Glass-Rubber Transitions of Cellulosic Polymers by Dynamic Mechanical Analysis TUGRULT. KARARLI*',JEFFREY 6. HURLBUT*, AND THOMAS E. NEEDHAM* Received July 5, 1989, from *Searle Research and Development, 4901 Searle Parkway, Skokie, IL 60077, and SMonsanto Company, 730 Worcester, Street, Sprmgfield, MA 07 757. Accepted for publication November 14, 1989. Abstract The glass-rubber transition temperatures (T,) of several cellulosic polymers [hydroxypropyl methylcellulose (HPMC), hydroxypropyl cellulose (HPC), hydroxyethyl cellulose (HEC)] have been examined using dynamic mechanical analysis (DMA). The melting temperatures of the above polymers were examined using a hot stage melting point apparatus. The primary Tg of three different grades of HPMC (3,6, and 15 cps) were determined to be 160, 170, and 175 "C, respectively. The primary T, of the HEC film was determined as 120 "C. The HPC film did not indicate a primary T,. These cellulosic polymers also displayed secondary transitions. Hot stage melting of HPMC and HPC was observed at 225 to 254 "C and 190 to 195 "C, respectively. The HEC powder did not exhibit a melting temperature, but became darker at temperatures > 150 "C.
The mechanical strength of tablet coatings,' water sorption and diffusion,z stability,3 and many other physical performance-related properties are related to glass-rubber transitions in the amorphous materials used in pharmaceutical and food formulations. The collapse phenomenon in lyophilized products is considered to be the result of transition from the glass to rubber state in the formulation components.4 Cellulosic polymers, such as hydroxypropyl methylcellulose (HPMC), hydroxypropyl cellulose (HPC), and hydroxyethyl cellulose (HEC), are commonly used in the pharmaceutical industry. The glass-rubber transition temperatures (T ) of some cellulosic polymers have been investigated by speciffc volume measurements,5 differential scanning calorimetry (DSC),6-9torsional braid pendulum (TBP),s and thermomechanical analysis (TMAP techniques. The results of the studies for HPMC by DSC,6-9 TMA,7-8and TBPS techniques were somewhat variable. Further, the DSC technique was insensitive to detection of any secondary transitions in HPMC.9 For HPC, the TMA and dynamic mechanical analysis @MA) studies did not clearly establish a primary Tg.S011J2 To our knowledge, no glass-rubber transition is reported for HEC. Further, most of the literature data for cellulosic polymers were collected a t temperatures >O "C, except the study by Rials et al. for HPC.11 Therefore, there is no information in the literature regarding transitions in HPMC a t temperatures 220 >220 >220
225-252 239-254
-
-
-
190-1 95
>150
-
a By TMA (expansion mode) of the film cast from water (ref 9). By TMA (penetration mode) of the film cast from water (ref 9). 'By DSC of the film cast from dich1oromethane:methanol (50:50,v/v; ref 6). dBy DSC of the film cast from water (ref 7). By DSC of powder (ref 8). 'By DSC of the film cast from dich1oromethane:methanol (5050,v/v; ref 8). By differential thermal analysis (DTA) of powder (ref 8). By TBP (ref 8). By DSC of the film cast from water (ref 9).' By TMA (penetration mode) of the film cast from dich1oromethane:methanol:isopropylalcohol (4:3:3; ref 10). By DMA of the film cast from dioxane or acetone (ref 11). 'By DMA of the film cast from N,N-dimethylacetamide (ref 12).
2-2 t B I
1
\I
11
'h
I
.
'.
\
1011,
10'0 -
5>
100-
i a
Y
L
108-
these transitions correspond to P- and ytransitions, respectively. Overall, the observation of such secondary transitions is consistent-with the transitions measured in the literature using TBPS and DMA11J2 (Table I). Besides the secondary transitions, the literature also indicates other transitions for HPC at higher temperatures11.12 (Table I). Some of these transitions may correspond to the primary transition.12 The presence or absence of higher temperature transitions in HPC may stem from the use of different molecular weight samples and different solvents to prepare the films (Table I). The HEC film indicated a primary T, and at least two secondary transition peaks. The secondary transition temperatures which are listed in Table I may again be due to p and ytransitions. During the DMA measurements, the film showed residual stiffness following the primary transition at 120 "C. This is consistent with the hot stage melting determinations which showed that the HEC polymer does not melt a t temperatures up to 250 "C, but rather becomes darker. Not much is known about the relevance of the secondary transitions in the formulations. The small scale relaxation in the polymer chains and side groups caused by @transition can be important in the diffusion of relatively small molecules, such as oxygen and water. This may then affect the stability of formulations. In summary, by using the DMA technique, the primary and secondary glass-rubber transition temperatures of HPMC, HEC, and HPC films were determined. Among these polymers, only HPMC and HEC exhibited a primary T,, and all the polymers showed secondary transitions. Microscopic examination of the above cellulosic polymers revealed birefringency, and all the polymers except HEC had melting temperatures.
I id
10'
References and Notes
0
100
200
300
Temperaturs ('C)
Flgure &The DMA properties of HPC as a function of temperature. Key: (-) storage moduli (E');(- -) loss moduli (E"); (- .-) tan 6.
225°C. At such high temperatures, decomposition of the polymer is also possible. In our study, no primary T, was detected for HPC. The storage moduli (E') of the sample film dropped gradually until the film became deformed. However, the HPC film showed two secondary transitions, one between 0 and 40 "C and the second between -60 and - 100 "C. Again,
1. Okhamafe, A. 0.; York, P. J. Pharm. Pharmacol. 1983, 35, 409-415. 2. Berens, A.R.;Hopfenberg, H. B. J.Membr. Sci. 1982,10,283303. 3. Gejl-Hansen, F.; Flink, J. M. J. Food Sci. 1977,42,1049-1055. 4. Levine, H.; Slade, L. In Water Science Reviews, Vol. 3;Franks, F., Ed.; Cambridge University: Cambridge, U.K., 1988; pp 79-185. 5. Mandelkern, L.; Flory, P. J. J. Am. Chem. Soc. 1951,73,32063212. 6. Entwistle, C.A.;Rowe, R. C. J. Pharm. Pharmacol. 1979,31, 269-212. I . Okhamafe, A. 0.;York, P. Pharm. Res. 1985,2,19-23. 8. Sakellariou, P.;Rowe, R. C.; White, E. F. T. Int. J. Pharm. 1985, 27,261-211. 9. Okhamafe, A. 0.; York, P. J. Pharm. Sci. 1988,11,438-443. 10. Masilungan, F. C.; Lordi, N. G. Znt.J.Pharm. 1984,20,295-305. Journal of Pharmaceutical Sciences I 047 Vol. 79, No. 9, September 1990
11. Rials T. G.; Glasser, W. G. J . Appl. Polym. Sci. 1988, 36, 749-+58. 12. Suto, S: Kudo, M.; Karasawa, M. J . Appl. Polym. Sci. 1986,31, 1324-1d41. 13. Murayama, T. In Dynamic Mechanical Pro erties o Polymeric Materidls. Material Science Monographs; Egevier: dmsterdarn, 1978; Vol. 1. 14. Nielson, L. E. In Mechanical Pro erties of Polymers and Composites; Marcel Dekker: New Yo&, 1974; Vol. 1.
848 I Journal of Pharmaceutical Sciences I/nl
7a
&In
a
Cantamha.
idan
15. Boyer, R. F. In Polymeric Materials: Relationshi between Structure and Mechanical Behavior; Bear, E.; RadcEffe, S. V., Eds.; American Society for Metals: Metals Park, OH, 1974; pp 227368.
Acknowledgments The valuable discussions with Dr. R. Mendelson and Prof. G. Zografi are acknowledged.
The mechanical strength of tablet coatings,' water sorption and diffusion,z stability,3 and many other physical performance-related properties are related to glass-rubber transitions in the amorphous materials used in pharmaceutical and food formulations. The collapse phenomenon in lyophilized products is considered to be the result of transition from the glass to rubber state in the formulation components.4 Cellulosic polymers, such as hydroxypropyl methylcellulose (HPMC), hydroxypropyl cellulose (HPC), and hydroxyethyl cellulose (HEC), are commonly used in the pharmaceutical industry. The glass-rubber transition temperatures (T ) of some cellulosic polymers have been investigated by speciffc volume measurements,5 differential scanning calorimetry (DSC),6-9torsional braid pendulum (TBP),s and thermomechanical analysis (TMAP techniques. The results of the studies for HPMC by DSC,6-9 TMA,7-8and TBPS techniques were somewhat variable. Further, the DSC technique was insensitive to detection of any secondary transitions in HPMC.9 For HPC, the TMA and dynamic mechanical analysis @MA) studies did not clearly establish a primary Tg.S011J2 To our knowledge, no glass-rubber transition is reported for HEC. Further, most of the literature data for cellulosic polymers were collected a t temperatures >O "C, except the study by Rials et al. for HPC.11 Therefore, there is no information in the literature regarding transitions in HPMC a t temperatures 220 >220 >220
225-252 239-254
-
-
-
190-1 95
>150
-
a By TMA (expansion mode) of the film cast from water (ref 9). By TMA (penetration mode) of the film cast from water (ref 9). 'By DSC of the film cast from dich1oromethane:methanol (50:50,v/v; ref 6). dBy DSC of the film cast from water (ref 7). By DSC of powder (ref 8). 'By DSC of the film cast from dich1oromethane:methanol (5050,v/v; ref 8). By differential thermal analysis (DTA) of powder (ref 8). By TBP (ref 8). By DSC of the film cast from water (ref 9).' By TMA (penetration mode) of the film cast from dich1oromethane:methanol:isopropylalcohol (4:3:3; ref 10). By DMA of the film cast from dioxane or acetone (ref 11). 'By DMA of the film cast from N,N-dimethylacetamide (ref 12).
2-2 t B I
1
\I
11
'h
I
.
'.
\
1011,
10'0 -
5>
100-
i a
Y
L
108-
these transitions correspond to P- and ytransitions, respectively. Overall, the observation of such secondary transitions is consistent-with the transitions measured in the literature using TBPS and DMA11J2 (Table I). Besides the secondary transitions, the literature also indicates other transitions for HPC at higher temperatures11.12 (Table I). Some of these transitions may correspond to the primary transition.12 The presence or absence of higher temperature transitions in HPC may stem from the use of different molecular weight samples and different solvents to prepare the films (Table I). The HEC film indicated a primary T, and at least two secondary transition peaks. The secondary transition temperatures which are listed in Table I may again be due to p and ytransitions. During the DMA measurements, the film showed residual stiffness following the primary transition at 120 "C. This is consistent with the hot stage melting determinations which showed that the HEC polymer does not melt a t temperatures up to 250 "C, but rather becomes darker. Not much is known about the relevance of the secondary transitions in the formulations. The small scale relaxation in the polymer chains and side groups caused by @transition can be important in the diffusion of relatively small molecules, such as oxygen and water. This may then affect the stability of formulations. In summary, by using the DMA technique, the primary and secondary glass-rubber transition temperatures of HPMC, HEC, and HPC films were determined. Among these polymers, only HPMC and HEC exhibited a primary T,, and all the polymers showed secondary transitions. Microscopic examination of the above cellulosic polymers revealed birefringency, and all the polymers except HEC had melting temperatures.
I id
10'
References and Notes
0
100
200
300
Temperaturs ('C)
Flgure &The DMA properties of HPC as a function of temperature. Key: (-) storage moduli (E');(- -) loss moduli (E"); (- .-) tan 6.
225°C. At such high temperatures, decomposition of the polymer is also possible. In our study, no primary T, was detected for HPC. The storage moduli (E') of the sample film dropped gradually until the film became deformed. However, the HPC film showed two secondary transitions, one between 0 and 40 "C and the second between -60 and - 100 "C. Again,
1. Okhamafe, A. 0.; York, P. J. Pharm. Pharmacol. 1983, 35, 409-415. 2. Berens, A.R.;Hopfenberg, H. B. J.Membr. Sci. 1982,10,283303. 3. Gejl-Hansen, F.; Flink, J. M. J. Food Sci. 1977,42,1049-1055. 4. Levine, H.; Slade, L. In Water Science Reviews, Vol. 3;Franks, F., Ed.; Cambridge University: Cambridge, U.K., 1988; pp 79-185. 5. Mandelkern, L.; Flory, P. J. J. Am. Chem. Soc. 1951,73,32063212. 6. Entwistle, C.A.;Rowe, R. C. J. Pharm. Pharmacol. 1979,31, 269-212. I . Okhamafe, A. 0.;York, P. Pharm. Res. 1985,2,19-23. 8. Sakellariou, P.;Rowe, R. C.; White, E. F. T. Int. J. Pharm. 1985, 27,261-211. 9. Okhamafe, A. 0.; York, P. J. Pharm. Sci. 1988,11,438-443. 10. Masilungan, F. C.; Lordi, N. G. Znt.J.Pharm. 1984,20,295-305. Journal of Pharmaceutical Sciences I 047 Vol. 79, No. 9, September 1990
11. Rials T. G.; Glasser, W. G. J . Appl. Polym. Sci. 1988, 36, 749-+58. 12. Suto, S: Kudo, M.; Karasawa, M. J . Appl. Polym. Sci. 1986,31, 1324-1d41. 13. Murayama, T. In Dynamic Mechanical Pro erties o Polymeric Materidls. Material Science Monographs; Egevier: dmsterdarn, 1978; Vol. 1. 14. Nielson, L. E. In Mechanical Pro erties of Polymers and Composites; Marcel Dekker: New Yo&, 1974; Vol. 1.
848 I Journal of Pharmaceutical Sciences I/nl
7a
&In
a
Cantamha.
idan
15. Boyer, R. F. In Polymeric Materials: Relationshi between Structure and Mechanical Behavior; Bear, E.; RadcEffe, S. V., Eds.; American Society for Metals: Metals Park, OH, 1974; pp 227368.
Acknowledgments The valuable discussions with Dr. R. Mendelson and Prof. G. Zografi are acknowledged.
