Hydroxypropyl Methylcellulose As A Novel Tool For Isothermal Solution Crystallization of Micronized Paracetamol

Subscriber access provided by FONDREN LIBRARY, RICE UNIVERSITY
Article
HPMC as a novel tool for isothermal solution
crystallisation of micronized Paracetamol
Nuno Miguel Reis, Zizheng K Liu, Cassilda M Reis, and Malcolm Mackley
Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/cg4009637 • Publication Date (Web): 23 May 2014
Downloaded from http://pubs.acs.org on June 9, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Crystal Growth & Design is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 22 Crystal Growth & Design
1
2 HPMC as a novel tool for isothermal solution crystallisation of micronized
3 Paracetamol
4
5
6 Nuno M Reisa,b,*, Zizheng K Liua, Cassilda M Reisa, and Malcolm R Mackleya
7 a
8 Department of Chemical Engineering and Biotechnology, University of Cambridge, New Museums
9 Site, Pembroke Street, Cambridge CB2 3RA, UK
10
b
11 Department of Chemical Engineering, Loughborough University, Loughborough LE11 3TU, UK
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 *
Corresponding author: Phone +44 (0) 1509 222 505; Fax +44 (0) 1509 223 923; e-mail n.m.reis@lboro.ac.uk
60
ACS Paragon Plus Environment 1
Crystal Growth & Design Page 2 of 22
1
3 Paracetamol
4
5
6 Nuno M Reisa,b,†, Zizheng K Liua, Cassilda M Reisa, and Malcolm R Mackleya
7 a
9 Site, Pembroke Street, Cambridge CB2 3RA, UK
10
b
11 Department of Chemical Engineering, Loughborough University, Loughborough LE11 3TU, UK
12
13
14 Abstract
15
16 Pulmonary inhalation is increasingly being selected as a preferred route for the delivery of both
17
18 small and large drug macromolecules for the treatment of a range of pathologies. The direct
19 crystallisation of micronized powders, in particular Paracetamol remains difficult, as it requires the
20
21 ability to work in high solution supersaturations where agglomeration, wall crusting and
22
23 heterogeneous nucleation hinders the control of crystal size and crystal size distribution. Polymer
24 additives are recognised to help driving the production of a given polymorph or controlling crystal
25
26 shape by means of adsorption on crystal surface. With the aim of exploiting the polymer-control
27
28 nucleation and growth of crystals for enhanced direct crystallisation of micronized powders, batch
29 cooling crystallisation of Paracetamol in water was carried out in the presence of 0.1-0.8% w/w
30
31 hydroxypropyl methylcellulose (HPMC). In the presence of polymer the onset of nucleation was
32
33 delayed and extended beyond the cooling time of the solution, resulting in an isothermal cooling
34 crystallisation and the production of micronized Paracetamol with a mean crystal size D50, in the
35
36 range of 15-20 µm and an improved crystal size distribution. Equally, the rate generation of
37
38 solution cloudiness was reduced by over 3-fold for the highest HPMC concentration tested, with no
39
detectable impact on final product yield. The mechanisms for nucleation delay and growth
40
41 inhibition by HPMC is unknown, however a modification of crystal’s shape observed upon the
42
43 addition of HPMC to the solution suggested it might be related to mass transfer limitations and
44
inter-molecular hydrogen bounding between the large HPMC and the small drug molecules. This
45
46 technique can potentially be used for direct crystallisation of other micronized drugs.
47
48
49
Keywords: Paracetamol, micronized, cooling crystallisation, isothermal crystallisation, HPMC
50
51
52
53
54
55
56
57
58
59 †
60
1
2 Introduction
3
4 Pulmonary inhalation is being increasingly selected as a preferred route for the delivery of both
5
6 small and large drug macromolecules for the treatment of a range of pathologies. The fast
7 expanding dry powder inhalers (DPI) market is expected to reach over US$13bn per annum by
8
9 2016, benefiting from the remarkable development in drug formulation and inhalation device
10
11 designs mainly in recent years.1 Dispatch such technological advancement, the production of
12 micronized (∼1-10 µm) active pharmaceutical ingredients (APIs) still relies on traditional milling
13
14 and grinding techniques for size reduction of crystals. These methods apply high energy that result
15
16 in a final product with broad size distribution, limited crystallinity and poor flowing properties with
17 limited dispersability.2 An alternative technique that is in increasing adoption by industry is spray
18
19 drying.3 It allows producing micronized APIs of desirable size however particles are amorphous and
20
21 have a larger tendency to re-crystallise or degrade.2 It is therefore recognised that direct
22 crystallisation of micronized crystals solution is advantageous, however producing small crystal
23
24 sizes requires operation with the high supersaturations at which the control of agglomeration, wall
25
26 crusting, crystal size and crystal size distribution (CSD) becomes extremely challenging.4
27
28 Crystallisation is an important process in the chemical, pharmaceutical, biotechnological and allied
29
30
industries, as it is used extensively for separation and purification of organic fine chemicals or APIs
31 and production of microsized APIs for drug delivery.5 Important product characteristics such as
32
33 crystal size, CSD, and crystal morphology are determined by the operating conditions during the
34
35
crystallization process, which include: supersaturation, temperature profiling, the presence of
36 additives, air-water interface, anti-solvent addition rate, fluid mixing, residence time, materials of
37
38 crystalliser, and agglomeration/breakage phenomena.
39
40 Stagnant crystallisation experiments have shown that polymers and other additives can help driving
41
42 the production of a given polymorph,6-9 reducing the crystal size10 or controlling the crystal shape.9
43 The list of polymers tested is extensive and includes both hydrophilic and hydrophobic powders,
44
45 such as ethylcellulose, methylcellulose (CM), polyethylene glycol (PEG), polyvinyl alcohol (PVA),
46
47 agar, gelatine, polyvinyl pyrrolidone (PVP), poly(ethylene oxide) (PEO), corn starch, carrageenan,
48 polyethylene (PE), polyvinyl chloride (PVC), polymethyl methacrylate (PMMA) and
49
50 hydroxylpropyl methylcellulose (HPMC) to name a few.3, 8-12
Generally, crystallisation additives
51
52 are effective in producing polymorphs of pharmaceuticals while avoiding others by encouraging the
53 growth of a desirable form or disrupting the growth of the other form.8, 9, 13, 14
Modification of
54
55 crystal shape occurs as the polymer deposition on the crystal surface inhibits the growth of specific
56
57 surfaces of the crystal.13 Reduction of particle size by means of long molecular weight additives was
58 only marginally reported in the literature, and exception is perhaps the work of Femi-Oyewo and
59
60
1
2 Spring10, although a recent work by Xie et al.12 showed that salbutamol sulphate crystal size was
3
reduced to less than 10 µm upon presence of polymer additives HPMC, polyvinylpyrrolidone (PVP
4
5 K25), lecithin and Span 85 during anti-solvent water-ethanol crystallisation.
6
7 There is also very few studies touching the effect of organic polymers in the kinetics of
8
9 crystallisation. An exception is the work of Raghavan et al.15 who have studied anti-solvent
10
11 crystallisation of hydrocortisone acetate (HA) in the presence of four different polymers (HPMC,
12 PVP, MC and PEG400). In the absence of solution additive, the nucleation of HA was observed to
13
14 be spontaneous; however in the presence of a polymer the nucleation was delayed by several hours
15
16 and a level of growth inhibition of HA was observed. The mechanism of nucleation retardation was
17 explained by the authors in terms of association of HA with the polymer through hydrogen bonding
18
19 whilst the growth inhibition was related to the adsorption of polymer to the surface of the crystal.
20
21 HPMC is a natural polymer extensively used in the pharmaceutical industry as a tableting ingredient
22
23 and as a binder16 because of its regulatory approval status. Steckel et al.2 have shown that HPMC
24
can act as a stabilising hydrocolloid during the production of micronized fluticasone-17-propionate
25
26 by spray drying, resulting in one order of magnitude reduction in particle size.
27
28 In this work small concentrations of HPMC were used for manipulating the kinetics of cooling
29
30 crystallisation of Paracetamol, which offered a new level of control over the nucleation and crystal
31
32 growth under well mixed conditions and allow reduced crystal size and improved CSDs. The mean
33 crystal size, CSD and crystal morphology have been experimentally evaluated for a concentration of
34
35 HPMC in range of 0 - 0.8% w/w using standard powder characterisation techniques.
36
37
38 Experimental section
39
40 Materials
41
42 Paracetamol or acetaminophen (CAS number, 103-90-2, min 99% purity) and HPMC (CAS 9004-
43
44
65-3, typical MW = 10,000) were supplied by Sigma-Aldrich, Inc (Dorset, UK).
45
46 Crystallisation procedure
47
48 Paracetamol was crystallised in water by dissolving a given amount of Paracetamol at 70 oC
49
50 corresponding to a saturated solution at 65 ºC, and rapidly cooling it to room temperature (20 ºC).
51 For each experiment, 3.0 grams of Paracetamol were dissolved in 60 grams of deionised water in
52
53 the presence of 0-0.8% w/w of HPMC (percentage of mass of HPMC per mass of water), giving a
54
55 maximum drug to polymer ratio of 6.25:1 in weight or 413:1 in molar basis. The crystallisation
56 vessel consisted of 42 mm internal diameter, d, glass-jacketed unbaffled tank with a liquid height, h
57
58 = d corresponding to a working volume of 60 ml (Figure 1). The solution was magnetically stirred
59
60
1
2 at 70 ºC for 1 hour to ensure complete dissolution of Paracetamol, by circulating hot water (HAAK,
3
Fisons DCS B3, Pacific Diagnostic, Inc. USA). The maximum working volume of the tank was 83
4
5 ml, and all experiments were done in the presence of a free surface. Mixing was provided using an
6
7 18 mm length magnetic stirrer at 300 rpm, controlled by a stirrer plate (KIKA, Labortechnik, Janke
8
& Kunkel Gmbh & Co KG. UK).
9
10 The Paracetamol solubility estimated from Granberg&Rasmuson19 for working temperature of 65
11
12 ºC and 0% HPMC was ∼48.3 grams per kilogram of water, i.e. ∼2.9 grams per 60 ml of water.
13
14 The cooling crystallisation was started by quickly switching circulating hot water in a 3-way valve
15 to chilled water at 20 ± 0.1 ºC from a refrigerated/cooling circulator (Grant LTD20, Cambridge,
16
17 UK), which induced a rapid temperature quenching in the crystallisation vessel. The experiments
18
19 were continued for about 45 minutes from the start of the cooling process, which was found
20 sufficient to complete of crystals growth and fully deplete supersaturation in the solution. The
21
22 supersaturation ratio of pure paracetamol solution defined as S = C/C* was estimated as equal to 3.8
23
24 based on the solubility data for pure paracetamol reported by Granberg&Rasmuson.19 The initial
25 paracetamol concentration remained constant for all HPCM concentrations tested.
26
27 The crystallisation vessel was equipped with a thermocouple (TME 2050, k-type) and a reflectance
28
29 fibre optical probe for on-line monitoring of the temperature and turbidity of crystallisation
30 solution, respectively (Figure 1). Details of the optical probe are given in section 2.3.
31
32 At the end of each crystallisation run, the slurry was vacuum filtered through a 0.2 µm cut-off PTFE
33
34 membrane (Whatman Inc, USA) and the crystals washed with 3*1.5 ml of deionised water at room
35 temperature. Paracetamol crystals were then dried in a desiccator for about 12 hours after which the
36
37 dry weight was determined in an analytical balance. The crystal size and CSD were then
38
39 characterised using an automated optical microscopy system, and crystal morphology analysed by
40 Scanning Electron Microscopy (SEM). Crystal system was identified using powder X-ray
41
42 diffraction (XRD) and the thermal behaviour using Differential Scanning Calorimetry (DSC)
43
44 techniques.
45
46 Monitoring of solution turbidity
47
48 Solution cloudiness or turbidity was monitored real-time using a reflectance optical fibre probe. The
49
50 optical probe (FCR 7UV200, Avantes, Eerbeek, Netherlands) consisted of 6 × 200 µm diameter
51
optical fibres carrying the light from a deuterium halogen light source and a 7th fibre in the core of
52
53 the probe delivering the scattered light to a spectrometer connected to a PC, which was controlled
54
55 by Avasoft (Avantes, Eerbeek, Netherlands). The probe was inserted though the lid of the
56
crystalliser at a certain angle to avoid direct reflection from the walls of the crystalliser. The
57
58 crystalliser was fully covered with aluminium foil to impede interference of environment light. The
59
60
1
2 reflected light spectrum was continuously scanned through the experimental time, t, with a CCD
3
spectrometer using a proper integration time. An integration function was then defined to monitor
4
5 the total reflected light, Iw or the photon counts in the spectrum for a wavelength range, w, between
6
7 585 and 615 nm, where crystal particles were found to scatter light strongly:
8 w= 615nm
9
10
R( t ) = ∫I w
w=585 nm
(t ) (1)
11
12 The relative solution transmittance, T(t) defined as the fraction of incident light that passed through
13
14 the solution in the monitored wavelength range w, was then calculated as T(t) = R(t)/R(t=0) and
15 normalised for each experiment by assuming 100 % light transmittance at time zero, i.e. T(t=0) = 1
16
17 and 0% transmittance towards the end of the crystallisation, i.e. T(t=∞) = 0.
18
19
20 Crystal characterisation
21
22 The mean crystal size and CSDs of Paracetamol powder were determined using an automatic
23 optical system, the Morphologi G3 (Malvern Instruments, Malvern, UK). Optical microscopy
24
25 offered a more robust CSD method for sizing Paracetamol crystals in comparison with laser
26
27 diffraction measurements because it offers the possibility of analysing the sample in the dispersed
28 powder form using compressed air at 1.5 barg and avoided sonication process for dispersion of
29
30 crystals, which has shown to result in the partial dissolution or agglomeration of Paracetamol
31
32 crystals (results not shown). Paracetamol is very prone to agglomeration, therefore CSD analysis is
33 especially difficult. Nevertheless, the Morphologi G3 system allows the reproducible, automatic
34
35 analysis of a large number of particles or crystals and the possibility of using different types of
36
37 filters to manually discard large agglomerates resulting from crystal caking during the sampling and
38 powder drying process.
39
40 For CSD analysis using Morphologi G3 1.5 mg of dry powder were loaded in 20 µm thick
41
42 aluminium foil and dispersed into the clean, glass surface of the microscope using compressed air at
43 1.5 barg. The G3 software measured the projected area of the crystals/particles and assigned it the
44
45 equivalent diameter of a circle, CE, through a standard operating procedure (SOP). The crystal size
46
47 computed this way was based on CE3, which would only fit laser scattering measurements for
48 perfectly spherical particles. A typical run comprised a sample of 5,000-10,000 crystals to ensure
49
50 the statistical significance of the acquired CSD. The reproducibility of CDS measurements was
51
52 found very high upon SOP optimisation (results not shown).
53 For SEM of the micronized powders, the dried crystals were spread over a surface and dried
54
55 overnight in He atmosphere. The samples were then coated with a 20 nm thick gold coating and
56
57 scanned at 5 Kev in FEI Philips XL30 FEGSEM equipped with an Oxford Instruments INCA EDX
58 system running a 30 mm2 SiLi thin window pentafet EDX detector.
59
60
1
2 The effect of HPMC on the thermal behaviour above ambient of dry Paracetamol powder was
3
recorded using a Perkin-Elmer Pyris 1 scanning differential calorimeter. For that purpose,
4
5 approximately 5 mg of sample were heated at 10 ºC/min from 40 to 180 ºC in perforated crimped
6
7 aluminium pans while being purged with dry nitrogen. The heating and cooling cycles were
8
performed, intercalated by 1 min holding time.
9
10 For analysis of the effect of HPMC on crystal system, powder X-ray diffractions in dried powder
11
12 were acquired at room temperature on a Philips PW1820 diffractometer using Copper-k-alpha
13
14
radiation (tube operated at 40 kV, 40 mA), a θ-θ goniometer, automatic divergence and 0.2 mm
15 receiving slits, a silicon secondary monochromator, and a scintillation counter. The XRD trace was
16
17 scanned in 2-θ from 5º to 60º, at a rate of 0.050º of 2θ per 2 seconds. The powder samples were
18
19 prepared as flat surfaces in aluminium sample holders.
20
21 Results and discussion
22
23 Effect of HPMC on the kinetics of Paracetamol crystallisation
24
25 Paracetamol was crystallised in water in a small batch stirred glass vessel containing a reduced
26
27 HPMC concentration, by quickly dropping the temperature of a slightly undersaturated solution
28
from 70 oC to 20 oC. Turbidity was on-line monitored during the cooling process using an optical
29
30 microprobe immersed in the solution, which allowed sensitive detection of the onset of nucleation
31
32 and quantitative detection of the rate of generation of cloudiness, which is ultimately linked to the
33
generation and growth of crystals. Figure 2 shows the variation of solution temperature and relative
34
35 transmittance (the lower the transmittance the higher the solution cloudiness) for a selected number
36
37 of HPMC concentrations. The HPMC concentrations tested in this study were 0.0, 0.1, 0.3, 0.4, 0.7
38
and 0.8% w/w but only a selection of these is shown in Figure 2 for simplicity in clarity of data
39
40 presented. The temperature cooling profile was highly reproducible in the whole set of experiments,
41
42 therefore only one cooling curve is shown in Figure 2. HPMC effectively delayed the onset of
43
nucleation and reduced the rate of generation of cloudiness in the solution as can be confirmed from
44
45 the plots. With pure Paracetamol, a shower of crystals was detected within 1-2 minutes, and 90% of
46
47 the solution cloudiness generated within the following minute, and crystal growth was completed
48
within 5 minutes from the beginning of the cooling process, which compares well with the cooling
49
50 time of the solution in the vessel. In the presence of HPMC, the time of onset of nucleation
51
52 extended to up to 9 minutes depending on the HPMC concentration, so well beyond the cooling
53
time of the solution in the vessel. This ultimately means that the crystallisation in the presence of
54
55 HPMC occurs entirely under full isothermal conditions, at which the solution supersaturation is at
56
57 its maximum and therefore a shower of crystals beneficial for the production of micronized powders
58
is expected. The ability of running solution crystallisations in full isothermal conditions as shown in
59
60
1
2 this manuscript is novel and might find broad applications in the direct crystallisation of micronized
3
APIs.
4
5
6 The turbidity profiles in Figure 2 did not allow the full deconvolution of crystal nucleation and
7 crystal growth rates, however some quantitative information was extracted in order to quantify the
8
9 effect of polymer additive in the overall batch cooling crystallisation kinetics. For that purpose, an
10
11 arbitrary induction time, tind and crystal growth completion time, tG were defined corresponding to
12 the time required to give a 20% and 80% decrease, respectively, in solution transmittance. For this
13
14 analysis the whole range of HPMC concentrations tested was considered. Equally, a maximum
15
16 generated cloudiness rate, Rc, was calculated from the initial gradient of relative solution
17 absorbance versus cooling time, which is summarised in Figure 3a. During the initial stage of
18
19 crystal growth, when submicron nuclei grow to a size to be optically detected and therefore the
20
21 concentration of crystals is proportional to the absorbance of the solution, the increase in solution
22 absorbance (defined as –log10 of transmittance) is essentially related to the increase in the
23
24 concentration of particles or the formation of new crystals, therefore Rc should capture the relative
25
26 effect of polymer additive concentration on kinetics of crystal nucleation. Figure 3a showed that tind
27 increased by up to 5.4-fold and tG by up to 9.5-fold upon the addition of HPCM to the solution, with
28
29 the maximum effect being observed at the highest HPMC concentration tested of 0.8% w/w.
30
31 Equally, Rc decreased up to 73.4% with increasing HPMC concentration, being the inhibitory effect
32 more noticeable at low HPMC concentrations. This represents a major effect of polymer additive in
33
34 respect to the overall kinetics of crystallisation, from where a major reduction in crystal size could
35
36 be expected.
37
38 Effect of HPMC on mean crystal size and CSD
39
40 A common motivation for controlling crystallisation processes in pharmaceutical industries is the
41
42 production of uniform particle sizes with a given mean crystal size. Therefore, the Paracetamol
43 powder produced in the presence of different HPMC concentrations has been characterised in
44
45 respect to particle size using the Morphologi G3 system, which applies compressed air for
46
47 dispersing dry powders on a glass microscope slide and advanced imaged analysis to determine a
48 circle equivalent diameter, CE of the crystals. This allowed confirming that the presence of HPMC
49
50 in the initial crystallisation solution resulted in a significant reduction in the mean crystal size, D50
51
52 and an improvement in CSD as shown in Figures 4a and 4b. With pure Paracetamol the D50
53 obtained was 39.6 µm. This was per si considerably smaller than the mean sizes reported by other
54
55 authors for batch cooling crystallisation from solution. For example, Chew et al.17 and Fujiwara et
56
57 al.5 reported a D50 in the range of 100-250 µm for batch crystallisers, and Zarkadas & Sirkar18
58
59
produced crystal sizes with 50-150 µm in a continuous hollow fibre device. The smaller crystal
60
1
2 sizes obtained in this study are related to the large surface area-to-volume ratio of the glass vessel
3
used in the crystallisation runs, which delivered high cooling rates for high supersaturation and
4
5 nucleation rates that cannot be mimicked in larger crystallisers. Nevertheless, the CSD in Figure 4a
6
7 was broad as can be seen from be great deviation of the values represented by percentile 10, D10 =
8
15.5 µm and percentile 90, D90 = 62.1 µm. This could be associated with the non-isothermal
9
10 conditions in the crystallisation vessel at the time of onset of solution cloudiness. The addition of
11
12 0.1% w/w of HPMC resulted in D50 decreasing from 39.6 to 15.4 µm and the volume-based CSD
13
14 becoming remarkably sharper, with D10 = 8.3 µm and D90 = 22.6 µm. The frequency-based CSD in
15 Figure 4b confirmed a larger number of fines in the dried powder crystallised in the presence of
16
17 0.1% w/w HPMC, and smaller number of large crystals. The dashed vertical line represents the
18
19 smallest crystal size that could be detected in with the used setup with Morphologi G3. That
20 significant decrease in the mean crystal size was confirmed by SEM images in Figures 4c and 4d
21
22 (note the different scales in the SEM microphotographs).
23
24 The effect of varying initial concentrations of HPMC on particle size distribution in respect to
25
26 percentile 10, D10, percentile 50, D50 and percentile 90, D90 is fully summarised in Figure 3b. The
27 use of HPMC concentrations above 0.1% w/w resulted in no further improvement in particle size
28
29 reduction or CSD, and it appeared that the presence of a very small concentration of HPMC in the
30
31 solution was sufficient to provide a good level of particle size control.
32
33 Another aspect that contributed to smaller crystal sizes was the fact that Paracetamol solubility
34
35
decreases very quickly with HPMC content, as shown in supplementary data (Figure S1). For a
36 0.1% mass ratio of HPMC to Paracetamol (corresponding to approximately 0.03% w/w of HPMC
37
38 in respect to mass of solution) the solubility measured at 30 oC dropped by 40%. This means that in
39
40
the presence of HPMC the crystallisation is isothermal but also supersaturation is higher than for a
41 pure Paracetamol system. This should in theory return higher crystallisation yields, and in respect to
42
43 particle size control this represents enhanced supersaturation in the presence of HPMC which
44
45
favours production of smaller crystals and more fines.
46
47 Overall and based on the turbimetry data, it is not possible to identify the main mechanic leading to
48 a significant reduction in crystal size and improvement in CSD in the presence of HPMC, but it
49
50 appears linked to a higher supersaturation and tighter supersaturation control resulting from the
51
52 delay of onset of nucleation, higher number of fines and reduction in Paracetamol solubility.
53
54 Effect of HPMC on crystal morphology
55
56 A number of recent studies mainly performed at static conditions have linked the use of organic
57
58 polymer additives with crystal polymorphism as reviewed in the Introduction section. A given
59 crystal form can be inhibited upon the presence of polymer or “impurities”, therefore polymer
60
1
2 additives can induce selective production of a given polymorph.6-9 Equally, other studies have
3
reported a modification of crystal habit upon the addition of polymer additives.8,9 Mainly with
4
5 isentropic crystals like Paracetamol,20 polymers can present higher selectivity to bind given faces a
6
7 crystal, so selectively inhibiting the growth on a given direction. In order to test for possible effects
8
of HPMC additive on crystal structure, Paracetamol powders were characterised regarding their
9
10 thermal behaviours and X-ray diffraction patterns. Optical and electronic microscopic analysis of
11
12 the crystals showed that crystals shape changed to an elongated-prismatic shape upon the addition
13
of polymer. The DSC and XRD data in Figures 5 and 5b, respectively, confirmed that the
14
15 crystalline product obtained with pure Paracetamol and 0.1% w/w of HPMC were in fact from the
16
17 same crystal form I, i.e. monoclinic Paracetamol. The experimental XRD patterns of the two
18
Paracetamol samples crystallised with 0% and 0.1% w/w of HPMC compared in Figure 4 agreed
19
20 well with XRD data of e.g. Martino et al.21 and Nichols and Frampton22 by showing peaks unique
21
22 to the monoclinic form I such as 2θ = 12.025, 15.425 and 26.475. Also, the melting point
23
24 (endothermic) at 164 ºC and two crystallisation (exothermic) points at 88–89ºC and 135–139 ºC
25 shown by DSC analysis in Figure 5b also confirmed that the two samples were of the same crystal
26
27 form. This confirmed that the detected difference in kinetics of crystallisation is not due to the
28
29 production of a different polymorph.
30
31 Detailed SEM analysis to all Paracetamol samples revealed a side effect with the use of HPMC
32 concentrations of 0.3% w/w and above, with extended agglomeration and a large number of cavities
33
34 or holes on the crystal surface. Figure 6 shows cavities up to 100-200 nanometers in diameter that
35
36 become more prominent as more HPMC was added to the solution. The reason behind the
37 formation of these cavities is unknown, however it is suspect to be linked to the intermolecular
38
39 interactions of HPMC and Paracetamol, with HPMC adsorbing the crystal surface as suggested by
40
41 Thompson et al.23 for Metacetamol (a small molecule with –OH groups). Using AFM and SEM,
42 they have shown surface features in pure Paracetamol crystals between 1 and 20 nm, that changed
43
44 to 15 nm steps interspersed with holes in the presence of Metacetamol.
45
46 Effect of HPMC on final product yield
47
48 The product yield calculated in terms of dried powder weight was found independent of the HPMC
49
50 concentration used, and in average 59.5 ± 1.62% as shown in Figure 3c. This was slightly lower
51
52 than the theoretical yield expected from pure Paracetamol (74.1%) based on the solubility data of
53 Granberg&Rasmuson.19 Note that the addition of HPMC to the solution can only increase the
54
55 theoretical yield up to a maximum of 3.6% assuming all HPMC added to the solution ends up in the
56
57 dried powder. Based on the solubility data shown in Figure S1 the addition of HPMC to the
58 crystallisation solution should have resulted in higher yield than for pure Paracetamol, as HPMC
59
60
1
2 was found to decrease significantly the solubility of Paracetamol in water. This might be related to
3
the dissolution of some material during the sample washing or to surface poisoning. Femi-Oyewo &
4
5 Spring10 reported 91% reduction on yield of Paracetamol crystallised in the presence of HPMC
6
7 under static conditions, which was shown to be related to the irreversible adsorption of polymer of
8
the surface of the crystals. This data herein reported shows that under well mixed conditions the
9
10 adsorption of polymer to the surface of the crystals is reversible however it cannot entirely
11
12 overcome surface poisoning that leads to the terminal size of the crystal.
13
14 Remarks on the effect of HPMC on kinetics of Paracetamol crystallisation
15
16 The mechanics for delaying the onset of nucleation and reducing the rate of solution cloudiness in
17
18 cooling crystallisation of Paracetamol crystals by HPMC is still unknown, however recent studies
19
have proposed different hypothesis spanning from inhibition of heterogeneous nucleation to
20
21 increased mass transfer resistances both related to the hydrogen bounding between the polymer and
22
23 Paracetamol molecules. This section offers an overview to those important mechanisms that support
24
the possible impact of HPMC on the nucleation and/or growth of Paracetamol crystals.
25
26
27 Heterogeneous nucleation is thermodynamically more favourable and is known to be related to the
28 presence of interfaces, such as crystalliser’s wall, impeller or air-waste free surface. Zarkadas
29
30 &Sirkar18 have found that a combination of homogeneous and heterogeneous nucleation is possible
31
32 in a cooling crystallisation process of Paracetamol in water at high supersaturation ratios, i.e. S =3
33 and above, which was the case in this present study. In order to validate this hypothesis a further set
34
35 of experiments was carried out in a combination of experimental conditions that were recognised to
36
37 promote heterogeneous nucleation, such is the case of stirring, glass wall material and air-water
38 interface, as summarised in Table 1. For each crystallisation run, the time required to detect the
39
40 onset of solution cloudiness was noted (in these set of experiments the induction time was
41
42 determined using naked eye). It was observed that the addition of 0.1 % w/w HPMC has similar
43 effect in respect to the delay on induction time to the removal of free surface during the
44
45 crystallisation vessels, resulting in delayed induction time from 4.0 to 9-10 min. This is likely to be
46
47 associated with the fact that HPMC reduced the frequency of collisions between the solute particles,
48 which, in turn, led to the delay of nucleation. The rate for primary nucleation is linearly
49
50 proportional to the diffusion coefficient4 of Paracetamol, having a molecular weight of 151.2 g/mol.
51
52 When the diffusion is sufficiently rapid the nucleation is not diffusion controlled. This is normally
53 the case at high values of S, and explains the very short induction time obtained with pure
54
55 Paracetamol (run 1 in Table 1). Nevertheless, with the presence of HPMC in the crystallisation
56
57 solution the nucleation became increasingly limited as more polymer additive is added to the
58 solution. This was explained by Zarkadas&Sirkar18 and Tian et al.24 based on inter-molecular
59
60
1
2 hydrogen bonding between Paracetamol and the polymer additive. HPMC has both a hydrogen
3
bonding donor and acceptor groups on its ring structure (Figure S2), therefore showing a high H-
4
5 bonding ability and network association with drug molecules as concluded from FT-IR results of
6
7 Tian et al.24 Similar conclusion was obtained by Sahoo et al.25 using Raman spectroscopy. The last
8
have identified intra-molecular hydrogen bounding between HPMC and Ciprofloxacin, resulting
9
10 from hydroxyl OH- groups present in both polymer additive and drug molecules. Raghavan et al.15
11
12 has also reported strong effect of cellulose polymers on morphology of hydrocortisone acetate. This
13
means HPMC effectively can work as a “molecular transporter” therefore limiting both the
14
15 diffusivity of drug molecules and reducing the impact probability for the formation of nuclei
16
17 through collisions.
18
19 The production of smaller crystals and the reported morphological changes could also be explained
20
21 based on several growth-inhibition hypothesis, such as i) mass transfer limitations, ii) impurities
22 deposition,25iii) polymer adsorption on the surface of the crystals, or iv) a combination of those.
23
24 During growth some impurities can adsorb to the surface of the crystals “blocking” the active
25
26 growth sites (sometimes only blocking growth of a given face of the crystal) or entirely suppresses
27 crystal growth.26 HPMC was previously shown capable of adsorbing the surface of Paracetamol
28
29 crystals.9 This has been linked to a reduction in the growth rate of steps when compared to pure
30
31 Paracetamol. Using AFM and SEM, Thompson et al.23 showed that metacetamol changed the
32 growth steps of Paracetamol and disrupting the crystal lattice by adsorbing to the surface and
33
34 inducing the formation of defects or cavities in the crystal surface.
35
36 A parallel observation from Table 1 is the relevant effect of mixing and interfaces on the
37
38 crystallisation of Paracetamol in respect to induction time in the presence of 0.10% w/w HPMC.
39
Although it is difficult to separate the effect of interfaces from that of mixing, it was observed that
40
41 induction time approximately doubled in the absence of headspace in the crystallisation vessel.
42
43 Interfaces are known for creating temperature gradients therefore being preferential zones for
44
nucleation. On the other hand, the removal of agitation resulted in further increased induction times
45
46 (rows 6 and 7), however this is believed to be related to the lower cooling rate in unstirred vessel.
47
48
49 Conclusions
50
51 Batch cooling crystallisation experiments of Paracetamol in water in the presence 0.1-0.8% w/w of
52
53 HPMC revealed the ability of the natural polymer additive modifying the overall kinetics of
54 crystallisation by slowing the rate of generation of cloudiness in the solution and delaying the onset
55
56 of nucleation beyond the cooling time of a crystalliser. This allowed direct production of
57
58 micronized Paracetamol crystals from solution, with a D50 in the range of 15-20 µm and sharp CDS,
59 which was linked to the isothermal crystallisation conditions in the presence of HPMC and
60
1
2 reduction of Paracetamol solubility in the presence of polymer. The observed habit change induced
3
by the presence of HPMC suggested the polymer acted by inhibiting growth on specific surfaces of
4
5 the crystals, as previously reported for other crystallisation systems. The high density of cavities
6
7 observed on the surface of the crystals for higher HPMC concentrations strongly suggests that
8
HPMC adsorbed to surface of the crystals disrupting the hydrogen bounding network. This
9
10 technique can potentially be used for enhancing direct crystallisation of other micronized drugs for
11
12 pulmonary drug delivery.
13
14
Acknowledgment
15
16
17 The authors are grateful to Dr Jeremy Skepper from Multi Imaging Centre, Cambridge, for help
18 with SEM; Dr Eric Rees and Dr Joanna Stasiak for help with XRD; Zlatko Saracevic for help with
19
20 DSC analysis, and Henrietta Adomako at Loughborough University for assistance with Paracetamol
21
22 solubility experiments. NMR is grateful to Fundacao para a Ciencia e Tecnologia, FCT
23 (SFRH/BPD/26904/2005) and European Commission Marie Curie IEF programme (project
24
25 221768) for financial support.
26
27
28 Supporting Information Available
29
30 Solubility data for Paracetamol in the presence of different mass ratios of HPMC, and chemical
31
32 structure of Paracetamol and HPMC molecules. This information is available free of charge via the
33
34
Internet at http://pubs.acs.org/.
35
36 References
37
38 (1) Pulmonary Drug Delivery Systems: Technologies and Global Markets; Rockville, MD, USA,
39 2012; p. 222.
40
41 (2) Steckel, H. International Journal of Pharmaceutics 2003, 258, 65–75.
42 (3) Martin, G. P.; Zeng, X. M. Pharmaceutical composition for pulmonary delivery.
43 WO/2001/058425, 2001.
44
(4) Garside, J.; Mersmann, A.; Nyvlt, J. Measurement of crystal growth and nucleation rates; 2nd
45
46 editio.; IChemE: Rugby, UK, 2002.
47 (5) Fujiwara, M.; Nagy, Z. K.; Chew, J. W.; Braatz, R. D. Journal of Process Control 2005, 15,
48 493–504.
49
50 (6) Grzesiak, A. L.; Matzger, A. J. Crystal growth & design 2008, 8, 347–350.
51 (7) Price, C. P.; Grzesiak, A. L.; Matzger, A. J. Journal of the American Chemical Society 2005,
52 127, 5512–7.
53 (8) Lang, M.; Grzesiak, A. L.; Matzger, A. J. Journal of the American Chemical Society 2002,
54
55 124, 14834–5.
56 (9) Capes, J. S.; Cameron, R. E. CrystEngComm 2007, 9, 84.
57 (10) Femi-Oyewo, M. N.; Spring, M. S. International Journal of Pharmaceutics 1994, 112, 17–28.
58
59
60
1
2 (11) Pawar, A.; Paradkar, A. R.; Kadam, S. S.; Mahadik, K. R. Indian Journal of Pharmaceutical
3 Sciences 2007, 69, 658–664.
4 (12) Xie, S.; Poornachary, S. K.; Chow, P. S.; Tan, R. B. H. Crystal Growth & Design 2010, 10,
5
6 3363–3371.
7 (13) Davey, R. J.; Blagden, N.; Potts, G. D.; Docherty, R. Journal of American Chemical Society
8 1997, 7, 1767–1772.
9
(14) Gu, C.-H.; Chatterjee, K.; Young, V.; Grant, D. J. Journal of Crystal Growth 2002, 235, 471–
10
11 481.
12 (15) Raghavan, S. L.; Trividic, a; Davis, a F.; Hadgraft, J. International journal of pharmaceutics
13 2001, 212, 213–21.
14
15 (16) Weiner, M. L.; Kotkoskie, L. A. Excipient Toxicity and Safety (Drugs and the
16 Pharmaceutical Sciences); Marcel Dekker: New York, USA, 2000.
17 (17) Chew, C. M.; Ristic, R. I.; Dennehy, R. D.; De Yoreo, J. J. Crystal Growth & Design 2004, 4,
18 1045–1052.
19
20 (18) Zarkadas, D. M.; Sirkar, K. K. Industrial & Engineering Chemistry Research 2007, 46, 2928–
21 2935.
22 (19) Granberg, R. A.; Rasmuson, Å. C. Journal of Chemical & Engineering Data 2000, 45, 478–
23
24
483.
25 (20) Heng, J. Y. Y.; Williams, D. R. Langmuir : the ACS journal of surfaces and colloids 2006, 22,
26 6905–9.
27 (21) Martino, P. Di; Guyot-Hermann, A.-M.; Conflant, P.; Drache, M.; Guyot, J.-C. International
28
29 Journal of Pharmaceutics 1996, 128, 1–8.
30 (22) Nichols, G.; Frampton, C. S. Journal of pharmaceutical sciences 1998, 87, 684–93.
31 (23) Thompson, C.; Davies, M. C.; Roberts, C. J.; Tendler, S. J. B.; Wilkinson, M. J. International
32
journal of pharmaceutics 2004, 280, 137–150.
33
34 (24) Tian, F.; Baldursdottir, S.; Rantanen, J. Molecular Phamaceuticals 2009, 6, 202–210.
35 (25) Sahoo, S.; Chakraborti, C. K.; Behera, P. K. International Journal of Applied
36 Pharmaceuticals 2012, 4.
37
38 (26) Myerson, A. Handbook of Industrial Crystallization; Myerson, A., Ed.; Butterworth
39 Heinemann: Boston, USA, 2001.
40 (27) Mullin, J.W. Industrial Crystallization. Butterworth- Heinemann: London, 1993.
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2 Tables
3
4
5
Table 1. Induction times detected by naked eye, tind’ for batch cooling crystallisation of
6
7 Paracetamol with HPMC under different nucleation conditions
8
9 Run Wall Water-air Mixing HPMC concentration tind’
10 material interface (w/w) (min)
11 1 Glass Yes Yes 0.00% 1.7
12
13 2 Glass Yes Yes 0.10% 4.0
14 3 Glass No Yes 0.10% 9.5
15 4 LLDPE Yes Yes 0.10% 4.4
16 5 LLDPE No Yes 0.10% 10.5
17
18 6 LLDPE Yes No 0.10% 15.2
19 7 LLDPE No No 0.10% 20.5
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
Figures list
4
5 thermocouple
6
7
8
9 light
10
11 source
12
13 spectrometer
14
15 optical probe
16
17
18
19 3-way
20
21
valve
22
23
24 hot
25 water
26
27
magnetic plate
28
29 cold water
30
31
32 Figure 1. Experimental setup used in the batch cooling crystallisation experiments.
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2 80
Temp. [ºC]
3
4
5 60
6
7
8
40
9
10 20
11
12 1.2
13
14
Relative transmitance [a.u.]
15 1.0
16
17
18 0.8
19
20 0.8% w/w
21 0.6
22
23
24 0.4 0.4% w/w
25
26 0.1% w/w
27 0.2
28 0% w/w
29
30 0.0
31
32
0 10 20 30 40
33 cooling time [min]
34
35 Figure 2. Real-time profiles temperature and relative solution transmittance (or turbidity) in the
36 vessel during batch cooling crystallisation. For clarity, only HPMC concentrations of 0%, 0.1%,
37
38 0.4% and 0.8% w/w were shown.
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2 tind and tG -Rc
3 [min] [a.u./min]
4 25 0.04
5
6 tG
7 20
0.03
8
9 15
10
11 0.02
12 10 tind
13
14 0.01
15 5 -Rc
16
17
0 0
18 0.00 0.20 0.40 0.60 0.80 1.00
19
20 HPMC concentration [gHPMC/100 g solvent]
21 (a)
22
100
23
24
25 80
26
27
Crystal size [µm]
28 60
29
30
31 40
32
33 D90
20
34 D50
35
D10
36 0
37 0.00 0.20 0.40 0.60 0.80 1.00
39 (b)
40
100
41
42
43 80
Product yield [% w/w]
44
45
60
46
47
48 40
49
50
20
51
52
53 0
54 0.00 0.20 0.40 0.60 0.80 1.00
56 (c)
57
58
59
60
1
2 Figure 3. Effect of initial HPMC concentration on (a) induction time, tind crystal growth time, tG
3
and maximum rate of generation solution cloudiness, Rc as determined from turbidity plots in
4
5 Figure 2; (b) crystal size distribution of dried crystals, and (c) final product yield, for 45 min of
6
7 cooling. For more details calculating tind, tG and Rc in (a) see text. The product yield expected from
8
the solubility data of Granberg&Rasmuson19 was 73.5% w/w, which is plotted in (c) as a guideline.
9
10
11
12
13 1.0
1.2
14
0.1% w/w HPMC
15
16 1.0 0% w/w HPMC 0.8
Frequency number distribution [-]

17
18 0.8
19
volume [%]
0.6 0.1% w/w HPMC

20
21 0.6
22
0.4
23 0.4
24 0% w/w HPMC
25
0.2 0.2
26
27
28 0.0
29 0.0
0.1 1 10 100 1000 0.1 1 10 100 1000
30 Equivalent diameter of crystal, CE [µm]
31 Equivalent diameter of crystal, CE [µm]
32 (a) (b)
33
34
35
36
37
38
39
40
41
42
43
44
45 200 µm
46 20 µm
47
48 (c) (d)
49
50
51 Figure 4. Crystal size distribution for Paracetamol crystallised in 0% w/w and 0.1% w/w of HPMC
52 based on (a) volume and (b) frequency. SEM images of Paracetamol samples crystallised in (c) 0%
53
54 w/w HPMC and (d) 0.1% w/w HPMC. Note the different scale bars used in (c) and (d).
55
56
57
58
59
60
1
2
3
0% w/w HPMC
4
5
6
7
8 Intensity [a.u.]
9
10
11
12 0.1% w/w HPMC
13
14
15
16
17
18
19
20
21
22
23 5 15 25 35 45
24
25 º2θ
26
27
(a)
Endothermic
28
29 0% w/w HPMC
30
31
32
33
34
35
36
37
38
39
0.1% w/w HPMC
40
41
42
43
44
45
46
47
48
49 50 80 110 140 170
50
51 Temperature [ºC]
52
53 (b)
54
55
56 Figure 5. (a) XRD patterns and (b) DSCs for Paracetamol crystallised in 0% w/w and 0.1% w/w of
57 HPMC.
58
59
60
1
2
3
(a) (b)
4
5
6
7
8
9
10
11
12
13
14
15
16
17 (c) (d)
18
19
20
21
22
23
24
25
26
27
28
29
30
31 Figure 6. SEM images of Paracetamol crystallised in the presence of (a) 0.1% w/w, (b) 0.3% w/w,
32
33 (c) 0.7% w/w and (d) 0.8% w/w of HPMC showing microporosity in crystal’s surface.
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2 For Table of Contents Only
3
4
6
7
Paracetamol
8
9 Nuno M Reisa,b,‡, Zizheng K Liua, Cassilda M Reisa, and Malcolm R Mackleya
10
a
12
13
Site, Pembroke Street, Cambridge CB2 3RA, UK
14 b
Department of Chemical Engineering, Loughborough University, Loughborough LE11 3TU, UK
15
16
17 0% w/w HPMC
1.2
18
1.0
19
0.8
20 0.1% w/w HPMC
volume [%]
0.6
21 200 µm
0.4
22
23 0.2
24 0.0
0.1 1 10 100 1000 20 µm
25 Equivalent diameter crystal [µm]
26
27 Micronized Paracetamol crystals were directly produced from solution crystallisation in water by
28
29 quickly dropping the temperature in the presence of 0.1-0.8% w/w of HPMC. The polymer reduced
30 the solubility of Paracetamol and delayed the unset of nucleation beyond the cooling time of the
31
32 vessel, consequently allowing to produce fine crystals with sharp CSDs.
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 ‡
60

Hydroxypropyl Methylcellulose As A Novel Tool For Isothermal Solution Crystallization of Micronized Paracetamol

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hydroxypropyl Methylcellulose As A Novel Tool For Isothermal Solution Crystallization of Micronized Paracetamol

Uploaded by

Copyright:

Available Formats

Subscriber access provided by FONDREN LIBRARY, RICE UNIVERSITY

Frequency number distribution [-]

0.6 0.1% w/w HPMC

25 Equivalent diameter crystal [µm]

You might also like