pubs.acs.org/JPCL
“pH Swing” in Frozen Solutions;Consequence of Sequential Crystallization of Buffer Components Prakash Sundaramurthi,† Evgenyi Shalaev,‡ and Raj Suryanarayanan*,† †
Department of Pharmaceutics, University of Minnesota, Minneapolis, Minnesota 55455, and ‡Pfizer Inc., Groton, Connecticut 06349
ABSTRACT Succinate buffer solutions of different initial pH values and concentrations were cooled. The solution pH and the phases crystallizing from solution were monitored as a function of temperature. In a solution buffered to pH 4.0 (200 mM), the freeze-concentrate pH initially increased to 8.0 and then decreased to 2.2. On the basis of X-ray diffractometry (synchrotron source), the “pH swing” was attributed to the sequential crystallization of succinic acid, monosodium succinate, and disodium succinate. A similar swing, but in the opposite direction, was seen when a solution with an initial pH of 6.0 was cooled. In this case, crystallization of the basic buffer component occurred first. The direction and magnitude of the pH shift depended on both the initial pH and the buffer concentration. In light of the pH-sensitive nature of a significant fraction of pharmaceuticals (especially proteins), extreme care is needed, both in the buffer selection and in its concentration. SECTION Biophysical Chemistry
T
he freezing behavior of aqueous solutions is a subject of both fundamental and technological importance and attracts attention from numerous disciplines including physical chemistry, biotechnology, cryobiology, pharmaceutical, and food science.1-3 Freezing is a critical step, both in the cryostabilization of biologicals and in the lyophilization (freeze-drying) of pharmaceuticals and food products.4,5 However, it is widely recognized that freezing can induce solute destabilization.1,6-9 Characterization of the freeze-concentrated liquid provides an avenue for understanding the destabilization mechanisms. In protein solutions, changes in pH, ionic strength, and solute concentration brought about by ice crystallization are known to cause destabilization. When a buffered solution is cooled, ice crystallization can be followed by selective crystallization of a buffer component, resulting in a significant pH shift (up to ∼3 pH units) in the freeze-concentrate.10 Such pH shifts are known to accelerate drug degradation in frozen solutions.11 van den Berg had investigated the pH shift under equilibrium conditions by seeding the frozen systems.12 However, during freeze-drying of pharmaceuticals, supersaturated solutions are often formed, and the systems are far from equilibrium.13 As a result, there is potential for crystallization of more than one buffering species with the potential to cause a “pH swing”. Here, we report such a phenomenon of an increase in pH followed by a decrease, or vice versa, in frozen succinate buffer systems due to sequential crystallization of buffer components. This observation was based on direct pH measurement coupled with X-ray diffractometry of the frozen systems.
r 2009 American Chemical Society
A low-temperature pH electrode, with a working temperature range of 80 to -30 °C, was placed in the center of a beaker. The buffer solutions, containing succinic acid and sodium succinate, were equilibrated at 0 °C and then cooled to -25 at 0.5 °C/min. Both the temperature and solution pH were monitored continuously. The phases crystallizing from the frozen solution were monitored by X-ray diffractometry (XRD), using both laboratory (Cu KR radiation) and synchrotron (Advanced Photon Source, Argonne National Laboratory) sources. Succinic acid [(CH2COOH)2] is a dicarboxylic acid with pKa values of 4.21 and 5.64 at room temperature (RT). Figure 1a shows the solution temperature and pH when a succinate buffer solution (200 mM; buffered to pH 4 at RT) was cooled. At ∼-8 °C, ice crystallization was evident from the abrupt increase in sample temperature. After a substantial fraction of the water had crystallized as ice, the sample temperature decreased again. The pH of the freeze-concentrate increased, slightly at first, and in a more pronounced manner as the sample temperature reached ∼-23 °C. By this time, the cooling was complete, and annealing had been initiated. While the sample and bath temperature remained approximately constant, the pH first exhibited an abrupt increase and then a sharp decrease. The small initial increase in pH, from 4.0 to ∼4.3, may be attributed to the temperature effect on pKa. When the temperature was decreased from 50 to 0 °C, the pKa1 of succinic acid increased from 4.19 to 4.29, while Received Date: October 20, 2009 Accepted Date: November 16, 2009 Published on Web Date: November 30, 2009
265
DOI: 10.1021/jz900164q |J. Phys. Chem. Lett. 2010, 1, 265–268
pubs.acs.org/JPCL
Figure 1. Low-temperature pH measurement of succinate buffer solution during cooling followed by isothermal hold at -25 °C. The bath and sample temperatures were simultaneously measured. The solutions had been buffered to pH values of 4.0 (a) and 6.0 (b) at RT.
Figure 2. Overlaid XRD patterns of 200 mM succinate buffer solution with an initial pH of 4.0, recorded during cooling (from bottom to top). Characteristic peaks of ice, β-succinic acid, and monosodium succinate are indicated.
Figure 3. The magnitude and direction of pH shift observed in the frozen buffer solutions. The solid horizontal lines at the bottom and top represent, respectively, the pH of 200 mM succinic acid and disodium succinate solution at RT. For the observed pH swing, the first pH shift is shown in bold typeface.
pKa2 was essentially unchanged.14,15 In addition to the temperature effect, an increase in buffering species concentration would cause a small increase in pKa values of succinic acid, though the magnitude of this effect is unknown.16-18 Moreover, the discussion of pKa is strictly valid only for dilute solutions,16-19 whereas the freeze-concentrates represent highly concentrated systems.13,20,21 From the speciation profile of succinic acid, it is evident that at pH 4, the solution will consist of unionized succinic acid (61% H2A) and monoprotic succinate (38% HA-). As a result of the freeze-concentration caused by ice crystallization, the system is likely to be supersaturated with respect to succinic acid. As seen in Figure 1a, the initial increase in the freezeconcentrate pH from ∼4.3 to 5.8 to 8.0 ((0.3) could then be attributed to crystallization of succinic acid, followed by monosodium succinate. XRD (Figure 2) provided direct evidence of sequential crystallization of ice (peaks at 22.7, 24.4, and 25.8° 2θ), succinic acid (20.0° 2θ), and monosodium succinate (14.0 and 17.8° 2θ). At the high pH value of ∼8.0, the freeze-concentrate is expected to be supersaturated with respect to disodium succinate (A2-). Crystallization, first of disodium succinate followed by that of monosodium succinate, can explain the sharp drop in pH from 8.0 to 2.2. However, there was no XRD
evidence of crystalline disodium succinate in the system. This can be explained by the fact that a substantial fraction of the solute had already crystallized, both as succinic acid and monosodium succinate. Therefore, the crystalline disodium succinate concentration in the frozen system may be below the detection limit of XRD. When the buffer solution (200 mM) with an initial pH value of 5.0 was cooled, the pH remained constant during initial ice crystallization (data not shown). As shown in Figure 3, the pH of the freeze-concentrate increased sharply to 7.1 ((0.2) and then decreased to 3.1 ((0.1). On the basis of XRD, this was attributed to the sequential crystallization of monosodium succinate followed by disodium succinate (data not shown). Interestingly, the latter crystallized as a hexahydrate (Na2(CH2COO)2 3 6H2O). There was no evidence of succinic acid crystallization in this system. Similar experiments were conducted in succinic acid solutions (200 mM) buffered to pH 6.0 (Figure 1b). A small increase in pH (0.3 units) was followed by a sharp drop, from 6.3 to 4.3 ((0.1), when the solution was cooled to -23 °C. Upon annealing at -25 °C for 100 min, there was a small increase in pH from 4.3 to 4.8 ((0.2). Upon annealing for another 50 min, the pH increased to 5.7 (data not shown). This initial pH drop followed by its increase is attributed to the
r 2009 American Chemical Society
266
DOI: 10.1021/jz900164q |J. Phys. Chem. Lett. 2010, 1, 265–268
pubs.acs.org/JPCL
Figure 4. Two-dimensional synchrotron XRD image of the succinate buffer (50 mM) with an initial pH value of 6.0 recorded after annealing at -25 °C for 20 min. Debye rings, unique to each component, have been pointed out.
sequential crystallization of disodium succinate, monosodium succinate, and succinic acid. The two-dimensional synchrotron XRD images recorded during annealing (50 mM) provided direct evidence of crystalline disodium succinate (as a hexahydrate), monosodium succinate, and succinic acid in the frozen system (Figure 4). To summarize, in solutions buffered to initial pH values of 4.0 or 5.0, the 200 mM buffer solution showed a pH swing, first in the basic and then in the acidic direction (Figure 3). Such a swing, reported for the first time during lyophilization, was attributed to the sequential crystallization of acidic followed by basic components of succinate buffer. A similar swing, but in the opposite direction, was seen when a solution with an initial pH of 6.0 was cooled. In this case, crystallization of the basic buffer component occurred first. The extent of pH shift depended both on the initial pH value and the buffer concentration. This “pH swing” was not evident at buffer concentrations e100 mM (Figure 3). The 100 and 50 mM solutions buffered to an initial pH value of 4.0 attained final values of 5.4 ((0.1) and 4.9 ((0.2), respectively, whereas the 200 mM solution reached 8.0 ((0.3). Solutions at initial concentrations of 200, 100, and 50 mM and buffered to pH 6.0 experienced net pH shifts of 1.7, 1.6, and 1.5 units, respectively. In summary, the sequential crystallization of succinate buffer components in frozen systems was characterized using low-temperature pH measurement and XRD. The phase crystallizing from solution and the consequential pH shift depended on both the initial pH and the buffer concentration. At high initial buffer concentrations (>100 mM), due to the sequential crystallization of the buffer components, a “pH swing” was observed. During lyophilization of pharmaceutical products, the buffer concentration is typically low (e50 mM). However, freeze-dried formulations are multicomponent
r 2009 American Chemical Society
systems which contain, in addition to the drug, several additives. The presence of other readily crystallizable solutes, such as glycine and mannitol, can facilitate buffer crystallization even when the initial buffer concentration is low. Moreover, NaCl, a commonly used additive to adjust solution osmolarity, can alter the ion activity product, which can further facilitate buffer salt crystallization. On the other hand, our ongoing work indicates that noncrystallizing excipients, such as sucrose and trehalose, could completely inhibit buffer crystallization.22 In order to develop an unambiguous understanding of the behavior of the buffer system, we did not include any model drug or other additives commonly found in freeze-dried pharmaceutical products. However, we recognize that the concentration and physical form of these components can influence the crystallization behavior of the buffer components and vice versa. The direct pH measurement allowed the continuous monitoring of frozen solution pH. The use of the synchrotron source enabled the identification of all of the crystallized phases. In the pharmaceutical community, in light of the pH sensitivity of a significant fraction of pharmaceuticals, the use of buffers is widespread. However, the potential for pH shift during lyophilization makes this practice questionable. Such pronounced shifts, brought about by selective crystallization of buffer components, can be avoided by excluding a buffer from the dosage form. This will also result in a simpler formulation, a trait particularly desirable in drug products intended for parenteral (injectable) administration.
SUPPORTING INFORMATION AVAILABLE Details of the experimental procedure, speciation of succinic acid as the function of pH at room temperature and at -25 °C, and the solubility of succinic acid as a function of temperature are provided. This material is available free of charge via the Internet at http://pubs.acs.org.
267
DOI: 10.1021/jz900164q |J. Phys. Chem. Lett. 2010, 1, 265–268
pubs.acs.org/JPCL
AUTHOR INFORMATION
(14)
Corresponding Author: *To whom correspondence should be addressed. Phone: 612-6249626. Fax: 612-626-2125. E-mail:
[email protected].
(15) (16) (17)
ACKNOWLEDGMENT S. Kumar (Kent State University), D.
Robinson (Argonne National Laboratory), and N. RodríguezHornedo (University of Michigan) are thanked for their help. Part of the work was conducted at the I.T. Characterization Facility (partial support from NSF).
(18) (19) (20)
REFERENCES (21)
(1)
Fennema, O. Reaction Kinetics in Partially Frozen Aqueous Systems. In Water Relations of Foods; Duckworth, R. B., Ed.; Academic Press: London, 1975; pp 539-556. (2) Takenaka, N.; Ueda, A.; Maeda, Y. Acceleration of the Rate of Nitrite Oxidation by Freezing in Aqueous Solution. Nature 1992, 358, 736–738. (3) Bauerecker, S.; Ulbig, P.; Buch, V.; Vrbka, L.; Jungwirth, P. Monitoring Ice Nucleation in Pure and Salty Water via HighSpeed Imaging and Computer Simulations. J. Phys. Chem. C 2008, 112, 7631–7636. (4) Heller, M. C.; Carpenter, J. F.; Randolph, T. W. Protein Formulation and Lyophilization Cycle Design: Prevention of Damage Due to Freeze-Concentration Induced Phase Separation. Biotechnol. Bioeng. 1999, 63, 166–174. (5) Dong, J.; Hubel, A.; Bischof, J. C.; Aksan, A. Freezing-Induced Phase Separation and Spatial Microheterogeneity in Protein Solutions. J.Phys. Chem. B 2009, 113, 10081–10087. (6) Hatley, R. H. M.; Franks, F.; Day, H. Subzero-Temperature Preservation of Reactive Fluids in the Undercooled State II. The Effect on the Oxidation of Ascorbic Acid of Freeze Concentration and Undercooling. Biophys. Chem. 1986, 24, 187–192. (7) Hatley, R. H. M.; Franks, F.; Day, H.; Byth, B. SubzeroTemperature Preservation of Reactive Fluids in the Undercooled State I. The Reduction of Potassium Ferricyanide by Potassium Cyanide. Biophys. Chem. 1986, 24, 41–46. (8) Reategui, E.; Aksan, A. Effects of the Low-Temperature Transitions of Confined Water on the Structure of Isolated and Cytoplasmic Proteins. J. Phys. Chem. B 2009, 113, 13048– 13060. (9) Takenaka, N.; Ueda, A.; Daimon, T.; Bandow, H.; Dohmaru, T.; Maeda, Y. Accelerated Mechanism of Chemical Reaction by Freezing: The Reaction of Nitrous Acid with Dissolved Oxygen. J. Phys. Chem. 1996, 100, 13874–13884. (10) Gomez, G.; Pikal, M. J.; Rodriguez-Hornedo, N. Effect of Initial Buffer Composition on pH Changes During Far-From-Equilibrium Freezing of Sodium Phosphate Buffer Solutions. Pharm. Res. 2001, 18, 90–97. (11) Lam, X. M.; Costantino, H. R.; Overcashier, D. E.; Nguyen, T. H.; Hsu, C. C. Replacing Succinate with Glycolate Buffer Improves the Stability of Lyophilized Interferon-γ. Int. J. Pharm. 1996, 142, 85–95. (12) van den Berg, L. pH Changes in Buffers and Foods During Freezing and Subsequent Storage. Cryobiol. 1966, 3, 236– 242. (13) MacKenzie, A. P. Non-Equilibrium Freezing Behavior of Aqueous Systems. Philos. Trans. R. Soc. London, Ser. B 1977, 278, 167–189.
r 2009 American Chemical Society
(22)
268
Pinching, G. D.; Bates, R. G. First Dissociation Constant of Succinic Acid from 0° to 50°C and Related Thermodynamic Quantities. J. Res. NBS 1950, 45, 444–449. Pinching, G. D.; Bates, R. G. Second Dissociation Constant of Succinic Acid from 0° to 50°C. J. Res. NBS 1950, 45, 322–328. Bates, R. G. Determination of pH: Theory and Practice; John Wiley & Sons, Inc.: New York, 1964. Bates, R. G.; Vosburgh, W. C. The Activity Coefficients of Cadmium Iodide. J. Am. Chem. Soc. 1937, 59, 1583–1585. Galster, H. pH Measurement: Fundamentals, Methods, Applications, Instrumentation, 1st ed.; VCH: New York, 1991. Kielland, J. Individual Activity Coefficients of Ions in Aqueous Solutions. J. Am. Chem. Soc. 1937, 59, 1675–1678. Paabo, M.; Bates, R. G.; Robinson, R. A. Buffer Solutions of Potassium Dihydrogen Phosphate and Sodium Succinate at 25°C. J. Res. NBS 1963, 67A, 573–576. Pikal, M. J. Freeze Drying. In Encyclopedia of Pharmaceutical Technology; Swarbrick, J., Boylan, J. C., Eds.; Marcel Dekker: New York, 2002; pp 1299-1326. Sundaramurthi, P.; Varshney, D.; Kumar, S.; Kang, S.-W.; Suryanarayanan, R. Effective Inhibition of Buffer Component Crystallization in the Frozen Solution by Lyoprotectants. AAPS J. 10, S2; AAPS annual meeting, Atlanta, GA, November 2008.
DOI: 10.1021/jz900164q |J. Phys. Chem. Lett. 2010, 1, 265–268