| Rantanen J, Jørgensen A, Räsänen E, Luukkonen P, Airaksinen S, Raiman J, Hänninen K, Antikainen O, Yliruusi J.
Process Analysis of Fluidized Bed Granulation.
AAPS PharmSciTech. 2001; 2(4): article 21.
|
Jukka Rantanen,1,2
Anna Jørgensen,2
Eetu Räsänen,2
Pirjo Luukkonen,2
Sari Airaksinen,2
Johanna Raiman,2
Kaisa Hänninen,2
Osmo Antikainen,2
and Jouko Yliruusi1,2
1Viikki Drug Discovery Technology Center, Pharmaceutical Technology Division, University of Helsinki, Finland
2Pharmaceutical Technology Division, University of Helsinki, Finland
Correspondence to:
Jukka Rantanen
Tel: +358-9-191 59141;
Fax: +358-9-191 59144;
Email: jukka.rantanen@helsinki.fi | Submitted: August 7, 2001; Accepted: October 2, 2001; Published: October 17, 2001 | Keywords:
formulation, fluidized bed granulation, near-infrared (NIR) spectroscopy, moisture measurement, process control | This study assesses the fluidized bed granulation process for the optimization of a
model formulation using in-line near-infrared (NIR) spectroscopy for moisture determination.
The granulation process was analyzed using an automated granulator and optimization of the
verapamil hydrochloride formulation was performed using a mixture design. The NIR setup with
a fixed wavelength detector was applied for moisture measurement. Information from other
process measurements, temperature difference between process inlet air and granules
(Tdiff), and water content of process air (AH), was also analyzed. The
application of in-line NIR provided information related to the amount of water throughout
the whole granulation process. This information combined with trend charts of Tdiff and AH enabled the analysis of the different process phases. By this means, we can obtain
in-line documentation from all the steps of the processing. The choice of the excipient
affected the nature of the solid-water interactions; this resulted in varying process times.
NIR moisture measurement combined with temperature and humidity measurements provides a tool
for the control of water during fluid bed granulation.  |
The development of process analytical chemistry (PAC) will provide a window to the
physicochemical phenomena occurring during pharmaceutical manufacture. The border between
PAC and more traditional laboratory analysis is quite ambiguous. The terms in-line, on-line, at-line, off-line, noninvasive are often referred to in the literature. Definitions of these classes are as follows: in-line,
the sample interface is located in the process stream; on-line, automated sampling and sample transfer to an automated analyzer; at-line, manual sampling with
local transport to analyzer located in the manufacturing area; off-line, manual sampling with transport to a remote or centralized laboratory.1,2 Solid-water interactions are one of the fundamental issues in the pharmaceutical
technology.3 The state of water in a solid material may be characterized using x-ray diffraction,
microscopic methods, thermal analysis, vibrational spectroscopy, and nuclear magnetic
resonance spectroscopy.4 Traditionally, the control of fluidized bed granulation is based on indirect
measurements. These control methods use the properties of process air.5-7 The nondestructive character of vibrational spectroscopy techniques, such as
near-infrared (NIR), makes it a novel tool for in-line quality assurance.8,9 NIR has been widely applied for the measurement of water in various applications.10,11 In the case of wet granulation of pharmaceutics, NIR can be applied for both
quantitative analysis of water12-16 and for determining the state of water in solid materials.17,18 This enables us to understand the molecular level phenomena during manufacture of
pharmaceutics. Further, NIR has been applied for studying the nature of water-solid
interactions within various materials.19,20 In this study, the NIR moisture measurement during fluidized bed granulation was applied for the
optimization of a model formulation. The in-line moisture measurement was performed using a fixed wavelength (4-wavelength) NIR setup. Other
process information (granule temperature and process air properties) was also analyzed. A model
formulation of verapamil hydrochloride was optimized using a mixture design.
| | Granulations All granulations were performed in an automated bench-scale fluid bed granulator
(Glatt WSG 5, Glatt GmbH, Binzen, Germany). Details of instrumentation and in-line NIR setup are given elsewhere.21 The in-line NIR setup is a prototype developed in cooperation with the
Technical Research Centre of Finland (VTT Electronics, Oulu, Finland, http://www.vtt.fi/indexe.htm). The integrated
detector technique was used instead of the traditional filter-wheel solution. A model
formulation (batch size 3500 g) consisting of 48% wt/wt of verapamil hydrochloride
(Orion Pharma, Espoo, Finland) and 48% wt/wt of filler was applied. The fillers studied
were microcrystalline cellulose (MCC) (Emcocel 50M, Penwest Pharmaceuticals, Nastola,
Finland), lactose monohydrate (Pharmatose 200M, DMV Pharma, Veghel, the Netherlands),
and pregelatinized starch (Starch 1500, Colorcon, Indianapolis, IN). The composition
of the filler in a formulation was optimized using mixture design (Figure 1). Seventeen granulations were performed in randomized order (coding A-Q in Figure 1). Polyvinylpyrrolidone (PVP) (Kollidon K25, BASF, Ludwigshafen, Germany) was used
as a binder in the formulation (4% wt/wt). Solutions in water were prepared using 10%
wt/wt of PVP (1400 g of granulation liquid used in 1 granulation). Granulations were
performed with identical process set values (inlet air temperature 50°C, granulation
liquid flow rate 0.1 kg/min, spraying pressure 0.15 MPa). The flow rate of inlet air
was varied between 0.040 and 0.055 m3/s for optimum bed performance.
Material characterizationOff-line NIR spectra of the materials were measured with a Fourier
Transform (FT)-NIR spectrometer (Bomem MD-160 DX, Hartmann & Braun, Quebec,
Canada) using Bomem-GRAMS software (v. 4.04, Galactic Industries Inc, Salem, NH)
and Teflon as a reference (99% reflective Spectralon, Labsphere Inc, North Sutton,
NH). The spectra were recorded over a range of 4000 to 10 000 cm-1 with a resolution of 8 cm-1, and an average of 32 scans was used. The moisture contents of the fillers were determined by Karl Fisher (KF)
titration on a Mettler KF titrator (model DL35, Mettler Toledo AG, Greifensee,
Switzerland) using Hydranal Solvent and Hydranal Titrant 5 (Riedel-deHaën
Laborchemikalien GmbH, Seelze, Germany). The titration vial was warmed to 50°C and
samples of 500-1000 mg were used. The determinations were performed in triplicate.
Granule particle size was determined by sieve analysis. The sample match was
confirmed using a rotary sample divider (Fritsch laborette 27, Idar-Oberstein,
Germany). Granules were first sieved through a 3.15-mm sieve; thereafter, samples
(20 g) of the undersize fraction were vibrated with an automatic sieve shaker
(Fritsch analysette, Idar-Oberstein, Germany) for 5 minutes. The sieve analysis
(range 45-2000 µm with v2 increment) was performed in triplicate for each batch
of granulation and average values for mass mean particle size were determined. Design of experimentsThe experiments were performed according to the mixture design (Figure 1). This centroid cubic design included 13 experiments. In addition, 3 corner
point and 1 center point experiments were replicated. Thus, the number of
experiments was 17 (Figure 1, batches A-Q). The effect of varying composition on the mean particle size
was modeled using Modde software (Modde for Windows, v. 3.0, Umetri AB, Sweden).
Mass fractions (0-1) of fillers (MCC, lactose monohydrate, starch) in a formulation
were used as variables. Regression modeling was performed using the backward
selection technique. The original mixture model with quadratic; interaction and
linear terms was simplified by removing the least significant terms from the model
as long as the predictive power of model was increasing. The predictive powers of
the models were evaluated using cross-validation. The cross-validation was basing
on the leave-one-out principle with Q2 as a measure of predictive power (Modde for Windows). Q2 can be described as a fraction of the variation of the response that can be predicted by
a model. | Properties of starting materials The off-line FT-NIR spectra of the materials applied were measured to
understand the varying levels of in-line NIR signals (materials in dry powder form, Figure 2). The excipients used in formulations A-Q were all carbohydrates; therefore,
they showed to some extent similar spectral features in the region applied for in-line moisture measurement (1800-2200 nm). The band at around 1940 nm
(a combination of water OH stretching and bending vibrations) was applied for
the calculation of in-line spectral response, apparent water absorbance (AWA).21 This combination band is widely used for PAC applications as a result of its
relatively strong absorption and high selectivity. The application of the water band
at around 1450 nm (the first overtone of OH stretching vibrations) is difficult,
because this region has spectral features resulting from other OH groups available
in formulations. The reflectance at 1813 nm was used for the baseline correction and
the reflectance at 2214 nm for the normalization of water signal. The higher initial
moisture content of starch (9.28%) in comparison with lactose monohydrate and MCC
(4.74% and 4.43%, respectively) was seen as a higher water absorbance maximum (Figure 2B). The state of water affects the NIR spectra.17-20 The water of crystallization with lactose monohydrate resulted in a relatively
narrow water band at around 1940 nm in comparison with the water band of MCC and starch
(Figure 2). The broad band for adsorbed water of MCC and starch indicated a spread of
energies of interaction, whereas the monohydrate water band is typical of a more
uniform interaction. The binder effects have been previously evaluated.22
Evaluation of different phases of granulation The corner points of the mixture design were used for the evaluation of the
granulation behavior of each filler. The temperature difference (Tdiff)
between inlet air and granules is a widely used indicator of the state of the
granulation (Figure 3A).5 AWA is a useful tool for monitoring of water through all phases of
granulation (Figure 3B). The AWA value can be calibrated against a reference technique (eg, loss on
drying methods, KF titration), and in-line moisture content of granules may
be plotted during all 3 process phases. Further, the absolute water content (AH)
of process air (g H2O/m3 of dry air) was applied for the
process monitoring (Figure 4. The measured values of Tdiff and AH are not directly obtained
from the granules, as is the case with NIR. The evaporation of water requires energy,
which is seen as a decrease in granule temperature (increase in Tdiff).
The water evaporated is seen as an increase in outlet air AH (the spraying phase in Figure 4). The typical behavior of each formulation can be identified by combining
information of Tdiff, AWA, and AH. An experienced process operator combines
this information from various trend charts (eg, Tdiff, AWA, AH), and easily
defines the state of the granulation process. Multivariate process monitoring methods
can also be used.21,23
During the first phase of granulation (mixing), the moisture content of
formulations B and C decreased (300-second period plotted after settling of
air flow, Figures 3 and 4). The Tdiff of these formulations increased slightly in comparison
with the formulation containing lactose monohydrate (Figure 3A). The AWA of formulations B and C decreased during the mixing phase; that of
formulation A remained constant (Figure 3B). This indicated loss of water with formulations B and C. Further, the AH of
outlet air during the mixing phase was higher with the formulations containing MCC
and starch (Figure 4). With increasing inlet air temperature, the relative humidity of inlet air
decreased. The water of MCC and starch was loosened because of the decreased
equilibrium-moisture content of materials. The water of crystallization in
formulation A (lactose monohydrate) was not loosened during drying at 50°C. This
fact may be used for materials containing water of crystallization. The presence
of indicators described previously can be used to detect pseudopolymorphic changes.
Morris et al24 introduced the use of accelerated fluid bed drying basing on NIR and
temperature probes. They showed that no physical changes occurred in compounds
during the fluid bed drying with process inlet air temperature above the melting
point of a low-melting compound. The spraying phase was seen as a decrease in granule temperature (increase in
Tdiff) and an increase in the AH of outlet air (Figures 3A and 4). The moisture profile (AWA) with formulation C (starch) was displaced in
parallel (Figure 3B). This was due to the higher initial water content of starch (Figure 2B). A fixed wavelength NIR sensor (WET-EYE, Fuji Paudal Co., Ltd., Osaka, Japan)
has been previously used by Watano et al25,26 and Miwa et al.27 Watano et al26 evaluated the formulation effects on the NIR sensor output, and they found that
the mixing ratio of the model formulation (lactose monohydrate: corn starch) affects
the calibration. This was explained by the different states of water in the materials
used. The water-absorbing potential of corn starch was higher than that of lactose
monohydrate. Miwa et al27 evaluated the wet massing behavior of different excipients. They classified the
excipients they studied into 5 groups according to the output of the NIR sensor. The NIR
sensor output was related to the inside/surface distribution of water in at-line
samples. They also used this result to estimate the amount of granulation liquid needed
for wet granulation. In this study, no differences could be observed by visual perusal of AWA profiles
(rate of increase in Figure 3B) during the spraying phase. The result is in contrast to the earlier results
obtained by Watano et al,26 Miwa et al,27 and Luukkonen et al.17 In these studies the spectral changes were greatest with materials that had a
poor liquid retention capacity. In this case, however, the method of granulation was a
low shear system (fluid bed granulation). It has been shown previously that MCC is able
to immobilize more water during granulation if high shear forces are used compared to
low shear forces.28 More water is able to get into the internal structure of microcrystalline
cellulose, and less water is present on the surface of the particles. During fluid
bed granulation the shear forces are much lower compared to high shear granulation.
Because of this, the influence of the different water absorbing potentials of the
materials on the AWA profiles is minimized. Consequently, when evaluating water-solid
interactions with NIR, the effect of shear forces on the state of water should be
considered. During the drying phase, differences in the drying rates were observed (Figures 3 and 4). The drying time needed for formulation A (lactose monohydrate) was 800 and
1000 seconds less than those for formulations B (MCC) and C (starch), respectively.
The adsorbed water in the case of formulation A resulted in a faster drying rate in
comparison with formulations B and C containing water-absorbing excipient. In the
batch drying unit operations, a constant-rate drying period could be identified
from high Tdiff value (Figure 3A). The falling-rate drying period occurs during the decreasing Tdiff
value. In the case of formulation A, both these periods were shorter than for
formulations B and C. The information gained with traditional methods from the state of the granules
during the spraying phase and at the beginning of the drying phase is quite limited.
The levels of variables Tdiff and AH stayed at a constant level throughout
the spraying and some of the drying phases (Figures 3A and 4). The in-line information was measured with the NIR setup through all
stages of the granulation process (Figure 3B). Using this information, we can obtain documentation for all the steps in
the processing. Granule particle size The model for mass mean granule particle size is presented in Figure 5. With the increasing amount of MCC, the mean particle size decreased. The
largest granules were obtained with the formulations containing lactose monohydrate
as a filler (batches A and O). The mean granule size with formulations containing
starch was between these 2 cases. The water soluble filler (lactose monohydrate)
resulted in an enhanced wetting of the powder mass, and by that means, a faster
rate of agglomeration.
The shorter drying time with formulation A was one reason the particle size
was affected. Both attrition and granule breakage occurred during the longer drying
period, and the mean granule size of formulations B and C was smaller. The NIR signal
was not affected by the particle size differences. The simultaneous detection of all
wavelengths facilitates the application of NIR to process analysis.
| NIR moisture measurement combined with temperature and humidity measurements provides
a novel tool for the monitoring of water during fluid bed granulation, and subsequently, to
control the granulation process. The choice of excipient affected the solid-water interactions
resulting in different processing times. The varying behavior of formulations during processing
can be identified in a real-time mode. NIR spectroscopy offers unique information of granule
moisture content during all phases of granulation.
| This study was financially supported by the Graduate School in Pharmaceutical
Research (Ministry of Education, Finland), the National Technology Agency, TEKES
(Finland), and Orion Pharma (Finland).
|
1.
Callis J, Illman D, Kowalski B. Process analytical chemistry. Anal Chem. 1987;59:624A-637A.
2.
Beebe K, Blaser W, Bredeweg R, et al. Process analytical chemistry. Anal Chem. 1993;65:199R-216R.
3.
Zografi G. States of water associated with solids. Drug Dev Ind Pharm. 1988;14:1905-1926.
4.
Brittain H. Methods for the characterization of polymorphs and solvates. In: Brittain H, ed. Polymorphism in pharmaceutical solids. 1st ed. New York: Marcel Dekker Inc;1999:227-278.
5.
Schaefer T, Wørts O. Control of fluidized bed granulation III. Effect of inlet air temperature and liquid flow rate on granule size and size distribution. Control of moisture content of granules in the drying phase. Arch Pharm Chem Sci. 1978;6:1-13.
6.
Abberger T, Raneburger J, Egermann, H. Instrumentation of laboratory-scale fluid-bed granulator for critical moisture content and of free moisture. Sci Pharm. 1996;64:255-262.
7.
Wöstheinrich K, Schmidt P. Evaluation and validation of a fully instrumented Hüttlin HKC 05-TJ laboratory-scale fluidized bed granulator. Drug Dev Ind Pharm. 2000;26:621-633.
PubMed
8.
Workman J, A review of process near infrared spectroscopy: 1980-1994. J Near Infrared Spectrosc. 1993;1:221-245.
9.
Axon T, Brown R, Hammond S, Maris S, Ting F. Focusing near infrared spectroscopy on the business objectives of modern pharmaceutical production. J Near Infrared Spectrosc. 1998;6:A13-A19.
10.
Kamat MS, Lodder RA, DeLuca PP. Near-infrared spectroscopic determination of residual moisture in lyophilized sucrose through intact glass vials. Pharm Res. 1989;6:961-965.
PubMed
11.
Osborne BG, Fearn T, Hindle PH. In: Practical NIR Spectroscopy with Applications in Food and Beverage Industry Analysis. 2nd ed. Harlow, UK: Longman;1993:227.
12.
Watano S, Terashita K, Miyanami K. Development and application of infrared moisture sensor to complex granulation. Bull Univ Osaka Pref, Series A. 1990;39(2):187-197.
13.
White J. On-line moisture detection for a microwave vacuum dryer. Pharm Res. 1994;11:728-732.
PubMed
14.
Frake P, Greenhalgh D, Grierson S, Hempenstall J, Rudd D. Process control and end-point determination of a fluid bed granulation by application of near infra-red spectroscopy. Int J Pharm. 1997;151:75-80.
15.
Goebel S, Steffens K. Online-messung der Produktfeuchte und Korngröße in der Wirbelschnicht mit der Nah-Infrarot-Spektroskopie. Pharm Ind. 1998;60:889-895.
16.
Rantanen J, Lehtola S, Rämet P, Mannermaa J-P, Yliruusi J. On-line monitoring of moisture content in an instrumented fluidized bed granulator with a multi-channel NIR moisture sensor. Powder Technol. 1998;99:163-170.
17.
Luukkonen P, Rantanen J, Mäkelä K, Räsänen E, Tenhunen J, Yliruusi J. Characterization of silicified microcrystalline cellulose and a-lactose monohydrate wet masses using near infrared spectroscopy. Pharm Dev Technol. 2001;6:1-9.
PubMed
18.
Räsänen E, Rantanen J, Jørgensen A, Karjalainen M, Paakkari T, Yliruusi J. Novel identification of pseudopolymorphic changes of theophylline during wet granulation using near infrared spectroscopy. J Pharm Sci. 2001;90:389-396.
PubMed
19.
Iwamoto M, Uozumi J, Nishinari K. Preliminary investigation of the state of water in foods by near infrared spectroscopy. Budapest, Hungary. Intl NIR/NIT Conference;1987:3-12.
20.
Delwiche S, Pitt R, Norris K. Examination of starch-water and cellulose-water interactions with near infrared (NIR) diffuse reflectance spectroscopy. Starch/Stärke. 1991;43:415-422.
21.
Rantanen J, Känsäkoski M, Suhonen J, et al. Next generation fluidized bed granulator automation. AAPS PharmSciTech 2000;1(2) Available from: http://www.aapspharmaceutica.com/scientificjournals/pharmscitech/volume1issue2/014/manuscript.htm.
PubMed
22.
Rantanen J, Räsänen E, Mannermaa J-P, Yliruusi J. In-line moisture measurement during granulation with a four wavelength near infrared sensor: an evaluation of particle size and binder effects. Eur J Pharm Biopharm. 2000;50:271-276.
PubMed
23.
Rantanen J, Laine S, Antikainen O, Mannermaa J-P, Simula O, Yliruusi, J. Visualization of pharmaceutical unit operations using self-organizing maps (SOM). J Pharm Biomed Anal. 2001;24:343-352.
PubMed
24.
Morris K, Stowell J, Byrn S, Placette A, Davis T, Peck G. Accelerated fluid bed drying using NIR monitoring and phenomenological modeling. Drug Dev Ind Pharm. 2000;26:985-988.
PubMed
25.
Watano S. Mechanism and control of granule growth in fluidized bed granulation [dissertation]. Osaka, Japan: College of Engineering, University of Osaka;1995.
26.
Watano S, Takashima H, Sato Y, Miyanami K, Yasutomo T. IR absorption characteristics of an IR moisture sensor and mechanism of water transfer in fluidized bed granulation. Adv Powder Technol. 1996;7:279-289.
27.
Miwa A, Yajima T, Itai S. Prediction of suitable amount of water addition for wet granulation. Int J Pharm. 2000;195:81-92.
PubMed
28.
Schmidt C, Kleinebudde P. Influence of the granulation step on pellets prepared by extrusion/spheronization. Chem Pharm Bull. 1999;47:405-412.
|
|
View PDF Version of this article
Email This Article