- Email: [email protected]
The Influence of Relative Humidity on the Cohesion Properties of Micronized Drugs Used in Inhalation Therapy
The Influence of Relative Humidity on the Cohesion Properties of Micronized Drugs Used in Inhalation Therapy PAUL M. YOUNG,1 ROBERT PRICE,1 MICHAEL J. TOBYN,1 MARK BUTTRUM,2 FIONA DEY2 1
Pharmaceutical Technology Research Group, Department of Pharmacy and Pharmacology, University of Bath, Claverton Down, Bath, BA2 7AY, UK 2
Respiratory Technology, Aventis Pharma, Holmes Chapel, UK, CW4 8BE
Received 12 June 2003; revised 4 August 2003; accepted 4 August 2003
ABSTRACT: The influence of relative humidity (RH) on the cohesion properties of three drugs: salbutamol sulphate (SS), triamcinolone acetonide (TAA), and disodium cromoglycate (DSCG) was investigated using the atomic force microscope (AFM) colloidal probe technique. Micronized drug particles were mounted in heat-sensitive epoxy resin for immobilization. Multiple AFM force–distance curves were conducted between each drug probe and the immobilized drug particulates at 15, 45, and 75% RH using Force– Volume imaging. Clear variations in the cohesion profile with respect to RH were observed for all three micronized drugs. The calculated force and energy of cohesion to separate either micronized SS or DSCG increased as humidity was raised from 15 to 75% RH, suggesting capillary forces become a dominating factor at elevated RH. In comparison, the separation force and energy for micronized TAA particles decreased with increased RH. This behavior may be attributed to long-range attractive electrostatic interactions, which were observed in the approach cycle of the AFM force–distance curves. These observations correlated well with previous aerosolization studies of the three micronized drugs. ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 93:753–761, 2004
Keywords: AFM; separation energy/force; relative humidity; cohesion; individual particle–particle measurements
INTRODUCTION Particle–particle interactions in pharmaceutical systems inevitably dominate the performance and efficacy of a solid dosage form. For dry powder inhalation systems, which aim to deliver sub-5 micron particulates to the respiratory tract, the magnitude of these adhesion or cohesion forces are critical for both the interactive formulation and drug liberation during patient inspiration. Drug particles for inhalation therapy are commonly produced by size reduction of larger crystal-
Correspondence to: Paul M. Young (Telephone: þ44 01225 383103; Fax: þ 44 01225 386114; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 93, 753–761 (2004) ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association
line material using high-energy comminution processes such as micronization. Consequently, the resultant micron sized (<5 mm) particles have irregular surface morphologies with a high surface area. When combining such physical factors with the intrinsic chemical and crystal structure, the variation in adhesion or cohesion between individual particles becomes very complex. It is generally considered that three interparticulate forces are prevalent in dry powder inhalation systems: van der Waals, capillary, and electrostatic forces. The degree to which each force dominates will be dependent on the physical, mechanical, and chemical properties of the contiguous surfaces and the environment to which they are exposed. Recently, with advances in inhalation drug design, and the potential of systemic delivery, a fundamental understanding of such particle–
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
753
754
YOUNG ET AL.
particle interactions has become imperative. Because theoretical calculation of adhesion or cohesion forces in pharmaceutical systems are very complex (due to the stratifying variables), the simplest approach would be direct measurement of particle interactions. Indeed, such investigations are commonplace, and can be found in abundance in the literature.1–4 Generally, such measurements involve the removal of small quantities of powder from model surfaces using techniques such as centrifugation1–3 or reentrainment.4 However, such techniques may have limitations. For example, when considering centrifugation, the maximum theoretical removal force will approach or become smaller than the theoretical adhesion force. In addition, the determination of individual particle adhesion forces becomes very difficult, as such measurements are mainly dependent on the removal of a large quantity of particulates or agglomerates. A possible method for overcoming such issues is the use of an atomic force microscope.5 In simple terms, by mounting a particle on the apex of a microfabricated cantilever (with known spring constant), it becomes possible to construct a ‘‘probe,’’ which can be used to measure the forces of interaction.6 Furthermore, because the cantilever deflection, which is dependent on the acting force, is measured as a function of probe–substrate separation, it becomes possible to investigate influence of short-range attractive (or repulsive) forces.7 Previous investigations (applicable to DPI technology), using the AFM colloid probe technique include determination of lactose–silica,8 lactose– gelatin,9 and lactose–lactose10 interactions. However, little work has been conducted using drug probes of micron size (<5 mm), applicable to particles used in inhalation. The most relevant published work has investigated the influence of inhaler components11 and carrier morphology12 on drug adhesion. More recently, the influence of RH on drug carrier adhesion13 and drug cohesion14 has demonstrated the colloid probe technique a powerful tool in predicting in vitro behavior. Nonetheless, previous studies have mainly focused on measuring the interaction between micronized drug probes and relatively large planar surfaces (such as drug compacts or carrier surfaces). As part of an ongoing study, investigating the influence of RH on drug cohesion, three drugs were chosen for their formulation characteristics and analyzed using an individual particle approach. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
Three micronized drugs of comparable size were chosen for investigation: salbutamol sulphate (SS), triamcinolone acetonide (TAA), and disodium cromoglycate (DSCG). Previous in vitro investigations have shown RH (over the range 11 to 75% RH) to have a significant effect on the aerosolization properties of the micronized drugs.15 Because such in vitro studies can prove time consuming, the AFM colloid probe technique was employed for comparative study, to assess its potential as a rapid formulation screening technique.
MATERIALS AND METHODS Materials Micronized salbutamol sulphate (SS), triamcinolone acetonide (TAA), and disodium cromoglycate (DSCG) (all Aventis Pharma, Holmes Chapel, UK) were used as supplied. Physical Characterization Moisture sorption profiles for SS, TAA, and DSCG were determined using dynamic vapor sorption (DVS) (DVS-1, Surface Measurement Systems Ltd, London, UK). Approximately 10 mg of drug was weighed into the sample cell and subjected to two 0–90% RH cycles at 258C (10% RH steps). Equilibration at each RH was determined by a dm/dt of 0.002. The influence of RH on the surface morphology of the micronized drugs was investigated usingAFM in Tapping-ModeTM (Multimode AFM, Nanoscope IIIa controller, Digital Instruments, UK). Briefly, samples of the micronized drug were lightly brushed through a 45-mm aperture stainless steel sieve (BS410/1986, Endecotts, London, UK) onto a piece of freshly cleaved, grade 1, muscovite mica (SPI supplies, Pennsylvania). Samples were imaged with a high-aspect ratio silicon probe (OTESP, Digital Instruments, UK) at a scan rate of 0.7 Hz. Atomic force microscope images of the same region were taken at 15, 45, and 75% RH (at 258C) using a custom-built perfusion apparatus described elsewhere.13,14 Mounting of Micronized Drug Particles on Epoxy Resin Micronized particle samples were immobilized on steel AFM stubs with Tempfix (SPI Supplies) at 408C using a custom-built Peltier connected to a thermocouple temperature controller (SE5000,
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
Marlow Industries, Texas). Briefly, a steel AFM sample stub was mounted on the Peltier and heated to 1208C. A small quantity of Tempfix was melted onto the stub and cooled to 408C, at which point the resin becomes tacky or adhesive. A small quantity of the powder was lightly brushed through a 45 mm aperture stainless steel sieve (BS410/1986, Endecotts, London, UK) onto the AFM sample stub. The sieving procedure was found to remove most agglomerates. The AFM sample stubs, containing Tempfix-mounted particles, were left to cool to room temperature before removing loose material with a filtered nitrogen air stream. The mounted drug samples were stored in tightly sealed containers for 24 h prior to use.
755
RESULTS AND DISCUSSION Physical Characterization To fully quantify the effect of RH on cohesive interactions, the three drugs were first characterized for moisture sorption and morphology. Dynamic vapor sorption isotherms for SS, TAA, and DSCG are shown in Figure 1a,b, and c, respectively. The water sorption isotherms for SS
Atomic Force Microscope Colloid Drug Probe Measurements All AFM investigations were conducted at specific environmental conditions using a custom-built perfusion system described elsewhere.13,14 Micronized particles of each of the three drugs were mounted onto the apex of nominal 0.58 Nm1 tipless AFM cantilevers using a micromanipulation process detailed elsewhere.14 Multiple force– distance curves were conducted between each drug probe and the corresponding Tempfix mounted particle substrates using Force–Volume (FV) imaging,16 over a 25 25-mm area. This provided a block of 64 64 force–distance curves corresponding to the relative X-Y position in the Force–Volume scan. In addition, by recording the relative piezo height at the constant compliance limit, a low-resolution ‘‘relief image’’ of the FV region could be produced.16 Unlike previous investigations, which utilized batch-conversion software to process FV data,12–14 force–distance curves were analyzed using an individual approach. Force–distance curves, conducted on individual micronized drug particles, were exported and manually analyzed/integrated to produce both separation force and separation energy values. Approximately 30 randomly selected force–distance curves across six micronized particles were analyzed for each drug (at each humidity). For comparative purposes, drug probe–Tempfix interactions were also analyzed. Cohesion measurements for each drug were conducted over the same FV region at 15, 45, and 75% RH (at 258C). A stabilization period of at least 4 h was instigated with each incremental increase in RH to obtain dynamic equilibrium.
Figure 1. Moisture sorption isotherms for (a) SS, (b) TAA, and (c) DSCG. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
756
YOUNG ET AL.
(shown in Fig. 1a) followed a sigmoidal class L-3 curve,17 suggesting Langmuir-type monomulti layer water sorption onto the crystal surface. In comparison, analysis of the moisture sorption curves for TAA and DSCG suggested non-Langmuir type isotherms. The relative percent moisture sorption for TAA was approximately one order of magnitude greater than that of salbutamol sulphate, suggesting TAA had a higher affinity for water. Thus, it is proposed that the moisture sorption curve for TAA may be a class H curve.17 Class H curves are essentially a ‘‘high-affinity’’ form of the L type curve in which the adsorbate may be chemisorbed. Moisture sorption isotherms for DSCG (Fig. 1c) suggest a class C sorption curve.17 Class C curves are indicative of porous materials with a high affinity for the adsorbate and are associated with absorption into the crystal structure. This is to be expected, however, as DSCG is essentially a solid solution, whose crystal structure contains channels that allow the up to nine molecules of water per molecule of DSCG (depending on the RH), before collapsing into a series of liquid crystal mesomorphs.18 Dynamic vapor sorption isotherms for the three micronized materials showed no characteristic hysterisis relating to the presence of amorphous material. Representative AFM images of SS and TAA at 45% RH are shown in Figure 2a and b, respectively. It can be seen from Figure 2 that SS exhibits a columnar shape (Fig. 2a), while TAA (Fig. 2b) appears to have a more flattened platelet morphology. Similar low-magnification AFM images of DSCG (not shown here) suggested a columnar shape. In general, the low-magnification AFM images agreed well with previously reported studies of the micronized drugs using conventional imaging methods such as optical microscopy and scanning electron microscopy.15 However, it is important to note that limited information was available in the literature regarding the general morphology of TAA. Both SS and TAA showed no macroscopic changes in particle morphology over the humidity range 15–75% RH. In comparison, the morphology of DSCG clearly changed as humidity was increased from 15 to 75% RH. suggesting sorbed moisture to integrally influence particle morphology. Representative high-resolution AFM images of a DSCG particle after exposure to 15, 45, and 75% RH (for approximately 4 h at each humidity) are shown in Figure 3a,b, and c, respectively. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
Figure 2. Atomic force microscope Tapping mode images of (A) SS and (B) TAA.
Variations in particle morphology of the micronized DSCG with respect to RH were expected, as DSCG is effectively a nematic choromonic liquid crystal that below 93% RH forms a solid solution.18–20 As previously discussed, the planar cromoglycate molecules are ‘‘stacked’’ to form long rods that, in turn, form an interstitial solid held together by water molecules. These water channels can subsequently contain up to approximately nine molecules of water per unit cell before collapsing to form a liquid crystal mesophase (at 93% RH). Adsorption or desorption of water molecules into the crystal structure will subsequently cause changes in the torsional angles of the cromoglycate molecules,20 resulting in a variation in the unit cell dimensions. Consequently, it is likely that such a change in crystal structure with respect to RH would have a significant impact on the macroscopic particle morphology.
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
757
Figure 4. Atomic force microscope Tapping mode height image of a 25 25 mm area of the salbutamol sulphate particles mounted on the Tempfix polymer surface.
Figure 3. High-resolution AFM Tapping mode images of DSCG taken at 15, 45, and 75% RH (258C).
Atomic Force Microscope Colloid Drug Probe Measurements The influence of RH on drug particle cohesion was investigated by conducting FV scans (25 25 mm) between a drug probe and corresponding drug particles mounted on Tempfix polymer resin. A representative AFM tapping mode ‘‘height’’ image of the mounted SS particles is shown in
Figure 4. The salbutamol sulphate drug particulates had an irregular morphology, and were clearly proud of the Tempfix polymer surface. In addition, AFM tapping mode images on randomly chosen areas of a ‘‘control’’ Tempfix surface (not shown here), prepared without drug particle mounting, showed no such exposed irregularities. Atomic force microscope height images for TAA and DSCG mounted in Tempfix produced similar results. Essentially, such variations in height allow the identification of individual mounted drug particles and enable subsequent cohesion measurement. An example of a Force–Volume scan, conducted between an SS drug probe and SS mounted particles, is shown in Figure 5. The white areas on the FV ‘‘relief’’ height image correspond to regions with a Z-axis height of greater than 2.5 mm. It would be reasonable to assume that such undulations in the surface could be identified as micronized drug particles, as they correspond to observations made using conventional AFM tips an imaging methods (Fig. 4). Comparison of the ‘‘relief’’ height image with the FV image (Fig. 5) suggest the raised ‘‘drug particle’’ areas have a low adhesion force (dark areas on FV image) in comparison to the polymer surface. A representative force curve for the SS interactions is included in the FV scan (Fig. 5) with its relative position marked with a cross in the ‘‘relief’’ height and FV images, respectively. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
758
YOUNG ET AL.
Figure 5. Force–volume data for salbutamol sulphate drug probe interactions with salbutamol sulphate particles mounted in Tempfix. Lighter areas represent elevated topography and lower adhesion forces in the ‘‘relief’’ height image and FV image panes, respectively.
To quantify such drug probe interactions, randomly chosen force curves taken across five distinct ‘‘raised particle’’ regions were exported and manually integrated to obtain both cohesion force and separation energy values. It is important to note that although an individual analysis approach does not allow such comprehensive statistical analysis as batch reprocessing, it was deemed necessary to ensure differentiation between micronized particle and Tempfix polymer interactions. A detailed summary of the SS, TAA, and DSCG cohesive parameters is given in Table 1. ANOVA
one-way analysis of variance indicated RH to have a significant effect (p < 0.05) on the cohesion force of all three micronized drugs. However, such an observation is effectively dependent on a normal distribution. Further analysis of the cohesion force data (Table 1) suggested this not to be the case, as the large standard deviations and positive skewness values indicate a non-normal distribution. Analysis of the cohesion force values indicated a log-normal distribution to best fit the data. Furthermore, this agreed with previous reports using multiple colloid drug probe measurements on irregular surfaces.14–16
Table 1. Summary of Cohesion Force Measurements, Conducted between Individual Drug Probes and Six Corresponding Micronized Drug Particles Mounted in Tempfix Resin (n ¼ 30)
Drug SS
TAA
DSCG
a
Humidity (% RH)
Mean Cohesion Force (nN)
15 45 75 15 45 75 15 45 75a
39.941 54.795 65.731 65.088 36.491 27.339 66.292 116.491 —
St Dev
Median (f0.5) (nN)
10th Percentile (f0.1) (nN)
90th Percentile (f0.9) (nN)
GSD
Skewness
24.155 28.144 30.910 21.421 18.297 19.871 45.269 60.086 —
32.887 48.531 58.747 61.154 30.599 19.903 55.041 98.431 —
14.198 25.767 31.417 37.923 12.821 6.090 25.749 43.662 —
76.174 91.405 109.853 100.818 73.027 65.045 117.656 221.901 —
1.924 1.638 1.629 1.451 1.961 2.516 1.808 1.884 —
0.748 1.021 0.725 1.100 0.756 1.066 1.604 0.221 —
DSCG drug probe was ‘‘lost’’ during measurements at 75% RH (repeated with two further probes, refer to text).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
In general, the cohesion force values for salbutamol sulphate agreed with previously published data using both the AFM colloid probe technique.13,14 To the authors’ knowledge, no literature relating to fundamental cohesion measurements on DSCG or TAA exist. Analysis of lognormal probability plots for SS, TAA, and DSCG suggested to have a significant effect (CI ¼ 0.95) on the cohesion force. The median cohesion force (f0.5) for SS increased with respect to RH (32.9, 48.5, and 58.7 nN for 15, 45, and 75% RH, respectively), suggesting capillary interactions to become more dominant at higher relative humidities. Such observations correlate well with previous in vitro investigations, which demonstrated a decrease in SS fine-particle aerosolization as RH was increased.15 Comparison of the both the mean and median SS cohesion values with Tempfix adhesion measurements suggested the polymer drug adhesion to be significantly greater than drug cohesion. All adhesion measurements conducted between the ‘‘flat’’ polymer regions of the FV scan were of a higher value than that of SS cohesion (142.1, 152.6, and 164, at 15, 45, and 75% RH, respectively). Such observations are to be expected, however, as adhesive resins have high surface energies due to the polar nature of groups present on the polymer chains.21 In contrast to the SS cohesion profile, the median separation force for TAA decreased as RH was increased (61.2, 30.6, and 19.9 nN for 15, 45, and 75% RH, respectively). One possible explanation for such a profile would be the presence of dominant electrostatic forces at lower relative humidities.9 Analysis of Individual TAA force–piezo distance curves taken at 15 and 75% RH are shown in Figure 6a and b, respectively. At low RH (Fig. 6a) a negative curve in cantilever deflection as a function of piezo displacement suggests a long-range attraction force to be acting over the Z scan axis (2 mm). It is reasonable to assume that such a long-range attraction force would be due to electrostatic interactions. As the humidity was increased to 75% RH (Fig. 6b) the presence of such a long-range attraction was not observed. Again, this would be expected from an electrostaticaly charged material, as progressive sorption of water at high relative humidities would increase surface electron mobility, and hence, dissipate charge. In addition, TAA has negligible water solubility and is poorly wetted. Thus, at higher humidities, the presence of water may act as a physical barrier between TAA particulates.
759
Figure 6. Representative force–piezo distance curves for TAA drug probe interactions with micronized TAA particles, taken at (A) 15 and (B) 75% RH (258C).
In general, such observations correlate well with previous in vitro studies of TAA, which have reported a significant increase in fine particle fraction at high RH.15 As with SS, the adhesion between TAA and Tempfix polymer resin was greater than the cohesion force in all cases (113.6, 349.1, and 470.2 nN at 15, 45, and 75% RH). Again, this was expected. However, it is interesting to note that the TAA–Tempfix adhesion increases with increased RH suggesting capillary interactions dominate. The cohesion force for DSCG at 15% RH was greater than that for SS at equivalent humidity. Such observations may be due to the presence of water in the crystal structure (>5%), leading to elevated capillary interactions at relatively low RHs. Increasing the humidity from 15% RH to 45% RH resulted in an increased median cohesion force from 55.0 to 98.4 nN. By normalizing the data, it becomes clear that the relative/decrease in cohesive forces for DSCG is greater than SS and TAA, JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
760
YOUNG ET AL.
with a þ78.8% increase between 15 and 45% RH for DSCG in comparison to a þ47.6% and 67.5% change for SS and TAA over the same humidity range. This increase can most likely be attributed to further water absorption in combination with surface morphological changes of both the drug probe and Tempfix mounted particles. Elevation of humidity from 45 to 75% RH resulted in discontinuous FV measurements for DSCG interactions. Essentially, the force–piezo distance curves conducted between the DSCG drug probe and micronized particles became erratic, with multiple break points being observed during the retraction cycle of individual measurements. Such observations continued throughout the scan until the approach and retraction cycles were ‘‘separated’’ lines with no observed constant compliance region. This was indicative of drug probe loss, because the tipless cantilever surface will produce a contact area far greater than the spring constant tolerance. Subsequent observation of the DSCG drug probe tip, using optical microscopy, indicated the micronized drug to be ‘‘lost.’’ It is envisaged that such a loss may be due to morphological changes of the drug probe as water is absorbed into the crystal lattice. Furthermore, as DSCG is an interstitial solid, effectively held together by water molecules, the adhesion force between the individual drug particles may become great enough to ‘‘tear’’ apart or fracture the crystal lattice. Such observations were confirmed further, by analysis of two additional DSCG drug probes, which were also lost between 45 and 75% RH. Consequently, the cohesion force between DSCG particles could only be investigated by the AFM colloid probe technique at 15 and 45% RH. Nevertheless, such observations correlate well with previous in vitro investigations, which have reported the aersolization of DSCG above 60% RH to be effectively terminated.15 As with SS and TAA, the median separation force for DSCG–Tempfix was greater than the cohesion force in all cases (110.1 and 136.8 nN for DSCG–Tempfix interactions at 15 and 45% RH, respectively). By integration of the piezo–displacement cantilever–deflection curve, the energy of separation (esep) between a drug probe and surface can be obtained. The advantage of expressing particle interactions in terms of esep is that it takes into account the variations in mechanical properties of the two contacting surfaces. Although the force of adhesion/cohesion allows us insight into the influence of physicochemical properties, roughness JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
Figure 7. Cohesion energy distributions for SS, TAA, and DSCG as a function of relative humidity.
and environmental conditions on the interactive mechanisms of two surfaces, it does not directly take into account the elastic moduli of the contacting surfaces. A graphical representation of the individual cohesion energy data values for all three drugs, as a function of RH, is given in Figure 7. As with the cohesion force values, a positive skew in the separation energy was observed, suggesting the irregular morphology to produce a log-normal separation energy distribution. Again, as with the cohesion force measurements, the influence of RH had a similar significant impact on the cohesion profile of each drug. An increase in median separation energy was observed for SS (853.0, 1603.9, and 2786.5 nJ at 15, 45, and 75% RH, respectively) and DSCG (2623.4 and 8271.7 nJ), while a decrease was observed for TAA (2308.9, 571.5, and 259.5 nJ). Comparative analysis between force and separation energy values as a function of RH suggested good correlation. However, significant differences (ANOVA p < 0.05) in the force–energy factors were observed (for SS and TAA), suggesting, as expected, the presence of sorbed water to alter the elastic response of the separating surfaces.
CONCLUSION The variation in cohesion properties of three micronized drugs (SS, TAA, and DSCG) may be attributed to the balance of interparticulate forces that exist within the respective systems. Clearly, such a balance is dependent on a multitude of factors including chemistry, crystal structure,
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
and surface morphology. Although such studies can be undertaken using conventional in vitro techniques, they often prove time consuming. In comparison, the AFM, if used in the early stage of product development may prove an invaluable tool for rapid screening of formulation components.
REFERENCES 1. Booth SW, Newton JM. 1987. Experimental investigation of adhesion between powders and surfaces. Int J Pharm 39:679–684. 2. Podczeck F, Newton JM, James MB. 1994. Assessment of adhesion and autoadhesion forces between particles and surfaces 1. The investigation of autoadhesion phenomena of salbutamol xinafoate and lactose monohydrate particles using compacted surfaces. J Adhes Sci Technol 8:1459–1472. 3. Clarke MJ, Peart J, Cagnani S, Byron PR. 2002. Adhesion of powders for inhalation: An evaluation of drug detachment from surfaces following deposition from aerosol streams. Pharm Res 19:322–329. 4. Staniforth JN. 1995. Performance modifying influences in dry powder inhalation systems. Aerosol Sci Technol 22:346–353. 5. Binnig G, Quate CF, Gerber C. 1986. Atomic force microscope. Phys Rev Lett 56:930–933. 6. Ducker WA, Senden TJ, Pashley RM. 1991. Direct measurement of colloidal forces using an atomic force microscope. Nature 353:239–329. 7. Heinz WF, Hoh JH. 1999. Spatially resolved force spectroscopy of biological surfaces using the atomic force microscope. Trend Biotechnol 17:143–150. 8. Louey MD, Mulvaney P, Stewart PJ. 2001. Characterisation of adhesional properties of lactose carriers using atomic force microscopy. J Pharm Anal 25:559–567. 9. Ibrahim TH, Burk TR, Etzler FM, Neuman RD. 2000. Direct adhesion measurements of pharmaceutical particles to gelatin capsule surfaces. J Adhes Sci Technol 14:1225–1242.
761
10. Sindel U, Zimmermann I. 2001. Measurement of interaction forces between individual powder particles using an atomic force microscope. Powder Technol 117:247–254. 11. Eve JK, Patel N, Luk SY, Ebbens SJ, Roberts CJ. 2002. A study of single drug particle adhesion interactions using atomic force microscopy. Int J Pharm 238:17–27. 12. Young PM, Cocconi D, Colombo P, Bettini R, Price R, Steele DF, Tobyn MJ. 2002. Characterisation of a surface modified dry powder inhalation carrier prepared by ‘‘particle smoothing.’’ J Pharm Pharmacol 54:1339–1344. 13. Price R, Young PM, Edge S, Staniforth JN. 2002. The influence of relative humidity on particulate interactions in carrier-based dry powder inhaler formulations. Int J Pharm 246:47–59. 14. Young PM, Price R, Tobyn MJ, Buttrum M, Dey F. 2003. Investigation into the effect of humidity on drug–drug interactions using the atomic force microscope. J Pharm Sci 92:815–822. 15. Young PM, Price R, Tobyn MJ, Buttrum M, Dey F. 2003. Effect of humidity on the aerosolisation of micronised drugs. Drug Del Indust Pharm 29:959– 966. 16. Applications note AN20. Cambridge, UK: Digital Instruments. Web address http://www.di.com 17. Giles CH, MacEwan TH, Nakhwa SN, Smith D. 1960. Studies in adsorption. Part XI.* A system for classification of solution adsorption isotherms, and its use in diagnosis of adsorption mechanisms and in measurement of specific surface areas of solids. J Chem Soc 3973–3993. 18. Hartshorhe NH, Woodard GD. 1973. Mesomorphism in the system disodium chromoclycate–water. Mol Cryst Liq Cryst 23:343–368. 19. Cox JSG, Woodard GD, McCrone WC. 1971. Solidstate chemistry of cromolyn sodium. J Pharm Sci 60:1458–1465. 20. Chen RL, Young VG, Lechuga-Ballesteros D, Grant DJW. 1999. Solid-state behaviour of cromolyn sodium hydrates. J Pharm Sci 88:1191–1200. 21. Pizzi A, Mittal KL. 1994. Handbook of adhesive technology. New York: Marcel Dekker.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
Pharmaceutical Technology Research Group, Department of Pharmacy and Pharmacology, University of Bath, Claverton Down, Bath, BA2 7AY, UK 2
Respiratory Technology, Aventis Pharma, Holmes Chapel, UK, CW4 8BE
Received 12 June 2003; revised 4 August 2003; accepted 4 August 2003
ABSTRACT: The influence of relative humidity (RH) on the cohesion properties of three drugs: salbutamol sulphate (SS), triamcinolone acetonide (TAA), and disodium cromoglycate (DSCG) was investigated using the atomic force microscope (AFM) colloidal probe technique. Micronized drug particles were mounted in heat-sensitive epoxy resin for immobilization. Multiple AFM force–distance curves were conducted between each drug probe and the immobilized drug particulates at 15, 45, and 75% RH using Force– Volume imaging. Clear variations in the cohesion profile with respect to RH were observed for all three micronized drugs. The calculated force and energy of cohesion to separate either micronized SS or DSCG increased as humidity was raised from 15 to 75% RH, suggesting capillary forces become a dominating factor at elevated RH. In comparison, the separation force and energy for micronized TAA particles decreased with increased RH. This behavior may be attributed to long-range attractive electrostatic interactions, which were observed in the approach cycle of the AFM force–distance curves. These observations correlated well with previous aerosolization studies of the three micronized drugs. ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 93:753–761, 2004
Keywords: AFM; separation energy/force; relative humidity; cohesion; individual particle–particle measurements
INTRODUCTION Particle–particle interactions in pharmaceutical systems inevitably dominate the performance and efficacy of a solid dosage form. For dry powder inhalation systems, which aim to deliver sub-5 micron particulates to the respiratory tract, the magnitude of these adhesion or cohesion forces are critical for both the interactive formulation and drug liberation during patient inspiration. Drug particles for inhalation therapy are commonly produced by size reduction of larger crystal-
Correspondence to: Paul M. Young (Telephone: þ44 01225 383103; Fax: þ 44 01225 386114; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 93, 753–761 (2004) ß 2004 Wiley-Liss, Inc. and the American Pharmacists Association
line material using high-energy comminution processes such as micronization. Consequently, the resultant micron sized (<5 mm) particles have irregular surface morphologies with a high surface area. When combining such physical factors with the intrinsic chemical and crystal structure, the variation in adhesion or cohesion between individual particles becomes very complex. It is generally considered that three interparticulate forces are prevalent in dry powder inhalation systems: van der Waals, capillary, and electrostatic forces. The degree to which each force dominates will be dependent on the physical, mechanical, and chemical properties of the contiguous surfaces and the environment to which they are exposed. Recently, with advances in inhalation drug design, and the potential of systemic delivery, a fundamental understanding of such particle–
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
753
754
YOUNG ET AL.
particle interactions has become imperative. Because theoretical calculation of adhesion or cohesion forces in pharmaceutical systems are very complex (due to the stratifying variables), the simplest approach would be direct measurement of particle interactions. Indeed, such investigations are commonplace, and can be found in abundance in the literature.1–4 Generally, such measurements involve the removal of small quantities of powder from model surfaces using techniques such as centrifugation1–3 or reentrainment.4 However, such techniques may have limitations. For example, when considering centrifugation, the maximum theoretical removal force will approach or become smaller than the theoretical adhesion force. In addition, the determination of individual particle adhesion forces becomes very difficult, as such measurements are mainly dependent on the removal of a large quantity of particulates or agglomerates. A possible method for overcoming such issues is the use of an atomic force microscope.5 In simple terms, by mounting a particle on the apex of a microfabricated cantilever (with known spring constant), it becomes possible to construct a ‘‘probe,’’ which can be used to measure the forces of interaction.6 Furthermore, because the cantilever deflection, which is dependent on the acting force, is measured as a function of probe–substrate separation, it becomes possible to investigate influence of short-range attractive (or repulsive) forces.7 Previous investigations (applicable to DPI technology), using the AFM colloid probe technique include determination of lactose–silica,8 lactose– gelatin,9 and lactose–lactose10 interactions. However, little work has been conducted using drug probes of micron size (<5 mm), applicable to particles used in inhalation. The most relevant published work has investigated the influence of inhaler components11 and carrier morphology12 on drug adhesion. More recently, the influence of RH on drug carrier adhesion13 and drug cohesion14 has demonstrated the colloid probe technique a powerful tool in predicting in vitro behavior. Nonetheless, previous studies have mainly focused on measuring the interaction between micronized drug probes and relatively large planar surfaces (such as drug compacts or carrier surfaces). As part of an ongoing study, investigating the influence of RH on drug cohesion, three drugs were chosen for their formulation characteristics and analyzed using an individual particle approach. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
Three micronized drugs of comparable size were chosen for investigation: salbutamol sulphate (SS), triamcinolone acetonide (TAA), and disodium cromoglycate (DSCG). Previous in vitro investigations have shown RH (over the range 11 to 75% RH) to have a significant effect on the aerosolization properties of the micronized drugs.15 Because such in vitro studies can prove time consuming, the AFM colloid probe technique was employed for comparative study, to assess its potential as a rapid formulation screening technique.
MATERIALS AND METHODS Materials Micronized salbutamol sulphate (SS), triamcinolone acetonide (TAA), and disodium cromoglycate (DSCG) (all Aventis Pharma, Holmes Chapel, UK) were used as supplied. Physical Characterization Moisture sorption profiles for SS, TAA, and DSCG were determined using dynamic vapor sorption (DVS) (DVS-1, Surface Measurement Systems Ltd, London, UK). Approximately 10 mg of drug was weighed into the sample cell and subjected to two 0–90% RH cycles at 258C (10% RH steps). Equilibration at each RH was determined by a dm/dt of 0.002. The influence of RH on the surface morphology of the micronized drugs was investigated usingAFM in Tapping-ModeTM (Multimode AFM, Nanoscope IIIa controller, Digital Instruments, UK). Briefly, samples of the micronized drug were lightly brushed through a 45-mm aperture stainless steel sieve (BS410/1986, Endecotts, London, UK) onto a piece of freshly cleaved, grade 1, muscovite mica (SPI supplies, Pennsylvania). Samples were imaged with a high-aspect ratio silicon probe (OTESP, Digital Instruments, UK) at a scan rate of 0.7 Hz. Atomic force microscope images of the same region were taken at 15, 45, and 75% RH (at 258C) using a custom-built perfusion apparatus described elsewhere.13,14 Mounting of Micronized Drug Particles on Epoxy Resin Micronized particle samples were immobilized on steel AFM stubs with Tempfix (SPI Supplies) at 408C using a custom-built Peltier connected to a thermocouple temperature controller (SE5000,
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
Marlow Industries, Texas). Briefly, a steel AFM sample stub was mounted on the Peltier and heated to 1208C. A small quantity of Tempfix was melted onto the stub and cooled to 408C, at which point the resin becomes tacky or adhesive. A small quantity of the powder was lightly brushed through a 45 mm aperture stainless steel sieve (BS410/1986, Endecotts, London, UK) onto the AFM sample stub. The sieving procedure was found to remove most agglomerates. The AFM sample stubs, containing Tempfix-mounted particles, were left to cool to room temperature before removing loose material with a filtered nitrogen air stream. The mounted drug samples were stored in tightly sealed containers for 24 h prior to use.
755
RESULTS AND DISCUSSION Physical Characterization To fully quantify the effect of RH on cohesive interactions, the three drugs were first characterized for moisture sorption and morphology. Dynamic vapor sorption isotherms for SS, TAA, and DSCG are shown in Figure 1a,b, and c, respectively. The water sorption isotherms for SS
Atomic Force Microscope Colloid Drug Probe Measurements All AFM investigations were conducted at specific environmental conditions using a custom-built perfusion system described elsewhere.13,14 Micronized particles of each of the three drugs were mounted onto the apex of nominal 0.58 Nm1 tipless AFM cantilevers using a micromanipulation process detailed elsewhere.14 Multiple force– distance curves were conducted between each drug probe and the corresponding Tempfix mounted particle substrates using Force–Volume (FV) imaging,16 over a 25 25-mm area. This provided a block of 64 64 force–distance curves corresponding to the relative X-Y position in the Force–Volume scan. In addition, by recording the relative piezo height at the constant compliance limit, a low-resolution ‘‘relief image’’ of the FV region could be produced.16 Unlike previous investigations, which utilized batch-conversion software to process FV data,12–14 force–distance curves were analyzed using an individual approach. Force–distance curves, conducted on individual micronized drug particles, were exported and manually analyzed/integrated to produce both separation force and separation energy values. Approximately 30 randomly selected force–distance curves across six micronized particles were analyzed for each drug (at each humidity). For comparative purposes, drug probe–Tempfix interactions were also analyzed. Cohesion measurements for each drug were conducted over the same FV region at 15, 45, and 75% RH (at 258C). A stabilization period of at least 4 h was instigated with each incremental increase in RH to obtain dynamic equilibrium.
Figure 1. Moisture sorption isotherms for (a) SS, (b) TAA, and (c) DSCG. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
756
YOUNG ET AL.
(shown in Fig. 1a) followed a sigmoidal class L-3 curve,17 suggesting Langmuir-type monomulti layer water sorption onto the crystal surface. In comparison, analysis of the moisture sorption curves for TAA and DSCG suggested non-Langmuir type isotherms. The relative percent moisture sorption for TAA was approximately one order of magnitude greater than that of salbutamol sulphate, suggesting TAA had a higher affinity for water. Thus, it is proposed that the moisture sorption curve for TAA may be a class H curve.17 Class H curves are essentially a ‘‘high-affinity’’ form of the L type curve in which the adsorbate may be chemisorbed. Moisture sorption isotherms for DSCG (Fig. 1c) suggest a class C sorption curve.17 Class C curves are indicative of porous materials with a high affinity for the adsorbate and are associated with absorption into the crystal structure. This is to be expected, however, as DSCG is essentially a solid solution, whose crystal structure contains channels that allow the up to nine molecules of water per molecule of DSCG (depending on the RH), before collapsing into a series of liquid crystal mesomorphs.18 Dynamic vapor sorption isotherms for the three micronized materials showed no characteristic hysterisis relating to the presence of amorphous material. Representative AFM images of SS and TAA at 45% RH are shown in Figure 2a and b, respectively. It can be seen from Figure 2 that SS exhibits a columnar shape (Fig. 2a), while TAA (Fig. 2b) appears to have a more flattened platelet morphology. Similar low-magnification AFM images of DSCG (not shown here) suggested a columnar shape. In general, the low-magnification AFM images agreed well with previously reported studies of the micronized drugs using conventional imaging methods such as optical microscopy and scanning electron microscopy.15 However, it is important to note that limited information was available in the literature regarding the general morphology of TAA. Both SS and TAA showed no macroscopic changes in particle morphology over the humidity range 15–75% RH. In comparison, the morphology of DSCG clearly changed as humidity was increased from 15 to 75% RH. suggesting sorbed moisture to integrally influence particle morphology. Representative high-resolution AFM images of a DSCG particle after exposure to 15, 45, and 75% RH (for approximately 4 h at each humidity) are shown in Figure 3a,b, and c, respectively. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
Figure 2. Atomic force microscope Tapping mode images of (A) SS and (B) TAA.
Variations in particle morphology of the micronized DSCG with respect to RH were expected, as DSCG is effectively a nematic choromonic liquid crystal that below 93% RH forms a solid solution.18–20 As previously discussed, the planar cromoglycate molecules are ‘‘stacked’’ to form long rods that, in turn, form an interstitial solid held together by water molecules. These water channels can subsequently contain up to approximately nine molecules of water per unit cell before collapsing to form a liquid crystal mesophase (at 93% RH). Adsorption or desorption of water molecules into the crystal structure will subsequently cause changes in the torsional angles of the cromoglycate molecules,20 resulting in a variation in the unit cell dimensions. Consequently, it is likely that such a change in crystal structure with respect to RH would have a significant impact on the macroscopic particle morphology.
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
757
Figure 4. Atomic force microscope Tapping mode height image of a 25 25 mm area of the salbutamol sulphate particles mounted on the Tempfix polymer surface.
Figure 3. High-resolution AFM Tapping mode images of DSCG taken at 15, 45, and 75% RH (258C).
Atomic Force Microscope Colloid Drug Probe Measurements The influence of RH on drug particle cohesion was investigated by conducting FV scans (25 25 mm) between a drug probe and corresponding drug particles mounted on Tempfix polymer resin. A representative AFM tapping mode ‘‘height’’ image of the mounted SS particles is shown in
Figure 4. The salbutamol sulphate drug particulates had an irregular morphology, and were clearly proud of the Tempfix polymer surface. In addition, AFM tapping mode images on randomly chosen areas of a ‘‘control’’ Tempfix surface (not shown here), prepared without drug particle mounting, showed no such exposed irregularities. Atomic force microscope height images for TAA and DSCG mounted in Tempfix produced similar results. Essentially, such variations in height allow the identification of individual mounted drug particles and enable subsequent cohesion measurement. An example of a Force–Volume scan, conducted between an SS drug probe and SS mounted particles, is shown in Figure 5. The white areas on the FV ‘‘relief’’ height image correspond to regions with a Z-axis height of greater than 2.5 mm. It would be reasonable to assume that such undulations in the surface could be identified as micronized drug particles, as they correspond to observations made using conventional AFM tips an imaging methods (Fig. 4). Comparison of the ‘‘relief’’ height image with the FV image (Fig. 5) suggest the raised ‘‘drug particle’’ areas have a low adhesion force (dark areas on FV image) in comparison to the polymer surface. A representative force curve for the SS interactions is included in the FV scan (Fig. 5) with its relative position marked with a cross in the ‘‘relief’’ height and FV images, respectively. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
758
YOUNG ET AL.
Figure 5. Force–volume data for salbutamol sulphate drug probe interactions with salbutamol sulphate particles mounted in Tempfix. Lighter areas represent elevated topography and lower adhesion forces in the ‘‘relief’’ height image and FV image panes, respectively.
To quantify such drug probe interactions, randomly chosen force curves taken across five distinct ‘‘raised particle’’ regions were exported and manually integrated to obtain both cohesion force and separation energy values. It is important to note that although an individual analysis approach does not allow such comprehensive statistical analysis as batch reprocessing, it was deemed necessary to ensure differentiation between micronized particle and Tempfix polymer interactions. A detailed summary of the SS, TAA, and DSCG cohesive parameters is given in Table 1. ANOVA
one-way analysis of variance indicated RH to have a significant effect (p < 0.05) on the cohesion force of all three micronized drugs. However, such an observation is effectively dependent on a normal distribution. Further analysis of the cohesion force data (Table 1) suggested this not to be the case, as the large standard deviations and positive skewness values indicate a non-normal distribution. Analysis of the cohesion force values indicated a log-normal distribution to best fit the data. Furthermore, this agreed with previous reports using multiple colloid drug probe measurements on irregular surfaces.14–16
Table 1. Summary of Cohesion Force Measurements, Conducted between Individual Drug Probes and Six Corresponding Micronized Drug Particles Mounted in Tempfix Resin (n ¼ 30)
Drug SS
TAA
DSCG
a
Humidity (% RH)
Mean Cohesion Force (nN)
15 45 75 15 45 75 15 45 75a
39.941 54.795 65.731 65.088 36.491 27.339 66.292 116.491 —
St Dev
Median (f0.5) (nN)
10th Percentile (f0.1) (nN)
90th Percentile (f0.9) (nN)
GSD
Skewness
24.155 28.144 30.910 21.421 18.297 19.871 45.269 60.086 —
32.887 48.531 58.747 61.154 30.599 19.903 55.041 98.431 —
14.198 25.767 31.417 37.923 12.821 6.090 25.749 43.662 —
76.174 91.405 109.853 100.818 73.027 65.045 117.656 221.901 —
1.924 1.638 1.629 1.451 1.961 2.516 1.808 1.884 —
0.748 1.021 0.725 1.100 0.756 1.066 1.604 0.221 —
DSCG drug probe was ‘‘lost’’ during measurements at 75% RH (repeated with two further probes, refer to text).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
In general, the cohesion force values for salbutamol sulphate agreed with previously published data using both the AFM colloid probe technique.13,14 To the authors’ knowledge, no literature relating to fundamental cohesion measurements on DSCG or TAA exist. Analysis of lognormal probability plots for SS, TAA, and DSCG suggested to have a significant effect (CI ¼ 0.95) on the cohesion force. The median cohesion force (f0.5) for SS increased with respect to RH (32.9, 48.5, and 58.7 nN for 15, 45, and 75% RH, respectively), suggesting capillary interactions to become more dominant at higher relative humidities. Such observations correlate well with previous in vitro investigations, which demonstrated a decrease in SS fine-particle aerosolization as RH was increased.15 Comparison of the both the mean and median SS cohesion values with Tempfix adhesion measurements suggested the polymer drug adhesion to be significantly greater than drug cohesion. All adhesion measurements conducted between the ‘‘flat’’ polymer regions of the FV scan were of a higher value than that of SS cohesion (142.1, 152.6, and 164, at 15, 45, and 75% RH, respectively). Such observations are to be expected, however, as adhesive resins have high surface energies due to the polar nature of groups present on the polymer chains.21 In contrast to the SS cohesion profile, the median separation force for TAA decreased as RH was increased (61.2, 30.6, and 19.9 nN for 15, 45, and 75% RH, respectively). One possible explanation for such a profile would be the presence of dominant electrostatic forces at lower relative humidities.9 Analysis of Individual TAA force–piezo distance curves taken at 15 and 75% RH are shown in Figure 6a and b, respectively. At low RH (Fig. 6a) a negative curve in cantilever deflection as a function of piezo displacement suggests a long-range attraction force to be acting over the Z scan axis (2 mm). It is reasonable to assume that such a long-range attraction force would be due to electrostatic interactions. As the humidity was increased to 75% RH (Fig. 6b) the presence of such a long-range attraction was not observed. Again, this would be expected from an electrostaticaly charged material, as progressive sorption of water at high relative humidities would increase surface electron mobility, and hence, dissipate charge. In addition, TAA has negligible water solubility and is poorly wetted. Thus, at higher humidities, the presence of water may act as a physical barrier between TAA particulates.
759
Figure 6. Representative force–piezo distance curves for TAA drug probe interactions with micronized TAA particles, taken at (A) 15 and (B) 75% RH (258C).
In general, such observations correlate well with previous in vitro studies of TAA, which have reported a significant increase in fine particle fraction at high RH.15 As with SS, the adhesion between TAA and Tempfix polymer resin was greater than the cohesion force in all cases (113.6, 349.1, and 470.2 nN at 15, 45, and 75% RH). Again, this was expected. However, it is interesting to note that the TAA–Tempfix adhesion increases with increased RH suggesting capillary interactions dominate. The cohesion force for DSCG at 15% RH was greater than that for SS at equivalent humidity. Such observations may be due to the presence of water in the crystal structure (>5%), leading to elevated capillary interactions at relatively low RHs. Increasing the humidity from 15% RH to 45% RH resulted in an increased median cohesion force from 55.0 to 98.4 nN. By normalizing the data, it becomes clear that the relative/decrease in cohesive forces for DSCG is greater than SS and TAA, JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
760
YOUNG ET AL.
with a þ78.8% increase between 15 and 45% RH for DSCG in comparison to a þ47.6% and 67.5% change for SS and TAA over the same humidity range. This increase can most likely be attributed to further water absorption in combination with surface morphological changes of both the drug probe and Tempfix mounted particles. Elevation of humidity from 45 to 75% RH resulted in discontinuous FV measurements for DSCG interactions. Essentially, the force–piezo distance curves conducted between the DSCG drug probe and micronized particles became erratic, with multiple break points being observed during the retraction cycle of individual measurements. Such observations continued throughout the scan until the approach and retraction cycles were ‘‘separated’’ lines with no observed constant compliance region. This was indicative of drug probe loss, because the tipless cantilever surface will produce a contact area far greater than the spring constant tolerance. Subsequent observation of the DSCG drug probe tip, using optical microscopy, indicated the micronized drug to be ‘‘lost.’’ It is envisaged that such a loss may be due to morphological changes of the drug probe as water is absorbed into the crystal lattice. Furthermore, as DSCG is an interstitial solid, effectively held together by water molecules, the adhesion force between the individual drug particles may become great enough to ‘‘tear’’ apart or fracture the crystal lattice. Such observations were confirmed further, by analysis of two additional DSCG drug probes, which were also lost between 45 and 75% RH. Consequently, the cohesion force between DSCG particles could only be investigated by the AFM colloid probe technique at 15 and 45% RH. Nevertheless, such observations correlate well with previous in vitro investigations, which have reported the aersolization of DSCG above 60% RH to be effectively terminated.15 As with SS and TAA, the median separation force for DSCG–Tempfix was greater than the cohesion force in all cases (110.1 and 136.8 nN for DSCG–Tempfix interactions at 15 and 45% RH, respectively). By integration of the piezo–displacement cantilever–deflection curve, the energy of separation (esep) between a drug probe and surface can be obtained. The advantage of expressing particle interactions in terms of esep is that it takes into account the variations in mechanical properties of the two contacting surfaces. Although the force of adhesion/cohesion allows us insight into the influence of physicochemical properties, roughness JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004
Figure 7. Cohesion energy distributions for SS, TAA, and DSCG as a function of relative humidity.
and environmental conditions on the interactive mechanisms of two surfaces, it does not directly take into account the elastic moduli of the contacting surfaces. A graphical representation of the individual cohesion energy data values for all three drugs, as a function of RH, is given in Figure 7. As with the cohesion force values, a positive skew in the separation energy was observed, suggesting the irregular morphology to produce a log-normal separation energy distribution. Again, as with the cohesion force measurements, the influence of RH had a similar significant impact on the cohesion profile of each drug. An increase in median separation energy was observed for SS (853.0, 1603.9, and 2786.5 nJ at 15, 45, and 75% RH, respectively) and DSCG (2623.4 and 8271.7 nJ), while a decrease was observed for TAA (2308.9, 571.5, and 259.5 nJ). Comparative analysis between force and separation energy values as a function of RH suggested good correlation. However, significant differences (ANOVA p < 0.05) in the force–energy factors were observed (for SS and TAA), suggesting, as expected, the presence of sorbed water to alter the elastic response of the separating surfaces.
CONCLUSION The variation in cohesion properties of three micronized drugs (SS, TAA, and DSCG) may be attributed to the balance of interparticulate forces that exist within the respective systems. Clearly, such a balance is dependent on a multitude of factors including chemistry, crystal structure,
RELATIVE HUMIDITY INFLUENCE ON MICRONIZED DRUGS FOR INHALATION THERAPY
and surface morphology. Although such studies can be undertaken using conventional in vitro techniques, they often prove time consuming. In comparison, the AFM, if used in the early stage of product development may prove an invaluable tool for rapid screening of formulation components.
REFERENCES 1. Booth SW, Newton JM. 1987. Experimental investigation of adhesion between powders and surfaces. Int J Pharm 39:679–684. 2. Podczeck F, Newton JM, James MB. 1994. Assessment of adhesion and autoadhesion forces between particles and surfaces 1. The investigation of autoadhesion phenomena of salbutamol xinafoate and lactose monohydrate particles using compacted surfaces. J Adhes Sci Technol 8:1459–1472. 3. Clarke MJ, Peart J, Cagnani S, Byron PR. 2002. Adhesion of powders for inhalation: An evaluation of drug detachment from surfaces following deposition from aerosol streams. Pharm Res 19:322–329. 4. Staniforth JN. 1995. Performance modifying influences in dry powder inhalation systems. Aerosol Sci Technol 22:346–353. 5. Binnig G, Quate CF, Gerber C. 1986. Atomic force microscope. Phys Rev Lett 56:930–933. 6. Ducker WA, Senden TJ, Pashley RM. 1991. Direct measurement of colloidal forces using an atomic force microscope. Nature 353:239–329. 7. Heinz WF, Hoh JH. 1999. Spatially resolved force spectroscopy of biological surfaces using the atomic force microscope. Trend Biotechnol 17:143–150. 8. Louey MD, Mulvaney P, Stewart PJ. 2001. Characterisation of adhesional properties of lactose carriers using atomic force microscopy. J Pharm Anal 25:559–567. 9. Ibrahim TH, Burk TR, Etzler FM, Neuman RD. 2000. Direct adhesion measurements of pharmaceutical particles to gelatin capsule surfaces. J Adhes Sci Technol 14:1225–1242.
761
10. Sindel U, Zimmermann I. 2001. Measurement of interaction forces between individual powder particles using an atomic force microscope. Powder Technol 117:247–254. 11. Eve JK, Patel N, Luk SY, Ebbens SJ, Roberts CJ. 2002. A study of single drug particle adhesion interactions using atomic force microscopy. Int J Pharm 238:17–27. 12. Young PM, Cocconi D, Colombo P, Bettini R, Price R, Steele DF, Tobyn MJ. 2002. Characterisation of a surface modified dry powder inhalation carrier prepared by ‘‘particle smoothing.’’ J Pharm Pharmacol 54:1339–1344. 13. Price R, Young PM, Edge S, Staniforth JN. 2002. The influence of relative humidity on particulate interactions in carrier-based dry powder inhaler formulations. Int J Pharm 246:47–59. 14. Young PM, Price R, Tobyn MJ, Buttrum M, Dey F. 2003. Investigation into the effect of humidity on drug–drug interactions using the atomic force microscope. J Pharm Sci 92:815–822. 15. Young PM, Price R, Tobyn MJ, Buttrum M, Dey F. 2003. Effect of humidity on the aerosolisation of micronised drugs. Drug Del Indust Pharm 29:959– 966. 16. Applications note AN20. Cambridge, UK: Digital Instruments. Web address http://www.di.com 17. Giles CH, MacEwan TH, Nakhwa SN, Smith D. 1960. Studies in adsorption. Part XI.* A system for classification of solution adsorption isotherms, and its use in diagnosis of adsorption mechanisms and in measurement of specific surface areas of solids. J Chem Soc 3973–3993. 18. Hartshorhe NH, Woodard GD. 1973. Mesomorphism in the system disodium chromoclycate–water. Mol Cryst Liq Cryst 23:343–368. 19. Cox JSG, Woodard GD, McCrone WC. 1971. Solidstate chemistry of cromolyn sodium. J Pharm Sci 60:1458–1465. 20. Chen RL, Young VG, Lechuga-Ballesteros D, Grant DJW. 1999. Solid-state behaviour of cromolyn sodium hydrates. J Pharm Sci 88:1191–1200. 21. Pizzi A, Mittal KL. 1994. Handbook of adhesive technology. New York: Marcel Dekker.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 93, NO. 3, MARCH 2004