- Email: [email protected]
Effect of Shear on the Apparent Viscosity of Amorphous Carbonate Ion Containing Aluminum Hydroxide Suspensions
Effect of Shear on the Apparent Viscosity of Amorphous Carbonate Ion Containing Aluminum Hydroxide Suspensions ELAINE M. MOREFIELD', THOMAS J. KONECHNIK', GARNET E. PECK*,JOSEPH R. FELOKAMP~, JOE L. WHITE*,AND STANLEY L. HEM" Received February 15, 1985, from the Depatiments of 'Industrial and Physical Pharmacy and *Agronomy, Purdue University, West Lafayetfe, IN 47907. Accepted for publication December 3, 1985. __
_----
~~~
Abstract 0 The application of shear to carbonate ion containing
aluminum hydroxide suspensions caused a change in the apparent viscosity by two possible mechanisms: change in the surface charge because of desorption of specifically adsorbed carbonate ion, and aggregate dispersal and formation of more extensive particle networks. The desorption of specifically adsorbed carbonate ion is related to the expansion of the air-liquid interface during shear. Shear-inducing processing equipment which generates a minimal amount of new airliquid interface was found to produce the least change in pH and, consequently, in surface charge. However, viscosity increases caused by aggregate dispersal and formation of more extensive particle networks may occur without a shear-induced change in surface charge. ~
Amorphous carbonate ion containing a l u m i n u m hydroxide is a component of m a n y commercial antacid suspensions.' The physical properties of suspensions are influenced by formulation a n d processing factors. A n u m b e r of studies have examined the effects of formulation factors on the properties of amorphous carbonate ion containing a l u m i n u m hydroxide s u s p e n s i o n s . ~ S H o w e v e r , l i t t l e w o r k has been reported on the effect of processing factors. The shear induced by processing equipment is expected to influence the degree of dispersion of the suspension which m a y be important i n terms of physical properties such as the a p p a r e n t viscosity. Therefore, an investigation of the effect of shear o n the apparent viscosity of amorphous carbonate ion containing aluminum hydroxide suspensions w a s undertaken.
Experimental Section Two amorphous carbonate ion containing aluminum hydroxide suspensions (Chattem Chemical Co., Chattanooga, TN; lot numbers 1451 and 1452) were diluted with distilled water to obtain a volume fraction of 0.055. The two lots will be identified as lots A and B, respectively. Samples were prepared a t various pH values by adjustment with either HCl o r NaOH. For example, <3 mL of 1M HCI was required to adjust 1 L of the 0.055 volume-fraction suspension to pH 5.8. The equivalent aluminum oxide content was determined by chelatometric titration. lo A gasometric technique was used to determine the carbonate content.l1 The point of zero charge (PZC) was determined by potentiometric titration.'* The density of the suspensions was determined by a pycnometer. The ratio of carbonate to aluminum ions was used to describe the empirical formula of the carbonate ion containing aluminum hydroxide in order to calculate the volume fraction of the suspensions. The apparent viscosity was determined by a rotational viscometer (model LVT, Brookfield Engineering Laboratories, Inc., Stoughton, MA) using spindle #1 at 12 rpm. The mean of three measurements was reported. Suspensions of lot A a t different pH conditions were exposed to high-intensity shear in a blender (model FCT 14, Waring Corp., New Hartford, CT) a t 10,000 rpm. The blender was stopped every 10 min, and the pH and apparent viscosity was measured after the temperature returned to 25°C. This intermittent exposure to shear in the blender minimized the temperature increase to
Homo Mixer, Gifford-WoodCo., New York, NY). The fixed clearance sheadmixer was stopped every 5 min, and the pH and PZC were measured after the temperature returned to 25°C. This intermittent exposure to shear also limited the temperature increase to < 5 T . The effects of various shear-inducing processing devices on the pH and apparent viscosity of lot A were determined by exposure to a fixed clearance shear/mixer (Eppenbach Homo Mixer, Gifford-Wood Co., New York, NY) for 5 min; a high-intensity low-pressure shear cell (model HS-2, HS-ZC nozzles, Gaulin Corp., Everette, MA) for one pass; a water-cooled colloid mill (Tri-Homo Disperser, Tri-Homo Corp., Salem, MA) a t a gap setting of 76.2 pm and a rotor speed of 6000 rpm for one pass; a sonicator (model W-370, Heat Systems, Ultrasonic, Inc., Plainview, NY) at a power setting of 4 for 5 min; and a homogenizer (Gaulin Corp., Everett, MA) a t 4500xg for one pass.
Results and Discussion A s seen i n Fig. 1, high-intensity shear caused an increase i n the apparent viscosity of carbonate ion containing alumin u m hydroxide suspensions, although the magnitude of the effect w a s dependent o n the initial pH. The suspensions which were adjusted to p H 5.80, 7.80, and 8.50 showed little increase i n apparent viscosity d u r i n g the 3.5-h exposure period. Suspensions at p H 6.30, 6.80, a n d 7.20 all exhibited significant increases i n apparent viscosity upon exposure to shear with the sample adjusted to p H 6.30 showing the largest increase. The d a t a in Fig. 1 also indicate that p H affects the apparent viscosity as the apparent viscosity before shear w a s applied varied from 17 t o 213 cps. Recent studies h a v e related the apparent viscosity of carbonate ion containing aluminum hydroxide suspensions to the surface charge as determined by the PZC-pH relationship.12 The apparent
I-
200 YI
" F 100
e 0
~
"
"
" 50"
"
" WEARING " "I00" TIME, ' ~ min' 150
200
Flgure 1--Effect of high intensity shear on the apparent viscosity, Q, of a 0.055 volume-fraction suspension of lot A. Key: (0) initial pH 5.80;(0) 6.30; ( ) 6.80; (A) 7.20;(0)7.80; (V) 8.50. Journal of Pharmaceutical Sciences / 297 Vol. 75, No. 3, March 1986
viscosity increased as the pH approached the PZC. This behavior was attributed to the formation of particle networks due to attractive interactions. These studies have shown that particle interactions in carbonate ion containing aluminum hydroxide suspensions are not symmetrical around the PZC. For example, the apparent viscosity did not drop as rapidly when the pH was above the PZC as when the pH was below the PZC.12 In addition, fiber optic Doppler anemometry revealed that the attractive particle network did not disperse as readily when the pH was above the PZC as when the pH was below the PZC.13 The effect of pH on apparent viscosity was investigated by preparing a series of suspensions of lot A at a volume fraction of 0.055 but having pH's ranging from 5.80 to 8.80 (Fig. 2).As seen in Fig. 2, the apparent viscosity of lot A was also related to the PZC-pH relationship. In agreement with earlier studies,12J3 this relationship was asymmetrical in terms of the PZC as the apparent viscosity exhibited a maximum from pH 6.70 to 7.80 even though the PZC was 6.75. Since the apparent viscosity of lot A ranged from -20-200 cps as a function of (PZC-pH) (Fig. 21, the effect of highintensity shear on the pH and PZC was investigated. The pH increased during shearing when the initial pH was between pH 5.80 and 7.80 (Fig. 3). The samples at pH 7.80 and 8.50 showed an initial decrease in pH; the sample at pH 7.80 showed a net increase in apparent viscosity over the 3.5-h
7
8
I
6.0
D
.
I
I
71)
-
.
I
I
~~~~1
I
8.0
pH
v
V
Y
d
*,
8.0 7.6
period, while the sample a t pH 8.50 remained below the initial value. The PZC of all the samples increased as a result of exposure to high intensity shear (Table I). Since the surface potential, & (in mV a t 25'0, is determined by the point of zero charge-pH relationship according to the Nernst equation14J6(eq. 11, the value of (PZC-pH) was determined before and after 3.5h of high-intensity shear and is presented in Table I. $0 =
RT
F In [ ( ~ H + ) / ( ~ H + ) P z c=] 5WZC-pH)
The effect of high-intensity shear on the apparent viscosity observed in Fig. 1can be largely explained by changes in the surface charge as characterized by (PZC-pH). The sample adjusted to pH 5.8 had an initial (PZC-pH) of 0.95. Based on the relatively low initial viscosity, 17 cps, this surface charge was large enough to minimize attractive particle interactions. Shearing increased both pH and PZC causing (PZCpH) to decrease to 0.53. This change in surface charge produced a minimal increase in apparent viscosity which is consistent with the effect of change in (PZC-pH) shown in Fig. 2. The bar labeled A in Fig. 2 indicates a change in (PZC-pH) from 0.95 to 0.53 and is associated with a small increase in the apparent viscosity. Thus, a shear-induced change in surface charge accounts for the small increase in viscosity. The sample adjusted to pH 6.30 had an initial (PZC-pH) value of 0.45.Shearing caused the (PZC-pH) to pass through zero and reach a final value of -0.06. The apparent viscosity increased by a factor of 4, the largest increase produced by shear in any sample. This increase in apparent viscosity can also be largely explained by examining the bar labeled B in Fig. 2, which portrays the shear-induced change in (PZCpH). This (PZC-pH) region encompasses the steepest slope in the (PZC-pH) versus apparent viscosity relationship. The observation that the final apparent viscosity (225 cps) was slightly higher than predicted by Fig. 2 (185 cps) suggests that a second mechanism, in addition to a change in the surface charge, may be contributing to the apparent viscosity after shear. It is speculated that shear may have produced a temporary dispersal of the attractive particle network and that a more extensive network may have formed when shear was stopped. The apparent viscosity of the sample at pH 6.80 increased from 178 to 276 cps during the 3.5 h of high-intensity shear. This increase is also slightly more than predicted by the change in (PZC-pH) noted by bar C in Fig. 2. Thus, the two possible mechanisms of shear-induced viscosity change, i.e., change in surface charge and aggregate dispersal and formation of a more extensive particle network, are believed to be acting in the sample with an initial pH which almost coincided with the PZC. The apparent viscosity of the sample with an initial pH of 7.20 increased slightly as a result of shear, even though the Table I-Effect of Hlgh-Intensity Shear for 3.5 h on the Apparent Vlscoslty and Surface Charge as Determlned by (PZC-pH) of a 0.055 Volume-Fractlon Suspenslon of Lot A
72
& 6.8
PZC
PH
Initial 6.0
Final
Initial
~~~~
l
0
(1)
l
50
l
l
l
l
l
l
l
I
1
100 SHEARING TIME,
1
1
150
1
1
1
1
zoo
1
min
Flgure 3-€ffecr of high-intensity shear on the pH of a 0.055 volumefraction suspension of lot A. Key: (0) initial pH 5.80; (0)6.30;( ) 6.80; (A)7.20; (0)7.80;(0)8.50. 298 /Journal of Pharmaceutical Sciences Vol. 75, No. 3, March 1986
5.80 6.30 6.80 7.20 8o
Apparent viscosity, cps
("C-pH)
Final
Initial
Final
7.11 7.07 6.99 7.08 7.1 6.95
0.95 0.45 -0.05 -0.45 -1,05 - .75
-0.06 -0.50 -0.72
Initial
Final
17 68 178 213
22 225 276 267 161 99
~
6.58 7.13 7.49 7.80
6.75 6.75 6.75 6.75 6.75 6.75
0.53
-
.52
134
92
change in (PZC-pH) noted in Fig. 2, bar D, suggests that a small decrease should have occurred. The high initial apparent viscosity of this sample suggests that attractive particle interactions predominate a t pH 7.20. Thus, as was hypothesized for the samples at initial pH values of 6.30 and 6.80, both potential mechanisms of shear-induced viscosity change appear to be operating. The (PZC-pH) value decreased slightly during shear for the suspension initially a t pH 7.80. Bar E in Fig. 2 indicates that this decrease should produce a small increase in apparent viscosity. The increase in apparent viscosity which was observed after shear was slightly larger than predicted by the (PZC-pH) versus apparent viscosity relationship. It is believed that both mechanisms of shear-induced viscosity also operate at pH 7.8. The (PZC-pH) value also decreased during shear for the suspension initially a t pH 8.50. Bar F in Fig. 2 indicates that the small increase in apparent viscosity which was observed (Table I) can be attributed to the change in surface charge. A likely mechanism for the increase in both pH and PZC during shear is desorption of specifically adsorbed carbonate ion. Desorbed carbonate ion will form bicarbonate anion in the pH range from 5.8 to 8.1 by the equilibrium shown in eq. 2. Thus, an increase in pH is predicted.
COT
+ HzO + HCO, + OH-
mechanism creates large amounts of air-liquid interface. As seen in Fig. 4, the ratio of carbonate to aluminum ions decreased steadily from 0.35 to 0.30 during the 30 min of shear. The loss of carbonate ion was accompanied by an increase in pH and PZC (Fig. 5). The effect of shear on the surface charge of carbonate ion containing aluminum hydroxide may be a complicating factor during processing operations. It would be desirable to be able to apply shear without also causing a change in the surface charge. Therefore, several types of shearing devices were evaluated. Samples of lot A a t a volume fraction of 0.055 and PZC of 6.75 were adjusted to pH 5.7, 6.2, and 6.7 prior to processing. The pH and apparent viscosity were measured after the sample had returned to 25°C following processing. Table I1 ranks the processing equipment in order of decreasing effect on pH. The fixed clearance shear/ mixer produced the largest change in pH while the colloid mill, sonicator, and homogenizer caused almost no change in pH. These results are consistent with the hypothesis that the pH change produced by shear is due to the loss of carbon dioxide because of the increased air-liquid interface. Since the tixed clearance sheadmixer operates on a mass transport mechanism for inducing shear, it creates the greatest in-
(2)
Likewise, the PZC of carbonate ion containing aluminum hydroxide has been recently shown to be inversely related to the amount of specifically adsorbed carbonate ion.lB Thus, desorption of carbonate ion produces an increase in the PZC until all carbonate ion is desorbed and a PZC of 10.4 corresponding to gibbsite, i.e., Al(OH)3, is reached. Recent studies have shown that carbonate ion could be desorbed from amorphous carbonate ion containing aluminum hydroxide by purging with nitrogen17 or by dilution with water.18 Both studies concluded that a series of interrelated equilibria exist in the suspensions which involve specifically adsorbed carbonate ion, carbonate ion in solution, carbonic acid, and carbon dioxide gas. Thus, it is believed that shear causes an increase in the air-liquid interface which accelerates the evolution of carbon dioxide gas. The interrelated equilibria are again established as carbonate ion is desorbed from the aluminum hydroxide surface. The hypothesized mechanism for the shear-induced change in pH and PZC was tested by exposing a suspension of lot B a t a volume fraction of 0.055 to shear in a fixed clearance sheadmixer. The initial pH was 6.65 and the PZC was 6.90. A fixed clearance sheartmixer was selected because its shear
- 7.2 - 71 - 70 0 N
- 6.9 a
6.61
I
0
5
30 29
0
High-intensity low-pressure shear cell Colloid mill Sonicator 5
1
0
1
5
2
SHEARING TIME
0
2
5
3
0
(mid
Figure &Effect of shear in a fixedclearance shearlinixer on the carbonate content of a 0.055volume-fraction suspension of lot 6.
I
1
I
I
68
-
67
I66
30
Table ICEffect of Various Types of Shear-lnduclng Processlng Equlpment on the pH and Apparent Viscosity of Lot A at a Volume Fractlon of 0.055
Fixed-clearance shear/mixer
tt
I
10 15 20 25 SHEARING TIME (min)
-
Figure 5-€ffect of shear in a fixed-clearance shearhixer on the pH (0)and PZC (0)of a 0.055 volume-fraction suspension of lot 6.
Processor
.31
- 73
7 3 1
Homogenizer
Apparent Viscosity,
PH
CPS
Initial
Final
A
5.7 6.2 6.7 5.7 6.2 6.7 5.7 6.2 6.7 5.7 6.2 6.7 5.7 6.2 6.7
6.3 6.7 7.2 5.9 6.3 6.7 5.8 6.3 6.7 5.8 6.3 6.7 5.8 6.2 6.7
0.6 0.5 0.5 0.2 0.1 0.0 0.1 0.1 0.0 0.1 0.1 0.0 0.1 0.0
0.0
20 34 170 12 38 181 20 28 164 24 39 162 20 46 175
Final
A
19 124 231 19 64 203 19 38 175 23 74 199 26 135 271
-1 90 61 7 26 22 -1 10 11 -1 35 37 6 89 96
Journal of Pharmaceutical Sciences / 299 Vol. 75, No. 3, March 1986
crease in the air-liquid interface, thereby accelerating the evolution of carbon dioxide. In contrast, there is no opportunity to increase the air-liquid interface during the operation of a homogenizer. The increase in apparent viscosity noted in Table I1 can be correlated with the change in pH for the fixed clearance sheadmixer, the high-intensity low-pressure shear cell, and the colloid mill. However, the sonicator and homogenizer each produced a significant increase in apparent viscosity even though the pH was virtually unaffected by processing. This increase in apparent viscosity is believed to be caused by shear-induced aggregate dispersal and formation of a more extensive particle network. The effect of initial pH on the increase in viscosity produced by the various types of processing equipment (Table 11) is consistent with the results obtained when the suspension was sheared in a blender (Fig. 1). The pH of the sample a t pH 5.7 increased on processing, but little, if any, change in the apparent viscosity resulted. Under these pH and PZC conditions the surface charge is high enough to prevent attractive particle interactions from occurring. The surface charge is small enough at pH 6.2 or 6.7 to allow attractive particle interactions to occur. The increase in apparent viscosity under these pH conditions is due to a reduction in surface charge and/or aggregate dispersal and formation of a more extensive particle network if the shear increases the airliquid interface. Other manufacturing operations, such as transfer between tanks and bottle filling, will expose carbonate-containing aluminum hydroxide suspensions to shear. Therefore, consideration must be given to the potential for changes in the apparent viscosity due to either a change in surface charge and/or aggregate dispersal and formation of a more extensive particle network.
2. Kerkhof, Nicholas J.; Hem, Stanley L.; White, Joe L. J . Pharm. Sci. 1975,64,2030-2032. 3 . Heyd, Allen; Dhabhar, Dadi J . J.Pharm. Sci. 1975,64, 16971 fiQQ * Y Y Y .
4. Nail, Steven L.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1976,65,1192-1195. 5. Wang, Ming K.; White, Joe L.; Hem, Stanley L. J . Phurm. Sci. 1980,69,668-671. 6. Shah, Dhiren N.; White, Joe L.; Hem, Stanley L. J . Phurm. Sci. 1981,70, 1101-1104. 7. Vanderlaan, Roger K.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1982. ~ . 71. _ _ ,780-786. _ 8. Shah, Dhiren N.; Feldkamp, Joseph R.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1982,71,266-268. 9. Zapata, Mary I.; Feldkamp, Joseph R.; Peck, Garnet E.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1984,73, 3-8. 10. “U. S.Pharmaco oeia,” 21 rev.; U.S.Pharmacopoeia1 Convention: Rockville, 8D,1985; 30. 11. “Official Methods of AnaGsis of the Association of Official Analytical Chemists”; Horowitz, W., Ed.; Association of Official Analytical Chemists: Washingon, D.C., 1970; 139. 12. Feldkam , Joseph R.; Shah, D men N ,Meyer, usan L.; White, Joe L.; d m , Stanley L. J . Pharm. Sci. 1981,70,638-640. 13. Wu, Paul P.;Feldkamp, Joseph R.; White, Joe L.; Hem, Stanley L. J . Colloid Inte ace Sci. in press. 14. Van Raij, Bernar 0 ; Peech, Michael. Soil Sci. Soc. Amer. Proc. 1972,36,587-593. 15. Gast, Robert G. in “Minerals in Soil Environments”; Dixon, Joe B.; Weed, Sterling B., Eds.; Soil Science Society of America: Madison, WI, 1977;p 36 43 16. Scholtz, Edward C.;Peldkamp, Joseph R.; White, Joe L.; Hem, Stanley L. J. Pharm. Sci. 1985,74,478-481. 17. Scholtz. Edward C.: Feldkamo. Joseoh R.: White.’ Joe L.: Hem. Stanley L. J . Pharm. Sci. 198h; 73,909-212. 18. Kerkhof, Nicholas J.; White, Joe L.; Hem, Stanley L. J . Phurm. Sci. 1975,64,940-942.
g
ll
References and Notes
Acknowledgments
1. Hem, Stanley L,; White, Joe L.; Buehler, John D.; Luber, Joseph R.; Grim, Wayne M.; Lipka, Edward A. Am. J . Hosp. Pharm.
This study was supported in part by William H. Rorer, Inc. This report is Journal Pa er 10420, Purdue University Agricultural Experiment Station, 8 e s t Lafayette, IN 47907.
1982,39,1925-1930.
300 /Journal of Pharmaceutical Sciences Vol. 75, No. 3, March 1986
_----
~~~
Abstract 0 The application of shear to carbonate ion containing
aluminum hydroxide suspensions caused a change in the apparent viscosity by two possible mechanisms: change in the surface charge because of desorption of specifically adsorbed carbonate ion, and aggregate dispersal and formation of more extensive particle networks. The desorption of specifically adsorbed carbonate ion is related to the expansion of the air-liquid interface during shear. Shear-inducing processing equipment which generates a minimal amount of new airliquid interface was found to produce the least change in pH and, consequently, in surface charge. However, viscosity increases caused by aggregate dispersal and formation of more extensive particle networks may occur without a shear-induced change in surface charge. ~
Amorphous carbonate ion containing a l u m i n u m hydroxide is a component of m a n y commercial antacid suspensions.' The physical properties of suspensions are influenced by formulation a n d processing factors. A n u m b e r of studies have examined the effects of formulation factors on the properties of amorphous carbonate ion containing a l u m i n u m hydroxide s u s p e n s i o n s . ~ S H o w e v e r , l i t t l e w o r k has been reported on the effect of processing factors. The shear induced by processing equipment is expected to influence the degree of dispersion of the suspension which m a y be important i n terms of physical properties such as the a p p a r e n t viscosity. Therefore, an investigation of the effect of shear o n the apparent viscosity of amorphous carbonate ion containing aluminum hydroxide suspensions w a s undertaken.
Experimental Section Two amorphous carbonate ion containing aluminum hydroxide suspensions (Chattem Chemical Co., Chattanooga, TN; lot numbers 1451 and 1452) were diluted with distilled water to obtain a volume fraction of 0.055. The two lots will be identified as lots A and B, respectively. Samples were prepared a t various pH values by adjustment with either HCl o r NaOH. For example, <3 mL of 1M HCI was required to adjust 1 L of the 0.055 volume-fraction suspension to pH 5.8. The equivalent aluminum oxide content was determined by chelatometric titration. lo A gasometric technique was used to determine the carbonate content.l1 The point of zero charge (PZC) was determined by potentiometric titration.'* The density of the suspensions was determined by a pycnometer. The ratio of carbonate to aluminum ions was used to describe the empirical formula of the carbonate ion containing aluminum hydroxide in order to calculate the volume fraction of the suspensions. The apparent viscosity was determined by a rotational viscometer (model LVT, Brookfield Engineering Laboratories, Inc., Stoughton, MA) using spindle #1 at 12 rpm. The mean of three measurements was reported. Suspensions of lot A a t different pH conditions were exposed to high-intensity shear in a blender (model FCT 14, Waring Corp., New Hartford, CT) a t 10,000 rpm. The blender was stopped every 10 min, and the pH and apparent viscosity was measured after the temperature returned to 25°C. This intermittent exposure to shear in the blender minimized the temperature increase to
Homo Mixer, Gifford-WoodCo., New York, NY). The fixed clearance sheadmixer was stopped every 5 min, and the pH and PZC were measured after the temperature returned to 25°C. This intermittent exposure to shear also limited the temperature increase to < 5 T . The effects of various shear-inducing processing devices on the pH and apparent viscosity of lot A were determined by exposure to a fixed clearance shear/mixer (Eppenbach Homo Mixer, Gifford-Wood Co., New York, NY) for 5 min; a high-intensity low-pressure shear cell (model HS-2, HS-ZC nozzles, Gaulin Corp., Everette, MA) for one pass; a water-cooled colloid mill (Tri-Homo Disperser, Tri-Homo Corp., Salem, MA) a t a gap setting of 76.2 pm and a rotor speed of 6000 rpm for one pass; a sonicator (model W-370, Heat Systems, Ultrasonic, Inc., Plainview, NY) at a power setting of 4 for 5 min; and a homogenizer (Gaulin Corp., Everett, MA) a t 4500xg for one pass.
Results and Discussion A s seen i n Fig. 1, high-intensity shear caused an increase i n the apparent viscosity of carbonate ion containing alumin u m hydroxide suspensions, although the magnitude of the effect w a s dependent o n the initial pH. The suspensions which were adjusted to p H 5.80, 7.80, and 8.50 showed little increase i n apparent viscosity d u r i n g the 3.5-h exposure period. Suspensions at p H 6.30, 6.80, a n d 7.20 all exhibited significant increases i n apparent viscosity upon exposure to shear with the sample adjusted to p H 6.30 showing the largest increase. The d a t a in Fig. 1 also indicate that p H affects the apparent viscosity as the apparent viscosity before shear w a s applied varied from 17 t o 213 cps. Recent studies h a v e related the apparent viscosity of carbonate ion containing aluminum hydroxide suspensions to the surface charge as determined by the PZC-pH relationship.12 The apparent
I-
200 YI
" F 100
e 0
~
"
"
" 50"
"
" WEARING " "I00" TIME, ' ~ min' 150
200
Flgure 1--Effect of high intensity shear on the apparent viscosity, Q, of a 0.055 volume-fraction suspension of lot A. Key: (0) initial pH 5.80;(0) 6.30; ( ) 6.80; (A) 7.20;(0)7.80; (V) 8.50. Journal of Pharmaceutical Sciences / 297 Vol. 75, No. 3, March 1986
viscosity increased as the pH approached the PZC. This behavior was attributed to the formation of particle networks due to attractive interactions. These studies have shown that particle interactions in carbonate ion containing aluminum hydroxide suspensions are not symmetrical around the PZC. For example, the apparent viscosity did not drop as rapidly when the pH was above the PZC as when the pH was below the PZC.12 In addition, fiber optic Doppler anemometry revealed that the attractive particle network did not disperse as readily when the pH was above the PZC as when the pH was below the PZC.13 The effect of pH on apparent viscosity was investigated by preparing a series of suspensions of lot A at a volume fraction of 0.055 but having pH's ranging from 5.80 to 8.80 (Fig. 2).As seen in Fig. 2, the apparent viscosity of lot A was also related to the PZC-pH relationship. In agreement with earlier studies,12J3 this relationship was asymmetrical in terms of the PZC as the apparent viscosity exhibited a maximum from pH 6.70 to 7.80 even though the PZC was 6.75. Since the apparent viscosity of lot A ranged from -20-200 cps as a function of (PZC-pH) (Fig. 21, the effect of highintensity shear on the pH and PZC was investigated. The pH increased during shearing when the initial pH was between pH 5.80 and 7.80 (Fig. 3). The samples at pH 7.80 and 8.50 showed an initial decrease in pH; the sample at pH 7.80 showed a net increase in apparent viscosity over the 3.5-h
7
8
I
6.0
D
.
I
I
71)
-
.
I
I
~~~~1
I
8.0
pH
v
V
Y
d
*,
8.0 7.6
period, while the sample a t pH 8.50 remained below the initial value. The PZC of all the samples increased as a result of exposure to high intensity shear (Table I). Since the surface potential, & (in mV a t 25'0, is determined by the point of zero charge-pH relationship according to the Nernst equation14J6(eq. 11, the value of (PZC-pH) was determined before and after 3.5h of high-intensity shear and is presented in Table I. $0 =
RT
F In [ ( ~ H + ) / ( ~ H + ) P z c=] 5WZC-pH)
The effect of high-intensity shear on the apparent viscosity observed in Fig. 1can be largely explained by changes in the surface charge as characterized by (PZC-pH). The sample adjusted to pH 5.8 had an initial (PZC-pH) of 0.95. Based on the relatively low initial viscosity, 17 cps, this surface charge was large enough to minimize attractive particle interactions. Shearing increased both pH and PZC causing (PZCpH) to decrease to 0.53. This change in surface charge produced a minimal increase in apparent viscosity which is consistent with the effect of change in (PZC-pH) shown in Fig. 2. The bar labeled A in Fig. 2 indicates a change in (PZC-pH) from 0.95 to 0.53 and is associated with a small increase in the apparent viscosity. Thus, a shear-induced change in surface charge accounts for the small increase in viscosity. The sample adjusted to pH 6.30 had an initial (PZC-pH) value of 0.45.Shearing caused the (PZC-pH) to pass through zero and reach a final value of -0.06. The apparent viscosity increased by a factor of 4, the largest increase produced by shear in any sample. This increase in apparent viscosity can also be largely explained by examining the bar labeled B in Fig. 2, which portrays the shear-induced change in (PZCpH). This (PZC-pH) region encompasses the steepest slope in the (PZC-pH) versus apparent viscosity relationship. The observation that the final apparent viscosity (225 cps) was slightly higher than predicted by Fig. 2 (185 cps) suggests that a second mechanism, in addition to a change in the surface charge, may be contributing to the apparent viscosity after shear. It is speculated that shear may have produced a temporary dispersal of the attractive particle network and that a more extensive network may have formed when shear was stopped. The apparent viscosity of the sample at pH 6.80 increased from 178 to 276 cps during the 3.5 h of high-intensity shear. This increase is also slightly more than predicted by the change in (PZC-pH) noted by bar C in Fig. 2. Thus, the two possible mechanisms of shear-induced viscosity change, i.e., change in surface charge and aggregate dispersal and formation of a more extensive particle network, are believed to be acting in the sample with an initial pH which almost coincided with the PZC. The apparent viscosity of the sample with an initial pH of 7.20 increased slightly as a result of shear, even though the Table I-Effect of Hlgh-Intensity Shear for 3.5 h on the Apparent Vlscoslty and Surface Charge as Determlned by (PZC-pH) of a 0.055 Volume-Fractlon Suspenslon of Lot A
72
& 6.8
PZC
PH
Initial 6.0
Final
Initial
~~~~
l
0
(1)
l
50
l
l
l
l
l
l
l
I
1
100 SHEARING TIME,
1
1
150
1
1
1
1
zoo
1
min
Flgure 3-€ffecr of high-intensity shear on the pH of a 0.055 volumefraction suspension of lot A. Key: (0) initial pH 5.80; (0)6.30;( ) 6.80; (A)7.20; (0)7.80;(0)8.50. 298 /Journal of Pharmaceutical Sciences Vol. 75, No. 3, March 1986
5.80 6.30 6.80 7.20 8o
Apparent viscosity, cps
("C-pH)
Final
Initial
Final
7.11 7.07 6.99 7.08 7.1 6.95
0.95 0.45 -0.05 -0.45 -1,05 - .75
-0.06 -0.50 -0.72
Initial
Final
17 68 178 213
22 225 276 267 161 99
~
6.58 7.13 7.49 7.80
6.75 6.75 6.75 6.75 6.75 6.75
0.53
-
.52
134
92
change in (PZC-pH) noted in Fig. 2, bar D, suggests that a small decrease should have occurred. The high initial apparent viscosity of this sample suggests that attractive particle interactions predominate a t pH 7.20. Thus, as was hypothesized for the samples at initial pH values of 6.30 and 6.80, both potential mechanisms of shear-induced viscosity change appear to be operating. The (PZC-pH) value decreased slightly during shear for the suspension initially a t pH 7.80. Bar E in Fig. 2 indicates that this decrease should produce a small increase in apparent viscosity. The increase in apparent viscosity which was observed after shear was slightly larger than predicted by the (PZC-pH) versus apparent viscosity relationship. It is believed that both mechanisms of shear-induced viscosity also operate at pH 7.8. The (PZC-pH) value also decreased during shear for the suspension initially a t pH 8.50. Bar F in Fig. 2 indicates that the small increase in apparent viscosity which was observed (Table I) can be attributed to the change in surface charge. A likely mechanism for the increase in both pH and PZC during shear is desorption of specifically adsorbed carbonate ion. Desorbed carbonate ion will form bicarbonate anion in the pH range from 5.8 to 8.1 by the equilibrium shown in eq. 2. Thus, an increase in pH is predicted.
COT
+ HzO + HCO, + OH-
mechanism creates large amounts of air-liquid interface. As seen in Fig. 4, the ratio of carbonate to aluminum ions decreased steadily from 0.35 to 0.30 during the 30 min of shear. The loss of carbonate ion was accompanied by an increase in pH and PZC (Fig. 5). The effect of shear on the surface charge of carbonate ion containing aluminum hydroxide may be a complicating factor during processing operations. It would be desirable to be able to apply shear without also causing a change in the surface charge. Therefore, several types of shearing devices were evaluated. Samples of lot A a t a volume fraction of 0.055 and PZC of 6.75 were adjusted to pH 5.7, 6.2, and 6.7 prior to processing. The pH and apparent viscosity were measured after the sample had returned to 25°C following processing. Table I1 ranks the processing equipment in order of decreasing effect on pH. The fixed clearance shear/ mixer produced the largest change in pH while the colloid mill, sonicator, and homogenizer caused almost no change in pH. These results are consistent with the hypothesis that the pH change produced by shear is due to the loss of carbon dioxide because of the increased air-liquid interface. Since the tixed clearance sheadmixer operates on a mass transport mechanism for inducing shear, it creates the greatest in-
(2)
Likewise, the PZC of carbonate ion containing aluminum hydroxide has been recently shown to be inversely related to the amount of specifically adsorbed carbonate ion.lB Thus, desorption of carbonate ion produces an increase in the PZC until all carbonate ion is desorbed and a PZC of 10.4 corresponding to gibbsite, i.e., Al(OH)3, is reached. Recent studies have shown that carbonate ion could be desorbed from amorphous carbonate ion containing aluminum hydroxide by purging with nitrogen17 or by dilution with water.18 Both studies concluded that a series of interrelated equilibria exist in the suspensions which involve specifically adsorbed carbonate ion, carbonate ion in solution, carbonic acid, and carbon dioxide gas. Thus, it is believed that shear causes an increase in the air-liquid interface which accelerates the evolution of carbon dioxide gas. The interrelated equilibria are again established as carbonate ion is desorbed from the aluminum hydroxide surface. The hypothesized mechanism for the shear-induced change in pH and PZC was tested by exposing a suspension of lot B a t a volume fraction of 0.055 to shear in a fixed clearance sheadmixer. The initial pH was 6.65 and the PZC was 6.90. A fixed clearance sheartmixer was selected because its shear
- 7.2 - 71 - 70 0 N
- 6.9 a
6.61
I
0
5
30 29
0
High-intensity low-pressure shear cell Colloid mill Sonicator 5
1
0
1
5
2
SHEARING TIME
0
2
5
3
0
(mid
Figure &Effect of shear in a fixedclearance shearlinixer on the carbonate content of a 0.055volume-fraction suspension of lot 6.
I
1
I
I
68
-
67
I66
30
Table ICEffect of Various Types of Shear-lnduclng Processlng Equlpment on the pH and Apparent Viscosity of Lot A at a Volume Fractlon of 0.055
Fixed-clearance shear/mixer
tt
I
10 15 20 25 SHEARING TIME (min)
-
Figure 5-€ffect of shear in a fixed-clearance shearhixer on the pH (0)and PZC (0)of a 0.055 volume-fraction suspension of lot 6.
Processor
.31
- 73
7 3 1
Homogenizer
Apparent Viscosity,
PH
CPS
Initial
Final
A
5.7 6.2 6.7 5.7 6.2 6.7 5.7 6.2 6.7 5.7 6.2 6.7 5.7 6.2 6.7
6.3 6.7 7.2 5.9 6.3 6.7 5.8 6.3 6.7 5.8 6.3 6.7 5.8 6.2 6.7
0.6 0.5 0.5 0.2 0.1 0.0 0.1 0.1 0.0 0.1 0.1 0.0 0.1 0.0
0.0
20 34 170 12 38 181 20 28 164 24 39 162 20 46 175
Final
A
19 124 231 19 64 203 19 38 175 23 74 199 26 135 271
-1 90 61 7 26 22 -1 10 11 -1 35 37 6 89 96
Journal of Pharmaceutical Sciences / 299 Vol. 75, No. 3, March 1986
crease in the air-liquid interface, thereby accelerating the evolution of carbon dioxide. In contrast, there is no opportunity to increase the air-liquid interface during the operation of a homogenizer. The increase in apparent viscosity noted in Table I1 can be correlated with the change in pH for the fixed clearance sheadmixer, the high-intensity low-pressure shear cell, and the colloid mill. However, the sonicator and homogenizer each produced a significant increase in apparent viscosity even though the pH was virtually unaffected by processing. This increase in apparent viscosity is believed to be caused by shear-induced aggregate dispersal and formation of a more extensive particle network. The effect of initial pH on the increase in viscosity produced by the various types of processing equipment (Table 11) is consistent with the results obtained when the suspension was sheared in a blender (Fig. 1). The pH of the sample a t pH 5.7 increased on processing, but little, if any, change in the apparent viscosity resulted. Under these pH and PZC conditions the surface charge is high enough to prevent attractive particle interactions from occurring. The surface charge is small enough at pH 6.2 or 6.7 to allow attractive particle interactions to occur. The increase in apparent viscosity under these pH conditions is due to a reduction in surface charge and/or aggregate dispersal and formation of a more extensive particle network if the shear increases the airliquid interface. Other manufacturing operations, such as transfer between tanks and bottle filling, will expose carbonate-containing aluminum hydroxide suspensions to shear. Therefore, consideration must be given to the potential for changes in the apparent viscosity due to either a change in surface charge and/or aggregate dispersal and formation of a more extensive particle network.
2. Kerkhof, Nicholas J.; Hem, Stanley L.; White, Joe L. J . Pharm. Sci. 1975,64,2030-2032. 3 . Heyd, Allen; Dhabhar, Dadi J . J.Pharm. Sci. 1975,64, 16971 fiQQ * Y Y Y .
4. Nail, Steven L.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1976,65,1192-1195. 5. Wang, Ming K.; White, Joe L.; Hem, Stanley L. J . Phurm. Sci. 1980,69,668-671. 6. Shah, Dhiren N.; White, Joe L.; Hem, Stanley L. J . Phurm. Sci. 1981,70, 1101-1104. 7. Vanderlaan, Roger K.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1982. ~ . 71. _ _ ,780-786. _ 8. Shah, Dhiren N.; Feldkamp, Joseph R.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1982,71,266-268. 9. Zapata, Mary I.; Feldkamp, Joseph R.; Peck, Garnet E.; White, Joe L.; Hem, Stanley L. J . Pharm. Sci. 1984,73, 3-8. 10. “U. S.Pharmaco oeia,” 21 rev.; U.S.Pharmacopoeia1 Convention: Rockville, 8D,1985; 30. 11. “Official Methods of AnaGsis of the Association of Official Analytical Chemists”; Horowitz, W., Ed.; Association of Official Analytical Chemists: Washingon, D.C., 1970; 139. 12. Feldkam , Joseph R.; Shah, D men N ,Meyer, usan L.; White, Joe L.; d m , Stanley L. J . Pharm. Sci. 1981,70,638-640. 13. Wu, Paul P.;Feldkamp, Joseph R.; White, Joe L.; Hem, Stanley L. J . Colloid Inte ace Sci. in press. 14. Van Raij, Bernar 0 ; Peech, Michael. Soil Sci. Soc. Amer. Proc. 1972,36,587-593. 15. Gast, Robert G. in “Minerals in Soil Environments”; Dixon, Joe B.; Weed, Sterling B., Eds.; Soil Science Society of America: Madison, WI, 1977;p 36 43 16. Scholtz, Edward C.;Peldkamp, Joseph R.; White, Joe L.; Hem, Stanley L. J. Pharm. Sci. 1985,74,478-481. 17. Scholtz. Edward C.: Feldkamo. Joseoh R.: White.’ Joe L.: Hem. Stanley L. J . Pharm. Sci. 198h; 73,909-212. 18. Kerkhof, Nicholas J.; White, Joe L.; Hem, Stanley L. J . Phurm. Sci. 1975,64,940-942.
g
ll
References and Notes
Acknowledgments
1. Hem, Stanley L,; White, Joe L.; Buehler, John D.; Luber, Joseph R.; Grim, Wayne M.; Lipka, Edward A. Am. J . Hosp. Pharm.
This study was supported in part by William H. Rorer, Inc. This report is Journal Pa er 10420, Purdue University Agricultural Experiment Station, 8 e s t Lafayette, IN 47907.
1982,39,1925-1930.
300 /Journal of Pharmaceutical Sciences Vol. 75, No. 3, March 1986