712 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS organomercurials and quaternaries may not be of practical importance but 50•o loss of a phenolic can be very serious. Thus, when considering the loss of preservative from a solution, through no matter what cause, it is not sufficient to state the weight lost or percentage lost, but it is imperative that this information be linked with the concentration exponent because it is the loss of activity and not the loss in weight that is of concern. Preservative activity is influenced by temperature, the higher the ambient temperature the more active is the preservative. It is usual to relate preservative temperature to the death-rate constant by the expression 0(50: - 50x ) - K•. (Ill) Ki where 0 = temperature coefficient of the preservative Kx = death-rate constant at temperature Tx K,. = death-rate constant at temperature T1 and since by equation (II) Kx tz K2 tl 0(50,- 50•) - tx (IV) ß Thus, if the temperature coefficient of a preservative is known the change in sterilization time may be calculated for a stated temperature change. Unfortunately, temperature coefficients vary according to the temperature range over which they are measured and the test organism (19, 20). For the sake of an example, a representative 0 xø for phenol over the temperature range 10-20 ø would be 5.0 (0•ø=5-1.175øC -•) which means that at 20 • a solution of phenol would kill E. coli five times as fast as at l0 ø. It is sometimes useful to be able to predict the effect on preservative activity of changing both the concentration and the temperature of a pre- servative. If from equation (II) is abstracted C1 n re-arranged as Clntl = Cs": and this is combined with equation (IV) re-arranged as tl = t2 0(50:-501)

PRESERVATIVES FOR PHARMACEUTICALS 713 we have Cx•tx = C•.,,t•. 0(v:- Vl ) which is an equation incorporating all the terms necessary to predict the effect on extinction time of changing both concentration and temperature. To illustrate the use of the equation, a multiple dose injection contained at the time of manufacture 0.1•o chlorocresol which provided an approxi- mate sterilization time for an infection of vegetative cells of 10 min at 25øC. After several months storage and as a result of absorption by the closure the chlorocresol concentration in the solution had fallen to 0.04•o. If the container was stored in a cupboard at 10 ø the activity of the solution would have fallen as a result of both the drop in concentration and drop in temperature. A new estimate of the sterilization time would be provided by Cxntl 0(•':- tz = C2 n (0.1) ø x 10 x -- -- (0.04)0 = 27 470 min. Such an interminable time may be without real meaning but it indicates a change from a solution possessing considerable preservative activity to one virtually devoid of preservative activity--a potentially dangerous solution. INFLUENCE OF FORMULATION The formulation of a product may substantially modify the chemical and biological availability of an included preservative and thereby modify the activity of the preservative. The major formulation factors affecting activity are pH, complexation with emulgents or dispersing agents and partitioning of the preservative between the components of an emulsified product. Whilst pH is by no means unimportant as a factor influencing pre- servative activity, there have been very few systematic and comprehensive studies of the problem and it is very difficult to quantify because changes in pH may modify the ionization of chemical groupings on the bacterial surface, produce ionization of the preservative and influence partitioning of the preservative between the product and the microbial cell and all three factors may interact. In very general terms there is evidence that a rise in the pH of solution of a quaternary produces increased uptake of the quaternary leading to enhanced antimicrobial activity (21). On the other