Polymorph stabilization in processed acetaminophen powders

Powder Technology 236 (2013) 52–62

Contents lists available at SciVerse ScienceDirect

Powder Technology journal homepage: www.elsevier.com/locate/powtec

Polymorph stabilization in processed acetaminophen powders Anna Łuczak a, Laila J. Jallo b, Rajesh N. Dave b, Zafar Iqbal a,⁎ a

Department of Chemistry and Environmental Science, New Jersey Institute of Technology, 138 Warren Street, Newark, NJ 07102‐1982, USA Department of Chemical, Biological and Pharmaceutical Engineering, and New Jersey Center for Engineered Particulates, New Jersey Institute of Technology, 138 Warren Street, Newark, NJ 07102‐1982, USA

b

a r t i c l e

i n f o

Available online 29 May 2012 Keywords: Polymorphs Raman spectroscopy Dry coating Stabilization Dissolution

a b s t r a c t A novel approach has been developed to stabilize the metastable form II of acetaminophen (APAP) prepared from the melt of micronized and nano-silica coated APAP I. It is suggested that stabilization of form II APAP occurs by pinning via van der Waals-type interactions on low energy defects created by micronization and impinging particles during dry coating in form I APAP, heating to the melt phase, and cooling. The defects are likely to persist into the near-molten isotropic phase due to a permanent memory effect similar to that previously observed in silica-liquid crystal composites and re-crystallize together with form II APAP. Raman spectroscopy, scanning electron microscopy (SEM), powder X-ray diffraction (PXRD), and conventional as well as intrinsic dissolution have been used to characterize the stabilized APAP II obtained. Raman measurements up to 18 months and PXRD data for micronized form II APAP show no indication of APAP I formation. Aged samples of dry-coated APAP II however display Raman features indicative of APAP I within 5 months. PXRD and SEM techniques also detected small fractions of APAP I in dry-coated APAP II. Both micronized APAP II and dry-coated APAP II give rise to a time-dependent background scattering superimposed on the Raman spectra which increases in integrated intensity with time, saturates at 180 and 90 days, respectively, and then decreases in intensity with further aging. The time-dependent scattering background occurs in the stabilized form II APAP samples only after cooling from the melt and is more intense in micronized APAP II compared to dry-coated APAP II. It is assigned to disorder associated with slowly diffusing defects created by micronization and dry coating with impinging nano-silica particles in form I APAP. Intrinsic dissolution profiles for the stabilized APAP II samples showed measureable increases, particularly for micronized APAP II. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Acetaminophen (APAP) is a widely used analgesic drug which provides an excellent example of an organic polymorphic system since it is known to exist in three forms with different crystalline structures in addition to its structurally disordered amorphous phase. Form I APAP is monoclinic, form II is orthorhombic, and form III is too unstable for its structure to be determined by X-ray methods. The commercial form I is the most stable at room temperature, but it has poor binding and densification properties [1]. Both form I and II can be grown out of the molten state of APAP crystals; however the unstable form III can only be grown from the molten state at specific conditions, for example, between two flat surfaces [2] or by use of a nanoporous host [3]. From the commercial point of view, it is desirable to stabilize the metastable form II because of its enhanced compression behavior and higher intrinsic dissolution rate [4,5] compared to form I. The compression properties of the form I and II polymorphs relate to different

⁎ Corresponding author. Tel.: + 1 973 596 8571. E-mail address: [email protected] (Z. Iqbal). 0032-5910/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.powtec.2012.05.046

molecular packing in the two forms. The molecules form puckered layers in the form I polymorph and flat sheets or layers in form II, as evident from Fig. 1(a) and (b), respectively. Van der Waals intermolecular bonding occurs between the sheets and stronger intermolecular hydrogen bonding exists among the molecules in the sheets. As shown in Fig. 2(b), two types of hydrogen bonding, NH···O and OH···O, which differ significantly in their corresponding bond lengths — 2.91 Å and 2.65 Å in form I and 2.97 Å and 2.72 Å in form II, are observed in APAP crystals [6]. The shorter (stronger) intra-layer hydrogen bonding in form I results in its higher melting point and greater thermodynamic stability relative to the form II [7,8]. Another interesting feature is the higher plastic deformability of form II relative to the form I, which is consistent with its flat layered architecture in contrast to the puckered architecture of the molecules in the form I. The relative orientation of layers of acetaminophen molecules in each structural type relates to compaction properties and behavior of each phase under pressure. The more planar orientation of the sheets in form II is responsible for its well-developed slip planes and higher plastic deformability [9]. Form I does not deform plastically since the relative orientation of the molecular sheets are in a puckered arrangement. As a result, form I crystallites break by brittle fracture and require

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

53

Fig. 1. Crystal unit cells of acetaminophen in: (a) form I, and (b) form II.

in the second method new interfaces are produced between nanosilica and APAP particles. Form II crystallites nucleated from the melt of these processed form I APAP powders showed enhanced structural stability. Since form II APAP has improved tabletting properties the newly developed stabilization process would have important pharmaceutical implications. Raman spectroscopy and analysis by scanning electron microscopy (SEM), powder X-ray diffraction (PXRD), and conventional and intrinsic dissolution were conducted to characterize the stabilized form II APAP. The PXRD experiments also confirmed Raman data showing that no structural change to the form II occurs when the form I is initially micronized or dry-coated with nanosilica. Raman spectroscopy moreover showed a time-dependent defect-induced scattering background in stabilized APAP II similar to the time-independent scattering observed in cryomilled griseofulvin [13]. The results are discussed here together with a qualitative explanation for the stabilization process.

binding agents during manufacture to ensure that the tablets remain intact after compression [9]. APAP can form unstable polymorphic crystal structures compared to the drug griseofulvin which has one stable form and no known polymorphs except for its amorphous form. It is of importance to understand how process design can potentially benefit from understanding polymorphic transitions of a crystal. A large body of work has been devoted to APAP's structural behavior during pharmaceutical processing [1,4,5,10,11] since its polymorphism can influence its mechanical properties and dissolution rate. In particular, the more elastic but metastable form II APAP can be compressed directly but it transforms to the stable form I, typically within days. Several techniques have been employed to stabilize form II APAP by compression using stabilization agents, such a gelatin, polyvinylpyrollidone (PVP) or starch [12]. For example, Di Martino et al. prepared the pure orthorhombic form II of APAP for direct compression with stability up to 11 months [11]. Here, two novel approaches to enhance the stability of form II APAP are discussed. In the first method, metastable form II APAP is produced from the melt using micronized form I APAP as the starting material. In the second method, as-received coarse form I APAP is first dry-coated with nano-silica and then form II APAP is generated from it via the molten phase. For the reference samples, form II APAP was also generated using as-received coarse form I APAP that was not dry-coated, (see schematic in Experimental section). In the first method increased surface area in the starting powder is obtained by micronization, and

2. Experimental Table 1 lists the volume average particle size D10, D50 and D90 distributions and the commercial sources of the starting raw materials used. The coarse APAP I, for example, has a broad size distribution with a weighted average of about 50% of particles smaller than 29.5 μm (cohesive) and 10% greater than 228.9 μm (non-cohesive). The micronized APAP I contains on a weighted average basis about 90% cohesive particles. In this work, only

(b)

(a)

O NH

CH3

N

O

O

HO

H

O H

O

H

Type B hydrogen bonding OH···O

N H O H H N

Type A hydrogen bonding NH···O

O Fig. 2. (a) Molecular structure of acetaminophen, and (b) type A and type B hydrogen bonding between acetaminophen molecules in form I and form II [6].

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

54

Density Distribution

0.8

2. Preparation of stabilized form II APAP: As-received micronized APAP I (MAPAP I) →Melt MAPAP II As-received Coarse APAP I (CAPAP I) →Dry−coating Dry-coated CAPAP I →Melt Dry-coated APAP II

0.6 0.4 0.2 0

- 0.2

Uncoated CAPAP Dry Coated CAPAP 1

10

100

1000

Particle Size (µm)

2.3. Conventional dissolution rate (CDR)

Fig. 3. Bimodal size distribution of the uncoated and dry-coated as-received coarse acetaminophen (CAPAP I).

the coarse APAP I was dry-coated with 1 wt.% nano-silica. The amount of silica deposited was sufficient to cover a large proportion of the surface to prevent particle–particle contact between the API particles as evident from Fig. 6. Raman spectroscopy and PXRD measurements performed after dry coating showed no resulting chemical and structural changes, respectively. 2.1. Dry coating method Detailed description of the dry coating method can be found elsewhere [14,15], and is briefly discussed here. The method employs the technique of Magnetically Assisted Impaction Coating (MAIC). A measured amount of magnetic, host and guest particles at a determined ratio is placed in a glass jar, which is placed in a collar coil. Voltage is applied, creating an oscillating magnetic field, which accelerates and spins the magnets along with the host and guest particles in the jar. This promotes the collisions among the particles and with the wall of the vessel. During this process, guest particles get coated onto the surface of the host particles mainly due to impaction of fluidized magnetic particles. This device may be used with only a small amount of material, typically 5 g. However, the device can be reconfigured to provide continuous, scaled up operation, which has been done by Aveka, Inc., who manufactures this continuous device and has coated various powders at industrial scale. 2.2. Preparation of form II APAP Form II APAP was generated from as-received coarse APAP I (CAPAP I), as-received micronized APAP I (MAPAP I), and nano-silica dry-coated coarse APAP I as presented schematically below. The resulting form II APAP samples are labeled as: CAPAP II, MAPAP II, and dry-coated APAP II, respectively. 1. Preparation of reference form II APAP: As-received Coarse APAP I (CAPAP I) →Melt CAPAP II

Table 1 Particle size and sources of the raw materials used. Material

Sources

Coarse acetaminophen (CAPAP Form I) Micronized acetaminophen (MAPAP Form I) Nano-silica (hydrophilic)

AnMar International Ltd, USA Mallinckrodt Inc., U.S.A Cabot Inc., U.S.A

The procedure involves placing a few milligrams of a sample on a glass slide and melting at 190 °C. The hot melt was rapidly cooled in a freezer resulting in a thin film of form II APAP on the glass slide. The formation of the orthorhombic form II was confirmed by Raman spectroscopy in the thin films or crushed powders (for dissolution studies). All form II APAP samples prepared as described above were stored under ambient temperature and humidity conditions during the stability studies.

Particle size D10

D50

D90

3.7 μm

29.5 μm

228.9 μm

2.6 μm

10.7 μm

37.3 μm

–

16 nm

–

Dissolution analysis was performed using a basket dissolution tester Distek 2100A USP apparatus 1 (North Brunswick, New Jersey). Samples were tested according to the USP29-NF24 monograph for acetaminophen capsules. The dissolution profile was obtained as per specified conditions in monograph using a dissolution tank with six dissolution vessels, 900 ml each, equipped with rotating baskets with samples instead of paddles. Approximately 50 mg of sample was enclosed in a gelatin capsule and placed in each of the dissolution baskets, immersed into the vessel with deionized (DI) water at 37 °C, and rotated at a constant speed of 50 rpm. The water bath temperature was constant during the experiment. Aliquots of 5 ml were removed at time intervals, t = 0, 5, 10, 15, 20, 25 and 30 min and analyzed using a Varian Cary 50 Bio UV–visible spectrophotometer (Palo Alto, California). A calibration curve was generated by a set of freshly prepared standard solutions of 0.005 mg/ml, 0.010 mg/ml, 0.025 mg/ml, 0.050 mg/ml, and 0.060 mg/ml concentrations of as-received, coarse form I APAP (CAPAP I) in water. All samples were run in triplicate to create a dissolution profile and a reliable data set. Dissolution analyses were performed for the following forms of acetaminophen indicated in schematic above: CAPAP I, CAPAP II, MAPAP II, and dry-coated APAP II. 2.4. Intrinsic dissolution rate (IDR) A Vankel, VK 7000 dissolution apparatus equipped with compacted and rotating disk was employed to perform intrinsic dissolution on four batches of acetaminophen: CAPAP I, CAPAP II, MAPAP II, and dry-coated APAP II. Sampling and experimental conditions were the same as described above for the conventional dissolution rate study. A typical IDR apparatus consists of a steel plunger and die, and a stainless steel sample holder for direct compaction. Approximately 40 mg powder sample of CAPAP I was directly compressed using a hydraulic Carver Press Model C, applying a pressure of 3 kN for 30 s to form a 0.6 mm diameter disc inside the sample holder. The same procedure was repeated for the remaining batches, where each batch analysis was performed in triplicate. The sample holder was attached to the rotating shaft and immersed into the dissolution bath. Prior to analysis, the air pockets formed at the bottom of the sample holder during immersion were carefully removed. Using a Cary 50 UV–visible spectrometer, and a set of standards of the same concentrations as above, the mass of dissolved acetaminophen was calculated after correcting for the change in volume and including the disk area. For both types of dissolution testing, thin films were gently crushed to generate powders which were then confirmed to be form II APAP by Raman spectroscopy. 2.5. Raman spectroscopy Two Raman instruments were used to record Raman spectra: (a) A confocal Horiba–Jobin Yvon LabRam micro-Raman spectrometer with a 20 mW He-Ne laser source emitting at a wavelength of 632.8 nm

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

focused to a spot size of 10 μm with a 10× lens; and (b) a Mesophotonics SE 1000 Raman spectrometer calibrated to 2 cm− 1 with a 250 mW nearinfrared laser operating at 785 nm with a 130 μm diameter spot size. Typical acquisition time for all spectra recorded was 10 s per scanned spectrum. Each spectrum acquired using the Horiba–Jobin Yvon LabRam micro-Raman spectrometer was averaged over 2 scans. Data acquired with the Mesophotonics SE 1000 Raman spectrometer represent an average of 5 scans, for a total of 50 s. All Raman data shown are as-recorded by the instruments used.

55

were all non-conducting, were mounted on aluminum stubs using double-sided carbon tape and sputter-coated with carbon to avoid charging. 3. Results and discussion The dry coating method breaks up agglomerates formed of the uncoated powders as well as it disperses the nano-silica on the surface of the individual particles as shown in Fig. 6(c). Previous work by Chen et al. [16–20], showed that the presence of these nanoparticles on the surface of the host particles changes the contact geometry between the particles by reducing the contact area, resulting in a decrease in the adhesion force and improvement in flow properties and dispersability of the particles. The particle size distributions of the uncoated and dry‐coated CAPAP I are shown in Fig. 3. The plot indicates a bimodal size distribution. The figure also shows the effect of the dry coating process on the particle size distribution. After dry coating CAPAP I there is a visible change of the mode in the left peak and a slight increase in the density distribution of the smaller particles after dry coating. This could be attributed to some attrition taking place during the

2.6. Powder X-ray diffraction X-ray diffraction patterns for the powders were obtained using a wide-angle Philips X'pert PW3040 X-Ray Diffractometer with CuKα radiation (λ = 1.5406 Å) operating at 40 kV and 30 mA, with 2θ ranging from 5° to 50°. 2.7. Scanning electron microscopy (SEM) SEM images were obtained with a VP-1530 Carl Zeiss LEO (Peabody, MA) field emission scanning electron microscope. The samples, which

(a) Scattering background due to defects/disorder Arbitrary Intensity (counts/s)

20000

15000

10000

5000

0 50

550

1050

1550

2050

2550

3050

(c)

9000

12000

6000

3000

0 50

100

150

200

250

Raman Shift (cm-1)

300

(d) Arbitrary Intensity (counts/s)

(b) Arbitrary Intensity (counts/s)

Arbitrary Intensity (counts/s)

Raman Shift (cm-1)

6000

0 1210

1230

1250

Raman Shift (cm-1)

1270

12000

6000

0 1540

1590

1640

1690

Raman Shift (cm-1)

Fig. 4. Raman spectra of acetaminophen in its different polymorph and amorphous forms prepared from as-received, stable form I, in the spectral regions: (a) 50 to 3200 cm− 1, (b) 50 to 300 cm− 1, (c) 1210 to 1270 cm− 1, and (d) 1540 to 1690 cm− 1. The amorphous, form I, form II and form III are colored green, blue, red and black, respectively. The approximate peak of the broad background scattering region in the form III spectrum is indicated by an arrow.

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

56

dry coating process. However, the distribution profile of the uncoated and dry coated are similar. It must also be noted that although the device used here can only coat small quantities of material, it can be reconfigured to provide continuous, industrially-relevant operation. 3.1. Raman spectra of acetaminophen polymorphs Raman scattering is one of the most widely used techniques employed to study polymorphism in solids because of its high sensitivity to crystalline structure. It is therefore used here as the primary tool to monitor changes in form II APAP prepared from micronized APAP I and dry coated CAPAP I as a function of elapsed time. Raman spectra shown in Fig. 4, together with the frequencies and their possible assignments in Table 2, for all four polymorphic types including the amorphous form of APAP have been measured down to 100 cm − 1

and are in essential agreement with the data previously reported by Szelagiewicz et al. [21] and Kauffman et al. [22]. Raman lines below 200 cm − 1, studied for the first time to the best of our knowledge, are assigned to inter-molecular or lattice modes which are particularly sensitive to the lattice structure and clearly show differences between the four APAP forms. In form I, the lowest frequency lines at 136 cm − 1 and 155 cm − 1 are assigned to largely translational motions of the acetaminophen molecule, similar to observations in griseofulvin [13]. Both of these lines shift to lower frequencies (127 cm − 1 and 145 cm − 1, respectively) in form II APAP. In the amorphous form, long-range crystalline order is lost and the molecules are randomly oriented. The low frequency Raman lines however persist in the amorphous phase due to remnant short-range order but are substantially broadened due to the effects of long-range disorder. In form III, the translational mode features are not evident in the spectrum and

(a) Scattering Background

Arbitrary Intensity (counts/s)

50000

Intensity (counts/s)

35000

0

1

90 180 300 550 Time (Days)

0 month 6 months 18 months

0 200

400

600

800

1000

1200

1400

1600

1800

2000

Raman Shift (cm-1)

(b)

45000

0 month 6 months

Arbitrary Intensity (counts/s)

18 months

0 1200

APAP I

1250

1300

1350

Raman Shift (cm-1) Fig. 5. Raman spectra of stabilized form II acetaminophen as a function of time for sample prepared by melting from micronized MAPAP I in the spectral region from (a) 200 to 2000 cm− 1 and (b) 1200 to 1350 cm− 1; and dry-coated CAPAP I in the spectral region from (c) 200 to 2000 cm− 1 and (d) 1200 to 1350 cm− 1. Insets in Figs. 5(a) and 5(c) show the integrated intensity of the background scattering as a function of time in days. The full line curve in Fig. 5(a) is an approximate guide to the eye of the background scattering.

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

(c)

57

Intensity (counts/s)

10000

20000

Arbitrary Intensity (counts/s)

0

1 90 150 Time (Days)

0 month 3 months 5 months

0 200

400

600

800

1000

1200

1400

1600

1800

2000

Raman Shift (cm-1)

(d)

20000

0 month

Arbitrary Intensity (counts/s)

3 months 5 months APAP I

0 1200

1250

1300

1350

Raman Shift (cm-1) Fig. 5 (continued).

appear to be embedded in a broad scattering background similar to that seen in cryomilled griseofulvin [13]. Raman lines attributed to intra-molecular vibrational modes are observed at frequencies above 200 cm− 1 for the different polymorphs and the amorphous form as shown in Fig. 4(a and b). Changes in these lines occur in the different phases due to changes in local potentials and are more subtle than those for the Raman features assigned to intermolecular or lattice vibrations discussed above. More detailed views of the Raman features in the spectral regions between 1200 cm− 1 to 1280 cm− 1 and between 1500 cm− 1 to 1650 cm− 1 are indicated in Fig. 4(c) and (d), respectively. The Raman lines in the 1200 cm− 1 to 1260 cm− 1 frequency region correspond to C\N, C\H, and phenyl ring vibrations [23]. The characteristic C\N stretching Raman line at 1237 cm− 1 for form I APAP develops into two well resolved peaks at 1221 cm− 1 and 1245 cm− 1 in form II APAP due to change in local symmetry. Furthermore, there is substantial disorder-induced broadening of the 1237 cm− 1 line in the amorphous phase with almost no frequency shift relative to that of the line in the form I. In the form III this line

also broadens but increases in frequency by 6 cm− 1 to 1243 cm− 1 indicating an increase in force constant associated with the C\N bond. Additionally, other changes in the intra-molecular vibrational lines are observed in the Raman spectra as a function of crystalline polymorph or amorphous form in the 1500 cm − 1 to 1650 cm− 1 frequency region as evident from Fig. 4(d). Here changes occur in the relative intensities and peak positions of Raman lines that are attributed to –N\H deformation and C_O stretching modes. The strongest lines for form I APAP are seen at 1611 cm − 1 and 1649 cm− 1 and are assigned to –N\H deformation and C_O stretching modes, respectively. In the amorphous form both of these lines shift up in frequency by 9 cm− 1 together with significant broadening as well as an increase and decrease in relative peak intensity, respectively. These changes are most likely due to disorder in the amorphous form and type A and B hydrogen bonding explained in the Introduction. The upshift in the peak positions indicate that the force constants for N\H deformation and C_O stretching increase probably because inter-molecular interactions are disrupted in the amorphous form. In form III APAP intermolecular

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

58

Table 2 Raman frequencies and qualitative mode assignments for acetaminophen in its three polymorph and amorphous forms. Form I

Form II

Form III

Amorphous

(cm− 1)

(cm− 1)

(cm− 1)

(cm− 1)

135 155 – 205 218 326 388 462 501 649 796 839 856 1169 – 1237 – 1278

127 145 172 208 – 330 388 452 5050 649 796 – 856 1169 1221 – 1245 1278

– 146 183 210 – 330 388 457 505 646 799 – 862 1172 – – 1242 1278

– 143 – – 215 330 388 458 508 650 800 831 864 1174 – 1238 – 1280

1324 1371 1559 – 1611 – 1649 2937 3016 – 3058 – 3072 3083

1326 1376 – 1578 1609 1623 1649 2940 3018 – 3069 – – 3083

1321 1374 1562 – 1614 – 1647 2928 3009 3045 – 3067 – 3095

1328 1378 1562 – 1620 – 1658 2938 3018 – – – 3072 –

– 3110

– 3111

3095 –

– 3107

3163 3326

3166 3328

3159 3329

3155 –

3.2. Monitoring the stability of APAP II by Raman spectroscopy

Qualitative assignments

Translational lattice modes

Out-of-plane ring deformation Out-of-plane ring deformation Out-of-plane ring deformation In-plane ring deformation Phenyl ring stretch [22] C\H out-of-plane bend Phenyl ring C\H stretching C\H in-plane bend C\C stretch carbon phenyl ring C\N stretch O\H and C\O combination, (ar) C\N stretch C\H bend –CH3 umbrella mode –H\N\C_O stretch [22] –N\H deformation [22] C_O stretch [22] –CH3 symmetric stretch –CH3 symmetric stretch –CH3 symmetric stretch –CH3 symmetric stretch –CH3 asymmetric stretch –CH3 asymmetric stretch –CH asymmetric stretch, phenyl ring –CH asymmetric stretch, phenyl ring O\H stretch N\H stretch

interactions persist, and the line at 1611 cm− 1 upshifts by 3 cm− 1 whereas the line at 1649 cm− 1 downshifts by 2 cm− 1 due to changes in the local inter-molecular potentials. In form II, the form I APAP line

(a)

(b)

10µm

at 1611 cm− 1 splits into two peaks at 1578 cm− 1 and 1623 cm− 1 indicating substantial changes in the hydrogen bonding interactions.

The Raman spectra in the 200 to 2000 cm − 1 range over a period of several months for form II APAP prepared from micronized and drycoated form I APAP displayed in Fig. 5(a and c), respectively, show the Raman spectral features of form II APAP discussed above and a time-dependent background scattering superimposed on the Raman spectra. The expanded view of the Raman spectra as a function of time in the 1200 to 1350 cm − 1 range for APAP II from micronized APAP I together with the spectrum from APAP I (in blue) shown in Fig. 5(b) indicate that the lines of APAP I in this spectral range near 1230 and 1260 cm − 1 are not observed even after 18 months, although small frequency downshifts are observed at 18 months. This confirms the long-term stability of APAP II prepared via micronization. However, the corresponding spectra for APAP II from nano-silica coated APAP I in Fig. 5(d) show larger downshifts in frequencies at shorter times after synthesis as well as the appearance of weak spectral features due to the formation of APAP I consistent with PXRD and SEM results discussed below. The broad time-dependent scattering background superimposed on the Raman spectra appears in the Raman spectra of stabilized MAPAP II and dry-coated APAP II samples. The scattering is similar to that seen previously in cryomilled griseofulvin and related materials [13,24]. The plots of the integrated intensities of this scattering as a function of time for MAPAP II and dry-coated APAP II are shown as insets in Fig. 5(a) and (d), respectively. The integrated intensities are much higher in MAPAP II compared to dry-coated APAP II. In both MAPAP II and dry-coated APAP II the intensities increase with time and then decrease after saturation at 180 days and 90 days, respectively. By analogy with Raman observations in cryomilled griseofulvin [13], the background scattering observed can be assigned to defect-induced disorder induced by the two processes used to prepare the form II APAP. The first process involves micronization of APAP I followed by melting and crystallization on cooling and the second process involves dry coating by impinging nano-silica particles on APAP I followed by melting and crystallization. The defect density is expected to be higher in MAPAP II compared to dry-coated APAP II in agreement with the higher background scattering intensity for MAPAP II (Fig. 5). The scattering intensity for MAPAP II and dry-coated APAP II increases with time followed by saturation and decrease in intensity after 180 and 90 days, respectively, the

(c)

10µm

Fig. 6. SEM images of (a) As-received CAPAP I, (b) As-received micronized APAP I, and (c) Nano-silica dry-coated CAPAP I showing uniform and substantial surface coverage of a crystal.

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

59

Fig. 7. SEM images of APAP II thin film from as-received CAPAP I with increasing magnification from left to right. Pores discussed in the text are indicated by arrows.

time-dependence of the scattering is probably due to slow diffusion of the defects under ambient conditions during aging. Although a quantitative model needs to be developed, it can be conjectured that long-term stabilization of MAPAP II and shorter-term stabilization of dry-coated APAP II occurs by pinning via van der Waals-type interactions on these low energy defect structures. Higher stabilization is achieved in MAPAP II compared to dry-coated APAP II because of higher defect densities in the former. This is also in agreement with the Raman spectra which show very little indication of form I in aged MAPAP II but weak features of form I appear in drycoated APAP II aged for 5 months. Also, since the precursor defects are created in form I, they have to persist into the near-molten isotropic phase during preparation of form II APAP from the melt. Persistence of the defects initially created in form I APAP into the melt phase is likely to occur due to a permanent memory effect similar to that observed in silica-liquid crystal composites [24] to provide sites for “pinning” the APAP II structure crystallizing out of the melt. The observed background scattering however only appears in MAPAP II and dry‐coated APAP II after crystallization from the melt. The reason for this is not entirely clear and needs further investigation.

3.3. Characterization of stabilized APAP II polymorph by SEM imaging, and PXRD and dissolution measurements 3.3.1. Scanning electron microscopy (SEM) SEM studies were carried out to image the physical features of the samples since various sample types with different morphologies were used, such as as-received powders, thin films, and powders from crushed films. The orthorhombic and monoclinic form II and form I polymorphs, respectively, can be easily recognized due to differences in their crystal habits, especially at higher magnifications. Fig. 6(a), (b), and (c) show representative images of as-received coarse CAPAP I, as-received MAPAP I, and dry-coated CAPAP I, respectively. Figs. 7, 8, and 9 show SEM images of thin films of APAP II prepared from as-received CAPAP I, MAPAP II, and dry-coated APAP, respectively. Interestingly, APAP II from as-received CAPAP I in Fig. 7 showed a large number of pores formed during film preparation. Higher surface area around the pores would accelerate the transition back to APAP I. Pores, however, were not observed in MAPAP II and from dry-coated APAP II. Well-established crystal habits of both form I APAP (monoclinic) and form II APAP (orthorhombic) were observed on the surface of dry-coated APAP II films shown in

Fig. 8. SEM images at different magnifications for APAP II from micronized APAP I fabricated as a thin film.

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

60

Type II

Type I

Fig. 9. SEM images of dry-coated CAPAP II fabricated as a thin film. Presence of monoclinic (form I) and orthorhombic (form II) crystals are evident in the sample as indicated by arrows.

3.3.3. Dissolution profiles By definition, “dissolution is defined as the process by which solid substances enters a solvent to yield a solution” [25]. Conventional dissolution rate (CDR) analysis was carried out to test the behavior of the stabilized form II APAP samples. Equal amounts of all samples used for CDR were enclosed in gelatin capsules. However, since CAPAP I was used as powder and CAPAP II, MAPAP II, and dry-coated APAP II samples were obtained by crushing the thin films, the gelatin capsules consisted of particles with varying surface areas. Although the polymorph form is one of the solid state characteristics affecting dissolution, in this case the CDR profile was strongly related to the disintegration of the solid particle of various sizes and shapes within the capsules rather than the physical properties of the drug, such as its polymorph form. Fig. 11(a) compares the dissolution profiles of CAPAP I and CAPAP II with those of MAPAP II and drycoated APAP II. It can be seen that almost complete mass release is seen in CAPAP I within the first 10 min of analysis; this was followed by CAPAP II. The two modified acetaminophen form II samples showed no improvement in dissolution rate, in fact it took almost three times as long, compared to APAP form I, for these samples to dissolve completely. Since modified APAP form II samples are generated from micronized and dry-coated CAPAP I, it is important to understand the influence of these solid state characteristics of the formulation on dissolution. The intrinsic dissolution method was used to deduce the relative solubilities of the metastable forms of APAP. During intrinsic dissolution, a constant surface area is exposed to the dissolution medium. IDR is described by the Eq. (1) referred to as the Noyes–Whitney equation [26]:

J¼

dm 1 · ¼ kC S dt A

(a) 8000

MAPAP II

MAPAP I

0 10

dm 1 · ¼ kðC s −C Þ dt A

ð1Þ

where dm/dt is the rate of increase of the mass of solute released per unit time t, k is the mass transfer coefficient, A is the surface area of

15

20

25

30

Diffraction Angle 2

35

40

-[°]

(b) 8000

Dry Coated APAP II

Dry Coated APAP I

0 10

J¼

ð2Þ

Intensity [arb units]

3.3.2. Powder X-ray diffraction (PXRD) Wide-angle PXRD diffraction data obtained using CuKα radiation are shown in Fig. 10(a) for MAPAP I and MAPAP II, and in Fig. 10(b) for dry-coated APAP I and APAP II, respectively. PXRD data for MAPAP I and II, and dry-coated APAP I agree with that of Welton et al. [23]. However the PXRD pattern of dry-coated APAP II shows the presence of small quantities of form I consistent with the Raman results.

the sample, Cs is the intrinsic solubility at the sample surface, and C is the concentration of solute dissolved in time t. Eq. (1) can be further reduced to Eq. (2), since for sink conditions Cs >>C,

Intensity [arb units]

Fig. 9 (indicated by arrows), consistent with the Raman spectra and XRPD results below.

15

20

25

30

Diffraction Angle 2

35

40

-[°]

Fig. 10. Wide-angle powder X-ray diffraction patterns of acetaminophen samples obtained using CuKα radiation, λ = 1.5406 Å, (panel a) micronized APAP I (MAPAP I) and MAPAP II from micronized APAP I, and (panel b) dry-coated coarse CAPAP I and APAP II from dry-coated CAPAP I.

A. Łuczak et al. / Powder Technology 236 (2013) 52–62

Total Mass Released (µm)

(a) 40000 CAPAP I CAPAP II

20000

MAPAP II Dry-coated APAP II

0

0

10

20 Time (min)

30

Total Mass Released (µg)

(b) 12000

61

and PXRD data for MAPAP II show no indication of APAP I formation. Aged samples of dry-coated APAP II however display Raman features indicative of APAP I within 5 months. PXRD and SEM techniques also detected small fractions of APAP I in dry-coated APAP II. Both MAPAP II and dry-coated APAP II give rise to a time-dependent background scattering superimposed on the Raman spectra which increases in integrated intensity with time, saturates at 180 and 90 days, respectively, and then decreases in intensity with further aging. The time-dependent scattering background which only appears in form II APAP samples after preparation from the melt is more intense in MAPAP II compared to dry-coated APAP II. It is assigned to disorder associated with slowly diffusing defects created by micronization and impinging particles during dry coating in form I APAP. Stabilization of MAPAP II and dry-coated APAP II occurs by pinning via van der Waals-type interactions on these low energy defects. Longer term stabilization is achieved in MAPAP II compared to dry-coated APAP II because of higher defect densities in the former. It is suggested that the defects formed in form I APAP persist into the nearmolten isotropic phase on heating due to a permanent memory effect similar to that previously observed in silica-liquid crystal composites [24] and then re-crystallize together with form II APAP on cooling. Measureable increases were also observed in the intrinsic dissolution profiles for the stabilized APAP II samples and particularly for MAPAP II.

8000

Acknowledgments 4000

0 0

10

20

30

Time (min) Fig. 11. (a) Conventional dissolution rate profiles for as-received CAPAP I, APAP II from CAPAP I, MAPAP II from micronized MAPAP I and dry-coated APAP II from CAPAP I, and (b) intrinsic dissolution rate profiles for the same samples as in (a).

This work was funded by the National Science Foundation Engineering Research Center for Structured Organic Particulate Systems (Grant no. EEC-0540855). Special thanks to Aveka, Inc., Woodbury, MN, for providing the use of the MAIC, Dr. Costas Gogos of the New Jersey Institute of Technology for providing the use of the conventional dissolution tester, and Dr. Shirlynn Chen of Boehringer-Ingelheim, Ridgefield, CT, for the use of the instrumentation for intrinsic dissolution testing. References

Eq. (2) indicates that the rate of dissolution J, is a function of the solid surface area. There are many known surface modification techniques, such as particle size (micronization), crystalline form of a solid (polymorphs), or coating that can act as a barrier to the dissolution rate. Since both micronization and surface coating were employed here to generate form II APAP samples, intrinsic dissolution with a compressed disk of known area was employed to eliminate surface area and surface electrical charges as dissolution variables. The dissolution rate so determined can be more closely correlated with the physical properties of the sample [26]. By holding the extrinsic factors such as agitation, pH, temperature, and surface area of the sample constant, the IDR of the pure solid was measured. From Fig. 11(b) it is evident that both MAPAP II and drycoated APAP II show measureable increases in dissolution relative to CAPAP I. Dry-coated form II APAP also showed improvement in mass release, but the improvement relative to CAPAP I is small. Different dissolution profiles indicate that the form I and II polymorphs dissolved at different rates. The calculated intrinsic dissolution rates for MAPAP II, dry-coated APAP II, and CAPAP II samples were 1.50, 1.33, and 1.31 mg/min/cm 2 respectively with %RSD value below 6% for all dissolution time points. Lower rate was obtained for CAPAP I sample equal to 1.14 mg/min/cm 2, with %RSD below 6%. As expected, all APAP form II samples resulted in higher intrinsic dissolution rates than the form I APAP sample. 4. Conclusions MAPAP II prepared from micronized APAP I and dry-coated APAP II prepared from coarse APAP I show long- and shorter-term stabilization of the form II APAP polymorph. Raman measurements up to 18 months

[1] M. Sacchetti, Thermodynamic analysis of DSC data for acetaminophen polymorphs, Journal of Thermal Analysis and Calorimetry 63 (2001) 345–350. [2] P.D. Martino, P. Conflant, M. Drache, J.-P. Huvenne, A.-M. Guyot-Hermann, Preparation and physical characterization of forms II and III of paracetamol, Journal of Thermal Analysis 48 (1997) 447–458. [3] M. Beiner, G.T. Rengarajan, S. Pankaj, D. Enke, M. Steinhart, Manipulating the crystalline state of pharmaceuticals by nanoconfinement, Nano Letters 7 (2007) 1381–1385. [4] G.L. Perlovich, T.V. Volkova, A. Bauer-Brandl, Polymorphism of paracetamol: relative stability of the monoclinic and orhtorhombic phase revisited by sublimation and solution calotimetry, Journal of Thermal Analysis and Calorimetry 89 (2007) 767–774. [5] T. Beyer, G.M. Day, S.L. Price, The prediction, morphology, and mechanical properties of the polymorphs of paracetamol, Journal of the American Chemical Society 123 (21) (2001) 5086–5094. [6] S.L. Wang, S.Y. Lin, Y.S. Wei, Transformation of metastable forms of acetaminophen studied by thermal Fourier Transform infrared (FT-IR) microspectroscopy, Chemical and Pharmaceutical Bulletin 50 (2002) 153–156. [7] M. Haisa, S. Kashin, H. Maeda, The orthorhombic form of p-hydroxyacetanilide, Acta Crystallographica Section B 30 (1974) 2510–2512. [8] M. Haisa, S. Kashino, T. Yuasa, K. Akigawa, Topochemical studies. IX. The crystal and molecular structure of p-aminoacetophenone, Acta Crystallographica Section B 32 (1976) 1326–1328. [9] G. Nichols, C.S. Frampton, Physicochemical characterization of the orthorhombic polymorph of paracetamol crystallized from solution, Journal of Pharmaceutical Sciences 87 (6) (1998) 684–693. [10] J.S. Capes, R.E. Cameron, Effect of polymer addition on the contact line crystallisation of paracetamol, Crystal Engineering Communications 9 (2007) 84–90. [11] P.D. Martino, A.M. Guyot-Hermann, P. Conflant, M. Drache, J.C. Guyot, A new pure paracetamol for direct compression: the orthorhombic form, International Journal of Pharmaceutics 128 (1996) 1–8. [12] J.-M. Fachaux, A.-M. Guyot-Hermann, J.-C. Guyot, P. Conflant, M. Drache, S. Veesler, et al., Pure paracetamol for direct compression Part I. Development of sintered-like crystals of paracetamol, Powder Technology 82 (2) (1995) 123–128. [13] A. Żarów, B. Zhou, X. Wang, R. Pinal, Z. Iqbal, Spectroscopic and X-ray diffraction study of structural disorder in cryomilled and amorphous Griseofulvin, Applied Spectroscopy 65 (2011) 135–143. [14] M. Ramlakhan, C.Y. Wu, S. Watano, R.N. Dave, R. Pfeffer, Dry particle coating using magnetically assisted impaction coating: modification of surface properties and

62

[15] [16]

[17] [18]

[19]

[20]

A. Łuczak et al. / Powder Technology 236 (2013) 52–62 optimization of system and operating parameter, Powder Technology 112 (2000) 137–148. J. Yang, A. Sliva, A. Banerjee, R. Dave, R. Pfeffer, Dry particle coating for improving the flowability of cohesive powders, Powder Technology 158 (2005) 21–33. Y. Chen, L. Jallo, M.A.S. Quintanilla, R. Dave, Characterization of particle and bulk level cohesion reduction of surface modified fine aluminium powders, Colloids and Surfaces A: Physicochemical and Engineering Aspect 361 (2010) 66–80. Y. Chen, J. Yang, R.N. Dave, R. Pfeffer, Fluidization of coated group C powders, American Institute of Chemical Engineers Journal 54 (2008) 104–121. L. Jallo, Y. Chen, J. Bowen, F. Etzler, R.N. Dave, Prediction of inter-particle adhesion force from surface energy and surface roughness, Journal of Adhesion Science and Technology 18 (2011) 367–384. L. Jallo, C. Ghoroi, L. Gurumurthy, U. Patel, R.N. Davé, Improvement of flow and bulk density of pharmaceutical powders using surface modification, International Journal of Pharmaceutics 423 (2011) 213–225. L. Jallo, M. Schoenitz, E.L. Dreizin, R.N. Dave, C.E. Johnson, The effect of surface modification of aluminum powder on its flowability, combustion and reactivity, Powder Technology 204 (2010) 63–70.

[21] M. Szelagiewicz, C. Marcolli, S. Cianferani, A.P. Hard, A. Vit, A. Burkhard, et al., In situ characterization of polymorphic forms: the potential of Raman spectroscopy, Journal of Thermal Analysis and Calorimetry 57 (1999) 23–43. [22] J.F. Kauffman, L.M. Batykefer, D.D. Tuschel, Raman detected differential scanning calorimetry of polymorphic transformations in acetaminophen, Journal of Pharmaceutical and Biomedical Analysis 48 (2008) 1310–1315. [23] G. Socrates, Infrared and Raman Characteristic Group Frequencies, 3rd ed. John Wiley & Sons, Ltd., England, 2000. [24] S. Relaix, R.L. Leheny, L. Reven, M. Sutton, Memory effect in composites of liquid crystal and silica aerosil, Physical Review E 84 (2011) 061705. [25] V. Kumar, D. Tewari, Dissolution, in: D.B. Troy (Ed.), Remington: The Science and Practice of Pharmacy, 21 ed., Lippincott Williams & Wilkins, Philadelphia, 2005, p. 672. [26] H.-K. Chan, D.J. Grant, Influence of compaction on the intrinsic dissolution rate of modified acetaminophen and adipic acid crystals, International Journal of Pharmaceutics 57 (1989) 117–124.