`
`Technical Note
`
`The Influence of Free Fatty Acid
`Formation on the pH of
`Phospholipid-Stabilized
`Triglyceride Emulsions
`
`Clifford J. Herman” and Michael J. Groves”
`
`ever, the value of this suggestion is unclear since, during the
`phospholipid hydrolysis process, lyso compounds are pro-
`duced in addition to FFA and these materials also contribute
`to the emulsion stabilization process (3,6).
`Measurement of pH could, therefore, provide an indi-
`rect method of determining the FFA content of phospho-
`lipid-stabilized emulsions. This concept was evaluated mea-
`suring the FFA content directly by potassium hydroxide ti-
`tration of the degrading emulsion at the same time as taking
`the pH of the system with a glass electrode.
`
`Received November 2, I 992,‘ accepted December I, 1992
`.KEY WORDS: phospholipids; thermal degradation; triglyceride
`emulsions; pH; free fatty acids.
`
`MATERIALS AND METHODS
`
`Materials
`
`INTRODUCTION
`
`The intravenous administration of triglyceride emul-
`sions stabilized with phospholipid emulsifiers has been em-
`ployed for parenteral nutrition for over 30 years (1). Termi-
`nally heat sterilized, these systems are required to be phys-
`ically and chemically stable in order to avoid harming the
`patient (2). Nevertheless, slow hydrolysis of the phospho-
`lipids is known to occur after the initial sterilization-induced
`degradation (3).
`It is implicitly assumed that the subsequent fall of prod-
`uct pH is due to the formation of free fatty acids (4,5). We
`have recently demonstrated that the initial hydrolysis of the
`phospholipids during the heat sterilization process paradox-
`ically promotes physical stabilization of the emulsion sys-
`tem, most probably because of the formation of liquid crys-
`talline structures at the oil/water interface (3,6). The princi-
`pal degradation process is due to the hydrolysis of the
`diacylphosphatidylcholines and diacylphosphatidylethanola-
`mines to their corresponding monoacyl (lyso-) derivatives
`and free fatty acid (FFA) moieties. In turn, the lyso deriva-
`tives can degrade to the corresponding glycerophosphoryl
`compounds, with the formation of additional FFA. FFA can
`also be formed by the hydrolysis of emulsified triglycerides
`to the corresponding mono- and diglycerides, although this
`reaction is believed to be relatively slow compared to the
`breakdown of the diacylphosphatidyl derivatives (3).
`The emulsion systems are unbuffered and the formation
`of FFA will inevitably lower the pH from the initial value of
`8.0 (1) over a period of time poststerilization. Hakansson (2)
`demonstrated that the degradation rate decreases until pH
`6.5 is reached, after which there is again an acceleration of
`the degradation process. This effect has been confirmed by
`Grit et al. (7). Stabilization may, therefore, be improved by
`the addition of extraneous FFA, a suggestion made by
`Washington and Davis (4), who evaluated the effect induced
`by the addition of oleic acid to their emulsion systems. How-
`
`
`
`‘ Institute for Tuberculosis Research, College of Pharmacy, Univer-
`sity of Illinois at Chicago (M/C 964), 840 West Taylor (2014 SEL),
`Chicago, Illinois 60607.
`2 Present address: Mallinckrodt Specialty Chemicals Company,
`P.O. Box 5439, St. Louis, Missouri 63147.
`3 To whom correspondence should be addressed.
`
`Purified egg phospholipid, Asahi Injectable grade (lot
`900-80201), was received as a gift from Austin Chemical,
`Chicago, IL. Dipalmitoyl phosphatidylethanolamine (Lot 1
`60PE-45) (DPPE) and hydrogenated egg phosphatidylcholine
`(Lot HEPC — 44) (HEPC) were purchased from Avanti Polar
`Lipids, Birmingham, AL. Pharmaceutical-grade and “super-
`refined”-grade soybean oils USP were received as gifts from
`Croda Inc., Edison, NJ. Glycerol, sodium hydroxide, potas-
`sium hydroxide, and potassium hydrogen phthalate were all
`used as received from Fisher Scientific, Itasca, IL. A Milli-Q
`ion-exchange water system was used.
`
`Model Emulsion
`
`The model emulsion was made to the following formula:
`soybean oil, 20 g; egg phospholipid, 1.2 g; glycerol, 2.25 g-,
`and water to 100 mL.
`The egg phospholipid was dispersed in the glycerol and
`about 95% of the water at 70°C, the oil mixed in, and the
`coarse emulsion passed through a Microfluidics Model 110T
`homogenizer at a pressure of 10,000 psig for a total of 10
`times to ensure minimal particle size (8). The system was
`washed through the homogenizer and made up to volume
`with water, and the pH adjusted to 8.0 with 0.01 N sodium
`hydroxide and packed and sealed in 2-mL volumes in 2-mL
`glass ampoules. Gas sparging with nitrogen or oxygen was
`carried out for 15 min when required. Hydrogenated phos-
`pholipid-stabilized emulsions were prepared by the same
`method, using 0.24 g DPPE and 0.96 g HEPC instead of the
`1.2 g egg phospholipids.
`Samples were stressed and analyzed in replicate; n = 4.
`
`pH Measurement
`
`An Orion Model 811 glass electrode pH meter was used.
`Potassium chloride was added to counteract the ion adsorb-
`ing effects of charged droplets (9), a process also used in the
`USP XXII for the measurement of the pH of Dextrose so-
`lutions. After experimentation to determine conditions re-
`quired to give reproducible results, measurements were
`made following the addition of 50 [IL of a saturated aqueous
`potassium chloride solution to 2 mL of emulsion sample.
`
`Free Fatty Acid Measurement
`
`FFA measurement of a solution of the emulsion sample
`
`0724-8741/93/0500—0774S07.00/0 © 1993 Plenum Publishing Corporation
`
`774
`
`000001
`
`IAKBM 1038
`
`000001
`
`
`
`FFA Formation and pH of Triglyceride Emulsions
`
`was by direct titration to neutrality with 0.01 N potassium
`hydroxide solution standardized using potassium hydrogen
`phthalate, with phenolphthalein as indicator (USP XXII).
`The IUPAC method (10) specifies ethanolzdiethyl ether (95:
`5) as the solvent of choice. However, for safety reasons, the
`solvent was changed to ethanohchloroform (2:1), each titra-
`tion being preceded by bringing the solvent to neutrality with
`0.01 N potassium hydroxide solution.
`
`Thermal Stress
`
`Thermal stress was applied by filling heating block
`chambers (Dry Baths, Fisher Scientific, Itasca, IL; 60 cham-
`bers per block, each 12 mm diameter and 50 mm deep) with
`oil and immersing the 2-mL ampoules containing the emul-
`sion at the desired temperature, covering the blocks with
`aluminum foil to minimize thermal fluctuation. Tempera-
`tures were determined with calibrated mercury-in-glass ther-
`mometers placed at random in the block chambers.
`In general, all ampouled emulsion samples were initially
`sterilized at 121°C (F0 = 18) using a Getinge BioFoOE au-
`toclave. Unautoclaved controls were stored at 5°C prior to
`evaluation.
`
`RESULTS AND DISCUSSION
`
`775
`
`‘Logk
`
`2.4
`
`2.6
`
`2.8
`
`3
`
`3.2
`
`1/T x 1000
`
`Fig. 2. The rate of pH change in a 20% soy oil emulsion stabilized
`with egg lecithin or saturated phospholipids and sparged with nitro-
`gen or oxygen before sterilization and storage over the range 25-
`90°C. Control: egg phospholipids, nitrogen sparged (estimated slope
`= -5.93), —-——O—-; saturated phospholipids, nitrogen sparged
`(estimated slope
`-5.30), I; egg phospholipids, oxygen sparged
`(estimated slope
`— 6.11), ----O----.
`
`The rate of pH change and the rate of FFA formation
`are compared in Fig. 1 for a model emulsion prepared with
`the pharmaceutical grade of soy oil. Results obtained using
`the “superrefrned” grade of oil were similar. It is evident
`that the rate of formation of FFA is slower than the rate at
`which the pH drops, suggesting that some other factors are
`involved. As shown in Fig. 2, emulsions prepared with un-
`saturated or saturated acyl groups on the phospholipid moi-
`eties and sparged with nitrogen prior to sterilization and stor-
`age changed pH at almost identical rates. However, it should
`be noted that these two emulsions had markedly different
`hydrolysis rates, (3), which suggested that the lowering of
`
`pH was due to some other factor not necessarily associated
`with phospholipid hydrolysis. As noted earlier, free fatty
`acids can also arise by hydrolysis of triglycerides but the ’
`total (titratable) FFA in the system, irrespective of its
`source, is clearly unable to account for the relatively rapid
`lowering of pH.
`When sparged with oxygen, the pH change in the con-
`trol emulsion was considerably increased (Fig. 2), and this
`observation suggests that the effect of oxygen on the emul-
`sion pH was more pronounced than that produced by the
`heat-induced hydrolytic degradation reaction.
`The solubility of gases in liquids is described by Henry’s
`law (1 1):
`
`LogkFormation -4
`
`2.5
`
`275
`
`3
`
`3.25
`
`3.5
`
`1/Tx1000
`
`Fig. 1. The rate of FFA formation (by KOH titration) and the fall of
`pH (glass electrode) in a 20% soy oil emulsion stabilized with egg
`lecithin after sterilization and storage over the range 25-90°C. pH
`(slope = —3.42),——I———; FFA (slope = ——1.47). '--C---. K, rate
`of formation (slope of change against time); T, absolute temperature
`(kelvins).
`
`000002
`
`PA = XA K3
`
`where PA is the vapor pressure of a solution containing sol-
`ute A, XA is the mole fraction of A, and K, is Henry’s law
`constant. Thus, intuitively, since the K, for gases in non-
`aqueous solvents is generally higher than the corresponding
`value in water, it would appear that the oxygen is likely to
`preferentially dissolve in the oil phase of the emulsion. At-
`kins (11) discussed this issue in relation to benzene at stan-
`dard temperature and pressure and observed that, in all
`cases, the gas was more soluble in the benzene than in the
`water.
`-
`
`Based on this consideration, it seems feasible to suggest
`that residual oxygen may remain dissolved in the triglyceride
`oil phase after preparation and manipulation of the emulsion.
`Some of the triglycerides contain unsaturated acyl centers,
`which could, therefore, become partially oxidized, although
`not necessarily to the point where they would be titratable
`with alkali and estimated as FFA. These moieties would
`affect the pH of the unbuffered system, in addition to the
`effects produced by the FFA resulting from hydrolysis of
`phospholipid and triglyceride entities. The evident complex-
`ity of the physical and chemical structure of phospholipid-
`
`000002
`
`
`
`776
`
`Herman and Groves
`
`stabilized emulsions makes this suggestion difficult to con-
`firm at present.
`
`ACKNOWLEDGMENTS
`
`Our joint thanks are due to Mr. Peter Cade and his
`colleagues at Croda (US) Inc. for partial support on this
`project.
`
`REFERENCES
`
`1. P. K. Hansrani, S. S. Davis, and M. J. Groves. The preparation
`and properties of sterile intravenous emulsions. J. Parent. Sci.
`Technol. 37:l45—150 (1983).
`2. I. Hékansson. Physico—chemical changes in artificial fat emul-
`sions during storage. Acta Chem. Scand. 20:2267-2281 (1966).
`3. C. J. Herman. The Influence of Thermal Stress on the Proper-
`ties of Phospholipid Stabilized Emulsions, Ph.D. thesis, Uni-
`versity of Illinois at Chicago, Chicago, 1992.
`4. C.‘Washington and S. S. Davis. Ageing efiects in parenteral fat
`emulsions: The role of fatty acids. Int. J. Pharm. 39:33-37 (1987).
`
`J. M. A. Kemps and D. J. A. Crommelin. Chemical stability of
`phospholipids in pharmaceutical preparations: Hydrolysis of
`phospholipids in an aqueous medium. Pharm. Weekblad.
`123:355—363 (1988).
`. C. J. Herman and M. J . Groves. Hydrolysis kinetics of phos-
`pholipids in thermally stressed intravenous lipid emulsion for-
`mulations. J. Pharm. Pharmacol. 44:539—542 (1992).
`. M. Grit, J. H. de Smidt, A. Struijke, and D. J. A. Crommelin.
`Hydrolysis of phosphatidylcholine in aqueous liposome disper-
`sions. Int. J. Pharm. 50:l—6 (1989).
`. D. M. Lidgate, R. C. Fu, and J. S. Fleitman. Using a n1icroflu—
`idizer to manufacture emulsions. BioPharm. 45:28-33 (1989).
`. G. Lee, 1). Dick, E. V. Vasquez, and K. Werner. pH measure-
`ments of suspensions. In M. H. Rubenstein (ed.), Pharmaceu-
`tical Technology: Drug Stability, Ellis Horwood, Chichester,
`UK, PP. 113-117.
`. C. Paquot. Determination of the acid value and the acidity. In
`Standard Methods for the Analysis of Oils, Fat, and Deriva-
`tives, Pergamon, New York, 1979, pp. 52-55.
`P. W. Atkins. Physical Chemistry, W. H. Freeman, New York,
`1986.
`
`11
`
`000003
`
`000003