1. Introduction Ibuprofen sodium dihydrate (ibuprofen sodium) is a fast-acting salt of ibuprofen free acid and one of the most widely used nonsteroidal anti-inflammatory drugs today (Rainsford et al., 2009; Sörgel et al., 2005). This sodium salt modification of ibuprofen was reported to be absorbed into plasma more rapidly than conventional ibuprofen and to have comparable tolerability and safety profile (Sörgel et al., 2005). A multiparticulate drug delivery system, such as ``direct to mouth'' granules of ibuprofen sodium, would facilitate the swallowability for geriatric or pediatric population patients suffering from dysphagia (Gandhi et al., 2013; Patel & Dhake, 2011; Patwekar & Baramade; 2012, Sharma & Chaurasia, 2013; Stegeman et al., 2011; Wahlich et al., 2013). A ``direct to mouth'' system containing ibuprofen sodium must be taste-masked due to the two major taste components of ibuprofen: bitterness and burning sensation (Higton 1999). Recently, our group presented a novel and cost-efficient approach to manufacturing a taste-masked multi3

particulate system with a stable immediate release profile by applying lipid-based excipients in a solvent-free hot-melt coating process (Becker et al., 2013; Becker et al., 2016). This technique uses fluid bed technology to apply molten coating material into fluidized particles (Jozwiakowski et al., 1990, Jannin & Cuppok 2013, Lopes et al., 2015). The re-solidification of molten material on the surface of particles creates the coating layer. However, hot-melt coating cannot be applied to commercially available ibuprofen sodium that has small particle size (<100 µm), low bulk density (< 0.5 g/ml) and dusty properties (Jannin & Cuppok, 2013; Jozwiakowski et al., 1990; Werner et al., 2007). Since these characteristics are common to the majority of APIs, a granulation step is required before hot-melt coating. In the pharmaceutical industry, using dry granulation via roller compaction and/or slugging dates back to the late 1940s (Miller 2005). However, in the 21st century roller compaction has attracted renewed scientific attention, especially due to the transition of equipment from manual to automatic that resulted in a higher level of process control (Guigon et al., 2007; Khorasani et al., 2016; Miller 2005; Osborne et al., 2013). Similarly to hot-melt coating, roller compaction is a solvent-free technology resulting in a fast and cost-effective manufacturing process, since timeconsuming evaporation steps or costly solvent recovery and disposal are not required (Guigon et al., 2007; Jannin & Cuppok, 2013; Jozwiakowski et al., 1990; Miller 2005). Furthermore, during roller compaction, the density of ribbon and granules can be increased by applying higher pressures (Hancock et al., 2003; Teng et al., 2009). These densified granules are suitable for hot-melt coating purposes. Additionally, it is critical that the resulting granules are resistant to the mechanical stress involved in the fluid bed hot-melt coating process since air velocity may result in pulverization of fragile granules (Jozwiakowski et al., 1990). During a roller compaction process, powder friction at the surface of the rollers first increases the pressure exerted to the powder. Next, this compression pressure alone is responsible for the bond4

ing of particles in different stages: (1) particle rearrangement; (2) particle deformation; (3) particle fragmentation; and (4) particle bonding (Miller 2005; Mollet & Grubenmann, 2001; Yu et al., 2012). Applying roller compaction technology to producing granules with a high content of ibuprofen sodium is known to be challenging (Gruber & Reher, 2004). The reasons are the extremely poor compactibility of ibuprofen sodium powder and the low frictional properties (internal and wall friction) that are not suitable for granule formation via roller compaction (Higton 1999; Gruber & Reher 2004). The aim of this work is to produce granules with high ibuprofen sodium content, dense and abrasion-resistant via roller compaction and investigate their resulting solid state. A second goal is to evaluate the effect of hot-melt coating parameters on taste-masking and in vitro immediate release profile of obtained ibuprofen sodium granules using a statistical design of experiments (DoE).

2. Material and methods

2.1. Material Ibuprofen sodium dihydrate was purchased from BASF (Ludwigshafen, Germany). The roller compaction excipients sorbitol (Parteck®SI 150) and mannitol (Parteck® M200) were purchased from Merck (Darmstadt, Germany). Isomalt (galenIQTM 721) was purchased from Beneo (Mannheim, Germany). Polyvinylpyrrolidone (PVP) (Kollidon® 25) was obtained from BASF (Ludwigshafen, Germany), and microcrystalline cellulose (Avicel® PH 105) was purchased from FMC (Philadelphia, PA USA). The coating material tripalmitin (Dynasan®116) was generously provided by IOI Oleo GmbH (Witten, Germany) and polysorbate 65 (Tween®65) was obtained from Croda GmbH (Nettetal Kaldenkirchen, Germany). All other chemicals were of analytical grade and purchased from Sigma-Aldrich (Steinheim, Germany).

2.2. Methods

2.2.1. Manufacturing of ibuprofen sodium granules 5

Table 1 lists the formulations selected for roller compaction excipient screening, throughput evaluation and process robustness. Raw material was de-agglomerated by sieving before further processing (sieve screen 1 mm). Blends for excipient screening trials (600 g) were prepared using a Stephan mixer (Stephan machinery GmbH, Hameln, Germany). Blends for throughput evaluation (3 kg) were manually mixed in a PET bag, while blends for process robustness (2x25 kg) were prepared in a Rego PM 100 (Rego-Herlitzius, Haan, Germany). The blending time was 8 minutes and the blending speed was 100 rpm. Roller compaction trials were carried out in an Alexanderwerk WP120 Pharma (Alexanderwerk, Remscheid, Germany) equipped with 40 mm (width) corrugated rollers. Selected process parameters for excipient screening trials were as follows: screw speed = 81 rpm, roller speed = 3.4 rpm and hydraulic pressure= 60 bar. The screw speed and hydraulic pressure were kept constant for the evaluation of the process throughput, while the roller speed increased from 3.4 to 14.8 rpm. The process robustness trials were undertaken using a roller speed of 9 rpm, keeping screw speed and hydraulic pressure constant at 3.4 rpm and 60 bar, respectively. The rollers were cooled to 10°C for all trials. Sieves of 1.6 and 0.8 mm were used for the pre-granulation and granulation steps, respectively.

2.2.2. Characterization of ibuprofen sodium granules

2.2.2.1. Friability of granules Friability can be defined as a reduction in the mass of granules occurring when they are subjected to mechanical strain during handling (tumbling, vibration, fluidization, etc.) (Ph. Eur. 8.0, 2.9.41). The selected friability test aims to characterize the propensity of granules to be pulverized when subjected to high shear, in general. In case of the excipient screening and evaluation of the process throughput trials, samples of 100 g granules were taken after finishing each roller compaction process. In case of process robustness trials, samples of 100 g granules were taken during the process. Each sample was sieved for 5 min in an AS 200 sieve tower, using a sieve screen of 250 µm (Rescht, Haan, Germany). The fraction below 250 µm was discarded. The friability of granules above 250 µm was characterized by subjecting 10 g of granules together with 200 glass-beads to falling shocks for 10 min at 25 rpm in a friabilator PTF E (Pharmatest, Hainburg, Germany) (Inghelbrecht & Remon, 1998). After the test, the granules were sieved with a 250 µm sieve for 2 min and weighted to find the mass loss during the friability test. All tests were conducted three times. The granules were discarded after the friability test.

2.2.2.2. Microstructure of granules

6

The structure of obtained granules was characterized using scanning electron microscopy (SEM). The SEM samples were sputtered with Pt/Pd prior to imaging. The micrographs were obtained at 5 kV on a scanning electron microscope (FESEM ZEISS Ultra 55 microscope, FELMI, Graz, Austria).

2.2.2.3. Thermal analysis The thermal behavior of the raw material, the physical mixtures of the API and sorbitol and isomalt, and the obtained granules were analyzed using differential scanning calorimetry (DSC). DSC was carried out using a DSC 204 F1 Phoenix (Netzsch, Selb, Germany) at a heating rate of 5 K/min with samples of approximately 3 mg.

2.2.2.4. Solid state The solid state of raw material, the physical mixtures of API and sorbitol and isomalt and the obtained granules were analyzed using both small- and wide-angle X-ray scattering (SWAXS). The experiments were performed with a point-focusing SWAXS camera system (Hecus-S3MICROSWAXS, Bruker AXS, Karlsruhe, Germany) powered by a Cu Kα radiation microsource with a wavelength of 1.542 Å. The system was equipped with a thermally controlled sample spinning and two linear position sensitive detectors to cover the real space resolution ranges of 10–1500 Å (SAXS) and 3.3–4.9 Å (WAXS). The samples were placed into a glass capillary with a diameter of ≈2 mm, which was later sealed with wax and placed into the capillary rotation unit. The measurements were performed at room temperature with an exposure time of 1300 sec.

2.2.3. Manufacturing of coated granules (fluid bed hot-melt spray coating) In this study, the fluid bed hot-melt coating was performed using an Innojet Ventilus® V-2.5/1 laboratory system fluid bed equipped with an IHD-1 hot-melt device (Innojet Herbert Hüttlin, Steinen, Germany) with a 1 L glass container. A batch size of 200 g of granules was used for all experiments. A formulation comprising tripalmitin (Dynasan® 116) and polysorbate 65 (Tween® 7

65) was applied for the hot-melt coating of granules. The temperatures of the melt of coating material and the spray air in the nozzle were set to 90°C. For each run, a batch of granules was fed into the glass container and fluidized using defined air flow. The melt of coating material was sprayed on the surface of fluidized granules by using defined spray rates and atomization air pressures. The coating was achieved after solidification of the melt on the surface of granules. MODDE 10 (Umetrics, Sweden) was used for creating a statistical model for design of experiments. An extended Rechtschaffner quadratic model with 37 runs, including three center points, was chosen to evaluate the main and two-factor interactions of the input parameters: the spray rate, the spray pressure, the coating ratio, the emulsifier ratio, the flow rate and the temperature. The selected ranges of parameter-settings are listed in Table 2. The output parameters (responses) were the time required (min) for release of 85% of ibuprofen sodium from coating (85% API release within 30 minutes was taken as a specification for an in vitro immediate release profile) and the drug amount released (%) after 5 minutes (less than 10% release within 5 minutes is an indicator for efficient taste masking) (Siewert et al., 2003).

2.2.4. Characterization of coated granules

2.2.4.1. Particle size and size distribution The particle size of the coated granules was measured using a high speed analysis sensor (QICPIC, Sympatec, Clausthal-Zellerfeld, Germany) with a dry disperser RODOS/L. The feeding rate was 30%, 400 frames/s were taken, the injector diameter was 4 mm and the air pressure was 1 bar.

2.2.4.2. In vitro release profile of API from coating The dissolution tests to evaluate the in vitro release profile of API from coating were performed in a USP apparatus II at 100 rpm coupled with an auto-sampler (Erweka DT820 LH, Heusenstamm, Germany). The dissolution medium was phosphate buffer (pH 6.8). 900 ml of medium 8

were pre-warmed to 37 ± 0.5°C. Samples of coated granules containing approximately 512.5 mg of ibuprofen sodium (400 mg of ibuprofen free acid) were tested. The samples of 1.5 mL of drug dissolved in the medium were automatically withdrawn at 5, 15, 30, 45, 60 and 90 minutes and assayed by HPLC. The withdrawn volumes were not replaced with an equivalent blank media. The dissolution tests were performed three times for each sample. The API content in the granules before and after coating was measured using a sample mass comprising approximately 384 mg of ibuprofen sodium. The sample of coated granules was weighed into a 250 mL volumetric flask, and 100 mL of 0.1 N potassium hydroxide (pH 12.8) were added. Afterwards, the samples were treated in an ultrasonic bath for 10 min at 50°C, shaken every third minute for 30 s to extract the ibuprofen sodium from coating. The suspension was then cooled down to room temperature and filtered (Nylon Membran, diameter 25 mm, pore size 0.45 μm). The obtained solution was diluted 10 times with the HPLC mobile phase in a 10 mL volumetric flask and filtered (MCE membrane, diameter 13 mm, pore size 0.22 μm) into vials and prepared for HPLC measurements. The same procedure was carried for uncoated granules, however in this case the samples were manually shaken at room temperature and not treated in the ultrasonic bath. HPLC measurements for the content assay and the dissolution test were performed using a Waters 2996 PDA Detector 195 HPLC system with an auto-sampler and a Purospher STAR RP-18 endcapped (5 µm) Hibar R 125-4 column (Merck, Millipore, Germany). The mobile phase consisted of acetonitrile (67%v/v) and phosphoric acid at 8.5%w/v (5%v/v), filled with ultra-purified water (Milli-Q, Merck Life Science, Darmstadt, Germany) to 100%v/v. Ibuprofen was detected at a wavelength of 231 nm, at a flow rate of 1 mL/min and at a column temperature of 21°C. The retention time of ibuprofen at this flow rate was around 10 min. The injection volume was maintained at 20 µL. 9

3. Results and discussion In this work, roller compaction was employed to produce granules of ibuprofen sodium for subsequent fluid bed hot-melt coating. The granules must be abrasion-resistant to prevent pulverization and generation of fine API particles within the coating layer, which could compromise tastemasking (Jannin & Cuppok, 2013; Jozwiakowski et al., 1990). Moreover, the space for excipients in the core granules formulation was limited because each final sachet should contain up to 1.600 mg (1.6 g) of the dosage form including: -

<-**1**>Matrix of flavors, sweeteners and thickening agents (~ 500 mg);

-

<-**1**>Core granules with 512.5 mg of ibuprofen sodium (equal to 400 mg ibuprofen);

-

<-**1**>Coating formulation for taste masking of API (up to 50%w/w of coated granules).

3.1. Roller compaction formulation and process development

3.1.1. Excipient screening Using mixtures of different grades of microcrystalline cellulose and ibuprofen free acid, Inghelbrecht & Remon (1998) demonstrated the direct relationship between granule strength (resistance to abrasion) and ribbon density (Inghelbrecht & Remon 1998). Eq. 1 written by Reynolds establishes the relationship between the ribbon density and the process parameters (Reynolds et al., 2010): γR=NS.cS./(NR.π.ρtrue.D.W.S) (Equation 1) where γR is the ribbon relative density (-), NS and cS are the screw speed (s-1) and the screw constant (kg), respectively. NR, D, W and S are the roller speed (s-1), diameter (m), width (m) and gap (m), respectively, and ρtrue is the true density (kg/m3) of the material. According to Eq.1, high screw speed, low roller speed and low roller gap yield dense ribbons and consequently granules with higher resistance to abrasion. This principle was employed during excipient screening: the screw speed was increased to its maximum (81 rpm) and the roller speed (3.4 rpm) was set to the 10

minimum. To achieve the minimal roller gap of 0.3 mm, a hydraulic pressure of 60 bar was used. This minimal roller gap was maintained due to the low frictional properties of the ibuprofen sodium powder, which do not exert sufficient roller separating force (Grubber & Reher, 2004). The same property of ibuprofen sodium powder enables the consistent flow of material through the rollers. A consistent material flow was observed for all formulations tested. An additional advantage of this parameter setting is the low throughput and consequently the smaller amount of material required during excipient screening. Note that the ribbon properties were not studied in this work. The ribbons were directly granulated and the friability of granules was analyzed as an indicator of the propensity of granules to be pulverized when subjected to high shear. Fig. 1 shows the variation of friability of ibuprofen sodium granules as a function of the excipient concentration in the formulation. Granules of pure ibuprofen sodium had high friability (57.7 ± 1.2%) due to the very poor compactibility and very poor frictional properties of this material (Grubber & Reher, 2004; Higton, 1999; Yu et al., 2012). Since the selected ratio of excipients (<30%) had no significant impact on these properties, the resulting roller gap was maintained at its minimum (0.3 mm) in all the formulations. As expected based on these very poor properties, the formulations containing up to 30%w/w microcrystalline cellulose resulted in weak granules with almost 42% friability, Surprisingly, a considerable reduction of granular friability was achieved for other excipients, despite the same poor compressible and frictional properties. This suggests that the bonding mechanism may be different from the traditional one described in the introduction section. When testing polyols excipients, i.e., sorbitol, isomalt and mannitol, the degree of granular friability reduction was correlated with the excipient’s solubility in water. Please note that ibuprofen sodium was used as dihydrate form in this study. Water molecules were expected to be released 11

due to the dehydration of ibuprofen sodium during the compaction, resulting in partial solubilization of excipients. Adding 2%w/w of the most soluble polyol (Bolhuis 2009), i.e., sorbitol, reduced the granular friability from 57.7 to 18.1% (± 0.4%). The least soluble polyol (Bohluis 2009), i.e., mannitol at a concentration of 20%w/w, reduced the granular friability only to 38.8% (± 1.2%). Furthermore, water-soluble PVP K25 also had a substantial reduction in friability compared with the insoluble microcrystalline cellulose (Takasaki et al., 2013). The roller compaction of formulations comprising more than 5%w/w of sorbitol had an increased stickiness, especially during the milling step when the friction between granulation blades and screen sieve created a paste. This behavior, which is most probably related to the high solubility of sorbitol in the released water molecules of ibuprofen sodium, was not observed with regard to the less soluble excipient isomalt (Bolhuis 2009). To reduce the stickiness, isomalt was added to sorbitol. The final formulation comprising ibuprofen sodium (94%w/w), isomalt (3.5%w/w) and sorbitol (2.5%w/w) had a friability of 22.5% (± 2.0%). This formulation was selected for further roller compaction studies.

3.1.2. Evaluation of the process throughput The parameter settings selected for the excipient screening resulted in a relatively low throughput of 3.6 kg/h. From a mass balance, the mass throughput of the roller compaction process, m, can be written as: m=ρtrue.γR.D.NR.W.S (Equation 2) (Reynolds et al., 2010). According to Equation 2, without up-scaling the equipment (D and W constants), only the roller gap (S) (m) and/or the roller speed (NR) (s-1) can be increased to improve the process throughput for a certain formulation (ρtrue constant). The effect of the roller gap on the throughput is not discussed in this study due to very low frictional properties of ibuprofen sodium that exert a very limited roller separation force. 12

Figure 2 shows the effect of roller speed on the throughput and friability of granules for the selected formulation, containing 94%w/w API, 2.5%w/w sorbitol and 3.5%w/w isomalt. The screw speed, the roller gap and the hydraulic pressure were kept constant at 81 rpm, 0.3 mm and 60 bar, respectively, while the roller speed increased from 3.4 to 14.8 rpm. The roller speed increment of up to 11 rpm was followed by a proportional increase in the throughput, suggesting that the ribbon density (γR) was constant (eq.2). However, the granules’ friability was not constant: it was the lowest at a lower roller speed and increased with the increasing roller speed. Furthermore, assuming that the selected screw speed (81 rpm) is feeding powder into the rollers at a constant rate, at roller speeds below 11 rpm the screw jams/overloads powder prior to the rollers. Because ibuprofen sodium has very low frictional properties, it was possible to complete the runs even with screw jamming occurring. These two observations suggest that the retention time of the powder before the compaction region was crucial for improving the granular strength. 3.1.3. Robustness of roller compaction process and re-compaction of fines In order to produce granules for hot-melt coating optimization, a roller speed of 9 rpm was selected while the remaining process parameters were the same as described above, i.e., the screw speed 81 rpm, the hydraulic pressure 60 bar and the roller gap 0.3 mm. At this speed, higher throughput is achieved (Fig. 2) without compromising the overloading of powder prior to the rollers required to improve granule strength. The granulation throughput was 9.1 kg/h (Fig. 2) and the friability was 29.4% (±1.6%) (Fig. 3). This friability was lower than that of the evaluation of process throughput (44.1%), which is shown in Fig. 2. The trials for the evaluation of process throughput were took 20 minutes, while the first sample of the trial for evaluation of the robustness of process was taken after one hour processing. This suggests that longer processing time can result in the granules with a lower friability. This phenomenon will be discussed in more detail below. Figures 3, 4 and 5 display the 13

variation of granule friability, sieve fractions, and ibuprofen sodium content, respectively, over 5.5 hours of production time. The only significant variation that occurred during the process was the reduction of particle size observed after the first 60 minutes (Fig. 4). This Figure shows a sustained reduction of the sieve fraction > 0.6 mm and an increase of the fraction < 0.25 mm. The fraction 0.25 – 0.6 mm remained appreciably constant. The reduction of particle size is explained by the deposition of dust in the milling sieves reducing the aperture of the sieve during the process. The mean particle size (x50) and the PSD (x90/x10) of the entire batch were 588 µm and 2.12, respectively. The ibuprofen sodium content (Fig. 5) and the granules’ friability (Fig. 3) were constant during the entire process, showing the robustness of the process. During the hot-melt coating process, the granules must have a narrow size distribution to avoid adhesion of fine on large particles (Jannin & Cuppok, 2013; Jozwiakowski et al., 1990). To that end, the granulation product was sieved in a mechanical sieve equipped initially with a sieve screen of 0.75 mm to remove large particles (2.8%w/w), and afterwards with a sieve screen of 0.3 mm to remove fine particles (53.2%w/w). The selection of a narrow size distribution (0.3-0.75 mm) resulted in a poor process yield of 44.8%w/w. Please note that despite the roller gap of 0.3 mm, which results in a relatively thin ribbon, it was possible to produce granules with a particle size of 0.75 mm and larger due to the selected lower screen sieve of the granulation step having a size of 0.8 mm. The ribbon properties were not studied in this work. In order to improve the yield, the re-compaction of fines; i.e. the fraction below 0.3 mm with the yield of 53.2%w/w by roller compaction was investigated. In the re-compaction of fines, the screw feed was reduced to 25 rpm since the torque generated at higher speeds was not supported by the equipment. Additionally, the roller gap was increased to 0.8 mm as a result of a higher rollers separating force created by the fine fraction of granules, compared with the powder formulation. The roller speed was 9 rpm. The resulting throughput 14

was 12.3 kg/h. Despite different process parameters, the granules’ friability (27.2 ± 3.3%) and the sieve fractions [f<0.25mm= 46.5 ± 0.9%; f0.25-0.6mm = 44.2 ± 1.1%; f>0.6mm=9.0 ± 0.1%] were within the same range as the initial granules. The ibuprofen sodium content of these granules was 94.9%w/w (± 0.2%w/w). From a practical point of view, the re-compaction of fines can be employed to improve the roller compaction yield significantly while providing granules resistant to abrasion. The performed re-compacting cycle improved the yield from 44.8 to 74.3%.

3.2. Solid state and bonding mechanism The effective bonding mechanism of sorbitol and isomalt and the solid state of ibuprofen sodium in the granules were investigated. Figure 6 displays SEM images of ibuprofen sodium granules comprising sorbitol (5%w/w) (Fig. 6a) and ibuprofen sodium granules comprising isomalt (3.5%w/w) and sorbitol (2.5%w/w) (Fig. 6b). Both surface images show densely packed granules that are covered with thin filaments characteristic of sorbitol (Bolhuis et al., 2009). In Fig. 6a these filaments fuse together into round shape structures, suggesting the re-solidification of sorbitol on the surface of granules after solubilization or deliquesce during the roller compaction process. This extensive change, which is most probably related to the high solubility of sorbitol in the released water molecules of ibuprofen sodium was not observed when adding less water-soluble isomalt to the formulation (Fig. 6b). Figure 7 displays the DSC analysis of the raw material, physical mixtures and granules. The thermogram of ibuprofen sodium shows a first endothermic event from 60 to 90°C corresponding to the crystal dehydration and a second endothermic event at 200°C corresponding to the melting of anhydrous ibuprofen sodium (Censi et al., 2013, Zhang & Grant, 2005; Zhang et al., 2003). Although pure crystalline sorbitol melts at 96°C, in the physical mixture its melting event seems to be shifted to and maybe partially overlaps the dehydration endotherm of ibuprofen sodium, around 70°C, due to the shouldering endotherm, which is apparent from the thermogram of phys15

ical mixture (Fig. 7a). The reduction of melting temperature of sorbitol is directly related to the water content of physical mixture (Quinquenet 1988). Isomalt (GalenIQ 721) is a mixture 3:1 of α-D-glucopyranosyl-1-6-sorbitol anhydrous (GPS) and α-D-glucopyranosyl-1-6-mannitol dihydrate (GPM) (Borde & Cesaro, 2001). The dehydration of isomalt occurs from 70 to 100°C, and anhydrous isomalt melts from 140 to 160°C. In the thermogram of the physical mixture it is not possible to visualize the dehydration of isomalt (Fig.7b). In the thermogram of granules, the dehydration endotherm event was broader than in the physical mixtures and spanned to higher temperatures (Fig. 7a-c). Additionally, in granules containing sorbitol (Fig 7a and c), there is a new endothermic event around 50°C that corresponds to the melting of sorbitol hydrate (Quinquenet et al., 1988). The low melting point of this entity can explain the increased stickiness of the formulation comprising 5%w/w sorbitol in the milling sieve and the round shape structures on the surface of its granules (Fig. 6a). Because sorbitol hydrate is formed from aqueous solutions of sorbitol (Quinquenet et al., 1988), we can assume that during roller compaction the pressure and generated heat (Guigon et al., 2007, Osborne et al., 2013) solubilized or deliquesced the sorbitol to a certain extent in a freely available fraction of water. The increase of powder temperature can be explained with the jamming prior to the rollers caused by the selected parameter setting, i.e., low roller speed and high screw speeds. The high screw speed produces thermal energy in this region, rising the temperature of the retained powder. Moreover, the friability of granules produced during the first 20 minutes was 44.1% while the friability of those produced during the first hour was reduced to 28.4% and was stable during the following hours, which can be correlated with the temperature rise and stabilization. Thus, a bonding mechanism is suggested to be as follows: 1- <1-**1**>The jamming prior to the rollers results in increased compacted powder temperature (~54°C); 16

2- <2-**1**>Compaction pressures and high temperatures of compacted powder initiate the dehydration of ibuprofen sodium; 3- <3-**1**>Water molecules released from ibuprofen sodium and high temperatures partially solubilize crystalline sorbitol; 4- <4-**1**>Solubilized sorbitol recrystallizes as sorbitol hydrate on the surface of ibuprofen sodium particles creating strong solid bridges, as shown in Fig. 6a. Although solid bridges are not typical in roller compaction processes (Mollet & Grubenmann, 2001; Summers & Aulton, 2006), high process temperatures and a high water content have been proven to create solid bridges via roller compaction of an amorphous dextrose syrup powder (Osborne et al., 2013). Most probably, the same mechanism occurs in the case of isomalt. However, in this work the partial dissolution of isomalt during roller compaction is not confirmed since dehydration enthalpy analysis was not performed. The suggested bonding mechanism involves API dehydration, solubilization of crystalline sorbitol and recrystallization into sorbitol hydrate. To complete the DSC data, small- and wide- angle X-ray scattering (SWAXS) were employed. Ibuprofen sodium is commercialized in a stable dihydrate form without any known polymorphs (Zhang & Grant, 2005). The dehydration of this crystal occurs from 60 to 90°C at atmospheric pressures, during which the crystal loses 13.6% of its weight (Censi et al., 2013). At higher temperatures, three polymorphs (α, β and γ) of anhydrous ibuprofen sodium were identified (Zhang et al., 2003). The most stable of these polymorphs is the γ-form that melts around 200°C. The αand β-forms are ``enantiotropically related,'' with a transition temperature of 113 °C. At temperatures below 113°C, the α-form is the most stable of the two. Above this temperature, the β-form is more stable, melting at around 190°C. The thermograms (Fig.7a-c) of the granules and physical mixtures show that the ibuprofen sodium anhydrous formed after dehydration melts in the βform and does not recrystallize into the γ-form. Other small endothermic events (130-170°C) can be due to the melting of the α-form. We employed SWAXS to study whether these ibuprofen 17

sodium anhydrous forms are only created during DSC analysis or if they are present in the granules. The most characteristic distance of ibuprofen sodium is in the SAXS region, where ibuprofen sodium dihydrate has a peak at 2θ = 3.7 ° (Zhang & Grant, 2005) and the anhydrous forms have a signal at 2θ = 4.0 (γ form); 2θ = 5.2 and 5.7 (α form); and 2θ = 5.9 (β form) (Zhang et al., 2003). Figure 8 displays the overlaps of SAXS patterns. This scattering demonstrates that only the dihydrated form of ibuprofen sodium is present in the granules. Censi et al. proved that, although it is easy to remove water from the crystal of ibuprofen sodium, the driving force of the hydration is higher than that for the dehydration (Censi et al., 2013). From a thermodynamic point of view, these observations explain why only the hydrated form of ibuprofen sodium is found after the roller compaction process. Additionally, the full reversal of transition from anhydrate to dihydrate of ibuprofen sodium was reported to be almost completed after about one hour in the air at room temperature (Rossi et al., 2014), which is in agreement with our results.

3.3. Hot-melt coating formulation and process development ``Direct to mouth'' granules of ibuprofen sodium are bitter and cause a burning sensation (Higton 1999). To mask the unpleasent taste of ibuprofen sodium, a coating formulation comprising tripalmitin and polysorbate 65 was applied during hot-melt coating. The taste-masking efficacy of this coating formulation was previously confirmed in in vivo studies with N-acetylcysteine (Becker et al., 2016). In the current work, organoleptic measurements were avoided and a dissolution criterion (≤10% API dissolved in 5 minutes) that largely depends on the taste intensity of the drug was selected for an in vitro evaluation of the taste-masking properties (Siewert et al., 2003).

18

The HMC formulation and process were evaluated and optimized using an extended Rechtschaffner quadratic model. The input parameter-settings and the related results of the output parameters are listed in Table 3. A statistical evaluation of the model showed its optimal fitting. For the output parameter ``time for the 85% release of API,'' R2 was 0.92, which demonstrates the model’s fit. Q2 was 0.89, which shows the future prediction precision. The model validity was 0.85 and the reproducibility was 0.92. For the output parameter ``API released within first 5 minutes,'' R2 was 0.90 and Q2 was 0.83. The model’s validity was 0.56 and reproducibility was 0.98. Models with a high reproducibility generally have a lower validity. The model indicated that the coating ratio (CoR) and the ratio of emulsifier/lipid (EMU) had the most significant impact on both of the output parameters (Fig. 9). The time required for the release of 85% of API decreased by reducing the coating ratio or by increasing the emulsifier’s ratio in the coating formulation. This means that the in vitro immediate release profile can be positively adjusted in this manner. The positive impact of the emulsifier (polysorbate 65) on the faster release of API from the coating is due to its HLB value of 10.5, making it suitable as an O/W emulsifier. The output parameter ``API released in 5 minutes,'' which is an indicator of the taste-masking efficiency, was positively affected by increasing the coating ratio (CoR) or by decreasing the emulsifier ratio (EMU) in the coating formulation. However, the taste-masking efficiency was negatively affected by decreasing the inlet air temperature and its interaction with the emulsifier ratio. Such a setting results in the release of higher amounts of API within the first 5 minutes of the dissolution test and affects the taste masking negatively. Note that the process temperature was always almost 5°C above the selected inlet temperature. This means that when selecting inlet temperatures of 25°C, 30°C and 35°C, the gained process 19

temperatures were 30°C, 35°C, and 40°C, respectively. A process temperature close to the crystallization temperature of the β-form of lipid (approximately 49°C for tripalmitin) can prevent the fast and punctual crystallization of lipid on the surface of the core, providing a better coating quality. Moreover, the effect of inlet temperature on the coating and the release profile can be due to the increased tortuosity of triacylglycerol’s microstructure when crystallized at higher temperatures (Lopes et al., 2015). The results show the requirements for main input parameters for achieving the product specification, i.e., the immediate release profile (85% release of API in max. 30 minutes) combined with efficient taste-masking properties (max. 10% release of API in first 5 minutes). Obtaining an in vitro immediate release profile was possible with 35%w/w coating ratio and the lowest emulsifier ratio (5%w/w). Increasing the coating ratio to 50%w/w had to be combined with an increased emulsifier ratio (Table 2). The optimal taste masking could be achieved with at least 38.5%w/w coating ratio and 5%w/w emulsifier ratio. By increasing the coating ratio to 50%w/w, it was possible to increase the emulsifier content to 15%w/w and still ensure optimal taste masking. Figure 10 shows the optimal design space for both of the output factors (in vitro immediate release and taste masking). The optimal space is colored black. Coating and emulsifier ratios were considered as variables for creating a design space. The inlet air temperature was set at the highest value (35°C) and other input parameters were set at their mean value (spray rate = 5 g/min, atomizing air pressure = 1.1 bar and air flow = 37.5 m3/h). As shown in Figure 10, the suggested lower border for the variables was the average of 40%w/w (± 1.5%w/w) coating with 5%w/w emulsifier. Increasing the coating ratio should be accompanied by increasing the ratio of emulsifier. However, it is advisable to use higher coating ratios than suggested by software to ensure the coating quality: 40%w/w, combined with an average of 10%w/w emulsifier. Combinations of tripalmitin and 10%w/w emulsifier with a coating ratio of more than 40% w/w (4420

48%w/w) should be reliable. The effect of 10%w/w polysorbate 65 on the polymorphism of tripalmitin and on the stability of the coating has been discussed in detail elsewhere (Becker et al., 2016). While N-acetylcysteine crystals could be taste masked with coating amounts of up to 40%w/w (Becker 2016), this study shows that larger coating amounts (44-48%w/w) are required for ibuprofen sodium granules. This is due to the more densely packed structure of N-acetylcysteine crystals compared with ibuprofen sodium granules, which is more resistant to abrasion during hot-melt coating. Conclusions In this work, an easy-to-swallow medicine of ibuprofen sodium was developed by combining hot-melt coating and roller compaction. High concentrations of ibuprofen sodium (≥94%w/w) in the roller compaction formulation were achieved by using the polyols sorbitol and isomalt as excipients. An investigation of the bonding mechanism showed that sorbitol is solubilized during roller compaction and recrystallizes as sorbitol hydrate producing strong solid bridges. Increased temperature and retention time of compacted powder prior to the rollers that are crucial for the bonding mechanism can be provided by increased screw to roller speed ratios. This nontraditional process resulted in the production of robust granules, suitable for subsequent hot-melt coating. The quality of granules was constant during the process. The main drawback was the small size of 0.3-0.75 mm fraction, resulting in a poor yield of 44.8%w/w. However, we demonstrated that this yield can be improved by re-compaction of the fines fraction without compromising the drug content of granules. Critical hot-melt coating process parameters and material attributes affecting the properties of coated granules were evaluated using a Rechtschaffner statistical design. In particular, the coating amount, emulsifier content and inlet air temperature were identified as the main critical parame21

ters. Taste-masking of ibuprofen sodium granules required coating amounts larger than 40%w/w emphasizing the need of high drug content and abrasion-resistant granules. A large amount of coating material is due to a larger specific surface area of granules as multi-particulate systems, compared with other oral dosage forms, such as tablets or capsules. Despite using molten lipid as a coating material, no drug degradation is expected during the coating process, which is due to the moderate inlet air temperatures applied (within the range of 25 °C-35 °C). Acknowledgement This work was funded through the Austrian COMET Program by the Austrian Federal Ministry of Transport, Innovation and Technology (BMVIT), the Austrian Federal Ministry of Economy, Family and Youth (BMWFJ) and by the State of Styria (Styrian Funding Agency SFG) The authors thank Hermes Arzneimittel GmbH, and IOI Oleo GmbH for kindly providing of material. References Becker, K., Saurugger, E., Kienberger, D., Lopes, D., Haack, D., Köberle, M., Stehr, M., Lochmann, D., Zimmer, A., Salar-Behzadi, S.,;1; 2016. Advanced stable lipid-based formulations for a patient-centric product design. Int. J. Pharm. 497, 136-149. Becker, K., Zimmer, A., Haack, D., Salar-Behzadi, S.,;1; 2013. Oral pharmaceutical compositions comprising taste-masked N-acetylcysteine. EP13163638.3. Bolhuis, G.K., Rexwinkel, E.G., Zuurman, K.,;1; 2009. Polyols as filler-binders for disintegration tablets prepared by direct compaction. Drug Dev. Ind. Pharm. 35, 671-677. Borde, B., Cesàro, A.,;1; 2001. A DSC study of hydrated sugar alcohols: Isomalt. J. Therm. Anal. Cal. 66, 179-195. Censi, R., Martena, V., Hoti, E., Malaj, L., Martino, P.,;1; 2013. Sodium ibuprofen dihydrate and anhydrous: study of the dehydration and hydration mechanisms. J. Therm. Anal. Calorim. 111, 2009-2018. Gandhi, B., Baheti, J.,;1; 2013. Multiparticulates drug delivery systems: a review. Int. J. Pharm. Chem. Sci. 2, 1620-1626. Gruber, P., Reher, M.,;1; 2004. Dosage form of sodium ibuprofen. CA 2500987 A1. Guigon, P., Simon, O., Saleh, K., Bindhumadhavan, G., Adams, M.J., Seville, J.P.K.,;1; 2007. Roll pressing, in: Salman, A.D., Hounslow, M.J., Seville, J.P.K. (Eds), Granulation. Elsevier, pp. 255-288. Hancock, B.C., Colvin, J.T., Mullarney, M.P. & Zinchuk, A.V.,;1; 2003. The relative densities of pharmaceutical powders, blends, dry granulations, and immediate-release tablets, Pharm. Tech. 27, 64-80. 22

Higton, F.,;1; 1999. The pharmaceutics of ibuprofen. In: Rainsford, K.D. (Ed), Ibuprofen: A critical review. Taylor & Francis, London, pp. 48-78. Inghelbrecht S., Remon, J.P.,;1; 1998. Roller compaction and tableting of microcrystalline cellulose/drug mixtures. Int. J. Pharm. 161, 215-224. Jannin, V., Cuppok, Y.,;1; 2013. Hot-melt coating with lipid excipients. Int. J. Pharm. 457, 480487. Jozwiakowski, M.J., Jones, D.M., Franz, R.M.,;1; 1990. Characterization of a hot-melt fluid bed coating process for fine granules. Pharm. Res. 7, 1119-1126. Khorasani, M., Amigo, J.M., Bertelsen, P., Sun, C.C., Rantanen, M.,;1; 2016. Process optimization of dry granulation based tableting line: Extracting physical material characteristics from granules, ribbons and tablets using near-IR (NIR) spectroscopic measurement. Powder Technol. http://dx.doi.org/10.1016/j.powtec.2016.03.004 Lopes, D.G., Becker, K., Stehr, M., Lochmann, D., Haack, D., Zimmer, A., Salar-Behzadi, S.,;1; 2015. Role of lipid blooming and crystallite size in the performance of highly soluble drugloaded microcapsules. J. Pharm. Sci. 104, 4257-4265. Miller, R.W.,;1; 2005. Roller Compaction Technology, in: Swarbrick, J. (Ed), Handbook of pharmaceutical granulation technology. Taylor & Francis Group, LLC. Pp. 159-190. Mollet, H., Grubenmann, A.,;1; 2001. Solid forms, in: Mollet, H., Grubenmann, A. (Eds), Formulation Technology: Emulsions, Suspensions and Solid forms. Wiley VCH Verlag GmbH, pp. 181-246. Osborne, J.D., Althaus, T., Forny, L., Niederreiter, G., Palzer, S., Hounslow, M.J., Salman, A.D.,;1; 2013. Investigating the influence of moisture content and pressure on the bonding mechanisms during roller compaction of an amorphous material. Chem. Eng. Sci. 86, 61-69. Patel, P.B., Dhake, A.S.,;1; 2011. Multiparticulate approach: an emerging trend in colon specific drug delivery for chronotherapy. J. Appl. Pharm. Sci. 1, 59-63. Patwekar, S.L., Baramade, M.K.,;1; 2012. Controlled release approach to novel multiparticulate drug delivery systems. Int. J. Pharm. Pharm. Sci. 4, 757-763. Quinquenet, S., Ollivon, M., Gabriele-Madelmont, C.,;1; 1988. Polymorphism of hydrated sorbitol. Termoch. Act. 125, 125-140. Rainsford, K.D.;1; 2009. Ibuprofen: pharmacology, efficacy and safety. Inflammopharmacol. 17, 275-342. Reynolds G., Ingale, R., Roberts, R., Kothari, S., Gururajan, B.,;1; 2010. Practical application of roller compaction process modeling. Comp. Chem. Eng. 34, 1049-1057. Rossi, P., Macedi, E., Paoli, P., Bernazzi, L., Carignani, E., Borsacchi, S., Geppi, M.,;1; 2014. Solid-solid transition between hydrated racemic compound and anhydrous conglomerate in NaIbuprofen: A combined x-ray diffraction, solid-state NMR, calorimetric, and computational study. Crys. Groeth Des. 14, 2441-2452. Sharma, A., Chaurasia, S.,;1; 2013. Multiparticulate drug delivery system: palletization through extrusion and spheronization. Int. Res. J. Pharm. 4, 6-9. Siewert, M., Dressman, J., Brown, C.K., Shah, V.P.,;1; 2003. FIP/AAPS Guidelines to dissolution/in vitro release testing novel/special dosage forms. AAPS PharmSciTech. 4, 1-10. Soergel, F., Fuhr, U., Minic, M., Siegmund, M., Maares, J., Jetter, A., Kinzig-Schippers, M., Romalik-Scharte, D., Szymanski, J., Goeser, T., Toex, U., Scheidel, B., Lehmacher, W.;1; 2005. Pharmacokinetics of ibuprofen sodium dihydrate and gastrointestinal tolerability of short-term treatment with a novel, rapidly absorbed formulation. Int. J. Clin. Pharmacol. Ther. 43, 140-149. 23

Stegemann, S., Klebovich, I., Antal, I., Blume, H.H., Magyar, K., Németh, G., Paál, T.L., Stumptner, W., Thaler, G., Van de Pitte, A., Shah, V.,;1; 2011. Improved therapeutic entities derived from known generics as an unexplored source of innovative drug products. Eur. J Pharm Sci. 44, 447-454. Summers, M., Aulton, M.,;1; 2006. Granulation. In: Aulton M. (Ed), Pharmaceutics: the science of dosage form design. Churchil Livingstone. pp. 365-378. Takasaki, H., Yonemochi, E., Messerschmid, R., Ito, M., Wada, K., Terada, K.,;1; 2013. Importance of excipients wettability on tablet characteristics prepared by moisture activated dry granulation (MADG). I. J. Pharm. 456, 58-64. Teng, Y., Qiu, Z., Wen, H.,;1; 2009. Systematical approach of formulation and process development using roller compaction. E. J. Pharm. Bio. 73, 219-229. Wahlich, J., Stegemann, S., Orlu-Gul, M.;1; 2013. Meeting commentary – „Medicines for older adults: Learning from practice to develop patient centric drug products``. Int. J. Pharm. 456, 251257. Werner, S.R.L., Jones, J.R., Paterson, A.H.J., Archer, R.H., Pearce, D.L.,;1; 2007. Air-suspension coating in the food industry: Part II – micro-level process approach. 171, 34-45. Yu, S., Gururajan, B., Reynolds, G., Roberts, R., Adams, M.J., Wu, C.,;1; 2012. A comparative study roll compaction of free-flowing and cohesive pharmaceutical powders. Int. J. Pharm. 428, 39-47. Zhang, Y., Grant, D.J.W.,;1; 2005. Similarity in structures of racemic and enantiomeric ibuprofen sodium dihydrates. Act. Cryst. C61, m435-m438. Zhang, G.G.Z., Paspal, S.Y.L., Suryanarayanan, R., Grant, D.J.W.;1; 2003. Racemic species of sodium ibuprofen: characterization and polymorphic relationships. J. Pharm. Sci. 92, 1356-1366. Figures caption:

Fig. 1. Impact of the concentration (%w/w) of selected pharmaceutical excipients on the friability (%) of ibuprofen sodium granules. Concentration of 0 [%w/w] indicates 100% API. Vertical bars represent the error (s.d.) among replicate measurements (n=3)

Fig. 2. Impact of the roller speed (rpm) on the process throughput [kg/h] and granules’ friability (%) of a formulation comprising ibuprofen sodium (94.0%w/w), isomalt (3.5%w/w) and sorbitol (2.5%w/w).

Fig. 3. Friability (%) of granules collected during roller compaction for evaluation of process robustness. Vertical bars represent the error (s.d.) among replicate measurements (n=3)

Fig. 4. Sieve fraction (f<0.6; f0.25-0.6; and f<0.25) characterization of granules collected during roller compaction for evaluation of process robustness.. Vertical bars represent the error (s.d.) among replicate measurements (n=3)

Fig. 5. Ibuprofen sodium content (%w/w) of granules collected during roller compaction for evaluation of process robustness. Content of 94% w/w ibuprofen sodium is related to the total content of granules and corresponds to the recovery of 100 % of API.

Fig. 6. SEM images of roller compacted ibuprofen sodium granules comprising: a) sorbitol (5.0%w/w); b) sorbitol (2.5%w/w) and isomalt (3.5%w/w). 24

Fig. 7. Differential scanning calorimetry of raw material, physical mixtures and granules of ibuprofen sodium formulations comprising: a) sorbitol (5.0%w/w); b) isomalt (5%w/w); and c) sorbitol (2.5%w/w) and isomalt (3.5%w/w).

Fig. 8. Small angle x-ray scattering of ibuprofen sodium, ibuprofen sodium granules comprising sorbitol (2.5%w/w) and isomalt (3.5%w/w), ibuprofen sodium granules comprising isomalt (5.0%w/w) and ibuprofen sodium granules comprising sorbitol (5.0%w/w). (2θ represents the scattering angle)

Fig. 9. Scaled and centered coefficient plot for a) taste-masking, b)immediate release profile. Input parameters are coating ratio (CoR), inlet air temperature (InT), emulsifier/lipid ratio (EMU) and their interactions (CoR*EMU; InT*EMU; CoR*EMU).

Fig. 10. Design space for achieving product specification; immediate release profile and taste masking Table 1. List of roller compaction formulations employed Ibuprofen Sodium (%w/w)

Microcrystalline cellulose(%w/w)

PVP K25 (%w/w)

Mannitol (%w/w)

Isomalt (%w/w)

Sorbitol (%w/w)

95; 90; 80;70

5; 10; 20; 30

-

-

-

-

90; 80; 70

-

10; 20; 30

-

-

-

95;90;80

-

-

5; 10; 20

-

-

95;90

-

-

-

5; 10

-

98; 95;92

-

-

-

-

2; 5; 8

94

-

-

-

3.5

2.5

94

-

-

-

3.5

2.5

Excipient screening

Throughput evaluation

Process robustness

Table 2. Input parameter-settings for the Rechtschaffner DoE. The midpoints were run three times (centerpoints) Input parameters

Parameter-setting -1 0 2.5 5 0.8 1.1 30 37.5 25 30 35 4.5

Spray rate (g/min) Spray pressure (bar) Airflow (m3/h) Inlet temperature (°C) Ratio of coating to batch of granules (%w/w)

25

+1 7.5 1.4 45 35 50

Ratio of emulsifier to tripalmitin (%w/w)

5

10

15

Table 3. Input parameter-settings and related results of the output parameters for the Rechtschaffner DoE, asterisks demonstrate the values, which reach the specifications for each output parameter, light blue highlighted runs reach both specifications for immediate release profile (85 % release of API within 30 minutes) and taste masking (up to 10% release of API in 5 minutes) Input parameters Run

Spray rate (g/min)

Spray pressure (bar)

Air flow (m3/h)

Coating ratio (%w/w)

Inlet air temperature (°C)

Emulsifier ratio (%)

1

2.5

0.8

30

35

25

5

Output parameters Drug Time of release 85% API at 5 release min (min) (%) 14.5* 41.8

2

2.5

1.4

45

50

35

15

23.2*

16.2

3

7.5

0.8

45

50

35

15

16.9*

24.4

4

7.5

1.4

30

50

35

15

29.3*

4.5*

5

7.5

1.4

45

35

35

15

11.4*

38.2

6

7.5

1.4

45

50

25

15

19.1*

8.3*

7

7.5

1.4

45

50

35

5

115

2.5*

8

7.5

1.4

30

35

25

5

18.7*

44.1

9

7.5

0.8

45

35

25

5

15*

58.3

10

7.5

0.8

30

50

25

5

73.6

4.2*

11

7.5

0.8

30

35

35

5

30.7

16.6

12

7.5

0.8

30

35

25

15

8.6*

65.3

13

2.5

1.4

45

35

25

5

9.2*

52.3

14

2.5

1.4

30

50

25

5

84.6

5.2*

15

2.5

1.4

30

35

35

5

25.7*

16.2

16

2.5

1.4

30

35

25

15

9.6*

47.6

17

2.5

0.8

45

50

25

5

44.1

4.5*

18

2.5

0.8

45

35

35

5

20.3*

24.3

19

2.5

0.8

45

35

25

15

8.2*

65.9

20

2.5

0.8

30

50

35

5

77.7

0.9*

21

2.5

0.8

30

50

25

15

15.6*

15.8

22

2.5

0.8

30

35

35

15

8.3*

70.1

23

5

1.1

37.5

42.5

30

10

19.2*

10.7*

24

5

1.1

37.5

42.5

30

10

16.5*

11.6*

25

5

1.1

37.5

42.5

30

10

26.3*

7.5*

26

7.5

1.4

30

50

25

5

123.01

1.3*

27

7.5

0.8

30

50

35

15

17.96*

24.3

28

7.5

0.8

45

35

25

15

12.17*

47.4

29

2.5

1.4

30

35

35

15

9.931*

51.8

26

30

7.5

1.4

30

35

25

15

9.066*

52.8

31

2.5

0.8

45

35

35

15

9.258*

58

32

7.5

1.4

45

35

35

5

13.4*

36.1

33

2.5

1.4

30

50

35

5

88.8

0.9*

34

2.5

0.8

45

50

35

5

65.03

0.5*

35

7.5

0.8

45

50

25

5

118.39

1.9*

36

2.5

1.4

45

50

25

15

21.09*

4.6*

37

7.4

1.4

45

50

35

15

20.6*

7.6*

TDENDOFDOCTD

27

×

Report "Improving the granule strength of roller-compacted ibuprofen sodium for hot-melt coating processing"

Your name

Email

Reason -Select Reason- Pornographic Defamatory Illegal/Unlawful Spam Other Terms Of Service Violation File a copyright complaint

Description

Close Send

Our partners will collect data and use cookies for ad personalization and measurement. Learn how we and our ad partner Google, collect and use data. Agree & close