Calcium pantothenate. Part 3.1 Process for the biologically active enantiomer of the same via selective crystallization and racemization

Article abstract of DOI:10.1021/op800189g

Optically pure calcium (R)-pantothenate has been obtained from the racemic compound via direct fractional crystallization of the single enantiomer. Efficient racemization of the biologically inactive calcium (S)-pantothenate without its decomposition to pantolactone has been developed, resulting in a simple, complete, and efficient technology. The results elaborated in the laboratory have been then successfully applied in the industrial scale process.

Full text of DOI:10.1021/op800189g

Organic Process Research & Development 2008, 12, 1238–1244
Calcium Pantothenate. Part 3.1 Process for the Biologically Active Enantiomer of
the Same via Selective Crystallization and Racemization
Ludwik Synoradzki,*,† Halina Hajmowicz,† Jerzy Wisialski,† Arkadiusz Mizerski,‡ and Tomasz Rowicki†
Laboratory of Technological Processes, Faculty of Chemistry, Warsaw UniVersity of Technology, ul. Noakowskiego 3,
00-664 Warszawa, Poland, and Grodzisk Pharmaceutical Works Polfa Ltd., ul. Poniatowskiego 5,
05-825 Grodzisk Mazowiecki, Poland
Abstract:
industrial purposes. The fundamental requirement for such a
process is the existence of conglomerates instead of the racemic
crystals in the solid phase. Fortunately, PTTCa possesses this
unique property. Particularly, its solvate containing four mol-
ecules of methanol and one of water crystallizes as a single
enantiomer from the racemic solution when seeded with
optically pure crystals. To date, the only drawback of the PTTCa
resolution has been the lack in the literature of an efficient direct
racemization method. According to the well-known procedures
the PTTCa racemization rather may be made after its hydrolysis
to pantolactone, which could be then racemized and reused in
(R,S)-PTTCa synthesis.7 Such a process, however, unreasonably
elongates the whole technology. Fortunately, an effective
procedure for direct PTTCa racemization has been developed
in our laboratory and hence allowing to overcome the problem.
Considering these facts we have undertaken an attempt to
develop an advantageous technology of calcium (R)-pantoth-
enate via direct resolution of the racemate. Since the research
has been carried out in co-operation with Grodzisk Pharma-
ceutical Works Polfa Ltd., the results obtained in the laboratory
have been subsequently realized in industrial scale reactors (up
to 3 m3), with considerable scale-up factors (crystallization 1500,
racemization 15000).
Optically pure calcium (R)-pantothenate has been obtained from
the racemic compound via direct fractional crystallization of the
single enantiomer. Efficient racemization of the biologically inactive
calcium (S)-pantothenate without its decomposition to pantolactone
has been developed, resulting in a simple, complete, and efficient
technology. The results elaborated in the laboratory have been
then successfully applied in the industrial scale process.
Introduction
Calcium (R)-pantothenate ((R)-PTTCa) and (R)-panthenol
are the most important commercial precursors of (R)-pantothenic
acid, a compound possessing a vitamin activity, denoted also
as vitamin B5. The significance of calcium (R)-pantothenate as
an ingredient of pharmaceutical and cosmetic compositions as
well as food and feed supplements caused the methods of its
production attract much attention since the first total synthesis
by Stiller.2 The most frequently used is the condensation of
ꢀ-alanine calcium salt with (R)-pantolactone,3 although the
separation of the racemic pantothenic acid with chiral amines
or by direct crystallization of its salts have also been described.4,5
Lately, the importance of biosynthetical methods has also
increased (Scheme 1).6
Among the methods shown in (Scheme 1), direct separation
of enantiomers is an extraordinarily attractive alternative for
Results and Discussion
Crystalline calcium pantothenate forms a number of poly-
morphs and solvates, depending on the solvent and temperature
applied for its crystallization.5b,8 Although the processes de-
scribed allow separation of the enantiomers of calcium pan-
tothenate, they suffer from some drawbacks when thoroughly
analyzed from the industrial point of view.5 Since the PTTCa
crystallizes as a methanol-water (4:1) solvate, the starting
solution must contain some water, with the minimum level
depending on the PTTCa concentration. For typical PTTCa
concentration of 30%, the solution should contain at least 0.9%
of water. The amount of water employed in the described
processes was slightly over the minimum to achieve satisfactory
yield of the product, since an increase of water content
simultaneously increases the PTTCa solubility. We have found
the process with such low water concentration to be, however,
* To
whom
correspondence
should
be
addressed.
E-mail:
Ludwik.Synoradzki@ch.pw.edu.pl. Fax: +48(22)6255317.
† Warsaw University of Technology.
‡ Grodzisk Pharmaceutical Works Polfa Ltd.
(1) For parts 1 and 2 see, respectively, (a) Rowicki, T.; Synoradzki, L.;
Włostowski, M. Ind. Eng. Chem. Res. 2006, 45, 1259–1265. (b)
Synoradzki, L.; Rowicki, T.; Włostowski, M. Org. Process Res. DeV.
2006, 10, 103–108.
(2) Stiller, E. T.; Harris, S. A.; Finkelstein, J.; Keresztesy, J. C.; Folkers,
K. J. Am. Chem. Soc. 1940, 62, 1785–1790.
(3) Paust, J.; Pfohl, S.; Reif, W.; Schmidt, W. Liebigs Ann. Chem. 1978,
1024–1029.
(4) (a) Kuhn, N.; Wieland, T. Chem. Ber. 1940, 73, 971. (b) Kuhn, N.;
Wieland, T. Chem. Ber. 1940, 73, 1134. (c) Stiller, E. T.; Wiley, P. F.
J. Am. Chem. Soc. 1941, 63, 1237.
(5) (a) Okuda, N.; Kuniyoshi, I.; Tsukamoto, H. GB Patent 1,124,619,
1968; Chem. Abstr. 1968, 69, 95968h. (b) Inagaki, M. Chem. Pharm.
Bull. 1977, 25, 1001–1009.
(6) See for example: (a) Hikichi, Y.; Moriya, T.; Nogami, I.; Miki, H.;
Yamaguchi, T. EP Patent 590,857, 1994; Chem. Abstr. 1994, 121,
55989b. (b) Nishimura, S.; Miki, H.; Matsumoto, J.; Shibutani, K.;
Yada, H. WO Patent 96/33283, 1996; Chem. Abstr. 1997, 126, 6565e.
(c) Moriya, T.; Hikichi, Y.; Moriya, Y.; Yamaguchi, T. WO Patent
97/10340, 1997; Chem. Abstr. 1997, 126, 263217d. (d) Yocum, R.;
Patterson, T.; Pero, J. G.; Hermann, T. WO Patent 02/061,108, 2002;
Chem. Abstr. 2002, 137, 15936g. (e) Beck, C.; Harz, H.-P. WO Patent
02/072,857, 2002; Chem. Abstr. 2002, 137, 231482x.
(7) (a) Klein, H. C.; Kapp, R. U.S. Patent 2,688,027, 1954; Chem. Abstr.
1955, 49, 13289f. (b) Klein, H. C.; Kapp, R. U.S. Patent 2,743,284,
1956; Chem. Abstr. 1956, 50, 16838e. (c) Garbarini, S. M.; Manz,
W. K.; McCormick, J. R. D. U.S. Patent 2,789,987, 1957; Chem. Abstr.
1957, 51, 12965g.
(8) Inagaki, M.; Tukamoto, H.; Akazawa, O. Chem. Pharm. Bull. 1976,
24, 3097–3102.
1238
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development
10.1021/op800189g CCC: $40.75
2008 American Chemical Society
Published on Web 11/01/2008

Scheme 1. Methods of calcium (R)-pantothenate synthesis.
Figure 1. Solubility of enantiomerically pure and racemic calcium pantothenate as a function of temperature and water content in
the solvent.
difficult to control because of at least two reasons. First, the
crystallization rate is relatively high; thus, the optimal time,
when the precipitate should be removed from the mother liquor
is relatively short, since the second enantiomer begins to
crystallize nearly immediately. It means that even a little delay
in crystal separation can cause considerable deterioration of the
product quality. Second, the initial solution is highly presaturated
so spontaneous crystallization may occur, especially if there is
a delay between the solution preparation and seeding with the
proper crystal form. In case of low water concentration not only
does the methanol-water solvate crystallize, but also an R-form
of PTTCa may crystallize from the mother liquor.9 The latter
crystal form causes rapid crystallization with a significant
thermal effect (increase of temperature by a few degrees has
been observed). In both cases an optically inactive precipitate
is a natural result.
disturbances in the conditions may cause failure in the quality
of the product. Hence, we have examined a number of factors
influencing the process to find the optimal conditions suitable
for a large-scale procedure.
We have found the water content in the solution to be the
most important factor determining the crystallization. Studies
on the PTTCa solubility showed that high water content is more
advantageous for the resolution process. The difference in
solubility between optically pure and racemic PTTCa increases
with an increase of water content in the solvent (Figure 1), thus
offering a better yield of single enantiomer crystallization.
Moreover, the general increase of solubility enables carrying
out the process at a higher PTTCa concentration that results in
higher total output of the product in a single step.
During our work we have found that optimal water content
is between 20 and 30%. An increase in water level, when
compared to the literature data (1-3%),5 prevents the spontane-
ous crystallization of the R-form of PTTCa and enables
application of ∼30% of a supersaturated PTTCa solution in
methanol-water mixture as a starting material. Solutions of
higher PTTCa concentration are difficult to operate because of
their high viscosity. The starting solution may be easily prepared
by dissolving the amorphous (R,S)-PTTCa in water, then adding
As can be seen, even from a brief analysis as well as from
our preliminary results an extraordinary accuracy is required
to realize the literature process successfully, and even little
(9) The R form is the thermodynamically most stable form of PTTCa; it
crystallizes as needles from methanol as well as other solvents. In
contrary, PTTCa water-methanol solvate (PTTCa·4MeOH·H2O)
crystallizes as prisms. Those and other crystal forms of PTTCa have
already been described in literature.5b,8
Vol. 12, No. 6, 2008 / Organic Process Research & Development
•
1239

Figure 2. Observed rotation of PTTCa mother liquor versus time for various amounts and sizes of seeds. All experiments were run
at 8 °C, with the starting solution containing 30% of PTTCa and 25% of water. (*Starting from racemic PTTCa.)
the desired amount of methanol, maintaining at 35-40 °C for
∼15 min. However, before seeding with a single enantiomer
solvate, the solution must be filtered to remove traces of the
PTTCa R-form that always remains as a fine suspension.
Relinquishment of this filtration results in uncontrolled crystal-
lization of the PTTCa R-form during cooling or standing,
whereas the filtered solutions could be stored for at least 18 h
at room temperature.
been run, giving (R)- and (S)-enantiomers alternatively (26-33%
yield of crystallization in a single step). The (S)-PTTCa has
been racemized in 5 batches in the recycling loop (83-93%
yield of racemization in a single step) (Scheme 2). Product
amounts and optical rotation received from some batches is
shown in Table 1. The consumption of raw material was 1.2
kg (R,S)-PTTCa for 1 kg (R)-PTTCa, the overall yield 83%,
quality in compliance with the European Pharmacopoeia
(Ph. Eur.) and the United States Pharmacopoeia (USP).
Special care should be taken only with the preparation and
handling of starting solution and seeding. Since the succeeding
batches are run without cleaning of the reactors, the filtration
off the PTTCa R-form is extremely important to ensure proper
mixture behavior during the crystallization. The solvate crystals
used as seeding must be stored wet at low temperature, i.e. under
5 °C. At higher temperatures slow-phase metamorphosis occurs,
resulting in the appearance of the R-form.9
The direct racemization of the biologically inactive (S)-
PTTCa is critical for the economic effectiveness of the whole
process. Although such a process has already been described
in the literature,10 we have found the conditions presented
difficult to repeat. In particular, we have never succeeded in
(S)-PTTCa racemization with calcium methoxide in boiling
methanol.
Application of sodium methoxide indeed allows to racemize
(S)-PTTCa under the reflux of methanol as described, due to
the difference in solubility between respective pantothenates
but not due to the different basic properties of calcium and
sodium methoxides. Contrary to sodium pantothenate, calcium
pantothenate is only slightly soluble in methanol so that at reflux
the racemization observed is negligible. The application of
sodium methoxide is unfortunately inconvenient for techno-
logical purposes, since sodium must be removed before (R,S)-
PTTCa crystallization.
The thus prepared solution can be then seeded with a single
enantiomer solvate to start the crystallization. The crystallization
rate depends mainly on four factors: temperature, seeding
particle distribution, the amount of the PTTCa in the starting
solution, and the enantiomeric excess of the PTTCa. The process
has been easily monitored by measuring the optical activity
(observed rotation) of the mother solution. We have found that
optimally it should be carried out at 8 °C, while the crystal-
lization is completed in ∼3 h. Increasing the temperature up to
12 °C causes the process to be twice as long. Decreasing the
temperature obviously increases the crystallization rate, thus
curtailing the process. Such an alternation makes, however, no
advantage since spontaneous crystallization of the second
enantiomer occurs rapidly, thus decreasing the optical purity
of the crystals. To complete the crystallization within 3 h, 0.5%
of well-ground solvate has been typically used as seed crystals
(amount calculated on PTTCa in solution, 6.5 µm average size).
The amount of seed crystals can be even decreased to 0.1%
with no change in the crystallization rate, if the solvate
micronized in a jetmill (3.1 µm average size) has been used.
The optical activity of the starting solution also influences the
crystallization rate. The higher the optical activity of the solution,
the higher the crystallization rate at the beginning; however,
the conditions applied enable completion of the process in ∼3
h even, when starting from the racemic solution (Figure 2).
The optimal conditions, i.e. crystallization temperature 8 °C,
PTTCa and water content in solution 28-31% and 23-26%,
respectively, and 3 µm seeding (0.1% calculated on PTTCa)
has been therefore applied in the industrial scale process.
Thirteen batches for (R)- and 12 batches for (S)-PTTCa have
(10) Okuda, N.; Kuniyoshi, I.; Kamada, M.; Nakagawa, K. (Daiichi Seiyaku
Co., Ltd.). FR Patent 1,524,617, 1968; Chem. Abstr. 1969, 70, 58291g
(for corresponding Japanese patent).
1240
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development

Scheme 2. Block diagram of the technological process
Table 1. Product amounts and specific rotation obtained
of water. However, the water content in the product from the
optical resolution process was always not lower than 2.9% since
it has been a solvate containing four molecules of methanol
and one molecule of water. Therefore, the starting material for
racemization must be dried to attain the amorphous, water-free
compound.
from some batches of the industrial process
R)-PTTCa
S)-PTTCa
R,S)-PTTCA
[R]D
batch
no.
[R]D
[R]D
(deg)
(kg) (deg)/ee (%) (kg)
(kg)
(deg)
Fractional Crystallization of (R,S)-PTTCa
(690-765 kg per batch)
Another problem has been the behavior of the reaction
mixture during the first part of racemization. The dried (S)-
PTTCa used for racemization has been an amorphous form that
readily dissolved in methanol before the reaction. However,
during heating up, the R-form of (S)-PTTCa crystallized rapidly
from the solution, causing problems with stirring and heat
transfer. A powerful stirrer had to be used to ensure that all the
suspension was stirred to avoid local overheating near to the
reactor’s surface. This is important, since the decomposition
of PTTCa has been found to be irreversible in the temperature
range around 120 °C, even in the absence of water. The reason
for such a phenomenon has been the further decomposition of
ꢀ-alanine calcium salt to form ammonia and the respective
acrylate, which polymerized in the reaction conditions. As a
result, a considerable increase of the reaction mixture viscosity
as well as a decrease in ꢀ-alanine concentration have been
observed. Excessive elongation of the reaction time caused the
problems mentioned to appear even at 110 °C. Therefore, the
time of heating the reaction mixture should be curtailed to the
minimum, the temperature of the heating medium should be
near to but not exceeding 90 °C, and the mixing has to be
effective enough to prevent local overheating. At around 90
°C the reaction mixture has become homogeneous again due
to racemization that has started at this temperature, transforming
the slightly soluble (S)-PTTCa into much better soluble racemic
PTTCa (Figure 3). Typically, the time of heating the suspension
to 90 °C has been in the range of 30-60 min, and the
racemization at 110 °C has been carried out for further 1.5 h.
The racemized PTTCa with the 84-90% yield has been
recycled to the resolution loop.
1
2
12
13
198
212
195
228
+26.5/99.6
+25.8/97.0
+26.8/100
+26.0/97.7
195
203
227
-
-25.8
-23.8
-23.4
Racemization of (S)-PTTCa (300 kg per batch)
1
3
5
258
248
254
-0.5
-1.2
-0.7
After solving some technological problems we have found
the conditions for the quantitative racemization of (S)-PTTCa
directly with calcium methoxide as a base. Calorimetric curve
for (S)-PTTCa suspension in methanolic calcium methoxide
exhibits an endothermic peak at 80-100 °C, certainly being a
result of the simultaneous dissolving/racemization process
(Figure 3).
Effective racemization can therefore be carried out in
substantially anhydrous conditions, at above 90 °C, which
means that the reaction proceeds under elevated pressure. The
necessity of water absence is obvious since it causes calcium
methoxide decomposition and irreversible hydrolysis of PTTCa
in basic as well as acidic conditions (Scheme 3). Elevated
temperature also causes partial decomposition of PTTCa even
in the absence of water, but in this case the process is reversible;
thus, cooling down the mixture and maintaining it at room
temperature allows recovery of the compound in high yield.
We have found that to perform the reaction quantitatively
and to recover ∼90% of the starting material the reaction
conditions should be as follows. The water content in methanol
has to be not higher than 0.04%; exceeding this value up to
0.5% increased the PTTCa decomposition by 10-15% (Figure
4). Therefore, to keep the water level reasonable in the reaction
mixture, the (S)-PTTCa used has to contain not more than 0.3%
Vol. 12, No. 6, 2008 / Organic Process Research & Development
•
1241

Figure 3. Calorimetric curve for (S)-PTTCa suspension in methanolic calcium methoxide. Unsubtracted heat flow endo up (mW):
Steps: (1) hold for 1:0 min at 25.00 °C, (2) heat from 25.00 to 150.00 at 10.00 °C/min.
Scheme 3. Reversible and irreversible decomposition of PTTCa under various conditions
Conclusions
The complete technology of (R)-PTTCa including racemate
ogy has been successfully scaled up in 3 m3 reactors with an
overall yield 83% of (R)-PTTCa (based on (R,S)-PTTCa used)
and quality in compliance with Ph. Eur. and USP.
resolution via selective crystallization of a single enantiomer
and direct racemization of biologically inactive (S)-enantiomer
in the recycling loop has been elaborated. A self-controlling
resolution process consisted of the alternate crystallization of
(R)- and (S)-enantiomers, in a methanol-water solvent system
with high content of the later. The (S)-PTTCa racemization
process was catalyzed by calcium methoxide in dry methanol
at elevated temperature under pressure. The complete technol-
Experimental Section
Commercially available solvents and reagents were used
without further purification. Calcium (R,S)-pantothenate was
supplied by Grodzisk Pharmaceutical Works Polfa Ltd. Water
and calcium content were determined by Karl Fisher method
and complexometric titration, respectively. Optical rotation was
measured on an Optical Activity Ltd. automatic polarimeter.
Determination of the crystals particle size distribution was
performed with Malvern Instruments Ltd. automatic particle
fraction analyzer. Calorimetric analysis has been carried out with
Perkin-Elmer PYRIS 1 calorimeter.
Separation of Calcium Pantothenate into Optical Iso-
mers: Laboratory Procedure. 1. Starting Cycle. Water (232.0
g) and methanol (350.0 g) were added to 300.0 g (0.629 mol)
of racemic PTTCa. The mixture was then maintained at 35 °C
until PTTCa dissolved (∼15 min). The insoluble material was
separated by pressure filtration, washed twice with 40 cm3 (31.6
g) of methanol, and the washes were combined with the mother
liquor. The solution obtained (982 g, water content 24.0%,
PTTCa content 29.7%) was transferred to a 2 dm3 reactor,
cooled to 8 °C with stirring (Bola-Stirrer shafts with one rigid
Figure 4. Decomposition of PTTCa during racemization as a
function of temperature and the water content in methanol used
for reaction.
1242
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development

paddle, 300 rpm), and seeded with 1.5 g (0.003 mol) of calcium
(R)-pantothenate solvate suspended in 3 g of methanol. Samples
were taken out and crystallization of (R)-PTTCa was monitored
by measurement of the observed rotation of the mother liquor.
After 175 min (when R0 reached -1.51°) the precipitate was
filtered off, washed twice with 40 cm3 (31.6 g) of cold methanol,
and dried at room temperature under reduced pressure (20 hPa).
Obtained was 81.7 g of pure calcium (R)-pantothenate, having
RD ) +27.0°. The washes were concentrated and combined
with the mother liquor to give 833 g of PTTCa solution (water
content 25.6%, PTTCa content 25.1%, R0 ) -1.42°).
2. Repeatable Cycle. (R,S)-PTTCa (89.0 g, 0.187 mol) was
added to the mother liquor from the previous cycle. The
suspension was then maintained at 35 °C until PTTCa was
dissolved (∼15 min) and purified by pressure filtration as
described above. The solution obtained (986 g, water content
26.09%, PTTCa content 30.2%, R0 ) -1.22°) was transferred
to a 2 dm3 reactor, cooled to 8 °C with stirring, and seeded
with 2.4 g (0.005 mol) of calcium (S)-pantothenate solvate
suspended in 4.8 g of methanol. Samples were taken out as in
the starting cycle, and crystallization of (S)-PTTCa was
monitored by measurement of the observed rotation of the
mother liquor until the R0 reached +1.5°. The precipitate was
then filtered off, washed twice with 40 cm3 (31.6 g) of cold
methanol, and dried at room temperature under reduced pressure
(20 hPa). Pure calcium (S)-pantothenate (80.4 g), having RD )
-26.5°, was obtained. The washes were concentrated and
combined with the mother liquor to give 843 g of PTTCa
solution (water content 26.1%, PTTCa content 25.8%, R0 )
+1.36°). The dissolving-crystallization cycles were then
repeated by alternate seeding with each of the enantiomers,
resulting in calcium (R)-pantothenate in every odd cycle, and
calcium (S)-pantothenate in every even cycle. Subsequent
analyses of R0, water, and Ca content are necessary for the
proper amount of (R,S)-PTTCa, methanol, and water being
added in every cycle, thus keeping the monitored values within
the safe range.
Separation of Calcium Pantothenate into Optical Iso-
mers: Technical Scale Procedure. Fifteen kg (31.5 mol) of
calcium (R)-pantothenate and 750 kg (1577 mol) of calcium
(R,S)-pantothenate along with methanol and water (1040 and
540 kg, respectively) were placed in a 3 m3 reactor and heated
up to 35 °C for ∼15 min to dissolve the solid. The insoluble
material was immediately separated by pressure filtration with
2.5 kg of diatomite, washed with 40 dm3 of water, and the
washes were combined with the mother liquor. The solution
obtained (∼2385 kg, water content ∼24%, PTTCa content
30.3%) was transferred to a 3 m3 reactor, cooled to 8 °C, and
seeded with 0.8 kg (1.68 mol) of calcium (R)-pantothenate
solvate suspended in 1.6 kg of methanol. Samples were taken
out, and crystallization of (R)-PTTCa was monitored by
measurement of the mother liquor observed rotation. When the
R0 reached -1.5° the precipitate was filtered off in a centrifuge,
washed with 50 dm3 (39.5 kg) of methanol, and dried in a
fluidized drier to give ∼200 kg of calcium (R)-pantothenate
(RD ) +26.5°). The washes were combined with the mother
liquor to give 2050 kg of PTTCa solution to which 200 kg of
(R,S)-PTTCa was added. The mixture was maintained at 35
°C for 15 min to dissolve the solid, then the PTTCa and water
content were adjusted to 31% and 24% respectively. While still
warm, the solution was pressure filtrated as previously described,
transferred to the reactor, cooled to 8 °C, and seeded with 0.8
kg (1.68 mol) of calcium (S)-pantothenate solvate suspended
in 1.6 kg of methanol. Samples were taken out, and crystal-
lization of (S)-PTTCa was monitored by measurement of the
mother liquor observed rotation. When the R0 reached +1.5°
the precipitate was filtered off in a centrifuge (without washing)
and dried in a fluidized drier to give 200 kg of calcium (S)-
pantothenate (RD ) -24°). The dissolving-crystallization
cycles were then repeated by alternate seeding with each of
the enantiomers, resulting in calcium (R)-pantothenate in every
odd cycle and calcium (S)-pantothenate in every even cycle.
Subsequent analyses of R0, water, and Ca content are necessary
for the proper amount of (R,S)-PTTCa, methanol, and water
being added in every cycle, thus keeping the monitored values
within the safe range.
Racemization of Calcium (S)-Pantothenate: Laboratory
Procedure. Due to the necessity of keeping the water and
carbon dioxide concentration under the allowable level all
reactions and operations must be carried out in the atmosphere
or under the flush of an inert gas. Dry methanol (70 g) and
calcium (0.330 g, 0.008 mol) were stirred with gentle heating
until the hydrogen emission ceased and then refluxed for a while
to make the calcium methoxide formation complete. The
solution was cooled to room temperature and transferred to a
200 cm3 pressure reactor (stainless steel with Bola-magnetic
stirring bars, dumbbell-shape) containing dry, amorphous
calcium (S)-pantothenate (35.3 g, 0.074 mol) and (R,S)-
pantolactone (1.54 g, 0.003 mol). (Amorphous PTTCa can be
obtained from the methanol-water solvate by heating it to 110
°C under 2.5 kPa for 1.5-2 h.) The reactor was closed and
immediately placed in an oil bath heated previously to 110 °C.
After the temperature and pressure in the reactor reached
105-108 °C and 0.30-0.32 MPa, respectively, the solution
was mixed at these conditions for further 1.5 h. The reactor
was then cooled to room temperature, and the transparent or
slightly cloudy solution was transferred to a round-bottomed
flask and stirred overnight. Water was added to the mixture to
a final concentration of 3.5%, the pH was adjusted to 8-9 with
acetic acid, and the solution was stirred with activated carbon
for 20 min. The insoluble material was filtered off, and the clear
solution was transferred to the reactor, cooled to 3 °C, and
seeded with 0.24 g (0.0005 mol) of the (R,S)-PTTCa. Crystal-
lization was then carried out at -13 °C for at least 12 h, the
precipitate was filtered off and dried at 20-80 °C for 8 h under
reduced pressure (20 hPa). Calcium (R,S)-pantothenate (41.3
g) that could be used as the starting material for the selective
enantiomer crystallization was obtained.
Racemization of Calcium (S)-Pantothenate: Technical
Scale Procedure. As in the laboratory procedure, all reactions
and operations must be carried out in the atmosphere or under
the flush of an inert gas due to the necessity of keeping the
water and carbon dioxide concentration under the allowable
level. Dry methanol (632 kg, 800 dm3) and 3.6 kg (90 mol) of
calcium were placed in a 3 m3 reactor with jacket (heating with
steam 140 °C) and refluxed for 0.5 h. After cooling the calcium
Vol. 12, No. 6, 2008 / Organic Process Research & Development
•
1243

methoxide suspension to 20 °C, 300 kg (630 mol) of (S)-PTTCa
was added to the reactor, and the mixture was quickly heated
to 105-108 °C (heating with steam 180 °C). The racemization
was carried out at 105-108 °C and pressure of 0.30-0.32 MPa
for 1.5 h. The solution of racemic PTTCa was then cooled to
20 °C and maintained at this temperature for 20 h. Water was
added to the mixture to a final concentration of 3.5%, the pH
was adjusted to 8-9 with acetic acid, and the solution was
stirred with activated carbon for 20 min. The mixture was
filtered through a pad of diatomite, and the clear solution was
transferred to another reactor, cooled to 3 °C, seeded with 1.4
kg of the (R,S)-PTTCa, and crystallized at -13 °C for 12 h.
The precipitate was filtered off and dried in a fluidized drier to
give 237 kg of amorphous calcium (R,S)-pantothenate that was
further used as a starting material for the selective enantiomer
crystallization.
Acknowledgment
Financial support from the State Committee for Scientific
Research (Grant 2508/C.T09-3/99), and from Warsaw Univer-
sity of Technology, Faculty of Chemistry is gratefully acknowl-
edged. We kindly thank Grodzisk Pharmaceutical Works Polfa
Ltd. for fruitful co-operation, and for permission for publication
of the data concerning calcium pantothenate technology.
Received for review August 7, 2008.
OP800189G
1244
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development