TEMPERATURE AND MICROBIAL DESTRUCTION CONDITIONS IN FLEXIBLE PACKAGE SEALS Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY Bruce D. Eder I966 1145515 LlBRAR Y ’ Mighigan Sun ytfivemq at“ 'k!‘ '...' If??? rant 'u'l’" I I ABSTRACT TEMPERATURE AND MICROBIAL DESTRUCTION CONDITIONS IN FLEXIBLE PACKAGE SEALS by Bruce D. Eder Flexible packages are now being used as containers for high moisture heat-preserved shelf-stable foods even though all the container and processing problems have not been solved. One of these problems is to determine if microorganisms trapped in the seal area can survive heat sealing and a normal heat process; viable organisms may be released from the seal area and spoil the food product if the seal peels. Another problem is in de- termining Optimum sealing conditions. A knowledge of the tem- perature profile in the seal at any given time will aid in de- termining these optimum conditions. The purposes of this investigation were to determine the rate of destruction of Spores in seals of flexible packages, and to determine the time-temperature conditions in the seal, during and after heat sealing of flexible packages. Fraction negative tests of Spores heated in simulated seals, on tin cups and on vinyl were conducted. Destruction rate of spores in simulated seals heated in steam.was deter- mined and compared to the dry heat destruction rate of Spores heated on tin cups and on vinyl and to the rate of wet heat destruction of Spores heated on tin cups. The rate of de- struction of microorganisms in seals subjected to dry heat was determined using survivor curves and compared to the rate of destruction of Spores subjected to dry heat on vinyl and cups, also determined using survivor curves. The results showed that dry heat conditions predominate in seals heated in wet steam. The results of dry heat destruc- tion studies of spores on cups, vinyl, and in simulated seals indicated that the resistance of the spores may be higher when heated in seals than on vinyl or cups. The 2 value of Spores Subjected to dry heat on vinyl was found to be higher than that for spores subjected to dry heat on cups. Temperature profiles established during heat sealing of flexible packages were calculated by deriving approximate equations for unsteady state heat transfer through incremental thicknesses and iterating these equations using small time increments. The results were plotted in the form of heating and cooling curves and temperature profiles. Heating and cooling rates at the seal interface were measured during heat sealing using a small thermocouple. These results were plotted as heating and cooling curves. The cooling curve calculated for h = 3.0 was found to correSpond closely with the measured ccOling curve; however, the calculated heating curves did not correspond as closely with the measured heating curve. The significantly lower j and somewhat higher f of the measured curve probably resulted from use of the relatively large thermocouple junction. The cal- culated heating curve is believed to be the most accurate. TEMPERATURE AND MICROBIAL DESTRUCTION CONDITIONS IN FLEXIBLE PACKAGE SEALS By Bruce D. Eder A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science 1966 ACKNWLEDGEMENTS The author wishes to express his appreciation to his major professor, Dr. I. J. Pflug, for his patience and guidance through- out this study and to Dr. D. K. Anderson and Dr. R. V. Lechowich who served on the committee for their suggestions in editing this manuscript. Thanks go to Dr. B. S. Schweigert, Chairman, Food Science Department, for his interest and support to this program; to Helen Erlandson for her assistance in the laboratory; and to Kenneth Fox and Dr. I. J. Kopelman for their help in editing this manuscript. The author is grateful to his wife, Janet, for her en- couragement and assistance throughout this study. ii In memory of my mother Marguerite Katherine Miller Eder TABLE OF CONTENTS I. INTRODUCTION II. LITERATURE REVIEW OF HEAT DESTRUCTION OF BACTERIAL SPORES A. The order of death B. Thermal destruction curves and temperature coefficients C. Methods of calculating D values D. Nature and mechanism of death 1. General review 2. Wet heat vs. dry heat E. Some factors affecting dry heat resistance 1. Initial number of organisms 2. The water content of the organisms 3. The moisture content of the heating medium 4. The chemical composition of the heating medium 5. The support medium 6. The original suspending medium F. Previous results of heat destruction of Spores in seals of flexible packages III. EXPERIMENTAL PROCEDURE A. Heat resistance of Bacillus subtilis 5230 Spores 1. Preparation of the spore suspension a. Growing of Spores b. Harvesting and cleaning of spores 2. Examination of spores and preparation of stock iii Page 11 ll 16 17 18 19 20 21 22 23 23 26 26 26 26 27 28 3. Preparation of spores for heat treatment 4. Enumeration of initial number of Spores 5. Heat treatments a. Fraction negative (FN) tests (1) Wet heat (2) Dry heat (3) Heating in simulated seals b. Survivor curves (1) Method (2) Statistical analysis of data B. Measurement of heating and cooling rates in flexible package seals 1. Preparation of the thermocouple arrangement 2. Heating and cooling of the seal 3. Plotting of curves and description of their parameters IV. DERIVATION OF EQUATIONS USED TO CALCULATE TEMPERATURE PROFILES IN SEALS A. Description of the model and assumptions made B. Heating C. Cooling D. Method of calculation V. RESULTS AND DISCUSSION A. Heat destruction of spores B. Heating and cooling of seals VI. CONCLUSIONS REFERENCES iv 29 3O 31 31 31 31 32 33 33 34 39 39 41 43 45 46 49 54 55 6O 6O 89 100 102 III-1 III-2 III-3 III-4 IV-1 IV-2 IV-3 IV-4 V-2 v-3 V-4 FIGURES Simulated seal in the clamping device Placement of cups, cups containing vinyl, and simulated seals in a TDT can Diagram of the thermocouple arrangement used to measure the temperature in the center of the seal Photomicrograph of the thermocouple junction used to measure temperatures in the seal taken on a vinyl background Cross section of the sealing system taken along the heat flow direction Schematic diagram used in derivation of equations for heating of the seal Schematic diagram used in derivation of equations for cooling of the seal Block diagram for computing temperature profiles established during heating and cooling of the seal Thermal resistance curve of B. subtilis 5230 Spores heated on tin cups in wet heat Thermal resistance curve of B. subtilis 5230 Spores heated on tin cups in dry heat Survivor curves of B. subtilis 5230 Spores dry heated at 280° F. on tin cups on vinyl and in simulated seals determined by linear regression Survivor curves of B. subtilis 5230 Spores dry heated at 305° F on tin cups, on vinyl, and in simulated seals determined by linear regression Calculated and measured heating curves for heating at the seal interface, plotted on semilogarithmic paper Page 35 35 40 42 46 51 51 58 63 64 87 88 93 V-6 V-7 V-8 Calculated and measured cooling curves for cooling at the seal interface plotted on semilogarithmic paper 94 Calculated temperature profiles established during heat sealing 95 Calculated temperature profiles established during cooling of the seal 96 vi IV-1 IV-2 IV-3 V-2 V-3 V-5 V-6 V-7 V-8 TABLES Thermal Properties List of Equations Used for Calculating Temperature Profiles Illustration of the Results of the Calculations of Temperature Profiles for the First Ten Time Increments Results of Wet Heat Fraction Negative (FN) Test of B. subtilis 5230 Spores on Tin Cups Results of Dry Heat Fraction Negative (FN) Tests of‘B. subtilis Spores on Tin Cups Results of Dry Heat Fraction Negatives Tests of B. subtilis 5230 Spores on Vinyl and in Simulated Seals Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heating of B. Subtilis 5230 Spores on Tin Cups at 280° F Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heatingo of B. Subtilis 5230 Spores on Vinyl at 2800 F Counts, Adjusted Counts and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heating of B. Subtilis 5230 Spores in Simulated Seals at 280° F Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of Adjusted Counts for Dry Heating of B, Subtilis 5230 Spores on Tin Cups at 305° F Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of Adjusted Counts for Dry Heating of B. Subtilis 5230 Spores on Vinyl at 305°F vii Page 48 56 59 62 66 73 75 77 79 81 V-10 V-11 V-12 V-13 V-14 Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of Adjusted Counts for Dry Heating of B. Subtilis 5230 Spores on Simulated Seals at 305° F 83 Results of Decimal Reduction Times (D values) and 0 Time Intercepts (NA) and Their 95 Percent Confidence Limits Obtained from Linear Regres- sion on the Second Portion of the Broken Survivor Curves 85 Results of Statistical Testing of Linearity and Differences Between D Values and 0 Time Inter- cepts (NA) of the Second Portion of the Broken Survivor Curves 86 Results of the Measurement of Heating and Cooling Rates at the Seal Interface 90 Results of Calculation of Temperature Profiles 91 f and j Values Calculated from Heating and Cooling Curves 98 viii NOMENCLATURE Bacteriology FNSO P water activity, ($3) T decimal reduction time; time required to re- duce a population of microorganisms by 90 percent. 95% CLD, 95 percent confidence limits of D. min thermal death time; time required to reduce a given level of microorganisms to a desired low number min time when 50 percent of the replicate units are sterile min death rate constant min cumulative number of samples surviving at an exposure time number of tubes showing partial survival between P = .16 and P = .84 cumulative number of samples not surviving at an exposure time number of viable microorganisms; N , initial number; 0 NU, number viable after being exposed to a lethal agent for U minutes; NA’ apparent initial number, as determined by the intercept of a survivor curve or linear portion thereof at U = 0. number of replicates showing survival at an exposure time ix pv vapor pressure of a solution at temperature, T pw vapor pressure of pure water at temperature, T P probability of sterility, 25%;? q number of sterile replicates at an exposure time r total number of replicates at an exposure time ZS difference between the dosage at P = .16 and P = .84 min T temperature 0F U equivalent heating time at heating medium temperature min UFNSO equivalent heating time when 50 percent of the units are sterile min i most probable number of organisms 2 change in temperature which results in a ten fold change in D or F 0F n number of D values required to reduce a given level of microorganisms to a desired sterility. Statistics a regression estimate of A A Y axis intercept b regression estimate B B slope of a line F variance ratio; ratio of two estimates of variance or mean squares. n number of groups of samples corresponding to different X values n. number of samples in the ith group 1 saand Sb estimates of the standard errors of a and b t ratio of the difference between two means to the standard deviation of the difference X independent variable Y dependent variable; Yij’ the jth sample in the ith group of samples; Ti , mean of the ith group of samples I regression value of the dependent variable; $13, regression value of the jth sample in the ith group of samples. a level of significance; probability of rejecting a hypothesis when it is true Heat Transfer A area normal to heat flow Cp heat capacity f temperature response parameter; time required for the asymptote of the heating (or cooling) curve to traverse one log cycle h heat transfer coefficient j lag factor, (TA - T1)/(To - T1) k thermal conductivity 2 p C Ax M a modulus of the form k At . hr NBi biot number, k xi ft2 Btu/lb OF m. sec Btu/ftzhr °F dimensionless Btu ft/ftzhr OF dimensionless dimensionless At Ax radius of a sphere or cylinder or half thick- ness of a slab Ax resistance to heat £10 ,1:- time time interval temperature; To, initial temperature; TA, apparent initial temperature determined by intersection of the asymtote to the heating or coaling curve with the ordinate; T1, the heating or cooling medium.temperature; Tjaw’ temperature of the heated jaw thickness thickness of an element thermal diffusivity (k/p Cp) density xii ft ft2 hr °F [Btu SEC 86C °F inches inches ftz/hr 3 1bm/ft I. INTRODUCTION Flexible packages are in general use as containers for dry, liquid nonprocessed, and frozen foods but are just starting to be commercially used as containers for high moisture heat- processed shelf-stable foods. The nature of the container has created many problems in adapting its use for heat processed foods; however, the problems are offset by the inherent advan- tages, such as: potentially low container cost, less inventory storage Space, little or no container diSposal problem, and the ability of the container to take irregular shapes. Flexible packages for heat processed foods are formed by heat sealing two laminates on three edges leaving one end Open for filling. After filling the pouch with a food product, the Open end is heat sealed. The three seals of the empty pouch are usually produced under conditions which may be described as clean, but no attempt is made to sanitize or sterilize the inner surfaces of the web that are sealed together. Viable microorganisms in the atmOSphere and on the pouch fabricating equipment may contaminate the web and become incorporated into the seal area. Microorganisms present in the seal area under dry conditions may remain viable after a normal heat process and may be released into the food product if the seal peels even a small amount during storage and handling. Because the seal is a critical part of the flexible package, the sealing conditions need to be carefully controlled. Optimum sealing conditions for a material are determined by trial and error. Knowledge of the relationship between the seal quality and the temperature profiles established in the seal during and subsequent to the sealing Operation will aid in establishing the Optimum sealing conditions. This study was initiated to determine the resistance of Spores in seals and to find methods of determining the time- temperature conditions in the seal during and subsequent to heat sealing. ‘II. LITERATURE REVIEW OF HEAT DESTRUCTION OF BACTERIAL SPORES Vegetative bacteria of the genera, Clostidium and Bacillus, are known to Sporulate under certain conditions; Spcrulation may be induced by restricting nutrients such as inorganic nitrogen, and organic compounds. A number of bacteria require Mn++ for sporulation. The mature Spore is fully refractile, differen- tiated, not stainable, thermoresistant and contains 5-15 percent dipicolinic acids (DPA); whereas, the vegetative cell is homog- eneous, not refractile, fully stainable, heat sensitive and contains no DPA (Halvorson, 1962). Spores are liberated from the parent vegetative cell by lysis of the vegetative cell wall. A. The order of death The criterion almost universally used by the microbiologist to define death of a microorganism is the loss of the ability to reproduce. Rahn (1945b) and others have proposed that unicellular microorganisms die according to a logarithmic order. The theory of the logarithmic order of death predicts that a plot of the logarithims of the number of survivors versus the time exposed to a lethal agent (heat, chemical disinfectants, and irradiation) will yield a straight line. In a monomolecular chemical reaction the rate of reaction is directly proportional to the concentration. According to the theory of the logarithmic order of death of microorganisms, the death rate is proportional to the number of viable organisms dN 'Cl—fi = -KN (II -1) where K = the death rate constant N = the number of viable organisms U = equivalent heating time at heating medium temperature The minus sign indicates a decrease in the rate of inactiva- tion. Equation II - 1 can be rearranged and integrated over initial and final conditions as follows: Initial conditions Final conditions U = O U = U N = No N = NU N U I U 9-11 =j‘ -KdU (II -2) ll N o o NU 1n N“ = -KU (II -3) 0 Solving equation II -3 for U yields No U = l/K 1n-- (II -4) NU The decimal reduction time [Sandhozer and Katzin (1942) as cited by Rahn (1945b)] is the time required to reduce the initial p0pulation by 90% or the time to traverse one log cycle of a survivor curve. Schmidt (1950) suggested using the term D to designate decimal reduction time. Then when N = .lNO U = D Substitution of these conditions into equation II - 4 yields N 1 o D “ K 1“ .lN o 2.303 D - K (II-5) . ._ 2.303 . . . When D is substituted for K in equation 11-4 we find after rearrangement that D = U (II-6) 10g (NO/NU) D is convenient to use in comparing the rates of destruction of microorganisms. The eXpression for D can be generalized so that D can be obtained if the number of survivors are known at any two times. U - U 2 1 D = (II-7) log NU - log NU 2 1 Much evidence appears to support the logarithmic order of death theory; however, many investigators have obtained survivor curves which are not linear. Frank and Campbell (1957), Walker, Matches and Ayres (1961) Obtained wet heat survivor curves which were concave upward, while Amaha and Ordal (1957) obtained wet heat curves which were concave downward. El-Bisi and Ordal (1956b) obtained wet heat curves which were concave downward at low temperatures and straight at high temperatures. They explained the low temperature curves as follows: During relatively short heat- ing times some of the viable spores were not heat activated and as a result did not germinate when plated. At longer heating times the number not activated diminished. Because the numbers of spores surviving at short times were underesti- mated, the survivor curves did not drop logarithmically until all the viable spores were heat activated. Rahn (1945b) ex- plained that concave downward curves could have resulted from cell repair at low temperatures. Licciardello and Nickerson (1963) found linear wet heat survivor curves for PA 3679, concave downward curves for B. subtilis var niger, and both concave upward and downward curves for Salmonella senftenberg 775W. Stumbo (1945), Rahn (1945b), and EleBisi and Ordal (1956b) stated the possible reasons for deviations from a straight line survivor curve. Concave downward curves were explained by clumping of cells. A clump of cells will appear as only one colony when plated even though all these cells may be living. Under the assumption that each viable organism produces one colony, the number of living macroorganisms will be under- estimated. Until the clumps are reduced to one living cell, the survivor curve will'not dr0p logarithmically, and we will observe a curve similar to that of multicellular organisms. Concave upward survivor curves were explained as having resulted from non-homogeneous cultures containing different resistant organisms. This type of curve will be the net result of a combination of the survivor curves for each group of different resistant organisms. El-Bisi and Ordal (1956b) explained that concave downward curves resulted from organisms becoming more exacting in their nutritional requirements at longer heating times. A survivor curve which begins steep, then becomes fairly flat and then becomes steep again is explained by Stumbo (1965) as being due to flocculation of cells during heating. The curve falls logarithmically, then the organisms clump, resulting in a decrease in the slcpe of the curve. When only one viable organism remains in each clump, the curve again becomes steep. Stumbo (1965) also eXplains that deflocculation during heating would result in a survivor curve which rose then fell loga- rithmically. B. Thermal destruction cruves and temperature coefficients Thermal death time curves were first plotted by Bigelow (1921) who obtained a straight line when the common logarithm of the thermal death time was plotted against temperature. A convenient measure of the slope of these curves is 2 which is defined as the change in temperature which will result in a ten fold change in the thermal death time (F) or the decimal reduction time, D. A plot of D values, the decimal reduction times, on semilog paper against temperature is called a thermal resistance curve and will have the same slcpe as the thermal death time curve. Since F is the time required to reduce a given level of organisms to a desired low number, and D is the time required to reduce a given pOpulation by 90 percent, then D and F can be related by a certain number, n F = ND F and D are related to z by the following equations: F1 T2 ' T1 F; = 10 Z D1 T2 ' T1 ? = 10 z 2 These relationships are only valid if the thermal death time and thermal resistance curves are straight. Esselen and Pflug (1956) tested the wet heat resistance of PA 3679 Spores in the temperature range, 2500 F - 2900 F, and found a higher 2 value in the range, 2700 F - 2900 F, than in the range, 250° F - 270° F. Secrist and Stumbo (1958) found the wet thermal resistance curve of PA 3679 Spores linear in the range, 2500 F to 2900 F, after applying a lag correction factor at 2900 F. Wet heat thermal resistance curves, determined by Licciardello and Nickerson (1963) for B. subtilis var niger in the temperature range, 1490 F - 2210 F (650 C - 1050 C), and Salmonella senftenberg 775W in the temperature range, 114.8o F - 150.8o F (460 C - 660 C), were linear; the curve for PA 3679 was concave downward in the temperature range, 176° F - 248° F (80° c - 120° C), but linear in the range, 221° F - 248° F (105° C - 120° C). C. Methods of calculating D values D values can be obtained directly from survivor curves or they can be calculated from fraction negative (FN) data obtained from.multiple samples run at successive time intervals. Schmidt (1957) described the Stumbo (1948a), Stumbo, Murphy and Cochrane (1950), and the Schmidt (1954) methods. When D values calculated by these three methods were compared, Schmidt (1957) found the D values agreed with 10 percent. Lewis (1956) criticized the Stumbo, Murphy, Cochrane (1950) method for its systematic bias in using an unweighted average. Lewis prefers the Reed-Muench or the Spearman-Karber methods discussed by Finney (1952) over the Schmidt (1954) method. Augustin (1964) compared D values calculated by the Modified Schmidt (1954), Reed-Muench, Stumbo, Murphy, and Cochrane (1950), and Spearman-Karber and found only small differences. The 2 values were identical. The Schmidt (1954) method involves two assumptions: First, if a sample shows survivors at a given exposure time, it will show survivors at a shorter exposure time. Second, if a sample shows no survivors at a given exposure time, it will show no survivors at longer exposure time. The probability of sterility at any exposure time is: (II -8) 10 where m.= cumulative number of samples surviving at the exposure time and n = cumulative number of samples nor surviving at the exposure time Using probability paper the probability of sterility is plotted vs. corrected exposure time and the best straight line is fitted to the points. The time where P = .5 as read from the curve was called LD50 in analogy to treatment of bioassay data by Schmidt (1954) and revised by Pflug (1966) to be called FN50 (fraction-negative 50 percent). UFNSO is the corrected time where 50 percent of the samples will be sterile. The equation for estimating most probable number of organisms derived by Halvorsen and Ziegler (1933) is: £_= total number of replicates at an eXposure time where . . . number of replicates shOW1ng no growth at an exposure time When 50 percent of the tubes are sterile, r/q = 2 and i = 1n-E = .69 organisms/unit. D can be calculated in the following manner: U log No - log NU 11 UFNSO UFNSO then = log NO - log .69 = log NO + .16 (II -9) The 95% confidence limits (95% CL) were derived from the standard error of UFNSO and are as follaws: 1.96 X ZS 954 CL = 2M (II ~10) where 28 = the difference between the dosage for P = .16 and .84. Schmidt (1954) chose M equal to the number of tubes show- ing partial destruction. Pflug (1962) modified this definition by choosing M.equal to the number of units showing partial survival between P = .16 and P = .84. D. Nature and mechanism Of death 1. General review A number of workers have pr0posed that bacterial death is caused by inactivation of enzymes. Virtannen (1934) found that enzyme activity was much higher in vegetative cells than in spores and that proteinase was much more resistant to heat in the presence of protein than alone. 0n the basis of these results he proposed that, in Spores, enzymes combine with cell protein in a way that the enzymes become far more resistant to heat. Rahn and Schroeder (1941) showed from experiments with Bacillus cereus that there was no correlation between death of the cell and loss of enzyme activity during heating. Consider- ing the death rate to be logarithmic they postulated that a definite molecule such as a very rare gene molecule or an 12 equally important molecule in cell division is inactivated to cause death. Rahn (1945b) disclaimed the theory of death from enzyme inactivation by illustrating that all bacteria would die at the same time if the enzymes had to be reduced to a certain level in each organism for death to occur. Considering the proba- bilities of death of an organism, he demonstrated that if more than a few molecules had to be inactivated to cause death, the death rate of bacteria would follow that of multicellular organisms. Because there are many molecules of the same enzymes present in each cell, he concluded that it is unlikely that death is caused by inactivation of one or a few of these molecules. Since the criterion for death is the inability to reproduce, he suggested that denaturation of one essential gene would result in a cell no longer being able to divide. Ingraham (1962) proposed that heat destruction of microorganisms was a result of ixuttivation of an essential gene or the destruction of a certain structure, such as a cell membrane. Several theories have been presented to explain the mech- anism of heat destruction. Ball and Olson (1957) reasoned that since the concept of temperature is no longer valid in a volume as small as a bacterial cell, something outside rather than in- side the cell must cause death. They further prOposed that the agents of death are molecules of high energies, momenta, or velocities which randomly strike the cells. By applying this 13 theory they were able to explain the effects of pH and salt concentration on the thermal destruction of bacteria. Tischer and Horwicz (1954) considered heat as particulate quanta of energy which are discontinuous and explained that the probability of affecting a site will then depend on both the concentration of organisms and quanta of energy. At one time the greater wet heat resistance of spores as compared to their correSponding vegetative cells was attributed to impermeability of the Spore coat. This hypoth- esis has since been refuted by numerous researchers [Lewis, Burr, and Snell (1960), Murrell (1961), Black and Gerhardt (1961), Gerhardt and Black (1961), Black and Gerhardt (1962), and others]. Murrell (1961) found that when Spores were suSpended in heavy water, D20, nearly all the water in the Spores mixed with the heavy water. Gerhardt and Black (1961) found that Spores were permeable to all types of molecules and that the extent of penetration depended on the charge, lipid solubility and molecular weight of a mmlecule. Penetration increased with increasing basic charge, higher lipid solubility and lower molecular weight. They speculated that the cortex was lipid- like with the core finely heteroporous and permeable to small molecules. They visualized the Spore coat as being layered, coarsely heterOporous, and fully permeable. Black and Gerhardt (1962) on the basis of determinations 14 of water content and uptake of tritium labeled water concluded that water in the Spore is freely exchanged and in equilibrium with external water. On the basis of these and previous results they concluded that water is unevenly distributed throughout the spore with the dense core containing relatively little water and hypothesized that the dense core is made up of a heat-stable gel with stable reversibly bonded macromolecules forming a high polymer matrix containing free entrapped water. They felt that this structure accounted for heat resistance in Spores. The levels of Ca and DEA have been shown to be much higher in Spores than in vegetative cells [Amaha and Ordal (1957), Church and Halvorson (1959), Powell (1953), Lechowich and Ordal (1962), Halvorson (1962)]. Amaha and Ordal (1957) found that calcium added to the Sporulation media consistently increased the heat resistance of the spores. Walker, Matches, and Ayres (1961) concluded from their results that accumulation of DEA in Spores beyond a certain point does not influence heat re- sistance although a certain amount of DEA is needed for the Spores to be heat resistant. Church and Halvorson (1959) found that the rate of in- activation decreased with increasing DEA concentration in the Spore. Their results showed that the rate of inactivation was higher at low concentrations than at high concentrations of DEA, Lechowich and Ordal (1962) suggested that the Ca-DEA chelate could unite with protein to retard denaturation of the protein. 15 They also suggested that this Ca-DEA-protein complex could be a component of the contractile cortex described by Lewis, Snell, and Burr (1960). Halvorson (1962) cites'Young (1959) who found that a DEA-calciumnamino acid chelate complex could be formed with a variety of amino acids. Halvorson (1962), who extensively reviewed the literature relating to thermostability and calcium and DEA content of spores, stated that one can conclude that both Ca and DEA are necessary for formation of heat-resistant Spores. Church and Halvorson (1959) cited Halvorson (1957) who found using Spores of Clostidium roseum.that during Sporulation refractility was followed several hours later by DEA synthesis. This was, in turn, followed an hour or more later by an increase in heat resistance. This indicated that the mere presence of DEA was not sufficient for heat resistance. Lewis, Burr, and Snell (1960) suggested that the heat stability of Spores is due to the presence of a dehydrated core. They visualized slow contraction of the cortex during maturation of the Spore which served to dehydrate the core and cortex it- self leaving the system in water-vapor equilibrium" They also stated that the assumption of an impermeable Spore coat was in- consistent with the known permeabilities of organic materials. Friedman and Henry (1938) used the cyroscoPic method des- cribed by Newton and Gortner (1922) to determine the bound water content of Spores and vegetative cells. They found the water 16 binding capacity of Spores to be far greater than that of vegetative cells and advanced the theory that heat resistance in spores was due, at least partly, to a high percentage of the water being bound and, consequently, inactive in coagulat- ing protein. Their results should be viewed sceptically because they found it necessary to employ an arbitrary correction factor to a constant in their equations in order to aVoid obtaining negative values for bound water. This they justified by using the results for comparitive purposes only. 2. Wet heat versus dry heat Lewith (1890) as cited by Augustin (1964), demonstrated that proteins are less sensitive to heat when held in a rela- tiVely dry state by heating egg albumin with various amounts of moisture. This led him to Speculate that the death of bacteria in dry and wet heat is due to protein coagulation and accounted for the extremely high resistance to dry heat when compared to wet heat. ‘Williams (1929) concluded from reviewing the leterature and examining his results that condi- tions which render a protein more difficult to coagulate result in increased heat resistance. Rahn (1945a) proposed that the mechanism of dry heat destruction is oxidation and supports this by stating that dried proteins will not coagulate even at 100°C. He attributes increased death rate to increased oxidation. In studying the effects of six different recovery media, 17 Augustin (1964) found that the order of recovery of Spores subjected to wet and dry heat was the same on four of the media but different on the other two. These differences were explained as due to certain parts of genetic material being more susceptible to denaturation in the dry state with others being more susceptible to denaturation in the wet state. E. Some Factors affecting dry heat resistance Murrell (1964) discussed the factors affecting heat re- sistance of Spores by dividing them into three categories: a. before heating, b. during heating, and c. after heating. Schmidt (1957) divides the factors affecting heat resistance into three general types: a. inherent resistance, b. envi- ronmental influences active during the growth and formation of cells or Spores, and c. environmental influences active during the time of heating of the cells or Spores. I shall discuss those factors relevant to this thesis namely those active during heating. Moist heating is readily defined as heating in an envi- ronment of water activity equal to 1.00 or of relative humidity equal to 100% (e.g. saturated steam or aqueous solution), while dry heating encompasses heating in environments having aw's of less than 1.00. Therefore, because dry heating in- cludes heating under a variety of different environmental conditions, these conditions (i.e. the water content of the organisms, the moisture content and chemical composition of 18 the heating medium, the prOperties of the support medium, and the composition of the original suSPending medium) must be Specified. The influence of some factors affecting the rate of dry heat destruction of microorganisms will be re- viewed even though their effects are not fully known. 1. The initial number of organisms If the logarithmic theory of death holds, a change in the initial number of a microbial pOpulation will not effec its D value or death rate constant. However, Esty and Meyer (1922) and Williams (1929) reported that increasing the initial number of organisms increased the survival time by an amount greater than predicted by applying the logarithmic theory of death. Rahn (1945b) attributed the results observed by Esty and Meyer (1922) at high initial numbers, ranging from 109 - 1011, to protection of living cells by dead cells. He sup- ported this contention by citing Lange (1922), who found that young dead cells of Bacterium.ggli and Staphlococcus aureus doubled or tripled the death time of the same Species, and Watkins and Winslow (1932) who observed that increasing the initial number in the range of 1-100 million decreased the death rate. After reviewing the literature, Murrell (1964) concluded that changes in Spore concentration displace the survivor curve, but do not change the shape. Pflug and Esselin (1954) found D values at 280° F for wet heating of EA 3679 Spores in phosPhate buffer for the initial numbers of 120, 10,000, and 1,000,000 were in good agreement. Sisler (1961) determined survivor curves for three different initial numbers of spores (approximately 10°, 107, 10°) heated in wet stream.at 250° F and in superheated steam at 320° F. When Spores were heated in superheated steam, he found the D value for 10° Spores was greater than the D values for 107 and 10° Spores; he found no significant difference between the D values for the initial numbers of 107 and 10° Spores. No significant difference in D values was observed among wet heat survivor curves having different pOpulations. He explained the difference in D values observed for superheated steam as being due to eXperimental error at the high temperature employed. 2. The4water content of thefiorganisms Murrell and Scott (1957) heated freeze-dried Spores at 110° C in. m; the Spores had been equilibrated in m at water activities ranging from 0 to .998. They found the highest D values for organisms equilibrated at water activities (aw) ranging from .7 to .9. Organisms equilibrated at aW = 0 showed significantly lower D values than those at higher water activ- ities up to .9. Hutton, Hilmoe, and Roberts (1951) in studying factors influencing survival of Brucella abortus during freezing found that as dryness of the organisms increased, survival after storage increased. Scott (1958) investigated the effect of residual water on the survival of dried bacteria during storage. Although the results were variable, a higher rate of death was found when the organisms were equilibrated with water activities near zero than with water activities in the range .05 - .2 eSpecially when stored in air as compared to vacuum. Sugars appeared to protect organisms equilibrated at very low aw's; Scott (1958) cited Fry and Greaves (1951) who suggested that the beneficial effects of glucose may have resulted from its ability to retain small amounts of water necessary for survival. Scott (1958) concluded that the loss of the most firmly held water molecules results in a loss of stability eXpecially in the presence of air, and that high concentrations of sugar substantially prevent the damaging effects of air. 3. The moisture content of the heating medium Jacobs (1963) heated Spores in atmOSpheres of varying water contents by placing various amounts of alum, which will release its water of hydration upon heating, in thermal death time cans containing tin cups inoculated with Spores. His results showed that the D values at 50 mole percent were lower than at 0 mole percent moisture. Above 50 mole percent the D values decreased as moisture content increased. At 25 percent moisture the D at 235° F was higher than at 0 percent, however, the D at 250° F was lower than at 0 percent. Jacobs, Nicholas, and Pflug (1965) in presenting this same work concluded that the presence of water 21 vapor at concentrations intermediate to that of saturated steam and dry air exerts an effect on the heat resistance, but were unable to explain the effect from the data. Silverman (1966) in heating Bacillus subtilis var. niger spores in dry air flow- ing at a rate of 3 liters per min., in non-flowing dry air, and in air flowing at a rate of 3 liters per min. containing water vapor flowing at a rate of 5 grams per hour, found the D values at 110° C to be .7, 2.0, and 3.5 min, reSpectively. These results suggest dry heat killing is related to the rate of water loss; the conditions which increased the driving force for mass transfer increased the death rate and correspondingly decreased the D value. In dry heat studies of organisms heated in closed TDT cans, Pheil'g£.gl. (1966) observed that as the humidity of the air in the can increased the resistance of the organisms increased. D. The chemical composition of the heatinggmedium Pheil.gg.gl (1966) tested the resistance of B.subtilis 5230 in helium, nitrogen, carbon dioxide, oxygen, and air, and found only small differences in resistance. Koesterer (1962) found that where bacteria had the greatest resistance in atmos- pheric air as compared to helium.and vacuum. The D value in air was twice the D value in helium. Bruch‘gt.al. (1963) observed that the D values of B. subtilis var. niger were much higher when the organisms were heated while entrapped in various solids than when heated on filter strips. These differences in D values could be due to differences in the 22 chemical composition of the support media and atmOSphere or differences in the ability of the solids and filter paper to restrict moisture loss. Pflug (1960) found the D 02F value of B. subtilis 5230 350 in super-heated steam at one atmosphere pressure to be .37 min with a z of 42° F. In contrast, when heated in wet steam, the D2500 F was .35 min and 2 was 14.80 F. Sisler (1961) observed D values of B. subtilis 5230 Spores in superheated steam at 350° F and moist heat at 250° F to be .38 and .32 min., respectively, and the 2 values to be 37.1 and 12.8° F, respectively. These compare favorably with the values obtained by Pflug (1960). 5. The support medium Augustin (1964) found that the dry heat resistance of putrefactive anaerobe No. 3679 was higher on tin than on alu- minwm. Pflug and Fox (1966) determined survivor curves for B. subtilis 5230 Spore heated in TDT can containing atmOSpheric air; The spores were supported on tin, glass, aluminum, and filter paper. The order of resistance on the materials from highest to lowest was tin, aluminum, glass, and filter paper. The organisms had a much lower resistance on filter paper than on the other materials. This might be attributed to the organisms being more exposed to the air since the organisms were distributed throughout the porous filter paper while they accumulated in one Spot on the other materials. 23 Bruch SE.El° (1963) found in testing the resistance of organisms on paper strips, sand, and glass the greatest re- sistance on sand, a lower resistance on glass and the lowest resistance on paper. 6. The original suspending medium A number of researchers have found that adding various sugars to the suspending medium have exerted a protective effect on organisms when dried and stored. The protective effect of sugars on storage of dried bacteria as found by Scott (1958) and Fry and Greaves (1951) is discussed under section 4.b. They suggested that the beneficial effect of sugars on storage of bacteria is due to its ability to retain small amounts of moisture which are necessary for survival of the organisms. Greaves (1959, 1960, 1962), as cited by Pflug (1966), supported this idea. F. Previous results of heat destruction of Spores in seals of flexiblefipackages. Pflug and Schmidt (1962) conducted experiments to determine whether or not the seal area of flexible packages could be sterilized during heat sealing. They were concerned with the possibility of microorganisms surviving in the seal after pro- cessing and being released to spoil the food product upon peel- ing the seal. A method was used to determine if microorganisms survive in seal areas of flexible packages under sealing conditions 24 which may exist in the field. Bacillus subtilis 5230 Spores were dried on pieces of packaging material and incorporated into heat seals. These seals were cut into strips and the strips threaded through tubes filled with growth media. After sealing, the strips and media were sterilized, and then the strips were pulled to Open the seals. A positive or negative result was recorded by observing the tubes for typical turbidity and pelicle formation after inoculation. The results showed that microorganisms can survive in the seal areas, but suggested that a sterile seal could be produced by using a jaw temperature of 525° F and dwell time of 2.5 sec on a single heated jaw sealer. Preliminary eXperiments were carried out by Pflug and Augustin (1963) to determine a feasible method for testing heat resistance 0f.§- subtilis in seals of flexible packages made up of mylar-foil-vinyl laminates. They used a method similar to that used by Pflug and Schmidt (1962), as described above. They found that when the laminates were tightly sealed together it was impossible to tear apart the seal; however, when the seals were sealed loosely enough to be pulled apart, it was questionable whether or not the seal was tight enough to prevent penetration of water. Therefore they employed a clamping device to hold the seal together during heating. Using this device they could simulate a tight seal which was easily pulled by placing the mylar side of one laminate to N U: the vinyl side of another laminate. Throughout their studies they modified the clamping device and found that a ring clamp with two metal blocks (1/2 in x 1/2 in x 1/8 in thick) produced the tightest seal. Each seal was clamped between the metal blocks. They used two methods to inoculate Spores into the seal. In both methods they inoculated Spores on the vinyl side of one of the two laminates. In one method they tried to Spread the spores evenly over the seal area by using surface active agents (e.g. triton X-lOO), but found inconsistent results and a large amount of contamination. In the second method the laminate was inoculated with .01 ml of the spore suspension and allowed to dry. The D values obtained from.the second method varied 320° F from 3.5 - 5.2 minutes and were equal to or higher than the D value for the spores heated on tin cups in a dry atmosphere. They postulated that the higher resistance found when the organisms were heated between simulated seals was due to the restricted volume surrounding the Spores which prevented loss of Volatiles from the spores. The effect of the support nmdium (vinyl vs. tin) was not considered. III. EXPERIMENTAL PROCEDURE A. Heat resistance of Bacillus subtilis_5230 spores 1. Prgparation of spore suspension The organism used in this study is a meSOphilic spore form- ing aerobe known as strain 5230. Strain 5230 is physiologically and morphologically the same as Bacillus subtilis except that it will grow anaerobically in the presence of fermentable carbohy- drates (Sisler, 1961). Dr. C.F. Schmidt of Continental Can Company provided the original culture which he obtained from H. R. Curran as a 15p strain. The number 5230 used by Schmidt was retained. a. Growing of spores An aliquot of the original stock suspension was heat shocked and inoculated into five screw capped tubes containing 10 ml of dextrose tryptone starch broth (10 g. dextrose, 5 g. tryptone, 5 g. starch, and .4 g. bromcresol purple indicator per liter) and incubated for 24 hours at 37° C. A loop from each tube was then transferred to five tubes of broth which were incubated 24 hours. The Sporulation media consisted of Difco nutrient agar containing .5% glucose and 1 ppmMn++ from.MnSO4. H20. First .5 percent flucose was added to the nutrient agar which was autoclaved for 15 minutes at 2500 F. Then a solution of MnSO . H20 (100 ppm Mn++) was prepared, sterlized and added to 4 the agar just prior to pouring the plates. After the plates 26 were poured and allowed to harden, they were streaked with the organisms from the second set of screw capped test tubes; the plates were incubated at 37° C for 108 hours. b. Harvesting and cleaning of Spores In the first step of harvesting the Spores, the plates were flooded with sterile distilled water and scraped with an "L" shaped glass rod to loosen the adhering spores. To remove clumps of agar, the spore suspension was then poured directly from the plates through a sterile funnel containing glass wool wrapped in cheese cloth. The initial centrifugation was carried out in four 250 ml centrifuge bottles containing glass beads; the bottles were filled 3/4 full with the spore suSpension, centrifuged and the supernatant discarded. The bottles were re- filled and process repeated until the Spores were concentrated in the bottom of the four bottles. These Spores were then con- centrated into two centrifuge bottles by filling two of the bottles 3/4 full with sterile distilled water, shaking for 10 minutes to resuspend, and combining these into the other two bottles. These were shaken for 10 minutes, centrifuged, and the supernatant discarded. The Spores were washed four times in sterile distilled water by resuSpending, centrifuging, and discarding the supernatant. The spores were then washed in M/15 phosphate buffer pH 7.0 three times and stored in this buffer at 35° F. 28 2. Examination of spores and preparation of stock A direct microscopic count was made of the Spore stock .suspension in order to approximate the number of Spores in the stock suspension. Slides were prepared by Spreading .01 m1 of the organisms, diluted 1/100 on an area of 1 cm2, allowing them to dry, and fixing them in a flame. The organisms were then stained by flooding the slide with 5% aqueous malachite green, slowly heating under a flame until steam appeared, and rinsing the slide with distilled water. The slide was then counter- stained with .5% aqueous safranin which was allowed to remain on the slide 30 sec; and was rinsed with distilled water and allowed to dry. Microsc0pic examination of the slide showed the presence of nearly 100% Spores. (Spores appear green and vegetative cells appear pink.) Ten fields were counted under the microscope and the counts averaged. The average was found to be 32. The spores/ml were calculated as follows: Spores/ml - average spores/field X microscope factor X dilution factor Spores/m1 - 32 x 660,000 X 100 = 2.1 x 109 An appropriate dilution (1/100) was made in MLngphOSQhate buffer 2H_Z‘Q to obtain a substock sOlution of approximately 107 spores per ml which would be used in the test of heat resistance. 3. .Ereparation of Spores for heat treatment Sample cups [11 mm outside diameter by 8 mm deep and formed from tinplate 0.008 in thick (Pflug and Esselen, 1953)] were 29 first washed in butanone, then in absolute alcohol, dried, placed in petri dishes, and dry heat sterilized at 350° F for two hours. Rectangular pieces (3/8 in X 3/4 in and 3/8 in X 3/8 in) of a mylar-foil-vinyl (MFV) laminate (The mylar, foil, and vinyl films are .0005 in, .00035 in, and .003 in thick, respectively.) were cut after being washed with alcohol; these could not be heat sterlized because they would curl and wrinkle. The 3/8 in X 3/8 in pieces Of the MFV laminate were put in cups with the vinyl side exposed and used for testing the rate of dry heat destruction of the Spores on the vinyl. The 3/4 in X 3/8 in MFV pieces were placed with the vinyl side exposed in sterile plastic petri plates and used for testing the rate of heat destruction of ngillus Subtilis 5230 in simulated seals. A micrometer syringe (Roger Gilmont Instruments, Inc.), with a capacity of 2.0 ml where the smallest division was .002 ml, was used for diSpensing .01 ml aliquots of the spore sub- stock suspension directly into the tin cups and also on the vinyl side of the laminates which had been previously placed in cups and petri dishes. The syringe was cleaned by immersing the precision glass part in sulfuric acid-dichromate cleaning solution for a few hours and rinsing in distilled water, and by cleaning the re- maining parts with detergent. The syringe was autoclaved at 250° F for 30 minutes and oven dried. 30 Prior to filling the syringe, the spore suSpension, held in a glass jar containing glass beads and a magnetic stirring bar, was stirred for 5 minutes while immersed in an ice bath. To avoid incorporation of air which would result in formation of air bubbles, the needle was not attached when the syringe was filled. After inoculating the cups and laminates With..01 ml aliquots of the Spore suSpension, they were placed in a vacuum drier and dried for 24 hours under 29 in vacuum.at room temperature. After removal from the drier, the cups and laminates were stored at room temperature in a desiccator containing CaClz. In no case were the cups or laminates containing dried Spores stored more than one week before use. 4. Enumeration of initial number of Spores Dextroseutryptone-starch agar (dextrose-tryptone-starch broth + 15 g. agar per liter) was used for the determination of initial counts. Counts were made by placing random inoc- ulated cups or laminates into 10 ml dilution blanks immediately following the diSpension of the Spores suSpension, heat shock- ing for 10 min at 212° F, shaking 5 minutes, making the prOper dilutions in distilled water, and plating 1 ml. After the agar was poured and allowed to solidify, a thin agar overlay was poured to prevent the formation of large Spread- ing colonies. Usually five samples were plated in triplicate. The plates were incubated 48 hours at 37° C and counted. The initial spore counts were determined periodically and served to 31 note possible changes in the initial Spore concentration. Counts of the dried spores were made to check for possible lowering of the initial number during drying. The counting procedure used was the same as previously described except that the 10 ml blanks containing the cups or laminates were hand shaken 15 minutes to insure complete removal of the spores. 5. Heat treatments a. Fraction neggtive (EN) tests After heating, the cups or laminates were subcultured in 8 ml of dextrose tryptone starch broth contained in 22331 pyrex tubes fitted with resilient plastic foam plugs. These tubes were incubated at 37° C for 2 weeks after the last positive tube appeared. Growth was evidenced by turbidity and a white pelicle formation on top of the broth. The results of these tests were used with the modified Schmidt method to calculate D values which were plotted in the form of thermal destruction curves. (1) Wet heat The rate of wet heat destruction of Bacillus subtilis 5230 was measured at 230° F, 240° F, 2500 F and 260° F. The thermo- resistometer described by Pflug (1960) was used for heating samples in aluminum trays at 230° F and 240° F. The equipment and procedures used are the same as described by Augustin (1964). (2) Dry heat Dry heat conditions were attained by placing the cups in 32 thermal death time (TDT) cans [Townsend g£.§l, (1938)],hermet- ically sealing them in air, and heating them in minature retorts. Ten cups were placed in each TDT can which had been punched to accomodate the cups and to hold them in position. The cans were sealed with a Dixie Automatic Can Sealer (Dixie Canner Go.) To insure that the cans would not leak, the seals were coated with a layer of epoxy resin (100 pbw Epocast 10-F and 7 pbw Hardener 951, manufactured by Furance Plastics Inc.), which was allowed to harden 8-10 hours. The cans were then subjected to Specified heat treatment and water cooled. The cups were subcultured by asceptically opening the cans with a flamed can Opener and placing the cups in the subculture media with a flamed tweezers. (3) Heating in simulated seals The purpose of these experiments was to subject the spores to an environment of essentially zero volUme where little or no volatiles would be lost from the Spores during heating. Pieces of the inoculated MFV laminate (3/4 in X 3/8 in) were aligned with similarly sized pieces of a sealed backing (made up of two pieces of the MFV laminate sealed together). The vinyl side of the inoculated MFV laminate was placed against the mylar side of the sealed backing. These were clamped to- gether with the clamping device as shown in Figure 111-1. The seal was put in the clamping device which consists of two metal blocks (1/2 in X 1/2 in and 1/8 in thick) held together with hose clamps. These were hand tightened. Ten samples were pre- 33 pared for each test time, placed in a perforated metal rack, and heated in minature retorts. At the end of the heating time, the retorts were exhausted, and the clamps containing the seals removed from.the baskets and placed in a large sterile petri plate. After each clamp was removed from the petri plate and loosened, the seal was then removed from the clamp and pulled apart; both the inoculated piece of MFV laminate and sealed backing were subcultured. b. Survivor curves (1) Method Because the results of the partial destruction tests were variable, survivor curves were determined. In these experiments simulated seals were heated in TDT cans to eliminate the possi- bility of steam penetration into the seals. Because the con- ditions (e.g. humidity) in different cans might be variable, the cup, vinyl, and simulated seal samples were all heated in the same can during each heating time. Three samples each of cups, vinyl, and simulated seals were placed in each can (making nine total samples) and arranged as shown in Figure III-2. The simulated seals consisted of a sealed backing (3/4 in X 3/8 in) pressed against the inoculated vinyl side of a 3/4 in X 3/8 in laminate. Each simulated seal was formed by first plac- ing two sealed backings in the bottom of a pressed TDT can, aligning* these backings on top of each other, placing the in- oculated laminate on tOp and aligning it with the backings. 34 A cup was placed on top of the seal area containing the organisms. When the can was sealed, the pressure of the can lid on the cup kept the seal pressed together. The extra sealed backing in- creased the pressure of the lid on the cup which in turn, in- creased the pressure on the seal. After placing the samples in TDT cans, the can were sealed as described above. After heating, the cans were Opened aseptically and the samples placed in screw cap tubes containing ten ml of sterile distilled water. These were hand shaken ten minutes, diluted, and plated. The plates were poured with dextrose-tryptone- starch broth and incubated at 37° C for 48 hours. The plates were then counted; the results were averaged and plotted in the form of survivor curves. (2) Statisgigal analysis of data The mean, standard deviation, and 95 percent confidence intervals (Dixon and Massey, 1957) of the replicate samples heated at each time were calculated by first taking the loga- rithm of each sample and then computing these values. Replicate samples which deviated from their mean by more than two standard deviations and were discarded; these were considered to be a result of experimental error. After discarding these samples: the means, standard deviations and 95 percent confidence limits were calculated for the remaining samples. The refined data were used to calculate the least square regression lines of the linear portion of the semilogarithmic Figure III-l. Simulated seal in the clamping device. Figure III-2. Placement of cups, cups containing vinyl, and simulated seals in a TDT can. 36 survivor curves. The general equation for a linear curve may be written as Y = bX + a (III -1) where a = a constant; the value of Y at X = 0 b = the slcpe of the line X = independent variable Y a dependent variable In contrast, the equation for the straight line portion of a survivor curve used for regression was ‘U D_+ log N (III -2) 1°3 N" = '7 A D where NA = the apparent initial pOpulation determined by the intersection of the straight line portion at U = 0 Comparing equations III -1 and III -2, we find that a = log NA b=1/D K = U Y = N u The survivor curves were tested for linearity by testing the null hypothesis that the curve is linear against the alter- native that the curve is not linear at some chosen level of significance. The F statistic for testing linearity is the 37 ratio of the mean squared deviation about regression to the mean squared deviation within groups of samples. n, n k 1 ~ __ 2 k i _' 2 F(k-2,n-k)= s Z(Y,,-Yi-) t 20:. -r1) (III-3) i=1 j=1 13 ° i=1 j=l 13 ' where F(k - 2, n - k) = F statistic with k-2, n-k degrees of freedom, resPectively. Yij = jth sample in the ith group of samples T3. = mean of the samples in the ith group °ij = regression value of the jth sample in the ith group n = total number of samples ni = number of samples in the ith group This F statistic is compared to the tabular F statistic, F1,a(k°2’ n-k), at the chosen level of significance, a. If F(k-2, n-k) S F14y(k-2, n-k), the hypothesis that the regression curve is a straight line is accepted; if not, the null hypothesis is rejected and the alternative hypothesis that the curve is not linear is accepted. (Dixon and Massey, 1957). The hypotheses that the slcpes of the survivor curves (-l/D) are equal and the zero time intercepts (log NA) are equal were tested using the data for spores heated on tin vs vinyl, on tin vs simulated seals, and on vinyl vs simulated seals at each temperature against the alternative hypothesis that these are not equal using a t statistic. The calculated t statistic was obtained by pooling the standard errors of the regression co- 38 efficients (slopes or 0 time intercepts) and dividing this pooled standard error into the difference in means. For example: if the null hypothesis (HO) that two lepes are equal is being tested against the alternative (H1) that they are not equal or o l 2 H1: B1 ¥ B2 then n +11 '4 b - b t(-—---12 2 ) = . 1 2 (111 -4) sb + Sb 1 2 where n1 + n2 - 4 n1 +n2 - 4 t( 2 ) = t statistic with 2 degrees of freedom B1 and B2 = the actual lepes of the survivor curves b1 and b2 = estimates of the lepes, B and B , reSpectively, obtained from sampling n1 = the number of samples used to determine b1 n2 = the number of samples used to determine b2 3 and s = the standard errors of b and b respectively b1 b2 1 2 b1 ' b2 J7=1f====1f = the pooled standard error 8 + 8 b1 b2 n1 + 112 - 4 The calculated t statistic, t( 2 ), was compared to the n1+n2-4 n1+n2-4 tabulated t statistics, t1/2a( 2 ) and t1—1/my( 2 ) 111 + n2 - 4 at the chosen level of significance, 0. If t < tl/Za( 2 ) 39 111 + n2 - 4 l-l/ZO/( 2 equal is rejected and we conclude that the lepes are not equal or t > t ),the hypothesis that the slopes are at the a level of significance; if not, the hypothesis that the slopes are equal is accepted. The means, 95% confidnece intervals, sums of squares, regression coefficients and their standard errors were computed using the Control Data 3600 Digital Computer. B. ‘Measurement.of heating andggpoling rates in flexible package seal . Heating and cooling rates were measured at the seal interface of a flexible package made up of two MFV laminates (described in section IIIeA-3) during heat sealing. 1. Preparation of the thermocouple arrangement The thermocouple arrangement is shown in Figure III-3. An uninsulated copper-constantan thermocouple with 3 inches of the .002 in diameter wires was placed on the vinyl side of a laminate so that the juction would be located in the center of the seal- ing area. The sealing area was indicated by outlining it with black wax-pencil. When the thermocouple was in place the COpper and constantan wires were Spread apart, each wire was glued in two places, just below the seal area and about 3 in farther down. Three feet of copper-constantan wire, with enameled, glass wrapped, fiberglass overwrapped insulation, was used as a lead and was shaped so it could be connected to the .002 in thermocouple unit. It extended out the side of the package near the bottom. After stripping both ends of insulation and separat- 40. thermocouple -- 5“" junction A 0'“ .002 in. Cu-Con thermocouple mre Mue-*<) C) x 3:. soldered junctions 30 gauge Cu-Con Ihermocouple wire \ Figure III-3. Diagram of the thermocouple arrangement used to measure the temperature in the center of the seal. (The diamgram.shows the top laminate removed to expose the thermo- couple arrangement on the bottom laminate.) 41 ing the copper and constantan wires, the 30 guage thermocouple was glued to the laminate in three places to hold it in place. The c0pper-c0pper and constantan-constantan junctions.of the 30 guage and .002 in thermocouple wires were then soldered. Another laminate, which had previously been marked to indicate the seal area, was aligned on t0p of this arrangment with the vinyl side on the inside and taped to the other laminate below the seal area. The thermocouple arrangement was checked for cold soldered joints by measuring a known temperature (in this case boiling water) and was found to read correctly. Figure IV-4 shows the .002 in thermocouple wire junction whose diameter was found to be .0065 in. 2. Heating and cooling of the seal The temperatures were measured and recorded every l/2 second by a multi-channel one minute cycle recording potentio- meter by attaching the 30 guage lead wire extending from the package to a cable which was connected to the potentiometer. The package was sealed with a single heated jaw sealer (Pack- Rite Poly-Jaw sealing heads) using an air pressure of 40 psia. The heated jaw temperature was controlled by a Brown direct control potentiometer (Brown Instrument Co., Philadelphia, Pa.). The cold jaw contained rubber insulation. The temperature of the jaw was recorded before the sealing Operation began and after cooling of the seal by a 30 guage thermocouple attached to the sealing bar. Temperatures at the seal interface of the 42 Lei .01 inches I I fl Figure III-4. Photomicrograph of the thermocouple junction used to measure temperatures in the seal taken on a vinyl background (The junction diameter is .0065 in.) 43 package were recorded during 6 sec. heating and cooling cycles. During cooling the package was removed from the jaws and held horizontally. 3. Plotting_of curves and description of their parameters Heating and cooling curves were plotted in a convenient form for practical application by plotting the unaccomplished temperature difference versus time on semilogarithmic paper. The ordinates of the graphs of heating and cooling curves were labeled with the actual temperature ccrresponding to the un- accomplished temperature difference (Ball and Olson, 1957). The equation for the straight line asymtote to heating and cooling curves is _.:E 15 - _ - log (T1 - T) - f + lug J (T1 To) (III 5) where f = the temperature response parameter j = the lag factor, (T1 - TA)/(T1 - To) t a time T a temperature at time, t T a apparent initial temperature; ordinate of the origin A of the asymtote. T1 = heating (or cooling) medium temperature To = initial temperature Both f and j are related to the terms appearing in the equations derived from conduction theory. At sufficiently large values of t all terms of the condition equations vanish except the first. A first term approximation of the conduction equations yields an expression which is equivalent to equation(III -5)- Comparison of these equations establishes the relationship of f and j with the prOperties of a system. The temperature response parameter, f, is the time required for the straight line asymtote to traverse one log cycle or, similarly, the time required for a 90 percent change in the log of the temperature difference in the straight line portion of the curve. f is a function of the Biot number, N31, thermal diffusivity, a, and geometry of the object being heated, but is independent of position and initial temperature distribution in the object (Kopelman, 1966). N = ._ ‘ (III -6) where r = the radius of a Sphere or cylinder or the half thick- ness of a slab. The lag factor, j, is dependent on the geometry of, initial temperature distribution in and position in an object, and the Biot number of the system. j is independent of p and Cp' IV. DERIVATION OF EQUATIONS USED TO CALCUIATE TEMPERATURE PROFILES IN SEALS Temperature profiles have been calculated for heat sealing of homogenous laminates (laminates made up of one type of film, e.g. polyethylene) using exact solutions for unsteady state heat conduction (Ridgeway, 1965). In the case where two or more films make up each laminate, the task of calculating the temperature using exact equations becomes quite complicated. For this case, I have applied numerical analysis using relaxation methods. Relaxation methods were applied by deriving approximate equations for unsteady state heat conduction throughlincrementalfithick- nesses and iterating these equations over small time intervals to obtain the temperature profiles. These methods are described in Bennett and Meyers (1962) and Dusinberre (1949). The profiles were calculated for sealing of laminates with a single heated jaw sealer. (Both are described in the experimental procedure.) A. Description of model and assumptions made FigureIV-I is a diagram of a cross section of the sealing system (taken along the heat flow direction) which consisted of a single heated jaw and an insulated cold jaw contacting the two laminates. The following were assumed in deriving the equations: 1. The heat flow is one dimensional. 2. Thermal contact resistance between the heated jaw and mylar, the vinyl films, the mylar film and rubber insulation, can be neglected. 46 wofimom 3 :3me ”.05 :33“..qu 30am uoofi. on”. macaw noxmu aoumhm magnum can no coauoom $95 .75.. 9.53m MI. mam 52:23.. .5 WW 3... 38 F $82. .5 a. . add»: . .3. 300. do“. .2. 389m . ._>z_> 4.. 8o. SEES: 2mm x ._>z_> .2. 3o. . .__o.._ .2. 38°.” ”24>: .2. 33. “HF 34.. om._. '01 'oa'Niw'co'o I 63 i = I5.2 °F l 1111 l I J 09 .08”- .07'- .06r- .OSF- .04 l 230 240 250 260 Temperature l°Fl Figure V-l. Thermal resistance curve of B. subtilis 5230 Spores heated on tin cups in wet heat (95% confidence limits are shown for each point.) D Volue (min) 64 '88 80 —- J 70 L- - —4 60 — — 50 - —3 z = 27.5 °F lg — 8 C 7 .— 6L. 5 r— 4 l—- 3 ._ 2 —- _ I285 295 305 315 Temperature l°Fl Figure V-2. Thermal resistance curve of B. subtilis 5230 Spores heated on tin cups in dry heat (95% confidence limits are shown for each point.) 65 Another possible source of error leading to a lower observed resistance was the slow cooling of the seal in the clamping device. After heating, the clamping devices containing seals were removed from the retort and placed in a large petri plate. One by one the seals were removed from the clamps and sub- cultured. Cooling took place by evaporation of condensed moisture, heat loss to the petri plate, and heat loss to the air by convection. An approximate calculation of the heat lost by convection to the air (h = 3) showed that the temperature of the seal would drop from 3150 F to 2800 F in about 5 minutes. If most of the cooling took place by heat loss to the air by convection, the lethality accumulated during cooling would be significant especially at 3150 F. However, the results did not indicate a tendency for the seals subcultured last to show a higher fraction negative than those subcultured first. (Re- moving, pulling, and subculturing ten seals required 15 to 20 min.) An approximate calculation of the time required for the seals to reach retort temperature showed that the temperature of the seals would rise from 800 F to 314.90 F in 6 sec. Another possible source of error which would result in a lower observed D value than the actual is penetration of steam into the seal by diffusion or possibly capillary action. All clamps were observed to be more loose after heating than before. If some clamps loosened during heating, steam could penetrate and wet heat conditions then would prevail. This would result in a higher rate of destruction than expected for dry heating alone. 66 Table V-3. Results of Dry Heat Fraction Negative Tests of B, Subtilis 5230 Spores on Vinyl and in Simulated Seals 315° ‘ Date 12116l65 12/30/65 Test 77 Test 78 Test 79 Test 80 Simulated Vinyl Simulated Vinyl Seals Seals Time Time (min) 13h: 13h: (min) p/r p/ r 5 10/10 10/10 18 1/10 7/10 10 9/10 10/10 20 2/10 6/10 15 10/10 10/10 27 1/10 1/10 20 10/10 9/10 24 0/10 3/10 25 3/10 0/10 26 2/10 1/10 :28 1_2/10 0(10 _Q(mig) 4.35 --- #f§.l4 4.25 9SZCLD --- --- --- 4.00-4.50 Tate Tf6/66 7 MM 3/4/66 Test 83 Test 84 Test 94 Test 93 Simulated Vinyl Simulated Vinyl Seals Seals Time Time (min) p/ r all: (min) p/ r p/ r 12 10/10 9/10 12 2/10 10/10 15 8/10 6/10 15 1/10 8/10 18 6/10 3/10 18 0/10 9/10 21 2/10 0/10 21 0/10 5/10 24 -2/10 2/10 24 0/10 0/10 27 0/10 0/10 27 0/10 0/10 30 1/10 0/10 . 30 0/10 0/10 kain) -3.ng 3.27 1.97 3.94 67 Table V-3. (continued) 305°F. Date 1 9166 l[Z/66 #211166 Test 86 Test 85 Test 89 Test 90 Simulated Vinyl Simulated Vinyl Seals Seals Time Time (min) p/r p/r (min) p/r p/r 20 10/10 10/10 20 10/10 10/10 30 2/10 7/10 24 10/10 10/10 40 . 1/10 1/10 28 2/10 10/10 50 0/10 0/10 32 1/10 8/10 60 0/10 0/10 36 0/10 5/10 70 0/10 --- 40 0/10 5/10 Dimin) 4048 6.43 4097 7.11 95%CLD 3.11-5.85" 5.35—7.51 4.12-5.82 6.52-7.70 Date 3/4/66 Test 92 Test 91 ' Simulated Vinyl Seals Time Time (min) p/r (min) Wt 22 4/10 25 10/10 25 1/10 30 10/10 28 0/10 35 10/10 31 0/10 40 5/10 34 0/10 45 5/10 37 1/10 50 2/10 . D(min) 4.25 8.41 68 Differences in tightness of the clamps (these were hand tightened) may account for some of the variability in results. To eliminate the possibility of steam penetration and reduce the variability of results, simulated seals were heated in TDT cans along with cups and vinyl. The raw and adjusted counts, means and 95 percent confidence limits of the adjusted data are given in Tables V-4, 5, 6, 7, 8, 9. The results of linear regression on the second portion of the broken survivor curves (Figures V-3 and V-4) are shown in Table V-10. In testing for linearity of regression (a a .01) on the second portion of the broken survivor curves, all curves were found to be linear except that obtained for heating of spores on vinyl at 2800 F (Table V-ll); however, this curve was drawn linear for sake of comparison (Figure V-3). The point shown at 25 min. may actually belong to the first portion of the broken survivor curve. This would explain why the curve was not statistically established to be linear. Another reason for the apparent non-linearity could be experiment error in determining the point at 50 min. The results of the D values determined by linear re- gression (Table V-lO) indicate that the D value of the Spores heated at 3050 F in simulated seals is higher than the D values of the organisms heated at 3050 F on tin cups and vinyl. When the D values were statistically tested for equality or inequality, (Table V-ll), the D value of spores heated in simulated seals was 69 found to not equal those of Spores heated on vinyl and tin cups (0 = .01). [Rejecting the hypothesis that the D values are equal at the .01 level is said to be "highly significant". (Dixon and Massey, 1957).] However, at 2800 F, the D value of Spores heated on cups appears to be higher than those of Spores heated on vinyl or in simulated seals. Statistical testing of equality or inequality of D values at 2800 F indicated the D value of Spores heated on cups was different than the Spores heated on vinyl at the .01 level, but no difference could be established between the D value of Spores heated in cups and that of spores heated in simulated seals or between D value of Spores heated on vinyl and that of Spores heated in seals. The D values determined by survivor curves were in all cases higher than those calculated from.FN tests conducted at the same temperature using the same support medium. The D values obtained from survivor curves for dry heating of Spores on cups, on vinyl and in simulated seals were 9.74, 10.47, and 24.47, re8pectively while those obtained from fraction negative tests were 7.31, 6.43-8.41, and 4.25-4.97, respectively. The lower D values calculated from FN tests partly resulted from assuming the survivor curves to be entirely linear. Figures V-4 and V-5 show that these curves are actually broken. From inSpecting these figures we can see that the D values of the second portions of these curves will be higher than those obtained from the straight lines drawn between the initial number, No’ and NU = .69 (the most probable number UFNSO). In de- I 70 . . . . - r riVing the equation for most probable number, x = ln- q) Halvorson and Ziegler (1933) assumed that only one viable organism was needed for growth to occur in a sample. If more than one organism is required, the i will be lower than the number of organisms surviving. This would result in a lower D value than the actual and may partially account for the lower D values obtained from FN tests. The reasons mentioned above probably do not explain why for heating in simulated seals at 3050 F the D value obtained from the survivor curves (24.47) was so much higher than those obtained from FN tests (4.25-4.97). The lower D values observed for FN tests probably resulted from steam penetration into the seals held in hose clamps. The first reason which probably comes to mind to explain the breaks in the survivor curves obtained in this study is that these broken curves result from having two populations of organisms with different resistances. However, we have found that the wet heat survivor curves of these same organisms vary from being straight at high temperatures (e.g., 2500 F) to concave downward at low temperatures (e.g., 2000 F). If the broken curves result from two populations of organisms, the zero time intercepts (NA) of the second portions of these curves would represent the population of the more resistant strain. (The zero time intercepts are listed in Table V-lO). At 3050 F the N for simulated seals was found to be signif- A icantly lower than that for vinyl or cups (0 = .01, Tahle 7l V-ll), but at 2800 F, although the NA for simulated seals was somewhat lower than that for vinyl or cups, no difference was established between them, by statistical testing. A lower NA for simulated seals can be explained by trapping of organisms in the vinyl or mylar caused by melting of the vinyl and pressure on the seal. Since steam may have penetrated into the simulated seals heated in steam, the results of dry heating of simulated seals in TDT cans will be used for comparing the rate of dry heat destruction of Spores in seals to that of Spores on vinyl. If survival of bacteria depends on keeping some part of them moist as proposed by Scott (1958) and Pflug (1966), increasing the rate of water loss from bacteria will increase their rate of dry heat destruction. If so, one would expect the resis- tance of Spores heated in seals to be higher than that of spores heated on vinyl. The results show that this is the case at 3050 F but not at 2800 F. It is possible that during the longer heating times (at 2800 F) more organisms become en- trapped in the vinyl portion of the simulated seals. This would result in a lower observed D value than the actual. Also, since the spores in the seal were exposed to vinyl on one side and mylar on the other, the heating of Spores on vinyl may not have been a good control. In the small seal area, release of water from the spores may have resulted in an increase in the water content of the small amount air inside (if any) during 72 heating. If so, variations in tightness of the seals would lead to different water contents of the air inside each seal. This could partially explain the erratic data obtained for heating in seals (Tables V-6 and V-9). The results of dry heating on cups and vinyl show the effect of heating on different support media. The D values for heating on vinyl compared to cups is higher at 3150 F, the same at 3050 F, and lower at 2800 F indicating that the 2 value for heating on vinyl is higher. The porous vinyl may protect the spores by preventing water loss at short heating times (high temperatures). At longer heating times (lower temperatures) this protective affect may be overridden by a release of toxic gases from the vinyl.- 73 Table V-4. Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heating of‘g. Subtilis 5230 Spores on Tin Cups at 280° F. Number of spores Time (min) Raw Adjusted Mean 95%Confidence Limits x 104 x 104 x 104 x 104 O 6.0 6.0 6.4 6.4 6.2 6.2 6.45 5.70 to 7.26 6.9 6.9 7.0 7.0 5.4 5.4 x 103 x 103 x 103 x 103 25 1.3 1.3 2.3 2.3 1.16 to 3.49 3.3 3.3 2.3 2.3 2 02 3.3 3.3 .9 .9 x 103 x 103 x 103 x 103 50 .66 1.34 1.34 1.24 .933 to 1.53 1.24 1.021 1 20 1.02 1.53 1.53 94 x 102 x 102 x 102 x 102 75 6.5 6.5 3.2 3.2 2.7 2.7 4.82 3.08 to 7.56 7.9 7.9 6.3 6.3 4.5 4.5 x 102 x 102 x 102 x 102 100 '3.6 1.2 1.2 1.4 1.16 7.35 to 18.3 1.4 1.9 1.9 .8 .8 .8 .8 74 Table V-4 (cont.) X 10 X 10 X 10 X 10 125 N N U1\IOMP|V\fl WNO-PNU! .809 .379 to 1.73 N N The means and 95% confidence limits were calculated logarithmically. Table V-5. Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry 0 Heating of g. Subtilis 5230 Spores on Vinyl at 280 F Number of spores Time (min) Raw .Adjusted Mean 95%Confidence Limits x 10 x 104 x 10 x 10“ O 5.1 5.1 7.3 7.3 6.4 6.4 6.21 5.42 to 7.08 6.0 6.0 6.9 6.9 5.8 5.8 x 10 x 103 x 10 x 103 25 1.2 1.2 3.4 3.4 1.6 1.6 2.29 1.45 to 3.62 2.6 2.6 307 3.7 2.3 2.3 x 102 x 102 X 10 X 102 50 2.8 2.8 2.4 2.4 3.28 2.28 to 4.72 3.1 3.1 5.2 5.2 13.7 3.5 3.5 x 102 x 102 x 10 x 102 75 1.4 1.4 1.4 1.4 1.8 1.8 1.71 1.32 to 2.21 2.3 2.3 1.8 1.8 5.1 Table V-5 (cont.) x 101 x 101 x 101 x 101 100 4 4 9 9 4 4 5.10 3.25 to 3.01 4 9 6 6 4 4 The means and 95% confidence limits were calculated logarithmically. Table V-6. Counts, Adjusted Counts, and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heating of B. Subtilis 5230 Spores in Simulated .Seals at 280° F Number of Spores Time (min) Raw Adjusted Mean 95% Confidence Limits x 104 x 104 x 104 x 104 0 6.1 6.1 6.3 6.3 6.3 6.3 6.00 5.17 to 6.94 4.8 4.8 7.3 7.3 5.5 5.5 x 103 x 103 x 103 x 103 25 2.5 2.5 .1 .1 .8 .8 .831 .167 to 4.13 2.2 2.2 .9 .9 x 103 x 103 x 103 x 103 50 1.89 1.89 .99 .99 .795 .178 to 3.56 .94 .94 .10 .10 1.81 1.81 x 102 x 102 x 102 x 102 75 0 0 3.7 3.7 .2 .2 .890 .114 to 6.95 1.7 1.7 .5 .5 78 Table V-6 (cont.) X 101 x 101 x 101 x 101 100 0 0 1 1 4 4 2.99 .925 to 9.67 5 5 20 4 4 The means and 95% confidence limits were calculated logarithmically. Table V-7. Counts, Adjusted Counts and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heating of g. Subtilis 5230 Spores on Tin Cups at 3050 F Number of Spores Time (min) Raw Adjusted Mean 95% Confidence Limits x 104 x 104 x 104 x 104 0 7.7 7.7 7.3 7.3 7.79 8.4 8.4 x 103 x 103 x 103 x 103 3 7.5 7.5 5.66 to 9.35 7.9 7.9 6.5 6.5 7.28 x 103 x 103 x 103 x 103 5 5.3 5.3 4.2 4.2 6.6 6.6 4.62 3.61 to 5.93 4.8 4.8 4.5 4.5 3.3 3.3 x 103 x 103 x 103 x 103 10 1.30 1.30 1.40 1.40 1.37 1.37 1.39 1.27 to 1.51 2.23 1.33 1.33 1.55 1.55 x 102 x 102 x 102 x 102 15 5.74 3.45 to 9.53 1d1d U1uac>c~uru wVO-l-‘O‘D MO-PWV DOC-PO‘D 80 Table V-7 (cont.) x 101 x 101 20 8 8 11 11 13 13 X 10 10.4 1 x 101 5.58 to 19.3 The means and 95% confidence limits were calculated logarithmically. T, . ‘“-. 81 Table V-8. Counts, Adjusted Counts and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heating of‘g. Subtilis 5230 Spores on Vinyl at 305° F Number of spores Time (min) Raw Adjusted Mean 95% Confidence Limits x 104 x 104 x 104 x 104 0 6.8 6.8 7.1 7.1 7.40 8.4 8.4 x 103 x 103 x 103 x 103 3 6.4 6.4 6.3 6.3 7.36 3.90 to 13.9 8.7 8.7 x 103 x 103 x 103 x 103 5 6.3 6.3 6.0 6.0 4.3 4.3 6.06 4.97 to 7.40 6.2 6.2 6.4 6.4 7.7 7.7 x 103 x 103 x 103 x 103 10 2.03 2.03 1.56 1.56 1.76 1.76 1.90 1.58 to 2.30 3.38 2.34 2.34 1.92 1.92 X102 X102 X102 X102 15 14.4 14.4 6.7 6.7 13.0 13.0 6.23 3.60 to 10.8 5.9 5.9 5.5 5.5 5.9 5.9 82 Table V-8 (cont.) x 10 x 102 x 102 x 102 15 14.4 14.4 6.7 6.7 13.0 13.0 6.23 3.60 to 10.8 5.9 5.9 5.5 5.5 5.9 5.9 x 102 x 102 x 102 x 102 20 3.3 3.3 1.7 1.7 1.99 .651 to 6.07 1.4 1.4 The means and 95% confidence limits were calculated logarithmically. with“ I - n . - 83 Table V-9. Counts, Adjusted Counts and Means and 95 Percent Confidence Limits of the Adjusted Counts for Dry Heating of g. Subtilis 5230 Spores in Simulated Seals at 305° F 7 Number of spores Time (min) Raw Adjusted ZMean 95% Confidence Limits x 104 x 101 x 104 x 104 0 5.3 5.3 6.3 6.3 5.88 4.59 to 7.21 6.1 6.1 x 103 x 103 x 103 x 103 3 4.3 4.3 6.6 6.6 1.42 .00463 to 43.3 .1 .l ‘ x 103 x 103 x 103 0 x 103 5 2.9 2.9 3.8 3.8 2.0 2.0 2.57 1.23 to 5.36 3.9 3.9 4.7 4.7 .7 .7 x 103 x 103 x 103 x 103 10 1.41 1.41 1.29 1.29 2.75 2.75 1.54 .565 to 4.12 .46 .46 3.60 3.60 x 102 x 102 x 102 x 102 15 10.2 10.2 5.6 C5.6 9.9 9.9 7.65 5.45 to 10.7 4.7 4.7 8.3 8.3 9.0 9.0 84 Table V-9 (cont.) x 102 x 102 20 3.0 3.0 9.6 9.6 3.3 3.3 X 10 4.54 2 x 102 .904 to 22.8 The means and 95% confidence limits were calculated logarithmically. a\H- . a .mHm>uouaH monmuwwaoo uamnu was dz uo manufiumwoa as» spam moundsoamo mums mafiafia ouaumfimaoo “Hana mam Auso «nu mo nomads onu scum woumasuamo mums muaafla moammamaoo Hausa was monam> a may Hv moa x n.5H ou MOH x oa.~ MOH x MN.¢ comma ou mo.mm mm.ne . mamom vmumwsafim .5 moa x mn.m ou moa x Nw.m moa x oa.o Heumm ou m~.H¢ mH.nq ahaa> a. moa x mm.m ou moa x m<.m moa x m¢.m ma.mm ou «N.nn mq.oo mmbo can owN moa x wm.~ ou moa x mm.a MQH x nu.m m.moH ou mm.ma sq.¢~ mason voumasaam sou x a~.N on sea x s~.H sea x no.4 ma.ma ou mo.m aq.oa Hana» «OH x om.H ou «OH x N~.H aoa x em.H Hn.oa ou mm.m an.m mane aau mom az «6 muaaaa Aaaavn mo ouaaaa Assay Amov muduvauaoo Nmm dz moamvamaou Nnm o=Hm> n abwvmz uuoamsm wnaumummama Nv av .mw>uao uo>fi>u5m amxoum onu mo coauuom vacuum ago no coaumunwom .umoaaq soum_wuaacuno.muaafin monopfimaou unmouom ma ufionh.vum Adzo numuuuduua oafih C van Amandu> av «QEHH nowuosvmm Hmaaomn mo muaammm .OHI> manmy 86 Table V-ll. Results of Statistical Testing of Linearity and Differences Between D Values and 0 Time Intercepts (NA) of the Second Portion of the Broken Survivor Curves. Temperature Support Linearity *D values *Intercepts, NA (0 F) medium a = .01 a = .01 a = .01 305 tin cups linear D1 = D2 NA1 = NAZ vinyl linear D2 ¥ D3 NA2 # NA3 simulated seals linear D3 3‘ D1 N‘o‘3 a! NA1 280 tin cups linear D1 # D2 NA1 = NA2 vinyl not linear D2 8 D3 NA2 = NA3 simulated seals linear D3 = D1 NA3 - NA1 *The subscripts l, 2, and 3 refer to cups, vinyl, and simulated seals, respectively. of Survivors Number 87 IG'E I —o— Tin cups - —-A—- Vinyl — .—-D-—- Simulated seals _. IO“: j C I _ I _ -4 > IOfi: E% : : ‘ \UD : .. N d _. \\ c> ‘ 1. l§\\\ 7‘ ._ \41 d \‘N\\ <3 2_ ' - IO : c&\ > : ~ \. ‘ 1.. 3 1 I00 25 50 75 I00 '25 Time (min) Figure V-3. Survivor curves of B. subtilis 5230 spores dry heated at 2800 F on tin cups, vinyl, and in simulated seals determined by linear regression (The sample means are shown for each time.) of Survivors Number 88 «----GF---' \Hnfl -——D-—- Simulolod ml: / o 5 I0 IS 20 Time (min) _ Figure V-4. Survivor curves of B. subtilis 5230 Spores dry heated at 305° F on tin cups, on vinyl and in simulated seals determined by linear regression (The sample means are shown for each time) 89 B. Heating and cooling of seals The results of the measurements of temperatures at the seal interface during heating and cooling are shown in Table V-12. Table V-13 show the results of the calculations of temperature profile in the seal during heating and cooling. The results are plotted in the form of heating and cooling curves (Figures V-5 and V-6) and as temperature profiles (Figures V-7 and V-8). The rapid initial rise and correspondingly low lag factors (j) (Table V-l4) of the heating curves shown in Figure V-S result from measurement of the temperature close to the heated surface of the laminates and the cold jaw. Although the theo- retical curve is initially steeper than the measured curve, the j of the measured curve (.293) is less than the j of the calculated curve (.502). 'The slower initial rise of the measured temperature may result from the relatively large thermocouple junction (.0065 inches in diameter) acting as a heat sink. At the beginning of heating the small amount of heat conducted to the junction will be rapidly distributed throughout it. This may result in a lower temperature than would be expected if the thermocouple was not present. The slower initial rise of the measured as compared to the cal- culated curve may also be caused by contact resistance be- tween the jaw and mylar film which was neglected in calculating the temperature response. 90 Table V-12. Results of the Measurement of Heating and Cooling Rates at the Seal Interface. T - 367° F and T = 94.1° F jaw a Temperature 0 F Time (sec) Heating Cooling 0.0 94.1 349.9 0.5 158.2 346.6 1.0 265.6 326.7 1.5 284.7 304.0 2.0 295.5 289.0 2.5 300.4 275.0 3.0 305.1 264.6 3.5 309.9 255.0 4.0 312.4 245.1 4.5 314.1 237.6 5.0 315.5 230.5 5.5 317.7 222.4 6.0 319.1 216.0 91 Table V-13. Results of Calculation of Temperature Profiles. T. . 367° F and T - 94.1 F jaw a Heating Temperature 0 F T1 T2 T3 T4 T5 T6 Time (jawsmylar (foil) (vinyl-vinyl (foil) (mylar-rubber (aluminum (sec) interface) interface) interface) bar) 0.00 367.0 94.1 94.1 94.1 94.1 94.1 0.01 367.0 289.0 106.6 94.6 94.1 94.1 0.02 367.0 329.4 127.6 96.4 94.1 94.1 0.03 367.0 339.2 146.6 98.6 94.2 94.1 0.04 367.0 342.7 162.5 100.7 94.4 94.1 0.05 367.0 344.6 175.4 102.4 94.6 94.1 0.06 367.0 346.1 185.9 104.0 94.8 94.1 0.07 367.0 347.2 194.4 105.3 95.1 94.1 0.08 367.0 348.1 201.4 106.4 95.4 94.1 0.09 367.0 348.8 207.0 107.4 95.7 94.1 0.10 367.0 349.4 211.6 108.3 96.0 94.1 0.20 367.0 351.7 230.0 114.0 99.8 94.1 0.30 367.0 352.2 234.0 118.0 103.8 94.1 0.40 367.0 352.5 236.2 121.7 107.7 94.1 0.50 367.0 352.7 238.2 125.4 111.6 94.1 1.00 367.0 353.7 247.5 142.8 130.0 94.1 1.50 367.0 354.7 256.1 159.0 14711 94.1 2.00 367.0 355.6 264.1 173.9 162.9 94.1 2.50 367.0 356.4 271.5 187.8 177.5 94.1 3.00 367.0 357.1 278.3 200.6 191.0 94.2 3.50 367.0 357.8 284.6 212.4 203.6 94.2 4.00 367.0 358.5 290.5 223.4 215.2 94.2 4.50 367.0 359.1 295.9 233.9 225.9 94.2 5.00 367.0 359.6 300.9 242.9 235.8 94.2 5.50 367.0 360.2 305.5 251.7 245.0 94.3 6.00 367.0 360.6 309.8 259.7 253.6 94.3 Cooling 92 Temperature (0 F) Time T T 1 2 (sec) (air-mylar (foil) interface) 0.0 367.0 360.6 0.1 322.3 322.1 0.2 310.7 310.7 0.5 302.7 302.8 1.0 294.0 294.1 1.5 288.4 288.5 2.0 281.7 281.8 2.5 275.2 275.3 3.0 268.9 269.0 3.5 262.0 263.0 4.0 257.1 257.2 4.5 251.4 251.5 5.0 246.0 246.1 5.5 240.6 240.8 6.0 235.6 235.7 0.0 367.0 360.6 0.1 321.5 321.3 0.2 309.1 309.2 0.3 304.2 304.4 0.5 299.0 299.2 1.0 288.4 288.6 1.5 278.4 278.5 2.0 268.9 269.1 2.5 259.9 260.1 3.0 251.4 251.5 3.5 243.3 243.4 4.0 235.6 235.8 4.5 228.4 228.5 5.0 221.4 221.6 5.5 214.9 215.0 6.0 208.7 208.8 T3 (vinyl-vinyl interface) 309.8 308.7 307.2 302.8 294.2 288.6 281.9 275.4 269.1 263.0 257.2 251.6 246.1 240.8 235.8 309.8 308.0 305.8 303.6 299.2 288.6 278.6 269.1 260.1 251.6 243.5 235.8 228.5 221.6 215.1 208.8 T4 (foil) 259.7 295.2 303.7 302.7 294.1 288.5 281.8 275.3 269.0 263.0 257.2 251.5 246.1 240.8 235.7 259.7 294.6 302.2 302.6 299.0 288.5 278.5 261.1 260.1 251.5 243.4 235.8 228.5 221.6 215.0 208.8 T5 (mylar-air interface) 253.6 294.8 303.5 302.5 294.0 288.4 281.7 275.2 268.9 262.9 257.1 251.4. 246.0 240.7 235.6 253.6 294.1 301.9 302.4 298.9 288.4 278.4 268.9 259.9 251.4 243.3 235.6 228.4 221.4 214.9 208.7 357 347 337 327 3|? 25. NM (no NN Temperature i°Fl l6? ”7 67 93 - 0 Measured 0 Calculated I l l l l 1 l. l 1 1 [1111111 0 l 2 3 4 5 Time (sec) Figure V—5. Calculated and measured heating curves for heating at the seal interface, plotted on semilogarithmic paper . .Hmama owamuanmwoauawm do mouuoam oommumuafi Hmmm on» um waaaooo you mo>uao mawaooo venomous mam vmumanoamo .ol> shaman 94 ( 3‘) ammadwal 3mm. 25 m m e m N _ o _ _ _ a _ _ e2 1&8 n . q n e a Les 4 I o a n o o a I .1¢¢N o a u . o 4 o 4 a .1¢m~ all], I d I O I e o .- as I O 4” fl/ Ivom . 2:382 G I, u II is.” 32. 6223.8 0 4 II can. 6223.8 0 6 Jim P — — P b - VOW 95 mcaammm umm: mawusm nonmwanmumm mmHHmonm muoumuuaamu vmumasoamo .nu> muowam 3:5 32 8.9.2 2: 5o: 33 2: 25 3855 ~ m m e m m o . a a .4 _ a [II / cod Too. 5.0 1 Ion. / 50/. . 66/ l a .J 8.0... / _.o 108“. no N . 0. n. 10.3 H. mm ow loom I.“ loam £5»: so“? ._>z_> ._>z_> .5 m3»: 96 3:5 37. 8.8; so: 63 25 887.5 m ¢ A.oumv 08H» wsafiooo n.u wcwaooo wawusv pmnmfianmumo Ao.m u : nomV mmHHmoum musumuoaawu umumasofimu m .flmmm man no .m-> seamen m. on. —tD — u :3»: 4.0.... ._>z_> ...>z_> 4.0m [own / fiiomm (3°) amiaiadma 1 97 The higher rise of the measured heating curve (as com- pared to the calculated) and correspondingly lower j probably results from the presence of the relatively large thermocouple in the seal. Since the thermocouple junction diameter (.0065 in) is of the same order as the thickness of the seal (.0077 in), the thermal resistance in the heat flow direction (through the seal) is appreciably smaller in the zone where the thermo- couple is located than in a zone not containing the thermo- couple. This lower resistance results in a higher measured temperature than actually exists if the thermocouple is not placed in the seal. We observed from numerical computations that reasonable changes in the thermal properties of the mylar or vinyl did not cause an appreciable change in the j of the calculated heating curve. The higher j observed for the cal- culated curve cannot be explained by neglecting the contact resistance since taking into account contact resistance will increase the j. The fs of the measured and calculated curves shown in Table V-14 are 25.3 sec and 15.7 sec respectively. The higher f of the measured curve may result from a difference between the actual thermal prOperties and those used for determining the calculated curve, by an air film.between the seal interface caused by the presence of the thermocouple during measurement, or by contact resistance between the heated jaw and mylar film. The slcpe of the so called "straight line portion" of the 98 measured curve appears to decrease with time. This may be due to a change in the thermal properties of either mylar or vinyl with temperature, absorption of heat by the vinyl during melting or a lowering of the bar temperature during heating. The fs of the measured cooling curve and the calculated curve for h assumed to be 3.0 are nearly the same (20.2 and 21.9, reSpectively) while the f of the calculated cooling curve for h assumed to be 2.0 is 33.0 (Figure V-6). An h of 3.0 is of reasonable magnitude for air flow in a room. Since the re- sistance of the air film to heat flow (1/h) is much higher than that of the seal, the rate of cooling is controlled by the mag- nitude of h as can be seen from the results; therefore, after the initial cooling period, we would expect the temperatures to be the same throughout the seal as is Shown by the temperature profiles for cooling (Figure V-8). For the first few tenths of a second of cooling the temperature rose in the cooler part of the seal due to heat being conducted from the part which had been closest to the heated jaw. The js of the measured and theoretical cooling curves are .905 and 1.0, respectively. This difference in js could have resulted from not starting the measurement at the exact time the seal was removed from the jaw, but probably resulted from different initial temperature profiles. As shown in the calculated cooling temperature profiles and as discussed above, a significant temperature gradient does 99 not develop in the seal after the initial cooling period due to the comparatively high thermal resistance of the air film. Therefore, the presence of the thermocouple junction, which has a much lower heat flow resistance than that of the air film, would not be expected to alter the temperature profiles or cooling rate of the seal significantly. We anticipated that the thermocouple junction would have more influence on the heating rate (as discussed above) than on the cooling rate of the seal, therefore, it was not surprising to see that the calculated (h - 3.0) and measured cooling curves coincided more closely than the calculated and measured heating curves. The calculated temperature profiles for heating shown in Figure V-7 demonstrate that the seal temperature is not uniform even at times as long as 5 sec. The vinyl portion closest to the heated jaw will melt long before the vinyl at the seal interface. Some of the melted vinyl will probably then be extruded from the seal changing the seal thickness. Whether this extrusion will be large enough to affect the rate of heating or cooling is a matter for Speculation. Table V-14. f and j Values Calculated from Heating anc Cooling Curves Heating Coolingfi Calculated ‘Measured Calculated Measured h = 2 h = 3 f 15.7 25.3 33.0 21.9 20.2 j .502 .293 1.0 1.0 .905 100 VI. CONCLUSIONS From the study of heat resistance of B. subtilis 5230 Spores heated on tin cups, vinyl, and in simulated seals, the following are concluded: 1. The dry heat resistance of the microorganisms in seals as com- pared to that on vinyl was found to be higher at 3050 F, but the same at 2800 F. These results suggest that the resistance of Spores in seals may be higher than on vinyl; however, no definite conclusion can be made. If the dry heat resistance of Spores is higher in seals, this supports the idea that con- ditions which restrict water loss may increase dry heat resis- tance. 2. The z value of Spores on vinyl subjected to dry heat appears to be higher than for Spores on tin cups subjected to dry heat; at 3150 F the resistance on vinyl was higher, while at 2800 F the resistance on tin cups was higher. 3. Dry heat conditions predominate in seals heated in wet steam. 0n the basis of the results of the calculation of temperature profiles in the flexible seals and the measurement of heating and cooling rates at the seal interface, the following conclusions are made: 1. The j of the calculated heating curve (0.502) of the seal interface is probably closer to being correct than the measured 101 j (.293). 2. The temperature response parameter, f, of the measured heating curve (25.3 sec.) was somewhat higher than that of the calculated (15.3 sec.). Because this difference was probably caused by the relatively large thermocouple junction used in measurement, the f of the calculated heating curve is probably more accurate. 3. The measured cooling curve and cooling curve calculated for h = 3.0 correSponded closely. 4. The actual thermal properties of the flexible package films, and the magnitude of the errors involved in using the assump- tions made in this study to calculate the temperature profiles should be determined. 5. A smaller thermocouple and thermocouple junction or perhaps another device should be used to measure the temperatures in the seal. 102 REFERENCES Amaha, M. and 2. J. Ordal. ' 1957. Effect of divalent cations in the Sporulation media on the thermal death rate of Bacillus coagulans var. thermoacidurans. Journal 2§_Bacteriology 74: 59 . Augustin, J. A. L. 1964. Recovery patterns of putrefactive anaerobe no. 3679 in various subculture media following moist and dry heat treatment. Unpublished Ph. D. thesis. Michigan State University. Ball, C. O. and F. C. W. Olson. 1957. Sterilization i3 Food Tech- nology. McGraw~Hill Book Company, Inc. New York. Bennett, C. 0. and J. E. Meyers. 1962. Numerical, graphical and analog methods in the analysis of heat conduction. Momentum Heat and Mass Transfer. McGraw—Hill Book Company, Inc. New York. Bigelow, W. D. 1921. The logarithmic nature of thermal death time curves. Journal 9; Infectious Disegses. 27:602. Black, S. H. and P. Gerhardt. 1961. Permeability of bacterial spores I. Characterization of glucose uptake. Journal 2; Bacteriology. 82:743. Black, S. H. and P. Gerhardt. 1962. Permeability of bacterial spores IV. Water content, uptake and distribution. Journal ,,.of Bacteriology. _83:960. Bruch, C. W., M. G. Koesterer, and M. W. Bruch. 1963. Dry heat sterilization: its deve10pment and application to components of exobiological space probes. DevelogmentsIig_lndustria1 Microbiology, 4:334. Church, B. D. and H. 0. Halvorson. 1959. Dependence of the heat resistance of bacterial spores on their dipicolinic acid content. Nature 183:124. Dixon, W. J. and F. J. Massey, Jr. 1957. Introduction.£g_Statis- tical Analysis. McGraw~Hill Book Company, Inc. New York. Dusinberre, G. M. 1949. Numerical_Analysis 2; Heat Flow. McGraw—Hill Book Company, Inc. New York. 103 El-Bisi, H. M. and Z. J. Ordal. ll956b. The effect of sporulation temperature on the termal resistance of Bacillus coagulans var. thermoacidurans. Journal‘gvaacteriology. 71:10 Esselen, W. B. and I. J. Pflug. 1956. Thermal resistance of putrefactive anaerobe no. 3679 spores in vegetables in the temperature range 2500 - 2900 F._ Food Technologyn 10:557. Esty, J. R. and K. F. Meyer. 1922. The heat resistance of spores of Bgcillus botulinus and allied anaerobes. Journal .2: Infectious Diseases. 31:650. Frank, H. A. and L. L. Campbell, Jr. 1957. The non-logarithmic destruction of spores of Bacillus coagulans. Applied Micro- . biology.,_5:243. Friedman, C. A. and B. S. Henry. 1938. Bound water content of vegetative and spore forms of bacteria. Journal g£_Bacteriology. 36:99-105. Gerhardt, P. and S. H. Black. 1961. Permeability of bacterial spores II. Molecular factors affecting solutepermeat‘ion. _Journal_g§ Bacteriology. 83:750. ' ' Halvorson, H. 0. 1962. Physiology of sporulation. The Bacteria ‘11 edited by I. C. Gunsalus and R. Y. Stavier. Academic Press, Inc. New York. p. 223. Halvorson, H. O. and N. R. Ziegler. 1933. Application of statistics to problems in bacteriology I. A means of determining bacterial pOpulation by the dilution method. Journal 2: Bacteriology. 25:101. Halvorson, H. O. and N. R. Ziegler. 1933. Application of =statistics to problems in bacteriology II. A consideration of the accuracy of dilution data by using a single dilution. Journal of_Bacteriology. 26:331. Henry, B. S. and C. A. Friedman. 1937. Water content of bacterial spores. Journal gbeacteriology. 33:323. Hodgeman, C. D. 1952. Handbook 2; Chemistry and Physics. 38 ed. Chemical Rubber Publishing Company. Cleveland. Hodgeman,.C. D. 1965. Handbook 9; Chemistry and Physics. 46 ed. Chemical Rubber Publishing Company. Cleveland. 104 Hunnell, J. W. and Z. J. Ordal. 1961. Cytological and chemical changes in heat killed and germinated bacterial spores. Spores II. edited by H. O. Halvorson. Burgess Publishing Company, Minneapolis. 101. Hutton, R. S., R. J. Himoe, and J. L. Roberts. 1951. Some physical factors that influence the survival of Brucella abortus during freeze drying. Journal,o§ Bacteriology. 61:309. Ingraham, J. L. 1962. Temperature relationship. The Bacteria IV edited by I. C. Gunsalas and R. Y. Stavier. Academic Press, Inc. New York. p. 265. Jacobs, R. A. 1963. Heat resistance of Bacillus subtilis spores ‘ in atmospheres of different water contents. Unpublished M. S. thesis. Michigan State University. Jacobs, R. A., R. C. Nicolas, and I. J. Pflug. 1965. Heat resistance of Bacillus subtilis spores in atmospheres of different water contents. Michigan Agricultural Eyperiment. Station., Quarterly Bulletin. 48:238. . Koesterer, M. G. and C. W. Bruch. 1962. Resistance of dry bacterial spores to sterilization by moist and dry heat. Bacteriological Proceedings of the American Society for Microbiology. 30. Kopelman, I. J. 1966. Transient heat transfer and thermal properties in food systems. unpublished Ph. D. Thesis. Michigan State University. Lechowich, R. V. and Z. J. Ordal, 1962. The influence of the sporulation temperature on the heat resistanCe and chemical 'composition of bacterial spores. Canadian Journal 2£_Micro- biology. 8:289. Lewis, J. C. 1956. The estimation of decimal reduction times. ,Applied Microbiology. 4:211. - Lewis, J. C., H. K. Burr, and N. S. Snell. 1960. Water permeability of bacterial spores and the concept of a contractile vortex. Science. 132:544. Licciardello, J. J. and J. T. R. Nickerson. 1963. Some observa- tions on bacterial death time curves. Applied Microbiology, 11:476. 105 Murrell, W. G. 1961. Discussion. Spores II. edited by H. 0. Halvorson._ Burgess Publishing Company, Minneapolis. Murrell, W. G. .1964. Heat resistance of micro-organisms. Journal_g§_Pharmacy. Murrell, W. G. and W. J. Scott. 1957. Heat resistance of bacterial spores at various water activities. Nature. 179:481. Perry, J. H. 1963. 'Chemical Engineers Handbook. 4 ed. McGraw-Hill Book.Company, Inc. New York. Peters, M. S. 1954. Elementary Chemical Engineering. McGraw- Hill Book Company, Inc. New York. Pflug, I. J. 1960. Thermal resistance of micro-organisms to dry heat: design of apparatus, operational problems and preliminary results- Food Technology. 14:483. Pflug, I. J. and J. L. Augustin. 1963. Unpublished data. Pflug, I. J., J. H. Bock, and P. E. Long. 1963. Sterilization of food in flexible packages. Food Technology. 17:87. Pflug, I. J. and W. B. Esselen.' 1953. Development and application of an apparatus for study of thermal resistance of bacterial spores and thiamine and temperatures above 250°F. Food . ‘Technology._ 7:237. Pflug, I. J. and W._B- Esselen. 1955. Heat transfer in open metal thermoresistometer_cups. JFood Research. 20:237. Pflug, I. J. and W. B- Esselen. 1954. Observations on the thermal resistance of putrefactive anaerobe no. 3679 spores in the temperature range of 250°-300°F. Food Research. 19:92. Pflug, I. J. and K, I. Fox. 1966. Unpublished data. Pflug, I. J. and C. F. Schmidt. 1962. Unpublished report. Pheil, C. G., I. J. Pflug, R. C. Nicholas, and J. A. L. Augustin. 1966. Unpublished data. Powell, J. F. 1953. Isolation of dipicolinic acid (pyridine-2:6-- dicarboxylic acid) from spores of Bacillus megatherium. Biochemical Journal. 54:210. 1C6 Rahn, O. 1945b. Injury and death of bacteria by chemical agents. BiodynamicalMonograph‘Ng._§. Biodynamica. Normandy, Missouri. Rahn, O. 1945a. Physical methods of sterilization of micro— organisms. Bacteriological Reviews. 9:1. Rahn, 0. and W. R. Shroeder. 1941. Inactivation of enzymes as the cause of death of bacteria. Biodynamics. 3:199. Ridgeway, J. and S. Kavesh. 1965. Heat sealing of polyaelifins. "Modern Packaging. 38:139. * . Ross, K. F. A. and E. Billings. 1957. The water and solid content of living bacterial spores and vegetative cells as indicated by refractive index measurements. Journal gpreneral Micro- biology.“ 16:418. Salton, M; R. J. and B. Marshall. 1959. The composition of the spore wall and the wall of vegetative cells of Bacillus subtilis. Journa1_of_GeneralkMicrobiology, 21:415 Schmidt, C. F. 1957. Antiseptics, Disinfectants, Fun icides, and Chemical and Physical Sterilization by C. F. Reddish. 2 ed. Lea and Febiger. Philadelphia. 831. Schneider, P. S. 1955. Conduction Heat Transfer. Addison- Wesley Publishing Company, Inc. Reading, Massachusetts. Scott, W. J. 1958. The effect of residual water on the survival of dried bacteria during storage. Journal‘g§;Genera1 Micro- biology. 19:624. Secrist, J. L. and C. R. Stumbo. 1958. Some factors influencing ‘thermal resistance values obtained by the thermoresistometer method. Food Research. 23:51. Silverman, G. 1966. Unpublished data. Sisler, W. A. 1961. A study of the thermal resistance of bacterial spores to moist and dry heat by use of the thermoresistometer. Unpublished M. S. Thesis. Michigan State University. Stumbo, C. R. 1965. Thermobacteriology ig_Food Processing. Academic Press, Inc. New York. Tischer, R. G. and H. Hurwicz. 1954. Thermal characteristics of bacterial populations. Food Research. 19:80. Townsend, C. T., J. R. Esty, and F. C. Baset. 1938. Heat re- sistance studies on spores of putrefactive anaerobes in rela- tion to determination of safe process for canned foods. Food Research. 3:323. Virtannen, A. I. 1934. Enzymes of bacteria and bacterial metabo- lism. _Journal of Bacteriology. 28:447. Waldman, D. G. and H. 0. Halvorson. 1954.. Studies on the rela- tionship between equilibrium vapor pressure and moisture content of bacterial endospores. .Applied Microbiology. 2:333. Walker, H. W., J. R. Matches, and J. C. Ayers. 1961. Chemical composition and heat resistance of some aerobic bacterial spores. _Journal‘g§_Bacteriology. 83:960. Williams, 0. B. 1929. The resistance of bacterial spores. Journal .Qi Infectious Diseases. 44:421. MICHIGAN STATE UNIVERSITY 1 3 1293 0307] 6 LIBRARIES 95