A facile microwave-mediated drying process of thermally unstable / labile products

Article abstract of DOI:10.1021/op100102m

The drying behavior of (S)-N-acetylindoline-2-carboxylic acid, precipitated (1a, 17 wt %) and nonprecipitated (1b, 5 wt %), and N-acetyl-(S)-phenylalanine ((S)-2-acetamido-3-phenylpropanoic acid, 2), both pharmaceutical intermediates, and of cocarboxylase hydrochloride (thiamine pyrophosphate, 3), a coenzyme, a bioactive form of vitamin B₁, being a thermolabile substance, has been determined in straightforward drying setups. The method of supplying energy to the system had a profound influence on the drying rate and on the internal temperature of the samples during drying. The drying time of (S)-N-acetylindoline-2-carboxylic acid (1b) with the low moisture content (5 wt %) could be reduced by a factor 4 using microwave irradiation instead of conventional heating, while keeping the sample temperature under 35 °C. N-Acetyl-(S)-phenylalanine (2) with a higher moisture content (22 wt %) demonstrated a decrease in drying time by a factor 2.5 to 4 depending on the applied microwave powers. A reduction in drying time of the precipitated (S)-N-acetylindoline-2-carboxylic acid (1a, 17 wt % moisture) by a factor 2 was demonstrated for drying at 150 W of microwave irradiation instead of using a water bath at 70 °C. A dramatically shorter drying time by a factor 10 was found for cocarboxylase hydrochloride (3, 15 wt % water) on lab-scale which could be reproduced on pilot-plant scale. To achieve with conventional heating similar drying times as under microwave irradiation for the four examples, extremely high energy inputs should be applied, necessitating extremely high temperature differences between the heating source and the sample. The results reveal that microwave irradiation is less energy-consuming and is particularly useful for effective drying of thermally unstable materials in short periods of time.

Full text of DOI:10.1021/op100102m

Organic Process Research & Development 2010, 14, 1130–1139
A Facile Microwave-Mediated Drying Process of Thermally Unstable / Labile
Products
N. A. Pinchukova,*,† A. Yu. Voloshko,† O. V. Shyshkin,† V. A. Chebanov,† B. H. P. Van de Kruijs,‡ J. C. L. Arts,‡
M. H. C. L. Dressen,‡ J. Meuldijk,‡ J. A. J. M. Vekemans,‡ and L. A. Hulshof*,‡
National Academy of Sciences of Ukraine, Lenin AVenue, 60, KharkiV 61001, Ukraine, and EindhoVen UniVersity of
Technology, Den Dolech 2, 5612 AZ EindhoVen, The Netherlands
Abstract:
heat2-9 and has become a popular heating technique for various
processes including synthesis, solution concentration and drying
of food,10-14 fruits,15 chemicals,16,17 agricultural products,18
polymers,19 ceramics,20 pulp and paper,21 textiles,22 in mineral
processing23 as well as in wood processing industries.24
Although drying of pharmaceutical powders with microwave
heating has been shown to increase drying rates and product
stabilities during the drying process,25,26 application in the
pharmaceutical industry is still limited. The main property of
microwaves is to directly deliver energy from the source to the
microwave-absorbing sample. Field inhomogeneity of domestic
microwave ovens, culminating in uneven heating rates, dictate
the use of dedicated equipment for microwave processing.27
Conventionally heated drying processes are limited by the
thermal conductivity of the substances. When the thermal
conductivity is low, the drying process becomes slow, and
The drying behavior of (S)-N-acetylindoline-2-carboxylic acid,
precipitated (1a, 17 wt %) and nonprecipitated (1b, 5 wt %), and
N-acetyl-(S)-phenylalanine ((S)-2-acetamido-3-phenylpropanoic acid,
2), both pharmaceutical intermediates, and of cocarboxylase
hydrochloride (thiamine pyrophosphate, 3), a coenzyme, a bioac-
tive form of vitamin B1, being a thermolabile substance, has been
determined in straightforward drying setups. The method of
supplying energy to the system had a profound influence on the
drying rate and on the internal temperature of the samples during
drying. The drying time of (S)-N-acetylindoline-2-carboxylic acid
(1b) with the low moisture content (5 wt %) could be reduced by
a factor 4 using microwave irradiation instead of conventional
heating, while keeping the sample temperature under 35 °C.
N-Acetyl-(S)-phenylalanine (2) with a higher moisture content (22
wt %) demonstrated a decrease in drying time by a factor 2.5 to
4 depending on the applied microwave powers. A reduction in
drying time of the precipitated (S)-N-acetylindoline-2-carboxylic
acid (1a, 17 wt % moisture) by a factor 2 was demonstrated for
drying at 150 W of microwave irradiation instead of using a water
bath at 70 °C. A dramatically shorter drying time by a factor 10
was found for cocarboxylase hydrochloride (3, 15 wt % water)
on lab-scale which could be reproduced on pilot-plant scale. To
achieve with conventional heating similar drying times as under
microwave irradiation for the four examples, extremely high
energy inputs should be applied, necessitating extremely high
temperature differences between the heating source and the
sample. The results reveal that microwave irradiation is less
energy-consuming and is particularly useful for effective drying
of thermally unstable materials in short periods of time.
(2) Be´langer, J. M. R.; Pare´, J. R. J.; Poon, O.; Fairbridge, C.; Ng, S.;
Mutyala, S.; Hawkins, R. J. MicrowaVe Power Electromagn. Energy
2008, 42, 24.
(3) Loupy, A., Ed. MicrowaVes in Organic Synthesis, 2nd ed.; Wiley-
VCH: Weinheim, 2006.
(4) Kappe, C. O.; Stadler, A. MicrowaVes in Organic and Medicinal
Chemistry; Wiley-VCH: Weinheim, 2006.
(5) Polshettiwar, V.; Varma, R. S. Acc. Chem. Res. 2008, 41, 629.
(6) Kappe, C. O. Angew. Chem., Int. Ed. 2004, 43, 6250.
(7) Strauss, C. R.; Trainor, R. W. Aust. J. Chem. 1995, 48, 166.
(8) Varma, R. S. Green Chem. 1999, 1, 43.
(9) Varma, R. S. Pure Appl. Chem. 2001, 73, 193.
(10) Zhang, M.; Tang, J.; Mujumdar, A. S.; Wang, S. Trends Food Sci.
Technol. 2006, 17, 524.
(11) Torringa, E.; Esveld, E.; Scheewe, I.; Van den Berg, R.; Bartels, P. J.
Food Eng. 2001, 49, 185.
(12) Regier, M.; Knorzer, K.; Erle, U. Chem. Ing. Tech. 2004, 76, 424.
(13) Hamer, A.; Puschner, H. A. Chem. Ing. Tech. 1997, 69, 480.
(14) Puschner, H. A. Elektrotech. Z. B 1968, 20, 278.
(15) Wojdyło, A.; Figiel, A.; Oszmian´ski, J. J. Agric. Food Chem. 2009,
57, 1337.
(16) Esveld, E.; Chemat, F.; Van Haveren, J. Chem. Eng. Technol. 2000,
23, 279.
Introduction
(17) Esveld, E.; Chemat, F.; Van Haveren, J. Chem. Eng. Technol. 2000,
23, 429.
Drying is one of the oldest and most common unit operations
in chemical engineering and is an essential procedure for
purifying and isolating products.1 Drying is one of the highest
energy-consuming and most expensive processes in the phar-
maceutical industry. Microwave irradiation has attracted much
attention in different fields dedicated to studying benefits and
drawbacks of microwaves as an unconventional source of
(18) Askari, G. R.; Emam-Djomeh, Z.; Mousavi, S. M. Drying Technol.
2009, 27, 831.
(19) Bakass, M.; Mokhlisse, A.; Lallemant, M. Thermochim. Acta 2000,
356, 159.
(20) Clark, D. E.; Sutton, W. H. Annu. ReV. Mater. Sci. 1996, 26, 299.
(21) Kumar, P.; Mujumdar, A. S.; Koran, Z. Drying Technol. 1990, 8, 1061.
(22) Haghi, A. K. Asian J. Chem. 2005, 17, 639.
(23) Haque, K. E. Int. J. Miner. Process. 1999, 57, 1.
(24) Vongpradubchai, S.; Rattanadecho, P. Chem. Eng. Process. 2009, 48,
997.
* To whom correspondence should be addressed. E-mail: L.A.Hulshof@tue.nl
and pinchukova@isc.kharkov.com.
(25) Chee, S. N.; Johansen, A. L.; Gu, L.; Karlsen, J.; Heng, P. W. S.
Chem Pharm. Bull. 2005, 53, 770.
† SSI “Institute for Single Crystals” NAS of Ukraine, Kharkiv, Ukraine.
‡ Laboratory of Macromolecular and Organic Chemistry, Applied Organic
Chemistry.
(26) McMinn, W. A. M.; McLoughlin, C. M.; Magee, T. R. A. Powder
Technol. 2005, 153, 23.
(1) Mujumdar, A. S.; Devahastin, S. Fundamental Principles of Drying;
Exergex Corp: Brossard, 2002.
(27) Kelen, A.; Pallai-Varsanyi, E.; Ress, S.; Nagy, T.; Pintye-Hodi, K.
Eur. J. Pharm. Biopharm. 2006, 62, 101.
1130
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development
10.1021/op100102m  2010 American Chemical Society
Published on Web 08/24/2010

Figure 1. (S)-N-acetylindoline-2-carboxylic acid (1a, 1b), N-acetyl-(S)-phenylalanine (2), and cocarboxylase hydrochloride (3).
Figure 2. Setup for the drying experiments under conventional and microwave heating conditions. With a capillary tube, a constant
air purge provided a constant level of humidity in the apparatus.
Table 1. Temperatures, corresponding theoretical power
moreover, considerable temperature gradients may induce
inputs, and actual power inputs in the microwave-assisted
overheating of the sample surface which can be extremely risky
for thermosensitive materials from a quality and a safety point
of view. Vacuum drying is often used to reduce the boiling
temperatures of volatiles and to reduce the risk of overheating
as well. However, drying under reduced pressure becomes even
more challenging due to ineffective convection in many cases.
The problem of heat transfer may be solved by introducing
microwave dielectric heating.28-34
experimentsa
T (°C) (water bath temperatures)
45
60
75
QCH, estimated, t)0 (W)
QMW, used (W)
QMW, used effectively (W)
13
50
∼5
42
71
100
∼10
150
∼15
a MW: microwave heating; CH: conventional heating (oil-bath). See
Experimental Section for the calculation method.
The present paper deals with microwave-assisted vacuum
drying of two pharmaceutical intermediates, based on earlier
work,35 and of a coenzyme of vitamin B1 with different
thermosensitivities, see Figure 1. The first one, N-acetylindoline-
2-carboxylic acid, precipitated (1a) and nonprecipitated (1b),
an intermediate in the synthesis of perindopril (an ACE inhibitor
drug working to lower blood pressure),36,37 is a relatively
thermostable substance. The second intermediate, N-acetyl-(S)-
phenylalanine ((S)-2-acetamido-3-phenylpropanoic acid, 2), a
versatile building block for the synthesis of various active
pharmaceutical ingredients (APIs) and intermediates, is also a
relatively thermostable substance. The third, cocarboxylase
hydrochloride, thiamine pyrophosphate 3), a coenzyme and
bioactive form of vitamin B1, is a thermolabile substance. The
study was aimed at establishing the advantages of microwave
over conventional heating in terms of drying efficiency, product
quality, and energy savings.
Drying Behavior of (S)-N-Acetylindoline-2-carboxylic Acid
To determine the influence of microwave irradiation on the
drying of pharmaceutical intermediates and to clarify whether
its application beneficially influences the drying process, a
comparison of conventional with microwave heating was made
for this substance. The following straightforward drying pro-
cedure was selected for small-scale experimentation, see Figure
2. A three-necked flask, containing the static sample, was held
at a pressure of 5 kPa and heated by either a water bath or by
a multimode microwave apparatus. An air-inlet capillary tube
ensured a constant air flow over the sample. This setup
guaranteed a constant removal of evaporating moisture from
the upper layer of the sample and excluded mass transport
limitations in the gas phase. In this way, the diffusion rate of
the moisture in the sample and the evaporation rate of the
moisture governed the rate of the drying process.
(28) Boldor, D.; Balasubramanian, S.; Purohit, S.; Rusch, K. A. EnViron.
Sci. Technol. 2008, 42, 4121.
(29) Bowman, M. D.; Holcomb, J. L.; Kormos, C. M.; Leadbeater, N. E.;
Williams, V. A. Org. Process Res. DeV. 2008, 12, 41.
(30) Roberts, B. A.; Strauss, C. R. Acc. Chem. Res. 2005, 38, 653.
(31) Vaucher, S.; Unifantowicz, P.; Ricard, C.; Dubois, L.; Kuball, M.;
Catala-Civera, J.-M.; Bernard, D.; Stampanoni, M.; Nicula, R. Physica
B 2007, 398, 191.
(32) Salagnac, P.; Dutournie´, P.; Glouannec, P. Ind. Eng. Chem. Res. 2008,
47, 133.
(33) Hettenbach, K.; Am Ende, D. J.; Dias, E.; Breneck, S. J.; Laforte, C.;
Barnett, S. M. Org. Process Res. DeV. 2004, 8, 859.
To compare the different heating methods, the power
supplied by the water bath to the flask at t ) 0 was estimated
for the applied temperatures, see Experimental Section. The
actual power input of the water bath is lower than these values.
The water bath temperatures used and the microwave power
inputs are collected in Table 1.
(34) Kocakus¸ak, S.; Ko¨roglu, H. J.; Go¨zmen, T.; Savas¸ci, O. T.; Tolun, R.
Ind. Eng. Chem. Res. 1996, 35, 159.
(35) Dressen, M. H. C. L.; Van de Kruijs, B. H. P.; Meuldijk, J.; Vekemans,
J. A. J. M.; Hulshof, L. A. Org. Process Res. DeV. 2007, 11, 865.
(36) Sheldon, R. A. Chirotechnology: Industrial Synthesis of Optically
ActiVe Compounds; Dekker: New York, 1993.
(37) Souvie, J.-C., NoVel method for the synthesis of (S)-indoline-2-
carboxylic acid and application thereof in the synthesis of perindopril;
(Servier Lab). Patent number EP 1348684, 2003.
The energy efficiency of the microwave irradiation in the
used setup is variable during the drying process. Also the
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1131

input of 5 W). The seemingly very short, constant drying-rate
period could indicate that (S)-N-acetylindoline-2-carboxylic acid
mostly contained moisture in the interstitial space. Also in this
case, removal of pore moisture was dominant in time and in
energy consumption during the drying process. The drying time
could be drastically reduced by increasing the microwave power
input to 100 W (15 min drying time) and 150 W (10 min drying
time). Increasing the temperature of the water, and thus
increasing the power input supplied by the water bath to the
flask, also decreased the drying time.
Temperatures of 60 and 75 °C corresponded to a 25- and
20-min drying time for conventional heating, see Figure 5 (left).
A striking difference between both heating methods is the
temperature of the sample. Higher water bath temperatures also
increased the sample temperature, shown most pronounced for
the final equilibrium temperature at the end of the drying
process. To gain a similar drying rate as for the 150 W
microwave drying experiments, a temperature exceeding 200
°C should be applied for conventional drying. This would lead
to a sample temperature of more than 150 °C, which is
impractical and could be detrimental for the stability of the
sample.
The temperature measurements were performed at the center
of the sample for both heating methods. When applying a
conventional heat source, the sample is heated from the outside
which creates a temperature profile and a drying front. The
temperature has the lowest value at the center of the sample.
Therefore, the recorded temperatures of the sample with
conventional heating are lower than the average temperature.
This is in sharp contrast with drying under microwave irradia-
tion. The volumetric heating distributes the energy relatively
evenly over the sample, especially with low loss tangent
materials which have a large penetration depth. The insulation
of the central point of the sample by the surrounding material
causes a minimal heat loss in the center, and therefore, the
temperature in the center of the sample in the microwave
experiments is the highest. The dielectric properties of the
sample also determine the energy distribution in the sample,
see Figure 6.41
Figure 3. Microwave energy efficiency as a function of the
occupancy of the microwave resonator; Vsubstance ) the volume
of the substance loaded in the resonator; Vresonator ) the volume
of the resonator.
efficiency of a microwave is usually proportional to the degree
of charging the cavity which makes a direct comparison of the
heating techniques cumbersome.38
The larger the sample is with respect to the cavity, the higher
the microwave energy efficiency becomes; see Figure 3.
The results of preliminary experiments showed that, although
the energy efficiency was assumed to be 10%, a power input
of 420 W39 (corresponding to the energy input at t ) 0 of a
water bath with a temperature of 60 °C) strongly outperformed
the conventionally heated experiments. Therefore, the used
microwave power inputs were determined empirically. The
microwave equipment used in the experiments was inaccurate
with power settings below 40 W. The selected power inputs of
50, 100, and 150 W were adequate to get comparable drying
rates for microwave and conventional heating.
To study the effect of higher moisture contents, and in
particular the influence of microwave irradiation on the constant
drying interval, i.e. the removal of surface moisture, and on
the most crucial falling rate period usually determining the
overall drying time, i.e. the removal of moisture in the inner
pores, a sample of (S)-N-acetylindoline-2-carboxylic acid40 with
a moisture content of 5 wt % was subjected to a base/acid
precipitation procedure. This resulted in an initial moisture
content of 17 wt %.
The particle sizes of the precipitated (S)-N-acetylindoline-
2-carboxylic acid and N-acetyl-(S)-phenylalanine (Vide infra)
are approximately 3 and 5 times smaller than that of the original
grade of (S)-N-acetylindoline-2-carboxylic acid, respectively,
see Figure 4.
The results of the drying experiments using conventional
and microwave heating are shown in Figure 5.
The application of microwaves reduced the drying time
significantly. The experiments with conventional heating showed
a drying time of 35 min at 45 °C which was also obtained with
a microwave power of 50 W (corresponding to an actual power
The dielectric loss factor (ε′′) of a substance strongly depends
on the moisture content of the sample. The dielectric loss factor
only increases slightly with a moisture content below the critical
moisture content (mc).
For moisture contents above mc, ε′′ steeply increases with
increasing moisture content. As a consequence, a higher
moisture content results in a higher loss tangent. So, the higher
moisture content regions of the sample are heated more
efficiently causing higher evaporation rates than in the regions
where the moisture content is relatively low. Finally, as
compared to conventional heating the moisture distribution will
be more homogeneous when using microwave heating. Con-
sequently, the energy uptake will be relatively even.42 The even
temperature distribution (i.e., no hotspots) and the relatively high
drying rate make microwaves of great interest for drying
thermally unstable compounds.
(38) Hoogenboom, R.; Wilms, T. F. A.; Erdmenger, T.; Schubert, U. S.
Aust. J. Chem. 2009, 62, 236.
(39) The filling level of the microwave was in the range of 5%. Figure 3
shows an efficiency of 10% for this situation. The theoretical power
input with a water bath at 60 °C was calculated at 42 W. Therefore,
the preliminary experiments were performed with a microwave setting
of 420 W.
(41) Perkin, R. M. Int. J. Heat Mass Transfer 1980, 23, 687.
(42) Monzo´-Cabrera, J.; D´ıaz-Morcillo, A.; Catala´-Civera, J. M.; De los
Reyes, E. MicrowaVe Opt. Technol. Lett. 2001, 30, 166.
(40) Kindly provided by DSM, Venlo, The Netherlands.
1132
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development

Figure 4. Microscopic images. Left: original grade of (S)-N-acetylindoline-2-carboxylic acid. Middle: precipitated (S)-N-acetylindoline-
2-carboxylic acid. Right: N-Acetyl-(S)-phenylalanine.
Figure 5. Drying and temperature profiles for drying of the original grade of (S)-N-acetylindoline-2-carboxylic acid (5 wt % water)
at 5 kPa, at different temperatures and with different microwave powers. Left: conventional heating (CH). Right: microwave heating
(MW).
heat supplied by the microwave irradiation and the heat
consumed by evaporation of the moisture. The microwave
energy, converted into heat in the sample, at 150 W was too
high to be compensated by the heat consumption due to
evaporation, leading to higher temperatures. When the air flow
over the sample is insufficient to remove all of the evaporating
moisture, then the air is saturated leading to a higher equilibrium
temperature. The higher loss tangent is beneficial for the drying
rate of (S)-N-acetylindoline-2-carboxylic acid, which has rela-
tively good thermal stability. Although the higher energy
absorption initially caused faster heating and drying rates, as
compared to those of the original grade (S)-N-acetylindoline-
Figure 6. Dielectric loss factor (ε′′) as function of the moisture
2-carboxylic acid, the higher bulk temperature indicates that
content, with mc as the critical moisture content.
microwave irradiation should be applied with due caution for
thermally unstable materials.
The drying behaviors of the 5 wt % and the 17 wt % samples
have been studied for conventional heating (75 °C) and
microwave heating (150 W), and the results are collected in
Figure 7.
Drying Behavior of N-Acetyl-(S)-phenylalanine
To gain more insight into the drying behavior of other solid
organic substances using microwave irradiation, the drying
behavior of N-acetyl-(S)-phenylalanine was studied on small
scale, see the setup in Figure 2. The sodium salt of (S)-
phenylalanine was acetylated with acetic anhydride under
Schotten-Baumann conditions resulting in a product with a
moisture content of 22 wt %, which was immediately suitable
for the drying experiments. The particle size of N-acetyl-(S)-
phenylalanine was smaller than that of the original grade of
(S)-N-acetylindoline-2-carboxylic acid, see Figure 4. The drying
behavior of N-acetyl-(S)-phenylalanine with microwave and
conventional heating is depicted in Figure 8.
The precipitated (S)-N-acetylindoline-2-carboxylic acid with
17 wt % water displayed drying times of 70 and 35 min for
conventional (75 °C) and microwave heating (150 W), respec-
tively. The original (S)-N-acetylindoline-2-carboxylic acid grade
(5 wt % water) required a drying time of 20 min for
conventional heating (75 °C) and 10 min for microwave heating
(150 W). The measured bulk temperature during the drying
experiments with the precipitated (S)-N-acetylindoline-2-car-
boxylic acid was higher (45-60 °C) with microwave heating
than with conventional heating (25-50 °C). This contrasts with
the observations for the original grade of (S)-N-acetylindoline-
2-carboxylic acid containing 5 wt % water. The higher moisture
content of the precipitated grade caused a higher loss tangent
of the sample. Consequently, the microwave energy was more
readily converted into heat, resulting in an imbalance in the
The moisture content of the N-acetyl-(S)-phenylalanine was
even higher than that of the precipitated (S)-N-acetylindoline-
2-carboxylic acid. Such high water content could also lead to
superheating of the sample, as described above. This is indeed
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1133

Figure 7. Weight loss curves and temperature curves for conventional and microwave heating. Left: nonprecipitated (S)-N-
acetylindoline-2-carboxylic acid (5 wt % water). Right: precipitated (S)-N-acetylindoline-2-carboxylic acid (17 wt % water).
Figure 8. Drying- and temperature profiles of drying N-acetyl-(S)-phenylalanine at various temperatures and powers for the
conventional at 5 kPa. Left: conventional heating (CH). Right: microwave heating (MW).
observed for the microwave drying experiment at 150 W,
leading to a sample temperature of 60 °C. The drying time for
this high-energy input was 25 min, which is extremely short
compared to the drying times with conventional heating, for
which drying times of 130, 100, and 65 min at 45, 60, and 75
°C, respectively, were registered. At a high bulk temperature
of 63 °C N-acetyl-(S)-phenylalanine remained stable and hence
microwave irradiation is a very suitable heating technique for
drying this substance. The advantage of lower bulk temperatures
combined with a high drying rate under mild microwave
irradiation was demonstrated by the drying experiments with a
lower microwave power input. The drying times were 40 and
50 min for 100 and 50 W microwave power, respectively. These
drying times were shorter than observed for the conventionally
heated samples using a water bath of 75 °C, where the sample
temperature increased to 55 °C, while the sample temperature
under these low microwave power inputs remained below 45
°C. The latter observation indicates that even at high moisture
contents microwave irradiation can lead to higher drying rates
with lower bulk temperatures, as long as the microwave power
is mildly applied.
question was raised how practical the drying efforts could
become on larger scales, in particular when more material of
thermally unstable substances should be handled. For that
purpose, the drying behaviour of a pharmaceutical grade of
cocarboxylase hydrochloride was investigated. The thermosen-
sitive enzyme was dried according to different sources from
ambient to moderately high temperatures but not higher than
50 -55 °C to avoid degradation.43-45 The drying times for the
enzyme are long enough and can reach even 80 h46 with
conventional heating. Since microwave drying can be most
efficient in the case of thermally unstable substances, it was a
matter of particular interest to investigate the drying behavior
of cocarboxylase hydrochloride with microwave heating.
The drying experiments for this substance were carried out
on two scales: in a laboratory environment and on pilot-plant
scale.
Laboratory-Scale Experimentation. On laboratory scale
a standard 1.5 kW electric vacuum oven for thermal drying of
150-200 g amounts of cocarboxylase hydrochloride was
applied. This vacuum convective drying procedure was per-
(43) Berezovsky, V. M.; Koltunova, V. I.; Pekel, N. D.; Shlimovich, Y. A.
Russ. J. Gen. Chem. 1963, 33, 49.
(44) Budz, J.; Karpinski, P. H.; Mydlarz, J.; Nyvit, J. Ind. Eng. Chem.
Prod. Res. DeV. 1986, 25, 657.
Drying Behavior of Cocarboxylase Hydrochloride
Having gained a better understanding of the benefits of
microwave heating compared to oil bath heating during
exploratory drying of some pharmaceutical intermediates in a
specially designed small-scale setup in the laboratory, the
(45) Weijlard, J.; Tauber, H. J. Am. Chem. Soc. 1938, 60, 2263.
(46) Technological requirements for production of cocarboxylase hydro-
chloride TR 64-00205096-031-2002; Kharkov Plant for Chemical
Reagents, National Academy of Sciences of Ukraine: Kharkiv,
Ukraine, 2002.
1134
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development

of the adsorbed moisture, approximately 30-40 min for 3 kg
of material, the temperature rose due to absorption by the
product and was determined by the microwave absorbance
capacity of the constituents. At this stage much less energy was
required. Therefore, after reaching a predetermined range of
50-55 °C for the enzyme, the temperature was maintained by
an on/off working regime of the microwave oscillator through-
out the drying process. The average value of the microwave
power input was about 70 W at this stage and was 5-fold less
than at the constant rate period. A rotation of 6-8 rpm was
maintained during the whole drying process to prevent any
product caking.
To assess the energy efficiency of the microwave drying
process with respect to the conventionally heated procedure the
actual energy consumption figures were calculated (see Ex-
perimental Section for the calculation method). For this purpose
the dielectric properties of the enzyme, dielectric constant and
dielectric loss, were measured using special equipment,47 and
amounted to 1.95 and 0.03, respectively. The results are
presented in Table 2.
The energy efficiencies for both microwave and conventional
batch methods were calculated from eq 14 in the Experimental
Section. The energy consumption values in Table 2 reveal a
tremendous difference between microwave-heated and conven-
tionally heated drying, the efficiency of microwave energy
uptake (43%) being in a good correlation with Figure 3
considering that the filling of drying chamber in our runs was
only at the level of 15%. Based on this result it may be firmly
stated that the level of 70% efficiency may be reached at
appropriate filling of drying chamber. Evaluation of the energy
uptake Qabsorbed during the heating stage was done only for
laboratory and batch microwave drying. The closely related
thermodynamic and electro-physical Qheating and Qabsorbed values
(Table 2), the latter being dependent on both the dielectric
constants of the substances and electro-technical parameters of
the microwave equipment, show that the utilization of micro-
wave energy may be very high under proper conditions. The
most reasonable explanation for such an immense difference
between microwave and conventional heating is the limited
efficiency of conventional heating due to the heat capacity of
the material. Therefore, considerably more energy should be
supplied for heating than theoretically required. In contrast,
microwave heating is not limited by heat resistance due to direct
coupling of microwaves with the material and is predominantly
determined by the microwave-absorption capacity of the
substance, thus minimizing losses to the surroundings. It should
be noted that efficient microwave heating was possible in this
case despite the relatively low loss tangent of the enzyme.48
Figure 9. Schematic image of the experimental laboratory setup
used for drying of cocarboxylase hydrochloride with 1: multi-
mode microwave cavity, 2: microwave generator (oscillation
frequency 2.45 GHz), 3: temperature sensor (IR - pyrometer),
4: pressure sensor, 5: cold trap, 6: vacuum pump.
formed with the oven connected to a vacuum system using a
water pump, an exhausting vacuum pump and a cold trap. For
the microwave-heating experiments an experimental laboratory
microwave setup was used, as shown in Figure 9.
Microwave-assisted vacuum drying in both laboratory and
larger-scale experiments was conducted under the same condi-
tions including material amounts, initial humidity, temperature,
and pressure as in the case of conventional heating. The
operating microwave power was within 30-300 W, and the
heating rate was approximately 1 °C/min. The moisture content
was determined by loss on drying measurements. The drying
curves were generated for both conventional and microwave
heating, see Figure 10.
The curvature in the range of 1.5-2% critical humidity is
characteristic for both curves, showing that the drying rate
drastically decreases during the falling rate period. It is evident
that the overall drying time with microwave heating is shorter
by a factor 10 compared to the conventional process. Appar-
ently, removal of the bound moisture (1.5-2%) is a limiting
phase during the whole drying process. Such a dramatic
shortening of the drying process can be rationalized by the
influence of microwaves on the diffusion times of diffused
moisture.
Larger-Scale Runs. The satisfying results in the laboratory
encouraged us to move to pilot-plant scale aimed at establishing
the scalability and energy consumption efficiency of microwave
heating in drying. The larger-scale drying experiments of 3 kg
amounts of the enzyme were performed with a rotary vacuum
microwave dryer of 35 L chamber capacity (with an optimal
filling of about 50% by volume), see Figure 11. The operating
microwave power was within 70 and 350 W during the runs,
and the rotation rate varied from 6 to 8 rpm, being satisfactory
for a thorough agitation of the product. The product temperature
was measured with a thermo sensor fixed at the drying chamber
wall (bottom part) where the electromagnetic field intensity is
close to zero level.
Discussion
During the constant rate period when the adsorbed moisture
at the surface of the particles is removed, see Figure 12, the
energy primarily is consumed for evaporation, causing a
considerable energy input. For 3 kg of wet product (15.5 wt %
moisture) microwave power input ranging from 300 to 350 W
was required. The temperature of the product at this stage was
determined by the vapour temperature, corresponding to the
boiling temperatures of ethanol and water, which did not exceed
30 °C (boiling temperature of water at 30 Torr). After removal
The similarity in chemical structure between (S)-N-acetyl-
indoline-2-carboxylic acid and N-acetyl-(S)-phenylalanine can-
(47) Voloshko, A. Yu.; Kramskoy, G. D.; Kramskoy, Ye. D.; Samoilov,
V. L.; Seminozhenko, V. P; Chepkiy, A. A.; Shishkin, O. V. Method
for measurement of dielectric constant and dielectric loss, Ukrainian
patent number 34146, 2008.
(48) It is generally accepted that substances with tg δ < 0.1 are considered
low microwave absorbing, e.g. the value for water which is considered
to have a moderate microwave absorbance capacity is 0.123.
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1135

Figure 10. Kinetics of conventional drying (left) and microwave drying (right) of cocarboxylase hydrochloride (the longer falling
rate period is shown with higher resolution in the inset).
Figure 11. Industrial rotary vacuum microwave dryer “Pharma-Micro”. Right: photograph. Left: principal scheme. (1) Rotary
drying chamber with microwave resonator. (2) Microwave oscillator. (3) Control unit.
make it hard to predict the effect of a smaller particle size on
the drying behavior.
The results of the microwave drying behavior of all three
substances show a clear similarity with respect to a faster
decrease in moisture contents compared to conventional heating
during the constant rate period where free moisture is removed.
This observation can easily be explained by the volumetric and
inertia-less nature of microwave heating. However, for the
enzyme this difference between both heating methods is more
expressive than for the pharmaceutical intermediates. Moreover,
the effect of instant heating allowed sufficiently precise tem-
perature control, thus avoiding the risk of overheating, the latter
being a valid argument in favor of microwave drying of
thermolabile substances.
Figure 12. Microwave drying and temperature curve of
cocarboxylase hydrochloride including removal of both free
(first stage) and bound (occluded water during second stage)
moisture.
Conclusions
not be directly translated into a similar drying behavior. Also
the particle size may influence the drying behavior. A smaller
particle size corresponds to a higher specific surface and a
smaller channel or pore size. The higher surface area could
increase the drying rate, particularly for the unbound water. A
smaller pore size may decrease the amount of pore moisture
but could inhibit the removal of the moisture from these pores.
The packing of the particles of various sizes is also different,
leading to a different pore size and a difference in diffusion
rate of the moisture through the sample. All of these factors
A comprehensive study of drying processes for substances
of various natures and properties has been carried out, compar-
ing microwave with conventional heating techniques. From the
results described for the drying of (S)-N-acetylindoline-2-
carboxylic acid (both original and precipitated grades), of
N-acetyl-(S)-phenylalanine, and of the enzyme cocarboxylase
hydrochloride under conventional and microwave heating, it
can be concluded that microwave irradiation leads to consider-
ably higher drying rates. The power input when using micro-
wave irradiation should be chosen carefully. A high power input
1136
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development

Table 2. Drying characteristics of cocarboxylase hydrochloride during conventional and microwave heating
heating method
microwave lab
microwave batch
conventional batch
heating time to predetermined temp. (h)
heater power (W) at first drying step (removal of adsorbed
water)
0.25
300
0.5
5
350
150a
overall drying time (h) (average)
heater power (W) at second drying step (removal of bounded
water)
8
30
8
70
70
150a
Q inputb (kJ) (heat source and vacuum pump)
Q inputb (kJ) (heat source only)
Q heatingc (kJ)
15.5 × 103
45 × 103
2.5 × 103
380
420 × 103
40 × 103
380
103
18
48
21
10
Q vaporizationd (kJ)
700
700
Q absorbede (kJ)
400
-
∼3
energy efficiency (%)
43
a Average value for the whole process. b Experimental value comprises heater/vacuum pump power input, calculated from eq 8 in the Experimental Section. c Calculated
from eq 10 in the Experimental Section. d Calculated from eq 11 in the Experimental Section. e Calculated from eqs 12 and 13 in the Experimental Section for
microwave-assisted drying only.
enables a dramatic decrease of the drying time of the sample
but can also lead to overheating and thermal product decom-
position. A balance between both overheating and fast drying
should be aimed at. However, even for mild conditions the
drying rate is much higher than for similar heat-flow rates with
conventional heating. It has even been shown that substances
with low microwave absorbance may be successfully dried.
To obtain similar drying rates, conventional heating demands
much higher heat-flow rates than microwave heating. Higher
heat-flow rates lead to rather large temperature differences
between the heating element and the sample and to very high
temperatures inside the sample, which may be detrimental for
thermally unstable materials. Lower temperatures during drying
by microwave irradiation make this heating technique extremely
appropriate for drying in short periods of time.
Estimations of the energy consumption figures for cocar-
boxylase hydrochloride revealed a high efficiency for micro-
wave heating which can be up to 70% of energy utilization at
50% filling of the drying chamber. This value can hardly be
imagined for conventional technologies.
In brief, the results of this study revealed a number of
benefits of microwave heating such as scalability, high drying
rates, energy savings (by a factor of 10), accurate temperature
control, and better product quality due to a minimized risk of
product degradation regardless of the chemical structure and
physicochemical properties of the substances.
flask was charged with 15 g of either (S)-N-acetylindoline-2-
carboxylic acid or N-acetyl-(S)-phenylalanine. A temperature
probe was inserted in the central neck and placed at a constant
height in the exact center of the sample. An air-inlet tube was
connected to another neck of the flask. The air-inlet tube was
connected to a capillary tube. The capillary tube ensured a
constant air flow. The flask was held at a constant pressure of
50 mbar by an oil pump connected to the third neck. The flask
was heated by either a magnetically stirred (750 rpm) water
bath of the indicated temperature for the conventionally heated
experiments or the power-controlled microwave with the
indicated power settings. The weight was measured at constant
time intervals. Water losses (in weight) were quantified by:
(mwet - mdry) · 100
m% ) 100 +
mwet
where m% ) decrease in mass (%), mwet ) mass of wet
substance (g), mdry ) mass of dry substance (g). The temperature
of the sample was recorded at constant time intervals by either
a thermocouple for the experiments with conventional heating
or by a fiber-optic sensor in the microwave oven. The dry
substances showed no traces of degradation after the drying
procedure.
Again, no stirring and a sufficient air purge guaranteed that
the evaporation rate and the diffusion rate of the surface and
inner-pore bound water governed the rate of the drying process.
This time during the runs, not the temperature but the power
was chosen as the variable. A commercially available, auto-
mated multimode microwave oven MicroSynth from Milestone
srl (Italy) was used. This oven operates at 2.45 GHz and is
temperature controlled by a fiber-optic sensor. The maximum
power input could be adjusted between 0 and 1000 W.
Drying Procedures of Cocarboxylase Hydrochloride. For
the laboratory experiments 150-200 g amounts of wet enzyme
material (humidity 10-15%) were taken for both conventional
and microwave experiments. Drying was performed at 25-30
Torr at temperatures below 55 °C. At the constant drying rate
interval when the removal of free moisture took place, the
product temperature was e30 °C at the given pressure, which
was determined by the vapour temperature (ethanol and water).
Experimental Section
All starting materials including wet cocarboxylase hydro-
chloride (moisture content of ∼15 wt %) were obtained from
commercial suppliers and used as received unless indicated
otherwise. In case of need the wet cocarboxylase hydrochloride
was kept in a refrigerator at T e 10 °C for no longer than 24 h
before drying.
All moisture-sensitive reactions were performed under an
atmosphere of dry argon. 1H NMR spectra were recorded on a
Varian Gemini (400 MHz). Proton chemical shifts were reported
in ppm downfield from tetramethylsilane (TMS).
Drying Procedures of (S)-N-Acetylindoline-2-carboxylic
Acid and N-Acetyl-(S)-phenylalanine. During the convention-
ally and microwave-heated experiments, a three-neck 250-mL
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1137

Then the temperature was increased gradually, to avoid
overheating, to 50-55 °C by regulating the power input of an
electric heating unit in the case of conventional drying and of
a microwave oscillator in the case of microwave-assisted drying.
A constant air flow was provided to remove water vapours and
to maintain a constant humidity in the drying chamber.
During the microwave pilot-plant scale runs, 3 kg of wet
enzyme material after centrifuging was placed in the drying
chamber (which was then evacuated to 25-30 Torr) and was
dried under the same conditions as those for the laboratory
experiments. For a better mass transfer of the moisture, rotation
of the drying chamber was provided in the range of 6-8 rpm.
The drying curves were generated by measuring the loss on
drying at 100 °C in the samples taken during drying at 1.5-2
h intervals (except for the first sampling which was after 0.5 h
of drying). The moisture content was determined by measuring
loss on drying for each sample in duplicate. After reaching the
critical moisture content (1.5-2%) the samples in both con-
ventional and microwave experiments were subjected to me-
chanical milling for product homogenization and breakage of
the agglomerates that might be formed during drying due to
particle aggregation.
Base-Acid Precipitation Procedure of (S)-N-Acetylin-
doline-2-carboxylic Acid. A 2500-mL three-neck, round-
bottomed flask was charged with (S)-N-acetylindoline-2-
carboxylic acid40 (300 g, 1.47 mol) and was dissolved in NaOH
solution (770 mL, 2.0 M). Using a dropping-funnel this mixture
was brought to pH 3 with 4 M HCl (approximately 385 mL)
under stirring. The precipitated product was then filtered under
vacuum and consecutively washed with saturated brine (200
mL) and demineralised water (100 mL). The filtrate was dried
in air for two days, yielding 95% (S)-N-acetylindoline-2-
carboxylic acid (99% ee). The moisture content was 17 wt %
upon drying in air for two days.
Q ) U · A · ∆T
A ) 2πr∂h
(1)
(2)
Q
) power supplied to the flask (W)
) overall heat transfer coefficient (W/(K.m2)
) heat transfer area (m2)
U
A
∆T
dh
) temperature difference (K)
) submerged depth of the flask in the water bath (m)
Nu · kwater
1
U
1
l
k
)
+
hwater
)
(3)
(4)
hwater
D
Nu
k
) Nusselt number
) thermal conductivity coefficient (W/m K)
) diameter (m) (0.085 m)
D
Nu ) 2.0 + 0.66Re1/2 · Pr1/3
F · υ · D
998 · 3.34 · 0.085
Re )
⇒
≈ 3.2 × 105
(
)
µ
H2O
0.000894
(5)
cp · µ
4200 · 0.000894
Pr )
⇒
≈ 6.47 (6)
(
)
k
H2O
0.58
n
750
60
V ) 2πr ·
⇒ 2π · 0.0425 ·
≈ 3.34 m/s
(
)
(
)
60
(7)
Re
Pr
F
) Reynolds number
) Prandtl number
) density (kg/m3) (998 kg/m3
)
ν
) velocity (m/s)
Synthesis of N-Acetyl-(S)-phenylalanine. A 2500-mL
three-neck, round-bottomed flask was charged with (S)-phenyl-
alanine (198 g, 1.2 mol). (S)-Phenylalanine was dissolved in
NaOH solution (300 mL, 4.0 M). Under vigorous stirring and
ice bath cooling, acetic anhydride (122.4 g; 1.2 mol) was added
dropwise with simultaneous addition of NaOH solution (300
mL, 4.0 M). The pH was maintained at 10. After 2 h, additional
acetic anhydride (122.4 g; 1.2 mol) was added dropwise with
simultaneous addition of NaOH solution (300 mL, 4 M). Using
a dropping-funnel this mixture was brought to pH 3 with 6 M
HCl (approximately 25 mL). The precipitated product was then
filtered under vacuum and washed with brine (200 mL) and
demineralised water (100 mL). The solid was dried in air for
two days, yielding 295 g of crude product with a moisture
content of 22 wt % and finally pure product (230 g dry product,
93%). 1H NMR (400 MHz, CD3OD) δ: 7.51 (m, 5H) 4.92 (m,
1H), 3.45 (dd, 2H), 3.20 (dd, 2H), 2.15 (s, 3H) confirmed that
all (S)-phenylalanine was acetylated. The specific rotation of
the dry product was [R]23D ) +28 (c ) 1.0, MeOH), literature:
r
) radius flask (m) (0.0425 m)
n
) rotational speed stirrer (750 rpm)
) diameter (m) (0.085 m)
) viscosity (Pa/s) (8.94 × 10-4 Pa/s)
) heat capacity (J/kg K) (4200 J/kg K)
D
µ
cp
Calculation of Energy Consumption Values for Cocar-
boxylase Hydrochloride.
Qinput ) P × t
(8)
Qinput
) energy input during drying (J)
) heat source/vacuum pump power (W)
) drying time (s)
P
t
In the case of microwave heating:
Qinput ) PMW × t
(9)
[R]23 ) +33 (c ) 1.0, MeOH).49
D
Qinput
) energy input during drying (J)
Calculation of Power Input for (S)-N-Acetylindoline-2-
carboxylic Acid and N-Acetyl-(S)-phenylalanine.
PMW
t
) microwave power input (W)
) action (on) period of microwave generator (s)
(49) Smith, H. E.; Neergaard, J. R. J. Am. Chem. Soc. 1997, 119, 116.
1138
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development

ε′′
) tan δ × ε/ (dielectric loss factor)
) dielectric loss tangent
Qheating ) m × c × ∆T
(10)
tan δ
ε/
Qheating ) energy required for heating (J)
) dielectric permittivity
m
) weight of the material (g)
) heat capacity (J/g ·K)
c
The efficiency of the energy utilization was estimated by:
∆T
) difference between the initial and final temperatures (K)
Qheating + Qvaporization
× 100%
(14)
Qinput
Qvaporization ) m × Qboil
(11)
Q heating ) energy for heating (J)
Qvaporization ) energy required for vaporization of volatiles (J)
Qvaporization ) energy for vaporization (J)
Qboil
m
) specific boiling heat of volatiles (J/g)
) weight of the material (g)
Q input
) energy consumed by heat source only (J)
Qabsorbed ) P × V × t
(12)
Acknowledgment
The authors of Eindhoven University of Technology are
grateful to SenterNovem for funding this research and to Dr.
E. Esveld of the Food Technology Center, Wageningen, The
Netherlands for his critical comments. We thank DSM Pharma
Chemicals, Geleen, The Netherlands, and Syncom B.V.,
Groningen, The Netherlands, for their support. The authors of
the National Academy of Sciences of Ukraine are grateful to
Kharkov State Plant for Chemical Reagents, Ukraine, for
providing them with wet cocarboxylase hydrochloride along
with the data regarding conventional batch drying.
Qabsorbed ) energy absorbed by dielectric irradiated by microwaves (J)
P
) microwave power absorbed by dielectric in a unit of volume
(W/cm3
)
50
V
t
) bulk volume (cm3
)
) time (s)
P ) 0.3 × E2 × f × ε′′ × 10-6
(13)
E
f
) microwave field strength (V/cm)
) microwave generator frequency (MHz)
Received for review April 15, 2010.
OP100102M
(50) Okress, E. C. MicrowaVe Power Engineering; Academic Press: New
York and London, 1968.
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1139