Research and development of the catalytic oxidation of methylacrylate to 3,3-dimethoxy methyl propionate

Article abstract of DOI:10.1021/op900001p

A selective synthesis of 3,3-dimethoxy methyl propionate has been developed using a green approach, in which the key step is a palladium-catalyzed oxidation of methylacrylate in methanol using oxygen as oxidant. The relationship between several reaction p

Full text of DOI:10.1021/op900001p

Organic Process Research & Development 2009, 13, 548–554
Research and Development of the Catalytic Oxidation of Methylacrylate to
3,3-Dimethoxy Methyl Propionate
⊥
§
Yoshiyuki Tanaka, Jun P. Takahara, and Hans E. B. Lempers*
Mitsubishi Chemical Corporation, 3-10 Ushiodori, Kurashiki, Okayama 712-8054, Japan, Mitsubishi Chemical Corporation,
1000 Kamoshida-cho, Aoba-ku, Yokohama 227-8502, Japan, and Rainbow Oxidations, Niels Bohrweg 11-13,
2333 CA Leiden, The Netherlands
Abstract:
Scheme 1. Pd/Fe/Cu-catalyzed oxidation of methylacrylate
to 3,3-dimethoxy methyl propionate using oxygen as oxidant
A selective synthesis of 3,3-dimethoxy methyl propionate has been
developed using a green approach, in which the key step is a
palladium-catalyzed oxidation of methylacrylate in methanol using
oxygen as oxidant. The relationship between several reaction
parameters including catalyst composition, oxygen pressure,
substrate/solvent ratio, reaction temperature and reaction perfor-
mance have been discussed. The data obtained in combination with
data on explosion limits enabled successful scale-up to a 2 L scale,
giving the desired product in 82% yield and 99% purity. Special
attention is paid to catalyst activity, catalyst cost contribution,
solvent recovery and safety. The product is a valuable industrial
fine chemical for the synthesis of pharmaceuticals, functionalized
polymers and adhesives, as it contains a protected aldehyde
functional group, an ester functional group and an activated
methylene function.
as catalyst components under oxygen or air pressure. Generally,
the Wacker oxidation combines the stoichiometric oxidation
5
of an olefin by Pd(II) in aqueous solution with the reoxidation
of Pd(0) in situ by molecular oxygen in the presence of copper
salts. The overall reaction constitutes a palladium-catalyzed
oxidation of olefins to aldehydes or ketones. In our case the
reaction differs from conventional Wacker oxidation as reaction
6
takes place in an alcohol as solvent. The Pd(II) oxidizes the
methylacrylate substrate to produce 3-methoxy methylacrylate,
while being reduced to Pd(0). Under the acidic reaction
conditions the 3-methoxy methylacrylate is only an intermediate
and is quickly converted to 3,3-dimethoxy methyl propionate.
Reoxidation of Pd(0) is achieved by Cu(II) oxidation as is
similar to conventional Wacker oxidation. The Cu(I) which is
formed can be reoxidized to Cu(II) by oxygen, but we have
Introduction
We became interested in the synthesis of 3,3-dimethoxy
methyl propionate because we think it has high potential as a
synthetic building block in numerous chemical processes, due
to the presence of a protected aldehyde functional group, an
ester functional group and an activated methylene functional
3
found it advantageous to reoxidize the Cu(I) using Fe(III) and
then to reoxidize the Fe(II) using oxygen as this is faster,
enhancing the stability of the palladium catalyst. In principle,
this oxidation of methylacrylate can be carried out in water,
which would lead to the free aldehyde as product. However,
aldehydes are very susceptible to overoxidation, leading to a
decrease in product selectivity. By carrying this reaction out in
alcohol solvents the aldehyde product is protected as an acetal,
minimizing overoxidation. Not only is alcohol a good solvent
for this reaction based on reactivity and miscibility, but also
through the formation of protective acetals it has a positive effect
on product selectivity.
This process makes use of a catalytic reaction and an
environmentally friendly oxidant and fits nicely into the context
of green chemistry. However, there are three key problems with
this type of reaction: the danger of an explosion due to the
oxygen-rich gas phase, the danger of runaway reactions due to
1
group,. In fact, we have previously developed a route for the
synthesis of methyl cyanoacetate from 3,3-dimethoxy methyl
propionate. Methyl cyanoacetate is a key intermediate in the
synthesis of cyanomethyl methacrylate, which finds wide
2
application in the adhesive industry. We envisaged the synthesis
of 3,3-dimethoxy methyl propionate by a Wacker-type oxidation
of the cheap starting material methylacrylate, using methanol
as solvent (Scheme 1).
This type of oxidation reaction is well established at the
Mitsubishi Chemical Corporation in, for example, the oxidation
3
of cyclohexene in glycols to give cyclohexanone ketals and
the oxidation of acrolein in 1,3-propanediol to give the
4
corresponding propanedialdehyde diacetal. A general procedure
for the methylacrylate oxidation consists of reacting the substrate
with a methanol solution containing palladium, copper and iron
(
5) (a) Sheldon, R. A.; Kochi, J. K. Metal-Catalyzed Oxidations of Organic
Compounds; Academic Press: London, 1981. (b) Smidt, J.; Hafner, W.;
Jira, R.; Sedlmeier, J.; Sieber, R.; Ruttinger, R.; Kojer, H. Angew. Chem.
1959, 71, 176. (c) Smidt, J.; Hafner, W.; Jira, R.; Sedlmeier, J.; Sabel,
A. Angew. Chem., Int. Ed. Engl. 1962, 1, 80. (d) Smidt, J. Chem. Ind.
1962, 54. (e) Shioyama, T. K. U.S. Patent 4,507,506, 1985. (f) Clement,
W. H.; Selwitz, C. M. J. Org. Chem. 1964, 29, 241. (g) Fahey, D. R.;
Zuech, E. A. J. Org. Chem. 1974, 39, 3276.
*
To whom correspondence should be addressed. Telephone: +31-71-
5
288702. E-mail: hanslempers@live.nl.
⊥
Mitsubishi Chemical Corporation.
Rainbow Oxidations.
§
(
(
1) Lempers, H. E. B.; Takahara, J. P. JP2005/097233, 2005.
2) Cary, R. Concice International Chemical Assessment, Document 36,
2
001.
(6) (a) Lloyd, W. G.; Luberoff, B. J. J. Org. Chem. 1969, 34, 3949. (b)
Hosokawa, T.; Ohta, T.; Kanayama, S.; Murahashi, S. J. Org. Chem.
1987, 52, 1758.
(
(
3) Lempers, H. E. B.; Setoyama, T. WO/018276, xxxx.
4) Takahara, J. P.; Setoyama, T. WO/0249999, xxxx.
5
48
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development
10.1021/op900001p CCC: $40.75  2009 American Chemical Society
Published on Web 03/18/2009

Chart 1. Internal reaction temperature of the oxidation of
methylacrylate to 3,3-dimethoxy methyl propionate against
time, illustrating the exothermic nature of the reaction
Table 1. Minimization of the catalyst cost to the cost of
3,3-dimethoxy methyl propionateFeCl
2
the exothermic nature of the reaction and the above-mentioned
overoxidation of the product, especially at high conversions.
While studying the palladium-catalyzed oxidation of methy-
lacrylate in methanol using oxygen as oxidant, to form 3,3-
dimethoxy methyl propionate, we encountered all three of these
problems. The initial experiments were conducted using pure
oxygen, and for the sake of safety we were constrained to using
small robust autoclaves equipped with bursting discs. High substrate
and catalyst concentrations resulted in the internal reaction tem-
perature rising considerably faster than the temperature of the
external heating source due to the exothermic nature of the
oxidation. This effect was exacerbated by Teflon inserts used to
protect the autoclave against the corrosive catalyst solution.
Containing the reaction mixture in a good insulator such as Teflon
led to runaway reactions with internal temperatures exceeding the
external heating bath by up to 80 °C (see Chart 1).
a
Arbitrary unit.
methylacrylate ) 8.9 and Pd/Cu/Fe ) 1/230/230. Because the
Na PdCl was used in such a small amount compared to the
2
4
substrate and it is one of the cheapest palladium sources
available, we decided not to investigate its contribution to the
product cost price. The Cu and Fe cocatalyst on the other hand,
are used in larger amounts and contribute significantly to product
price (Table 1).
From the several Cu and Fe combinations used it can be
seen that the activity of the system increases with decreasing
amount of chloride anion. Taking the price of the catalyst
components into consideration it appears that the combination
of elemental copper and FeCl
3
gives the lowest contribution to
In this particular experiment methylacrylate was reacted with
the Pd/Cu/Fe catalyst at 0.9 MPa oxygen pressure while heating
the autoclave at 45 °C. In about 10 min the internal reaction
temperature reached 120 °C. At this point the reaction started
to stop as oxygen depletion led to palladium-catalyst deactiva-
tion by palladium-black formation. In addition, product selectiv-
ity was decreased due to overoxidation and other side reactions.
Experiments like this on large scale could very easily lead to
dangerous situations involving pressure buildup and explosions.
In this contribution we report our investigations of the
relationships between reaction parameters such as catalyst
composition, catalyst ratio, oxygen pressure, temperature,
catalyst stability and product selectivity. Combining our un-
derstanding of these factors with data obtained concerning
explosion safety, we successfully scaled the reaction up to 2 L
scale. The reaction kinetics are in good agreement with our
proposed mechanism, and this can be used to model different
reaction conditions. We believe that we can use the model to
scale up this process to 1000 tonnes per year.
3
,3-dimethoxy methyl propionate cost price. The physical form
of metallic copper used as catalyst component does not influence
the reaction negatively. This is because the catalyst solution is
prepared prior to use by mixing FeCl
methanol. FeCl is a strong enough oxidant to oxidize elemental
Cu, causing it to dissolve. In all the future experiments we used
the combination of Na PdCl , FeCl and Cu as the catalyst
combination.
3
and metallic copper in
3
2
4
3
In the previous experiment the Pd/Cu/Fe catalyst ratio of
/230/230 was chosen arbitrarily. We investigated the effect
1
of changing the Pd/Cu/Fe ratio on the reaction rate at different
oxygen pressures. The conditions were as follows; Methylacry-
late was oxidized in methanol to 3,3-dimethoxy methyl propi-
2 4 3
onate at 70 °C using the Na PdCl , FeCl , Cu catalyst
combination. The Cu/Pd and Fe/Pd molar ratios were varied
from 100 to 800, and reactions were done at oxygen pressures
of 0.2 MPa, 0.3 and 0.4 MPa (Chart 2a). At (Cu, Fe)/Pd molar
ratios of lower than 400 the reaction rate is strongly dependent
on the oxygen pressure and catalyst ratio. In other words, under
these conditions the reaction is first order in (Cu, Fe)/Pd ratio
and oxygen pressure and zero order in palladium and substrate
concentration. At (Cu, Fe)/Pd molar ratios higher than 400 the
reaction rate is independent of the oxygen pressure. Under these
conditions the reaction is zero order in (Cu, Fe)/Pd molar ratio
and oxygen pressure, while it is first order in palladium and
substrate concentration. This has significant implications for
further development of this process as we now can keep high
reaction rates even at low oxygen pressure, because the reaction
is zero order in oxygen at (Cu, Fe)/Pd molar ratios of higher
Results and Discussion
Small-Scale Reaction Optimisation. The first part of the
optimization was to reduce the cost contribution of the catalyst
to the price of the final product to a minimum. The conditions
for catalyst optimization involved the oxidation of methylacry-
late in a methanol solution containing a Pd/Cu/Fe catalyst at
7
0 °C and 0.2 MPa oxygen pressure while stirring at 1000 rpm
on a 50 mL scale. The molar ratio of the reaction components
was as follows; methylacrylate/Na PdCl ) 40000, methanol/
2
4
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
549

Chart 2. Investigation of reaction parameters: (a) effect of catalyst composition on reaction rate at different pressures; (b)
product selectivity at different reaction temperatures but with same reaction profiles; (c) conversion curves at different MeOH/
MA ratios; (d) selectivities at different MeOH/MA ratios
than 400. For example, using this high (Cu, Fe)/Pd molar ratio
catalyst solution at an industrial scale would reduce the
manufacturing capacity required. Furthermore, the catalyst
performance will be less sensitive to fluctuations of partial
oxygen pressure due to, for example, inefficient stirring. In the
oxidation of methylacrylate to 3,3-dimethoxy methyl propionate.
Under the acidic reaction conditions the intermediate 3-meth-
oxyacrylate equilibrates with water and the hemiacetal, 3-meth-
oxy-3-hydroxy methyl propionate. This hemiacetal is unstable
under these reactions conditions and undergoes side reactions,
which are primarily irreversible overoxidations. This occurs
either on the hemiacetal itself or by deprotection to the free
aldehyde followed by oxidation. All of these reactions rates will
increase with temperature, but the irreversible nature of the
oxidation will pull the equilibrium to overoxidized byproduct,
reducing the yield of the desired product. Increasing the amount
of methanol in the system should reduce hemiacetal formation,
thereby blocking this overoxidation route and increasing the
overall yield of the reaction. The results of variation of
methanol/methylacrylate ratio on activity and selectivity are
depicted in Chart 2c and Chart 2d, respectively. Oxidizing
methylacrylate in a methanol solution containing a Pd/Cu/Fe/
methylacrylate molar ratio of 1/500/500/25000 at an oxygen
pressure of 0.2 MPa using different methanol/methylacrylate
ratios does not show any differences in the conversion curves.
As predicted, at the high conversion of 90% using a low
methanol/methylacrylate ratio of 4.7 a drastic decrease in 3,3-
dimethoxy methyl propionate selectivity is observed, falling as
low as 50%. Increasing the methanol/methylacrylate ratio to
32 gives a selectivity of 90% at more then 94% conversion.
2 4 3
following experiments we will use Na PdCl /FeCl /Cu molar
ratios of 1/500/500 using an oxygen pressure of 0.2 MPa.
An attempt was made to find the optimum temperature for
this reaction through the comparison of reaction profiles
measuring selectivity against time at different temperatures. This
was done by careful adjustment of the methylacrylate/Pd ratio.
It appeared that, when the methylacrylate/Pd ratio is decreased
from 40000 to 25000 while decreasing the reaction temperature
from 80 to 70 °C, the same conversion curve is obtained (Chart
2b).
For up to 60% methylacrylate conversion for both the 70
and 80 °C reactions, selectivity is nearly quantitative and the
yield curve approximates the conversion curve. From 60%
conversion to about 95% conversion selectivity to product starts
to decrease in both cases, but the decrease in selectivity is more
pronounced in the 80 °C experiment, leading to a product yield
lower than 60% at 95% conversion. Product yield at 95%
conversion in the case of the 70 °C reaction is above 75%. The
reason for the decrease in selectivity at higher conversions can
be explained by the equimolar coproduction of water in the
5
50
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development

Chart 3. Behavior of reaction rate with decreasing oxygen
partial pressure at constant total pressure
explosion limits. The reaction could now be tested on a 2 L
scale. The reaction was done by heating a 2 L mixture of
methylacrylate in methanol (methanol/methylacrylate ) 16) in
a 3-L autoclave to 70 °C while feeding a N
2 2
/O mixture
(
containing 13% oxygen) at 0.61 MPa and stirring at 1000 rpm.
When the mixture was at reaction temperature, the reaction was
started by injecting a 10 mL methanol solution containing the
catalyst components (Na PdCl /Cu/FeCl ) 1/500/500, methy-
2 4 3
lacrylate/Pd ) 25000). The reaction was allowed to proceed
for 9 h, after which the methylacrylate conversion was 95%,
and 3,3-dimethoxy methyl propionate selectivity was 86%. The
reaction mixture also contained minor amounts of methyl
formate, dimethoxymethane and 3-methoxy methylacrylate.
After cooling, 48.1 g 18.5 wt % solution of NaOH was added
to neutralize the reaction solution. This was a 3-fold excess to
However, this high ratio of 32 makes the reaction very dilute
and will result in very poor plant utilization. A methanol/
methylacrylate ratio of 16 is sufficient for the large-scale
experiments as selectivity drops just a few percent compared
to the ratio of 32.
So far, all the experiments were conducted using pure
oxygen under pressure, resulting in conditions where an
explosion could occur. For the sake of safety we investigated
ensure complete neutralization. Other bases, such as NaHCO
or Na CO could also effectively be used for this neutralization
but suffer from the disadvantage of generating stoichiometric
quantities of CO If the crude material is acidic, side reactions
3
2
3
2
will occur during distillation, such as the demethoxylation of
the 3,3-dimethoxy methyl propionate to form 3-methoxy
methylacrylate. The reaction mixture (∼1600 g) was transferred
to a distillation flask, and a 726 g fraction containing methanol,
methylacrylate, some water, methyl formate and dimethyl ether
was distilled while heating from 42 to 50 °C at 160 mmHg. A
second fraction weighing 324 g was distilled while heating from
50 to 61 °C at 160 mmHg which contained mainly methanol,
some water and a small amount 3,3-dimethoxy methyl propi-
onate. At this point, the contents of the flask separate into two
layers. It proved to be beneficial to separate the water layer
from the organic phase before continuing the distillation.
Distilling the two-phase system gives a slight decrease in yield
comparing to first-phase separation and then product distillation.
Furthermore, distilling the two-phase system leaves an alkaline
catalyst cake in the reaction flask which proved to be difficult
to remove. Analysis of the 161 g water layer showed that it
contained more than 95% of the used catalyst. In principle this
catalyst can be reused after workup, but we did not attempt to
investigate this as calculations show that on a 1000 tonnes/
year scale that the financial benefit would be minimal. On the
other hand, there may be an environmental case for recycling
the catalyst in case this process goes into industrial production.
After removal of the water layer the distillation of product was
continued while increasing the temperature from 61 to 91 °C
at 20 mmHg, giving a 20 g forerun product fraction, followed
by a 207 g product fraction which boiled at 79.9-80.1 °C at
bubbling with pressurized N
methylacrylate molar ratio of 1/500/500/25000 at 70 °C.
Reducing the total N /O pressure from 0.9 MPa to 0.29 Mpa
and simultaneously increasing oxygen concentration from
2.9% to 40% and consequently keeping the oxygen partial
pressure constant at 0.116 MPa, gives a constant reaction rate
2 2
/O mixtures with a Pd/Cu/Fe/
2
2
1
-1
of 0.58 h . Decreasing the oxygen partial pressure from 0.116
to 0.031 MPa by decreasing total pressure from 0.90 to 0.24
MPa with constant oxygen concentration of 12.9% led to a
-1
decrease in reaction rate from 0.54 to 0.28 h . A linear decrease
in reaction rate is observed (Chart 3) when the total pressure is
kept constant at 0.90 MPa and the total oxygen partial pressure
is decreased from 0.12 to 0.03 MPa by reducing the oxygen
concentration from 13 to 3.4%.
The linear decrease in reaction rate means that, even when
the oxygen concentration is very low in the reaction solution,
no deactivation of the palladium catalyst takes place. This is of
major importance for larger-scale reactions where the N
reaction mixture feed will have an oxygen concentration just
below the explosion limit, but the outlet N /O mixture will be
2 2
/O
2
2
much lower in oxygen due to the consumption during the
reaction. This outlet oxygen concentration must be high enough
to ensure that active palladium catalyst is not deactivated by
palladium black formation. The results from Chart 3 show us
outlet gas oxygen concentrations of as low as 3% can be
tolerated without deactivation of the catalyst.
Explosion limits were determined experimentally on both a
starting reaction mixture of 14% methylacrylate and 86%
methanol and on a product mixture 1% methylacrylate, 73%
methanol, 23% 3,3-dimethoxy methyl propionate and 3% water
at 75 °C and a total gas pressure of 0.45 MPa. The result was
for the starting reaction mixture no explosion could occur at
oxygen percentages below 17.7%, and for the final reaction
mixture the value was 13.8%.
20 mmHg. The remaining catalyst layer in the flask was easily
washed off with water. The mass balance of distillation of the
total reaction mixture was 98.6%, while the mass balance of
the combined forerun and main fraction of 3,3-dimethoxy
methyl propionate was virtually 100%, giving an isolated yield
of 82% based on methylacrylate charged. GC purity of the 3,3-
dimethoxy methyl propionate was 99%. In this distillation 90%
of methanol is recovered which can be distilled and be reused.
Reaction Simulation Studies As Aid for Further Scale-
up. The reaction mechanism scheme of palladium-catalyzed
oxidation of methylacrylate to 3,3-dimethoxy methyl propionate
is presented in Scheme 2.
Large-Scale Methylacrylate Oxidation Reaction and
Workup. So far we have optimized the catalyst composition,
temperature, pressure, methanol/methylacrylate ratio, and we
know how we can carry the reaction out safely in terms of
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
551

Scheme 2. Reaction mechanism of palladium-catalyzed oxidation of methylacrylate to 3,3-dimethoxy methyl propionate.
The palladium oxidizes the methylacrylate in the presence
of methanol to form 3-methoxy methylacrylate and an equimo-
lar amount of water with a reaction rate R3. The reduced Pd is
reoxidized by the Fe/Cu cocatalyst with a reaction rate R2 giving
a reduced Fe/Cu cocatalyst which is reoxidized by oxygen gas.
The oxidized palladium catalyst can also form byproduct
originating from the oxidation of two molecules of methanol
to form methylformate or from the oxidation of three molecules
of methanol to form dimethoxymethane. The 3-methoxy me-
thylacrylate reacts either with a molecule of methanol to give
the desired product 3,3-dimethoxy methyl propionate or with a
molecule of water to form 3-hydroxy-3-methoxy methyl pro-
pionate. Both product molecules can lead to the formation of
impurities, giving a decrease in product selectivity. It should
be noted that the hemiacetal is less stable than the full acetal
and is thus more prone to giving overoxidized byproduct. This
proposed reaction mechanism in Scheme 2 was fitted with the
obtained experimental data using the reaction equations belong-
ing to the different reaction pathways together with the fitted
values of activation energy and pre-exponential factors as are
shown in Scheme 3. The reaction is 1.32 order in methylacrylate
and 2.42 order in methanol.
tion prompted us to investigate in the laboratory and the results
are shown in Chart 2a (above). This is a good example how
the simulation of reaction conditions led to improvement of the
design of the catalytic system. We showed that fitting our
reaction kinetics with our proposed reaction mechanism gave
good agreement with experimental data, and we are now able
to simulate different reaction conditions, possibly using this tool
for further scale-up to a 1000 tonnes/y process.
Conclusion
In conclusion, we developed a safe synthetic route for the
production of 3,3-dimethoxy methyl propionate by a palladium-
catalyzed oxidation of methylacrylate in methanol using oxygen
as oxidant. In 2 L scale experiments we obtained an 82% yield
of 3,3-dimethoxy methyl propionate based on methylacrylate
charged. Purity of the obtained product was 99%. Solvent and
catalyst recovery were as high as 90% and 95%, respectively.
This synthetic procedure will aid the development of adhesives,
pharmaceuticals, agrochemicals and functionalized polymers.
Experimental section
The experimental data fit very well to the proposed reaction
mechanism in Scheme 2. This gives us a good model for the
palladium-catalyzed oxidation of methylacrylate to 3,3-dimethoxy
methyl propionate. As an example the results from a reaction
using a Pd/Cu/Fe/methylacrylate molar ratio of 1/500/500/25000
at 70 °C and oxygen pressure of 0.2 MPa at a methanol/
methylacrylate ratio of 8.90 are compared with data generated
by the model. (Chart 4).
We used the model to predict how the reaction would
perform with different ratios of Cu and Fe to Pd. This showed
that with a low ratio of Cu and Fe to Pd the reaction rate
becomes first order with respect to the amount of Cu and Fe
and oxygen and is zero order in palladium and methylacrylate.
The reverse is true in the case of a high ratio of Cu and Fe.
Hence, at high ratio of Cu and Fe to Pd the reaction rate is
independent of the oxygen pressure. This means that under those
conditions it may be possible to drop to a safe level of oxygen
without deactivating the catalyst. The results from this simula-
General. All reagents and solvents were obtained from
commercial sources and were purified appropriately if necessary.
Purity of starting materials and products and progress of reaction
were determined by GC. GC was performed on a Varian Star
3600 gas chromatograph with a flame-ionization detector and
a Chrompak CP Sil 5 CB wide-bore column (50 m × 0.53
mm) using nitrogen as carrier gas.
Small-Scale Experiments. A representative small-scale ex-
periment is given below. The example chosen is a reaction using
a Pd/Cu/Fe/methylacrylate molar ratio of 1/500/500/25000 at
70 °C and oxygen pressure of 0.2 MPa at a methanol/
methylacrylate ratio of 8.90 (the experiment shown in Chart
4).
A 1.296 mg quantity of Na
methanol, and 140.1 mg Cu and 358.2 mg FeCl
2
PdCl
4
was dissolved in 0.26 g
were dissolved
3
in 1.16 g methanol. The two solutions were combined and
charged into an HPLC-pump with outflow into an autoclave.
The Teflon-lined autoclave was charged with 9.486 g methy-
5
52
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development

Scheme 3. Reaction equations and the corresponding activation energies and pre-experimental factors
Chart 4. Experimental data compared to simulated data
lacrylate, 0.5 g decane (internal standard) and 30 g methanol
and heated to 70 °C while stirring at 1000 rpm under an oxygen
pressure of 0.2 MPa (and kept at those conditions throughout
the reaction). When the reaction mixture temperature had
reached 70 °C, the catalyst solution was injected, the reaction
was continued and samples were taken at stated times. GC
analysis after 10 h reaction time showed 94.8% methylacrylate
conversion with a 3,3-dimethoxy methyl propionate selectivity
of 70%.
2. Litre Scale. A 31.8 mg quantity of Na
dissolved in 6.32 g methanol, and 3.44 g Cu and 8.79 g
FeCl were dissolved in 28 mL methanol. The two
solutions were combined and charged into an HPLC-pump
with outflow into an autoclave.
A 233 g quantity of methylacrylate and 1363 g
methanol were charged into a glass, 3 L autoclave,
equipped with a condenser, catalyst feeding tube, gas
bubbler and a heating jacket at 70 °C while dosing a 13%
2 4
PdCl was
3
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
553

O
2
87% N
2
mixture at a total pressure of 0.61 MPa and
separated from the distillation mixture. Distillation was
continued by heating to 91 °C at 20 mmHg pressure while
collecting a forerun of 20.1 g of 3,3-dimethoxy methyl
propionate, followed by a constant boiling fraction of
207.5 g of 3,3-dimethoxy methyl propionate at 79.9-80.1
°C at 20 mmHg. A 39.2 g quantity of residue remained
in the flask, and it was easily dissolved in water. The trap
between collection flasks and oil-pump contained a 95.5 g
methanol/water mixture. Yield of the combined fractions
was 82% with a purity of 99%.
stirring at 1000 rpm. When reaction temperature was
reached, the catalyst solution was injected. The reaction
was continued for 9 h under the conditions mentioned
above. After cooling to room temperature, 48.1 g of 18.5
wt % aqueous NaOH was added, and 1596 g of the
reaction mixture was transferred to a 3 L distillation flask.
The flask was heated to 50 °C at 160 mmHg while a 726 g
fraction containing methanol, methylacrylate, water, meth-
yl formate and dimethyl ether was taken off. A second
fraction of 324 g was obtained while heating from 50 to
6
1 °C at 160 mmHg which contained mainly methanol,
Received for review January 5, 2009.
OP900001P
some water and a small amount of 3,3-dimethoxy methyl
propionate. Stirring was stopped, and 161.1 g water was
5
54
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development