30 bar, and then the pressure was released
to 8 bar. In cycles 2 and 3 the pressure was raised to 35 bar
and then released back to 11 bar. Following this, the vessel
was loaded with 8 bar of carbon monoxide (1.08 equiv based
on six 50 mmol reactions) followed by 31 bar of nitrogen, this
giving a total pressure of 50 bar. An identical heating program
was used as in the previous experiments (heat to 125 °C over
15 min and hold for 30 min). After cooling and venting, each
product mixture was worked up individually using the same
protocol as before but scaling down volumes of solvents and
water by a factor of 20.
Acknowledgment
We thank the American Chemical Society Petroleum
Research Foundation (45433-AC1) for funding and Milestone
s.r.l. for access to the UltraCLAVE microwave unit.
Received for review November 26, 2008.
OP800296D
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
637