Note: Descriptions are shown in the official language in which they were submitted.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
Process for the Preparation of
3,5,5-Trimethyl-Cyclohex-2-eae-1,4-Dioae
Field of the Invention
This invention relates to an improved process for the
preparation of 3,5,5-trimethylcyclohex-2-ene-1,4-dione
(ketoisophorone) by oxidation of 3,5,5-trimethyl-cyclohex-
3-en-1-one (~i-isophorone).
Background of the Invention
Ketoisophorone (KIP) is an important intermediate product
in the synthesis of trimethylhydroquinone or
trimethylhydroquinone esters, which in turn are an
intermediate product in the synthesis of vitamin E. KIP is
furthermore an intermediate product for the preparation of
various carotenoids, such as, for example, astaxanthine,
zeaxanthine and canthaxanthine.
The production of KIP by oxidation of oc-isophorone or (3-
isophorone ((3-IP) is known.
According to Hosokawa et al. CChem. Lett., 1983, 1081-
1082), oxidation of o~-isophorone to KIP is achieved by
tert-butyl hydroperoxide in the presence of 10 mold
palladium acetate and an auxiliary base in yields of a
maximum of 55~. In addition to the high amount of catalyst
and the low yield which can be achieved, in particular the
use of the expensive oxidizing agent tert-butyl
hydroperoxide makes the process unattractive for an
industrial reaction.
The same applies to the reaction according to WO 96/154094,
in which compounds of the sub-group metals of groups Ib,
V'b, VIb or VIII, inter alia vanadium(V) oxide and iron(III)
chloride, are used as catalysts for the oxidation with
tert-butyl hydroperoxide.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
2
Oxidation by means of oxygen or an oxygen-containing gas
mixture is considerably more economical here. DE 25 26 851
discloses a corresponding process for the oxidation of oc-
isophorone under catalysis by e.g. phospho- or
silicomolybdic acid, possibly with the addition of
copper(II) salts or molybdenum trioxide. However, very
long reaction times of 96 hours (h) and high temperatures
of 100°-C are necessary to achieve complete conversions.
The yield which can be achieved under these conditions is
only 45~.
Comparable results with the use of phosphomolybdic acid as
the catalyst are described by Freer et al. CChem. Lett.,
1984, 2031-2032) and in EP 0 425 976.
According to JP Sho 61-19164'5/1986, the selectivity of this
process can be increased to up to 96~ by employing organic
amines or alkali metal salts as additives, in addition to
catalytic amounts of phospho- or silicomolybdic acid.
However, the maximum conversion which can be achieved in
this way is only 59~, which necessitates a working up of
the product solution which is expensive and not very
desirable from economic aspects.
According to JP Hei 10-182543/1998, similar results are
also achieved when a catalyst system comprising salts of
the platinum metals and hetero-polyacids or salts thereof
is employed. However, the high price of platinum metal
salts additionally contributes towards the unprofitability
of this process.
According to DE 24 59 148, acetylacetonates of various
transition metals, preferably vanadium acetylacetonate, can
be employed as catalysts of the oxidation of a-isophorone
by molecular oxygen. Here also, however, long reaction
times of more than 40 h and high temperatures of 100°-C to
130qC are necessary, and only unsatisfactory yields of 20~
to 40~ are obtained.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
3
Overall, no economical process is known for the direct
oxidation of a-isophorone to KIP, since the yield of the
reaction by the processes described is low. The oxidation
of (3-IP, which can be obtained from a-isophorone by known
processes, can be carried out considerably more
efficiently. The most economical variant of this oxidation
here also is the procedure using oxygen or an oxygen-
containing gas as the oxidizing agent.
Thus, according to DE 24 57 157, the yields in the
oxidation reaction catalyzed by transition metal
acetylacetonate can be increased to up to 56~ by prior
isomerization, mediated by sodium acetate, of a-isophorone
to (3-IP, at the same time somewhat lower temperatures of
25°-C - 75°-C and shorter reaction times of > 26 h being
necessary. Nevertheless, these results are still
unsatisfactory.
According to DE 25 15 304, a significant improvement was to
be achieved by adding pyridine or pyridine derivatives, in
addition to the transition metal acetylacetonate, as a
result of which the reaction time is shortened to 2 h to
3.5 h and KIP is obtained in yields of 70~ to 80~, and in
one example even 91%. It proves to be a disadvantage here
that large amounts of base (up to 250 mold, based on the (3-
IP) and of catalyst (up to 10 per cent by weight (wt.~),
based on the (3-IP) are required.
According to DE 38 42 547, the above-mentioned amounts of
base and catalyst employed can be reduced significantly
with comparable KIP yields if specifically copper
acetylacetonate is used in the presence of pyridine.
The oxidation of ~i-IP to KIP catalyzed by active charcoal
with the addition of triethylamine (DE 26 57 386) or
heterocyclic nitrogen bases, such as pyridine (JP Hei 11-
49717/1999) in acetone as the solvent is also described.
However, the flash point of the solvent acetone of -17~C is
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
4
exceeded by far at the reaction temperatures of 100°-C,
which is not acceptable for an industrial realization for
safety reasons.
An improvement in the oxidation of (3-IP to KIP in respect
both of the amounts of catalyst employed and of the
reaction conditions (low temperature, short reaction time)
and the yields which can be achieved at high conversions
(high selectivities) resulted from the use of transition
metal complexes with polydentate ligands as catalysts of
the reaction.
EP-B 0 311 408 employs Mn tetraphenylporphyrin as the
catalyst, in addition to triethylamine as a base and water
as an additive. Optimum crude yields of 98o are obtained
here using a solvent mixture of ethylene glycol dimethyl
ether and methylene chloride. The use of a solvent mixture
is not appropriate for an industrial realization both from
economic and from safety aspects. -
In a publication by the same authors which appeared later
(Ito et al., Synthesis 1997, 2, 153-155), a selectivity of
only a maximum of 93~ when ethylene glycol dimethyl ether
is employed as the solvent under optimum conditions is
reported. The use of ethylene glycol dimethyl ether as the
solvent is not desirable for an industrial realization
because of the low flash point of -6~C and the associated
risk of explosion. Furthermore, the use of porphyrin
catalysts, which are very expensive to synthesize and must
be prepared in a separate two-stage process in low yields,
appears to be a disadvantage. If manganese(III)salen
chloride is used as the catalyst, a yield of only 81~ is
obtained under the conditions described.
According to JP Sho 64-9015011989 and JP Hei 01-
175955/1989, yields of KIP of just above 90~ can be
achieved using manganese(III)-salen compounds and
derivatives under optimum, selected conditions. Ethylene
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
glycol dimethyl ether is also used here as the solvent,
which in turn involves the disadvantages already described.
DE 26 10 254 discloses. the oxidation of (3-IP to KIP using
manganese(II)- or cobalt-salen or related compounds as the
5 catalyst. A selectivity of the reaction catalyzed by
manganese-salen of 100 is reported here in one example.
Under the conditions described, this corresponds to a
space/time yield of 0.09 kilograms of KIP per hour-liter
(kg/h*1). Some years after the patent application was laid
open, this result is no longer mentioned by the same
authors in a scientific publication (M. Constantini et al.,
J. Mol. Catal., 1980, 7, 89-97). A maximum KIP yield of
850, which can be achieved with manganese-salen as the
oxidation catalyst under optimum conditions, is reported
there. According to this publication, it was possible to
increase the space/time yield to 0.16 kg KIP/(h*1) under
optimum conditions, but this is still unsatisfactory.
An investigation of the influence of the solvent on the
space/time yield and the selectivity of the reaction is
also presented in this publication. The authors come to
the conclusion that if aprotic solvents are employed, as
the polarity and basicity increase the rate of reaction
indeed increases, but not the selectivity. Various ethers
and ketones, in particular ethylene glycol dimethyl ether
and acetone, are acknowledged as optimum solvents for the
selectivity. However, the use of acetone with a flash
point of -17QC or ethylene glycol dimethyl ether with a
flash point of -6°-C as the solvent for an industrial
oxidation process in the temperature range described is
eliminated by consideration of safety aspects, since the
risk of explosion of the reaction mixture can be avoided
only by very expensive safety precautions, which in turn
means a considerable economic outlay. Moreover, the use of
ethers as solvents in the presence of bases and oxygen
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
6
results generally in the risk of the formation of highly
explosive peroxides, which involves a further safety risk.
US 5,874,632 addresses in detail for the first time the
connection between the educt concentration in the reaction
mixture and the selectivity which can be achieved. It is
found that in the reaction catalyzed by manganese-salen in
diethylene glycol dimethyl ether (diglyme) in the presence
of triethylamine as the base and water as a reaction
accelerator, good selectivities of 91~ can be achieved only
at low (3-IP concentrations of a maximum of 10 wt.~. A
dramatic drop in selectivity takes place at a higher
concentration. The addition of a catalytically active
substance from the class of organic acids with pKa values
of between 2 and 7 or the corresponding aldehydes, various
enolizable compounds or lithium sulfate, in particular the
addition of acetylacetone, is disclosed here as a solution
to the problem, ethers and ketones, in particular diglyme,
being used as the solvent. Significantly higher educt
concentrations can be realized by this measure, without a
significant reduction in the selectivity having to be
accepted. Space/time yields of up to 0.34 kg KIP/(h*1) can
be realized in this way.
For the profitability of a catalytic process, in particular
using a homogeneous catalyst, in addition to the spaceltime
yield (product in kg per hour and reaction volume in
liters) and the amount of catalyst needed, the choice of
additives and solvent is of central importance. When the
known processes are reproduced, it is found that by using
the preferred solvents and additives described,
considerable amounts of by-products are formed in the
reaction matrix, which have the effect of a reduction in
catalyst output during recycling and therefore a reduction
in the selectivity for the desired product KIP if the
process is operated in circulation. A considerable
consumption both of the solvent and of the additive is
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
7
associated with this, and at the same time the need arises
to remove the by-products formed by suitable process
technology operations with considerable expenditure on
apparatus.
Of the above-mentioned processes, the reaction of (3-IP with
oxygen in the presence of manganese-salen and metal
porphyrin and phthalocyanine catalysts of the composition
known in the prior art shows the best selectivities and
space/time yields. The process according to US 5,874,632
appears to be the most suitable for industrial realization
on the basis of the high space/time yield which can be
achieved, the relatively simple catalyst, which is
inexpensive to prepare and which already leads to complete
conversions with high selectivities when added in small
amounts, and on the basis of the relatively high flash
point of the preferred solvent diglyme of 53°-C.
Nevertheless, the process has some disadvantages, which are
to be explained in the following.
Tnrhen this process is operated as a circulation process with
distillative working up and removal of the solvent diglyme
from the product, it is found that the use of diglyme as
the solvents indeed renders possible very good yields and
selectivities of > 90~, but decomposition of the solvent,
base and additive leads to the formation of carboxylic
acids, which are recycled to the reactor again during
recycling of the solvent or must be separated off with
considerable outlay. Formic acid and acetic acid are to be
mentioned in particular here, but also methoxyacetic acid
and 2-methoxyethoxyacetic acid. These by-products become
concentrated in the circulation solvent and lead to a
continuous decrease in the selectivity of the oxidation
reaction.
Another considerable disadvantage for an industrial
procedure is the instability of the catalyst components
premixed in the solvent. Since direct metering of the
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
8
catalyst as a solid is usually to be avoided, it is
desirable to initially introduce the catalyst into the
reaction vessel in the solvent together with the catalyst
base and the catalytically active co-additives before the
reaction and to bring this solution continuously into
contact with (3-IP and to feed it to the reaction part. If
ethylene glycol ethers, which are optimum for the
selectivity, are used, a considerable aging of the catalyst
solution as the service life of the premixture increases is
found, which manifests itself in a drastic decrease in the
selectivity of the reaction in a continuous process.
Another problem which has so far not been solved
satisfactorily is the formation of by-products during the
oxidation, in particular hydroxyisophorone, which is formed
in an amount of between 50-20~ in the preparation by
conventional processes, depending on the reaction
procedure. Under optimum conditions, a minimum formation
of hydroxyisophorone of 5~ (based on the (3-IP employed) can
be obtained if manganese-salen is used as the catalyst in
the system NEt3/water/diglyme. The formation of by-
products in this order of magnitude is not desirable from
economic aspects.
The use of ethers as the solvent for the oxidation
reactions moreover involves the risk of formation of highly
explosive peroxides already described. Furthermore,
diglyme is a very expensive solvent, which has an adverse
effect on the preparation costs of the process, so that the
use of diglyme as the solvent for the reaction investigated
is not very desirable from economic aspects.
Summarizing, no satisfactory overall concept for the
industrial reaction has yet been found which takes into
account the following criteria simultaneously:
a) use of a favorable, readily accessible solvent,
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
9
b) use of a stable solvent which is inert under the
reaction conditions,
c) use of only one uniform solvent to avoid expensive
separations of substances during working up,
d) use of solvents which do not tend to form explosive
peroxides under the reaction conditions,
e) use of a suitable solvent which stabilizes the
reaction matrix (the solution of any additives and
catalyst) and thus forms a solution which is stable to
storage over a relatively long period of time.
Summary of the Invention
The object of the present invention is thus to discover, on
the basis of the prior art, a suitable, stable reaction
system, in particular solvent, which avoids the
disadvantages described for the processes already known,
which are in come cases considerable, while retaining or
increasing the good selectivities and reaction yields,
specifically in the catalyst system manganese-salen l
auxiliary base / optionally water / co-additive.
This invention provides an improved process for the
preparation of 3,5,5-trimethylcyclohex-2-ene-1,4-dione
(ketoisophorone, KIP) by oxidation of 3,5,5-trimethyl-
cyclohex-3-en-1-one ((3-isophorone, (3-IP) in the presence of
an oxidizing agent and a catalyst system comprising a
transition metal complex catalyst, an auxiliary base,
possibly water, and a catalytically active co-additive
chosen from the group consisting of:
1. an organic acid with a pKa of between 2 and 7 or the
corresponding aldehyde;
2. an aliphatic alcohol with 1-4 C-atoms or phenol;
3. compounds which can form an ,enol structure; and
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
4. lithium sulfate;
characterized in that carboxylic acid amides are employed
as the solvent.
Brief Description of the Figures
5 Figure 1 depicts diagram showing the KIP yield as a
function of the (3-IP concentration.
Figure 2 depicts diagram showing the KIP yield as a
function of the oxygen supplied.
Detailed Description of the Invention
10 The reaction is illustrated clearly in the following
equation:
O O
Catalyst: Transition metal complex/auxiliary
base%pt. water/co-additive
Oxidizing agent
Solvent: Carboxylic acid amides O
~i-Isophorone fCetoisophorone
Weak organic acids or bidentate complexing compounds have
proved to be particularly advantageous co-additives in
respect of the reaction kinetics. Particularly preferred
co-additives are acetic acid, butyric acid, salicylic acid,
oxalic acid, malonic acid, citric acid and further
aliphatic or aromatic mono-, di- or tricarboxylic acids.
Amino acids, such as e.g. glycine, leucine, methionine or
aspartic acid, are also suitable.
It has also been found that aliphatic alcohols, such as
methanol, ethanol, butanol, isobutanol and tert-butanol, or
phenol serve as the co-additive. Co-additives which can
form an enol structure, such as e. g. acetoacetic esters,
phenylacetone and, in particular, acetylacetone, are
particularly advantageous. Acetylacetone is particularly
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
11
preferably employed as the co-additive, since it
additionally allows higher reaction selectivities to be
achieved than with other suitable co-additives.
The use of acetylacetone has shown that higher
concentrations of (3-IP, based on the total amount of the
mixture, can be employed and therefore higher space/time
yields can be achieved without a serious reduction in the
selectivity occurring.
A molar ratio of the co-additive to the catalyst of 1:1 to
100:1, preferably 4:1 to 40:1, based on the catalyst, can
be employed.
The outstanding properties of carboxylic acid amides as the
solvent for the oxidation reaction under consideration have
not hitherto been acknowledged and found in any of the
publications. Instead, ethers and ketones have preferably
been mentioned and employed as the optimum solvent in all
the disclosures relevant to the present invention,
resulting in the above-mentioned disadvantages.
Suitable carboxylic acid amides in the process according to
the invention are dimethylformamide, diethylformamide,
dimethylacetamide, diethylacetamide or mixtures thereof.
Dimethylformamide is particularly preferred as the
carboxylic acid amide. The amount in which the suitable
carboxylic acid amides can be used is not critical for
carrying out the process according to the invention, but it
is preferable to use carboxylic acid amides in amounts of
50 wt.~ to 95 wt.~, preferably 65 wt.~ to 85 wt.~, based on
the total amount of the reaction mixture.
It has been found that the KIP selectivity which can be
achieved in the oxidation in carboxylic acid amides
unexpectedly reacts very sensitively to a lower supply of
oxygen. In a pilot bubble column with very efficient
introduction of gas via a fine-pored frit, the enormous
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
12
potential of this solvent class began to become visible
only from the relatively high metering in of oxygen of 0.5
liters of oxygen per hour and gram of (3-IP (a specific
value for the pilot apparatus used). A drastic drop in
selectivity took place below this value. Such an effect is
not to be observed to this extent with the solvents
conventionally employed, such as diglyme.
To achieve optimum yields, the oxygen supply in the
carboxylic acid amides even had to increased up to at least
about 0.3 liter of oxygen per hour and gram of (3-IP. Not
taking account of or not knowing these circumstances thus
unavoidably leads to poor results in the reaction procedure
in carboxylic acid amides, from which a supposedly lower
suitability of carboxylic acid amides as the solvent for
the reaction under consideration compared with the ethers
and ketones described as having priority has been
incorrectly assumed.
The fact is that not only does carrying out the oxidation
in carboxylic acid amides allow KIP to be prepared in high
selectivities and yields with a simultaneously reduced
formation of by-products, but also in carboxylic acid
amides the catalyst system reacts considerably less
sensitively to the presence of carboxylic acids unavoidably
obtained when the process is operated in circulation than
is the case with ethers.
The catalyst system also has a significantly better
stability in carboxylic acid amides than in ethers, which
allows premixing of the reaction matrix in continuous
operation of the process, without a drop in the selectivity
of the reaction taking place over a period of time.
Another great advantage of carboxylic acid amides is that
the effect of the catalytically active co-additive
described on the educt concentration which can be realized
without a serious loss in selectivity is significantly more
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
13
pronounced than, for example, in diglyme, which results in
economic advantages in respect of the space/time yield of
the reaction. Thus, even at a ~i-IP concentration of
40 wt.~, a selectivity of more than 85o is still observed,
a value which can be achieved in diglyme only at educt
concentrations up to a maximum of 20 wt. o.
At the same time, due to the weakly basic properties of the
carboxylic acid amide solvent, the amount of auxiliary
based employed can be reduced down to 10 mold, based on the
(3-IP employed, without serious losses in the selectivity of
the reaction which can be achieved occurring.
Furthermore, the use of carboxylic acid amides as the
solvent, in particular inexpensive dimethylformamide,
offers an enormous economic advantage over the use of the
very expensive ethylene glycol ethers, such as diglyme and
ethylene glycol dimethyl ether, which have hitherto been
described as the preferred solvents for achieving high
selectivities.
To carry out the reaction, (3-IP is continuously or
discontinuously brought into contact with the reaction
matrix, which comprises the catalyst and the auxiliary
base, as well as a catalytically active co-additive and
optionally water and which is dissolved or suspended in a
carboxylic acid amide as the solvent, and reacted with
oxygen or an oxygen-containing gas mixture under normal
pressure or increased pressure.
Catalysts which are used are the transition metal-
containing complex catalysts mentioned in the prior art,
such as manganese-salen, manganese-tetraphenylporphyrin and
manganese-phthalocyanine, manganese-salen being preferred.
The catalyst is conventionally added in amounts of 0.001 to
3 wt.~, based on the (3-IP, preferably in amounts of 0.05 to
1 wt.~.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
14
The organic and inorganic bases known according to the
prior art can be used as the auxiliary base, such as e.g.
alkylamines, di- and trialkylamines, aromatic and aliphatic
heterocyclic bases, sodium or potassium hydroxide solution
or alcoholates, or a quaternary ammonium hydroxide,
preferably trialkylamines, in particular triethylamine.
These bases can be employed in conventional amounts, such
as e.g. 5 to 60 molo, based on the (3-IP, amounts of 10 to
35 molo being particularly preferred.
Carboxylic acid amides, such as e.g. dimethylformamide
(DMF), diethylformamide (DEFA) and the corresponding
acetamides, such as dimethylacetamide or diethylacetamide,
are employed as the solvent in the process according to the
invention. In a particularly preferred embodiment, the
reaction is carried out in dimethylformamide. The content
of carboxylic acid amide in the reaction mixture is
conventionally 50 wt.~ to 95 wt.~, and amounts of 65 wt.~
to 85 wt.~ are preferably employed.
The water content in the total reaction mixture can vary
between 0 and 30 wt. o. Without the addition of water, very
high selectivities are achieved, but with uneconomical ,
reaction times. Water is therefore preferably employed as
a reaction accelerator, in particular between 0.05 wt.~ and
wt.~, preferably between 0.5 and 20 wt.~, particularly
25 preferably between 0.5 wt.~ and 5 wt. o, based on the total
weight of the reaction mixture.
Oxidizing agents which can be employed in this invention
are oxygen or oxygen-containing gas mixtures, such as e.g.
air or oxygen diluted by addition of an inert gas, such as,
30 for example, nitrogen.
The reaction can be carried out under normal pressure or
increased pressure. For example, the reaction can be
carried out under between 1 and 12 bar, depending on the
volume content of oxygen in the oxidizing agent employed.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
The reaction temperature can be between -30~C and 80°-C,
preferably between 10~C and 45°-C.
The process according to the invention is simple to carry
out and gives the reaction product in a good yield and high
5 purity. The reaction product can be isolated from the
product mixture by the usual processes, in particular by
vacuum distillation.
The yields were determined on an HP 5890 or an HP 6890 gas
chromatograph using a J&W DB-5 capillary column of 30 m
10 length, 0.32 mm internal diameter and 1 Eim film thickness.
Diethylacetamide was used as the internal standard. KIP,
which was purified by distillation, was used as the
reference substance.
The HPLC measurements were carried out on a system
15 comprising a Biotronik BT 3035 UV detector, a Jasco 880 PU
pump and a Spectra Physics Chrom Jet integrator. The
column used was an RP 18, 5 ~,, 250 x 4 mm internal
diameter. The KIP reference substance described above was
used as the external standard.
The following examples are intended to illustrate the
invention in more detail.
Examples 1 to 12
58.0 g of the solvent stated in table 1, 1.2 g water,
0.16 g acetylacetone, 2.53 g triethylamine and the amount
of manganese-salen which can be seen from Table 1 are
initially introduced into a glass beaker and stirred for
either 15 min (variant A) or 16 h (variant B). 15.4 g (3-IP
are then added and the reaction mixture is stirred briefly
and transferred to a bubble column. It is gassed with
oxygen (12 1/h) for 2.5 hours at 35gC under normal pressure
and the yield of KIP is then determined by gas
chromatography with an internal standard. The results are
listed in Table 1.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
16
Comparative Examples A to F
58.0 g diglyme, 1.2 g water, 0.16 g acetylacetone, 2.53 g
triethylamine and the amount of manganese-salen which can
be seen from Table 1 are initially introduced into a glass
beaker and stirred for either 15 min (variant A) or 16 h
(variant B). 15.4 g (3-IP are then added and the reaction
mixture is stirred briefly and transferred to a bubble
column. It is gassed with oxygen (12 1/h) for 2.5 hours at
35°-C under normal pressure and the KIP yield is then
determined by gas chromatography with an internal standard.
The results are listed in Table 1.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
17
Table 1
Example/ Solvent Variant Amount of catalystYield of KIP
Comp. [wt.~ based on [~]
Example (3-
IP]
1 DMF A 0.2 91.4
2 DMF B 0.2 89.7
3 DEFA A 0.2 90.5
4 DEFA B 0.2 89.6
A diglyme A 0.2 90.3
B diglyme B 0.2 85.3
DMF A 0.3 92.2
6 DMF B 0.3 91.9
7 DEFA A 0.3 92.0
8 DEFA B 0.3 90.3
C diglyme A 0.3 90.1
D diglyme B 0.3 87.0
9 DMF A 0.4 92.8
DMF B 0.4 92.0
11 DEFA A 0.4 92.1
12 DEFA B 0.4 92.5
E diglyme A 0.4 90.5
F diglyme B 0.4 88.8
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
18
Examples 13 to 15
72.5 g DMF, 1.56 g water, 0.21 g acetylacetone, 3.18 g
triethylamine, the amounts of acid stated in Table 2 and
75 mg manganese-salen are initially introduced into a glass
beaker and stirred for 15 min. 19.25 g (3-IP are then added
and the reaction mixture is stirred briefly and transferred
to a bubble column. It is gassed with oxygen (16 1/h) for
2.5 hours at 35$C under normal pressure and the yield of
KIP is then determined by gas chromatography with an
internal standard. The results are listed in Table 2.
Comparative Examples G to I
72.5 g diglyme, 1.56 g water, 0.21 g acetylacetone, 3.18 g
triethylamine, the amounts of acid stated in Table 2 and
75 mg manganese-salen are initially introduced into a glass
beaker and stirred for 15 min. 19.25 g (3-IP are then added
and the reaction mixture is stirred briefly and transferred
to a bubble column. It is gassed with oxygen (16 1/h) for
2.5 hours at 35°-C under normal pressure and the yield of
KIP is then determined by gas chromatography with an
internal standard. The results are listed in Table 2.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
19
Table 2
Example! Solvent Acid added Yield of KIP
Comp. Example [mg/kg solvent] [
13 DMF formic acid: 2059 91.7
14 DMF acetic acid: 2011 88.5
15 DMF formic acid: 1031 92.3
acetic acid:'1007
G diglyme formic acid: 1999 84.6
H diglyme acetic acid: 1742 86.5
I diglyme formic acid: 1044 86.0
acetic acid: 1028
It has again been confirmed that higher yields can be
achieved with the process according to the invention.
Examples 16 to 22
About 100 g of a solution of (3-IP, triethylamine, water,
acetylacetone and manganese-salen in the concentrations
which can be read off from Table 3 in DMF are gassed with
oxygen (16 1/h) in a laboratory bubble column for 2.5 hours
at 35°-C under normal pressure and the yield of KIP is then
determined by gas chromatography with an internal standard.
The results are listed in Table 3.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
w
H
x
p
N v-I 01 L~ In O d~
O
O
G
N
r-1
I rl ~ O
N \ ~ d1 L~ 00 dl W -i
N ri 00 l~ Ll7 M v-I
~, M M LC7 L~ ~ ~ c-I
~,
~'
O
O
N '-I
U \ Lf1 M L~ N M Lf7 c-I
-I v-I rl N N M t!1
y
r
N
U
M a,'
r-I
fa ~I
rl
p i-W -I o~ 00 oW t1
C-~N
y p co 0o t t~ t t~
o O O O O O O
N
O O O o1 01 01 op
M M M N
5y
r-I N N N
0 0 0 0
~ 0 0 0
.rl
H
d~ CO G1 d N L!1 c-1
H ~ LI1 o tt1 c-I t0 OD l0
I
f~. O v-I v-I N N M It1
~
N
r-I
vO L~ 00 01 O ~-I N
c-I e"'I~-I c-I N N N
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
21
The yields of KIP in examples 16 to 22 were compared with
the yields of KIP according to table 3 of US Patent
5,874,632. The result is shown in Figure 1.
In Figure 1, it can be clearly seen that the yields of KIP
of this invention are significantly improved, i.e.
increased, at the same (3-IP concentrations, compared with
the prior art.
Examples 23 to 26
58.0 g dimethylformamide, 1.2 g water, 0.16 g
acetylacetone, 2.53 g triethylamine and 45 mg manganese-
salen are initially introduced into a glass beaker and
stirred for 15 min. 15.4 g (3-IP are then added and the
reaction mixture is stirred briefly and transferred to a
bubble column. It is gassed with oxygen in the metered
amount which can be seen from Table 4 for 2.5 hours at 35sC
under normal pressure and the yield of KIP is then
determined by HPLC with an external standard. The results
are listed in Table 4.
Comparative Examples J to M
58.0 g diglyme, 1.2 g water, 0.16 g acetylacetone, 2.53 g
triethylamine and 45 mg manganese-salen are initially
introduced into a glass beaker and stirred for 15 min.
15.4 g (3-IP are then added and the reaction mixture is
stirred briefly and transferred to a bubble column. It is
gassed with oxygen in the metered amount which can be seen
from Table 4 for 2.5 hours at 35°-C under normal pressure
and the yield of KIP is then determined by HPLC with an
external standard. The results are listed in Table 4.
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
22
Table 4
Example/ Solvent Oxygen metered Yield of
Com in ItIP
Exam
le
p. fl/hl fl/h*g ~i-IPl(~l
p
23 DMF 6 0.40 75.8
24 DMF 8 0.53 87.4
25 DMF 10 0.66 89.2
26 DMF 12 0.80 90.2
J diglyme 6 0.40 86.8
K diglyme 8 0.53 87.9
L diglyme 10 0.66 89.0
M diglyme 12 0.80 88.9
Examples 23 to 26 illustrate the enormous influence the
oxygen supply has on the ketoisophorone selectivity in the
solvent DMF which can be achieved. Comparative Examples
to M demonstrate that such an effect is to be observed to
only a small extent in the solvent diglyme, which is
particularly preferred according to the prior art. Not
knowing this sensitivity of the selectivity of the reaction
in DMF to a deficient supply of oxygen thus unavoidably
leads to poor results (Examples 23 to 24), from which a
supposedly lower suitability of carboxylic acid amides as
the solvent for the reaction under consideration compared
with the ethers and ketones described as having priority in
the prior art results.
On the other hand, if a sufficiently high supply of oxygen
(Examples 25 to 26) is ensured, the unexpectedly high
potential, described in the present invention, of
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
23
carboxylic acid amides as the solvent for the oxidation
under consideration becomes clear. See Figure 2.
Examples 27 to 30
72.5 g dimethylformamide, 1.5 g water, 0.22 g
acetylacetone, the amount of triethylamine which can be
seen from Table 5 and 75 mg manganese-salen are initially
introduced into a glass beaker and stirred for 15 min.
19.3 g (3-IP are then added and the reaction mixture is
stirred briefly and transferred to a bubble column. It is
gassed with oxygen (16 1/h) for 3 hours at 35°-C under
normal pressure and the yield of KIP is then determined by
gas chromatography with. an internal standard. The results
are listed in Table 5.
Comparative Example N
In this Comparative Example, diglyme was used as the
solvent instead of DMF with the same relative composition
of the solution as in Examples 27 to 30. The yield of KIP
was determined by HPLC with an external standard. The
results are listed in Table 5.
Table 5
Examplel SolventTriethylamine Yield of
concentration KIP
Comp.
Example [mol~ with [wt.~ with
respect to (3-IP]respect to (3-IP]
27 DMF 5 3.6 84.5
28 DMF 10 7.3 90.1
29 DMF 15 10.9 91.1
DMF 20 14.6 91.2
N diglyme11.4 8.3 86.8
CA 02473941 2004-07-21
WO 03/062184 PCT/EP02/00621
24
It has been proved again that better yields are achieved
with the process according to the invention, i.e. using
dimethylformamide as the solvent instead of diglyme. See
Example 28 versus Comparative Example N.
