Sustained-release delivery systems, I: Phase diagram studies of dapsone and selected derivatives

Sustained-Release Delivery Systems, I: Phase Diagram Studies of Dapsone and Selected Derivatives

TSONG-TOH YANG* AND JAMES SWARBRICK' Received July 9, 1985, from the .Division of Pharmaceutics, School of Pharmacy, Universily of. North Carolina at Chapel Hill, Chapel Hill, NC 27514. Accepted for publication October 7, 1985. Present address: 'Schering Corporation, Kenilworth, NJ 07033.

Abstract 0 In order to develop slowly dissolving particles containing the antileprotic drug dapsone (4,4'-sulfonylbisbenzamine, 1) that would be suitable for a sustained-release intramuscular injection, the dilauryl (2) and monolauryl (3) derivatives of dapsone, N,N'-didodecanoyl-4,4'- sulfonylbisbenzarnine and Ndodecanoyl-4,4'-sulfonylbisbenzamine, respectively, were studied for their ability to form solid dispersions with the parent compound. The 1:2 binary phase diagram showed these two compounds were partially miscible in the liquid state, leading to the coexistence of a monotectic and a eutectic system in the phase diagram. The 1:3 phase diagram showed that these compounds were completely miscible in the liquid state and formed discontinuous solid solutions in the solid state. At a cooling rate of 25°C rnip-', eutectic mixtures lying on the dapsone side of the eutectic point formed glass solutions. Taken as a whole, the results demonstrate that the molecular interactions between 1 and 3 are stronger than those between 1 and 2. Accordingly, 3 would appear to be the better carrier for reducing the rate of dissolution of dapsone from a solid dispersion.

Dapsone (4,4'-sulfonylbisbenzamine, 1) is an effective antileprotic drug. However, the problem of patient noncompli- ance',* and the emergence of resistant strains of Mycobacte- rium Zeprae3 as a consequence of low blood levels make it desirable to develop a sustained-release formulation that can be given as infrequently as once a month. Solid dispersions have been used to increase the dissolution of poorly soluble drugs in soluble carrierscG as well as to sustain the release of soluble compounds in less-soluble matrices.'-g

The use of a dispersion of active drug with a less-soluble derivative has been proposed as a viable approach to produc- ing parenteral sustained-release dosage forms of relatively low toxicity.10 The less-soluble derivative dissolves slowly and thus retards the release of the active compound. If the less-soluble derivative is then transformed to the active drug, it can supplement the activity of the drug. Under these circumstances, the sustained-release formulation can be re- garded as having a potential drug content of 100%. Questions of toxicity arising from the use of other matrices, especially those that are nonbiodegradable, are minimized and may be avoided. The bioavailability of this type of solid dispersion is dependent on the derivative used, its solubility, and its degree and rate of conversion to the bioactive species, plus subsequent metabolism. Other factors involved include the pharmacological activity of the less-soluble derivative and the ratio of derivative to drug needed to change the release pattern of the drug.

To investigate the possibility of developing an intramuscular sustained-release dosage form of dapsone by forming a solid dispersion with a less-soluble derivative, the dilauryl (2) and monolauryl(3) derivatives of dapsone, N,"-didodecanoyl-4,4'-sulfonylbisbenzamine and N-dodecanoyl-4,4'-~~1- fonylbisbenzamine, respectively, were selected for formulation with 1. The potential advantages of using these compounds include: (a) the solubilities of 2 (c0.2 FglmL) and 3 (<0.5 pg/mL) in pH 7 phosphate buf€erll are sufficiently low so that slow dissolution in vivo can be expected; ( 6 ) the melting points of 2 and 3 (-150°C and 160"C, respectively) are lower than the melting point of 1 (175-177°C); and (c) as derivatives of dapsone (11, they can be expected to hydrolyze in vivo to dapsone once they have dissolved. The phase

equilibria existing in 1:2 and 1:3 binary systems were there- fore investigated in order to elucidate and better understand the interactions between these pairs of compound in the solid state, in the belief that such interactions will affect the dissolution rates of the components.

Experimental Section Materials-Dapsone (4,4'-sulfonylbisbenzamine, 1, Aldrich

Chemical Company, Inc., Milwaukee, WI) was used as received without further purification. N,"-Didodecanoyl-4,4'-sulfonylbisbenzamine (2) was synthesized according to the method of Buttle et a1.12 The final product was recrystallized from absolute alcohol and had a melting point of 149-150°C. N-Dodecanoyl-4,4'-sulfonylbisbenzamine, (3) was synthesized according to the method of Miura and Bando,lS in which the final product was recrystallized from impropyl alcohol. The melting point was 159.5-160.0"C. Chemical structures were confirmed by IR and NMR spectroscopy.

Physical mixtures were prepared by triturating various propor- tions of 1 and 2 or 3 in an agate mortar and pestle. Comelts of 1:3 were prepared by placing known mixtures of 1 and 3 in screw-capped glass tubes which were heated slowly in an oil bath to 182"C, at which point the heater was turned off. While stirring, the system was cooled to room temperature at a rate of -2.5Wmin. Under these conditions, comelts containing 45% to 86.56, w/w, of 1 formed glass solutions; all other comelts solidified to form crystals. Since prelimi- nary dissolution studies showed the crystalline form to dissolve more elowly than the glass form, comelts containing 45-86.5%, w/w, of 1 were seeded with a small crystal of dapsone during cooling.

Polymorphism and Enthalpy Changes of Pure Compounds-IR spectra of dapsone powder and of dapsone melt were obtained using a grating IR spectrophotometer (model 257, Perkin-Elmer Corpora- tion, Norwalk, CT). A KBr disk containing -2% of the sample was used.

Differential scanning calorimetry (model DSC-2 with Scanning Auto Zero, Perkin-Elmer) was used for the thermal analysis, and indium (mp = 156.60 "C) and lead (mp = 327.47"C) were used as standards. The sample and reference pans were continuously purged with nitrogen gas at a pressure of 20 psi. Differential scanning calorimetry (DSC) thermograms of 1, 2, and 3 (3 mg each) were obtained by scanning at 5"/min from 50°C to 187°C. The sample was then held at 187°C in situ for 20 min, cooled at 5"Clmin to 27T, and held at room temperature. After 6 d, the sample was scanned again using the same operating conditions. The peak transition temperature and area under the peak were obtained and the enthalpy change was calculated by comparing the area under the transition peak with that of a standard indium sample (AHf = 6.80 cal/g).

Phase Diagram Determinations-Differential scanning calorimetry thermograms of various physical mixture blends of 1 and 2 (2 mg each) were obtained by scanning the sample at Wmin from 108°C to 182°C. The melting points of physical mixtures containing 12.5 and 15% wlw of dapsone were determined microscopically using a micro hot-stage. Thermograms of various physical mixtures and comelts of 1 and 3 (5 mg each) were obtained in a manner similar to that with 1:2 physical mixtures except that the comelts were held in situ at 182°C for 20 min, cooled at 2.5Ymin to 27"C, and then reheated immediately.

Determination of Glass Transition Temperature (T,)-Thermo- grams were obtained by shock cooling the samples at 320"Clmin in situ from 182°C to -32"C, and then reheating the sample at 20°C/ min. The TB was taken as the midpoint in the thermogram, as measured from the extensions of the pre- and post-transition base- line.

OOZ-3549/86/0 1 OO-OO53$01.OO/O Q 1986, American Pharmaceutical Association

Journal of Pharmaceutical Sciences / 53 Vol. 75, No. I , January 1986

Results and Discussion Polymorphism and Enthalpy Change-The transition

temperatures and enthalpy changes for 1,2,'and 3 are shown in Table I. The thermogram of 2 showed two solid-solid transition endotherms a t 109.3"C and 148.8"C, and a melting endotherm a t 151.8"C during the first heating. The second heating thermogram showed only a melting endotherm a t 151.8"C, implying that 2 recrystallized from absolute alcohol does not possess the same crystal form as that obtained from a melt. Compound 3 showed only one melting endotherm; this occurred at 161.5"C with both heatings. Compound 2 showed a similar thermogram with both heatings; however, the solid-solid transition endotherm at 78.8"C was enhanced on the second heating. The enthalpy changes for the solid- solid transition a t 78.8"C were 77 and 330 cal/mol from the first and second heating, respectively. A third heating pro- duced a thermogram identical to that from the second heating. The heats of fusion calculated from studying the raw material and the melt were 4740 and 4770 callmol, respectively, in good agreement with a value of 4710 caUmol reported previo~sly.1~

The polymorphism of 1 has been studied by several investi- gators in recent years.1s17 The data in Table I are in good agreement with the findings of Bernabei et al.17 who observed that the stable form at room temperature (form I) underwent an enantiotropic conversion to form I1 at -80°C. The existence of a hydrated form a t room temperature was also confirmed. These findings are consistent with the present work where IR studies on powder of dapsone and melt showed spectra similar to the hydrated form and form I, respectively. Work undertaken to determine the 1:3 phase diagram (see later) also suggests the existence of two poly- morphs in the raw material of dapsone used.

Phase Diagrams-Dapsone:Monolauryl Derivative of Dap- gone--Thermograms of various comelts containing 1 and 3 shown in Fig. 1 were used to construct the 1:3 binary phase diagram. While the sample containing 80% wlw of dapsone showed three well-defined maxima, it is likely that the second peak is not a phase transformation but is an artifact arising from the shape of the phase bOundaries.18

The eutectic heat of fusion as a function of the weight fraction of 1 to 3 was calculated and is plotted in Fig. 2. Such a plot can be helpful in calculating the composition of unknown mixtures, although a given eutectic heat of fusion will correspond to two compositions, one on each side of the eutectic. However, the terminal melting endotherm will usually be different and can be used to determine which of the two straight lines of such a plot should be used.19 From Fig. 2, the eutectic composition was found to contain 36.5%

Table I-Thermal Properties of 4,4'-SuHonylblsbenzamlne (l), N,N-Dldodecanoyl-4,4'-Sulfonylblsbenzamlne (2), and N Dodecanoyl-4,4'-Sulfonylblsbenzamlne (3)

Transition Enthalpy Change, Temperature, "C cal/mol

Compound First Second First Second Heat" Heat" Heat" Heat"

C

C

- 3020 - 871 - 2 109.3

148.8 - 151.8 151.8 9020 9290

3 161.5 161.5 7740 7550 1 78.8 78.8 77 330

175.3 175.3 4740 4770

'Heated to 187°C at 5"C/min. "Following first heating, the sample was held at 187°C for 20 min, cooled at 5"C/min to 27"C, kept at room temperature for 6 d, and then reheated to 187°C at 5"C/min. 'Peak disappeared.

180

y 160

120 1, wt. o/o

Figure 1- Differential scanning calorimetry thermograms of 1 3 comelts (numbers on x-axis represent the weight percent of 1 present in comelts).

1, wt. o/o

Flgure 2-Enthalpy of fusion of the eutectic portion of 1:3 comelts as function of composition.

w/w 1 with terminal solid solubilities of 11% w/w 3 in 1 and 2.8% w/w 1 in 3 a t the eutectic temperature. The phase diagram was then constructed as shown in Fig. 3. It shows that the eutectic composition contains equimolar amounts of 1 and 3 and that there is no pure 1 or 3 existing in the comelts. Those comelts with >89% of 1 may be described as "1-rich" solid solutions. Likewise, comelts with <2.8% exist as "3-rich" solid solutions. The eutectic is a mixture of "1- rich" and "3-rich" solid solutions. Comelts which lie on the 1 side of the eutectic point consist of excess "1-rich" solid solution and the eutectic mixture, whereas comelts on the 3 side of the eutectic point consist of excess "3-rich" solid solution and the eutectic mixture.

1

160 9 g- c

B140

P ,G

1 100 80 60 40 20 0

1, wt. v o

Flgure &Phase diagram of 1:3 comelts: (0) first heating; (0) second heating. Solid solution a contains 3 dispersed in 1 (referred to in text as "1-rich" solid solution. Solid solution p contains 1 dispersed in 3 (referred to in text as "3-rich" solid solution. Key: (A) solid solution a; (B) solid solution a plus liquid; (C) liquid, (D) solid solution p plus liquid; (E) solid solution a plus solid solution a:solid solution p; (F) solid solution p plus solid solution a:solid solution p; (G) solid solution p.

54 /Journal of Pharmaceutical Sciences Vol. 75, No. 7, January 7986

At a cooling rate of 2.5"C/min, eutectic mixtures on the 1 side of the eutectic point form glass solutions. Thermograms from the second heating of such comelts within these compositions did not show a melting endotherm. Peak temperatures of the differential thermograms obtained from a second heating are shown (open symbols) in Fig. 3. The glass transition temperatures were determined to be between 30°C and 40°C.

The 1:3 binary phase diagram constructed from DSC thermograms for physical mixtures is shown in Fig. 4. A eutectic exists a t a composition of -30% wlw of 1. The phase diagram was complicated due to the existence of two poly- morphic forms in the 1 powder, a point which will be discussed later.

According to Vasil'ev,*O this type of phase diagram arises when the strength of bonding (U) between two compounds, A and B (1 and 3, respectively), is in the order Ub > Uk4 > UbB for the liquid state, and U k > U2B > 0 for the solid state. The first inequality is necessary for the complete miscibility in the liquid state. An essential condition for the formation of limited solid solutions is the presence of some bonding between different molecules (U2B > 0). However, it should be recognized that 3 is amphiphilic and steric factors may now be decisive. Even so, 3, which has a much lower solubility than 1, may have the potential to reduce the release of 1 into solution.

The possible existence of polymorphism in the 1 powder used to prepare 1:3 physical mixtures was studied further. Using a system containing 15% wlw of 1 and 85% wlw of 3, various combinations of raw materials and/or melts and a comelt were examined by DSC (Fig. 5). Thermogram A resulted from the physical mixture of the two raw materials. Changing 3 from a raw material to a melt reslllted in a similar thermogram (B). In contrast, changing 1 from a raw material to a melt changed the peak shape from A and B to C and D. This indicates that the three peaks in the original thermogram were due to 1, rather than 3, raw material.

Thermograms C and D were obtained from physical mixtures. When, as here, a mixed sample capable of forming a eutectic is heated, it requires a temperature higher than the eutectic temperature to form the melt. This is because the individual melting points of the two compounds are above the eutectic temperature. Accordingly, the endotherm for the melt is broader than the eutectic physical mixture due to the time (or temperature) delay beyond the eutectic point.2' This accounts for the slightly different shape of the comelt (thermogram E, Fig. 5).

At a cooling rate of 2.5"C/min, eutectic mixtures lying on the 1 side of the eutectic point formed glass solutions. This was evidenced by the transparency and brittleness of the sample in the test tube, and the disappearance of the crystallization exotherm from the DSC thermogram during cooling.

r-

e . - . - - . I 80 6b 40 20

1, wt. 010

Flgure &Phase diagram of 1:3 physical mixtures constructed from DSC data.

11111111 120 140 160 180

Temperature, "C Flgure 5- Differential scanning calorimetry thermograms of 1:3 (15:85) physical mixfures on first heating: (dH/dt is heat flux in mcal. s- ; sample size = 5 mg; scan rate = 5 "C min- '). Key: (A) 1 (raw material) and 3 (raw material); (B) 1 (raw material) and 3 (melt); (C) 1 (melt) and 3 (melt); (D) 1 (melt) and 3 (raw material); and (E) 1 3 comelt.

The glass transition temperatures (T,) of these samples were obtained from the DSC thermograms. In order to enhance the second-order transition in the thermogram, the sample was shock-cooled in situ to -32"C, then reheated at 2O"Clmin. "he glass transition temperature of 1:3 comelts as a function of the weight percent of 1 is shown in Fig. 6. As the amount of 3 in the comelt increases, Tg decreases. According to Sum- mers and Enever,22 for a series of glasses containing the same constituents at a fixed temperature above the transition points of all glasses, the glasses possessing the lowest transition temperature will be the least viscous. The lower the viscosity of the melt, the easier it is to crystallize. This provides an explanation for the finding that as the amount of 3 in the comelt increased, the crystallization rate increased.

Dapsone:Dilauryl Derivative of Dapsone-Peak temperatures from the DSC thermograms of 1:2 physical mixtures are plotted as a function of 1 in Fig. 7. In the region of 10- 20% wlw of 1, no endothermic peaks corresponding to the steep liquidus line were observed. This is due to the small amount of excess 1 melting over a range of temperature, which results in a small, broad endothermic peak, undiffer-

20-1 100 80 60 40

1, wt. Yo Flgure 6-Glass transition temperature of 1:3 comelts.

Journal of Pharmaceutical Sciences / 55 Vol. 75, No. 1, January 1986

%!

q c .

D

...

G

3 80 60 40 20 1, wt. 010

Flgure 7-Phase diagram of 1:3 physical mixtures: (0) from DSC; (I) from hot-stage microscope. Liquid 7 contains 79% WAV of 1 in 2; liquid 2 contains 7% w/w of 2 in 1. Key: (A) liquid; (6) liquid 7 plus liquid 2; (C) solid 1 plus liquid 2; (D) solid 1 plus liquid; ( E ) liquid; (F) solid 2 plus liquid; (G) solid 1 plus solid 2.

entiated from the base line. Accordingly, a micro hot-stage microscope was used to observe the melting points of the physical mixtures containing 12.5 and 15% wlw of 1 (range symbols in Fig. 7). A monotectic and a eutectic system coexist in this phase diagram. The eutectic contained 11% wlw of 1 and melted at -143°C. The phase boundary between solid 1 and liquid and two conjugate liquids is located at 1735°C; the compositions of the latter are 19% of 1:81% of 2 (liquid 1) and 7% of 2:93% of 1 (liquid 21, respectively. At this temperature, each sample containing >19% by weight of 1 coexisted in three phases (solid 1 and two liquids); the system was invariant.

The degree of mutual miscibility in the liquid state may affect thedissolution rate. For this type of phase diagram, the bonding strength (U) between compounds A and B (1 and 2 in this instance) is in the order Ub > Ub > 0 for the liquid state and U h = 0 for the solid ~tate.2~ Mutual miscibility increases as the strength of bonding between different mol- cules ( U h ) increases. A solution of liquid l in liquid 2 is possible only up to a limit which corresponds to the permissi- ble degree of deformation of the energy field of the solvent. If two compounds are immiscible in the liquid state, they will not disperse well within each other in the solid state after coaling. Under these circumstances, the particle size of one compound dispersed in the other will depend on the degree of agitation during cooling. On the other hand, two completely miscible comopunds will solidify into a well-dispersed comelt; some solute can even disperse in solvent a t the molecular level.

In conclusion, the potential value of any carrier to reduce the dissolution of the molecule being carried will depend primarily on the ability of the former to interact with the latter. The partial miscibility of 1 and 2 in the liquid state is an indication of relatively weak bonding between these two molecules that would be expected to lead simply to the physical entrapment of 1 in 2. On the other hand, 3 and 1 are completely miscible in the liquid state, demonstrating a relatively strong bonding between this pair of molecules. Accordingly, 3 can be considered as a potentially better carrier than 2 when attempting to reduce the dissolution rate of 2 from solid dispersions.

1.

2. 3.

4.

5.

6. 7. 8.

9.

10. 11.

12.

13. 14.

15.

References and Notes Ellard, G. A.; Gammon, P. T.; Helmy, H. S.; Rees, R. J. W. Am. J . Tro . Med. 1974,23,464. Low, J. M.; Pearson, J. M. H. Lepr. Rev. 1974,45,218. Pearaon, J. M. H.; Haile, G. S.; Barnetson, R. S. C.; Rees, R. J. W. Lepr. Rev. 1979,50, 183. Sekiguchi, K.; Obi, N.; Ueda, Y. Chem. Pharm. Bull. 1964,12, 134. Goldber A. H.; Gibaldi, M.; Kanig, J. L.; Mayersohn, M. J . Phurm. Sci. 1966,55,581. Chiou, W. L.; Rie elman, S. J . Phurrn. Sci. 1971,60, 1281. S'ogren, J. Actu €?hum. Suec. 1971,8, 153. d u b , F.; Simmonelli, A. P.; Higuchi, W. I. J . Phurm. Sci. 1972, 61, 1381. Goodhart, F. W.; McCoy, R. H., Ninger, F. C. J . Pharrn. Sci. 1974.63. 1748. Swkbrick J. U.S. Patent 4 461 776. 1984.

S.; Stephenson, D. Bwihem: J Miura, K.; Bando, Y. J . Pharrn. Soc: Japan 1943,63, 75. Grady, L. T.; Hays, S. E.; King, R. H.; Klein, .H. R.; Mader, W. J.; Wyatt, D. K.; Zimnerer, R. 0. J . Phurm. Scz. 1973, 62,456. F$Ed,ert-Brandstatter, M.; Moser, I. Mikrochim. Actu 1979, L O i l l l J .

16. Kuhnert-Brandstatter, M.; Moser, I. Mikrochim. Actu 1981, 421(D.

17. BeGabei, M. I.; Gramberini, G.; Ferioli, V.; Cameroni, R. Far-

18. Robinson, P. M.; Scott H. G. Nature (London) 1972,238, 14. 19. 20. 21. Visser, M. J.; Wallace, W. H. Du Pont Thermogram 1966,3, 9. 22. Summers. M. P.; Enever, R. P. J . Phurm. Sci. 1977,615,825.

m o Ed. Pmt. 1982,37, 55.

R. L. Instrum. News 1975,21, 1. asil'ev, M. E. R w s J . Phys. Chem. 1964,38,473.

Acknowledgments This investigation received financial su ort from the Chemo-

therapy of Le rosy (THELEP) component of%e UNDPNorld Bank/ WHO Speciaf Programme for Research and Training in Tropical Diseases.

56 /Journal of Pharmaceutical Sciences Vol. 75, No. 1, January 1986

Documents

Sustained-release delivery systems, I: Phase diagram studies of dapsone and selected derivatives