THE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING ADSORPTION OF NITROGEN-METHANE ON LINDE MOLECULAR SIEVE TYPE 5A Peter B b -Lederman A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in The University of Michigan 1961 March, 1961 IP-502

Doctoral Committeeo Professor Brymer Williams, Chairman Associate Professor Lee 0O Case Professor Donald Lo Katz Professor Guiseppe Parravano Associate Professor Mo Rasin Tek ii

ACKNOWLEDGETMENT The author wishes to take this opportunity to thank the many people who have assisted in the completion of this thesis~ Among these are Dr o TO Lo Thomas of the Linde Company, who kindly supplied the Molecular Sieve and whose advice was most helpfulMro RF. Lo Porter of Autoclave Engineers for his as well as his associates' valuable time in aiding to solve some valve problems; Mro Foglesong of The Babcock and Wilcox Company for his help in obtaining material for the adsorption vessel; The Kolene Corporation for Tufftriding some vital parts and the Dow Chemical Company for its gift of the necessary Styrofoam insulati on.. Special thanks are due to Messerso Cleatis Bolen, John Wurster, William Hines, and Peter Severn, who assisted in building the equipment and were always ready with helpful. suggestions. MrO Frank Drogoz also deserves many thanks as do his assistants, Richard Naragon and David Williams for the help in the mass spectrometer analysis~ Mr~ Jude Sommerfeld assisted with, the surface area determinationso The staff of the University Computing Center and the use of the IBM 704 Computer materially lightened thle task of reducing the experimental data to a useful fortmo To his committee the author desires to extend a special note of thanks for their patience, ever open doors and encouragement~ The Industry Program of the College of Engineering was of great assistance in preparing the final manuscript and in arranging for its presentation in this form. The encouragement given me by my wife, as well as her aid in preparing the final draft was most helpful~ Finail.y, to mny parents, for making this task possiblee and for their continuing encouragement and guidance, I dedicate this work. iii

TABLE OF CONTENTS Page ACKNOWTiDGMET.....................o.o ii LIST OF TABLES.............................. vi LIST OF FIGURES..., viii LIST OF APPENDICES..................4.......... x NOMENCILATURE9....................................,..,....0 xi ABSTRACT..... x.........,........................... xiii INTRODUCTION............................ 1 DISCUSSION OF THEORY.........*................... 2 Pure Component Adsorption Isotherms.................... 2 Adsorption from Mixtures............... 11 EQUIPMENT..........*........... *..... 15 Adsorption Apparatuses..............................9 15 The Present Apparatus............................. 17 The Adsorption Cell................... 22 The Pressure Transducer.............................. 24 Material Selection....................... 28 Valve Operation at Low Temperature,..........,...... 28 Temperature Measurement and Control,.................. 30 Safety Considerations,........................ 31 Leakage in the System............................ 32 Calibration of Equipment e....-....*.......j... 34 EXPERIMENTAL METHOD............................... 37 Adsorbent,o............................. 37 Gases............,..... 38 Experimental Method................. 40 Sampling..................................... 42 Method of Calculating Adsorption Equilibrium............ 44 Errors in Measurement............. 47 EXPERIMENTAL RESULTS........................ 52 Pure Component Adsorption................ 52 Binary Adsorption,.......,.*...-........... 57 iv

TABLE OF CONTENTS CONT'D Page DISCUSSION OF RESULTS,..........~ 63 Pure Component Adsorption Calculations,............. 63 Correlations for Total Adsorption from a Mixture....... 70 Adsorption of Components from a Mixture,............... 72 Computing Adsorption Loading from the Correlations,...,, 75 An Adsorption Model......,,........,..... 80 CONCLUSIONS AND RECOMMENDATIONS.........,. 4,,.......... 81 BIBLIOGRAPHY................. 130

LIST OF TABLES Table Page I Norwood Pressure Transducer Specifications........ 26 II Calibrated Volumes................,.,...e....., 35 III Analyses of Pure Feed Gases..................... 39 IV Paired Samples from Adsorption Cell............... 43 V Replicate Analyses of a Single Sample............. 43 VI Replicate Samples from a Uniform Mixture -.......... 43 VII Accuracy of Measuring Devices.................... 49 VIII Maximum Error Propagated in Pure Component Equilibrium Data,.,..........., 50 IX Maximum Error Propagated in Binary Equilibrium Data....*....*a.a................................ 50 X Pure Methane Adsorption..................... 54 XI Pure Nitrogen Adsorption........................... 56 XII Mixed Adsorption at 295~K Average Temperature.,..o 58 XIII Mixed Adsorption at 195 K Average Temperature..,, 59 XIV Mixed Adsorption at 1750K Average Temperature,.-.. 60 XV Mixed Adsorption at 1235K Average Temperature..... 60 XVI Confidence Limits on Equilibrium Values.,,........ 61 XVII Actual and Predicted Binary Adsorption...,....,.. 79 XVIII Explanation of Experimental Data.................. 86 XIX Experimental Data................ 87 XX Thermocouple Calibration Data.............,..,0 99 XXI Calibration Data for Pressure Gauges.............. 100 XXII Calibration Data for 2000 Psi Pressure Transducer, 104 XXIII Calibration Data for 2000 Psi Pressure Transducer, 105 vi

LIST OF TABLES CONT'D Table Page XXIV Calibration Data for 500 Psi Pressure Transducer.. 108 XXV Calculation of "BET" Surface Area Determination 1. 112 XXVI Calculation of "BET" Surface Area Determination 2. 114 XXVII Sample Print Out From Computer Calculation,.,...., 127 XXVIII Adsorption of Helium........................ 128 vii

LIST OF FIGURES Figure Page 1 Flow Diagram of Adsorption Apparatus....,........ 18 2 Top View of Closed Bath,.,......................... 19 3 Open Constant Temperature Bath...........,...... 20 4 Instrument and Control Section................... 21 5 Adsorption Cell in Cross Section.................. 23 6 Schematic Diagram of Norwood Pressure Transducer............^...^.......,........, 25 7 Simplified Schematic of Electrical Measuring Circuit.a...........a.......................... 27 8 Cross Section of Main Block Valve...,...~ 29 9 Methane Adsorption Isotherms on 5A Molecular Sieve..................... a.a e *...... 53 10 Nitrogen Adsorption Isotherms on 5A Molecular Sieve...,,CI..~............v. *.*.0.o*... 55 11 Total Mixed Adsorption as a Function of Pressure and Temperature................... 62 12 Pure Component Adsorption on 5A Molecular Sieve... 64 13 Constants for the Freundlich Equation for Nitrogen and Methane and their Mixtures........... 65 14 Langmuir Constants for Pure Component Adsorption.. 66 15 Volume of Nitrogen Adsorbed as a Function of the Adsorption Driving Force....................... 68 16 Volume of Methane Adsorbed as a Function of the Adsorption Driving Force............. 69 17 Volume of Nitrogen-Methane Mixtures Adsorbed as a Function of the Adsorption Driving Force......., 71 18 Langmuir Constants for Adsorption from NitrogenMethane Mixtures............................ 74 * * -

LIST OF FIGURES CONT'D Figure Page 19 Total Adsorption Correlation on Molecular Sieve 5A.,a. o.0.. **.. a*.a......*.............. 76 20 Relative Volatility as a Function of Pressure..... 77 21 Calibration for 300 Psi Pressure Gauge........... 101 22 Calibration for 800 Psi Pressure Gauge,........... 102 23 Calibration for 2000 Psi Pressure Gauge........... 103 24 Calibration for 2000 Psi Pressure Transducer...... 106 25 Calibration for 2000 Psi Pressure Transducer...... 107 26 Calibration for 500 Psi Pressure Transducer...... 109 27 Calibration for 500 Psi Pressure Transducer....... 110 28 "BET" Plot for Surface Area Determination 1,...., 113 29 "BET" Plot for Surface Area Determination 2....... 115 30 Computer Flow Diagram for Equilibriun Adsorption Loading Calculation,,..,...... o...... 125 31 Computer Flow Diagram for B-W-R Equation Solution for Density and Fgacity... 126 ix

LIST OF APPENDICES Appendix Page A Experimental Data..................... 84 B Calibrationso..................... 98 C Calculation of Surface Area.................... 111 D Pressure Vessel Design Calculations.,..,.... 116 E Sample Calculation of Adsorption Equilibrium...... 119 F Adsorption of Heliumn....,................ 128

NOMENCLATURE a constant in single component Langmuir equation; milligram moles/gm. atm. a' constant in multicomponent Langmuir equation; milligram moles/gm. atm, b constant in single component Langmuir equation; l/atm. b' constant in multicomponent Langmuir equation; 1/atm. b" constant in multicomponent Langmuir equation; l/atm. bv van der Waals' constant; cc./gm. mole E Schay's mixing coefficient f fugacity k constant in Freundlich equation; milligram moles/gmo N amount adsorbed; milligram moles/gm. N' amount adsorbed as pure component under the same pressure as the component partial pressure in the gas phase mixture. n constant in Freundlich equation P pressure; atm. R universal gas constant T temperature ~Kelvin v volume adsorbed x mole fraction adsorbed phase y mole fraction gas phase Y constant C adsorption potential; calories/gm. mole p density; gm. moles/cc. xi

Subscripts cr critical F final FS free space (adsorption cell) I initial i ith component m maximum s at saturation conditions xii

ABSTRACT The adsorption of nitrogen and methane and their mixtures on Linde Type 5A Molecular Sieve has been studied in a batch system from -140~C to room temperature and at pressures up to 85 atmospheres0 The amount adsorbed at equilibrium is reported for the pure components. Data indicating the equilibrium gas and adsorbate compositions, as well as the total amount adsorbed, are presented for mixtures of nitrogen and methane. Constants for the Langmuir equation, over the range of pressure and temperature studied, have been obtained for pure component adsorption. The Freundlich equation constants are also presented for the adsorption of pure nitrogen and methane and for total adsorption of mixtures as a function of temperature. The equations with the appropriate constants represent the capacity of the Molecular Sieve for pure components within five per cent and for mixtures within seven per cent of the total amount adsorbed. Individual component adsorption from mixtures may be expressed as a function of the pressure of the components in the gas phase in equilibrium with the adsorbate. Constants for the system studied are presented as a function of temperature between 100'Kelvin and 300'Kelvin, Limited data at room temperature for some other adsorbents, such as silica gel and some activated carbons, indicate that Molecular Sieves have a higher capacity for methane and nitrogen. A study made on another activated charcoal which had a slightly greater capacity shows that available surface area is a critical factor in determining capacity.

INTRODUCTION The present study was undertaken to analyze the adsorption system nitrogen - methane - Linde Molecular Sieve Type 5A*. A method of predicting the adsorbate loading from both pure and mixed gas streams was to be developed. If possible, a model for the adsorption mechanism was to be postulated, The nitrogen - methane system is of ever growing interest because these two gases appear together in many industrial applications. In addition, the present program to conserve helium will require removal of nitrogen and methane from that gas in varying quantities, for subsequent recovery The study, as undertaken, included a wide range of temperatures from room temperature to below the critical of the two gases as well as the pressure range of industrial interest, Temperatures varied from 1230 Kelvin to 2959Kelvin and pressures ranged from essentially one atmosphere to as high as 50 atmospheres. The range of compositions at the higher temperatures varied from pure methane to pure nitrogen, while at the lower temperatures, only one or two mixtures were evaluated in addition to the pure components. For purpose of completeness, a discussion of the various adsorption theores has been included, followed by a discussion of the experimental equipment. The experimental method and results are described and evaluated together with a series of conclusions and recommendations, * Registered Trade Mark -1

DISCUSSION OF THEORY Pure Component Adsorption Isotherms Several basic approaches to physical adsorption equilibrium have been presented since the phenomenon was first observed by Scheele in 1773.(13) These may roughly be distinguished as the kinetic and the thermodynamic approaches. This distinction is somewhat artifical as the isotherms based on the kinetic approach, first advanced by Langmuir(33,934) can be readily developed from thermodynamic or statistical principles. In the present context an isotherm defines the amount adsorbed as a function of pressure at constant temperature. Langmnuir first derived the isotherm which bears his name: N Nmb (1) 1 + bP or N = aP (la) 1 + bP from kinetic principles(34) This derivation requires that arbitrary forward and reverse rates be balanced. Fowler and Guggenheim(20) have. derived the Langmuir isotherm on statistical grounds, considering that the adsorbed molecules do not interact with each other and had no trans'lational motion and were thus confined to localized sites, or points of active adsorption. If one examines the Langmuir isotherm, one can readily see that at high pressures N approaches Im which corresponds to the amount of adsorbate in a complete monolayer. While the Langmuir isotherm fitted a great deal of experimental data, the three basic assumptions:

1, Uniform sites with no interaction between adsorbate molecules, 2. No translational motion on the surface, 3. The maximum adsorption corresponds to a complete monolayer~ placed severe restrictions on the theory. Brunauer, Emmett and Teller(15) in their well-known "B E T" equation made the first significant advance -when they considered multilayers, The "B E T" equation takes several forms, of which N c P/P 2.) Nm (1-P/Ps )[1 + (c-1)P/Ps] where the number of layers, or stacks of molecules above the solid surface are finite, is the most widely used form. Several modifications of the original "B E T" equation were proposed by Brunauer et al (15), where the number of layers are limited, and Brunauer and Demming(l4) for the case of adsorption on capillary walls. Joyner(31) has discussed in detail a method of determining'the number of adsorption layers in the complete "B E TE" equation by graphical methods, The original derivation of the "B E T" isotherm suffered from tne same limitation that Langmuir's derivation did, in that it was based on kinetic equilibria. Hill(24) and others have derived the equation from statistical principles, based on a model allowing vertical but no horizontal interaction between molecules. The "B E T" equation still suffers from the same basic deficiencies as the Langmuir isotherm: no interaction of adsorbate molecules and no lateral motion of the adsorbate molecules in any given layer0 More recently, a multi-molecular adsorption model, allowing for lateral interaction, has been suggested by Lee(35)o

-4Unfortunately, no experimental work has been done to confirm the postu. lated isotherm, Peierls (46) and Wang(61) derived the Langmuir equation from statistical principles for adsorbed atoms, exhibiting attractive and repulsive interaction respectively. Schay(56) has derived Langmuir's isotherm on statistical grounds in the case of a mobile monolayer, where the adsorbed molecules exhibit a covolume in the adsorbed state, thus removing the last serious theoretical restriction, The Freundlich isotherm(21) N = KPl/n (3) has been derived from the Langmuir isotherm by Zeldowitsh(63)0 By assuming that the adsorption surface is heterogeneous, the Langmuir equation may be written as JN z= Z aiP (4) i P + i where ci depends mainly on the heat of adsorption on the adsorption spaces and ai is a function of the number of these spaces, If "c" is said to vary continuously over the surface and the distribution function of "a" assumed the form a(c) Acl/n-l (5) then Zeldowitsh arrives at the Freundlich isotherm. The isotherm may also be derived for multi-molecular adsorption, as was shown by Baly(3), who assumed that each adsorbed layer obeys a Langm uir isotherm with different constants, and that each succeeding layer has a slightly lower heat of adsorption than the one below. These assumptions lead to the Freundlich isotherm for moderate pressures.

-5More recently, Sips(57,58) has shown that a generalized Freundlich isotherm: N = (P/(P + a))c O < c < 1 (6) may be obtained by assuming a localized adsorption model with no interaction. Equation (6) reduces to the Freundlich equation for small "P" and when "c" is equal to one the Langmuir equation results, Unfortunately, this theory does not appear to fit the experimental data. Sips' assumptions of localized sites and no molecular interaction may be necessary from a mathematical point of view, but present a simplified picture of the physical process, thus leaving much to be desired. In an excellent review of the integral inversion approach used by Sips, Honig(28) indicates that the isotherm is very insensitive to changes in the chosen distribution function. In addition, all attempts to obtain an isotherm by this method require a model with no molecular interaction. He further states that many of the most useful isotherms do not satisfy all of the mathematical restrictions found to be necessary in applying this method of analysis. The isotherm suggested by Sips is, as he states, the simplest satisfying all the necessary restrictions. A real physical restriction, that at very low pressures, the amount adsorbed is proportional to the pressure is not satisfied except in the case of "c" equal to one. It would seem, then, that Sips' model is too simple to explain the total adsorption isotherm, Real improvements in obtaining valid adsorption isotherms must come from advanced or improved models of the adsorption phenomena,

Redlich and Peterson(52) recently proposed an isotherm N = aP/(l + cpV/) (7) where O < y < 1 which fits much of the experimental data well. It further reduces to the phenomenalogically sound) limiting Langmuir equation N = ap (8) at low pressure. At moderate pressure the equation reduces to the Freundlich isotherm N = ai-' (9) c which appears to be sound in this range, The equation,however, breaks down as the pressure becomes very high. At present the kinetic-statistical approach, while indicating correct avenues, has, due to the limitations imposed by the simplified models, not fully explained the adsorption process, About the same time that Langmuir first advanced his kinetic adsorption model, Polanyi suggested a model based essentially on thermodynamic principles. As first proposed by Polanyi( 4748' 950), the theory assumed that long range forces controlled the adsorption. This basic assumption was later modified by Polanyi and Goldman(22) so that the theory would conform with the concept of molecular forces. Basically, the Potential Theory of Adsorption assumes that work must be done to bring the molecules from the gas phase into the adsorbed phase, and that this work, e, is work of compression: = VdP (10) Pg

-7Then, if any layer "i" is considered, one can obtain the work required, Gin which is called its potential, and the equipotential surfaces will vary from a maximum of co, the potential of the adsorbent-adsorbate interface, to zero, the potential at the gas-adsorbate interface. Each equipotential surface will enclose a given volume, vi, around the adsorbent. Thus it was shown that the volume adsorbed and the work expended in the adsorption process are related, Polanyi denoted this relationship Ci = p(vi) (11) as the characteristic curve dependent only upon the adsorbent-adsorbate system. Although Polanyi first applied the theory, Berenyi(7'8) improved the method considerably, Both found that the computation of the characteristic curve, although fundamentally the same over the whole range of interest, required slight alterations near the critical point and above the critical temperature. Well below the critical temperature, the characteristic curve may be computed, if one assumes an ideal gas model; 1) by integrating the compressive work integral between the limits of the system and saturation pressures respectively; thus: Ps i Vda (10) = RT ln(PS/P) (lOa) and 2) if one assumes an incompressible liquid, the adsorbed volume

-8becomes: vi Ni/pS (12) where p is taken as the density of the saturated liquid at the adsorption temperature. Berenyi suggested corrective terms for both the potential integral and the density near or above the critical temperature, so that near the critical: e = RT Ln (PS/P) + p (13) v = N/Ps (12) and above the critical, C RT 1n A14T (14) Pbv where bv is van der Waals' constant, The model as originally suggested has been used effectively; more recently modifications have been introduced, in particular for the conditions near or above the critical to remove the trial and error part of the solution for the correction terms originally introduced by Berenyi. Nikolaev and Dubinin(43) have proposed that above the critical the potential be computed by: = RT in r2( /P) (15) where T = T/Tcr (16) and the adsorbate volume: v = Nbv, They obtain this result by showing at the critical temperature the characteristic curve obtained using the normal method of evaluating E;

~= 1r ln (Por/P) (17) and using Berenyi's modification above the critical C = RTCr ln C (PCr/P) (18) will coincide when = (Pcrbv/Tcr) T (18a) In addition, Nikolaev and Dubinin(43'17) suggests that above the normal boiling point a density: P = P ((PSP/tr-ts~))(t-ts) (19) where pv is equal to van der Waals' constant, would better represent the density of the adsorbate, This density then will take into account the existence of the adsorbate in a compressed state due to the attractive forces of the adsorbento The deviation from ideality of the gas in equilibrium with the adsorbate should influence the adsorption potential. Although Polanyi in his original work suggests this, he did not develop the idea. Lewis and co-workers(27), in correlating some adsorption isotherms of hydrocarbons at moderate pressures, found that the deviations were sufficient to warrant introducing corrections for the deviations from ideality. They suggest the use of fugacity, the most reasonable approach, instead of pressure. Thus, e = zRT ln (f5/f) (20) where, from their calculations, fs is the fugacity at the saturation pressure, obtained by using the generalized f/p plots. Above the critical

-10f was set equal to the fugacity at the pseudo-saturation pressure as determined by extrapolating the usual vapor pressure curve: ln P = (-A/T)+ B (21) where "A" and "B" are constants, Lewis and his co-workers suggested that the density of the adsorbate was equal to the saturated liquid density at the adsorption pressure, It would. seem that their analysis is quite satisfactory except for their assumption regarding density, as liquid density is principally a function of the system temperature. A combination of the work of Dubinin and co-workers with that of Lewis should give a better representation of the adsorption potential. Thus, the use of fugacities, to represent the driving force, evaluated at the system pressure and saturation or the critical condition, together with the density evaluated by Dubinin's method should give a good characteristic curve; that is a single function E= P(v) (22) independent of temperature. Although the Potential Theory does not clearly explain the adsorption phenomena, it does present a satisfactory explanation for the gross process, The process above the usual critical point is identical in nature to that below the critical point in that a dense phase forms. This phase, while not a liquid as normally envisioned, does possess many of the same properties. The theory does have the very fine property of generalizing the system from one isotherm, Lack of knowledge regarding the dense or adsorbed phase limits the general applicability of the

11theory at present, Hopefully, added knowledge in this area will make this theory more useful, The deterministic approach, illustrated by Hill, Schay and many others may lead to a number of models and isotherm equations. At present, no totally satisfactory model has been suggested to explain in particular the adsorption phenomena at moderate pressures. Most probably, this is the result of the many simplifications necessary to enable one to treat the phenomena from the statistical viewpoint and still obtain mathematical models which may be solved analytically. It may be that by the application of numerical methods at some future date more complex models, which allow for all types of molecular interactions, may be proposed and a more definitive test for determining the applicability of specific models at moderate pressures may be found, Adsorption From Mixtures Much of the work done to date on adsorption from gaseous mixtures falls into the category of dehumidification. Some work has been done on removal of trace impurities from gaseous streams. In this area two works, that of Johnson(9) and Hiza(28) are of some interest as they deal with the removal of trace amounts of nitrogen and methane by silicon gel adsorption from a high purity hydrogen stream, A good review of the eatly work in mixed adsorption is given by Brunauer.(13) Very little work has been done in this area, The first work of any significance is that of Markham and Benton(40), which extend Langmuir's isotherm to mixtures, Assuming the same model as Langmuir, the component isotherms resulting from the solution of the

-12rate equations (two adsorption and two desorption) are: N - Nm bjPl. (23a) 1 1 + b2P2 + blP1 N = N2m b2P2 (23b) 1 + b2P2 + blP1 Equations (23a) and (23b) indicate a mutual decrease in the adsorption of both components at a given partial pressure. This has been shown to be true in general; the equation does not take into account the possibility of interaction between the two components in the adsorbed state. The relationship appears to hold well at low concentrations, but breaks down at the higher concentrations reinforcing the problems caused by interaction. Schay et al57 have developed essentially the same equation as Markham and Benton: N1 N 1 aP/E E) (24) adsorbed state and is defined by Schay as: In Ek = [ n% P - PJkj/ Znj%[ IA. ln(l - >A) + 1] P = two dimensional residual molecular volume A = adsorbent area n = number of molecules adsorbed at the asymptotic value of the isotherms'P1.r a binary component, if E1 - E2 - l,the result is identical to that of Markham and Benton. In addition, Schay and co-workers have shown that the

empirical relationship, for a binary mixture, suggested by Williams(62) and observed by Lewis and Gilliland(36) a + N2 1 (25) Nt N' 1 2 can be shown to be: 1 (P1 + P Pil1 N1 + N2 1 +.EE2 (26) N' N' a1 a2 N1 2 1 +e P1 E (P E P1) for their model. Thus Equation (25) is strictly correct for E1 = E2 = 1 which indicates essentially no interaction of the species in the adsorbed state. In practice, Schay has found that E1 and E2 each differ slightly from unity, one being greater the other less and Lewis' observation is still valid in practice. Redlich and Peterson(52) have suggested that their equation for pure components (Equation 8) is also valid for mixtures if it is modified in a manner analogous to the Markham modification of Langmuir's isotherm for multi-component adsorption. This results in: N1 a 1 (27) 1+ c pIY + c P Y 11 22 which does not account for any interactions if ai and ci are identical values for the mixed isotherm as for the pure component. A modification for interaction of the molecular species, similar to that suggested by Schay, can be introduced so that the form of the equation would be:

aiPi/Ei = +.,.....j... (28) Very little work has been carried out in the area of co-adsorption of mixtures so that the phenomena of competing adsorption has not been studied except as an extension of pure component adsorption, At the present time it would appear that very little can or should be said in this area until the phenomenon of pure component physical adsorption has been more clearly defined. A detailed sub-microscopic study of the solid adsorbent surface and the surface forces must be undertaken before a clear picture of the adsorption phenomena can be developed. In addition the molecular state of the adsorbate, both for pure and multicomponent systems, requires examinationo Until the adsorption phenomena is clearly established, a semiempirical approach such as that discussed above, must suffice for engineering purposes,

EQUIPMENT Adsorption Apparatuses There are basically three methods of measuring adsorption equilibria, the continuous once-through flow system, the recirculating flow system and the closed system. Each system or method has inherent advantages and disadvantages, The continuous flow system requires careful control of the flow rates and continuous stream analysis. Both of the requirements can easily present experimental difficulties as they must often be determined and controlled within very small tolerance limits. A modification, used by Linde(51), eliminates these difficulties to some extent, In this modified flow system, adsorbate is passed over the adsorbent bed until flow rates at the entrance and exit of the cell are equal. The cell effluent is then sampled and the cell taken off the line. The adsorbate is desorbed; the desorbed material is collected and its mass and composition determined. Knowing the gas phase composition and volume, the adsorbate volume and composition may be determined by material balance, Both the once-through and modified once-through flow systems require large amounts of adsorbate gas and a great deal of coolant at low temperature. They do, however, enable the experimenter to obtain kinetic data, as well as equilibrium data. The recirculating flow system, such as the one described by Lewis and co-workers(37), is quite satisfactory for equilibrium data but can present problems as there is a greater possibility of leaks, As it recirculates the gas, a uniform system will be achieved and by appropriate -15

-16pressure measurements and knowing the free space volume one can determine the amount adsorbed, As it has the additional feature of a large gas space not in contact with the adsorbent, it is easier to control the gas phase equilibrium composition and even the pressure by adding small increments of gas as the adsorption process proceeds. This system requires mechanical agitation, or rocking, which can present problems when considered in respect to a system that must be maintained at low temperatures and high pressures. In addition, the entire system must be leak-free which in itself can be a difficult problem under severe operating requirements and constant movement. This system is probably best adapted to operating pressures below ten atmospheres and temperatures in the ambient region, By far, in many respects, the simplest system is the closed system in which a measured amount of sample is admitted to the equilibrium vessel and in the case of the mixture, the gas phase is sampled when equilibrium is reached, Then, knowing the amount of each constituent in the gas phase, the equilibrium adsorbate can be determined by material balance, Several variations of this type of apparatus have been used. At low pressures and particularly for pure components, a gravimetric adsorption apparatus, as typified by the sorption balance of McBain and Bakr(38), is very useful in that the amount adsorbed can be determined directly. Sawyer, Josefowitz and Otbmer(29) have made modifications to the spring and general assembly which make the apparatus more durable, At high pressure, modifications of this balance have been developed by McBain and Britton(39) and Morris and Mass(42). Volumetric apparatuses for low pressure work are used frequently. The simplest apparatus of this type is probably that of Pease(44) and is

-17m similar to ones used by Homfray, Titoff and others. Coolidge(l6) developed an apparatus of this type which had no stopworks. A high pressure volumetric adsorption apparatus which was suitable for work over a wide temperature range was built by Antropff and co-workers,(l) This apparatus is excellent for pure component adsorption studies at high pressure and was used to determine the adsorption isotherms of Nitrogen and Argon. The equipment included a mercury pressure transmitting device to keep the gas volume constant, At low temperatures a temperature gradient in the gas in the pressure transmitter could present difficulties. In addition, there was some gas volume which never came in contact with the adsorbent; this presents no difficulties when working with pure components but makes it impossible to use this apparatus for multicomponent adsorption studies. The Present Apparatus A schematic diagram of the system used in the present study can be found in diagram I. There are two large feed tanks (A), size 1A cylinders, which contain pure component. These are connected to the main feed reservoir (B), which is a high pressure vessel fabricated of 2 1/2-in. extra heavy-duty pipe and has a volume of about 940 cc. The feed reservoir is connected by a quarter inch high pressure line to the feed pressure gauge (C), and the distribution block. From the distribution block, lines lead to the sampling connection (D) and the vacuum and expansion system (E). A third line leads through the sampling connection (K), through the low temperature cell block valve (L), to the adsorption cell (F) which is discussed in greater detail below, The vacuum and

METERING VALVE (H -2) SAMPLE CONNECTION FEED PRESSU (G3) CELL BLOCK VALVE(L) GGE 1CONSTANT TEMPERATURE BATH (G-6) (G-5) SAMPLE HEATER7, (F) ~(G ~6)SAMPLE ":?'....TO BRIDGE FIEED SAMPLE SECTION (K) AMPLIFIER (G-I) RESERVOIR (G-2) i1/16". D.LINE "B (G-4;H. ADSORPTION CELL PRESSURE GAUGE 300 Psi (A) METERING VALVE (E ) ( H-lI') 0 O.I.I.BAROMETER MAIN GAS CYLINDER TO VACUUM PUMP TO MC LEOD GAUGE.... _-TO VACUUM I/4" S.S TUBING PUMP.. —. GLASS TUBING ---------— ELECTRICAL CONNECTIONS MERCURY EXPANSION BURET VOLUME Figure 1. Flow Diagram of Adsorption Apparatus.

- H~~~~~~~~~~~~~~~~~~~~~~ Figur'e 2. Thp View of C:oseh path (showing protective steel shield). |~~~~~~~~~~~~~ Fig Fiur 2 Cp tiw f ~lse Bt (hoin potctvestelshel)

i-iiii~~~~~i~~~iii —~R:::~::?::;~0 Figure 5. Open Constant Tenperature Bath.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:::

E-APL: jili | i 0 f)l t~:~::i:~:;:i:_:::::i::'l:::i:ii i a) ~ Z cop_'>~~~~~~~~~~~~~~r

-22expansion systems are made of glass and are protected against high pressure surges by standard Autoclave 10,000 psi. valves (G3, G4) and Hooke metering valves (Hi, H2) in series. A low pressure gauge (I) is placed in between the block and metering values on the vacuum line as an additional precaution. The entire system is at room temperature except for the adsorption cell and cell sampling line (K) which are in the constant temperature bath. The Adsorption Cell The major modifications from other similar system were made to the adsorption cell. The cell, shown in detail in Figure 3, was made of 316 stainless steel and designed according to the ASME-API Code for Unfired Pressure VeSsels.(2) Detailed design calculations may be found in Appendix D. The two ends were machined from bar stock and the center cylinder is a piece of 1 1/2-in, schedule 180 seamless tubing donated by Babcock and Wilcox. The ends were joined to the cylinder using 308 stainless steel welding rod in order to prevent the formation of delta ferrite, The vessel was then x-rayed to insure the soundness of the welds. In order to release any stresses produced during welding or machining, the entire vessel was heat treated as 750-8000F for two hours and then furnace cooled. The quarter inch tubing connects to the main block valve (L) and is fitted with a fine stainless steel screen to prevent particulate matter from fouling the valve. The thermocouple well is standard 1/8 inch high pressure tubing. The 18 mm. threaded aperture serves both for filling the vessel with adsorbent and as the connector for a

12 FACED FOR THERMOCOUPLE 5 GASKET SEAT WELL __ _ _ ['"__ - _ i /WELD SCREEN WELDIID I.D DIAMETER TIP ro FEED LINE -- I-"+ 2 41" *|' 6 4 ------- I — S TAP FOR N ORWOOD PRESSURE TRANSDUCER MATERIAL - 316 S. S. Figure 5. Adsorption Cell in Cross Section, (F).

pressure transducexr-used for determining the pressure in the cell. This end is faced for a gasket and a teflon gasket is used* Thirty foot-pounds torque is sufficient to seal the gasket and insure a tight system. The use of a-:.pressure transducer elminates the need for any long connections where gas not in contact with the adsaorent can collect, In order to further reduce any "dead" space the quarter inch outside diameter connecting line is filled with clean copper wire. The cell sampling connection (K) is a length of 1/8 inch outside diameter tubing, also filled with copper wire so that the sample size can be kept to a minimum, Thus, the main feature of this system is the low "dead" space, assuring good contact between the entire gas sample and the adsorbent. The Pressure Transducer Pressure transducers are basically strain gauge devices, Several manufacturers make a variety of these devices and the Norwood Controls Transducer Model 101, of which Figure fj is a schematic diagram, was picked because of its ease of operation, good response and the wide choice of pressure ranges available. The pressure signal is received by a diaphragm which compresses a strain tube which supports the center of the diaphragm, Any changes in the strain tube are immediately detected by two strain gauges, one wound circumferentially and the other longitudinally. The gauges stretch and shorten respectively with an increase in pressure,

-25 - FLUSH CATENARY DIAPHRAGM CIRCUMFERENTIAL WINDING STRAIN TUBE LONGITUDINAL 2 1 3 WINDING 2 Figure 6. Schematic Diagram of Norwood Pressure Transducer. Courtesy of Norwood Unit - Controls Division American Standard Corporation.

The output produced by changes in the strain gauge is transmitted to an Ellis Bridge Amplifier which contains the other elements of the bridge circuit as well as an amplifier. The output signal from the amplifier is proportional to the induced strain and is read on a Model 130B Hewlett Packard Oscilloscope. The complete circuit including amplifier and scope are shown schematically in Figure 7. TABLE I NORWOOD PRESSURE TRANSDUCER SPECIFICATIONS -15 to -15 to 500 psig 2000 psig Tranisducer Transducer Accuracy + 1% of rated pressure +50 psi +200 psi Linearity + 1o over entire range Resolution 0.1% of rated pressure 5 psi 20 psi Repeatability l,1/o of full scale 12.5 psi 50 psi Temperature Effect 0*02%/o may be balanced out at any given temperature The specifications listed in Table I and the calibrations found in Appendix B, clearly show that these transducers are acceptable as secondary pressure measuring devices, if used with a linear amplifier and sensitive display device such as the oscilloscope mentioned previousty. There is, however, a very definite need to choose the range of the transducer carefully so that the rated pressure is not exceeded by fifty per cent and that the pressure to be measured is greater than ten per cent of the rated pressure. Within these bounds the transducer will perform most satisfactorily.

OSCILLOSCOPE ELLIS BRIDGE HEWLETT PACKARD NORWOOD AMPLIFIER MODEL DB 130 TRANSDUCER IA MODEL BA-2. ICTIHI J<~~~~~O __RESo G __E BIE IN TERMINALD B GAGES BALANC 1/2 etc COMP I I I I. ~ BRIDGE ( TERMINAL Figure 7. Simplified Schematic of Electrical Measuring Circuit.

In order to cover the desired range of pressures adequately, two transducers were used, a 2000 psi transducer and a 500 psi transducer. Material Selection As low temperature operation was anticipated, all high pressure lines, vessels and connectors were made of austenitic stainless steel, All welds were made with 308 rods; to prevent delta-ferrite formation, Wherever possible, standard parts of 316 stainless steel were purchased. Special stainless steel packing washers and packing nuts, rather than the standard bronze parts, were used in the valves which were exposed to low temperature. All valve packings used in valves at low temperature were made of Dilecto, a glass impregnated with teflon, This material is dimensionally more stable than teflon and proved an excellent packing. Some difficulty was encountered at low temperature in the main bl6ck valve (L). This is discussed in some detail below, The iriher liner of the constant temperature bath is made of an 1/8 inch copper plate. The cooling coils were also constructed of copper. Around the copper tank there are six inches of Styrofoam insulation. Valve Operation at Low Temperature The operation of valves at extreme temperatures can present a problem. The basic problems encountered are the differences in thermal expansion of various metals and packings and the lack of good, low temperature lubricants.

-29HANDLE-ALUM. SLEEVE-410 S.S. 7'-GLAND 17 -4 PH THRUST WASHER 17-4-P.H. X LOCK NUT-416 S.S PACKING WASHER 316 SS HEATED SECTION PACKING- DILECTO BOTTOM WASHER-304 S.S. SPACER-304 S.S STEM ASSEMBLY -316 SS. STEM HOUSING 316 S.S. LENS RING-316 S.S. BODY 316 S.S Figure 8. Cross Section of Main Block Valve (L). Courtesy of Autoclave Engineers Inc.

-30A standard Autoclave high pressure valve with stainless steel used throughout proved unsatisfactory, possibly because of the incorrect choice of material as seizing occurred between the stem (420 stainless steel) and the bottom washer made of 17-4PH Armco steel0 A high temperature valve, Figure 8, was inserted and with several modifications worked well throughout the tests. This valve has a two piece stem, the sleeve of which is made of 410 stainless steel. At low temperatures, where it is very critical that all contacting parts be clean and dry, this sleeve tends to gall in the packing nut threads. In order to eliminate this problem the stem must be treated so that its surface has properties different from those of the packing nut. This may be accomplished by either nitriding the surface or putting on a surface film which can withstand low temperatures. Both methods were tried, the former by the Tufftride* process, and the latter with a coating of teflon about 0,002 inches thick, Both methods appeared to work well, although no extensive tests were made, During the course of the tests, the Tufftrided stem was used and the long stem was wound with a heater to keep the valve packing and gland nut at temperatures no lower than -lOOC. Temperature Measurement and Control Temperature in the cell was measured with a thermocouple made of matched 30 gauge copper and constantan wire, A continuous record of the cell temperature was made and the recorded reading was checked periodic cally with a Leeds and Northrup Semi-Precision Potentiometer, * Trademark of the Kolene Corporation.

-31Another copper-constantan thermocouple was attached to the outside of the cell. The output from this thermocouple was recorded by a Micromax recorder-controller. The control element was an on-off switch controller Skinner three way valve, This valve, when on, allowed gaseous nitrogen, under pressure, to force liquid nitrogen into the constant temperature bath. With the aid of a 192 watt heating element wound around the cell, and controlled by a Variac, temperatures could easily be kept at + 0o50C within the cell. The temperature of the reservoir was measured with a copperconstantan thermocouple and Leeds and Northrup student type potentiometer. The temperature of the expansion volume and gas burette were determined with standard thermometers. Temperatures taken with the thermocouples could be determined to + 05~0C under the most severe conditions, -1929C and within + 0.10C at ambient conditions, The thermometers could be read to + 0,20C. Safety Considerations Whenever experimental work is carried out at high pressures and extreme temperatures the safety of both personnel and equipment becomes an important consideration. The adsorption cell, after welding, x-raying and stress relieving was hydraulically tested to 8,500 psi which was about one and a half times the design pressure. The entire high pressure system was then hydraulically tested at 7500 psi. In addition to the hydraulic testingS several safety shields were provided as can be seen in Figure 3. The inner shield (M) is a 2 1/2-inch schedule 80 stainless steel pipe

-32with the end near the transducer closed, to provide a shield should the transducer fail. A 1/8 inch copper bath (N) provided additional protection, One quarter inch steel plate (O) served as a final shield with space for the gas to escape between the hinged cover and the fixed tank, The temperature of this shield was checked with the bath at liquid nitrogen temperature (-196'C) in order to insure that its temperature was above the brittle fracture temperature, about 30QF. The temperature was about 650F with a room temperature of about 72Fo, A blow out disk was provided in the high pressure lines and was rated at 3,000 psi or one and a half times the maximum normal operating pressure. In order to protect the glass vacuum system and sample bottles, a block valve (G) and a metering valve (H) were placed in series and the two were never opened together if the pressure upstream from the block valve exceeded two atmospheres, These precautions have proved sufficient; to date no accident has occurred. Leakage in the System In a static system it is important that the test cell in particular be leak-free. In order to determine the leakage, if any, in the test cell, a series of pressure tests were conducted over a twenty-four hour period each. In addition, tests at vacuum were conducted in order to determine if any leakage at low pressure was present. The cell was also tested using the pressure method with a helium detector,

The high pressure tests were conducted at 2,000 psi. It was found that no visible leakage, as determined by readings on the pressure gauge, occurred in a twenty-four hour period. This would mean that any losses would be less than one part in a thousand during that period of test, This was deemed sufficient to consider the system free of any leakage, Vacuum tests were conducted on the cell. The cell was evacuated to a pressure of ten microns and was then closed, The rest of the system was kept under a vacuum of ten microns and after a period of eight, sixteen and twenty-four hours the pressure in the cell was determined, The pressure in the cell was found to be no greater than twenty microns after a twenty-hour period. This was considered sufficient to consider the cell free of any leaks,. The helium leak detector was used to find any possible small leaks. In this particular test the cell was filled with helium to a pressure of 2,000 psi and all joints were examined with the probe from the helium leak detector. No leakage indications were discovered. The loading system was also checked for any leaks. This was done in a manner similar to the one used for the test cell, The duration of the tests, however, was only two and four hours. No leakage was found over that period of time as determined by readings taken on a pressure gauge. The vacuum system was tested for possible leaks and the leakage in this part of the system was found to be less than one micron per hour. This was considered satisfactory for the purposes of the experiment.

Calibration of Equipment All vessels and lines, with the exception of the large expansion volume, were calibrated using the volume expansion method. Helium, as well as dry air, were used as the calibrating gases except in the free space determination in which only helim was used, The large expansion volume was calibrated by filling it with a measured amount of water, The calibrated volumes are listed in Table II, The effect of pressure and temperature on the cell volume is negligible. Complete calculations may be found in Appendix Do Several pressure gauges were used in the study in order to cover adequately the range of pressures studied, The two feed pressure gauges, 0o800 psi and 0-2000 psi and the -30 to 300 psig were calibrated against a dead weight tester. The calibration curves may be found in Appendix B. The calibration data was subjected to regression analysis using a program developed for the Bendix G-15 computer after the method of Milne(47) A second order fit represented the data very well. The pressure transducers, 2000 psi and 500 psi, were calibrated in the system against the pressure gauges, These data were also fitted with a regression curve of second order. The calibrations may be found in Appendix B. Excellent reproducibility was achieved at a given temperature range and although they are compensated for changes in temperature over part of the range, a change in the calibrations occurred over the range of interest. No drift in the calibrations occurred with time and it appeared, therefore, that the slight decrease in battery power had no effect on the readings,

-35TABLE II CALIBRATED VOLUMES Volume Volume Calibrated CC. Constant Volume Reservoir 923 + 3.cc Reservoir Connections 38.2 + 1,cc Central Distribution "T" 8,63 + 0.1cc Cell Sampling Line Section 2.46 + 0,05CC Vacuum Safety Gauge and Connections 12.7 + 1,cc Glass Lines (To top of barameter) 127.7 + 5,cc Barometer 0.196 + 0004cc Expansion Volume 11,780cc + 50.cc Adsorption Cell-Filled-Free Space Runs 1-107 8947 + 0o7cc Runs 107 - 88.8 + 1,Occ

-36The thermocouple used to determine the cell temperature was checked against a nitrogen-filled pentane thermometer at the ice point, in dry ice, and in liquid nitrogen, At all three temperatures the readings checked within 0.5QC. As this was well within the accuracy of the equipment the temperatures recorded by the thermocouple were assumed to be accurate,

EXPERIMENTAL METHOD Adsorbent The adsorbent is a calcium substituted Aluminosilicate commercially known as Linde Molecular Sieve 5A. This material is similar to the natural zeolites, chabizite and analcite, Natural materials of this type and related synthetic zeolites were studied by Barrer and coworkers The sieve used in this study has been studied in some detail. by Breck and co-workers of the Linde Company. The basic chemical com(12,53) position of the hydrated crystal is Ca6(Al.02)1(SiQO12) 513H20. The basic structure of the unit cell has been described as a framnework of alumina (A10) and silica (SiO4) tetrahedra linked at their apexes, providing a large interior cavity linked to other cavities by eight sided and six sided passages. The edge of the tetrahedra form eight membered oxygen rings with a diameter of 5,0 to 5,6A, In addition there are small six 0 membered rings of approximately 3A in diameter leading from the interior cavityo The sieve is supplied in the form of 1/16 inch or 1/8 inch pellets and in this study the former were used, These pellets are about 80 per cent actual crystal, The remainder is binder. The sieve has a surface area of 500 square meters per gram (see Appendix C) as measured by the "BET" method and agrees with the data of Breck (12). The adsorbent will rapidly adsorb water vapor upon exposure to the atmosphere, It is therefore necessary to activate the sieve before any equilibrium determinations are made, The recommended activation conditions are 350-C at one atmosphere or lower temperatures at -37

-38lower pressures. The sieve used in these experiments was activated for a minimum of 12 hours at 175~C and at a pressure of 10 mm.Hgo Between individual equilibrium determinations, the adsorbent was held for 4 hours at 175~C and 10-2mm Hg after nominally complete desorption had been achieved. In order to determine the effect of regeneration temperature and pressure, the surface area was determined after desorbing samples -6 of sieve at 300~C and 10 mm Hg. and at 125~C and 10 2mm.Hg for eight hours. There was no significant difference; the specific surfaces of the samples ($ee Appendix C) were within three per cent of each other which is well within the accuracy of the method, In addition, a sample of the adsorbent which had been regenerated in the cell at the normal desorption conditions, 175~C, 10- 2mm.Hg for a minimum of 4 hours, was desorbed for 12 hours at 350~C, The weight difference was 1.45 grams per 100 grams of adsorbent, equivalent to about five per cent of the water capacity of the sieve. Gases The Methane used in this study was obtained through the courtesy of Phillips Petroleum Company in standard, seventy cubic cylinders at 2000 psi. The gas received was pure grade, 99 plus per cent methane. A number of analyses of the methane were made during the course of the study with the mass spectrometer. A typical analysis can be found in Table III. The nitrogen used was prepurified grade nitrogen purchased from the Matheson Company in standard 220 cubic foot cylinders at 2000 psi. This gas is sold with a specification of 99,99 per cent nitrogen

-39TABLE III ANALYSES OF PURE FEED GASES Component Per Cent METHANE Methane 98 6 98.7 Nitrogen 1,0 0.9 C + 0,2 0,2 C02 0,2 0,2 H20 None Evident* NITROGEN N2 99.9+ H20 None Evident* 02 None Evident** * No. 18 peak on mass spectrometer ** No. 32 peak on mass spectrometer

-40o content and a maximum dew point of -96~Co Each cylinder was sampled and a mass spectrometer analysis made; a typical analysis can be found in Table IIIo As both gases were essentially pure and free of water, they were used as receivedo Experimental Method Of the two basic experimental methods, dynamic or flow and static or batch, the latter is in many ways the easier to execute and was chosen as the method to obtain the equilibrium data. Besides avoiding some of the experimental difficulties, such as keeping flow rates constant and measuring both flow rates and composition accurately for the entire run, the static method is more economical in regard to material consumption. After activation of the adsorbent which is discussed above, any number of adsorption-desorption cycles can be made. The gas mixture to be adsorbed is mixed in the constant volume reservoir, The mixture is allowed to come to equilibrium for 24 hours. This was focund sufficient to provide a feed mixture of constant composition in the reservoir. Before any run the adsorption cell is brought to the adsorption temperature and the initial pressure must be less than 0.02 mm. mercury~ The pressure and temperature of the reservoir are then determined~ The pressure transducer circuit is balanced so that at full vacuum the oscilloscope display shows a balanced bridge circuit. The main block valve (G-4) and metering valve (H-l) to the vacuum system are closed and the main block v-alve from the reservoir (G-6)

is opened. The cell is loaded with the gas mixture and the main cell valve (L) is closed. The pressure and temperature of the reservoir are again determined and a sample of the feed gas is taken for analysis. The system is allowed to come to equilibrium. Temperature is recorded continuously during the run. At the beginning of any run, as the gas is admitted and adsorbed, the system temperature is above the equilibrium temperature. The system temperature, depending upon the conditions, will be as much as 500C above the equilibrium temperature. Within 15 minutes, the temperature dropped to within five degrees of the equilibrium temperature. The equilibrium temperature is reached within one hour. Periodic pressure readings are made during the run. The pressure decreases rapidly during the first fifteen minutes. The pressure normally will decrease to the equilibrium pressure within the first hour. In order to insure that equilibrium is achieved, at least eight hours is allowed. A series of tests over a longer period of time indicated that essentially no change in the gas phase occurred after four hours, the time the first sample was taken. After eight hours, two pressure readings taken half an hour apart, at constant temperature, must be identical before the system is said to be at equilibrium, At equilibrium, the cell pressure and temperature are recorded and a gas sample is taken for analysis. The adsorbate is desorbed into the large, 11,780 cc., expansion chamber and the pressure, temperature, and composition determined. Desorption for purposes of material balance is said to be complete when the equilibrium pressure is less than 1 mm. Hg. It requires between three and six desorption steps to obtain essentially complete desorption. In

order to speed this process, temperature cycling is used with a maximum temperature of 1700C. After the equilibrium pressure is below 1 mm., the system is evacuated further to an equilibrium pressure of 0,01 mm, Hg. at 1700C. Sampling For ease of construction, only one line connects to the ad$orption vessel. A small, approximately 2.5 cc. sampling section exists between the main cell valve (L) and the first block valve (G-5). The main distribution system then connects with the sample outlet and the glass sampling tube connection, In order to obtain a gas phase sample, a 2,5 Cc. purge sample is first taken. The line is then evacuated and the sampling section block valve (G-5) is closed. The sample is then obtained by opening the main cell valve (L). It is expanded into the glass sample tube. It is imperative that a true gas sample be obtained. The total sample is about five per cent Of the total gas phase, In several tests several samples were taken in rapid succession, As indicated in Table IV, there is no difference in samples taken in rapid order; thus it can be assumed that the sampling technique gives a representative sample of the gas phase. On the other hand, if some time, say 20 minutes, elapses between samples, a new equilibrium is being approached. The rapid sampling is also accompanied by a temporary drop in pressure which increases again to a new equilibrium value with time; this supports the opinion that the gas phase sample is a representative sample of the gas phase only, The samples were analyzed on a Consolidated Electrodynamics Corporation Model 21-103B Mass Spectrometer, This instrument is said

-43TABLE IV PAIRED GAS SAMPLES FROM ADSORPTION CELL TAKEN AS A FUNCTION OF TIME TO DETERMINE SAMPLING VALIDITY Sample Mole Fractions Deviation from Aver. Number Time CH4 N2 for CH4 Analysis P-289 8:00 AM.6512.3488 - 0.0049 P-290 8:20 AM.6561.3439 0.0049 P-318 9:55 AM.1182.8818 0.0007 P-319 10:00 AM.1169.8831 - 0.0007 P-340 7:20 AM.5030.4970 0.0010 P-341 7:25 AM.5010.4990 - 0.0010 P-266 2:15 PM.1459.8541 0.0000 P-267 2:18 PM.1459.8541 0.0000 TABLE V REPLICATE ANALYSES OF A SINGLE SAMPLE Analysis % Deviation Mole Fraction From Average for Sample No. CH4 N2 CH4 Analysis P-570a 0.1731 0.8269 - 0.34 P-570b 0.1735 0.8265 - 0.12 P-570c 0.1744 0.8256 o.4o Mean 0.1737 TABLE VI REPLICATE SAMPLES FROM A UNIFORM MIXTURE Analysis % Deviation Mole Fraction From Average for Sample No. CH4 N2 CH4 Analysis P-428 0.7100 0.2900 - 0o.14 P-431 0.7112 0.2888 0.03 P-435 0.7113 0.2887 o.o4 P-438 0.7115 0.2885 0.07 Mean 0.7110 St'd Deviation = 5.87 x o10

to be accurate to less than one-tenth of one per cent but replicate analysis of the same sample, found in Table V, indicate that the precision of 0.2 per cent can be expected with certainty. To determine the accuracy of analysis, a series of replicate samples were taken and analyzed. The results may be found in Table VI. This data also indicates that any given analysis is good to + 0.002 mole fraction or 0.2 per cent. Method of Calculating Adsorption Equilibrium The adsorption equilibrium cannot be directly calculated from the experimental measurements, pressure temperature, gas volume and composition, The amount actually adsorbed is the difference between the amount of gas admitted to the cell and the gas remaining in the free space of the adsorption cell at equilibrium and may be represented for a given component by: i (NI,i - NF,i) - FS, i (29) Equation (29) may be rewritten in terms of volume, density and compositions: i (ViPI VFPF)YI,) - VFSPFSYi (30) but the density is some function of pressure, temperature and composition: P = f(P,TY) (31) In the regions of low pressure and high reduced temperatures the ideal gas law will represent the P-V-T data adequately. At high pressures and lower temperatures the gas behavior cannot be adequately represented by the ideal gas law and some correction must be made. Therefore, a more complex equation of state or a compressibility factor must be used. Of

-45 - these two methods, the equation of state lends itself to computer solution and was used here. Several equations of state have been developed to represent the experimental P-V-T data. Of these, the most adequate is the Benedict, Webb, Rubin equation, which is an empirical equation developed for hydrocarbons using eight constants(6,7) and is of the form: P = RT + Clp2,+ C2p3 + C3p6 (32) where C1 = BoRT - Ao - Co/T2 (32a) C2 = (bRT.a) + C/T2[(l+yp2)/eYP2] (22b) C3 = ao (32c) The B-W-R equation as developed originally was intended to predict vaporliquid equilibria of hydrocarbon mixtures. In the original work no, constants for nitrogen were published. Bloomer and co-workers fitted a modified B-W-R equation of state for nitrogen(l) The equation added two constants to better represent the data, changing the first and second virial coefficients, so that: C = BoRT - Ao - Co/T2 - Do/T4 (32d) C 22/ep2 (32e) C2 = bRT - a + ( + )[(l+yp)/eP] (3e) In later work, based on experimental data of Sault(54), Darby,(l7) Pace, (44) and Keyes and Burks(34) for nitrogen-methane mixtures, Ellington and coworkers(l9) recommended the following combinatorial rules for the tenconstant modified B-W-R equation; o = xiBoi (32fl) Com =( xiCoil/ ) (32f2)

-46Dm = (Z xiai/2 2 (32f3) = (Zx.a ) (32f3) bm = (Ixi i/ (2f) i13 3 am = (. iai ) (32f4 ) xjceYia11/5~j (32f7) Eym = ( Xii "1/33) (2f 8) with a special modification for the constant A: 0o A = (Z xiAoi ) - 0.1000 (XN2 x XCH4) (32f9) _i The equation, as modified, represents data with an average deviation under one per cent at pressures up to 1500 psi and temperatures as low as -135~Co As this equation is explicit in pressure, it is necessary to solve for the gas density by an iterative procedure, The Newton-Raphson method suits itself ideally to obtaining a solution rapidly in this case. In addition, the equation may be modified so that the fugacity of a gas may be computed. Once the density is determined, the amount of each constituent in each phase can be readily determined, using relationship (30). The relative volatility can then be obtained from its definition: ac (YN2/XN2)/(YCH4/XcH4) (33) 2 2~~~~~~~~~~(3

As part of the calculations necessary to obtain the equilibrium loading are iterative and lengthy, the entire calculation was programmed for the IBM 704 computer utilizing the MAD translator, The logical flow sheet, together with a sample calculation may be found in Appendix D. Errors in Measurement Two basic methods for error analysis exist. The first and perhaps most often used is statistical in nature and depends on the basic assumption that the measurements are Gaussian-distributed about the true value. In order to get a good measure of the errors, which are propagated in the results, a number of values of the result are necessary at a given parameter or it is necessary to assume a model for the data, The second method of evaluating the effect of errors in measurement on the results is to determine the effect of the maximum possible errors in measurement on the calculated results. This method will bound the maximum error and in all but a very few cases the error in the finally calculated result will be measurably less that the maximum indicated from this type of analysis. The second or maximum propagated error method has been used in analyzing the effect of errors in measurements on the results. In determining equilibrium values it is important that the degree of accuracy of the results be known with some certainty. This could be achieved by obtaining data for a number of runs at a given set of conditions, but it is often uneconomical to carry out this number of experiments. A more realistic approach, therefore, is to make only a single run at most condit-ions and ob-tain results over a wider range of the test parametersJ

It is then best to make the error analysis by the second method unless one can assume a model which suits itself readily to a statistical regression analysis. The maximum errors in the various measured variables are listed in Table VII. The maximum errors were determined by checking the values read against primary standard values or secondary standard values. In addition, the errors in reading due to parallax and other factors were taken into consideration. In order to determine the effect of the maximum errors in measurements, the values used in a number of calculations covering the spectrum of results and basic data were altered. These altered data were then used to compute results using the basic computer program discussed above, The results obtained were then compared with the results obtained using the original data. The conclusions of the error analysis may be found in Tables VIII and IX, The major effect on the calculated equilibria ensues from inaccuracies in the measurement of the feed pressures. This occurs because a relatively small net pressure is the difference between two rather large pressure values; thus small errors in the initial or final feed cell pressure can materially effect the total cell loading. The calculations indicate that the other errors in measurement propagate errors which are quite reasonable. The errors resulting from inaccuracies in the feed pressure determination should be well below the maximum as both pressures are read with the indicated pressure decreasing and within five minutes of each other. Any reading errors, therefore, should be in the same direction,

-49TABLE VII ACCURACY OF MEASURING DEVICES % Error of Max Min Measuring Device Accuracy Reading Reading 2000 psi Pressure Gauge (500-2000 psi) + 5 psi 0.25 1.0 800 psi Pressure Gauge (100-800 psi) + 2 psi 0.25 2.0 2000 psi Pressure Transducer + 20 psi 1.0 3.33 (600-2000 psi) 500 psi Pressure Transducer (700-l00 psi) + 5 psi 0,7 5.0 Temperature (Thermocouple) (123K-300ooK) + 0,50K 0.5 o.16 Temperature (Thermometer) (3000K) + 0.50K 0.16 0.16 Mass Spectrometer Composition Analysis (Mole Frac.) + 0.002

-50TABLE VIII MAXIMUM ERROR PROPAGATED IN PURE COMPONENT EQUILIBRIUM DATA Milligram % Moles Measurement Changed Adsorbed Deviation Original Run (Run 26) 0.266 0 Feed Pressure 0,.5 0.258 3 Feed Volume 1 0.265 0.5 Cell Pressure 3 0.260 2 Cell Temperature 0.2 0.267 0. 5 Cell Volume 1 0.265 0.5 TABLE IX MAXIMUM ERROR PROPAGATED IN BINARY EQUILIBRIUM DATA Milligram Moles CH4 Measurement Changed Adsorbed Deviation Original (Run 108) 0.0348 Feed Pressure 2 0.0338 3.0 Feed Volume 1 O.0356 1.8 Cell Pressure 4.0 0.0350 o.6 Cell Pressure 4.0 * 0.7 Cell Composition 1.3 0.0349 0.3 Cell Volume 1.0 Cell Pressure, 4.0 Composition 1.3 0.0350 0.57 and Temperature 0.16 * Total adsorption

-51The amount adsorbed should be within three per cent of the true value under all but the most unusual cases but even in these cases the value will be correct to plus or minus six per cent, The values of relative volatility in all cases are within six per cent of the true value.

EXPERIMENTAL RESULTS Pure Component Adsorption Pure component isotherms were obtained for methane at three temperatures and over a wide pressure range. Well above the critical temperature of methane at 295~K, nine equilibrium determinations were made up to 50 atmospheres or 760 psia. At 1950K, close to the critical temperature, nine determinations covered a pressure range of about 30 to 450 psia. Below the critical temperature, at 1750K, the equilibrium loading was evaluated at four pressures covering the range of pressures up to the saturation pressure. A tabulation of these results may be found in Table X. As discussed in detail below, the maximum error in any individual determination of the amount adsorbed is plus or minus four per cent. A statistical regression analysis of the data indicates that the isotherms are accurate to plus or minus five per cent 95 per cent of the time at the pressure extremes. The isotherms may be found in Figure 9. Twenty-one values of pure nitrogen equilibrium loading or amount adsorbed, were obtained, These may be found in Table XI. The isotherms were determined at 295 K, 195"K and below the critical temperature of nitrogen, at 123~Ko At the higher temperatures, data was obtained at pressures up to 1300 psi and 900 psi respectively. Below the critical, the amount adsorbed was determined up to the saturation pressure. The 95 per cent confidence limits on the extremes of the adsorption isotherms, which may be found in Figure 10, are plus or minus five per cent, The experimental data from which the equilibrium calculations were made may be found in Appendix A. -52

6.0 -I I I LEGEND x 295 OK 0 298 OK (LINDE DATA SHEET) 5.0 C - - I950K_5 SDO.9I 1980K (LINDE DATA SHEET) 5. 0 t175~5K..~......SD-91 l 175 oK 4Q0 ____ W CHECK RUN C,) W~~ w -j QX — o ~ 2 l9CHEK xU xx _1 0 2.0 a.0 C) 0 4 1.0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 PRESSURE, ATM. Figure 9. Methane Adsorption Isotherms on 5A Molecular Sieve.

-54TABILE X PURE METHANE ADSORPTION Run Pressure Temperature Amount Adsorbed No. ATKo ~K Milligram Moles/GM Average Temperature: 295 K 115 51.3 297.0 3.12 25-1 47,63 293*0 3,24 26 40.15 297,0 3.01 27 29,94 296,0 3*05 28 25,86 295,5 2,91 122 22,5 296.0 3*02 25-2 22,46 293 0 2, 94 124 9,78 296X5 2,46 123 6.40 296,0 2.52 Average Temperature: 195 ~K 114 31*30 195,5 4,70 52 29,70 196,o 4.61 56 24,84 194,o 4.56 58 13*13 195.5 4.39 92 9,25 19350 4,34 53 9,o8 195.5 4.66 57 3*54 195.5 4,17 93 2,93 19350 4,12 59 2,72 195,0 3,92 Average Temperature: 1750K 78 22o52 175*5 4,95 79-1 8,98 174,5 4,83 79-2 4,87 174,0 4.62 80 0*50 175 0 4.25

6.0 LEGEND -E 1 X 295 OK 5.0 D- _ _ _ _ _ _ _A 195 OK /I~~ 0I ( i I ( I ~~~~~~o 198 0 K (LINDE DATA SHEET E3 1230K SD 62) ~ ~ _ _ 0 CHECK RUN 195K -...-.-"" ~ - ~ ~ - - -_ -~........ 4.0 w 0 3 D x <I: (.9~~~~~~~~~~~~~~~~~?9 w Cr 0 tO~I 2 0 30 40 50 60 70 80 9?.95.._..~._.. i.n (nu,~~~~PRESSURE, ATM. Figure 10. Nitrogen Adsorption Isothens on 5A Molecular Sieve. o ro ~ ~~ I 20: 50 40 50 6 0 70 8 0 90 PRESSURE, ATM. Figure 10. Ni~trogen Adsorption Isotherms on 5A Molecular Sieve.

TAB3LE XI PURE NITROGEN ADSORPTION Run Pressure Temperature Amount Adsorbed No,, ATS,...K Milligram Moles/GC Average Temperature: 295 ~K 29 85,4 294.0 3.06 30 62,9 295*5 2,74 31 33,68 295*0 2.32 32 21*57 295.5 1,92 120 21,3 295,5 1o95 33 15,65 296,0 1.80 103 10,55 296.0 1,66 121 6*78 295,0 1.39 Average Temperature: 195 ~K 1'12 57,84 195X 5 4,54 113 44.91 195,5 4,35 71 33548 195.0 4.12 51 31X98 195 0 4.31 72 17.56 194,o 4,16 73 7,72 194,0 3*90 74 2,86 194,0 3554 Average Temperature: 123 ~K 84 24.82 123,0 5,24 85 17,60 12355 5.30 87 13,34 123,0 5.23 86 11,64 123o,0 5 19 88 5 78 123 0 5 14 89 1,59 123,0 4,93

-57Lewis(27) reported very limited data on methane adsorption on Davison Silica Gel and Columbia G Activated Carbon. The data indicate that the Molecular Sieve has a greater capacity for methane. Antropff(l) studied the adsorption of nitrogen on a Bayer-Werk Leverkusen activated charcoal. The data show that his particular adsorbent has a slightly higher capacity for nitrogen; for example, at O0C and 27 atmospheres it adsorbs 2.7 milligram moles per gram as opposed to the 2.2 milligram moles per gram capacity of the Molecular Sieve at 220C and 27 atmospheres pressure, Binary Adsorption Adsorption equilibria for mixtures of methane and nitrogen were evaluated at the four temperatures at which pure isotherms were determined, 295K,? 195"K, 175"K and 123"Ko At the higher two temperatures, equilibrium adsorption was determined at pressures as high as 1350 psi or 90 atmospheres and gas phase compositions ranging between ten and ninety per cent. Results for the sixty runs may be found in Tables XII and XIII. Just below the critical temperature of pure methane, at 1750K, the equilibrium loading increased from 4,75 to 5,04 milligram moles per gram of adsorbent with an increase of pressure from 7,6 to 30 atmospheres. The equilibrium amounts of each component adsorbed may be found in Table XIV. At 123 K, near the critical temperature of pure nitrogen, the quantity adsorbed. was determined up to 300 psia for an equilibrium gas mixture containing five per cent methane. The maximum pressure investigated was limited at the higher two temperatures by the cylinder pressure without any additional compression. At the lower two temperatures the dew point pressure was the upper pressure limit.

-58TABLE XII MIXED ADSORPTION AT 295 0K AVERAGE TEPERATURE Mole Fraction Amount Adsorbed Run Pressure CH4 Milligram Moles/(CM Relative No. ATM. Gas Phase NT NCH N2 Volatility 13-1 89*82 0*793 3.27 261 o0,68 2,2 13-2 89.82 0.802 3,09 2,48 0,61 2*31 13-3 83.02 0.787 3,30 2.60 0.70 2.08 13-6 57.16 0,790 3533 2.63 0.70 2,16 13-9 42.,19 0,793 322 2.55 o*67 2.14 13-11 30.28 0,818 3.19 2.61 0.58 2,46 14-1 76.21 o 0639 3.45 2.66 0 79 1.91 14-2 74.85 0*633 3,45 2.68 0.77 2.02 15-!l 28,58 0,593 2,70 2.04 o.66 2.12 15-2 27*90 0,544 2*70 2.10 o.60 2*94 16-2 14.29 0,590 2,36 1,74 0.62 1.94 17-2 90.84 0,161 3. 08 0,87 2*21 2.o6 17-4 61,92 0.161 3,16 0.88 2,28 2.02 17-5 59*98 0o,168 3,18 o0,87 2,31 1.87 18 47,29 0,155 2-.86 0,74 2.12 1.91 19 22.80 0o,146 2*05 0.53 1.53 2,03 20 74,85 0,705 3551 2,94 0,57 2.15 21 56.14 0,700 3,00 2*50 0.50 2o14 22 34.02 0,675 2.83 2.29 o054 2,05 23 24.50 0,667 2,38 1,94 o044 2.19 24 15.31 o,656 2*02 1.62 o,4o 2*15 104 30,62 0*271 2,86 1,28 1*58 2,16 105 21*71 0.274 2,56 1,10 1,46 2,00 106 9,12 0,270 1.89 1.78 1.11 1.90 107 21*07 0*150 2.41 o,6o 1.81 1.89 108 8*39 0,149 1,75 0*39 1.36 1.66 lo9 35.40 0.171 2,90 o.86 2.04 2.03 110 19.64 0,173 2,31 0,67 1.64 1.94 111 12,83 0.174 2.13 0,60 1.53 1*85

-59TABLE XIII MIXED ADSORPTION AT 195 OK AVERAGE TEMPIPTUfRE Mole Fraction Amount Adsorbed Run Pressure CHE4 Milligram Moles/G:M Relative No-, A4 Gas Phase NT NCH4 N Volatility 34-1 84.o4 0o150 4.44 1.55 2.89 3.03 34-2 81,65 0,153 4.54 1*55 2.99 2.86 35 46.61 0,120 4.50 1,40 3*10 3.30 36 28.92 0.110 4.34 1.26 3,08 3.29 37 16.47 0.112 4.18 1.18 3,00 2*93 39 47.63 0.606 4.29 3.53 0.76 3.0o4 4o 28.58 0,556 4.25 3,45 0o80 3.44 41 11,36 0,503 3*50 2,64 0,85 3.36 43 37*43 0,405 4.51 3,08 1,43 3.20 44 16*33 0,406 4,28 2.98 1,30 3.40 45 76,89 o,476 4.80 3,20 1.60 2.21 46 36,74 0,611 4.30 2,78 1,52 2,88 48 18.17 0,451 4.18 2.98 1,20 3.02 61 6o51 0,314 4.10 2.22 1.88 2, 58 62 1.97 0,334 3,75 1,97 1 78 2 21 64 20.48 0,121 4,23 1.18 3.b5 2,81 65 7.59 o,126 4.07 1,03 3.04 2*36 66 1,38 0,133 3,42 0,83 2,59 2,69 67 31,20 0,546 4,75 3,72 1*03 2,71 68 16,40 0,531 4.96 3,74 1.22 2#70 69 7,96 0,508 4,36 3.21 1*15 2.69 70 o0,95 0,542 3,54 2,53 1.01 2.34 94 33.4 0,201 4.33 1,89 2.44 3.05 95 13.27 0,176 4,19 1,59 2,60 2.87 96 4,76 0,188 3.92 1.40 2.52 2,42 97 28,58 0,357 4.63 2.91 1,72 3.05 98 19.53 0,334 4.44 2.67 1,77 3,03 99 6,98 0,326 4.13 2*30 1,83 2.74 100 24,36 0.687 4,60 3,96 o064 2,76 101 13*98 o,658 4,55 3,86 0,69 2.92 102 4,49 0*700 4,21 3.46 0.75 1.95

-60TABLE XIV MIXED ADSORPTION AT 175 "K AVERAGE TEMIPERATURE Mole Amount Adsorbed Fraction Milligram Moles/GM Run Pressure CH4 NT NCH NN Relative No. ATM, Gas Phase 4 2 Volatility 75 30.55 o0807 5.04 4.07 0.97 3.o4 76 7.60 0.754 4.75 3.59 1.16 3.18 118 17.08 0o786 4.90 3*85 1.05 3.10 TABLE XV MIXED ADSORPTION AT 123 K AVERAGE TEMPERATURE Mole Amount Adsorbed Fraction Milligram Moles/GM Run Pressure CR4 NT NCH N Relative No. ApTM Gas Phase 4 _2 Volatilty 81 11,80 0.050 5.03 1,07 3.96 5.20 82 1.27 0*050 5.02 0.92 4.10 4,30 90 18.98 0.051 5.16 1,28 3.88 6,10 91 5.99 0,047 5.07 1.03 4.o04 5.02

-61The individual determinations of the binary adsorption equilibria have a maximum error of plus or minus six per cent. It is apparent from Figure 11, which shows the total adsorption for the binary system, that the total equilibrium amount adsorbed is independent of composition. The 95 per cent confidence of these isotherms at their extreme is seven per cent. TALBLE XVI CONFIDENCE LIMITS ON EQUILIBRIUM VALUES 95 Per Cent Confidence on Isotherms at Pressure Extremes Methane 5% Nitrogen 5% Methane-Nitrogen Mixtures 7%

I1O 123 OKl 175 0 K I IrIIII - ~ cl i 1 95 w 3 I 98 - 6 -_ 0 V~~~~V 2r~ ~ ~~~~13~ -J ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - 6.__. —-- _ —-' 0~~~~~~~~~~~~~~~~~~. to"~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 789100 2 3 4 5678910' 2 3V 4 5 6-78910 _1~~~~~~~~~~~.1 9 8 7L 6 5 7 89 I0~ 2 3 4 5 6 7 8 9 I0"1 2 3 4 5 6 7 8910 PRESSURE, ATM. Figure 11. Mixed Adsorption as a Function of Pressure and Temperature.

DISCUSSION OF RESULTS Pure Component Adsorption Correlations A number of methods of correlating the data were tried in order to extend the usefulness of the data over the entire range of interest. For pure component adsorption, both methane and nitrogen, the data was correlated using both the Freundlich N = P /n (3) and Langmuir N aP (la) 1 + bP relationships as previously discussed. The constants for the correlations were obtained by the method of least-squares which is well detailed in standard references on numerical methods such as Milne.(47) A plot of the experimental results for the pure components after Freundlich may be found in Figure 12. A Langmuir plot of the results has not been included but this would be rather easily constructed by plotting "'P/N" in atmospheres per milligram mole per gram versus "P" in atmospheres. The constants, valid between 100 and 3000 Kelvin, for the Freundlich equation are shown as a function of temperature in Figure 13, while those necessary to complete the Langmuir relationship may be found in Figure 14. Analytical relations for the Langmuir constants as a function of temperature will also be found with the graphical representation, The dependence of the constants on temperature is quite similar to that found by Fowler and Guggenheim(20) in their statistical development of the Langmuir equation o -63

I0,, 8 - _ 1~_ __ _ - LEGEND 7 -- - o N2 6 6E C H,I I _'~~N 123'K WV)~~~~~~~~~~ CH4 5H4 - 195 K 0 19 _ _ _ _ _ _. _ _' 3 PRESSURE, ATM Figure 12 Pure Component Adsorption on 5A Molecular Sieve 0~ z 0.7 1.0 2 3 4 5 67 891I0 20 40 60 801I00 200 PRESSURE, ATM Figure 12. Pure Component Adsorption on 5A Molecular Sieve.

-65o I0 6 -I 4 66~~~~~~~EPRTR, /nCH4 0 i F~ 04 8 4 2 I0 2_ 0 50 100 150 200 250 300 350 TEMPERATURE, K Figure 13. Constants for the Freundlich Equation for Nitrogen and Methane and Their Mixtures (P in ATM.)

~ k -66- QCH 2 2 I0 f e~~~~~~~~~~~~~~~~~~~~I-'4 oCH 8~~~~~~~~~~~~~~ I2 bH aN - aJ 4 0 10 8 6 _!o %~ 4 r j LANGMUIR CONSTANTS 00 6 FOR PURE COMPONENT ADSORPTION AS A FUNC4 ~ —-— t —- TION OF TEMPERATURE aCH4 =5 510 x 1600129T bcH4 =502 x1(5~'~lla T a = 1360 x 160'0130 T _1~ 2 \ I0 10 I bN, -253x 1(~0.0122T 50 50 100 150 200 250 300 TEMPERATURE,OK Figure 14. Langmuir Constants for Pure Component Adsorption.

Both relations fit the experimental data rather well with an average correlation coefficient for the logarithmic form of the Freundlich relationship equal to 0~96 for nitrogen and 0~92 for methane, The standard error in the logarithm of the amount adsorbed is less than Ol6'milligram moles per gram (95 per cent confidence) when N is equal to 4 illigram moles per gram~ In the case of the Largmuir relationship, thie average correlation coefficient is 0.99 plus and a standard error is less than 0005 which represents a 95 per cent confidence limit of +0.1 in N, the amount adsorbed. That both relationshps represent the data well is not surprising. Trapnell(60) indicated that, with the proper choice of constants, either the Langnuir or the Freundlich relationship will represent adsorption data in the middle pressure range adequatelyo In addition to the Langmuir and Freundlich correlations, two correlations based. on the Polaenyi Potential Theory were attempted0 The first is the one first proposed by Lewis and co-workers(37) where N/ps, the volume adsorbed, is plotted as a function fo/f s a measure of the adsorption potential0 Here, fs, is the futacity at the saturation pressure, as determined from an extrapolated Clausius.Clapeyron vapor press sure equation, calculated with the help of the Benedict-Webb-Ruebin fugacity relationship. The gas phase fugacity, f, was also evaluated using the B-W-R relationships0 The density, ps, which is the important correlating factor, was evaluated at the temperature equal to the temperature of saturation at the equilibrium pressure, The actual density data for methane is due to Bloomer and Parent(lO) and for nitrogen is due to Bloomer and Rao. (11)

0 10 9 8 -- 6 LEGEND 5~~~~~~~~~~~~~~~~~~~ 2 T<Tcr 4 G T>Tcr 3 E 0 >2 z T-3. Tcr T< cr -I10~~~~~~~~ I0 -i- _"'_ _-.. 8 7 ____ 6 5 0 2 4 6 8 I 0 1 2 14 16 18 20 22 24 26 28 30 32 34 36 38 40 TPs In fs/f, K gm mol/cm3 Figure 15. Volume of Nitrogen Adsorbed as a Function of the Adsorption Driving Force.

E]t T <Tc 00 -CI I I0 8 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 T/os In fs/f,~K gm mol/cm3 Figure 16. Volume of Methane Adsorbed as a Function of the Adsorption Driving Force. Driving Force.

-70The correlation for pure nitrogen, as shown in Figure 15, indicates that a good correlation is obtained above the critical point independent of temperature; however, a different correlation appears to exist of adsorption below the critical temperature. This difference could easily be due to an incorrect choice for the density correlating variable. The methane correlation, Figure 16, by this method appears to be quite good for values both above and slightly below the critical with the exception of two points which lie noticeably above the correlating line. These points are obtained from equilibrium data obtained at 40 and 50 atmospheres where the present method of evaluating the adsorbed state density may not predict this density correctly. This same observation can be made about the single point lying well above the nitrogen correlating line. Correlations for Total Adsorption from a Mixture A correlation similar to the one detailed in the previous paragraph was attempted for the total amount adsorbed from mixtures of methane and nitrogen. The density in this case was assumed to be the molal average density of the two components in the adsorbed state, using the densities of pure components at their respective saturation temperatures as a basis for obtaining the average. It can readily be seen from Figure 17 that no single correlation could be achieved by this method. The correct choice of density again seems to be the major problem. The total amount adsorbed from a mixture of nitrogen and methane can also be predicted by either a Freundlich or Langmuir type relationship. The Freundlich equation appears to represent the data better; the necessary constants may be found as a function of temperature in Figure 13. The

0 9 v 8 V LEGEND 7~~~~~~~~~6 ___ ~~~0 123 OK E 175 OK 5 A 195 OK 4A W' V 295 oK " A VA295 O >2 z 1230K 1950K " V -I I0 9 8 v I I ~~~"'~.... 7 6 5 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 TPs In fs/f Figure 17. Volume of Nitrogen-Methane Mixtures Adsorbed as a Function of the Adsorption Driving Force.

-72plots of the total amount adsorbed as a function of pressure and temperature, as found in Figure 11, show that the total adsorption is independent of the composition of the binary gas in equilibrium with the adsorbate. This is not entirely unexpected at the higher temperatures where both-gases are well above their critical temperatures. Near the critical point of methane, this behavior is still present and can best be explained by assuming that both gases compete for a given amount of adsorption space which limits the number of molecules adsorbed under a given set of conditions. This assumption would then follow for lower temperatures. More data over a wider range of pressures and compositions at these temperatures would be helpful in confirming this theory. Another correlating method based on the work of Berenyi and Polanyi is that advanced by Dubinin and co-workers. This method has been previously discussed in some detail. Although this method appears to be fundamentally sound, no correlation was possible using Dubinints methods of evaluating adsorbed phase density as a function of the system temperature, normal boiling point density and van der Waals' density. Adsorption of Components from a Mixture Methods for predicting the adsorption of one component from a mixture have not received as much attention as pure component adsorption. The basic work on this area is that of Markham and Benton, which has been previously discussed, and which was examined by Schay. A correlation by these methods proved quite successful. The constants for the equation aN 2PN2 NN2= (23a)

and NCH4 =CH4PCH4 (23b) 1 + bN2PN2 + bCH4PC are plotted as a function of temperature in Figure 18. It is of some interest that the "a"t and "b" constants take on different values from those found in pure component adsorption. This indicates, in conformity with the ideas advanced by Schay, that the adsorbate is not an ideal solution and that certain interactions take place in the adsorbed phase. The constants were obtained using the standard least squares technique and represent the data within plus or minus seven per cent. No simple interaction coefficients, as suggested by Schay were obtained. A correlation to predict individual component adsorption from a mixture using a Freundlich type relationship log N1 = log K + 1/n1 log ylP + 1/n2 log Yl (34) was attempted. Although at given temperatures a fair degree of correlation could be obtained, it was not possible to obtain a good correlation of the coefficients as a function of temperature. This may in part be due to the lack of data at intermediate temperatures and a better picture might be obtained with a greater amount of data below the critical temperature. However, the nature of the equation makes it difficult to deduce any fundamental relationship between pure component coefficients and mixed coefficients making it difficult to predict the behavior of the equation constants when mixtures are considered. From the observation of Lewis and co-workers and the theoretical proof of Schay that Ni 1~~~~~~~~~~~~~~(5

0, MILLIGRAMMOLES/GM. ATM. OR b, I/ATM. OI) NU. d, Oo 0)000~ n> 4 0)00o N> OOD0 Nj 4 OD ) i Ou 0 OD 0 - ZQ_ IT. C) a 0 Zc2 z 0 l0 00 F-I~~~~~~~~~~~I o' o~ o-~ oa ob o,, rn o Z Z_ ): (D~~~~~~~~h C ~ ~ND N-, N -I I.I -I., H Ul 4 4 0 00 0 CD 0 0 0o O O O K K Kx K c O ob ob Ob o~~~06 06 6 o p O. 0 00, (D I/ IT _ -o~~~~~~~~~~~~~ _/ a~~~~~~~~~ 0; o u 0 0 / 2 I I OD N a 0 CD4 0, 0I C ~~~D ~~~~~~~-'~ ~ ~ ~ ~ 4 4 4 -4 - -4 (D F-/ HQ M 00 0 - H) rn o o l-, (D~~~~~~~~~~~~~~~~ z F-b o r 0 - 0 -Ol O /,:EJ 0~~~~~~~~~~~~~~~~~~~~~~~~~-0 I / 0, 0 ct~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ t t 0 (JW Ul 0

OW75 for ideal mixtures, a plot of NN2,/N2 as a function of NC4/N6e4 should result in a straight line tbrough the values 0, 1 and 1, 0 If the mixture is non-ideal then the data will lie on one side of the line, which is the case heres as seen in Figure 19o The relative volatility, as shown in Figure 20, is independent of composition0 At the higher pressures above the critical temperature, pressure appears to have little or no effect on the relative volatility* Below the critical point of the mixture, however, as is indicated by the values at 123 0K, there is a marked increase in relative volatility with increased pressure~ The relative volatility, together with the total loading, will yeld the equilibrium loading from a mixture0 Computing Adsorption Loading from the Correlations It is possible to calculate the amount of each constituent adsorbed from a mixture knowing the temperature, the total pressure and the composition of the'equilibrium gaso As outlined in the sample calculation below, the total amount adsorbed from a mixture can be determined by the Freundlich equation with the appropriate constants, as found in Figure 120 The relative volatility, obtained from Figure 2Q3, is independent of composition over the range of pressures of interest, From the gas phase data and the values of the relative volatility and total loading, the amount of each component may now be qomputedo These above calculations may be checked by determining the amount of p4ure component adsorbed at the system temperature and a pressure equal to its partial pressure in the mixture0 These values, together with

-76I.0 0.9 28 0.8 0.7 G z 0.5 - 0.4Q3 0.2 0.1 0 0.1 0.2 Q3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 ( N /N')CHY Figure 19. Total Adsorption Correlation on Molecular Sieve 5A.

8 81 A 1 1 11 1 11 1 - - 0 4_ r _I muI 1 ~ TI ~ Id t 1750K - 0 195 0K w ~ —'X~ - 0 5 10 15 20 25 30 55 40 45 50 55 60 65 70 75 80 85 90 PRESSURE, ATM. Figu-e 20. Relative Volatility as a Function of Pressure. (O= (YNw/XN )/(YC /xCH4/))) (a= 2Y 2/X2 [(C4XC4)

the amount of component adsorbed, may be used to evaluate the dimensionless factor N/NI. The values of this dimensionless factor for methane and nitrogen should satisfy the restriction placed upon them in Figure 19, a plot of the sum of the values of (N/N') for methane and nitrogen. The individual component Langmuir equations will also give individual component capacities from a mixture. Sample Calculation of Binary Adsorption P - 15 atmo YCH= 0o.6 PCH4 9 atm. PN2 = 6 atm. From Figure 20. a = 2.0 = (o.4/XN2)/(0.6/(1-XN2 ) then XN = 0.25 From Figure 11. NT = 2.2 milligram moles/gm. then NN2 = 2.2 x 0.25 = 0.55 milligram moles/gm, and NCH4 = 1.65 milligram moles/gm. This can now be checked by using Figure 19.

From Figure 12. NN2 1.35 milligram moles/gm at 6 atm. NCH4 = 2.6 milligram moles/gm at 9 atm. therefore (N/NI) N2 = 41 (N/Nt) = H o.64 and (N/N')N2 + (N/N )CH4 satisfieS the relationship illustrated in Figure 12. The Langmuir relationships predict NN2 = 0.59 milligram moles/gm. NCH4 171 milligram moles/gm. One or more points were checked at each temperature using both methods and they yielded results that were good to plus or minus ten per cent as indicated by the thre:e'?exles tabulatedtbeeow. TABLE XVII ACTUAL AND PREDICTED BINARY ADSORPTION Predicted Relative Volatility Langmuir Experimental Method Equation P T NCH4 NN2 Run atm, ~K YCH4 mgm. moles/gm. CH NCH4 NN2 16 14.29 295 0.590 1.74 0.62 1.65 0.55 1.71 0.59 64 20.48 195 0o.121 1.18 3.05 1.30 3.05 1.1~ 3 oD 81 11.80 123 0.050 1 07 3.96 1.12 3,98 0.72 4o 02

-80At the lowest temperature, 123~K, the Langmuir equation, with the predicted constants, gives low values for the amount of methane adsorbed. Data over a wider range of pressure and composition at this temperature should be evaluated before any definite conclusions regarding these predicted values are made. An Adsorption Model Many authors, as discussed previously, have suggested specific models for adsorption as a basis for theoretical development of various correlating equations. It would appear, however, that in light of the present knowledge and development, a modelistic approach based on data in the mid-pressure range is not justified. This is especially true since either the Langmuir or Freundlich relation can be developed from any of a number of models. The Langmuir equation suggests a mono-layer. In the present studies, the maximum amount adsorbed would cover approximately 95 per cent of the surface area (as measured by the "B E T" method). Many authors, among them Joyner(31) and Brunauer,(l4) suggest that a second layer begins to form when the first layer is 35 per cent fulled. In the case of definite porous structures of the type under study here, these assumptions are of little significance. Until a more detailed sub-microscopic study of the adsorption phenomena on sieve type adsorbents can be made, it appears unwise to draw any conclusions as to the method by which the adsorption proceeds.

CONCLUSIONS AND RECOMMENDATIONS The present study was undertaken to obtain adsorption data for the binary system methane-nitrogen on Molecular Sieves. For pure component adsorption, both the Langmuir and Freundlich equations will predict the adsorbent loading well, to the saturation pressure below the critical temnperature and to about 90 atmospheres above the critical point, over a wide range of temperature. A comparison of the resuits with limited data previously published by Lewis indicates that the Molecular Sieve has a greater capacity for methane than do either Davison Silica Gel or Columbia G Activated Carbon. Limited data obtained by von Antropff on an activated charcoal indicate that the capacity of the carbon for nitrogen is slightly higher than the capacity of the Molecular Sieve. A Polanyi type correlation, as those suggested by Dubinin or Lewis and co-workers, would be very useful from a theoretical as well as practical point. Unfortunately, the data did not correlate well using these methods. This appears to be due to the lack of a satisfactory method of predicting the density of the adsorbed state. Therefore, this method of correlation must wait for more fundamental studies than the one attempted here For mixtures, for practical purposes, the use of the total loading together with the relative volatility will give sufficiently accurate results. Somewhat more accurate results may be obtained using the individual Langmui.r isotherms with proper coefficients as proposed in the section on discussion of results. A study of the constants supports the ideas advanced by Schay regarding interaction of the components in the adsorbed state which -81

-82changes the values of the isotherm constants from those of the pure component. No simple interaction coefficients as those suggested by Schay could be obtained from a study of the present data. This suggests that more complex models must be assumed or that the interaction is more complex than can be analytically treated at this time. Much fundamental work remains to be done here before a general method of predicting adsorbate loading from mixtures from pure component adsorption data can be advanced. The present study indicates that there is equal competition for adsorption surface since the total amount adsorbed is independent of the equilibrium gas phase composition. It is quite possible that there is some interference between the molecular species. This future.:wOrk musi includee ~a better rrelationship Lfof the phase behavior of mixtures in both the gas and liquid state, a microscopic study of the adsorbate-adsorbent interface, and the phase behavior of the adsorbate in its adsorbed state. In the present study at low temperatures, the composition range was severely limited as it was difficult to obtain good equilibrium data below one atmosphere due to limitations imposed by the equipment. The study of this system at low pressures and over the whole composition range is recommended. From the present study it would appear that at the lower temperatures adsorption does not enjoy any advantage as a separative method for nitrogen from methane over vapor-liquid contacting since the relative volatilities of both methods are about equal and the adsorbent must be

regenerated constantly. At higher temperatures, or for final scrubbing of a stream of nitrogen, and met.a.ne, there may be applications where Molecular Sieve type ad6sorbents will prove to be uite useful0

APPENDIX A EXPERIMENTAL DATA The basic experimental data was converted to absolute pressure and temperature before being used in computations. The mass spectrometer analyses of the samples were reduced to mole fractions. These data, together with the applicable volumes, were then coded on cards which were read by the computer program and served as the basis for computation. The experimental data as reduced for the computer program may be found in Table XIX. In order to amplify the listing, a sample set of data with the necessary dimensions added may be found in Table XVIII. Each line corresponds to a data card and at the right of the line or card will be found an identification "Run 16". It will be noted that the first line of each series contains only a series of digits on the left. These identify the data set or run to the program as well as the number of samples taken plus desorptions made. The next two data cards or lines of data contain the information regarding the condition of the feed cell before (PI, TI, VI) and after (PF, TF, VF) loading as well as the composition of the gas in the cell (YI(CH4), YI(N2)). The cards following either refer to samples taken at equilibrium or to desorptiono They may easily be distinguished as the equilibrium samples will have a three-digit number (i.e., 002) preceeding the pertinent data, while the desorption data are preceeded by a two digit number preceeded by a minus sign. In addition, no cell pressure (PC) or cell temperature (TC) is recorded,for desorption. The data cards containing equilibrium data, contain the following information in order from left to right: 1) a three digit number identifying the sample, -84

2) the cell pressure'(;PC) in psia, 3) the cell temperature (TC) in ~K, 4) the pressure of the sample (PS) in psia, 5) the temperature of the sample (TS) in ~K, 6) the sample volume (VS) in cc, 7) the mole fraction methane (YC(CH4)) in the sample, 8) the mole fraction nitrogen (YC(N2)) in the sample. In the data cards which contain desorption information, as previously stated, no cell data will be found. The pressure, temperature and volume of the desorbate, followed by the composition where it was possible to obtain a sample or an assumed value in case no experimental value was available, are recorded. It will be noted that a number of runs were eliminated from considerationo Several of these were eliminated because the amount desorbed indicated a leak in the system. One run was not used because the liquefaction pressure had been exceeded. In addition, two pure component runs were not used because they did not agree with the remainder of the data as well as the check runs made to check these values. Runs 116 and 117 were blank runs made to check the adsorptive capacity of the empty cello

TABLE XVIII SAMPTLE OF EXPERIMENTAL DATA INPUT FOR COMPUTER Run No, No. of Data Pts. o16 004 Run 16 YI YI PI TI PF TF CH4 2 415.0 psia 297.0 ~K 317.0 psia 297.0 ~K 0.7062 0.2938 Run 16 VI VF 948.8cc 959.9cc, 89.72cc, PC TC PS TS Vs YCCR YCN2 001 210.0psia 292.75 ~K 29.40 psia 295.2 ~K 23.19cc 0,6612 0.3388 Run 16 002 210.0 292.75 29.4 296.3 23.79 0.5903 0.4097 Run 16 - 03 7.85 294.1 11771.0 1.0 0.0 Run 16 - 04 1.82 293,0 179.6 1.0 0.0 Run 16

TABLE XIX EXPERIMENTAL DATA (See Table XVIII for Explanation of Format) -87

-88013016 RUN 13 1845.0 296.0 1615.0 296.0 0.7052 0.2935 RUN 13 948.8 959.9 89.72 01u 1320.0 293.25 147.0 293.3 23.79 0.6240 0.3580 RUN 13 002 1320.0 293.25 147.0 298.0 23.79 0.6362 0.3638 RUN 13 003 1220.0 293.25 147,U 299.6 23.79 0.6396 0.3604 RUN 13 -04 855.0 299.7 36.89 0.6339 0.3661 RUN 13 005 880,0 295.25 100O. 293.9 23.79 0.6445 0.3555 RUN 13 006 840.0 295.0 45.0 298.7 23.79 0.6360 0.3640 RUN 13 -07 6u5,0 298.7 36.89 0.6392 U.3608 RUN 13 008 640,0 295.15 68.0 299.7 23.79 0.6306 0.3694 RUN 13 009 620,U 295.75 12.30 299.7 23.79 0.6412 0.3588 RUN 13 -10 445.U 298.7 36.89 0.4798 0.5202 RUN 13 011 455,0 292.75 50.7 295.2 23.79 0.6464 0.3536 RUN 13 -12 325,v 292.25 36.89 0.3780 0.9122 RUN 13 013 190. 293.25 36*7 296.9 23.79 0.6537 0.3463 RUN 13 -14 7.91 295.2 11770.0 0.7226 0.2774 RUN 13 -15 1.86 295.2 11770.0 1.0 0.0 RUN 13 -16 16.84 295.2 180.2 1.0 0.0 RUN 13 014005 RUN 14 1550.0 293.75 1340.0 293.75 0.7050 0.2950 RUN 14 948.8 959.9 89.72 001 1120.0 293.25 112.0 293.u 23.79 0.6387 0.3610 RUN i4 002 1100-0 294.5 117,0 295.4 23.79 0.6330 0.3670 RUN 14 -03 14.25 294.1 11770.0 0.67u0 u.3300 RUN 14 -04 3.11 293.0 11770.0 0.8377 u.1623 RUN 14 -05 14.59 295.2 179.6 1.0 0.0 RUN 14 015005 RUN 15 700.0 295.5 575*0 295.5 0.7051 0.2949 RUN 15 948.8 959.9 89.72 001 420.0 293.25 45.0 297.4 23.79 0.5912 0.4048 RUN 15 002 410.0 293.25 37.0 297.4 23.79 0.5443 0.4556 RUN 15 -U3 8.45 296.3 11774.0 0.6666 0.3334 RUN 15 -04 2.05 296.3 11770.0 1.0 0.0 RUN 15 -05 8.19 296.3 179.6 1.0 0.0 RUN 15 016004 RUN 16 415.0 297.0 317.0 297.0 0.7062 0.2938 RUN 16 948.8 959.9 89.72 001 210.0 292.75 29.40 295.2 23.79 0,6612 0.3388 RUN 16 002 210.0 292.75 29.4 296.3 23.79 0.5903 0.4097 RUN 16 -03 7.85 294.1 11771.0 1.0 0.0 RUN 16'04 1.82 293.0 179.6 1.0 0.0 RUN 16 017011 RUN 17 1890.0 295.75 1635,0 296.25 0.2145 0.7855 RUN 17 948.8 959.9 89.72 001 1335.0 293.25 83.0 296.3 23.79 0.1608 0.8392 RUN 17 002 1335.0 293.25 145.0 297.4 23.79 0.1610 0.8390 RUN 17 -03 879.0 297.4 36.89 0.1624 0.8376 RUN 17 004 910,0 293.25 100.0 296.3 23.79 0.1610 0.8390 RUN 17 005 880,0 293.0 100o0 296.3 23.79 0.1680 0.8320 RUN 17 -06 653.U 296.3 36.89 0.2586 0.7414 RUN 17 007 640.U 294.0 71.0 296.3 23.79 0.1661 0.8339 RUN 17 008 620.0 294.0 60.u 296.3 23.79 0.1752 0.8248 RUN 17 -09 10.34 295.2 11775.0 0.2237 0.7763 RUN 17 -10 1.7 296.3 11775.0 1.0 0.0 RUN 17 -11 1.62 295.2 179.6 1.0 0.0 RUN 17 018006 RUN 18 1140.0 297.0 970.0 297.0 0.2160 0.7840 RUN 18 948.8 959.9 89.72 001 695,0 296.25 79.0 298.6 23.79 0.1548 0.8452 RUN 18 002 670., 296.25 65.0 298.6 23.79 0.1526 0.8473 RUN 18 003 660,0 294.25 76.0 298.6 23.79 0.1595 0.8404 RUN i8 -04 10.79 295.2 11775.0 0.1995 0.8005 RUN 18 -05 1.6 295.2 11770.0 1.0 0,0 RUN 18 -06 3.74 295.2 179.6 1.0 0.0 RUN 18 019004 RUN 19 570.0 296.0 467.0 296.0 0.2218 0.7782 RUN 19 948.8 959.9 89.72 001 335.0 294.25 36.0 300.2 23.79 0.1459 0.8541 RUN 19 002 335.0 294.25 35.0 300.2 23.79 0.1459 0.8541 RUN 19 003 335.0 295.25 35.0 299.7 23.79 0.1456 0.8544 RUN 19 -04 8.005 296.3 11776.0 0.1983- 0.8016 RUN 19 020005 RUN 20 01530.0 0295.5 01325.0 0295.75 0.7722 0.2278 RUN 20 948.8 959.9 89.72 001 01100.0 296.0 127.0 300.8 23.79 0.7051 0.2948 RUN 20 002 1075.0 296.0 125.0 300.8 23.79 0.7038 0.2962 RUN 20 -03 13.42 298.0 11777.8 0.7473 0.2527 RUN 20 -04 1.96 296.4 11769.7 1.0 0.0 RUN 20 -05 0.11 296.4 179.7 1.0 0.O RUN 20 021005 RUN 21 01300.0 298.0 1128.u 298.0 0.7715 0.2285 RUN 21 948.8 959.9 89.72 001 845,0 295.5 98.0 297.5 23.79 0.6983 0.3016 RUN 21 002 825.0 293.j 83.0 297.5 23.79 0.7003 0.2997 RUN 21 -03 13.323 295.8 11776.7 0.7405 0.2594 RUN 21 -04 1.742 296.3 11769.8 1.0 0.0 RUN 21'05 0.319 296.3 179.7 1.0 0.0 RUN 21

-89 022005 RUN 22 857.0 296.5 722.0 296.5 0.7631 0.2369 RUN 22 948.8 959.9 89.72 001500.0 293.75 25.0 296.9 49.59 0.6747 0,3253 RUN 22 002500.0 294.25 55.0 299.1 23.79 0.6627 0.3373 RUN 22 -03 10.461 297.4 11775.1 0.7329 0.2671 RUN 22 -04 1.431 296.6 11769.8 1.0 0.0 RUN 22 -05 0.561 297.0 179.7 1.0 0.0 RUN 22 023004 RUN 23 705.0 295.5 595.0 295.5 0.7679 0.2321 RUN 23 948.8 959.9 89.72 O1l 400,0 293.75 45.0 298.6 23.79 0.6675 0.3325 RUN 23 002 360,0 293.75 45.0 298.6 23.79 0.6670 0.3330 RUN 23 -03 9.34 296.3 11774.6 0.7335 0.2665 RUN 23 -04 1.189 296.3 11769.9 1.0 0.0 RUN 23 024005 RUN 24 425.0 298.25 337.0 298.25 0.7671 0.2329 RUN 24 948.8 959.9 89.72 001 235.0 296*5 21.7 300.7 23.79 0.6512 0.3488 RUN 24 002 225.0 296.5 20.7 300.7 23.79 0.6561 0.3439 RUN 24 -03 7.135 298.6 11773.4 0.7280 0.2720 RUN 24 -04 0.880 298.6 11769.0 1.0 0.0 RUN 24 — 05 0.348 300.0 179.7 1.0 0.0 RUN 24 025003 RUN 25 1130.0 299.0 975.0 299.5 0.9895 0.0105 RUN 25 948.8 959.9 89.72 001 700.0 293.25 4.05 296.3 11772.2 0O9830 0.0170 RUN 25 002 330.0 293.25 36.7 30U.7 23.79 0.9891 U.0109 RUN 25 -03 10.036 298.6 11771.2 1.0 0.0 RUN 25 026002 RUN 26 917.0 298.5 775.0 298.5 0.9952 0.0047 RUN 26 948.8 959.9 89.72 001 590.0 297.5 1.0 0,0 RUN 26 -02 13.13 299.7 11771.6 1.0 0.0 RUN 26 027003 RUN 27 770.0 297.75 640.0 297.75 1.0 0.0 RUN 27 948.8 959.9 89.72 001 440,0 296.25 1.0 0.0 RUN 27 002 450,0 296.0 1.0 0.0 RUN 27 -03 11.776 298.6 11771.2 1.0 0.0 RUN 27 028004 RUN 28 600.0 297.0 477.0 297.0 1.0 0.0 RUN 28 948.8 959.9 89,72 001 380.0 298.25 1.0 0.0 RUN 28 002 380,0 296.25 1.0 0.0 RUN 28 003 380.0 295.75 1.0 0.0 RUN 28 -04 10.887 299.1 11771 1.0 0.0 RUN 28 029005 RUN 29 1895.0 297.75 1630.0 298.0 0.0013 0.9985 RUN 29 948.8 959.9 89.72 001 1255.0 293.25 0.0 1.0 RUN 29 002 1270.0 294.75 0.0 1.0 RUN 29 -03 915.0 299.7 36.89 0.0 1.0 RUN 29 004 910,0 295.0 0.0 1.0 RUN 29 -05 14.738 296.9 11772.0 0.0 1.0 RUN 29 029005 RUN 29A 1890.0 297.75 1632.0 298.0 0.0 1.0 RUN 29A 948.8 959.9 89.72 001 1255.0 293.25 0.0 1.0 RUN 29A 002 1270.0 294.75 0.0 1.0 RUN 29A -03 3.1519 298. 11771.2 0.0 1.0 RUN 29A 004 910.0 295.0 0.0 1.0 RUN 29A -05 14.738 296.9 11772.0 0.0 1.0 RUN 29A 030005 RUN 30 1380.0 298.5 1180.0 298.5 0.0 1.0 RUN 30 948.8 959.9 89.72 001 925.0 295.5 0.0 1.0 RUN 30 002 925.0 296.25 0.0 1.0 RUN 30 003 940.0 296.25 0.0 1.0 RUN 30 004 940 296.25 0.0 1.0 RUN 30 -05 14.835 297.0 11770.4 0.0 1.0 RUN 30

-90031004 RUN 31 758.0 295.75 627.0 295.75 0.0 1.0 RUN 31 948.8 959.9 89.72 001 505.0 299.5 0.0 1.0 RUN 31 002 490.U 293.15 0.0 1.0 RUN 31 003 495,0 295.25 0.0 1.0 RUN 31 -04 10.442 295.8 11771.0 0.0 0.0 RUN 31 032002 500.0 295.2 402.0 295.2 0.0 1.0 RUN 32 948.8 959.9 89.72 001 317.0 295.5 0.0 1.0 RUN 32 -02 7.87 296.9 11770.8 0.0 1.0 RUN 32 033002 RUN 33 397*0 296*2 312.0 296.25 0.0 1.0 RUN 33 948.8 959.9 89.7 RUN 33 001 230.0 296.0 10.0 297.6 0.0 0.0 1.0 RUN 33 -02 6.64 297.6 11769.9 0.0 1.0 RUN 33 034014 RUN 34 1705.00 300.5 1310.0 300.5 0.2291 0.7709 RUN 34 948.8 959.9 89.72 001 1235.0 194.75 160.0 305.2 23.79 0.1503 0.8497 RUN 34 002 1200.0 194.75 157.u 305.2 23.79 0.1530 0.8470 RUN 34 -03 880.0 305.2 36.89 0.2318 0.7682 RUN 34 004 1070.0 197.5 127.0 302.4 23.79 0.1464 0.8536 RUN 34 -u5 14.10 301.9 11777.1 0.1478 0.8522 RUN 34 -06 3.71 301.9 11771.6 0.2832 0.7168 RUN 34 -07 2.01 301.9 11770.2 100 0.0 RUN 34 -08 1.02 301.9 11769.6 1.0 0.0 RUN 34 -09 0.657 300.8 11769.4 1.0 0.0 RUN 34 -10 0.425 300.8 11769.2 1.0 0.0 RUN 34 -11 0o193 300.8 11769.0 1.0 0.0 RUN 34 -12 0.097 300.8 11769.0 1.0 0.0 RUN 34 -13 0.116 300.8 11769.0 1.0 0.0 RUN 34 -14 0.029 300.8 11769.0 1.0 0.0 RUN 34 035005 RUN 35 1002.0 300.75 733.00 300.75 0.2299 0.7701 RUN 35 948.8 959.9 89.72 001 685.0 194.75 88.0 299.1 23.79 0.1205 0.8795 RUN 35 002 670.0 194.75 86.0 299.1 23.79 0.1149 0.8851 RUN 35 -03 10.364 296.3 11775.0 0.1276 0.8724 RUN 35 -04 2.456 296.3 11770.8 1.0 0.0 RUN 35 -05 4.734 296.5 1176-.0 1.0 0.0 RUN 35 036009 RUN 36 760.0 305.0 538.0 305.0 0.2302 0.7618 RUN 36 948.8 959.9 89.72 001 425.0 191.33 50.0 306.3 23.79 0.1101 0.8899 RUN 36 002 412.0 191.33 50.0 306.3 23.79 0.1137 0.8863 RUN 36 -03 7.909 303.U 11773.8 0.1230 0.8770 RUN 36 -04 3.171 296.9 11771.2 1.0 0.0 RUN 36 -05 2.688 301.9 11771.0 1.0 0.0 RUN 36 -06 0.812 301.9 11770.0 1.0 0.0 RUN 36 -07 0.348 302.4 11770.0 1.0 0.0 RUN 36 -08 0.048 302.4 11770.0 1.0 0.0 RUN 36 -09 0.006 303.0 11770.0 1.0 0.0 RUN 36 037006 RUN 37 568.0 309.75 383.0 309.75 0.2480 0.7520 RUN 37 948.8 959.9 89.72 001 242.0 193.o7 25.0 303.5 23.79 0.1182 0.8818 RUN 37 002 240.0 193.67 25.0 303.5 23.79 0.1124 0.8876 RUN 37 -03 5.105 301.9 11772.3 0.1150 0.8850 RUN 37 -04 5.443 305.2 11772.6 0.2284 0.7716 RUN 37 -05 1.682 304.1 11770.5 1*0 0.0 RUN 37 -06 0.288 303.0 1176-.4 1.0 0.0 RUN 37 039006 RUN39 970.00 302.7 703.0 302.7 0.7159 0.2841 948.8 959.9 89.72 001 700.0 192.45 89.0 301.0 23.79 0.6058 0.3942 RUN39 -02 10.055 298.8 11774.9 0.5996 0.4004 RUN39 -03 3.016 297.1 11771.2 0.7094 0.2906 RUN39 -04 4.467 299.9 11772.0 0.8382 U.1018 RUN39 -05 2.204 299.9 11770.8 0*9175 0.0824 RUN39 -06 0.197 299.9 11770.0 1.0 0.0 RUN39 040004 RUN40 727.0 303.95 520.0 303.95 0.7256 0.2744 RUN 40 948.8 959.9 89.72 001 420.0 192.2 51.0 298.8 23.79 0.5558 0.4442 RUN40 -02 6.932 296.5 11773.2 0.5635 0.4365 RUN40 -03 7.986 296.5 11773.8 0.7896 0.2104 RUN40 -04 1.241 297.6 11770.0 1.0 0.0 RUN40 041007 RUN41 512.0 305.2 370.0 305.2 0.7101 0.2899 RUN41 948.8 959.9 89.72 001 167.0 191.9 20.0 302.1 23.79 0.5030 0.4970 RUN41 002 167,0 191.9 20.0 302.1 23.79 0.5010 0.4990 RUN41 -03 4.215 298.8 11772.0 0.5380 0.4620 RUN41 -04 4.796 298.8 11772.1 0.7292 0.2708 RUN41 -05 2.651 301.0 11771.1 1.0 0.0 RUN41 -06 0.657 301.0 11770.0 1.0 0,0 RUN41 -07 0.135 301.0 11770.0 1.0 0.0 RUN41

-91042003 RUN42 1810.0 305.7 1398.0 305.7 0.5093 0.4907 RUN42 948.8 959.9 89.72 001 580,0 194.45 79.U 296.5 23.79 0.4370 0.5629 RUN42 002 565.0 194.45 77.0 296.5 23.79 0.4369 0U.5531 RUN42 -03 16.881 297.6 11774.7 0O44i7 0.5573 RUN42 043005 RUN43 1295.0 301.25 1055.0 3v1.25 0.5759 0.4241 RUN43 948.8 959.9 89.72 001 550.0 191.9 72.0 297.1 23.79 0.4048 0.5952 RUN43 002 540,0 191.9 70.U 297.1 23.79 0.4054 0,5946 RUN43 -03 9.668 298.2 11774.3 0.4176 0.5824 RUN43 -04 7.696 298.8 11773.2 0.7011 0.2989 RUN43 -u5 1.025 297.6 11770.8 1.0 0.0 RUN43 044004 RUN44 948.00 302.75 772.0 302.75 0.6360 0.3640 RUN44 948.8 959.9 89.72 001 240,0 194.45 22.0 299.9 23.79 0.4059 0.5941 RUN44 -02 7.429 296.0 11772.8 0.4866 0.5134 RUN44 -03 5.279 296.0 11772.0 0.7479 0.2521 RUN44 -04 0.812 295.4 11769.6 1.G 0.0 RUN44 045006 RUN45 1867.0 301.5 1455.0 301.5 0.5499 0.4511 RUN45 948.8 959.9 89.72 001 1130.0 192.5 160.0 297.1 23.79 0.4764 0.5236 RUN45 -02 19.530 296.5 11779.6 0.4494 0.5506 RUN45 -03 8.160 296.5 11773.6 0.6571 0.3429 RUN45 -04 2.340 296.5 11770.5 0.75 0.25 RUN45 -05 0.367 297.6 11770.0 1.0 0.0 RUN45 -06 0.048 297.6 11770.0 1.0 0.0 RUN45 046005 RUN46 915.0 302.75 673.0 298.25 0.5466 0.4534 RUN46 948.8 959.9 89.72 O01 540,0 194.95 70.0 295.4 23.79 0.3887 0.6113 RUN46 -02 9.63G 296.5 11774.3 0.3954 0.6046 RUN46 -03 6.652 296.5 11772.7 0.6537 0.3463 RUN46 -04 2.224 296.5 11770.0 0.75 0.25 RUN46 -05 0.155 296.5 11769.0 1.0 0.0 RUN46 048005 RUN48 577.0 298.25 402.0 298.25 0.6512 0.3478 RUN48 948.8 959.9 89.72 001 267.0 192.5 35.0 298.8 23.79 0.4512 0,5488 RUN48 -J2 10.538 299.9 11774.8 0.5479 0,4521 RUN48 -03 4.041 299.3 11771.3 0.8133 0.1869 RUN48 -04 1.006 299.9 11769.7 1.0 0.0 RUN48 -05 0.048 299.9 11769.0 1.0 0,0 RUN48 049005 RUN49 1593,0 298 t15 1240.0 298.75 0.0007 0.9993 RUN49 948.8 959.9 89.72 001 1145.0 192.9 67.0 298.8 49.6 0*0007 0.9993 RUN49 -02 15.237 301.5 11777.2 00009 0,.9991 RUN49 -03 9.978 299.9 11774.4 0.0018 0.9982 RUN49 -04 0.309 299.9 11769.3 0.0 1.0 RUN49 -05 0.010 299.9 11769.0 0.0 1,0 RUN49 050004 RUN50 1213.0 305.0 940.0 305.0 0.0006 0.9994 RUN50 948,8 959.9 89.72 001 850,0 193.9 99.0 300.9 23.79 0.0016 0.9984 RUN50 -02 15.798 299.9 11779.6 0.0013 0,9987 RUN50 -03 4.738 300.4 11771.8 O0G 1.0 RUN5U -04 0.512 299.9 11769.5 0.0 1.0 RUN5O 051003 RUN51 825.0 298.0 600.0 298.0 0.0009 0.9991 RUN51 948.8 959.9 89.72 001 470,0 194.7 12.820 299.3 11775.9 0.0009 0.9991 RUN51 -02 4.670 298.8 11771.6 0.0 1.0 RUN51 -03 0.058 298.8 11769.0 0.0 1.0 RUN51 052004 RUN52 750.4 298.2 541.9 298.2 0.9920 0.0080 RUN52 961.1 972.2 89.72 001 436.4 195.9 60.0 302.0 23.79 0.9897 0.0103 RUN52 -02 7.677 299.8 11783.0 0.9842 0.0158 RUN52 -U3 10.364 298.7 11758.7 0*9940 0U0060 RUN52 - 04 0u590 298.7 11733.5 1.0 0.0 RUN52 053004 RUN53 508.0 298.6 345.0 298.65 0.9900 0.0100 RUN53 961*1 972.2 89.72 001 133,46 195.6 20.0 301.5 23.79 0.9860 0.0140 RUN53 -02 4.892 299.8 11786.1 0.9884 0.0116 RUN53 -03 9.127 300.4 11758.0 1.0 0.0 RUN53 -04 0.300 300.9 11753.0 1.0 0.0 RUN53 056005 RUN56 687.0 296.9 493.5 296.9 0.9904 0.0096 RUN56 961.1 972.2 89.72 001 365.0 193.8 47.0 299.3 23.79 0.9883 0.0118 RUN56 -02 6.710 299.3 11782.5 0.9862 0.0138 RUN56 -03 10.055 299.3 11783.0 1.G 0,0 RUN56 -04 1.006 299.8 11779.0 1.0 0.0 RUN56 -05 0.029 299.8 11779.0 1.0 0.0 RUN56

-92057004 RUN57 470.0 297.4 333.0 297.4 0.9887 0.0113 RUN57 961.1 972.2 89.72 001 52.0 195.5 20.0 299.8 23.79 0.9814 0.0186 RUN57 -02 11.254 299.8 11784.9 0.9919 0.0081 RUN57 -03 0.986 298.7 11779.0 1.0 0.0 RUN57 -U4 0.029 298.7 11779.0 1.0 0.0 RUN57 058006 RUN58 299.0 296.2 172.8 296.2 0.9895 0.0105 RuN58 961.1 972.2 89.72 -01 337.3 296.2 -972.2 0.9914 0.0086 RUN58 -02 296.9 296.2 972.2 0.9914 0.0086 RUN58 003 193,0 195.o 25.0 300.4 23.79 0.9849 0.0151 RUN56 -04 11.873 298.7 11780.0 0.9894 0.0106 RUN58 -05 2.881 299.3 11780.0 1.0 0.0 RUN58 -06 0.211 299.3 11780.0 1.0 0.0 RUN58 059003 RUN59 236.4 298.4 102.2 298.4 0.9797 0.0203 RUN59 961.1 972.2 89.72 0U1 40.0 194.9 40.0 300U.9 23.79 0.9830 0.0170 RUN59 -02 10.287 300.4 11780.0 0.9918 0.0082 RUN59 -03 0.85U 3uu.4 11780.0 1.0 0.0 RUN59 060005 RUN60 770.5 298.16 556.0 798.16 0.5241 0.4759 RUN6o 961.1 972.2 89.72 001 485.0 195.8 60.0 298.2 23.79 0.3517 0,6483 RUN60 — 02 7.793 299.3 11780.0 0.3572 0.6428 RUN60 -03 9.108 299.8 11780.0 0.500 0U.500 RUN60 -04 1.122 299.3 11780.0 1.0 0.0 RUN60 -05 0.072 299.3 11780.0 1.0 0.0 RUN60 060004 RUN6OA 537.9 295.8 369.6 296.2 0.5205 0.4795 RUN60A 961.1 972.2 89.72 Oul 178.0 194.4 20.0 298.7 23.79 0.5066 0.4934 RUN60A -02 13.014 299.8 11780.0 0.3060 0U.6940 RKUN60A -03 0.870 299.8 11780.0 1.0 0.0 RUN60A -04 0.012 299.8 11780.0 1.0 0.0 RUN60A 0610C4 RUN61 349.4 296.9 203.1 296.9 0.5208 0.4792 RUN61 961.1 972.2 89.72 Oul 95.6 195.5 13.26 297.0 23.79 0.3139 0.6861 RUN61 -02 00.887 293.2 11780.0 0.4959 0.5041 RUN61 -03 1.305 297.6 11780.0 0.500 0.500 RUN61 -04 0.0U64 297.6 11780.0 1.0 0.0 RUN61 062004 RUN62 193.0 296.2 66.8 296.2 0.5195 0.4805 RUN62 961.1 972.2 89.72 001 29.0 192.8 24.0 295.4 23.79 0.3344 0.6656 RUN62 -02 9.707 299.8 11780.0 0.5065 0.4935 RUN62 -03 0.831 299.3 11780.0 1.0 0.0 RUN62 -04 0.041 299.3 11780.0 1.0 0.0 RUN62 063005 RUN63 381.2 296.2 559.0 296.2 0.2411 0.7589 RUN63 961.1 972.2 89.72 001 550.0 200.9 65.0 295.4 23.79 0.1382 0.8618 RUN63 ~02 8.373 297.6 11780.0 0.1382 0.8618 RUN63 -03 8.412 298.7 11780.0 0.3078 0U.6922 RUN63 -04 0.928 298.7 11780.0 1.0 0.0 RUN63 -05 0.039 298.7 11780.0 1.0 0.0 RUN63 064005 RUN64 533.8 296.2 349.4 296.4 0.2405 0.7595 RUN64 961.1 972.2 89.72 001 301.0 192.8 37.0 294.3 23.79 0.1214 0.8786 RUN64 -02 12.105 298.7 11780.0 0.1962 0.8038 RUN64 -03 2.63U 298.7 11780.0 0.506 U.500 RUN64 -04 0.261 298.7 11780.0 1.0 0.0 RUN64 -05 0.014 298.7 11780.0 1.0 0.0 RUN64 065005 RUN65 336.2 296.9 186.0 296.9 0.2402 0.7598 RUN65 961.1 972.2 89.72 001 111,5 192.8 14.7 296.5 23.79 0.1261 0.8739 RUN65 002 106.0 192.8 14.7 296.5 23.79 0.1263 0.8737 RUN65 -03 11.505 299.8 11780.0 0.2274 0.7726 RUN65 -04 0.87U 298.7 11780.0 0.500 0.500 RUN65 -05 U.021 299.8 11780.0 1.0 0.0 RUN65 066004 RUN66 177.9 298.66 61.8 298.66 0.2395 0.7605 RUN66 961.1 972.2 89.72 001 20.3 190.5 8.35 298.2 23.79 0.1326 0.8674 RUN66 -02 8.798 301.5 11780.0 0.2282 0,.7718 RUN66 -03 0.657 301.5 11780.0 0.500 U.500 RUN66 -04 0.029 301.5 11780.0 1.0 0.0 RUN66 067005 RUN67 755.4 297.91 535.8 297.91 0.7100 0.2900 RUN67 961.1 972.2 89.72 001 458.5 193.83 60.0 294.3 23.79 0.5638 0.4362 RUN67 -02 8.257 299.8 11780.0 0.5740 0.4260 RUN67 -03 9.475 298.7 11780.0 0.8185 0.1815 RUN67 -04 0.503 298.7 11780.0 1.0 0.0 RUN67 -u5 0.029 298.7 11780,0 1.0 0.0 RUN67 068005 RUN68 515.7 298.16 323.2 298.16 0.7122 0.2878 RUN68 961.1 972.2 89.72 001 241.0 192.84 28.0 296.5 23.79 0.5312 0.4688 RUN68 -02 12.086 299.8 11780.0 0.6803 0.3196 RUN68 -03 2.765 299.8 11780.0 0.7500 0.2500 RUN68 -04 U.193 299.8 11780.0 1.00 0.0 RUN66 -05 0.014 299.8 11780.0 1.0 0.0 RUN68

-93069005 RUN69 311.0 300.66 152.0 300.66 0.7113 0.2887 RUN69 961.1 972.2 89.72 001 117., 192.84 16.U 299.8 23.79 0.5081 0.4919 RUN69 -02 11.119 300.9 11780.0 0.6802 0.3198 RUN69 -03 1.992 302.0 11780.0 0.7500 0.2500 RUN69 -04 0.213 302.0 11780.0 1.0 0.0 RUN69 -05 0.014 302.0 11780.0 1.0 0.0 RUN69 070004 RUN70 145.6 300.J1 26.5 300.91 0.7115 0.2885 RUN70 961.1 972.2 89.72 001 14.0 194.66 8.84 299.3 23.79 0.5415 0.4585 RUN70 -02 8.856 302.0 11780.0 0.6977 0.3023 RUN70 -03 1.006 301.5 11780.0 0.7500 0.2500 RUN70 -04 0.014 301.5 11780.0 1.0 U.0 RUN70 071004 RUN71 762.5 297.9 543.0 297.9 0.0 1.0 RUN71 961.1 972.2 89.72 001 492.0 192.16 56.0 297.6 23.79 0.0 1.0 RUN71 -U2 16.243 298.7 11780.0 0.0 1.0 RUN71 -03 1.102 299.8 11780.0 0.0 1.0 RUN71 -04 0.021 299.8 11780.0 0.0 1.0 RUN7i 072004 RUN72 520.4 300.28 339.3 300.28 0.0 1.0 RUN72 961.1 972.2 89.72 001 258.0 192.84 30.0 298.2 23.79 0.0 1.0 RUN72 -02 12.878 300.4 11780.0 0.0 1.0 RUN72 — 03 1.568 302.0 11780.0 0.0 1.0 RUN72 -04 0.073 302.0 11780.0 0.0 1.0 RUN72 073004 RUN73 323.2 300.0 175.2 300.0 0.0 1.0 RUN73 961.1 972.2 89.72 001 113.5 192.84 30.0 297.6 23.79 0.0 1.0 RUN73 -02 10.771 300.9 11780.0 0.0 1.0 RUN73 -03 1.431 302.0 11780.0 0.0 1.0 RUN73 -04 0.039 302.0 11780.0 0.0 1.0 RUN73 074004 RUN74 168.7 298.75 45.0 298.75 0.0 1.0 RUN 74 961.1 972.2 89.72 001 42.0 192.84 25.0 298.7 23.79 0.0 1.0 RUN74 -02 9.444 302.0 11780.0 0.0 1.0 RUN74 U03 0.870 302.0 11780.0 0.0 1.0 RUN74 -0a 0.019 302.0 11780.0 0.0 1.0 RUN74 075005 RUN 75 742.3 300.4 498.6 300.4 0.7260 0.2740 RUN 75 961.1 972.2 89.72 001 449.0 175.7 57.0 297.0 23.75 0.5791 0.4209 RUN 75 -02 12*105 298.7 11780.0 0.6278 0.3723 RUN 75 -03 6.226 298.7 11780.0 0.8760 0.1240 RUN 75 -04 1.044 298.7 11780.0 1.0 0.0 RUN 75 -05 0.07 298.7 11780.0 1.0 0.0 RUN 75 076004 RUN 76 328.2 300.4 155.7 300.4 0.7257 0.2743 RUN 76 961.1 972.2 89.72 001 111.7 174.87 15.7 297.6 23.79 0.4908 0.5092 RUN 76 -02 12.447 301.5 11780.0 0.6949 0.3051 RUN 76 -03 1.799 302.1 11780.0 1.0 0.0 RUN 76 -04 0.126 302.1 11780.0 1.0 0.0 RUN 76 077005 RUN 7 749.4 301.2 458.1 301.2 0.9990 0.0010 RUN 77 961.1 972.2 89.72 Oul 420.0 173.2 60.0 297.6 23.79 0.99E6 0.0014 RUN 77 -02 13.652 299.8 11780.0 0.9984 0.0016 RUN 77 -03 10.925 299.8 11780.0 1.0 0.0 RUN 77 — 04 1.856 300.4 11780.0 1.0 0.0 RUN 77 -05 0.102 300.9 11780.0 1.0 0.0 RUN 77 078004 RUN 78 584.2 301.2 366.5 301.2 0.9995 0.0005 RUN 78 961.1 972.2 89.72 001 331.0 175.2 48.0 299.8 23.79 0.9978 0.0022 RUN 78 -02 14.232 302.1 11780.0 0.9931 0.0069 RUN 78 -03 3.887 303.2 11780.0 1.0 0.0 RUN 78 -04 0.435 303.2 11780.0 1.0 0.0 RUN 78 079004 RUN 79 348.4 299.2 172.8 299.2 0.9931 0.0069 RUN 79 961.1 972.2 89.72 001 132.0 174.2 17.7 298.7 23.79 0.9883 0.0117 RUN 79 -U2 12.337 302.1 11780.0 0.9926 0.0072 RUN 79 -03 2.398 302.1 11780.0 1.0 0.0 RUN 79 -04 0.261 302.1 11780.0 1.0 0.0 RUN 79 080004 RUN80 165.8 300., 24.4 300.4 0.9924 0.0076 RUN80 961.1 986.9 89.72 001 7.4 173.2 7.4 173.2 7.37 298.7 23.79 09862 0.0138 RUN8 -02 10.345 302.6 11754.0 1.0 0.0 RUN8O -03 1.276 30U.9 11754.0 1.0 0.0 RUN80 -04 0.048 300.9 11754.0 1.0 0.0 RUN80 081006 RUN81 421.0 301.0 206.1 301.0 0.1779 0.8221 RUN81 961.1 972.2 89.72 -01 27.0 299.8 23.79 0.1766 0.8234 RUN81 002 173.4 123.2 24.0 300.9 23.79 0.0495 0.9505 RUN8 -03 24.0 300.9 2.3 0.0500 0.9500 RUN81 -04 13.748 302.6 11780.0 0.1348 0.8652 RUNBi -05 3.539 30u.9 11780.0 0.0 1.0 RUN81 -06 0.261 298.7 11780.0 0.0 1.0 RUN81

-94O82005 RUN 82 197.06 302.2 24.48 302.2 0.1805 0.8195 RUN 82 961.2 984,9 89.72 001 18.7 122.0 18.38 299.8 23.79 0.0498 0,9502 RUN 82 -02 12.240 303.2 11780.0 0.1694 0.8306 RUN 82 -03 1.315 303.2 11780.0 0.0 1.0 RUN 82 -04 0.097 3u3.2 11780U.0 0.0 1,0 RUN 82 — 05 0.01 303.2 11780.0 0.0 1.0 RUN 82 084007 RUN 64 724.2 296.7 419.96 296.7 0.0 1.0 RUN 84 961.1 972.2 89.72 001 364.7 123.0 100.0 298.0 0.0 0.0 100 RUN 84 002 374.8 123.0 100.0 298.0 00 0. 0 1.0 RUN 84 003 369.8 122.0 41.L 296.5 23.79 0.0 1.0 RUN 84 -04 18.525 299.8 11780.0 0.0 1.0 RUN 84 -05 5.434 298.7 11780.0 0.0 1.0 RUN 84.06 1.74v 298.7 11780.0 0.0 1.0 RUN 84 — 07 0.386 298.7 11780.0 0.0 1.0 RUN 84 085004 RUN b5 515.71 298.2 263.65 298.2 0.0 1.0 RUN 85 961.1 972.2 89.72 001 258.7 123.0 36.56 297.6 23.79 0.0 1.0 RUN 85 -02 17.694 300.9 11780.0 0.0 1.0 RUN 85 -03 1.798 30U.9 11780.0 0.0 1.0 RUN 85 -04 0.126 300.9 11780.0 0.0 1.0 RUN 85 086005 RUN 86 404.84 299.7 185.96 299.7 0.0 1.0 RUN 86 961.1 972.2 89.72 u001 171.0 123.0 100.0 298.0 0.0 0.0 1.0 RUN 86 002 173.0 123.0 101.0 298.7 23.79 0.0 1.0 RUN 86 -03 16.3ul 300.9 11780.0 0.0 1.0 RUN 86.04 1.16o 3uu.9 11780.0 0.0 1.0 RUN 86 -05 0.019 3ou.9 11780,0 0.0 1.0 RUN 86 087006 RUN 87 262.5 299.4 37.3 299.4 0.0 1.0 RUN 87 961.2 984.9 89.72 RUN 87 001 196.0 125.5 100.u 298.0 0.0 0.0 1.0 RUN 87 002 176.0 123.0 l10.0 298.0 0.0 0.0 1.0 RUN 87 003 196.0 123.0 3u.0 304.1 23.79 0.0 1.0 RUN 87 -04 16.378 301.5 11780.0 0.0 1.0 RUN 87 -05 1.47u 301.5 11780.0 0.0 1i.0 RUN 87 -06 0.048 301.5 11780.0 0.0 1.0 RUN 87 088005 RUN 88 350.4 303.0 155.7 303,0 0.0 1.0 RUN 88 961.2 972.2 89.72 RUN 88 001 87.0 125.5 lu0.0 298.0 0.0 0.0 1.0 RUN 88 102 85.0 123.0 49.0 296.5 23.79 0.0 1.0 RUN 88 -03 14.483 298.7 11780,0 0.0 1.0 RUN 88 -'04 1.102 299.8 11780,0 0.0 1.0 RUN 88 -05 0.029 299.8 11780.0 0.0 1.0 RUN 88 089005 RUN 89 257.6 298.0 89.1 298.0 0.0 1.0 RUN 89 961.2 972.2 89.72 RUN 89 001 23.3 123.0 100.o 298.0 0.0. 0 0 100 RUN 89 002 23.3 122.0 21.0 296.5 23.79 0.0 1.0 RUN 89 -03 12.047 30u.4 11780.0 0.0 1.0 RUN 89 -04 1.682 298.7 11780.0 0.0 1.0 RUN 89 -05 0.087 298.7 11780.0 0.0 1.0 RUN 89 090006 RUN 90 574.15 298.? 319.13 298.7 0.1818 0.8182 RUN 90 961.2 972.2 89.72 RUN 90 001 279.0 123.2 20.0 298.0 0.0 0.0511 0.9488 RUN 90 002 269,0 122.7 20.o 298.0 0.0 0.0511 0.9488 RUN 90 003 274,0 123.2 36.0 294.3 23.79 0.0511 0.9488 RUN 90 -04 19.279 299.3 11780.0 0.1643 0.8357 RUN 90 -05 0.889 299.3 11780.0 0.0 1.0 RUN 90 -06 0.042 299.3 11780.0 0.0 1.0 RUN 90 091007 RUN 91 297.6 299.0 110.0 299.0 0.1855 0.8145 RUN 91 961.2 972.2 89.72 RUN 91 001 84.0 124.2 20.0 298.0 0.0 0.0466 0.9534 RUN 91 002 88.0 124.2 20.0 298.0 0.0 0.0466 0.9534 RUN 91 003 88.0 124.0 80.0 297.0 23.79 0.0466 0.9534 RUN 91 -04 13.804 298.7 11780.0 0.1607 0.8393 RUN 91 -05 1.624 298.7 11780,0 0.0 1.0 RUN 91 -06 0.077 298.7 11780.0 0.0 1.0 RUN 91 -07 0.004 298.7 11780.0 0.0 1.0 RUN 91 09200C)7 RUN 92 411.0 298.2 255.0 298.2 0.9904 0.0096 RUN 92 961.2 972,2 89.72 RUN 92 001 166.0 200.2 20.0 298.0 0.0 0.9860 0.0140 RUN 92 002 136.0 192.2 20.0 298.0 0.0 0.9860 0.0140 RUN 92 003 161,0 200.2 115.0 297.0 23.79 0.9860 0.0140 RUN 92 -04 11.879 299.3 11780,0 0.991? 0.0083 RON 92 -05 1.76u 299.3 11780.0 1.0 0.0 RUN 92 -06 0.155 299.3 11780,0 1.0 0.0 RUN 92 -07 0.014 299.3 11780.0 1.0 0.0 RUN 92

-950930U5 RUN 93 233.0 299.7 92.6 299.7 0.9895 0.0105 RUN 93 961.2 972.2 89.72 RUN 93 001 43*0 192.9 20.0 298.U 0.0 0.9855 0.0145 RUN 93 002 37.0 191.5 33.0 298.2 23.79 0.9855 0.0145 RUN 93 — 03 10.887 30u.4 11780.0 0.9910 0.0090 RUN 93 -04 0.967 300.9 11780.0 1.0 0U.0 RUN 93 -05 0.039 300.9 11780.0 1.0 0.0 RUN 93 094004 RuN 94 751.4 302.2 529.8 3u2.2 0.3528 0.6472 RUN 94 961.2 972,2 89,72 RUN 94 001 490,0 196.9 52.0 30U.9 23.79 0.2010 0.7990 RUN 94 -02 16.746 3U3.2 11780.0 0.3528 0U6472 RUN 94 — 03 0.677 302.6 11780.0 0.0 1.0 RUN 94 - 04 0,029 302.6 11780.0 0.0 1.0 RUN 94 095004 RUN 95 413.0 302.2 244.5 302.2 0.344d 0.6552 RUN 95 961.1 972.2 89.72 001 195.U 195.9 43.0 298.7 23.79 0.1762 0.8238 RUN 95 -02 12.898 300.9 11780.0 0.3310 0.6690 RUN 95 — 03 c.657 300.9 11780.0 1.0 0.0 RUN 95 -04 0.029 3vu.9 11780,0 1.0 0.0 RUN 95 096004 RUN 96 232.4 300U.7 92.1 300.7 0.3457 0*6543 RUN 96 961.1 972.2 89.72 001 70.0 194.2 69.0 296.5 23.79 0.1876 0.8124 RUN 96 -02 10.248 3uu.4 11780.0 0.3317 0,6683 RUN 96 -03 0.851 300.4 11780.0 1.0 0.0 RUN 96 -04 0.048 300.4 11780.0 1.0 0.0 RUN 96 097004 RUN 97 744.3 300.7 529.8 300.7 0.5455 0.4545 RUN 97 961.1 972.2 89.72 001 420.0 194.7 102.0 298.7 23.79 0.3566 0.6434 RUN 97 -02 15.585 299.8 11780.0 0.5250 0.4750 RUN 97 — u3 1.276 30U0.4 11780.0 1.0 0.0 RUN 97 -04 0.058 300u.4 11780.0 1.0 0.0 RUN 97 0989005 RUN 98 505.6 302.2 317,1 302.2 0.5405 0.4595 RUN 98 961.1 972.2 89.72 001 287,0 195.2 66.0 298.2 23.79 0.3335 0.6665 RUN 98 -02 13.768 299.8 11780.0 0.5202 0.4798 RUN 98 -03 1.257 299.8 11780,0 1.0 0.0 RUN 98 -04 0.081 299.8 11780.0 1,0 0.0 RUN 98 -05 0.0097 299.8 11780.0 1.0 0.0 RUN 98 099004 RUN 99 277.5 301.0 126.6 301.0 0.5340 0.4660 RUN 99 961.1 972.2 89.72 0ul 102.6 195.2 94.0 297.6 23.79 0.3257 0.6743 RUN 99 -02 11.428 299.3 11780.0 0.5289 0.4711 RUN 99 -03 0.851 299.3 11780.0 1.0 0.0 RUN 99 — 04 0.035 299.8 11780.0 1.0 0.0 RUN 99 100005 RUN100 660.0 299.2 463,0 299.2 0.8123 0.1877 RUNlOO 961.1 972.2 89.72 001 358.0 195.2 81.u 297.0 23.79 0.6873 0.3127 RUNlO -02 15.276 299.8 11780.0 0.8062 0.1938 RUN100 -03 1.102 299.8 11780.0 1.0 0.0 RUNlOO -04 0.058 299.8 11780.0 1.0 0.0 RUNlO0 -05 Ou0039 299.8 11780.0 1.0 0.0 RUNlOO lU01005 RUNlI1 432.0 299.0 257.6 299.0 0.8156 0.1844 RUN101 961.1 972.2 89.72 001 205,5 194.7 67.0 298.7 23.79 0.6582 0.3418 RUNlO1 -02 12.956 300.4 11780.0 0.7842 0.2158 RUNlO1 -03 1.392 300.4 11780.0 1.0 0.0 RUN101 -04 0.073 30u.4 11780.0 1.0 0.0 RUNlO1 -05 J.015 30u.4 11780.0 1.0 0.0 RUNIO1 102005 RUNLO2 245.5 299.7 99.2 299.7 0.8138 0.1862 RUN102 961.1 972.2 89.72 001 66.0 196.2 51.7 298.7 23.79 0.7003 0.2997 RUNO02 -02 10e829 301.5 11780.0 0.805d 0.1942 RUN102 -03 1.044 301.5 11780.0 1.0 0.0 RUN102 -04 0.056 3u1.5 11780.0 1.0 0.0 RUN102 -05 u.0097 301.5 11780.0 1.0 0.0 RUN102 103004 RUN103 235.4 301.2 160.7 301.2 0*0349 0.9651 RUN103 961.1 972.2 89.72 001 155.0 295.9 132.5 297.6 23.79 0.0007 0.9993 RUN103 -02 5.356 300.9 11780.0 0.0 1.0 RUN103 -03 0.213 300.9 11780.0 0.0 1.0 RUN103 -04 0.014 300.9 11780.0 0.0 1.0 RUN103 104005 RUN104 643.6 301.2 502.6 301.2 0.3900 0.6100 RUN104 961.1 972.2 89.72 001 488.0 300.9 50.7 299.8 23.79 0.2801 0.7199 RUN104 002 450.0 295.7 82.0 302.1 23.79 0.2712 0.7288 RUN104 -03 10.016 305.2 11780.0 0.3828 0.6172 RUN104 -04 0.348 305.2 11780.0 1.0 0.0 RUN104 -05 0.019 305.2 11780.0 1.0 0.0 RUNI04

105004 RUN105 476.4 304.0 358.5 304.0 0.3887 0.6113 RUN105 961.1 972.2 89.72 001 319.0 296.7 134.5 302.1 61.94 0.273d 0.7262 RUNl05 — 02 8.044 302.6 11780.0 0.3845 0.6155 RUN105 -03 0.348 302.6 11780.0 1.0 0.0 RUN105 -04 0.014 302.6 11780.0 1.0 0.0 RUNl05 106004 RUN106 226.3 302.9 149.6 302.9 0.3881 0.6119 RUN106 961.1 972.2 89.72 001 134.0 296.4 81.8 302.1 23.79 0.2698 0.7302 RUN106 -02 5.588 302.1 11780.0 0.3776 0.6224 RUN106 — 03 0.251 302.1 11780.0 1.0 0.0 RUN106 -04 0.0097 302.1 11780.0 1.0 0.0 RUN106 107004 RUN 107 448.2 303.2 335.3 303.2 0.2226 0.7774 RUN 107 961.1 972.2 89.72 001 309.7 296.7 63.0 304.1 23.79 0.1495 0.8505 RUN 107 -02 8.121 303.2 11780.0 0.2121 0.7879 RUN 107 -03 0.329 303.2 11780.0 0.0 1.0 RUN 107 -04 0.014 303.2 11780.0 0.0 1.0 RUN 107 108004 RUN 108 210.2 298*9 139.5 298.9 0.2122 0.7878 RUN 108 961.1 972.2 89.72 OU1 123,3 296.4 40.0 302.6 23.79 0.1489 0.8511 RUN 108 -02 5.337 303.2 11780.0 0.2165 0.7835 RUN 108 -03 0.193 303.2 11780.0 0.0 1.0 RUN 1U8 -04 0.0096 303.2 11780.0 0.0 1.0 RUN 108 109004 RUN 109 682.9 302.9 532.8 302.9 0.2533 0.7467 RUN 109 961.1 972.2 89.72 Ou1 520.2 296.2 45.0 300.4 23.79 0.1714 0.8286 RUN 109 -02 10.132 302.1 11780.0 0.2496 0.7504 RUN 109 -03 0.483 302.1 11780.0 0.0 1.0 RUN 109 -04 0.021 302.1 11780,0 0.0 1.0 RUN 109 110004 RUN 110 415.9 300.4 31.0o 300.4 0.2583 0.7417 RUN 110 961.1 972.2 89.72 001 288.7 296.2 52.0 302.6 23.79 0.1731 0.8269 RUN 110 -02 7.870 303.2 11780.0 0.2405 0.7595 RUN 110 -03 0.348 303.2 11780.0 0.0 1.0 RUN 110 -04 0.015 303.2 11780.0 0.0 1.0 RUN 110 111004 RUN 111 296.9 303.0 206.1 303.0 0.2589 0.7411 RUN 111 961.1 972.2 89.72 001 188.5 296.2 50.8 301.5 23.79 0.1741 0.8259 RUN 111 — 02 6.613 303.2 11780.0 0.2474 0.7526 RUN 111 -03 0.309 303.2 11780.0 0.0 1.0 RUN 111 — 04 0.012 303.2 11780.0 0.0 1.0 RUN 111 112005 RUN112 1365.0 302.2 1055.0 302.2 0.0 1*0 RUN112 961.7 972.7 89.72 RUN112 301 850,0 197.5 95.0 303.2 23.79 0.0011 0.9989 RUN112 — 02 20.304 303.2 11780.0 0.0009 0.9991 RUN112 -03 3.616 303.2 11780.0 0.0 1.0 RUN112 -04 0.348 303.2 11780.0 0.0 1.0 RUN112 -05 0.004 303.2 11780.0 0.0 1.0 RUN112 113004 RUNl13 1005.0 304.4 742.0 304.4 0.0012 0.9988 RUN113 961.7 972.7 89.72 RUNl13 001 660,.0 197.5 80.0 305.2 23.79 0.0016 0.9984 RUN113 -02 18.000 305.8 11780.0 0.0012 0.9988 RUNI13 -03 0.948 305.8 11780.0 0.0 1.0 RUN113 -04 0.019 305.8 11780.0 0.0 1.0 RUN113 114005 RUN114 755.0 298.2 539.5 298.2 0.9915 0.0085 RUN114 961.7 972.7 89.72 RUN114 001 460.0 195.o 55.0 297.6 23.79 0.9873 0.0127 RUN114 -02 16.417 298.2 11780.0 0.9874 0.0126 RUN114 -03 3.461 298.2 11780.0 1.0 0.0 RUN114 -04 0.252 298.2 11780.0 1.0 0.0 RUN114 -05 0.019 298.2 11780.0 1.0 0.0 RUN114 115004 RUN115 910.0 296.2 755.0 296.2 0.9913 0.0087 RUNI15 961.7 972.7 89.72 RUNl15 001 750.0 297.9 80.0 297.6 23.79 0.9877 0.0123 RUNl15 -02 12.762 298.2 11780.0 0.9884 0.0116 RUN115 -03 0.793 298.2 11780.0 1.0 0.0 RUN115 -04 0.038 298.2 11780.0 1.0 0.0 RUNl15

-97116004 RUNl16 1585.0 295.4 1243.0 295.4 0.0 1.0 RUNl16 961.7 972.7 123.6 RUN116 001 1290,0 191.2 15.0 297.6 0.0 0.0 1.0 RUNl16 -02 24.152 299.6 11780.0 0.0 1.0 RUN1I6 -u3 u.425 297.6 11780U.0 0.0 1.0 RUN116 -04.0.019 297.6 11780.0 0.0 1.0 RUNI16 117004 RUNl17 885.0 295.7 698.0 295.7 1.0 0.0 RUN117 961.7 972.7 123.6 RUNl17 001 625.0 192.5 15.0 297.0 0.0 1.0 0.0 RUN1 7 -02 17.171 297.0 11780.0 1.0 0.0 RUN117 -03 0.309 297.0 11780.0 1.0 0.0 RUN117 -U4 0.019 297.0 11780.0 1.0 0.0 RUNli7 118004 RUNI18 525.0 30u.0 328.0 300.0 0.7360 0.2640 RUN118 961.2 972.2 89.7 RUN118 001 251.0 175.0 30.0 297.0 23.7 0.5489 0.4511 RUNll8 -02 12.5 298.0 11780.0 0.6298 0.3703 RUNl18 -03 3.5 298.0 11780.0 1.0 0.0 RUNl18 -04 1.5 298.0 11780.0 1.0 0.0 RUNl18 120004 RUN 120 469.0 298.2 370.0 298.2 0.0 1.0 961.7 972.7 88.91 RUN 120 Oul 312,0 298.4 27.0 300.9 23.79 0.0 1.0 RUN 120 -U2 7.851 297.7 11780.0 0.0 1.0 RUN 120 -u3 U.715 297.7 11780.0 0.0 1.0 RUN 120 -04 U.019 297.7 11780.0 0.0 1.0 RUN 120 121006 RUN121 174.0 296.7 117.0 296.7 0.0 1.000 RUN121 961.7 972.7 88.91 RUN121 001 99.6 297.9 10.u 302.6 0.0 0.0 1.000 RUN121 002 94.0 295.9 10.0 302.6 0.0 0.0 1.00 RUN121 003 99.6 297.2 90u. 302.6 26.0 0.0 1,0 RUN121 -04 4.351 298.8 11780.0 0.0 1.0 RUN121 -U5 0.193 298.8 11780.0 0.0 1.0 RUN121 -06 0.0u4 299.3 11780.0 0.0 1.0 RUN121 122005 RUN 122 553.0 298.2 429.0 298.2 1.0 0.0 RUN 122 961.7 972.7 88.91 RUN 122 001 330.8 295.9 39.0 300.0 0.0 1.0 0.0 RUN 122 002 326.0 293.4 39.0 299.3 23.79 1.0 0.0 RUN 122 -03 10.596 298.8 11780.0 1.0 0.0 RUN 122 -04 0.406 298.8 11780.0 1.0 0.0 RUN 122 -05 0.014 298.8 11780.0 1.0 0.0 RUN 122 123004 RUN 123 217.00 299.7 125.0 299.7 1.0 0.0 RUN 123 961.7 972.7 88.91 RUN 123 001 94.0 295.7 85.0 298.7 23.79 1.0 0.0 RUN 123 -02 6.052 297.6 11780.0 1.0 0.0 RUN 123 -u3 0.367 297.6 11780.0 1.0 0.0 RUN 123 -04 0.014 297.6 11780.0 1.0 0.0 RUN 123 124005 RUN124 265.0 293.9 170.0 293.9 1.0 0.0 RUN124 961.2 972.7 88.91 RUN124 001 115.0 296.4 30.0 297.61 0.0 1.0 0.0 RUN124 002 145.0 296.4 30.0 297.61 23.79 1.0 0.0 RUN124 -03 7.31 297.1 11780.0 1.0 0.0 RUN124 -04 0.406 297.1 11780.0 1.0 0.0 RUN124 -05 0.029 297.1 11780.0 1.0 0.0 RUN124

APPENDIX B CALIBRATI ONS -98

TAB LE X X THIERBMOCOUIPLTE CALIBRATION DATA Thermocouple Reading Reference Average* Average* Reference Temperature EMF. Temp.'C Average Point 0C MV. (58 Calib) Deviation Boiling H20 (739,5 mm Hg) 99.2 4,25 99.4 0.2 Ice Point 0 0,0 0 0.0 CO2 Sublimation Point -78.5 -2.71 -78.25 0.25 Boiling 02 -182.96 -5.25 -182.5 0,46 Boiling N2 -195.8 -5.47 -195.5 0.50 * 5 readings each at each point.

-100TABLE XXI CALIBRATION DATA FOR PRESSURE GAUGES Dead Weight Gauge Readings Tester PSIG PSIG Up Down Gauge C-2-522 50 75 100 120 125 200 220 225 300 325 325 400 420 425 500 520 525 750 800 805 1000 1050 1055 1100 1160 1165 1250 1305 L305 1350 1410 1415 1500 1585 1585 1750 182o Gauge C-2-464 20 27 26 30 36 37 40o 48 50 60 67 69 80 85 88 110 117 117 160 165 165 210 215 216 260 265 266 310 316 318 41o 415 416 510 518 519 560 570 Gauge C-2-82 20 26 26 40 48 47 80 86 85 1_0o 116 117 210 215

-101300 250 0 200 150 (I) Ciz a- 100 50 00 50 100 150 200 250 300 GAUGE READING, PSIG. Figure 21. Calibration of 300 Psi Pressure Gauge (C-2-82).

-1021000 900 800 >- 700 m ci 600 0o,,i tr 500 C,, C,, cr 400 a. 300 200 100 200 300 400 500 600 700 800 900 GAUGE READINGPSIG. Figure 22. Calibration for 800 Psi Pressure Gauge (No. c-2-464).

-1032000 1800 x 1600 1400 CD U5 a.0X/ 1 200 U) U) -K —w 1000, a800 0~~~~~~~~ 600i x 400 i /X /x 200 x~ 0 -t 200 400 600 800 1000 1200 1400 1600 1800 2000 GAUGE READING, PSIG Figure 23. Calibration for 2000 Psi Pressure Gauge (No. C-2-522).

-104TABLE XXII CALIBRATION DATA FOR 2000 PSI PRESSURE TRANSDUCER (Serial No. 110) Temperature: 26~C Reference Pressure: O.01mm Hg, GAUGE READING TRANSDUCER READING PSIA MV 290 7,75 1215 36.5 1305 39.0 955 28.5 76o 22,5 570 16.75 475 13.5 375 10.5 205 5.0 260 6,25 450 12,5 290 7,75

-105TABTE XXIII CALIBRATION OF 2000 PSI PRESSUIE TRANSDUCER (Serial No. 110) Temperature: 79.6~C Reference Pressure: 0.01rm Hg GAUGE READING TRANSDUCER READING PSIA MV 1435 39.0 114o 30*8 1055 27.5 932 25.0 741 20.0 612 16,0 360 9.2 245 6,o 162 4.0 87 2,0 1100 29.5 913 24,8 815 22.0 665 18,0 550 14,4 43o 11*2 330 8,4 220 _5,4 180 4,5 325 8,4 450 11.8 573 15,0 680 18,0 645 17%0 965 27.0 1419 39.2 1320 36.5 1050 29,0 945 26,0 680 21*0 620 16.7 455 12*2 240 6.o 180 4.6

-1o61400 1300 X REF PRESSURE 0.01 mm Hg 1200 AMP SETTING I TEMP. 260C 1100 REF GAUGE C 2-522 1000 900 800 (I) ~~~~~~~~~~~~~~~~~x a700 w~~~~~~~ 600 w x a. 500 x 400 Ox 300 -x x/ 200 X ____ 100 /:O0 2 6 10 14 18 22 26 30 34 38 42 TRANSDUCER READING, MV. Figure 24. Calibration for 2000 Psi Pressure Transducer (Serial No. 110).

-1072000 1800 REF PRESSURE O.OImmHg AMP SETTING I 1600 TEMP -79.7 0C REF GAUGE C2-522!0 1400 1200 I1000 800,/, X Itt CALIBRATION.+C~ ~~ + 2nfd CALIBRATION 400 4000 x3d CALIBRATION UP A6,,DOWN 200 5 10 15 20 25 30 35 40 45 50 TRANSDUCER READING,MV Figure 25. Calibration for 2000 Psi Pressure Transducer (Serial No. 110).

TABLE XX1Y CALIBRATION DATA FOR 500 PSI PRESSTRE TRANSDUCER (Serial No, 60) Gauge Reading Transducer Reading PSIA MV Temperature: -79*70C Reference Pressure: O.Ommn Hg 115 13*5 218 24.0 321 33.5 423 43.5 523 53.0 621 65, 656 67.0 566 57.0 519 52.5 420 4440 368 39o0 320 3345 216 23.9 150 1747 119 1440 67 8,4 421 44X0 115 13 2 Temperature: 20'C Reference Pressure: OO01mm Hg 138 16 256.25.5 254 25.0 419 45.0 455 49.0 322 35.0 218 24.0 138 15.6 67 8,0

-109600 550 - lREF PRESS 0.01 mm Hg AMP SET I 500 TEMP 200C REF GAUGE C-2-464 450 400 350 300 250 200 150 100 50 5 10 15 20 25 30 35 40 45 50 55 60 MV. Figure 26. Calibration for 500 Psi Pressure Transducer. (Serial No. 60)

-110700 650 REF PRESSURE 0.01 mm Hg. AMP SETTING I 600 | TEMP. -79.7 ~C REE GAUGE C - 2 -464 550 0 500 450 t-_ 400 300 0 356 250 -- 200 350 300 50 I00 5 10 15 20 25 30 35 40 45 50 55 60 65 70 MV. Figure 27. Calibration for 500 Psi Pressure Transducer. (Serial No. 6o)

APPENDIX C CALCULATION OF SURFACE AREA Two determinations of the surface area of the adsorbent were made using the standard "BET" method. The first was made after activating the adsorbent for 12 hours at 3,50~C under a vacuum of 10-6 mm. of Hg. For the second determination, the sample was activated at 1250C for 12 hours at 3 x 10-2 mm. Hg. pressure. The results were essentially identical being 494 and 512 square meters per gram for the respective "BET" determinations. The calculated data may be found in Tables XXV and XXVI and Figures 28 and 29. -111

-113-.014..O ~ ~ ~ ~~12~ /.012.010 0 / 404/.004 /// m= — = 0.0149 CC-'.002 0 0 0.1 0.2 0.3 0.4 0.5 0.6 P / Po Figure 28. "B E T" Plot for Surface Area Determination 1 Sample - 5A Molecular Sieve

-114TABLE XXVI CALCUIATION OF EET SURFACE AREA DETERMINATION DETERMINATION 2 Activation Temperature = 125 C Activation Pressure = 3 x 10 Weight of Sample = 0.3023 gms. Average Adsorption Temperatulre = 193* 5 ~C DATA Pressure P-Po V Ads Readings mm Hg. imm Hg ccSTP P/V Ads(P-Po) 1 23.0 712.3 39.5 0.00083 0*031 2 32.0 703.3 39.54 0,00115 0o044 3 41.0 694*3 39,77 0oo00148 0.056 4 50.0 685.3 40.02 0 00182 0o068 5 59.5 675.8 40o25 000219 0o081 6 78.5 656.8 40.53 0.00295 0.107 7 100,5 634,8 40o,. 0,00389 0,137 8 147.5 587.8 41,20 0.00609 0.201 9 206,5 529,3 41.6o 0oo00936 0280 10 255,Q 480*3 41.83 o,oG69 0,347 11 298,0 437,3 42,25 0o01617 0.4b5 12 398,0 337,3 43,03 0,02742 0,541 m = 0.0282cc'1 (From Figure 29) I = 0OO00cc m 1 =34.46 cc I+m m Total Surface Area = (34.46cc)(6.02x1023 )(16,2) 154 (22,400) (1010)2 Specific Surface = 154.4 sq. meters _ 511 sq, meters/gm. 0.3023 gins

-115 - 0.018 0.016 / / / / / 0.014 / / / ~~~0. ~012 it~~/ / 0.01/ / / 0.010 / / 0. / / o~~~~ / 0.008 In/ -a / / 0. / ~~0.1 02030.06/ 0.006 / / / ~ ~~~~~~~~~~~~/ // Y m:-~ —:0.0282CCO' 0.004 0.002 0 0.1 0.2 0.3 0.4 0.5 0.6 Figure 29. "B E T" Plot for Surface Area Determination 2 Sample - 5A Molecular Sieve

APPE1N7DIX D PRESSURE VESSEL DESIGN CALCULATIONS A. Design Calculation Using the thick-walled formula from the Unfired Pressure Vessel Code for computing the allowable working pressure (Para. VA-1) P = SE(Z-l/z+l) where z = (Ro/Ro-t)2 E = Weld efficiency = 1.0 for seamless tube P = allowable pressure Ro = outside radius (in.) = 1.9 in S = allowable stress (psi) = 18,750 psi t = thickness (in.) = 0,4 in. Then z (0.95/o95-0.4)2 = 2.99 P = 18,750 x 1.0(2.99 - 1/2*99 + 1) P = 9,350 psi. For the end design the Code (Para. UG - 34) specifies that for flat heads the thickness shall be determined using: t = d cP/ where t = minimum thickness d = the inside diameter c = appropriate constant = 0.25 as the criteria. -116

-117Then t = 1.1 40.25 x 9,350/18,750 = 0.125 inches minimum B, Effect of Temperature and Pressure on Free Volume The effect on the calibrated volume of the absorption cell by both high pressures and low temperatures may be considerable. Bartlett(5) found that: V E(r22 [(2u)(P2r2 - Plrl) + 2(+u)(P2-Pl)r2] E(r22 r12) where V0 = 124 cc at 1 atm. E - 29.3 x 106 lbs./in.2 u = 0.28 rI = 1.11 r2 = 1*.9 P2 = operating pressure Pi = test pressure represents the effect of pressure on the volume of a vessel at constant temperature. Calculations indicate that the change of volume resulting from a maximum operating pressure of 2000 psi, is 0.050 cc, or 0.04%. The effect of a change of temperature on the volume of the adsorption cell was determined using the standard volumetric expansion equation: VT= VT [1 + 3xa(T2 - T1)] 2 1 where the linear coefficient of expansion a = 16x10 0C for 316 stainless steel.

-118The minimum operating temperature of 1230K resulted in a change of 0.9 per cent in the volume. As the total change resulting from the maximum change in the operating conditions is less than one per cent, the approximate error in the calibration, no correction was made for temperature and pressure.

APPENDIX E SAMPLE CALCULATION OF ADSORPTION EQUILIBRIUM Sample Calculation Run No. 76 A. Calculation of Gas Density by B-W-R Equation at Initial Feed Conditions Experimental Data P = 328.+2psia T = 300.40K YCH4 = 0.7257 Calculation p = 22533 atm. 14.696 Mixed Coefficients Bom(BK)* = ~ B y = 0.0426000 x 0,7257 + 0,0484824 x 0.2743 = 0.0442135 A (AM) (, Aoi02y) - 0.1000 (YCH4 X Y i Y2 I 1 = ((1.8550)2 x 0.7257 + (1.27389)2 x 0.2743)2 - 0.1 (0.7257 x 0.2743) = 1.6648 Com(CM), Dom(DM) and ym(IM) follow squared combinatorial rule, without the added correction. Thus: 1 1 Com(CM) = ((22570.0)2 x 0.7257 + (4273o0)2 x 0.2743) = 16,648.6 * The letters in parentheses are used in the computer program. -119

-120Dom(DM) = (7.,61781x 106) x (0.2743)2 - 5.73168 x 105 y(IM) = ((o.oo6oo00) x 0.7257 + (0.006500)2 x 0.2743)2 = 0.006135 bm(EM) = (Zbi /3y)3 mi i i //3 ((0.00338004)1/3 x 0.7257 + (0.00322373) x 0.2743)3 - 0.00306385 am(FM), cn(GM), bm(HM) and am(JM) follow the cubic combinatorial rule also, therefore: am(FM) = ((0.0494000) /3x 0.7257 + (0,0178444)l/3x 0.2743)3 - 0.05385993 m(M) = ((2,545.o0)/3 x 0.7257 + (475.00oo)1/3 x 0o.2743)3 = 1,748.9 m(HM) -- 0.08320 x 106 x (0o2743)3 - 0.1717 x 105 on(JM) = ((0.000o 24359)1/3 x 0.7257 + (0o.oo00015300)1/3 x 0.2743)3 0o,ooo131824 The second and f~ourth virial coefficients may now be evaluated independent of the density. C1 = BomT - Aom - Co/T2 - Dom/T4 = 0,0442135 x 0.08206 x 300.4 - 1.6648 - 16,648.6 (300*4)2 5.73168 x 105 (300.4)4 z -0.753636

-121C3 = amm - 0,385993 x 0,000131824 = 0.508830 x 10-5 The third virial coefficient C2 = bRT a + (C/T/T4)( 1 + yp2)/eYP ) can only be evaluated after a density has been estimated, and if p = 0.93058 gm. moles/liter then C2 = 0.00306385 x 0.8206 x 300oo.4 - 0.0385993 + (,78.9 + 1.717 x 10 1 + (0o.93058)2x o. 006135 (300.4) (300o4)4 p(0.93058)2 x o006135 = 0,0553097 The Pressure (PI) may now be evaluated from the equation of state PI = RT + Clp2 + C3 + C3 = 0.08206 x 300.4 - 0,75363 x (0.930585)2 + 0.0563097 x (0.930585)3 + 5.0883 x 104 x (0.930585)6 = 22,332 atm. Since PI = P the density p = 0,930585 gmin, moles/liter is the correct density. B. Calculation of Amount Loaded 1. Initial Amount in Feed Reservoir Methane: GMCI = Vres X PINT x YCH4 = VI x DAI x MFCHI = 961.1cc. x 0.000930585 gm.moles/liter x 0,7527 = o064906 gm.moles,

-122Nitrogen: GMNI = 961.1 x 0.0009305.85 x 0.2473 - 0.24533 gm.moles. 2. Final Amount in Feed Reservoir Density: DAF = 0.43538 gmmoles/liter at 155.7 psia and 300.40K Methane: GMCF = 972.2 x 0.00043538 x 0.7527 -0.30718 Nitrogen: GMNF = 972.2 x 0.00043538 x 0.2473 - 0o11611 3. Net Amount Loaded Into Cell Methane: LOADC = GMCI - MCF = o064906 - 0*30718 = 0*34188 gm. moles Nitrogen: LOADN = 0.12922 C. Calculation of Equilibrium Amount Adsorbed 1. Experimental Data Pressure: PC - 111.7 psia Temperature: TC = 174.87~K Free Space: VC - 89.72 cc, Gas Phase Comp, CH4, MFCH = o.4908 N2, MFN = 0.5092 2. Computation Pressure: PC = 111.7/14.696 = 7.60 atm. Density: DAC = 0.0005558 gm, moles/liter Amount in Gas Phase

-123Methane: GMGC = VC x DAC x MFCH = 89.72 x 0.0005558 x 0.4908 = 0,024476 GMGN = 89.72 x 0.0005558 x 0,5092 = 0*02539 Amount in Adsorbed Phase GAC = LQADC - GMGC = 0o34188 - 0.02448 = 0.3174 GAN = 0.12922 - 0.025394 = 0,103828 XC = GAC/GAN + GAC XC = 0*7535 Relative Volatity x = (MFN/X)/(MFCH/XC) = 3.28 D. Desorption or Amount Lost in Sampling Sample calculation for first desorption 1. Experimental Data Pressure: PS = 12.447 psia Volume: VS = 11,780 cc. Temperature: TS = 301.5 Composition: MFCH = 0,6949, MFN = 0.3051 2, Calculation DAS = 0.000034 gm.moles/cc.

-124Desorbed DELC VS x DAS x MFCH - 11,780 x 3.4 x 10-5 x 0.6949 = 0.2805 gm. moles DELN -11,780 x 3.4 x 10'6 x 0.3051 = -0 1231 gm, moles Net Remaining in System Methane LOADC LOADC. - DELC i-i 1 0.34188- 0,2805 = 0.0613 gm,moles

AD INPUT RMAD INPUT AI=BWR. (PI, TI GMCI=VI*NDAI*MFCHI MCF=VFC DAFAMFCHI LOADC=GMCI-GMCF A _ RUN NO. =-N PI, TI, PF, TF, FC I NO. of SAMS MFCHI, MFNI, VI, DAF=BWR. (PFTF, GMNI=VI*DAI*MFNI NF=VF*DAF*MFNI LOADN=GMNI-GMNF PRINE DAC=BWR.(PC,TCMFCH) TREAD DATA I |, PF, TF, F FOR MFCHI, MFNI, D MGC=VC*DACy*MFCH N r ErL VER Ng PCG TC, PS, TS, 4, I LGMCIi GMNI, VS MFCHoe MFN GMCF, GMNFD GMGN=VC *DAC *MFN LOADC LOADN GAC= LOADC h GMGC AXC=GAC/(GAC+GAN) MFCHPRINT IA1PHA= (MFCH/X)/ N, GMGC, GMGN, GAC, FUG ACITY DESIRED A _ h GAN=LOADN GMGN XN=GAN/(GAC+GAN) (MFN/XN)GAN, MFCH XC MFN, XN, ALPHA, PC7 TC I BWRE. DAF - Density of Feed Mixture After Loading GM - gm moles in Loading Cell GAN - gm. moles Nitrogen Adsorbed V - Volume GMG - gm. moles inM Gas Phase (Next Symbol Denotes Component Figure 30. Computer Flow Diagram for Equilibrium- Adsorption Loading Calculation.

NOMENCLATURE: RBIW, ) DENI=Z/RT EXECUTE DEN. FUNCTION RETURN DENF. AM,.JM - Coefficients of BWR Equation (Function of Composition) DENF - Next Density Guess DENI=1'5 x ~ ~ 0 DENI - Present Density -Dvv-l IEN(TRYN) > (EXECUTE DEN. FUNCTION RETURN DEN N - Composition-Mole Fraction ENTRY PI - Present Pressure Atm. T ZI N. FUN, FUC T - Temperature OK. fZ - Pressure Atm. DENI=Z/RT EXECUTE DEN. TRANSFER TO ADD DENI=1.5 x COMPUTE LIeD.(T,N) EXECUTE DEN. DP=i/EXP.(IMxDENF2 FlC......F6C FUN=MFN x FFN=F1N+F2N+F3N COMPUTE FUC=N x FFC=FlC+F2C-F3C EXP.(FFN/RT) +F4N+F5N-F6N FiN. F6N EXT.(FFC/RT) +F4C+F5C-F6c FUNCTION RETURN I D COMPUTE MIXED C1=f (R,T,AM,BM,CM,DM) C2=f(R,T,EM,FM,GM F=RxTxDENI+ClxDEN | COEFFICI~ C3=FM x JM I~ 1 a HMIMDENI) +C2xDENI3+-3xDENI6 YES FUNCTION RETURN 0_~ DENF PI-Z 1 ~~) DENF-DEN DENF=DENI.- P ART=)PI/~DENI TRANSFER TO a DENI=DENF Figure 31. Computer Flow Diagram for B-W-R Equation Solution for Density and Fugacity

TIABLE~:XXVII SAMPLE PRINT OUT FROMZ CCNP~JUTER CAI~ULATION L' " """''' i< E'""" L:!L(L::''i.-"::.~ —-?. U:.'.-' "~C. I'i ~ 7i. -. ii.,.._._ ~,..,.....::. ~'i~~~~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,':': L_. n R'S r, -. ~z L i.-~. = 0.3:i i;5;i::-'-: ":'-:=- _. I ~.......~ i. I...i.,~ - i".:~_ -n i"~;,,',,,, _ i: i:. _i.-,,.~..=: = ~ ii ~ ~ ~ ~~~~~~~~~~_ i 7~'.~:,,,L......: -,._.: I~: ~-.,,.'i;;:,??'..~r- ~:,' ~-.-~~ i:-' -' 3;'":....... Ei, ""- }5':-;:'.:!M...-::RT~~~~~~~~~~~~~~C.i (,;R Ii ~, = f'~II i P.I 4..:;,2 4 u':. i-:::: _.;;_: -...._,:;.;~~~~~~ ~~~ ~~~~~~.......: _. i....... ii~-,.:2:'i:.'~...' z'.......':- "'"!"!-'.4:'S 3' hiI~.........'"'i "'i':"'": T'C.:::71 " ~:"........,:. i ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~: ~~~~ ~ ~ ~ _'_.."'.: ":. ~ "" i:. LH,:F.5..''.. ":,/ i 0 ~".'.;::3 i'?~i}.....CF i:.~ii']S...F]...:: i;..~, G i:-:: 2S6.:');-:i~'i:( CJ" i':.....F r- ~~~~~,_.: 1;i-:i; C':'.:3'0i'f::~ "'":'" ~~ i"~ 2 -- >;~:. (.)i;: i~ i'[}' 4..r ~,......!';'.J'}::::~:::"""'-L. T''.........':'....... is ~ ~~~~~~~~~~~~~~~_,..,,,.:::....-........ _.. _. _......_...-~.: ~ ~ ~ ~~~~~~~ ~: t -—...:...-:..-,. -........... -~r:: _. ti C!L.E —~. 4''C"'-C;: E:" i- Ci.......:: -::: ":, i._.................._:...:,_, ii~: E..: L.'r'.k: = 0,} }2d;'5i::.... ( " " ='::''': L~~~~~~,. i.....-....._.....:._,.-i~;:..:, —,, T H:':: i —jii~~~i.:':..'..-"''-.... CH4..-.(: }' 3 i,.:: ": i "' %... G., 0:-:-:; 3;i. L.Li' L.,.i L~.. L.:..~..,.. -- t~~i:;'~'"':5 (iI~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~3:: P,~i r"- ":~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

APPENDIX F ADSORPTION OF HELIUM During the early exploratory work, a number of runs was made to determine the adsorption of helium at higher pressures. Several runs were made at room temperature as well as one at -68OC and. one at -1970C. The results are listed in Table XXVII below. TABLE XXVIII ADSORPTION OF HELIUM Temp. Pressure Amount Adsorbed ~C; ATM milligram moles/gm. 21 71.45 0.0 21 41.17 0.0 -68 44.2 2.0 -197 27.2 2.9 The experimental procedure was identical to that used in determining methane and nitrogen equilibrium. The calculation of the equilibrium amount adsorbed was similar to that outlined in Appendix E which shows a sample calculation for methanenitrogen adsorption. The P-V-T behavior of helium at room temperature was represented by the Beattie-Bridgman equation with appropriate constants. At the lower temperatures, -68~C and -197~C, the reduced state correlation of Hamrin and Thodos(23) was used to obtain the gas phase density of helium. One determination of the adsorption equilibria of a mixture of nitrogen and helium was made at room temperature. The data from this run, though inconclusive, indicate that the helium acts as a diluent at this temperature. The amount adsorbed was, however, about fifteen per cent -128

-129greater that would be expected if one assumes an ideal mixture between the two components. The results obtained in this exploratory study are included to serve as a guide for future work as the helium-nitrogen system is of particular interest today from a conservation standpoint.

BIBLIOGRAPHY 1. Antropff, von A. Kolloid Z., 129, 11 (1952). 2. "Boiler and Pressure Vessel Code." ASME, Sect. VIII, 1956ed., 7-12, 109-113. 3. Baly, E. C. C. Proc. Roy. Soc., A160, 465 (1937). 4. Barrer, R. M. J. Soc. Chem. Ind., 64, 130 (1945). 5. Bartlett, E. P. J. Am. Chem. Soc., 49, 687 (1927). 6. Benedict, M., Webb, G. B. and Rubin, L. C. J. Chem. Phys., 8, 334 (1940). 7. Ibid, 10, 747 (1942). 8. Berenyi, L. L. Z. Physik. Chem., 94, 628 (1920). 9. Berenyi, L. L. Z. Angw. Chem., 35, 237 (1922). 10. Bloomer, O. T., Parent, J. D. I.G.T. Bull. No. 17, 1952. 11. Bloomer, O. T., Roa, K. N. I.G.T. Bull. No. 18, 1952. 12. Breck, D. W., Eversole, W. G., Milton, R. M., Reed, T. B., and Thomas, T. L. J. Am. Chem. Soc., 78, 5963 (1956). 13. Brunauer, S. The Adsorption of Gases and Vapors, Vol. I, Princeton Univ., Princeton, 1945. 14. Brunauer, S., Demning, L. S., Demming,W. E. and Teller, E. J. Am. Chem, Soc., 62, 1723-32 (1940). 150 Brunauer, S., Emmet, P. and Teller, E. J. Ani. Chem. Soc, 6o, 309 (1938). 16. Coolidge, A. S. J. Am. Chem. Soc., 46, 596 (1924). 17. Darby, H. S. Pressure - Volume - Temperature Relationship of Methane - Nitrogen Mixtures at Low Temperatures, Unpublished Master's Thesis, I.G.T., 1951. 18. Dubinin, M. M. Chem. Rev., 60, 235 (1960). 19. Ellington, R. T., Bloomer, O. T., Eakin, B. E. and Gami, D. C. "Thermodynamic and Transport Properties of Gases, Liquids and Solids," ASME, New York, 102 (1959). -130

-13120. Fowler, R. H. and Guggenheim, E. A. Statistical Thermodynamics, Macmillan, New York, 1939. 21. Freundlich, H. Colloid and Capillary Chemistry, Methuen and Co., London, 1926. 22. Goldman, F. and Polanyi, M. Z. Physik. Chem., A132, 321 (1928). 23. Hamrin, C. E. Jr. and Thodos, G. AIChE Journal, 4, 480 (1958). 24. Hill, T. L. J. Chem. Phys., 14, 263 (1946). 25. Ibid., 268-7. 26. Hill, T. J. Am. Chem. Soc., 72, 5347 (1950). 27. Hiza, M. J. Chem. Eng. Progr., 56, No. 10, 68 (1960). 28. Honig, J. M. Anal. N. Y. Acad. Sci., 58, 741 (1958). 29. Johnson, V. J. Adv. in Cryogenic Eng., 3, Paper A-2, Plenum Press, New York, 1960. 30. Josefowitz, S. and Othmer, D. F. Ind. Eng. Chem., 40, 739 (1948). 31. Joyner, L. G., Weinberger, E. B. and Montgomery, C. W. J. Am. Chem. Soc., 67, 2182 (1945). 32. Keyes, F. G. and Burks, H. G. J. Am. Chem- Soc. 50,-'1100 (1928). 33. Langmuir, I. Phys. Rev., 8, 149 (1916). 34. Langmuir, I. J. Am. Chem. Soc. 40, 1361 (1918). 35. Lee, V. J. Unpublished Paper, University of Michigan, 1959. 36, Lewis, W. K., Gilliland, E. R., Chertow, B. and Codogan, W. P. Ind. Eng. Chem,, 42, 1319 (1950). 37e Ibid., 1326 (1950). 38. McBain, J. W. and Bakr, A. M. Jr. J. Am. Chexm. Soc., 48, 690 (1926). 39. McBain, J. W. and Britton, G. T. J. Am. Chem. Soc., 52, 2198 (1930). 40. Markham, E. D. and Benton, A. F. J. Am. Chem. Soc., 53, 497 (1931). 41. Milne, W. E. Numerical Calculus, Princeton University, Princeton, 1949. 42. Morris, H. E. and Mass, O. Can. J. Res., 9, 240 (1923).

iDUN~pIyNaIIN~~III 2 -1323 9015 03466 5821 43. Nikolaev, K. M. and Dubinin, M. M. Izvest. Akad. Nauk SSSR, Otdel. Kheim. Nauk, 1165 (1958). 44. Pace, J. C. Compressibility Factors of Nitrogen, Methane, and Three of Their Mixtures. Unpublished Master's Thesis, I. G. T., 1951. 45. Pease, R. eId J. Am. Chem. Soc., 45, 1196 (1923). 46. Peierls, R. Proc. Camb, Phil. Soc., 32, 471 (1936). 47. Polanyi, M. Ber. Deut. Physik. Ges., 16, 1012 (1914). 48. Polanyi, M. Ver. Deut. Physik. Ges., 18, 55 (1916). 49. Polanyi, M. Z. Electrochemie, 26, 370 (1920). 50. Polanyi, M. Z. Physik. Chem., A138, 459 (1928). 51. Rabe, A. and Thomas, T. L. Private Communication. 52. Redlich, 0, and Peterson, D. L. J. Phys. Chem., 63, 1024 (1959). 53. Reed, T. B. and Breck, D. W. J. Am. Chem. Soc., 78, 5972 (1956). 54o Rideal, E. K. Surface Chemistry, Cambridge University, Cambridge, 1930. 55. Sault, R. A. Compressibility of Certain Methane - Nitrogen Mixtures. Unpublished Masterts Thesis, I.G.T., 1952. 56. Schay, G. J. Chem, Phys., (Hungary), 53, 691 (1956). 57. Schay, G. Fejes, P. and Szethmary, J., Acta Chim. Aca. Sci., Hungar-y, 12, 299 (1957). 58. Sips, R. J. Chem. Phys. 16, 490 (1948). 59. Ibid., 18, 1024 (1950). 60. Tarpnell, B. M. W. Chemisorption. Butterworthl, London,. 1955. 61. Wang, J. Proc. Roy. Soc., A161, 127 (1938). 62. Williams A. M. Proc. Roy. Soc., Edinburg, 38, 23 (1917-18). 63. Zeldowitsh, J. Acta Physiochim. USSR, 1, 961 (1934).