Noncontact Photo-Acoustic Defect Detection in Drug Tablets

Transcription

1 Noncontact Photo-Acoustic Defect Detection in Drug Tablets IVIN VARGHESE, CETIN CETINKAYA Department of Mechanical and Aeronautical Engineering, Center for Advanced Materials Processing, Wallace H. Coulter School of Engineering, Clarkson University, Potsdam, New York Received 16 February 2006; revised 18 September 2006; accepted 13 October 2006 Published online in Wiley InterScience ( DOI /jps ABSTRACT: Quality assurance monitoring is of great importance in the pharmaceutical industry for the reason that if defects such as coating layer irregularities, internal cracks, and delamination are present in a drug tablet, the desired dose delivery and bioavailability can be compromised. The U.S. Food and Drug Administration (FDA) established the Process Analytical Technology (PAT) initiative, in order to ensure efficient quality monitoring at each stage of the manufacturing process by the integration of analysis systems into the evaluation procedure. Improving consistency and predictability of tablet action by improving quality and uniformity of tablet coatings as well as ensuring core integrity is required. An ideal technique for quality monitoring would be noninvasive, nondestructive, have a short measurement time, intrinsically safe, and relatively inexpensive. In the proposed acoustic system, a pulsed laser is utilized to generate noncontact mechanical excitations and interferometric detection of transient vibrations of the drug tablets is employed for sensing. Two novel methods to excite vibrational modes in drug tablets are developed and employed: (i) a vibration plate excited by a pulsed-laser and (ii) pulsed laser-induced plasma generated shockwave expansion. Damage in coat and/or core of a tablet weakens its mechanical stiffness and, consequently, affects its acoustic response to an external dynamic force field. From the analysis of frequency spectra and the time frequency spectrograms obtained under both mechanisms, it can be concluded that defective tablets can be effectively differentiated from the defect-free ones and the proposed proof-of-concept techniques have potential to provide a technology platform to be used in the greater PAT effort. ß 2007 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 96: , 2007 Keywords: drug tablet monitoring; noncontact process monitoring; process monitoring; defective drug tablets; Process Analytical Technology (PAT) INTRODUCTION Physical properties and mechanical integrity of drug tablets often affect their therapeutic functions. Therefore, the monitoring for defects and the characterization of tablet mechanical properties are of great practical interest in the pharmaceuticals industry. The U.S. Food and Drug Administration (FDA) initiated the program Correspondence to: Cetin Cetinkaya (Telephone: ; Fax: ; cetin@clarkson.edu) Journal of Pharmaceutical Sciences, Vol. 96, (2007) ß 2007 Wiley-Liss, Inc. and the American Pharmacists Association Process Analytical Technology (PAT) for comprehensive quality assurance monitoring in the pharmaceuticals industry. PAT is often defined as a system for designing, analyzing, and controlling manufacturing through timely measurements (i.e., during processing) of critical quality and performance attributes of raw and in-process materials and processes with the goal of ensuring final pharmaceutical product quality. 1 Various methods have been introduced for process monitoring purposes. For example, noninvasive and passive techniques, such as acoustic emission (AE), have been widely utilized in the JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST

2 2126 VARGHESE AND CETINKAYA pharmaceutical industry for monitoring granular materials due to its cost effective and noninvasive properties. 2 By measuring AE it is possible to identify many of the phenomena that occur during powder compaction of pharmaceutical products such as granular rearrangement, fragmentation, visco-plastic deformation of grains or granules. 2,3 Acoustic relaxation emissions are detected and utilized during roller compaction of microcrystalline cellulose and maize starch. 4 Tablet coatings and core integrity play key roles in drug bioavailability. The mechanical integrity and uniformity of coating layers as well as cores is, consequently, of crucial concern for quality assurance. Tablet coatings serve a wide variety of purposes such as regulating the controlled release of active ingredients in the body, contributing to the bioavailability of a particular drug or combination of drugs, during certain times and locations within the body, protecting the stomach from high concentrations of active ingredients, extending the shelf life by protecting the ingredients from degradation by moisture and oxygen, and improving the tablet visual appeal. 5 If a coating layer is nonuniform and/or is with surface or subsurface defects, the desired dose delivery and bioavailability can be compromised. Some of the techniques presently used for analysis of tablet coatings are laser-induced breakdown spectroscopy, terahertz pulsed imaging, scanning thermal microscopy, atomic force microscopy, electron paramagnetic resonance, Fourier transform infrared spectroscopy, X-ray photoelectron spectroscopy, and confocal laser microscopy. 5 There are several types of defects that may occur in the manufacturing process of tablets such as deformations, cracks, excessive material, coating layer delamination, and nonuniform coating. Nondestructive testing and evaluation of drug tablets is, hence, essential in quality assurance process outlined within PAT. The objective of the current investigation was to develop a technology platform on which rapid, noncontact techniques for monitoring and evaluating drug tablets for mechanical defects such as coating layer irregularities, internal cracks, and delamination using a laser-acoustic approach 6,7 can be built for specific applications. In the presented testing platform, a pulsed laser is utilized as a source of acoustic fields to generate noncontact mechanical excitations and interferometric detection of transient vibrations of the drug tablets is employed for sensing. Two novel methods to excite vibration in drug tablets are developed and employed: (i) a vibration plate excited by a pulsed-laser and (ii) pulsed laser-induced plasma shockwaves. Nanometer-scale transient surface displacements of the drug tablets in the khz range frequencies are measured using the laser interferometer to detect stiffness change in a drug tablet due to coating and core irregularities and/or damage. Signal processing techniques are then applied to these transient displacement responses to differentiate the defective tablets from the defect-free ones. The potential of the presented platform in monitoring and evaluating drug tablets has been discussed. EXPERIMENTAL PROCEDURE In the current experiments, 200 mg tablets (P-tablets) with the characteristic dimensions of 6.01 mm width, mm length, and 3.84 mm thickness were utilized (Fig. 1). The thickness of the tablet coat is approximately 102 mm. Three types of tablets have been considered and classified in this study: A denotes a defect-free tablet, defect type C refers to tablets with cracks (Fig. 1a), and defect type H denotes tablets with a shallow circular hole with a diameter of approximately 500 mm drilled on one side (Fig. 1b). The depth of the holes are approximately equal to the thickness of the tablet coating layers (e.g., 102 mm). The main components in the laser acoustic experimental setup consist of an Nd:YAG pulsedlaser, a high energy plano-convex lens, the sample tablet, a laser interferometer (Polytec OFV 511), and a digital phosphor oscilloscope (Tektronix TDS 3052) (Figs. 2a and 3a). The interferometer is operated in the out-of-plane displacement mode in which its bandwidth is 50 khz 30 MHz for the amplitude range of 75 nm. Measurements below 50 khz in displacement mode are not linearly decoded. A Q-switched Nd:YAG laser (Quantel Brilliant series Q44) was employed with following characteristics: the fundamental wavelength of 1064 nm with the approximate pulse energy of 370 mj, the pulse duration of 5 ns, the repetition rate of 10 Hz and a beam diameter of 5 mm. The beam was focused by a 25 mm diameter, 100 mm focal length plano-convex lens with an antireflective coating at 1064 nm. Direct Pulsed-Laser Irradiation of a Vibration Plate In this excitation mechanism, a metal plate thermo-elastically excited by a pulsed laser beam JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007 DOI /jps

3 NONCONTACT PHOTO-ACOUSTIC DEFECT DETECTION 2127 Figure 1. a: Photographs of different tablet samples A denoting defect-free tablet and C denoting a cracked tablet (tablet defect type C ), and (b) the tablet samples with defect type H (small hole drilled at the back). The diameter of the hole is approximately 500 mm. without focusing is employed as an AE source. In the experimental set-up, the vibration plate (a cylindrical aluminum stud with 5.58 mm thickness and 8.14 mm diameter) is positioned in front of the Nd:YAG laser in such a manner that the laser pulse strikes the center of the vibration plate. The schematic instrumentation diagram for the experimental setup is depicted in Figure 2a. Due to the rapid thermal expansion created by the transient thermal field (laser pulse) deposited on the surface of the plate, the vibration plate oscillates at high frequencies and emits acoustic fields into its near proximity. In experiments, the pulsed-laser is operated in single-shot mode to avoid overheating of the vibration plate. Yet a sufficiently high number of shots were fired to ensure the repeatability of the tablet response. Shot-to-shot pulse energy stability of the laser was tested and found to be consistent within the time frame of the experiments. The sample tablet is fixed behind the vibration plate at a distance of approximately 22 mm, which provides sufficient vibrational amplitudes in the tablet (Fig. 2b and c). The dimensions of the vibration plate were chosen to restrict low frequency excitation components. The laser interferometer is positioned behind the sample tablet and the measuring laser beam is focused at the center of the tablet surface, as depicted in Figure 2b. The exact location of the interferometer laser spot on the tablet is less of a concern since the method under investigation is based on the modal properties of the tablet. The distance between the interferometer probe and the tablet is approximately 13.5 cm, which is Figure 2. a: Instrumentation diagram of the vibration plate excitation-based, (b) schematic of the laser excitation/detection mechanism, and (c) the photograph of the experimental set-up. DOI /jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007

4 2128 VARGHESE AND CETINKAYA determined by maximizing the detection sensitivity of the interferometer. The experiments are conducted on an optical bench to eliminate the influence of environmental vibrations. Various tests such as blank laser shots and acoustic blockage of the shock waves were conducted to overrule the influence of the environmental effects and the potential electrical cross-talk that might adversely affect the accuracy of the measurement system. A set of four defect-free tablets ( A ) and four defective tablets (defect type C ) were tested. Laser Induced Plasma (LIP) Based Tablet Excitation In this excitation mechanism, a shockwave field created by laser-induced plasma is utilized. In the experiments reported in the current study, the incident laser pulse is focused by a plano-convex lens; as a result of high power concentration near the focal point of the lens, an expanding plasma core is created by dielectric breakdown of air. This plasma expansion results in a strong shock wavefront generating a transient pressure field in the air. The wavefront propagates and exerts a transient pressure field on the sample tablet. The tablet vibrates at its various resonance frequencies as a result of its interactions with the shock wavefront. The laser interferometer captures the transient surface displacement response of the tablet. A set of six defect-free tablets ( A ) and six defective tablets (defect type H ) were tested in evaluating the effectiveness and potential of this excitation mechanism. The schematic instrumentation diagram for the experimental set-up for the experiments reported in this study is depicted in Figure 3a. The test tablet is fixed at a distance of approximately 30 mm away from the LIP core, perpendicular to the laser beam. At this distance the plasma core never achieves contact with the drug tablet since the expansion of the plasma core is constrained by a specific diameter. The laser interferometer is positioned behind the tablet and its interferometer beam is focused at the center of the tablet surface (Fig. 3b). EXPERIMENTAL RESULTS Vibration Plate Experiments The measured transient surface displacements for the four defect-free tablets indicate that the Figure 3. a: Instrumentation diagram of the excitation mechanism based on the pulsed LIP shockwave, and (b) the photograph of the experimental set-up. responses of the tablets under the same excitation conditions are quite consistent (Fig. 4a). Multiple measurements for each tablet were conducted to ensure the stability of the measurement system. This study and associated observation established repeatability and stability of the excitation mechanism and the optical measurement system. The defective tablets (defect type C ) are observed to have considerably higher transient surface displacement responses compared to the defect-free tablets (Fig. 4a). As depicted in Figure 4b, the locations of the resonance frequencies for the defect-free tablets used in the experiments are quite consistent while the resonance frequency locations for the defective tablets are widely varied from each other. The fast Fourier transform algorithm with appropriate windows from the transient JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007 DOI /jps

5 NONCONTACT PHOTO-ACOUSTIC DEFECT DETECTION 2129 Figure 4. a: Transient surface displacement responses excited by the vibration plate mechanism of the four defect-free ( A ) tablets (solid lines) and defective (defect type C ) four tablets (dashed lines) and (b) their frequency spectra. responses reported in Figure 4a is used for determining the frequency response functions of the tablets. The interferometer is operated in the out-of-plane displacement mode in which its bandwidth is 50 khz 30 MHz for the amplitude range of 75 nm. In addition, the frequency spectra of the defective tablets exhibited higher magnitudes than the defect-free tablets. Higher amplitudes are expected since the mechanical integrity of the defective tablets is compromised and their stiffness (rigidity) is lower than that of the defect-free tablet. In analyzing nonstationary acoustic signals, time frequency analysis is an effective computational tool. The time frequency spectra of the signals provide temporal and spectral resolutions, simultaneously. Wavelet-based time frequency method is employed in this study. From the time frequency spectrograms, it is observed that the defect-free tablets (Fig. 5a d) display very similar responses, whereas each defective tablet (Fig. 5e h) clearly has different distinctive responses due to the tablet-to-tablet defect variations within the tablet group. The time frequency spectrogram of tablet C3 having very small defects were found to be quite similar to the defect-free tablet responses within the first wave portion from 100 to 130 ms (Fig. 5g). However, it substantially DOI /jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007

6 2130 VARGHESE AND CETINKAYA Figure 5. Time frequency spectrograms of the vibration plate-excited defect-free tablets (a d), and the tablets with defect type C (e h). differentiates from the response of a defect-free tablet after 130 ms. The responses of tablets with more severe defects exhibit large deviations from the defect-free ones (Fig. 5). LIP Experiments The transient responses of tablets under the LIP generated shockwave excitation are depicted in Figure 6a. The defect-free and defective tablets (defect type H ) are observed to have visibly similar responses in the time-domain. This is because of the fact that the defect type H is merely a small shallow circular hole (diameter of approximately 500 mm) drilled on one side of a defect-free tablet. Clearly it is difficult to distinguish these small defects from the raw waveforms. From the frequency spectra of the responses (Fig. 6b) it is observed that the locations of the resonance frequencies for Figure 6. a: Transient surface displacement responses excited by the LIP shockwaves mechanism of the six defect-free ( A ) tablets (solid lines) and defective (defect type H ) six tablets (dashed lines) and (b) their frequency spectra. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007 DOI /jps

7 NONCONTACT PHOTO-ACOUSTIC DEFECT DETECTION 2131 Figure 6. (Continued) various defect-free tablets are quite consistent. However, in the case of the defective tablets, the resonance frequency locations visibly vary (Fig. 6b). Furthermore, the frequency spectra of the defective tablets exhibit higher magnitudes than the defect-free tablets. Cross-correlation coefficients for the transient displacement responses, between the response of a defect-free tablet (reference) and each individual defect-free (type A ) and defective tablet (type H ), are calculated. The normalized peak values of the cross-correlation coefficient, which measures the degree of similarity of a tablet response to the reference waveform, are summarized in Figure 7. It is observed that the defect-free tablet responses clearly exhibit higher degree of similarity to the reference defect-free tablet compared to the defective tablets. Moreover, from the time frequency spectrograms of the defect-free and defective tablets, it is observed that the defect-free tablets (Fig. 8a f) display similar responses, whereas the defective tablets (Fig. 8g l) have responses different from each other and those for the defect-free tablets. are effective in drug tablet defect detection by sensing the change in mechanical stiffness of a tablet due to irregularities and/or damage and the presented platform has potential in tablet integrity monitoring. In both mechanisms acoustic waves are imparted on the test tablets and their transient surface displacement responses are measured by a laser interferometer. In the first mechanism the acoustic waves are generated due to the vibrations of a metal plate thermally excited by direct pulsed-laser irradiation, whereas in the second mechanism, shockwaves generated due to the pulsed LIP are utilized as an CONCLUSIONS AND REMARKS Two pulsed-laser techniques for acoustic excitation of drug tablets have been proposed as noncontact excitation mechanisms in a monitoring platform and the results of a set of evaluative tablet defect detection experiments are reported. It is demonstrated that noncontact detection techniques based on these two types of excitation Figure 7. Time amplitude correlation coefficients between the transient responses of the defect-free tablets and tablets with defect type H excited by the LIP shockwaves. The horizontal dashed lines indicate the group averages. DOI /jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007

8 2132 VARGHESE AND CETINKAYA Figure 8. Time frequency spectrograms of the LIP shockwaves excited defect-free tablets (a f) and the tablets with defect type H (g l). acoustic excitation source. From the analysis of frequency spectra and the time frequency spectrograms obtained under both mechanisms it can be concluded that defective tablets can be effectively differentiated from the defect-free ones. From the vibration plate experimental results, it is possible to distinguish the degree of defect(s) in a tablet. From the frequency spectrum of the transient responses under LIP excitation, it is found that it is possible to distinguish even tablets with small shallow circular holes in tablet coats with diameters of approximately 500 mm (defect type H ). The correlation data presented depicts the degree of similarity of defect type H tablets with a defect-free tablet ( A ), thus the level of defect can be predicted. The main use of the correlation data is that even very small defective levels (defects that are undetectable in time-domain acoustic waveforms) can be detected in tablets. It is demonstrated that a small shallow hole (defect H ) can be detected using the LIP shockwave excitation mechanism. The JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007 DOI /jps

9 NONCONTACT PHOTO-ACOUSTIC DEFECT DETECTION 2133 amplitudes of vibrational motion of the tablets excited by the two methods are comparable, therefore, as observed, their performances are comparable. It is experimentally established that defect-free tablets exhibit quite consistent acoustic transient responses and, therefore, defective tablets having different results can be easily detected using either excitation mechanisms. The experimental results clearly indicates that the proposed excitation and sensing approaches utilizing either the vibration plate mechanism or the LIP mechanism are effective novel methods that can be utilized in a drug tablet quality assurance system. The technology platform presented in this study has the potential to be adopted in various tablet monitoring PAT applications and tools. The objective of the current investigation was to demonstrate that noncontact acoustic excitation and detection of a drug tablet is possible and to provide evidence that mechanical irregularities and/or defects on tablets leading to a change in structural stiffness can be detected without concerning the extent and nature of these defects. Characterizing defect and studying the nature of defects are the subjects of our ongoing research activities. ACKNOWLEDGMENTS The authors thank Dr. Dominic A. Ventura and Wyeth Pharmaceuticals for stimulating discussions and financial support as well as Center for Advanced Materials Processing (CAMP) at Clarkson University for partial funding. REFERENCES 1. Hussain AS, Watts C, Afnan AM, Wu H Foreword. J Process Anal Technol 1:3. 2. Whitaker M, Baker GR, Westrup J, Goulding PA, Rudd DR, Belchamber RM, Collins MP Application of acoustic emission to the monitoring and end point determination of a high shear granulation process. Int J Pharma 205: Serris E, Camby LP, Thomas G, Desfontaines M, Fantozzi G Acoustic emission of pharmaceutical powders during compaction. Powder Technol 128: Salonen J, Salmi K, Hakanen A, Laine E, Linsaari K Monitoring the acoustic activity of a pharmaceutical powder during roller compaction. Int J Pharm 153: Fitzgerald AJ, Cole BE, Taday PF Nondestructive analysis of tablet coating thicknesses using terahetz pulsed imaging. J Pharm Sci 94: Varghese I Laser acoustic techniques for drug tablet monitoring. (M.S. Thesis) Potsdam, New York, USA: Clarkson University. 7. Ban L Acoustic monitoring of industrial processes using signal processing methodologies. (Ph.D. Thesis) Potsdam, New York, USA: Clarkson University. DOI /jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 8, AUGUST 2007