RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Combined Study of Biphasic and Zero-Order Release Formulations with Dissolution Tests and ATR–FTIR Spectroscopic Imaging PATRICK WRAY, JING LI, LING QIAO LI, SERGEI G. KAZARIAN Department of Chemical Engineering, Imperial College London, London SW7 2AZ, United Kingdom Received 21 November 2013; revised 26 March 2014; accepted 7 April 2014 Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23987 ABSTRACT: In this study of multi-layer tablets, the dissolution of biphasic and zero-order release formulations has been studied primarily using attenuated total reflection–Fourier transform infrared (ATR–FTIR) spectroscopic imaging as well as UV–Vis detection of dissolved drug in the effluent stream and USP dissolution testing. Bilayer tablets, containing the excipients microcrystalline cellulose (MCC) and glucose, were used for biphasic release with nicotinamide and buflomedil as model drugs. ATR–FTIR spectroscopic imaging showed the changing component distributions during dissolution. Further experiments studied monolithic and barrier-layered tablets containing hydroxypropyl methylcellulose, MCC and buflomedil dissolving in a USP I apparatus. These data were compared with UV–Vis dissolution profiles obtained online with the ATR flow-through cell. ATR–FTIR imaging data of the biphasic formulations demonstrated that the drug release was affected by excipient ratios and effects such as interference between tablet sections. Tablets placed in the ATR–FTIR flow-through cell exhibited zero-order UV–Vis dissolution profile data at high flow rates, similar to barrier-layered formulations studied using the USP I apparatus. ATR–FTIR spectroscopic imaging provided information regarding the dissolution mechanisms in multi-layer tablets which could assist formulation development. The ability to relate data from USP dissolution tests with that from the ATR–FTIR flow-through cell could help C 2014 Wiley Periodicals, Inc. and the American Pharmacists spectroscopic imaging complement dissolution methods used in the industry.  Association J Pharm Sci Keywords: FTIR spectroscopic imaging; ATR–FTIR spectroscopy; drug release; biphasic delivery; zero-order release; imaging methods; dissolution; controlled release; tableting; spectroscopic imaging; zero-order release

INTRODUCTION Standard monolithic tablets containing a uniformly distributed drug generally provide continuously diminishing release rates.1,2 More complex tablet structures offer greater control and flexibility over the delivery of the drug in the patient and can create many different release regimes. This work investigates the application of FTIR imaging to study release from bilayer tablets and uses tri-layer formulations to help study release from constrained geometries. Tablets with a multi-layered structure offer an opportunity to modify release profiles. One example is bimodal release and although this may be achieved using other geometries such as core-in-cup tablets,3 Streubel et al.4 used hydroxypropyl methylcellulose acetate succinate-based tablets for pHdependent bimodal release. The multi-layer structure has also been successfully applied to create biphasic formulations for the purpose of quick/slow release, which may be achieved by using one fast release layer and another layer containing a controlled release matrix.5 Uekama et al.6 used a double-layer tablet for biphasic release of piretanide with $-cyclodextrin being used in the fast layer and ethyl cellulose and hydroxypropyl methylcellulose (HPMC) in the slow release layer. Maggi et al.7 used double-layer tablets for biphasic release of two different drugs, in which the speed of release from the fast release layer was determined by the addition of a super disintegrant.

Correspondence to: Sergei G. Kazarian (Telephone: +44-207-594-5574; Fax: +44-207-594-5604; E-mail: [email protected]) Journal of Pharmaceutical Sciences

 C 2014 Wiley Periodicals, Inc. and the American Pharmacists Association

A common purpose of using multi-layered tablets is to produce zero-order release tablets, which have the desirable property of exhibiting a constant release rate of drug.8–12 These multi-layer formulations employ barrier layers as a method for limiting the area of the tablet exposed to the dissolution medium. Zero-order release can also be attained using monolithic tablets.13 Baveja et al.13 used non-ionic HPMC and anionic sodium carboxymethylcellulose in a carefully controlled ratio with each other and the drug. This led to the swelling and expansion of the HPMC being balanced by the erosion of the gel layer, maintaining a constant diffusion length for the drug. Zero-order release has also been demonstrated using doughnutshaped tablets.14 Colombo et al.15 achieved a close to linear release profile by applying impermeable barrier coatings to the top and bottom faces of HPMC-based tablets to control the direction of swelling, restricting changes in the surface area for release. Further work found the manually applied film barriers could be replaced in favour of swellable HPMC barriers that would swell with the core maintaining the surface area of the exposed core.10 It is possible that the attenuated total reflection–Fourier transform infrared (ATR–FTIR) flow cell could be considered as a barrier layer to the tablet and it will be investigated whether the multilayer form may have some similar effect upon dissolution to the flow cell structure. Upon dissolution of these multi-layered tablets, complex changes to the structure of the tablet and the distribution of the components can occur. A better understanding of these processes and the effects they have on drug release can assist in the development of future formulations. FTIR spectroscopy can be used in conjunction with a focal plane array (FPA) detector to generate chemical images showing the distribution Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

1

2

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

of compounds within a sample.16 The acquisition of speed of the detector combined with the versatility of the ATR sampling methodology make this approach highly applicable for studying pharmaceutical formulations.17–21 Several custom-designed dissolution cells have been developed for use with the ATR–FTIR imaging approach which allow for in situ study of the compaction and dissolution of tablets. This approach may even be used in conjunction with a UV– Vis spectrometer to provide conventional dissolution profile data.22–25 A visibly transparent Perspex dissolution cell for use with tablets compacted ex situ has also been used for in situ ATR–FTIR imaging studies of dissolution.26 This cell can be used in conjunction with a visible optical video camera and non-cylindrical samples. The industry standard USP dissolution tests are the most commonly used approaches for assessing the dissolution profiles of formulations in pharmaceutical science by measuring the total amount of drug dissolved over time.27 These important tests were specifically designed as a necessary step in determining drug bioavailability and product performance.28 Although ATR–FTIR imaging has great potential for studying the release of drugs from tablets, it would be useful to compare the release data gathered from this method with that obtained from the industry standard USP dissolution tests. The major differences in these two approaches are found in the flow regimes and the geometry of the systems. The USP I and USP II type apparatuses are stirred tank systems, whereas the custom ATR dissolution cell is a flow-through system.20 In the USP systems, the tablet is exposed to the dissolution medium on all sides, whereas in the ATR flow cell, the top and bottom faces are covered, restricting the surface area, similar to a barrierlayered tablet. The differences between the geometries of the two dissolution approaches are shown in Figure 1. Although the USP I and USP II apparatuses are commonly used in industry, these apparatuses provide no information on the complex chemical and physical changes, such as polymer swelling and erosion, which occur during dissolution, which ATR–FTIR imaging can provide. In this work, ATR–FTIR imaging was used to study the processes which occur during the dissolution of bilayer tablets at the interface between tablet sections.29 This helped to understand how the more soluble layer-affected water ingress into the less soluble layer, and hence the release of the drug. The spatially resolved nature of the chemical information produced

by this approach allowed for the identification of the mechanisms of drug release and effects such as interference between tablet halves during dissolution. The second part of the results serves as a preliminary investigation of the effects of the differences between the flow-through cell and the USP apparatus. ATR–FTIR imaging was applied to help understand the effects of dissolution medium flow rate and the geometry of the flowthrough cell on tablet dissolution.

MATERIALS AND METHODS Tablet Formulations The biphasic tablets were designed to release two different soluble drugs, buflomedil (Lisapharma SpA, Erba Como, Italy) and nicotinamide (Sigma–Aldrich,Gillingham, Dorset, UK) from separate halves of the tablet. Each half of the matrix also contained microcrystalline cellulose (MCC; Merck Sharp & Dohme, Hoddesdon, Hertfordshire,UK) and glucose (Sigma–Aldrich, Gillingham, Dorset, UK) in varying amounts to produce differing dissolution properties for each half. The mass fraction of drug was 20 wt % in each section of the tablet, whereas the full range of concentrations of the remaining components used is shown in Table 1. The powders were ground down and sieved to particle sizes of less than 90 :m, and then mechanically mixed. A customdesigned compaction cell was created which produced rectangular multi-layered tablets. The tablets created by the cell were 3 mm wide and 2 mm deep, with each half of the tablet weighing 10 mg. To ensure good adherence of the tablet sections, the pressure was applied in stages. The first section of the tablet was compacted with 60 MPa, then the powder for the second section of the tablet was added and everything was compacted to 120 MPa. A monolithic tablet and a barrier-layered tablet were constructed for dissolution in the USP I type apparatus. The middle section of the barrier layered tablet was the same as the monolithic tablet: 45 wt % buflomedil (Lisapharma SpA), 35 wt % HPMC K4M (Colorcon, Dartford, Kent, UK), 20 wt % MCC (Merck Sharp & Dohme, Hoddesdon, Hertfordshire, UK). The tablets were compacted to be 10 mm in diameter using a cylindrical tablet press with a pressure of 120 MPa. The monolithic tablet and the middle section of the multi-layer tablet weighed 222 mg. The barrier-layered tablet also had two outer layers

Figure 1. (a) Diagram of ATR–FTIR flow cell containing monolithic tablet and a schematic of the data produced for a bi-layer tablet containing two different drugs. (b) USP I apparatus showing the tablet exposed to dissolution medium on all sides, as opposed to flow cell which is a sandwich structure. Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

DOI 10.1002/jps.23987

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Table 1.

3

Tablet Formulations Studied for Biphasic Release

Buflomedil side Nicotinamide side

Glucose (%) MCC (%) Glucose (%) MCC (%)

Tablet A

Tablet B

Tablet C

Tablet D

Tablet E

Tablet F

0 80 10 70

0 80 40 40

10 70 30 50

20 60 20 60

30 5 10 70

40 40 0 80

Concentrations of components in studied formulations are given in wt %.

of HPMC (100 mg each) compacted onto the flat faces of the middle section at 120 MPa. For comparison, two tablets with the same composition as the monolithic tablet above were dissolved using the ATR–FTIR imaging flow-through apparatus. These tablets were 3 mm in diameter and contained 7 mg of buflomedil. FTIR Spectroscopic Imaging The FTIR spectroscopic images were acquired from a continuously scanning FTIR spectrometer Equinox (Bruker, Ettlingen, Germany) connected to a macro-chamber with a 64 × 64 FPA infrared detector. FTIR spectra in the range of 3800– 1000 cm−1 were measured with 32 scans at 8 cm−1 spectral TM resolution. The Golden Gate diamond ATR accessory (Supercritical Fluid Analyser; Specac Ltd., UK), combined with the transparent dissolution flow-through cell made of Perspex was used in the macro-chamber.22,26 The imaging field of view using these optical arrangements was 690 × 610 :m2 . Dissolution Procedures Following compaction, the bilayer tablets were placed on the imaging crystal and the Perspex ATR flow-through dissolution cell was bolted into place on top. Tablets were positioned such that the ATR–FTIR image produced was of the interfacial area between the layers while also showing some of the dissolution medium as shown in Figure 1a. This allowed investigation of how the two drug release mechanisms proceeded simultaneously and their interactions with each other. Images of the dissolving tablets were acquired every 2.5 min during the experiments. The dissolution medium was deionised water and the flow rate was 1 mL/min. The larger tablets (10 mm diameter) were dissolved using a USP I dissolution test with rotation speed set to 50 rpm. Smaller tablets (3 mm diameter) were dissolved in the ATR flow cell at two flow rates: 1 and 9 mL/min.

RESULTS AND DISCUSSION Dissolution of Tablets A and B The bilayer tablets were designed to release two different readily soluble drugs, buflomedil pyridoxal phosphate and nicotinamide, although nicotinamide (691 mg/mL) is more soluble than buflomedil (69 mg/mL). The MCC was used to slow tablet dissolution because of its low solubility and good binding properties, whereas the glucose was used as a highly soluble agent, which expedited water ingress and caused the tablet to break up more rapidly. Each tablet contained buflomedil in one half of the tablet and nicotinamide in the other and as such the halves of the tablet will be referred to as either the nicotinamide or buflomedil sections. The wavenumber ranges for the characteristic bands used for generating the images showing the distribution of particDOI 10.1002/jps.23987

ular components of the tablets in the imaged area were as follows: buflomedil: 1166–1088 cm−1 , nicotinamide: 1412–1374 cm−1 , glucose: 922–903 cm−1 , MCC: 1026–1010 cm−1 , and water: 1665–1620 cm−1 . The data for Tablets A and B are presented in detail (highlighted in Table 1), as these formulations represent the extremes of the formulation range. Tablet B had a low loading of glucose in the buflomedil section of the tablet (0 wt %), whereas the nicotinamide section had a high loading of glucose. This formulation should highlight the effect of interference between the behaviour of the tablet sections, and FTIR imaging will be able to reveal how quickly the fast dissolving side disintegrates and whether this offers a new route of ingress for the dissolution medium into the slower dissolving side. Tablet A contained a relatively low loading of glucose in both sections of the tablet and so serves to compare against the performance of Tablet B. Figure 2a shows the dissolution of buflomedil, nicotinamide and glucose from the matrix of Tablet A. The data show that buflomedil was entirely located in one half of the tablet, whereas nicotinamide was clearly confined to the other half. It is also seen that glucose was distributed in the nicotinamide half of the tablet as would be expected by the formulation’s constitution. Nicotinamide dissolved slightly faster than buflomedil. In Figure 2b, the dissolution of the drugs and glucose from Tablet B’s matrix can be seen. This formulation had a high loading of glucose in the nicotinamide section of the tablet. It appears in the images as though there is less nicotinamide in the sample than buflomedil. As shown in Table 1, there was an equal loading of both drugs. The uneven distribution of nicotinamide was a result of the mechanical mixing of the formulation, of which this image displays a small fraction. It is immediately apparent from the data in Figure 2b that nicotinamide rapidly dissolved out of the matrix, as none of the drug remained in the imaging field of view after 5 min. This was much faster than buflomedil in the other section of the tablet, or nicotinamide in Tablet A. The buflomedil section of the tablet was not completely unaffected. Buflomedil in Tablet B only required 25 min (image not shown) to completely dissolve, whereas 40 min were required for the dissolution of buflomedil from Tablet A. These data indicated the high loading of glucose in Tablet B not only increased the dissolution rate of nicotinamide, but also increased the dissolution rate of buflomedil. To fully understand why this difference occurred, it is necessary to investigate the behaviour of the water and polymer components of Tablets A and B. The dissolution of polymer in Figure 3a shows that the MCC swelled slightly. It also reveals that a small channel, seen as the area of void in the MCC images, in the nicotinamide half of the matrix formed towards the end of the dissolution. If compared with glucose dissolution in Figure 2a, it can be seen that this region corresponds to a void created in the matrix by the dissolving glucose. The images in the second row of Figure 3a show Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

4

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Figure 2. Dissolution of buflomedil, nicotinamide and glucose from Tablets B and A. (a) Tablet A: buflomedil side—glucose: 0 wt %/MCC: 80 wt %, nicotinamide side—glucose: 10 wt %/MCC: 70 wt %. (b) Tablet B: buflomedil side—glucose: 0 wt %/MCC: 80 wt %, nicotinamide side—glucose: 40 wt %/MCC: 40 wt %. Image size is 690 × 610 :m2 .

that this void promoted the ingress of water into the tablet bulk, which would have accelerated the dissolution of nicotinamide. It is clear that in this tablet water ingressed into the more soluble nicotinamide half (containing the glucose) more rapidly than into the less soluble buflomedil section. Indeed, it appears that water ingress rates for each section were relatively decoupled. The dissolution performance seen in Figure 3b was significantly different to that seen in Figure 3a. Although the polymer did still undergo some expansion on the less soluble side of the tablet, it can be seen that by the end of the experiment, the polymer in the nicotinamide section of the sample had been severely eroded. The Buflomedil section of the tablet had also eroded more as compared with Tablet A. The greater level of disintegration was because of the greatly increased rate of water ingress. Water ingress progressed much more rapidly on the nicotinamide half of Tablet B. Faster water ingress was also seen in the buflomedil section relative to the buflomedil section of tablet A. In Tablet A, water ingress in each half of the tablet appeared quite independent, whereas in Tablet B, because of the longer period for which the tablet interface was wetted, there was a significant movement of water moved across interfacial line into the buflomedil section. Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

As the increase in glucose loading increased the rate of water ingress, which ultimately led to faster disintegration of the tablet, it is important to compare the water ingress rates for the two formulations in both tablet halves. Water ingress was measured by the wetted fraction of the matrix. This was calculated, in a semi-quantitative manner, based on the initial area of the tablet section before the initiation of dissolution and the remaining unwetted area at each time point. Wetted was defined as the point at which the intensity of the water band between 1665 and 1620 cm−1 reached 50% of the maximum value of the band in the lower half of the image in which there was no polymer. Results of this analysis are shown in Figure 4. Figure 4 clarifies the various rates of water ingress seen in Figure 3. The most significant difference between the formulations, as expected, was that the rate of water ingress into the nicotinamide section of the tablet was greatly increased in Tablet B over Tablet A, requiring approximately half the time for complete wetting. It is also seen from Figure 4 that the increase in the amount of glucose in the nicotinamide side of the tablet had a significant effect on the buflomedil dissolution. Indeed, the 40% increase in the rate of water ingress into the buflomedil half of the tablet from Tablet A to Tablet B was almost as much as that for the DOI 10.1002/jps.23987

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

5

Figure 3. Polymer dissolution and water ingress for Tablets B and A. (a) Tablet A: buflomedil side—glucose: 0 wt %/MCC: 80 wt %, nicotinamide side—glucose: 10 wt %/MCC: 70 wt %. (b) Tablet B: buflomedil side—glucose: 0 wt %/MCC: 80 wt %, nicotinamide side—glucose: 40 wt %/MCC: 40 wt %. Image size is 690 × 610 :m2 .

tablet tablet tablet tablet

time Figure 4. Comparison of wetted fraction for the buflomedil and nicotinamide sections of Tablets A and B. Data generated by extracting the wetted area as a function of time from Figures 6 and 7.

nicotinamide half. This was because as shown in Figure 3, the increased water ingress in the nicotinamide section of Tablet A almost doubled the wetted perimeter of the buflomedil section. These data have shown that the rate of the dissolution of the two drugs in these formulations can be controlled. Most significantly, they have shown the importance of gathering ATR–FTIR images with high chemical specificity and spatial resolution at the interfacial region of the tablets for the assessment of the dissolution of such formulations. Dissolution Performance Summary Figure 5 plots the time taken for total water ingress into a tablet section and the time taken for total dissolution of the DOI 10.1002/jps.23987

drug from that section for the relevant formulations in Table 1 (Tablets B–F). Complete water ingress was defined as the time at which the wetted fraction of that tablet half reached 1. Complete drug dissolution was defined as the point at which the maximum absorbance of the drug in the image was less than 20% of the maximum absorbance value in the dry image. Overall, nicotinamide dissolved significantly faster than buflomedil because of it being more soluble; hence, dissolution times for both drugs are plotted on separate axes. Tablets B and F had a 40 wt % loading of glucose in one tablet section and a 0 wt % loading in the other. Therefore, these formulations posed the greatest chance for interfacial interference, as one half dissolved rapidly potentially leaving the internal Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

6

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Figure 5. Summary of dissolution data for Tablets B–F shown in Table 1. The graph plots time required for the complete dissolution of the drug and time taken for the water ingress cover the entire visible area of the tablet half for both the buflomedil and nicotinamide sections of Tablets B–F. Two different axes are used because of the differences in time scales involved. Points are plotted as the mean value of three experiments; error bars generated from SD.

face of that tablet section exposed to the dissolution medium. Nonetheless, the dissolution rate of the drug was controlled across the range of formulations with the maximum release rate for each drug seen with the highest glucose loading and conversely the slowest release seen with the highest loading. Glucose had a smaller effect on the rate of nicotinmide dissolution because of its high intrinsic solubility. There was no significant change in the dissolution performance of nicotinamide until glucose loading reached 30 wt %. It is clear from Figure 5 that, as expected for two soluble drugs, the time taken for the drugs to dissolve and the time taken for the water to ingress across the whole of the imaged area for the tablets were linked, as both sets of curves are parallel. Indeed, for the nicotinamide, the curves are overlapped. This indicates that nicotinamide dissolved almost instantaneously (within the sampling rate of the images) upon contact with the dissolution medium. The match was less exact for the dissolution of buflomedil; there was a consistent lag of approximately 5 min between complete water ingress and complete dissolution of the drug. This is most likely because of the fact that buflomedil is significantly less soluble than nicotinamide and consequently required a greater concentration of dissolution medium and a longer exposure time to dissolve. The ATR–FTIR imaging data here demonstrate the applicability of this approach to studying these formulations. This work has shown that ATR–FTIR imaging can study the release of these two drugs from the tablet simultaneously. It has also shown that ATR–FTIR imaging has provided important information concerning the complex break up mechanisms of these structured formulations, such as cross-diffusion of the dissolution medium and identifying the dissolution front of the drug.

Table 2. Tests

Formulations Used for USP and FTIR Imaging Dissolution

Monolithic Tablet Constituents (wt %) Layered Tablet Core constituents (wt %) Barrier constituents (wt %)

Buflomedil

HPMC

MCC

45

35

20

45 –

35 100

20 –

The component loadings for the monolithic and barrier layer tablets are shown in Table 2. The monolithic tablet contained the soluble model drug buflomedil mixed with a swellable matrix of HPMC, a formulation which should have resulted in a tablet exhibiting a non-linear release profile for the drug,30 whereas the addition of swellable barriers to the outside of the layered tablet should have yielded a more linear release profile.10 The data for the monolithic tablet in Figure 6a show, as expected, a non-linear release profile, whereas the release profile for the barrier-layered tablet exhibited a significant shift towards constant release rate of the drug. These data are in agreement with the dissolution profiles seen in other studies using similar formulations.10,15 If the cell and diamond plating in the ATR flow-through cell can be considered as an approximation to barrier layers on the tablet, it may be expected that the release profile of these tablets in the ATR flow-through cell would tend towards linearity. Dissolution of Tablets in ATR Flow-Through Cell

Dissolution of Formulations in USP I Tri-layer tablets were used to explain how the constrained geometry of the flow cell may affect tablet dissolution for certain formulations relative to stirred tank systems. UV–Vis release profiles were used to compare the formulations. FTIR imaging was used to reveal the swelling and erosion behaviour of the tablets in the ATR flow cell to explain the effects of the flow cell on matrix dissolution. Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

The formulations tested for dissolution in the ATR flow cell had the following composition: 45 wt % buflomedil, 35 wt % HPMC, 20 wt % MCC and had a total mass of 15 mg. The dissolution in the ATR flow-through cell was performed at two different flow rates. A low flow rate of 1 mL/min still resulted in a non-linear dissolution profile; an increase in the flow rate to 9 mL/min caused the release rate to shift towards constant, as shown in Figure 6b. DOI 10.1002/jps.23987

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

7

Figure 6. (a) UV–Vis dissolution profiles for monolithic and barrier-layered tablets in USP I dissolution apparatus. (b) UV–Vis dissolution profiles for tablets dissolving in ATR flow cell at low flow rate (1 mL/min) and high flow rate (9 mL/min). The data in this graph were also fitted using the model in Eq. (1).

Modelling of Dissolution Profiles The data from these experiments were fitted using a model developed by Peppas and Sahlin31 which separated the Fickian and non-Fickian components of release, as shown in Eq. (1). Mt = k1 tm + k2 t2m. M∞

(1)

The term on the left hand side is the fraction of the drug released at time t and the first term on the right hand side describes the Fickian contribution to dissolution and the second term describes the relaxational (non-Fickian) contribution. It is a relatively simple model, but can give an indication as to the level of contribution from Fickian and non-Fickian release, via the magnitude of the k1 and k2 terms. The m term is the diffusion exponent and is dependent on the aspect ratio of the sample, with the value for cylindrical samples, such as the monolithic tablet dissolved in the USP I apparatus, being 0.45.31 The tri-layered tablets and those in the ATR flow cell had restricted geometries; therefore, the value for the cylinder was no longer appropriate. Indeed, these formulations were closer to slab geometries and so a value of 0.5 was used for m. The k values for the fitted curves in Figure 6 are shown in Table 3. Table 3 shows that the k1 value for the monolithic tablet dissolved in the USP I apparatus was the larger term, although, as this was a swelling and eroding tablet, there was a significant contribution from the polymer relaxation term. For the

Table 3.

Summary of k Values for Fitted Dissolution Curves k1

k2

USP Dissolution Monolithic fit Tri-layer fit

0.16 0.00024

0.094 0.0362

ATR–FTIR Flow Cell Low flow-rate fit High flow-rate fit

0.11 0.0045

0.044 0.11

DOI 10.1002/jps.23987

tri-layer tablet, results were markedly different, as would be expected for the data which so closely approximates a straight line, and the Fickian contribution was small compared with the non-Fickian term, indicating the dominance of the relaxational contribution. The modelling results from the ATR flow cell were similar to those from the USP dissolution apparatus; the low flow system produced a largely diffusion-dominated mechanism, whereas the high flow rate system produced a release profile in which release was controlled more by the relaxational term. Equation (2) was used to plot the fraction of release because of Fickian diffusion (F),31 the results of which are shown in Figure 7. F=

1 1+

k2 m t k1

(2)

The release of the drug from the non-zero-order formulations shown in Figure 7 relied significantly more on Fickian diffusion to release the drug. Over time, the contribution of Fickian diffusion did drop significantly during the dissolution of the tablets. Initially, the release was diffusion controlled as there had been little swelling. Once matrix swelling occurred, the regime of release shifted towards being swelling controlled as the drug was released from the relaxing polymer matrix,32 causing deviation from the standard square root of time kinetics.33,34 The mechanisms producing zero-order release for the barrier-layered tablet and the ATR flow-through cell at high flow rate were not identical, although the similar geometries played a role in producing similar release profiles. The barrier layers of the tri-layer tablet in the USP dissolution apparatus initially reduced the surface area for interaction of the drug and the dissolution medium and therefore the rate of water ingress into the core.5,10 The barrier layers restricted the swelling of the core to being primarily in the axial direction. Following this, as the polymers began to gel, swelling and erosion processes started to dominate. The core of the barrierlayered tablet swelled leading to an increase in the diffusion path length of the drug through the swollen polymer slowing drug release. This was partially limited by erosion of the gel Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

8

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Figure 7. Graph showing fraction of drug released by Fickian diffusion against total drug released. This graph illustrates the relative importance of the drug release mechanisms to the linear and Fickian release profiles. The fit of the model is only valid up to 60% release.

layer while the polymer barrier layers also began to erode, increasing the area of the core exposed to the dissolution medium. These two effects counterbalanced the increased diffusion path length, and as the effects were controlled and balanced correctly, it was possible to generate zero-order release.10 In the case of the ATR flow-through cell, although the flow cell and diamond plating acted as barrier layers, they were not erodible; therefore, the mechanism for maintaining zero-order release was slightly different. Although the barriers still had the same effects as the swellable barriers, of restricting water ingress and resulting in only axial swelling, they did not erode. Therefore, erosion of the radially swollen gel surrounding the core was more important, which is why zero-order release was only seen at higher flow rates when the rate of gel erosion was higher. This maintained the effective thickness of the gel layer through which the drug had to dissolve over time, resulting in a close to zero-order profile.35 It is for this reason that at a low flow rate, zero-order release was not achieved, as there was little erosion of the gel layer and so the diffusion path length became increasingly long for the drug.

To confirm this, it was necessary to use FTIR imaging to study how the swelling and erosion of the polymers changed between different flow rates. A comparison of the effects on polymer swelling can be seen in Figure 8. The increased flow rate, shown in the second row of images, is seen to erode the polymer faster than the lower flow rate, as total dissolution of the polymer from the field of view occurred about an hour earlier in the high flow-rate system. Dissolution of these formulations in the ATR–FTIR flowthrough cell does share some similarities with the dissolution profile features of barrier-layered tablets. Although the mechanisms responsible for the zero-order release were not identical, the similarities between the geometries of the two systems can be important. For this system and these formulations, at low flow rates, the level of Fickian diffusion involved in drug release was very similar to that of a monolithic tablet, whereas at high flow rates, the dissolution profile did shift towards being linear. The data presented do not serve as a rigorous comparison of the flow-through cell and the USP dissolution apparatus but do demonstrate the importance of the geometry of both systems and the flow rate of the dissolution medium.

f

f

Figure 8. Fourier transform infrared images of dissolution of tablets in ATR flow cell at low and high flow rates. The data show the initial 15 min of swelling and gel formation, and the last 3 h of final erosion of the swollen gel. Image size is 690 × 610 :m2 . Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

DOI 10.1002/jps.23987

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

CONCLUSIONS Attenuated total reflection–Fourier transform infrared spectroscopic imaging has been successfully applied to study the release of drugs from multi-layer formulations and has provided information which would be unavailable via other approaches. Bilayer tablets were applied to study biphasic release of drugs. It was found that ATR–FTIR imaging served as a useful approach for effectively studying the complex process of water ingress into structured formulations where the dissolution performance of one part of the tablet could have an effect on the performance of the other. In formulations with high loadings of glucose in one half of the tablet and low loadings in the other, the increased rate of water ingress into the more soluble section of the tablet caused it to dissolve rapidly. The high rate of dissolution increased the surface area of the less soluble section exposed to high concentrations of the dissolution medium. This led to an increased rate of drug dissolution being seen from the slow release section. Zero-order release formulations were explored as an analogue to the constricted geometry of the ATR–FTIR spectroscopic flow-through system. The barrier-layered formulations exhibited a similar geometry to the custom ATR flow-through cell. The dissolution cell and diamond plating above and below the tablet acted as barrier layers similar to those seen in zero-order release formulations. On the basis of the comparison of the levels of Fickian and non-Fickian release from the formulations studied here, it was found that in low flow-rate conditions, the dissolution performance of a tablet in the ATR flow cell possessed significant similarities to a monolithic tablet dissolving in a standard USP dissolution apparatus. The Fickian and non-Fickian contributions to the drug release mechanism were similar in extent. At higher flow rates, the dissolution performance did shift towards that of a barrier-layered zero-order release formulation because of an increase in the role played by polymer gel erosion. This mechanism was confirmed through the use of ATR–FTIR imaging to study polymer swelling and erosion in the cell. This work has shown that the geometry of the ATR–FTIR flow-through cell plays an important role when studying the dissolution of tablets with spectroscopic imaging. The data show that for the swellable formulations presented here at low flow rates, the Fickian component dominates in the release profile but that this is not the case at higher flow rates. This is an interesting observation which could facilitate further studies of these effects on other formulations in the future.

REFERENCES 1. Siepmann J, Kranz H, Bodmeier R, Peppas NA. 1999. HPMCmatrices for controlled drug delivery: A new model combining diffusion, swelling, and dissolution mechanisms and predicting the release kinetics. Pharm Res 16(11):1748–1756. 2. Siepmann J, Podual K, Sriwongjanya M, Peppas NA, Bodmeier R. 1999. A new model describing the swelling and drug release kinetics from hydroxypropyl methylcellulose tablets. J Pharm Sci 88(1):65–72. ¨ 3. Sirkia¨ T, Makimartti M, Liukko-Sipi S, Marvola M. 1994. Development and biopharmaceutical evaluations of a new press-coated prolonged-release salbutamol sulphate tablet in man. Eur J Pharm Sci 1(4):195–201. 4. Streubel A, Siepmann J, Peppas NA, Bodmeier R. 2000. Bimodal drug release achieved with multi-layer matrix tablets: Transport mechanisms and device design. J Control Release 69(3):455–468. DOI 10.1002/jps.23987

9

5. Conte U, Maggi L. 1996. Modulation of the dissolution profiles from GeomatrixR multi-layer matrix tablets containing drugs of different solubility. Biomaterials 17(9):889–896. 6. Uekama K, Matsubara K, Abe K, Horiuchi Y, Hirayama F, Suzuki N. 1990. Design and in vitro evaluation of slow-release dosage form of piretanide: Utility of $-cyclodextrin:cellulose derivative combination as a modified-release drug carrier. J Pharm Sci 79(3):244–248. 7. Maggi L, Machiste EO, Torre ML, Conte U. 1999. Formulation of biphasic release tablets containing slightly soluble drugs. Eur J Pharm Biopharm 48(1):37–42. 8. Chidambaram N, Porter W, Flood K, Qiu Y. 1998. Formulation and characterization of new layered diffusional matrices for zero-order sustained release. J Control Release 52(1–2):149–158. 9. Abdul S, Poddar SS. 2004. A flexible technology for modified release of drugs: Multi layered tablets. J Control Release 97(3):393–405. 10. Conte U, Maggi L, Colombo P, La Manna A. 1993. Multi-layered hydrophilic matrices as constant release devices (GeomatrixTM systems). J Control Release 26(1):39–47. 11. Qiu Y, Chidambaram N, Flood K. 1998. Design and evaluation of layered diffusional matrices for zero-order sustained-release. J Control Release 51(2–3):123–130. 12. Krishnaiah YSR, Karthikeyan RS, Gouri Sankar V, Satyanarayana V. 2002. Three-layer guar gum matrix tablet formulations for oral controlled delivery of highly soluble trimetazidine dihydrochloride. J Control Release 81(1–2):45–56. 13. Baveja SK, Ranga Rao KV, Padmalatha Devi K. 1987. Zero-order release hydrophilic matrix tablets of [beta]-adrenergic blockers. Int J Pharm 39(1–2):39–45. 14. Kim C. 1995. Compressed donut-shaped tablets with zero-order release kinetics. Pharm Res 12(7):1045–1048. 15. Colombo P, Conte U, Gazzaniga A, Maggi L, Sangalli ME, Peppas NA, La Manna A. 1990. Drug release modulation by physical restrictions of matrix swelling. Int J Pharm 63(1):43–48. 16. Lewis EN, Treado PJ, Reeder RC, Story GM, Dowrey AE, Marcott C, Levin IW. 1995. Fourier transform spectroscopic imaging using an infrared focal-plane array detector. Anal Chem 67(19):3377–3381. 17. Coutts-Lendon CA, Wright NA, Mieso EV, Koenig JL. 2003. The use of FT-IR imaging as an analytical tool for the characterization of drug delivery systems. J Control Release 93(3):223–248. 18. Kazarian SG, Chan KLA. 2003. “Chemical photography” of drug release. Macromolecules 36(26):9866–9872. 19. Kassis A, Bhawtankar VM, Sowa JR Jr. 2010. Attenuated total reflection infrared spectroscopy (ATR–IR) as an in situ technique for dissolution studies. J Pharm Anal 53(3):269–273. 20. van der Weerd J, Kazarian SG. 2004. Validation of macroscopic ATR–FTIR imaging to study dissolution of swelling pharmaceutical tablets. Appl Spectrosc 58(12):1413–1419. 21. Kazarian SG, Ewing AV. 2013. Applications of Fourier transform infrared spectroscopic imaging to tablet dissolution and drug release. Expert Opin Drug Deliv 10(9):1207–1221. 22. van der Weerd J, Chan KLA, Kazarian SG. 2004. An innovative design of compaction cell for in situ FT-IR imaging of tablet dissolution. Vibrat Spectrosc 35(1–2):9–13. 23. van der Weerd J, Kazarian SG. 2004. Combined approach of FTIR imaging and conventional dissolution tests applied to drug release. J Control Release 98(2):295–305. 24. Kazarian SG, Kong KWT, Bajomo M, van der Weerd J, Chan KLA. 2005. Spectroscopic imaging applied to drug release. Food Bioproducts Process 83(C2):127–135. 25. van der Weerd J, Kazarian SG. 2005. Release of poorly soluble drugs from HPMC tablets studied by FTIR imaging and flow-through dissolution tests. J Pharm Sci 94(9):2096–2109. 26. Kazarian SG, van der Weerd J. 2008. Simultaneous FTIR spectroscopic imaging and visible photography to monitor tablet dissolution and drug release. Pharm Res 25(4):853–860. 27. Achanta AS, Gray VA, Cecil TL, Grady LT. 1995. Evaluation of the performance of prednisone and salicylic acid USP dissolution calibrators. Drug Dev Ind Pharm 21(10):1171–1182. Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

10

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

28. Cohen JL, Hubert BB, Leeson LJ, Rhodes CT, Robinson JR, Roseman TJ, Shefter E. 1990. The development of USP dissolution and drug release standards. Pharm Res 7(10):983–987. 29. Wray PS, Clarke GS, Kazarian SG. 2013. Dissolution of tabletin-tablet formulations studied with ATR–FTIR spectroscopic imaging. Eur J Pharm Sci 48(4–5):748–757. 30. Lapidus H, Lordi NG. 1966. Some factors affecting the release of a water-soluble drug from a compressed hydrophilic matrix. J Pharm Sci 55(8):840–843. 31. Peppas NA, Sahlin JJ. 1989. A simple equation for the description of solute release. III. Coupling of diffusion and relaxation. Int J Pharm 57(2):169–172.

Wray et al., JOURNAL OF PHARMACEUTICAL SCIENCES

32. Siepmann J, Peppas NA. 2001. Modeling of drug release from delivery systems based on hydroxypropyl methylcellulose (HPMC). Adv Drug Deliv Rev 48(2–3):139–157. 33. Korsmeyer RW, Gurny R, Doelker E, Buri P, Peppas NA. 1983. Mechanisms of solute release from porous hydrophilic polymers. Int J Pharm 15(1):25–35. 34. Urtti A, Juslin M, Miinalainen O. 1985. Pilocarpine release from hydroxypropyl-cellulose-polyvinylpyrrolidone matrices. Int J Pharm 25(2):165–178. 35. M¨ockel JE, Lippold BC. 1993. Zero-order drug release from hydrocolloid matrices. Pharm Res 10(7):1066–1070.

DOI 10.1002/jps.23987