Heat Sterilization of Thermally Labile Solutions*J
`By T. HIGUCHI and L. W. BUSSE
`
`A general criterion for choosing high or low temperature for sterilizing heat-
`sensitive pharmaceutical preparations is presented. A mathematical analysis of
`the competing processes involved shows that the choice of temperature is largely
`dependent on the relative magnitudes of the heats of activation of the sterilization
`and the drug deterioration reactions. Results indicate, for example, that for pro-
`caine solutions, autoclaving at 120' is preferable to prolonged sterilization at 100'.
`
`R = the universal gas constant
`T = the absolute temperature of sterilization
`K = a constant depending on the number and
`kind of the most thermally resistant
`species present
`The above expression is a modified form of the
`classical Arrhenius equation.
`In Fig. 1 a typical relationship between In ta
`and temperature is shown. Also shown in the
`figure is the type of relationship found between the
`half life (the time required for 50y0 decomposition)
`of a thermally labile drug, such as procaine, and
`absolute temperature. This can be expressed
`similarly mathematically by the equation
`
`In t i / % = AH:/RT + C
`
`(2)
`
`where
`t i / , = the half life of the drug at temperature
`T
`AH,* = the heat of activation of the decomposi-
`tion reaction
`C = a constant
`
`RE LATlONSHl P OF STERILIZATION
`TIME AND HALF LIFE TO TEMPERA
`
`Lo 9
`time
`t
`
`HE PROBLEM of the relative merits of steriliz-
`ing heat-sensitive solutions a t one or another
`temperature is often encountered by those en-
`gaged in pharmaceutical
`investigations. The
`question, for example, whether a procaine solu-
`tion should be sterilized by autoclaving or
`sterilized a t 100" has often been debated. In
`the present study, a general criterion for deciding
`whether higher or lower sterilization temperature
`is to be preferred in preparation of a thermally
`sensitive solution is presented. Although the
`actual application is made to the case of procaine
`solutions, the method is generally applicable to
`nearly all thermally labile solutions.
`
`THEORETICAL CONSIDERATION
`
`The sterilization process involves, essentially, ir-
`reversible denaturation of certain vital enzyme-
`proteins in the microorganisms present. Since in-
`activation of enzymes and denaturation of proteins
`appear to be of
`the unimolecular type, we can
`reasonably expect the death rate of microorganisms
`to follow the first order reaction law (1). In case of
`ordinary sterilization, however, the picture is com-
`plicated by the fact that several varieties of micro-
`organisms are present. Furthermore, thermal sens -
`tivity of a given type of an organism may be ap-
`preciably different at different stages of its life cycle.
`Nevertheless it has been found experimentally that
`the first order law is followed reasonably well during
`the terminal part of the sterilization process. This
`can be rationalized in view of the fact that only
`the most thermophilic of the original flora is present
`during the terminal course of the process.
`On the basis of the above discussion and experi-
`inental evidence, we can write the following equation
`relating the death time, t d . necessary for total
`sterilization to the absolute temperature of steriliza-
`tion,
`
`In td = AH:/RT
`
`K
`
`(1)
`
`I / T
`Figure 1
`
`where
`td = the necessary sterilization time
`AH: = the heat of activation characteristic of
`It is evident from the curve that a t a low tem-
`killing most thermally resistant species
`perature ( A ) the half life of the drug may be much
`present
`shorter than the necessary sterilization time; yet
`at a higher temperature (B) sterilization of the drug
`* Received May 16, 1950, from the Research Laboratories
`may be achieved in a small fraction of its half life.
`of the School of Pharmacy, University of Wisconsin, Madi-
`The determining factor for such a situation is that
`son. t This project was supported in part by the Research
`the sterilization line has a much sharper slope than
`Committee of the Graduate School from funds supplied by
`the line representing the decomposition reaction.
`the Wisconsin Alumni Research Foundatcon.
`41 1
`
`MYLAN ET AL. - EXHIBIT 1019
`
`

`
`41 2
`ASSOCIATION
`JOURNAL OF THE AMERICAN PHARMACEUTICAL
`to be 14 kilocalories. Since for all practical
`Since the relative slopes of
`these lines depend
`the compound is
`directly on the magnitude of their respective AH*
`pharmaceutical preparations
`values, the above condition would be fulfilled if
`found in its salt form, only the former value need
`AH: > AH:.
`If we subtract Eq. 2 from Eq. 1
`be considered. These workers also found that the
`we obtain
`hydrolysis reaction of ionic procaine was iirst order
`In id - In t l l z = [AH? - AHf] F~ + K - C
`with respect to hydroxyl ions.
`It appears from arguments put forth in the pre-
`ceding section that since 12 kilocalories is much
`less than 50 kilocalories, sterilization of procaine
`1 - j. + Constant
`AH: -
`solutions is more advantageously carried out at
`t d
`In- =
`higher than at lower temperatures.
`In solutions
`R
`t . / 2
`buffered with acid type buffers, e.g., phosphate,
`AH: - A H a I +Constant
`log - =
`h
`borate, etc., the advantage of high temperature
`2.303 X 1.987 T
`ti/*
`sterilization is, however, considerably less. This is
`due to the fact that although these buffers maintain
`and if we let R = t d / t ‘ / z then at temperatures T A
`fairly constant pH values over wide temperature
`and TB we have
`ranges, hydroxyl ion concentration changes very
`rapidly with temperature. The increased hydrolysis
`rates resulting from higher hydroxyl concentration
`a t higher temperatures correspond to added heat of
`activation of 10-13 kilocalories (the heat of ioniza-
`
`) t>e exact value de-
`
`tion of water = 2.30R-
`d(1lT)
`pending on temperature.. In unbuffered or solutions
`buffered with an amine type buffer, the hydroxyl
`ion concentration is roughly independent of tem-
`perature and the correction may be omitted. In
`either case the apparent heat of activation of the
`hydrclysis reacticn is still well below the previously
`mentioned 50 kilocalories value for the sterilization
`reaction. The calculated Rjoo/Rim for unbuffered
`procaine solutions is roughly 12. Far solutions
`buffered with acid type buffers RioolRruo, calculated
`with Eq. 3, is approximately 8. For reasons
`stated above, sterilization by autoclaving for a short
`time is thus preferable to relatively long steriliza-
`tion a t 100” for procaine solutions.
`
`REFERENCES
`
`(1) Porter, J. R. “Bacterial Chemistry trod Physiology,”
`John Wiley and S o k , N e w York, 1946, pp. 172-192.
`(2) Higuchi, T.. Havinga. A. L., and Busse, L. W., THIS
`JOURNAL, 39,405(1950).
`
`or
`
`1
`
`The ratio RA/RB may be thought of as an improve-
`ment factor which expresses numerically the rela-
`tive rate advantage of sterilization over decom-
`position in going from T A to TLL If RA/RB is
`greater than unity higher temperature is to be pre-
`ferred. The value of RA/RB can be solved for any
`particular situation if numerical values of AH:
`and AH,* are known.
`AH: values obtained
`thermophilic
`for most
`microorganisms and spores appear to be over 50
`kilocalories. Those forms which have AH.* less
`than 50 kilocalories are relatively heat sensitive
`and do not influence the sterilization time. AH,*
`must be obtained for each drug.
`Application to Procaine Solutions
`Recently Higuchi, Havinga, and Busse (2) com-
`pleted a chemical kinetic study of the rate of hydrol-
`ysis of procaine in solution. The data obtained
`indicate that the rate of hydrolysis of the ionic
`form of the drug was entirely different from that of
`the free base form. The calculated heat of activa-
`tion for the hydrolysis of
`the salt form was 12
`kilocalories. That for the free base form was shown
`
`WHO MAKES IT?
`
`The National Registry of Rare Chemicals, Armour Research Foundation, 33rd, Ferleral and
`Dearborn Streets, Chicago, Ill., seeks information on sources of supply for the following chemicals:
`
`Trypiticase
`S-Benzylhomocysteine
`Chavicine
`Carene
`Neomenthol
`1.3-Diphosphoglyceric acid
`5,6-Dimethylbenzimidazole
`2- (d-Pantoylamino)ethylsulfono-4-chloroanilide
`m-Hydroxypropadrine
`bis-Dihydrocitronellyl ether
`
`Dodecyltriethylammonium bromide
`p-Fluoropropionic acid
`Soya lipase
`Carnegine
`Estriol glucuronate
`Barbatinic acid
`Hemipic acid
`Erythropleine
`Xanthotoxin
`Laurone