VAPOR-LIQUID EQUILIBRIA AT ELEVATED TEMPERATURES AND PRESSURES by DONALD JAMES WHITTLE B.A.Sc, University of Bri t i s h Columbia, 1956 M.A.Sc, University of British Columbia, 1958 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF Ph.D. in the Department of Chemical Engineering We accept this thesis as conforming to the required standard i THE UNIVERSITY OF BRITISH COLUMBIA March, 1962 In presenting t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f the requirements f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I f u r t h e r agree that permission f o r extensive copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head o f my Department o r by h i s r e p r e s e n t a t i v e s . I t i s understood t h a t copying or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be allowed without my w r i t t e n permission. Department of The U n i v e r s i t y of B r i t i s h Columbia, Vancouver 3 , Canada. Date TrU^Jl /(% /9CL The University of British Columbia FACULTY OF GRADUATE STUDIES PROGRAMME OF THE FINAL ORAL EXAMINATION FOR THE DEGREE OF DOCTOR OF PHILOSOPHY B.A.Sc. University of British Columbia M.A.Sc. University of British Columbia THURSDAY, MARCH 15th, 1962, at 2:30 P.M. IN ROOM 207, CHEMICAL ENGINEERING BUILDING of DONALD JAMES WHITTLE COMMITTEE IN CHARGE Chairman F.H. Soward W.A. BRYCE S.D. CAVERS H.M. DAGGETT, JR. B.A. DUNELL J.S. FORSYTH B.H. LEVELTON D.S. SCOTT-P.L. SILVESTON External Examiner: Benjamin C.-Y. LU University of Ottawa, Canada. VAPOR-LIQUID EQUILIBRIA AT ELEVATED TEMPERATURES AND PRESSURES ABSTRACT , A modified version of an apparatus by Sage and Lacey (Trans. A.I.M.E., 174: 102, 1948) for phase equilibrium measurements at elevated temperatures and pressures has been designed and built. Assembly and detail drawings of the apparatus, and a detailed description for its use have been included. The vapor pressure of methanol has been measured at temperatures from 100°C. to 225°C, and the vapor pres-sures of benzene and toluene have been measured at tem-peratures from 125°C. to 250°C, Good agreement with literature values was found for the vapor pressures of each of these materials. Vapor-liquid equilibrium composition and pressure data have been measured for the benzene-toluene system at 200°C. and compared to values reported for this system by v . Huhn (Forsch. Geb, Ingenieurw., A2: 129, 1931) and by Griswold, Andres, and Klein (Trans. A.I.Ch.E., 37: 223, 1943). Activity coefficients calculated for the liquid phase from the experimental data show that the liquid mixture forms a Lewis and Randall perfect solution. Isothermal vapor-liquid equilibrium composition and pressure data for the methanol-toluene system have been measured at 25-degree intervals from 125°C. to 250°C. An azeotrope was found for the system at temperatures of 125°C, 150°C. and 175°C., and i t was estimated that this azeotrope disappears at 193.4°C. Activity coefficients have been calculated from the experimental data using several approximate methods, and tested for thermodynamic consistency using the test proposed by Redlich and Kister (Ind. Eng. Chem., 40: 345, 1948). Coefficients calculated using the method proposed by Black (Ind. Eng. Chem., 50: 391, 1958) were the only ones which showed good thermody-namic consistency when tested in this way, and these have been correlated with a Redlich and Kister four-constant equation. GRADUATE STUDIES Field of Study: Chemical Engineering Chemical Engineering Thermodynamics Industrial Kinetics and Catalysis Momentum, Heat and Mass Transfer Solvent Extraction D i s t i l l a t i o n Fluid and Particle Dynamics Process Heat Transfer L.W. Shemilt D.S. Scott N. Epstein S.D. Cavers J.S. Forsyth N. Epstein N. Epstein Other Studies Applied Fluid Mechanics Theory and Application of Differential Equations St a t i s t i c a l Theory of Matter Physical Inorganic Chemistry A.W. Marris T. Hull W. Opechowsk C. Reid, J. Halpern ABSTRACT i i A m o d i f i e d v e r s i o n o f an a p p a r a t u s by Sage and Lacey"'" f o r phase e q u i l i b r i u m measurements a t e l e v a t e d t e m p e r a t u r e s and p r e s s u r e s has been d e s i g n e d and b u i l t . Assembly and d e t a i l d r a w i n g s o f t h e a p p a r a t u s , and a d e t a i l e d d e s c r i p t i o n f o r i t s use have been i n c l u d e d . The v a p o r p r e s s u r e o f m e t h a n o l has been measured a t t e m p e r a t u r e s f r o m 100 C t o 225 C , and t h e v a p o r p r e s s u r e s o f benzene and o f t o l u e n e have been measured a t t e m p e r a t u r e s f r o m 125°C. t o 250°C. Good agreement w i t h l i t e r a t u r e v a l u e s was f o u n d f o r t h e v a p o r p r e s s u r e s o f each o f t h e s e m a t e r i a l s . V a p o r - l i q u i d e q u i l i b r i u m c o m p o s i t i o n and p r e s s u r e d a t a have been measured f o r t h e b e n z e n e - t o l u e n e system a t 200°C and compared t o v a l u e s 2 3 r e p o r t e d f o r t h i s s ystem b y v. Huhn and b y G r i s w o l d , A n d r e s , and K l e i n . A c t i v i t y c o e f f i c i e n t s c a l c u l a t e d f o r t h e l i q u i d phase f r o m t h e e x p e r i m e n t a l k d a t a show t h a t t h e l i q u i d m i x t u r e forms a L e w i s and R a n d a l l p e r f e c t s o l u t i o n . I s o t h e r m a l v a p o r - l i q u i d e q u i l i b r i u m c o m p o s i t i o n and p r e s s u r e d a t a f o r t h e m e t h a n o l - t o l u e n e system have been measured a t 25-degree i n t e r v a l s f r o m 125^C. t o 250 C. An a z e o t r o p e was f o u n d f o r t h e system a t t e m p e r a t u r e s o f 125°, 150°, and 175°C, and i t was e s t i m a t e d t h a t t h i s a z e o t r o p e d i s a p p e a r s a t 193. 4°C A c t i v i t y c o e f f i c i e n t s have been c a l c u l a t e d f r o m t h e e x p e r i m e n t a l d a t a u s i n g s e v e r a l a p p r o x i m a t e methods, and t e s t e d f o r thermodynamic 5 c o n s i s t e n c y u s i n g t h e t e s t p r o p o s e d b y R e d l i c h and K i s t e r . C o e f f i c i e n t s 6 c a l c u l a t e d u s i n g t h e method p r o p o s e d b y B l a c k were t h e o n l y ones w h i c h showed good thermodynamic c o n s i s t e n c y when t e s t e d i n t h i s way, and t h e s e have been 5 c o r r e l a t e d w i t h a R e d l i c h and K i s t e r f o u r - c o n s t a n t e q u a t i o n . The f a i l u r e o f a c t i v i t y c o e f f i c i e n t s c a l c u l a t e d b y a p p r o x i m a t e methods o t h e r t h a n B l a c k ' s t o show good thermodynamic c o n s i s t e n c y i n d i c a t e s t h a t t h e s o l u t i o n t h e o r i e s o n w h i c h t h e s e methods a r e b a s e d a r e i n v a l i d f o r a n o n - i d e a l system o f t h e t y p e s t u d i e d . TABLE OF CONTENTS i i i INTRODUCTION Importance of Vapor-liquid Equilibrium Data 1 Methods of Measuring Vapor-liquid Equilibria Under Pressure 1 Methods of Testing Vapor-liquid Equilibrium Data k Calculation of Fugacities and Activity Coefficients from Experimental Data by Exact Methods 6 Calculation of Fugacities and Activity Coefficients from Experimental Data by, Approximate Methods 8 Systems Studied and Type °f Equipment Used 11 APPARATUS AND PROCEDURE - Apparatus Introduction 13 General Arrangement of Apparatus 13 Bath Temperature Control and Measurement 15 Pressure Measurement l6 Samples of Coexisting Phases 17 APPARATUS AND PROCEDURE - Experimental Procedure Preliminary Steps 17 Introduction of Material to Cell 18 Preparation for Sampling 19 Measurements 19 Analysis 20 MATERIALS Benzene 21 Toluene 21 Methanol 21 VAPOR PRESSURE MEASUREMENTS - Benzene Previous Work 23 Results and Discussion . 2k iv VAPOR PRESSURE MEASUREMENTS - Toluene Previous Work 27 Results and Discussion 28 VAPOR PRESSURE MEASUREMENTS - Methanol Previous Work 33 Results and Discussion 3^ BENZENE-TOLUENE SYSTEM Previous Measurements 39 Results hi Calculation of Activity Coefficients hi Discussion h9 METHANOL-TOLUENE SYSTEM 'Previous Measurements 5h Results 55 Calculation of Activity Coefficients 71 1. Activity coefficients in terms of deviations from Raoult 1s law 75 2. Activity coefficients by the Lewis and Randall rule 76 a) Generalized fugacity chart of Gamson and Watson 76 b) Generalized equation of state of Black 77 3. Generalized equation of state of Hirschfelder et a l and fugacity equation of Joffe 78 h. Generalized equation of state of Redlich and Kwong 79 5- Generalized equation of state of Black 80 Thermodynamic Consistency Check of Experimental Data 82 Correlation of Activity Coefficients Obtained from the Equations of Black 83 Discussion Presentation of experimental data 9^-Azeotrope 9^ V Activity coefficient calculations 96 Cause of apparent inconsistencies in activity coefficients calculated by Black's methodr 100 Activity coefficient correlation 101 CONCLUSIONS AND RECOMMENDATIONS Vapor Pressure Measurements 103 Benzene-toluene System 103 Methanol-toluene System 10k Recommendations for Future Work 105 TABLE OF SYMBOLS 107 LITERATURE CITED 111 APPENDIX I - APPARATUS Equilibrium C e l l I - l Volume Compensating C e l l I - l - Relative Location of the Cells and Method of Clamping 1-2 Measuring Head 1-3 Measuring Head Bridge 1-5 Measuring Head Rod Assembly 1-7 Measuring Head Rod Seal 1-7 Measuring Head Rod Drive 1-8 Measuring Head Rod Drive Circuit I-10 Bomb Tubing Connection 1-12 Equilibrium Cell Stirrer 1-12 Sampling Tube Assembly I-lk Constant Temperature Bath 1-15 Temperature Control of the Bath 1-17 Temperature Measurement 1-19 Measurement of Pressure 1-23 vi Associated Glassware 1-28 APPENDIX II - DETAILS OF METHOD Cleaning of Equipment II-l Introduction of Mercury and Definition of Gauge Zero II-l Conversion of Gauge Reading to Absolute Pressure in the Equilibrium Cell II-2 1. Correction for atmospheric pressure II-3 2. Correction for variation of mercury level inside the cell II-3 3- Correction for change of mercury * density with temperature II-3 Introduction of Materials II-k Preparation for Measurements II-6 Measurements and Taking of Samples II-8 Further Measurements II-9 Temperature Limits for the Use of the Equipment II-9 APPENDIX III - CHROMATOGRAPHIC ANALYSIS OF PURIFIED METHANOL III-l APPENDIX TV - ANALYSIS OF BENZENE-TOLUENE AND METHANOL-TOLUENE MIXTURES Benzene-toluene Mixtures IV-1 Toluene-methanol Mixtures IV-2 APPENDIX V - PURIFICATION OF MERCURY V-l APPENDIX VI - TABLES OF MEASURED AND CALCULATED DATA FOR METHANOL-TOLUENE SYSTEM VI-1 APPENDIX VII - PROCEDURES USED TO PURIFY BENZENE, TOLUENE, AND METHANOL FOR VAPOR PRESSURE MEASUREMENTS Benzene VII-1 Toluene VII-2 Methanol VII-3 LIST OF FIGURES v i i Figure Page 1. General Arrangement of Apparatus Ik 2. Vapor Pressure of Benzene 26 '3- Vapor Pressure of Toluene 31 k. Vapor Pressure of Methanol 37 5- Vapor-liquid Equilibrium "x - y" Composition Diagram for BenzEne-toluene System at 200 C. kk 6. Vapor-liquid Equilibrium Pressure Composition Plot for Benzene-toluene System at 200 C. k*? 7- Vapor-liquid Equilibrium "x-y" Composition Diagram for Benzene-toluene System at 200° C 50 8. Vapor-liquid Equilibrium Pressure Composition Diagram for Benzene-toluene System at 200° C 51 9- Vapor-liquid Equilibrium "x-y" Composition Diagram for Methanol-toluene System at 125° C. 56 10. Vapor-liquid Equilibrium Pressure Composition Diagram for Methanol-toluene System at 125° C 57 11. Vapor-liquid Equilibrium "x-y" Composition Diagram for Methanol-toluene System at I5O 0 C 58 12. Vapor-liquid Equilibrium Pressure Composition Diagram for Methanol-toluene System at I5O0 C 59 13- Vapor-liquid Equilibrium "x-y" Composition Diagram for Methanol-toluene System at 175° C 60 Ik. Vapor-liquid Equilibrium Pressure Composition Diagram for Methanol-toluene System at 175° C. 6l 15• Vapor-liquid Equilibrium "x-y" Composition Diagram for Methanol-toluene System at 200° C. 62 16. Vapor-liquid Equilibrium Pressure Composition Diagram for Methanol-toluene System at 200° C. 63 17- Vapor-liquid Equilibrium ."x-y" Composition Diagram for Methanol-toluene System at 225° C. 6k 18. Vapor-liquid Equilibrium Pressure Composition Diagram for Methanol-toluene System at 225° C 65 19- Vapor-liquid Equilibrium "x-y" Composition Diagram for Methanol-toluene System at 250° C. 66 v i i i Figure Page 20. Vapor-liquid Equilibrium Pressure Composition Diagram for Methanol-toluene System at 250° C. 67 21. Equilibrium Ratio as a Function of the Logarithm of the Pressure for Methanol 68 22. Equilibrium Ratio as a Function of the Logarithm of the Pressure for Toluene 69 23- Azeotrope Composition 72 2k. Pressure-temperature Diagram for Methanol-toluene Azeotrope 73 25- Ratio of Activity Coefficients in the System Methanol-toluene at 125° C. 81+ 26. Ratio of Activity Coefficients in the System Methanol-toluene at 150° C 85 27- Ratio of Activity Coefficients in the System Methanol-toluene at 175° C 86 28. Ratio of Activity Coefficients in the System Methanol-toluene at 200° C- 87 29- Ratio of Activity Coefficients in the System Methanol-toluene at 225 C 88 30. Ratio of Activity Coefficients in the System Methanol-toluene at 125° C. 89 31- Ratio of Activity Coefficients in the System Methanol-toluene at I5O0 C- 90 32. Ratio of Activity Coefficients in the System Methanol-toluene at 175° C. 91 I - l . Nichrome Detecting Wire Bridge 1-6 1-2. Measuring Head Drive Circuit I - l l 1-3- Temperature Control Circuit I-18 I-k. Thermocouple Position for Temperature Survey 1-21 1-5- Transfer Flasks 1-29 1-6. Intermediate Storage Flask and Mercury Storage Flask 1-30 1-7- Sample Collection Tubes 1-32 1-8. Assembly of Equilibrium Cell I-3U ix Figure Page 1-9- Assembly of Equilibrium Cell I-35 I-10. Assembly of Stirrer and Measuring Head I-36 1-11. Assembly of Lower Cell, Reduction Gear and Universal Joint I-37 1-12. Assembly of Measuring Head Rod Driving Equipment I-38 1-13. Part Number 1 I-39 1-14. Parts Number 5, 6, and 7 1-40 1-15. Parts Number 8 and ^ I-4l I-l6. Parts Number 2, 3, and 4 1-42 1-17- Parts Number 10, 11, 12, and 13 1-43 1-18. Parts Number 16 to 23 1-44 1-19. Parts Number 25 to 31 1-45 1-20. Parts Number 23, 36, 62, and 68 1-46 1-21. Parts Number 40 and 4 l 1-47 1-22. Parts Number 42 and 43 1-48 I-23. Parts Number 44, 45, 46, and 47 1-49 1-24. Parts Number 50, 51, 52, and 53 I-50 1-25. Parts Number 84, 88, 93, and 94 1-51 1-26. Parts Number 85, 86, 89, and 91 1-52 1-27. Parts Number 49, 75, 77, and 78 I-53 IV-1. Chromatographic Calibration for Benzene-toluene System IV-5 IV-2. Chromatographic Calibration for Benzene-toluene System IV-6 IV-3- Refractive Index Calibration Curve for the Methanol-toluene System IV-9 LIST OF TABLES x Table Page 1 . Vapor Pressure of Benzene 2 5 2 . Vapor Pressure of Toluene 2 9 3 - Measured and C a l c u l a t e d Vapor Pressures of Toluene 3 0 k. Vapor Pressure of Methanol 3 5 5 - Measured Vapor Pressure of Methanol 3 6 6 . V a p o r - l i q u i d E q u i l i b r i a o f Benzene-toluene System at 2 0 0 ° C 1+0 7 . V a p o r - l i q u i d E q u i l i b r i a of Benzene-toluene System at 2 0 0 ° C 1+2 1 8 . V a p o r - l i q u i d E q u i l i b r i a of the Benzene-toluene System at 2 0 0 ° C as Measured.in t h i s P r o j e c t 1+3 9 - A c t i v i t y C o e f f i c i e n t s f o r Benzene-toluene System at 2 0 0 ° C. 1+8 1 7 - C a l c u l a t e d and Measured Values f o r the Methanol-toluene Azeotrope 7 0 2 8 . Values of the I n t e g r a l (\r. <^ /Yt 6*-, f o r Methanol-toluene System •* ' 9 2 2 9 - C o e f f i c i e n t s of R e d l i c h and K i s t e r C o r r e l a t i o n f o r Methanol-toluene System 9 5 I - l . R e s u l t s o f Temperature Survey of E q u i l i b r i u m C e l l 1 - 2 0 1 - 2 . C a l i b r a t i o n Supplied by Manufacturer f o r Heise Pressure Gauge Number 2 1 8 7 0 ' I - 2 1 + 1 - 3 - C a l i b r a t i o n Determined Using the Barnett Dead Weight Tester Model Number MKT f o r Gauge Number 2187O 1 - 2 5 1-1+. C a l i b r a t i o n Supplied by Manufacturer f o r Heise Pressure Gauge Number 2 1 8 6 9 1 - 2 6 1 - 5 - C a l i b r a t i o n Determined Using the Barnett Dead Weight Tester Model Number MKT f o r Gauge Number 2 1 8 6 / 9 1 - 2 7 I V - 1 . Chromatographic C a l i b r a t i o n s f o r Benzene-toluene System I V - 3 I V - 2 . Accuracy of C a l i b r a t i o n of Chromatograph IV-1+ I V - 3 - T y p i c a l Refractometer C a l i b r a t i o n f o r Methanol-toluene System at 2 0 ° C I V - 6 xi Table Page 10. Vapor-liquid Equilibrium Data for the Methanol-toluene System at a Pressure of one Atmosphere VI-1 11. Vapor-liquid Equilibrium Data for the Methanol-toluene System at 125° C. VI-2 12. Vapor-liquid Equilibrium Data for the Methanol-toluene System at"150° c . VI-3 13. Vapor-liquid Equilibrium Data for the Methanol-toluene System at 175° c . Yl-k Ik. Vapor-liquid Equilibrium Data for the Methanol-toluene System at 200° C VI-5 15• Vapor-liquid Equilibrium Data for the Methanol-toluene System at 225° C. VI-6 l6. Vapor-liquid Equilibrium Data for the Methanol-toluene System at 250° C. VI-7 18. Activity Coefficients for Methanol at 125° C. VI-8 19. Activity Coefficients for Toluene at 125° C VI-9 . 20. Activity Coefficients for Methanol at I5O 0 c . VI-10 21. Activity Coefficients for Toluene at 1500 C. VI-U 22. Activity Coefficients for Methanol at 175° C. VI-12 23. Activity Coefficients for Toluene at 175° c . VI-13 2k. Natural Logarithm of the Ratio of the Activity Coefficients for Methanol and Toluene at 125° C VI-14 25- Natural Logarithm of the Ratio of the Activity Coefficients for Methanol and Toluene at 150° C VI-15 26. Natural Logarithm of the Ratio of the Activity Coefficients for Methanol and Toluene at 175° C VI-l6 27- Natural Logarithm of the Ratio of the Activity Coefficients for Methanol and Toluene Calculated by the Method of Black for Temperatures of 200° and 225° C. VI-17 30. Comparison of Activity Coefficients Calculated from Equation of Black and from Redlich and Kister Correlation for Methanol and Toluene at 125° C VI-18 Comparison of Activity Coefficients Calculated from Equation of Black and from Redlich and Kister Correlation for Methanol and Toluene at 1 5 0 ° rj. Comparison of Activity Coefficients Calculated from Equation of Black and from Redlich and Kister Correlation for Methanol and Toluene at 1 7 5 ° C. Comparison of Activity Coefficients Calculated from Equation of Black and from Redlich and Kister Correlation for Methanol and Toluene at 2 0 0 ° rj. Comparison of Activity Coefficients Calculated from Equation of Black and from Redlich and Kister Correlation for Methanol and Toluene at 2 2 5 ° C. ACKNOWLEDGEMENT - x i i i The author wishes to acknowledge the assistance and advice given by Dr. L. W. Shemilt, under whose direction this project was carried out, and to thank Dr. Shemilt for his constant encouragement. He wishes to acknowledge, as well, the help received from Dr. D. S- Scott in Dr. Shemilt's absence. The author is also deeply indebted to Dr. J. S. Forsyth/ for his extensive and very practical help in solving the many problems associated with the design and construction of the equipment used in this work. The Ph.D committee which supervised the work reported in this thesis consisted of Dr. L. W. Shemilt, Dr. D. S. Scott, Dr. J. S. Forsyth, Dr. H. M. Daggett, and Dr. N- Epstein. This committee was chaired by Dr. L. W. Shemilt and, in his absence, by Dr. D. S. Scott. The drawings and graphs included in this thesis have been copied from tracings prepared by,the author's wife, Mrs. D. Whittle. Personal assistance has been given to the author during the course of this work by The Standard O i l Company of British Columbia in the form of a Standard O i l Fellowship, by Canadian Industries Limited in the form of a C I. L. Fellowship, and by the National Research Council of Canada in the form of an N. R. C Studentship. Financial support for the work carried out in this project came from the National Research Council of Canada through a grant given to the Department of Chemical Engineering at this University. INTRODUCTION 1. Importance of Vapor-Liquid Equilibrium Data When a liq u i d mixture of two or more components is pa r t i a l l y vaporized, the composition of the vapor produced is normally different from that of the liquid i t leaves. It i s this change in composition with change in phase that forms the basis of such separation processes as d i s t i l l a t i o n and absorption, and a quantitative knowledge of the change i s therefore essential for the analytical treatment of these processes. Data descriptive of the composition of the li q u i d and vapor phases, as a function of temperature and pressure, when the two phases are in equilibrium with respect to the movement of the molecules of each component from one phase to the other, are used as the basis of this analytical treatment. The importance of the a v a i l i b i l i t y of such vapor-liquid equilibrium data is obvious. The accurate measurement of solution properties requires very careful work, and because i t is clearly impossible to investigate a l l systems at a l l conditions of temperature and pressure, many attempts have been made to develop a solution theory by means of which such data may be predicted from the properties of the pure components or from a limited amount of data on solutions. The testing of such theories required extensive tabulations of thermodynamic data for types of systems of no immediate industrial importance and for systems under conditions of temperature and pressure not normally encountered in industrial practice. The measurement of data to f i l l this need has become an important f i e l d of study. Methods of Measuring Vapor-Liquid Equilibria under Pressure Many methods have been used to experimentally determine vapor-liquid 7 equilibria-Robinson and G i l l i l a n d have cl a s s i f i e d these methods under the following headings: 2. a) Dynamic Flow Method b) Dynamic D i s t i l l a t i o n Method c) Circulation Method d) Continuous D i s t i l l a t i o n Method e) Dew-and-Bubble-Point Method f) Bomb Method Of these six methods, only the last four are important for measurements at pressures above atmospheric, and therefore only these four are discussed below. With the circulation method, the mixture to be studied is placed in an evacuated chamber. The vapor is collected from above the liquid and circulated back through the liquid u n t i l the composition of both phases becomes constant. Although the method is basically simple, several precautions have to be taken in order to obtain accurate results. The entire system must be leak-free to prevent the progressive loss of material and the resulting progressive change in equilibrium properties. The total quantity of each phase must remain constant,and therefore the apparatus must be carefully thermostated, and the pump used to circulate the vapor must have a very small displacement. Since the pressure at the bottom of the liquid phase, where the vapor enters, is different from that at the top of the phase, where the vapor leaves, the equilibrium composition is also different at the two levels. This change in equilibrium composition with pressure can be important in the c r i t i c a l region. Q The circulation method has "been successfully used by Aroyan and Katz at pressures up to 8000 pounds per square inch. The continuous d i s t i l l a t i o n method is a less accurate but simpler form of measurement. The vapor is collected from above the boiling liquid, condensed, and returned to the s t i l l as a liquid. The method suffers from two disadvantages: there i s some doubt as to whether or not a boiling liquid produces an equilibrium vapor, and since the condensed vapor returned to the s t i l l has a different composition than the s t i l l liquid, the two must be thoroughly mixed before any is vaporized. These d i f f i c u l t i e s can be minimized by vaporizing the condensate 3-before i t i s returned to the s t i l l , but care must be taken that a l l of the condensate is vaporized and that none of i t i s super heated. Measurements to 1000 pounds per square inch have been made with continuous d i s t i l l a t i o n equipment 3 by Griswold et a l . With the dew-and-bubble-point method, a vessel of variable volume is f i r s t evacuated and then f i l l e d with a sample of known composition. The volume of the c e l l is then varied in order to obtain the dew and bubble points as a function of pressure at constant temperature. These two points are determined either by observation or by plotting the pressure-volume isotherms. | Equilibrium data are obtained from the intersection of dew-and-bubble-point curves plotted as lines of constant composition on a temperature-pressure diagram. While extensive use has been made of this method for the deter-mination of vapor-liquid equilibrium data at elevated pressures, the method suffers from three disadvantages. Very pure material must be used and in particular i t must be free of permanent gases. Under certain conditions, the dew and bubble points are not sharply defined. Probably the most serious limitation of the-method is that i t i s restricted to the study of binary systems. Kay and co-workers ^' " L 1 have used this method with a visual type of apparatus. 12 Sage and Lacey have used a non-visual type at pressures as high as 10,000 pounds per square inch. In the bomb method,a sample is placed in an evacuated vessel, and the vessel is then agitated at constant temperature u n t i l equilibrium i s obtained between the liq u i d and vapor phases. Once the two phases are at equilibrium, samples of each are taken. Often the samples are displaced with an equal volume of an inert material such as mercury to prevent a change in the eq u i l i -brium properties while sampling. Although accurate results are possible with this method, care must be used in order to obtain them. Any trace of liquid in the vapor sample w i l l lead to serious errors in the vapor composition under most conditions because of the relative densities of the two phases. Sage and Lacey ^ have used this method to measure vapor-liquid equilibria at pressures as high as 10,000 pounds per square inch. Of the four methods discussed above, the last two, the static methods, have been used for the majority of the measurements reported at pressures significantly above atmospheric. A detailed review of equipment used with these two methods,and a shorter description of equipment used for the other methods was given earlier by the author. J Methods of Testing Vapor-Liquid Equilibrium Data The state of a single phase, apart from i t s size, can be described by the set of intensive properties consisting of the temperature, pressure and chemical potential of each component. Since the number of degrees of freedom ,"D" of a "C" component mixture with'Vphases is given by the phase rule ^ as D + N = C + 2 (l) the number of degrees of freedom of a single phase is "C + 1". The "C + 2" variables, temperature, pressure and chemical potential cannot therefore a l l be independent. The relationship between these quantities, the derivation of which 15 is given in most standard texts on thermodynamics, e.g. Guggenheim , i s known as the Gibbs-Duhem equation and may be written as SdT-VdP+Sxi 1/^=0 (2) The ;Gibbs-Duhem equation, or relationships derived from.it, is used as the basis of a l l thermodynamic consistency checks of vapor-liquid equilibrium data. Van Ness ^ has rearranged equation (2) using standard thermodynamic relations to a more convenient form for the 'testing of experimental data. In terms of fugacities: -^ TdP + i ^ H , d T - Z : x L d l n f t (3) or in terms of acti v i t i e s ^ d P - f ± , d T = Z * L d L * L (U) where A V and AH are the change of volume and change of enthalpy of mixing per mole of solution formed. IT 1 ft Carlson and Colburn and more recently E l l i s and Bourne have reviewed the methods of testing vapor-liquid equilibrium data. The most commonly used tests may be divided into two groups: the slope tests and the area tests. The slope test has been used for many years and is discussed in most 19 standard texts on solution thermodynamics - e.g. Dodge . Under conditions of constant temperature and pressure equations (3) and (k) may be written as 1 x-t d k - f =0 ( 5 ) I * L d \ o o ( 6 ) or for a binary solution dint'. _ d 1 " * ! (7\ X , ~ d ^ r " ^ "a^r ( 8 ) Although for a binary solution the data cannot be obtained with both temperature and pressure held constant, for many systems under isothermal conditions the effect of pressure changes can be ignored, and equations (7) and (8) can be used to check such isothermal data. The slope test has the advantage that the consistency of the data can be tested at every composition, but i t suffers from the serious disadvantage that i t is d i f f i c u l t to measure slopes accurately and thus that the u n c e r t a i n t y i n the slope measurements i s o f t e n as gr e a t r a s the i n c o n s i s t -encies i n the data. 20 The o r i g i n a l equations f o r the auea t e s t were d e r i v e d by Herington , 21 S Coulson and Herington , and R e d l i c h and K i s t e r . The equations have been g e n e r a l i z e d by Van Ness to give whereZ^F^ i s the excess f r e e energy of mixing. F o r . b i n a r y s o l u t i o n s a t constant temperature and pressure,equation (9) reduces to the f a m i l i a r R e d l i c h and K i s t e r y equation. In = O (|0) 18 E l l i s and Bourne recommend the area t e s t as the best method f o r checking the consistency of isot h e r m a l v a p o r - l i q u i d e q u i l i b r i u m data. I t does s u f f e r , however, from the disadvantage t h a t a p o i n t check of the data i s not obtained. C a l c u l a t i o n of F u g a c i t i e s and A c t i v i t y C o e f f i c i e n t s from Experimental Data by Exact Methods 22 The f u g a c i t y of a pure substance can be defined by the r e l a t i o n s h i p [ RT ln-£* J ( V - ^ ) d P ] (VO 0 T and the f u g a c i t y can t h e r e f o r e be evaluated from pressure-volume-temperature (PVT) data f o r the substance. Equation ( l l ) as w r i t t e n above i s g e n e r a l l y r e s t r i c t e d t o the e v a l u a t i o n of f u g a c i t i e s of substances i n the vapor s t a t e . The equation i s d i f f i c u l t to evaluate f o r l i q u i d s since the i n t e g r a l covers the pressure range from the h i g h l y attenuated gas s t a t e as the pressure approaches zerOjto the system pressure where the substance e x i s t s as a l i q u i d . The 7-fugacity of a liq u i d i s generally evaluated instead in two parts. The fugacity of the saturated vapor is f i r s t calculated, and thenjmaking use of the fact that at saturation the liq u i d and vapor fugacities are equal, the fugacity of the liquid i s found. Thus for a liq u i d R>ap 0 .P •Jo? ••• -Jp /n T- ~~ RT f V L 1 Q d P (13) Py/a p ^ P Pv/op 0M The fugacity of a component in a mixture can be defined in an ahalagous manner by the expression o If P V T data for the mixture at the specified composition at pressures from zero to that', of, the system are available, the integral in equation ( 1 5 ) can be evaluated and the fugacity calculated. Activity coefficients are calculated from fugacities according to the definition where f£ is the fugacity of the pure component at the same temperature, pressure and physical state as the component in the mixture. Fugacities and activity coefficients used in the testing of vapor-liquid equilibrium data are normally calculated for the liquid phase. The component fugacity f^ i s therefore that of the component in the liquid. Since, 8. however, the fugacity of a component in the li q u i d phase is equal to that of the component in the vapor phase under conditions of phase equilibrium, the fugacity i s evaluated from the vapor properties. Calculation of Fugacities and Activity Coefficients from Experimental Data by Approximate Methods Normally when vapor-liquid equilibrium data are measured, only the temperature, pressure, and composition of the co-existing phases are determined. Since the calculation of activity coefficients by exact methods requires a knowledge of part i a l molal volume data, and since such data are almost never available, the coefficients are usually calculated by some approximate method based on the properties of the pure components. Very often the problem is further complicated by,the fact that only a few of the properties of the pure components are known. An excellent review of the methods of calculation to be used under lU these circumstances is given by Reid and Sherwood Activity coefficients are calculated most simply in terms of deviations from Raoult's law. If i t can be assumed that the fugacity of each component in the vapor phase is equal to the partial pressure of that component as given by Dalton's law, and i f i t can be assumed that the fugacity of the pure li q u i d at the temperature and pressure of the mixture is given by thei vapor pressure of that liquid, then YA - M (17) This definition of activity coefficients i s generally only valid under low pressures, and then only for solutions with relatively ideal vapor behavior. The Lewis and Randall 4 rule gives a relation of more general validity. If a mixture forms an ideal solution, i.e. no volume changes occur on mixing, then i t can be shown that the fugacity of a component in a mixture is equal 9-to the product of the mole fraction of that component and the fugacity of the pure material at the same temperature, pressure, and physical state. f ^ - ^ (18) YA = (19) In order to use this rule to calculate activity coefficients, the fugacity of the pure vapor at the temperature and pressure of the mixture must he calculated for each component. If PVT data or equations of state are available for the pure components, equation ( l l ) can be used to calculate these fugacities. If such data is not available,then one of the generalized correlations of thermodynamic properties can be used. Several such correlations have been proposed. Charts based on reduced temperature and reduced pressure have been 23 Oh published by Newton J and Gamson and Watson . More recently, correlations based on an additional third parameter have been published. Lyderson, Greenkorr. and Hougen ^ used the c r i t i c a l compressibility, Pitzer et a l ^ used the 27 reduced saturation pressure at a reduced temperature of 0.7> and Riedel used the slope of the vapor pressure curve at the c r i t i c a l temperature for this additional parameter. The fugacities of the pure components can also be calculated from a 28 generalized equation of state. Redlich and Kwong have proposed such an equation based on reduced temperature and reduced pressure}•Black ^ , one based on reduced temperature and reduced pressure, and in the case of polar substances 29 one specific constant; and Hirschfelder et a l , one based on reduced temperature, reduced pressure, c r i t i c a l compressibility and the slope of the vapor pressure curve at the c r i t i c a l point. The Lewis and Randall rule has been recommended for use at values of 10. the reduced pressure of up to 0.6 At higher values of the reduced pressure or for non-ideal solutions, a different method must be used to evaluate the fugacity of a component in a solution. Most of the important methods involve either the calculation of the constants for an equation ,of state for the mixture using the constants for the pure components, or the calculation of mixture parameters from the parameters of the pure components and the use of these mixture parameters and a generalized correlation to calculate the properties of the mixture. If the specific constants for an equation of state are available for each of the components present in a mixture, these constants can be combined to give an equation of state for the mixture. A review of the methods of combining constants for the Van der Waals 3°, Keyes 3 1 , Beattie-Bridgeman 3 2 ; . Benedict-Webb-Ruben and v i r i a l - ^ equations of state, and of the accuracy of the results,has been given by Beattie ^5, jf the constants for a specific equation of state are not known, one of the generalized.equations can be used. Both Redlich and Kwong and Black give rules for combining the constants in their generalized equations. From an equation of state for the mixture, whether generalized or specific, the fugacity of each component can be calculated using standard thermodynamic relationships. If, instead of an equation of state, one of the generalized correlations is' used to evaluate the properties of a mixture, then the correlation patameters for the pure components must be combined to give those for the mixture. Hougen, Watson, and Ragatz ^ and Lewis and Randall 37 have discussed the combination of these parameters. The fugacity of the components of the mixture can then be calculated from the properties of the mixture using the relationship proposed by Joffe ^ . 11. Systems Studied and Type of Equipment Used A considerable amount of vapor-liquid equilibrium data is available in the literature as exemplified by the compilations of vapor-liquid equilibriun data of Ju Chin Chu^ and Hala et a l ^ , the compilation of physiochemical constants by Timmermans^ '"'", and the compilations of workers who have measured data at super atmospheric pressures by Comings^, Smith^, Newitt^- This is however a shortage of data for non-ideal systems at conditions of elevated temperatures and pressures. Since the testing of solution theories requires experimental data for systems known to be non-ideal, and data for conditions in which the deviations from ideality are large, and d i f f i c u l t to treat analytically, i t was decided that i t would be of interest to investigage such a system and to make vapor-liquid equilibrium measurements over most of the range of temperature and superatmospheric pressure in which the vapor and liq u i d phases could co-exist. The non-ideal system chosen for study was the toluene-methanol binary. This particular system was chosen for three reasons. Many measurements had previously been made at this university on systems formed from the normal alcohols with benzene aid toluene at atmospheric pressure ^5>^6,U7,48,U9,50^ i t seemed logical to continue to work with a system in this group. Both vapor-liquid equilibrium and enthalpy measurements 51>52,53 a + J elevated pressures had been reported for the closely related benzene-methanol binary system, and therefore measurements were available for comparison with those obtained in this project. Finally, to the author's knowledge, the only published data on the toluene-methanol system is the atmospheric pressure data of Benedict, Johnson, Solomon, and Rubin 12. A limited number of vapor-liquid equilibrium measurements were also made on the toluene-benzene system. These measurements were made because i t was planned to measure, at a later date, the benzene-toluene-methanol system, and therefore benzene-toluene binary data measured on the same equipment would be very valuable for checking ternary prediction methods. Since the benzene-2 3 toluene system had previously been investigated at elevated pressures ' J these measurements also served as a test of the experimental method and equipment. Several factors were considered when choosing the design of the equipment used in this project. Since i t was planned to obtain equilibrium values of theoretical interest, i t was important that the apparatus should give as accurate results as possible. As well, however, i t was desirable that the design be kept simple since the apparatus had to be bu i l t to withstand elevated temperatures and pressures. On the basis of these two considerations, i t was decided that of the various types of equipment used to measure vapor-liquid equilibrium data, one of the two static methods, the dew-and-bubble-point method and the bomb method, would be most suitable. Since the dew-and-bubble-point method is restricted to the measurement of binary systems, and since i t was planned that eventually the apparatus would be used for the study of ternary systems, a bomb-type design was chosen. APPARATUS AND PROCEDURE 13. Apparatus Introduction 13 55 In an earlier research project ->>•'•' involving the author, a simplified version of an apparatus designed by Sage and Lacy"^ " for phase equilibrium measurements was proposed and b u i l t . When this simplified apparatus was tested, i t was found to suffer from several major defects and therefore had to be rebuilt before i t could be used in the present project. Since the rebuilding was done by correcting each defect separately rather than by discarding the equipmeht and returning to the original design of Sage and Lacey f many changes from that design have occurred. Most of the features resulting from these changes are different to but not necessarily better than those described by Sage and Lacey"'". The f i n a l design provided an apparatus suitable for continuous use o at temperatures up to 200 c. and for intermittant use at temperatures up to 250° C. Pressures as high as 1500 p.s.i.a. and, with one very small modification, as high as 5000 p.s.i.a. could be used. A detailed description of the apparatus is given in Appendix I. General Arrangement of Apparatus The general arrangement of the apparatus is shown in Figure 1. The parts of the equipment designed to withstand elevated pressures are shown by heavy lines and associated glassware by light lines in this diagram. The equilibrium c e l l A was immersed in a constant temperature bath U. This c e l l contained a magnetically driven sti r r e r (not shown) and, at the end of tube K, a detecting device which could be used to measure the levels of the vapor-liq u i d and organic-mercury interfaces. Connected to the c e l l , and also located in the bath, was a sampling tube assembly D by means of which samples Figure 1 15-of the vapor and liquid phases could he taken. The bath and thus the c e l l temperature was measured with a platinum resistance thermometer and the c e l l pressure by means of two pressure gauges F and G. Bath Temperature Control and Measurement The o i l bath temperature was controlled with a c i r c u i t based on a Leeds and.Northrup platinum "thermohm" (resistance thermometer) and Mueller Bridge. The thermohm was located in the bath, and any change in i t s resistance from that corresponding, to the desired bath temperature resulted in an out-of-balance signal from the Mueller Bridge. This out-of-balance signal was used, after amplification, as the input to an off-on controller. The temperature which was recorded in this work was that indicated by a platinum resistance thermometer suspended from the bath top so that the sensing element was located near the centre of the equilibrium c e l l . The thermometer was calibrated in 1959 by the National Research Council of Canada. Since the temperature that was measured was that of the o i l surrounding the equilibrium c e l l and not that of the c e l l i t s e l f , an auxilliary temperature survey was carried out, using a sensitive thermocouple-bridge assembly in which the temperature at seven points on the c e l l and at one point adjacent to the platinum thermometer were measured. This survey showed that the temperature at the-thermometer location and that of the c e l l opposite that location were essentially identical. The results of the survey are given in Appendix I. The temperature of the bath was controlled to +0.05° C. and measured to +0.01° C However, because of the poss i b i l i t y of small temperature gradients in the c e l l walls under some conditions, temperature values are ^ o reported to ,+0.1 C. 16. Pressure Measurement The pressure in the equilibrium c e l l was measured with the aid of two Bourdon-tube type pressure gauges, F and G, connected to the c e l l through a mercury-filled line and valves V 6, V 7> and V-8. The gauges were calibrated for the-range 0 to 150 p.s.i.g. and 0 to 1500 p.s.i.g. respectively and were accurate - from +8 to -l6 mm. of mercury for the f i r s t gauge and +80 mm. for the second. Calibrations for both gauges are given,in Appendix I. Equilibrium C e l l Volume and Location of Interfaces The volume of the equilibrium c e l l which could be effectively used was that above the mercury surface in the c e l l , and this volume could be varied by the addition or removal of mercury through valves V 8 and V 9* The mercury for this use was stored in mercury storage flask J, connected to the equipment through valve V 11, and was transferred between the flask and c e l l using a pressure generator I. The level of the mercury surface-in the c e l l was determined by a resistance bridge technique using a detecting or measuring head welded to the upper end of the stainless steel tube K. A 0.005-in. diameter nichrome wire -was horizontally suspended from the head. This wire was connected to a Wheatstone bridge, and the sudden change in i t s resistance as i t was lowered into the mercury phase was used to detect the level of the interface. (This resistance change was caused by the difference in e l e c t r i c a l conductivity of the solvent and mercury phases.) The detecting head could also be used to determine the location of the vapor-liquid interface. Since the nichrome wire, when connected to the Wheatstone bridge, carried a small current, i t s temperature and thus i t s resistance was dependent upon the thermal conductivity of the surrounding 17-phase. The thermal conductivity of the liquid and vapor phases were normally different, and therefore the wire resistance changed when the wire was lowered from the vapor into the liquid phase. The measuring head could be raised or lowered inside the c e l l by means of a drive assembly. This assembly contained an electric motor which was used to rotate a lead screw connected to the measuring head tube K. As the lead screw rotated, i t also travelled vertically, causing the head to move vertical l y as well. The travel of the lead screw was limited, in both directions, by microswitches so that the head could not be accidentally damaged. The lower end of the lead screw was connected to a rod L which had the same outside diameter as the measuring head tube. This rod passed through a teflon packed gland into a mercury-filled volume compensation c e l l E. Since this cell,was connected, through valve V 5> "to the bottom of the equilibrium c e l l , no change occurred in the equilibrium c e l l net volume as the measuring head was raised or lowered. Samples of Co-existing Phases Samples of the vapor and of the liquid'phases could"be taken from the equilibrium c e l l through valves V,1 and V 2 respectively and isolated in the sampling tube aseembly D. Vapor samples were replaced, as they were taken, with an equivalent volume of mercury and thus with no disturbance of the phase equilibrium in the c e l l . Liquid samples were taken by a rapid removal of a small amount of the li q u i d phase. Experimental Procedure Preliminary Steps Several preliminary steps were required before the assembled equipment was used. It was f i r s t carefully pressure tested and cleaned and 18. was then evacuated. While s t i l l under vacuum, a l l of the pressure lines and fit t i n g s , with the exception of the equilibrium c e l l , were completely f i l l e d with deaerated mercury from storage flask J, and the equilibrium c e l l , i t s e l f was f i l l e d to a level 1 cm. above the lowest tubing connection. With this level as datum and with the c e l l open to the atmosphere, the scales on the two gauges were set to•read zero pressure. In use, the gauge readings had to be corrected in order to obtain the equilibrium c e l l absolute pressure. A detailed description of these preliminary steps, of the calculation of the gauge corrections, and of the use of the equipment i s given in Appendix II. Introduction of Material to Cell • The system to be studied was introduced to the apparatus through a vacuum d i s t i l l a t i o n transfer apparatus. This apparatus consisted of three identical sections, one of which is shown in Figure 1, (Flasks F 2, F 3, and F 10), and was provided with a single flask F 1, common to a l l three sections. One component of the system which was to be studied was f i r s t poured into the one-litre flask F 10 along with a small quantity of drying agent, and the flask was joined to the transfer apparatus through a mercury-sealed ground glass joint. After most of the air had been pumped out of the flask with a mercury diffusion pump, the solvent was d i s t i l l e d from the flask to the connected transfer vessel F 2 and frozen with liquid nitrogen. Any remaining a i r was pumped off from above the frozen solvent, and the solvent was vacuum d i s t i l l e d to the adjacent transfer vessel F 3- It was again frozen, evacuated, and f i n a l l y d i s t i l l e d to vessel F 1. The other component of the system to be studied was introduced in a similar manner using a different section of the transfer apparatus. It also was transferred to vessel F 1 in the f i n a l d i s t i l l a t i o n . The mixture prepared in F 1 in this manner was f i n a l l y vacuum d i s t i l l e d to the intermediate storage flask T. 19-The solvent mixture was transferred from the storage flask T to the previously evacuated equilibrium c e l l in the following manner. Stopcock S 13 was closed, valve V k and the mercury cut-off valve S 1 were opened, and hot water was poured into the center of the flask. The solvent, which was warmed by the water, was forced, by means of i t s own vapor pressure, into j the equilibrium c e l l . Preparation for Sampling Once the solvent mixture had been introduced to the c e l l , valve V h was< closed and the i n i t i a l steps required before phase equilibrim measurements could be made were carried out. The bath was heated to the desired temperature, and the lines connecting the sampling tube assembly D to the equilibrium c e l l were then purged with mercury." After this purging ?the sampling tube assembly was evacuated to remove any solvent that might have escaped from the cell ?and the assembly was r e f i l l e d with mercury. A check was made to see that the mercury level in the c e l l lay between the center and lowest sampling ports and that the liquid-vapor interface lay between the liq u i d and vapor sampling ports. Valve V 3 was then opened, connecting the mercury in the sampling tube to that in the equilibrium c e l l , and f i n a l l y the magnetic stirrer was turned on. Measurements A period of one hour was allowed after the f i n a l adjustments were made to the c e l l before any measurements were taken. At the end of this time, the c e l l temperature and pressure and room temperature and pressure were determined and recorded. A sample of the vapor phase was next taken by the following method. Valve V 1 was opened, allowing vapor from the c e l l to flow into the assembly and to displace an equivalent volume of mercury from the assembly through V 3 20. into the c e l l . Valves V 1 and V 3 were then closed. After the sample had "been isolated in this manner, i t was transferred by vacuum d i s t i l l a t i o n through valve V 13 to sample collection flask L. The sample was collected by freezing i t in the bottom of flask L and sealing off the flask above the frozen sample. When the vapor sample had been collected, the sampling tube assembly and sample collection flask were re-evacuated in preparation for taking a li q u i d sample. Valve V 2 was then cracked, and a small amount of the liquid phase was allowed to bleed out of the c e l l . The sample was collected in flask M and removed by the method used for the vapor samples. Since the samples obtained were completely sealed, they could be stored indefinitely before analysis. Analysis Samples from the two binary systems studied in this work were analysed using either a gas chromatograph (benzene-toluene system) or an Abbe refractometer (toluene-methanol system). Details of the method of analysis for both systems is given in Appendix IV. MATERIALS 21 Benzene Eastman "spectro-grade S777" benzene was used, after purification, for the vapor pressure and phase equilibrium measurements. This material was stored for several days over calcium chips and then charged, along with carbon boiling chips and fresh calcium, to the s t i l l pot of a Todd "Precise Fraction-ation Assembly". This s t i l l , using a 25-mm. diameter column packed for 90 cm. with U-mm. diameter glass helices, was equivalent to 42 theoretical plates at total reflux. The benzene was d i s t i l l e d at total reflux for several hours and was then removed at a reflux ratio of 50 to !• The f i r s t and f i n a l 20 percent cuts of the d i s t i l l e d solvent were rejected, and the center cut (boiling point approximately 80.1 to 80.2° C ) was stored over sodium-lead alloy u n t i l used. The purity of the center cut of the benzene was che.cked on a Beckman "GC2" gas chromatograph, using a six-foot column of 8N8 Flexol Plasticizer on firebrick. No impurities were detected, with the chromato-graph set for minimum attenuation, at either 70° or 100° C. Toluene Eastman "X-325" toluene was dried with sodium-lead alloy and used, without further purification, for vapor pressure and phase equilibrium measurements. No treatment other than drying was considered necessary, as a check of the purity of the dried material using the gas chromatograph and column described above, with the chromatograph set for minimum attenuation, showed no impurities. This result agreed with that obtained by the manufact-urer, who stated that the toluene had been tested by gas chromatography and that no impurities had been found. 22. Methanol Eastman spectro-grade methanol was purified for the methanol vapor pressure and phase equilibrium measurements. The solvent was f i r s t stored for several days over, "Drierite" (anhydrous calcium sulphate) and then charged, along with fresh "Drierite", to the s t i l l pot of the d i s t i l l a t i o n column described earlier. Purified nitrogen, stated by the manufacturer to be 99'99 percent pure, was introduced to the column just above the s t i l l pot to prevent the formation of formaldehyde . The methanol was d i s t i l l e d at total reflux for several hours and was then removed at a reflux ratio of 50 t o l . The i n i t i a l and f i n a l 20 percent cuts of the d i s t i l l a t e were rejected and the center cut was stored in a nitrogen atmosphere u n t i l used. A f i n a l drying of the purified alcohol occurred when it,was poured into a one-litre flask and sealed to the transfer apparatus. "Drierite" was added to the flask along with the methanol, and the alcohol and "Drierite" mixture was allowed to stand for at least one day before the methanol was d i s t i l l e d into the transfer vessels. The methanol purity was checked by gas chromatograph analysis according to the details given in Appendix III. The only impurity found was water, and the water content was found to be less than 0.1 mole percent. VAPOR PRESSURE MEASUREMENTS 23-Benzene Previous Work A considerable number of measurements have been made on the vapor pressure of benzene. Of these measurements, those of Bender, Furukawa, and 57 58 ? Hyndman ; Gornowski, Anick, and Hixson ; von Huhn ; and Griswold, Andres, and Klein^ are pertinent for comparison to the values reported in this project. Bender et al-^7 have measured vapor pressures at temperatures from 100° C. to the c r i t i c a l point. Reported values were calculated from an equation given by the authors which they estimated to be accurate to about 59 0.1 percent. Gornowski et a l reported carefully measured benzene vapor pressure values for temperatures from 130° C. to the c r i t i c a l point. -No 5 9 estimate was given of the accuracy of the measurements. Young measured the vapor pressure of benzene at temperatures from -10° C. to the c r i t i c a l point. Reported values were calculated from an equation. No estimate was given of 2 the accuracy of 'the equation. Von Huhn has reported benzene vapor pressures at-temperatures from 100° to 300° C. The vapor pressures were measured to within kO mm. of mercury (k mm. at pressures up to 7500 mm.),and temperatures o to within about 0.1 C. Reported values were obtained by that author from -3 smoothed curves. Griswold et a l J measured the vapor pressure of benzene at five temperatures between 121° and 255° C. in the course of their study of the benzene-toluene binary system. Pressures were determined to within 50 mm. of mercury and temperatures with an accuracy of approximately 0.5°'C The procedure used by each of the above workers to purify benzene for the vapor pressure measurements is described in Appendix VII. The vapor pressure data of the above workers for the temperature 57 range considered in this project i s given in Table 1. Bender et a l have compared these results (with the exception of those of Griswold et a l ) with 2k. their own and concluded that no one set could properly be cited as unequivocal support for any other. Results and Discussion The vapor pressure of benzene was measured at six temperatures using one sample of the purified material. The results of these measurements are given in Table 1 and shown plotted in Figure 2 . The reported pressures have been corrected for the par t i a l pressure of mercury vapor in the equilibrium c e l l on the basis of Dalton's law. Corrections were calculated for the 6o Poynting effect, but were found to be negligible, compared to the accuracy of the pressure measurements, at a l l temperatures. The measured values agree, within the accuracy of the pressure measurements^with the carefully determined values of Bender et al^7 a ^ every temperature studied. It is of interest to note that the vapor pressure measurements of Bender et al57 were also made over a mercury surface, and that the same corrections for the effect of mercury, in the vapor phase applied;, as were used with the values reported here. RowlinsonDJ- has shown that the simple Dalton and Poynting corrections are not always adequate, particularly at high pressures, to correct for the influence of mercury in the vapor phase, but any error introduced in this way w i l l be the same for both sets of results. p A comparison of the reported data with that of von Huhn and of Griswold, Andres, and Klein3 i s of particular interest since these workers have studied the phase equilibria of the benzene-toluene system. Since the values measured in this project agree so well with those calculated from the equation of Bender et a l ^ , this equation was used to calculate vapor pressure values for comparison purposes at temperatures at which measurements were not made. TABLE 1 25-VAPOR PRESSURE OF BENZENE Pressure - mm. of mercury Temperature - C. Temperature This Bender et a l 5' Gornowski p von Huhn Young59 Griswolc et a l 3 Work et a l 58 120.0 2249 2265 2230 121.0 2303 2260 125.0 2540 2540 2530 130.0 2841 2825 2821 140.0 35M+ 3518 35^ 5 3520 150.0 4385 4371 4331 4372 4335 'i 116O.O 5332 '5277 5310 5300 162.0 5541 5275 170.0 6440 6368 6386 175-0 7065 7056 7050 180.0 7712 7621 7690 7617 190.0 9159 9050 9045 200.0 10780 10800 I O 6 7 O IO780 IO65O 10860 210.0 12640 12500 12453 220.0 14702 14550 14700 14521 225.0 15860 15820 15870 230.0 17010 I685O 16825 240.0 19570 19425 19580 19352 19390 249.6 22340 22370 22270 250.0 22420 22310 22182 255-0 23950 24310 260.0 25570 25520 25560 25330 280.0 32870 32940 32630 32780 26 27-3 The five reported values of Griswold et a l differ from the measured (or calculated) values of, this work hy from -4.8 percent at 162° C., o to +1-5 percent at 255 C Some of the disagreement can he explained in terms of the relatively poor accuracy of the temperature measurements of 3 o Griswold et a l ; however, the disagreement for the values at l 6 2 C. and 255° C. is about twice what could be explained in this way. 3 The reported values of von Huhn agree very well with the present values. At each temperature the difference•in pressure is less than the uncertainty in the pressure measurements. The excellent agreement of the benzene vapor pressures measured in this work with those reported by other workers, particularly those of Bender 57 et a l , serves to verify both the accuracy of the method of measurements and the high degree of purity of the benzene used. On the other hand, the lack of agreement, at at least some temperatures, of the values reported by 3 Griswold et al with those of any other workers, would indicate that their values are in error. Toluene Previous Work Toluene vapor pressure measurements have been made at elevated 62 ? 63 pressures by Krase and Goodman ; von Huhn ; Zmaczynski ; and Griswold, 3 64 Andres, and Klein . The American Petroleum Institute Research project 44 has reviewed existing vapor pressure measurements and calculated, from c r i t i c a l l y selected data, the constants for a vapor pressure equation valid up to about 135° C. 62 Krase and Goodman have reported measured vapor pressures for temperatures from 129° C to the c r i t i c a l point. These workers have given 28. 2 no estimate of the accuracy of their measurements. Von Huhn has measured o o toluene vapor pressures from 120 to 300 C. Pressures were measured to within +k0 mm. of mercury (h mm. at pressures up to 7500 mm.) and temperatures to within about 0.1° C. Reported values were taken from a smoothed curve of the data by that author. Zmaczynski has reported the vapor pressure of toluene at temperatures from 80° to 155° C The results were obtained by comparing the temperatures at which water and toluene have the same vapor pressure, and therefore pressures were not measured directly. Vapor pressures were reported in the form of an equation relating the boiling temperature of toluene and of water under the same pressure. Temperatures calculated from this equation agreed with measured values to within 0 05° C. in most cases. 3 Griswold et a l have measured the vapor pressure of toluene at three ^o o temperatures between 130 and 227 C. Pressures were determined to within _ o 50 mm. of mercury and temperatures with an accuracy of approximately 0.5 C. Measured values were reported. The procedure used by each of these workers to purify toluene for the vapor pressure measurements is described in Appendix VII The toluene vapor pressure data of the above workers that is of interest in this project i s given in Table 2. Results and Discussion The vapor pressure of toluene was measured at nine temperatures using five samples. The results of these measurements are given in Table 3 and shown plotted in Figure 3- T n e reported pressures have "been corrected for the partial pressures of mercury in the equilibrium c e l l on the basis of 60 Dalton's law. The value of Poynting correction was calculated and found to be negligible compared to the accuracy with which pressure could be measured, for a l l points. The pressure measurements in runs 2, 2A, 3> 3A, and h agree within TABLE 2 29-VAPOR PRESSURE OF TOLUENE Temperature c. Pressure-mm. of mercury Temperature Calculated t h i s project API6" Zmaczynski^ Krase ag^ Goodman v. Huhn2 Griswold et al3 II6.9 906 906 123-3 1074 1075 125.0 1133 1124 129-5 1270 1262 1277 129-6 1275 1267 1268 136.1 1495 1488 1489 138.0 1566 1780 140.0 1643 1660 142.6 1746 1741 148.9 2021 149.1 2030 2026 149-5 2050 2128 150.0 2073 " T55-7 2350 2347 160.0 2585 2595 161.0 2641 2926 175-0 3530 175-5 3565 3940 176.7 3648 180.0 3897 3900 186.5 4416 4650 189.O 4629 4630 194.O 5078 5300 200.0 5659 566O 203.5 6021 6280 204.4 6120 214.4 7256 7520 220.0 7959 7960 225.0 8627 226.0 8766 9030 227.0 8906 9050 234.5 10010 10070 240.0 10880 10910 246.0 11890 12160 250.0 12600 12960 TABLE 3 MEASURED AND CALCULATED VAPOR PRESSURES OF TOLUENE Temp. Pressure - mm. of mercury Temperature - 0 C. Run 2 Run 2A Run 3 Run 3A Run 4 Run 30 Run 33 Avg. Avg. Deviation from Avg. Calculated 1123 1130 1130 4 1133 2008 2018 2030 11 2021 2062 2068 2065 2 2073 3519 3523 3520 2 3530 3641 3644 3655 12 3648 5646 5649 5650 2 5659 6097 6102 6120 12 6120 8613^ 8577 8600 18 8627 12600 12600 - 12601 125-0 148.9 150.0 175-0 176.7 200.0 204.4 225-0 250.0 2033 6133 2040 2022 2043 3672 6125 6128 2037 3661 6127 o 31 4.8r T T 4.7r 4.0-VAPOR PRESSURE OF TOLUENE 4.5k 'C.p. (64) 4AL 4.3r 5 4.2 o w 0 E •5 E E Id CC a. a. o o 4.0k to 3.9r CO 3.8L 3.7 3.6L 3.5L 3.4L_ 3.3L 3.2U 3.1 3.0L 1.6 1.7 1.8 L9 2.0 2.1 4 " X 10 3 2.2 2.3 2.4 Temperature = °K. 2.6 Figure 3 3 2 : the stated accuracy of the pressure gauges at each temperature as do the measurements in runs 3 0 and 3 3 - However, the maximum variation in the measured values;when both groups of runs are considered^is greater than that which would be expected from the uncertainty' in the gauge readings alone. The cause of the variation is not known, and because i t was impossible to decide i f one group of measurements was better than the other, a l l of the measurements were used for the determination of the averaged values. The constants for an Antoine equation have been determined from the experimental results and i t was found that the equation log P = 7 . 1 7 9 0 2 - I517.OO 242.76 + t where P i s the pressure in mm. of mercury o t i s the temperature i n C. allows the calculation of vapor pressures which agree with the average of the measured values to + 1 0 mm. at pressures below 85OO mm. of mercury and to + 3 0 mm. at pressures above 85OO mm. of mercury. Values calculated from this equation are shown in Table 2 for comparison with those of other workers. As shown in Table 2 , excellent agreement is found between the values of Zmaczynski^3 and those calculated from the equation given above. The fact 6^ that the values of Zmaczynski 2 also agree very well with the calculated API values gives additional confirmation of the accuracy of the reported values. Very poor agreement i s found between the calculated values of this 6 2 work and those of Krase and Goodman . At every, temperature the value of these two workers is>higher, and the difference varies from 0 . 6 to 1 1 . 2 percent (based on the calculated values). 2 The reported values of von Huhn agree with those of this work within the accuracy of the pressure measurements at a l l temperatures except 33> lkO° C; however, at this temperature the disagreement is less than that which can he accounted for • hy the uncertainty in both the pressure and temperature measurements. 3 The reported values of Griswold et a l agree closely with those of this work at only one temperature. At 189° C. the values are essentially identical; at 227°c. they dif f e r by 1.7 percent, and at I380 C. by 13.8 percent The difference at 227° C. can be explained by the uncertainty in the two values but the difference at 138°C. cannot be explained in this way, and the value for this temperature is almost certainly invalid. The excellent agreement of the toluene vapor pressures of this work go with those of Zmaczynski , and the good agreement of these two sets of values 6h with the API c r i t i c a l l y selected values serves to verify the accuracy of the vapor pressure values measured in this research. Additional confirm-ation of this accuracy is given by the good agreement between the present p values and those reported by v. Huhn . The lack of agreement of the values of 62 Krase and Goodman with those of any other worker'would indicate that these values are in error. Methanol Previous Work Methanol vapor pressure measurements have been made over the same 65 5Q temperature range as studied in this project by Kay and Donham , Young^, and Rao, Sarma, Swami, and Rao^. Kay and Donham^ have measured vapor pressures for temperatures from 130° C. to the c r i t i c a l point and reported values taken from a smoothed curve. They estimated that their temperature measurements were accurate to 0.05° C. and their pressure measurements to 25 mm. of mercury. Young has measured the vapor pressure of methanol at temperatures from -10° C. 34. the c r i t i c a l point. The results of these measurements are reported in the 52 form of an equation for which the accuracy is not stated. Rao et a l have reported measured vapor pressure values for temperatures from 90° to 1960 C. The accuracy of their measurements was not stated. Measured values were reported. The procedure used by each of the above workers to purify methanol for the vapor pressure measurements is described in Appendix VII. The vapor pressure measurements reported by each of the above groups of workers that are of interest for comparison to values determined in this work are given in Table k. Results and Discussion The vapor pressure of methanol was measured at six temperatures using four samples. The results of these measurements are given in Table 5 and shown plotted in Figure k. The reported pressures have been corrected for the partial pressure of mercury in the vapor phase on the basis of Dalton's law. The value of the Poynting correction was calculated and found to be negligible compared to the accuracy with which pressures were measured at a l l temperatures. The averaged values of the measured pressures have been included in Table k for comparison with those of other workers. These values agree within the uncertainty of the pressure measurements with the values of Kay and Donham^ at 150° and 200° C. They also agree, within the accuracy of the pressure o o 65 measurements, at 175 and 225 C. with values calculated from Kay and Donham's data, using temperatures 5° C. above and below the desired values and an equation of the form log P = A - B . T TABLE 4 VAPOR PRESSURE OF METHANOL Pressure - mm. of mercury Temperature - C. Temperature Averaged value this project Kay and Donnam^ Young59 Rao et a l 5 2 99-0 2590 100.0 2640 2621 2663* 105-3 3060 123-6 5172'" 125-0 5530 5454 5386* 127-0 5689 149.1 9825 150.0 10465 10441 10336 10030* 151.3 13400 170.0 16460 16292 173-7 17070 175-0 18315 18320 1803O 17510* 175.2 17580 180.0 20330 20089 200.0 30220 30190 29787 220.0 43480 42573 225.0 47540 47470* 46297 230.0 51735 50414 * Starred values calculated from two closest bracketting points using equation of the form log P = A - B . T 36. TABLE 5 MEASURED VAPOR PRESSURE OF METHANOL Pressure - nun, of Mercury Temperature - C. Temperature Run 15 Run 19 Run 20 Run 32 Avg. Avg. Deviation from Avg. 100.0 26U5 2635 2645 2640 5 125-0 5529 5519 5528 5530 5 150.0 10445 10446 10494 10480 10465 21 175-0 18330 18302 18315 14 200.0 30216 30221 30220 3 225-0 47522 47555 47540 16 37 38. The vapor pressure values of Young59 a r S } a t every temperature, lower than those measured in this research. The difference i s about 0-7 percent (based on the averaged values) at 100° C. and increases to 2.6 percent at 225° C 52 The starred values of the vapor pressures reported for Rao et a l were calculated from the closest two points of these workers using an equation of the form log P = A - B . The calculated value is higher at 100° C. than T the value measured in this project, and lower at every other temperature. While the values of these workers are on the whole closer to those of Young59 than to those of Kay and Donham^ ^^ o r to the present values, the agreement i s so poor (varying by from 1-5 to 3-1 percent) as to be of doubtful value as 59 support for Young's measurements. The agreement of the methanol vapor pressure of Kay and Donham^ and those of this work serves to verify the results obtained in this research. On 59 the basis of this agreement i t would appear that the values of Young^ are 52 slightly low and those of Rao et a l x are significantly in error. ( BENZENE-TOLUENE SYSTEM 39-Previous Measurements 3 Griswold, Andres and Klein have measured isothermal vapor-liquid equilibrium values for the benzene-toluene system at elevated pressures. The measurements were made with a d i s t i l l a t i o n type apparatus in which the condens-ate was returned to the boiler as a liquid. Pressures were measured to the nearest 1 p.s.i. (50 mm. of mercury) using a gauge and temperature with an accuracy of approximately 0.5° C using thermocouples. Thiophene-free benzene and nitration-grade toluene were purified for the measurements by fractionation through a laboratory column packed with l8 in. of single-turn glass helices. Heart cuts of 80 percent of the original charge were retained. The purity of each material was checked by vapor pressure measurements, the results of which are given and discussed in the section headed Vapor Pressure Measurements. Mixtures of the purified benzene and toluene were analysed by means of their boiling points. Temperature, pressure, and composition of the cb-existing phases were measured at temperatures of 120°, l60°, l80°, 200°, 2h0°, and 280° C. The values reported for 200° e. are given in Table 6. 2 Von Huhn has also measured isothermal vapor-liquid equilibrium values for the benzene-toluene system at elevated pressures. The equilibrium measurements were made using a bomb type apparatus in which both phases could be sampled. Pressures were measured with two pressure gauges accurate in the range 0 to 7500 mm. of mercury to +h mm., and in the range above 7500 mm. of mercury to +k0 mm. Temperatures were measured with the aid of thermocouples to about 0.01° C Benzene and toluene of the highest purity commercially available at the time of the measurements were dried by d i s t i l l a t i o n and stored over sodium wire. The purity of each material was checked by means of vapor pressure TABLE 6 VAPOR-LIQUID EQUILIBRIA OF BENZENE-TOLUENE SYSTEM AT 200° (Griswold, Andres, and Klein 3) Pressure Mole Percent Mole Percent (mm. of Benzene in Benzene in mercury) Liquid Vapor 6770 15-4 22.2 7240 27-0 38.4 7910 38.5 - 51.8 8690 56.6 68.2 9720 77-7 84.6 10860 100 100 kl. measurements. Smoothed values of the measured vapor pressures were reported, and these values are given and discussed in the section headed Vapor Pressure Measurements. Mixtures of the purified materials were- analyzed with the aid of an interferometer. 2 Phase equilibrium values taken by v. Huhn from smoothed curves of his data were reported for temperatures from 120°to 300°.-C. at 20° intervals. The values reported for 200° C- are given in Table 7-Results The vapor-liquid equilibria of the benzene-toluene system were measured at 200° C. The pressure and the composition of both the liquid and the vapor phases were determined at 20 points using six independent mixtures. Measured pressures were corrected for the presence of mercury in the vapor phase assuming Dalton's law. The results of the measurements are given in Table 8. Independent samples are represented by different run numbers in this table. A vapor-liquid equilibrium "x-y" plot of the data is given in Figure -5, and a pressure-composition plot in Figure 6 . Calculation of Activity Coefficients A c i t i v i t y coefficients were calculated for the measured points according to the definition The fugacity of each component in the vapor )f^, was calculated using the equation proposed by Joffe^ . | n A _ = Tcm,*-Tcl (HlH)+ Pc-ux-fe* ( 3 . , ) , * H f T I R T / r5 V 'rr\ix {21) TABLE 7 VAPOR-LIQUID EQUILIBRIA OF BENZENE-TOLUENE- SYSTEM AT 200' (von Huhn ) Pressure Mole Percent Mole Percent (mm. of Benzene in Benzene in mercury) Liquid Vapor 10,780 100 100 10,220 91.4 93-4 9780 82.5 87.O 9270 73-4 80.1 8750 63-9 72.3 8310 54.1 63.7 78OO - 44.0 54.9 7320 33-6 44.0 6800 22.8 32.1 6240 11.6 17.7 566O 0 0 TABLE 8 VAPOR-LIQUID EQUILIBRIA OF THE BENZENE-TOLUENE SYSTEM AT 200° C. AS MEASURED IN THIS PROJECT Run Pressure Mole Percent Mole Percent Number (mm. of Benzene in Benzene in mercury) Liquid Vapor 565O 0 0 8-200H 6105 10.5 16.3 G 6105 10.5 I6.5 F 6520 19.O 28.2 E 6520 19.0 28.4 D '6725 23.2 33-8 C 6725 23.2 33-8 A 6825 25.2 36.3 • B 6840 25-3 ,36.4 10-200E 7165 32.4 44.3 D 7445 38.0 50.3 C 7470 38.7 50.8 B 7840 45.4 57-4 A 7855 45.8 58.2 14-200C 8025 48.7 60.7 B 8450 56.1 68.2 A 8460 56.4 68.1 I1-200F 8800 62.6 74.1 D 9050 69-3 .79-0 C 9300 72.8 81.9 9-200B 9350 73-6 82.1 A ' 9350 73-5 82.2 11-200B 9450 76.4 84.2 A 9550 77-8 85.1 12-200D 10500 95-4 97-2 C 10500 95-4 97-2 B 10600 96.8 98.0 A 10600 97-0 98.0 10780 100 ' 100 kh 0 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT BENZENE IN LIQUID PHASE Figure 5 1 2 0 0 0 VAPOR-LIQUID EQUILIBRIUM PRESSURE COMPOSITION PLOT FOR B E N Z E N E - T O L U E N E S Y S T E M 11000. A T 2 0 0 ° C I 0 0 0 C _ 9000 8 0 0 C _ 7 0 0 C _ 6 0 0 C _ MOLE PERCENT B E N Z E N E 500OL 10 2 0 3 0 4 0 5 0 6 0 70 8 0 9 0 " 100 Figure 6 he. Pseudo-critical temperatures and pressures were calculated for use in equation (2l) assuming Kay's definitions TCm«= £ 3t TCU (22) PC(TM*~ I u^Pcc (23) and the c r i t i c a l data quoted by the A.P.I. Project hh^ (Benzene-Tc=289-5° C , Pc=U8.6 atm., Zc=.27*+; Toluene-Tc=320.8° C*., Pc=U0.0 atm., Zc=.263). Mixture properties for use in equation (2l) were calculated using the 25 generalized correlation of Lyderson et a l and reduced parameters based on the pseudo-critical temperature and pressures of equations (22) and (23)-The fugacity of the pure liquid at the temperature and pressure of O . X the'mixture, f ^ , (a hypothetical standard state in the case of benzeneJ was calculated from the equation p r * y i \i 0 V t - 'q J P 67 * In a recent paper, Ambrose and Grant have reported a c r i t i c a l temperature of 318.6 + 0.02° C for toluene. While the value measured by these two authors i s probably more accurate than that reported above, the error 6h introduced in the activity coefficients by the use of the A.P.I. value w i l l be very small. hi. Saturated fugacities and liq u i d volumes needed for the evaluation of equation (2h) were obtained from the following sources. Saturated fugacities for both benzene and toluene were calculated using the generalized 25 correlation of Lyderson et a l . The saturated liquid volume of benzene at 68 200° C. was obtained from the data reported by Organick and Studhalter , and the saturated liquid volume of toluene at 200° C. from the data reported by Driesback for 20° C.?and the method proposed for extrapolating liquid volume data by Lyderson et a l 2 - \ In the evaluation of equation (2h), i t was assumed that the liquid volume was independent of pressure and equal to the saturated value. ( i t i s of interest to note that a fugacity coefficient of 0.8U6 was obtained for pure saturated benzene at 200° C. using the generalized correlation of lyderson et a l 2 ^ while a value of O.815 was obtained from the 68 work of Organick and Studhalter which was based on experimental data. While 68 the value of Organick and Studhalter i s almost certainly more accurate than that calculated from the reduced correlation, the latter was used for the following reason. The fugacity coefficients for the mixtures of benzene and toluene had to be calculated using the reduced correlation. Since these mixture coefficients were calculated from the combined properties of benzene and toluene, and since these two substances are adjacent members of a homologous series, i t seems l i k e l y that the same sort of error would appear in the mixture coefficient as appeared in the pure component coefficient. The two fugacity coefficients appear as a ratio in the calculation of activity coefficients, and therefore, when they both are calculated on the same basis, much of the error in each should cancel. ) Activity coefficients calculated as described above for the data in Table 8 are given in Table 9- Activity coefficients calculated in terms of deviations from Raoult's law (equation 17) TABLE 9 48. ACTIVITY COEFFICIENTS FOR"BENZENE-TOLUENE SYSTEM AT 200° C Mole Percent Benzene Toluene Benzene Toluene Benzene in Raoult' s law Raoult's law Liquid 10.5 •993 •996 .876 1.010 10.5 1.005 •994 .886 1.008 19.0 1.001 •995 .894 1.022 19.0 1.008 •992 .901 1.020 23-2 1.007 .991 .904 1.025 23.2 1.008 •991 .905 1.026 25-2 1.007 •991 .908 1.029 25.3 1.010 •992 .909 1.031 32.4 •994 •995 .905 1.045 38.0 •992 •997 •917 .I.O56 38.7 .987 1.001 .906 1.06l 45-4 < .988 1.009 .916 I.O83 45.8 .994 •'•999 .922 1.072 48.7 •991 1.007 .924 1.088 56.1 - 1.007 •993 .949 1.086 56. 4 1.005 1.004 .944 1.098 62.a 1.011 • 975 .962 1.079 69-3 •995 •983 •953 I.O96 72.8 1.001 •975 .967 1.094 73-6 •999 •998 • 964 1.122 73-5 1.001 .988 .966 1.112 76. 4 •993 •993 .962 1.120 77-8 •994 1.003 .965 1.135 95-4 •992 .969 .989 1.131 95-4 •993. .969 •989 1.131 96.8 • 991 1.007 •992 1.172 97-0 •999 1.072 .989 1.119 h9. *<• x. P have also been calculated and are included in Table 9-t - M- ( 1 7 ) Discussion 2 The vapor-liquid equilibrium ."x-y" data of v. Huhn and of Griswold et a l for 200° C. and that obtained in this work are shown plotted in Figure 7 - Much better agreement was found between the present vapor-liquid 3 equilibrium composition data and that of Griswold et a l than between the 2 present data and that of v. Huhn . The maximum difference between the values of the mole fraction of benzene in the vapor for a given mole fraction 3 benzene in the liquid phase for this work and that of Griswold et al was about one mole percent, and most differences were less than this value. Differences of up to about three mole percent in the values of the mole fraction benzene in the vapor for a given mole fraction in the liquid were found between v. Huhn' 2 data and the values measured in this work, and a l l of v. Huhn's values were lower than the present ones. A pressure-composition diagram for the three sets of data i s given in Figure 8. In this plot, much better agreement was obtained between the 2 data of v. Huhn and the present data than between either of these sets and the 3 data of Griswold et a l . At every pressure, however, the difference in 2 composition between the liquid and vapor phases is less for v. Huhn's than for 2 the present work, and neither the liq u i d nor the vapor compositions of v. Huhn agree, over the entire pressure range, with the values obtained in this work. 52-The agreement between the pressure-composition values of Griswold et a l and the present-values is poor. At every composition,the pressure measured in this work is the lower of the two. The fact that better agreement should be found between the two sets of data on a vapor-liquid equilibrium plot than on a pressure-composition plot is not surprising, since the temperature •3 measurements of Griswold et a l J were relatively inaccurate, and since temperature has only a small effect on equilibrium composition and a large effect on pressure measurements. The difference, however,,is greater than expected, and at one point i s about four times that which can be explained on the basis of a difference in temperature alone. The greatest differences occur at the high toluene concentration end of the curve, and i t was noted in the section headed Vapor Pressure Measurements that one of the three reported vapor pressures of 3 Griswold et a l was significantly different than that of any other worker. A l l but 3 of the 27 activity coefficients calculated for benzene using Joffe'sequation are within +1.0 percent of 1-0 and a l l coefficients are within +1-5 percent of 1.0. A l l but 7 of the 27 coefficients for toluene are within +1.0 percent of l.O^and a l l but 1 are within +3-1 percent of 1.0. The last coefficient, which varies from 1.0 by 7-0 percent^is for a solution with a very low toluene concentration, and is thus subject to a very large experimental uncertainty. The coefficients for both benzene and toluene show no significant systematic change in value with concentration and are effectively equal to 1.0 for a l l concentrations. Since the activity coefficients are effectively equal to 1.0, the ratio of the activity coefficients w i l l also be equal to 1.0 and the log of the ratio equal to zero and independent of the mole fraction of benzene. The system benzene-toluene at 200° C. forms, h therefore, a perfect l i q u i d solution as defined by Lewis and Randall . Coefficients calculated using the Raoult's law deviation expression differ substantially from 1.0; the coefficients for benzene are a l l less than 53-1.0 and the coefficients for toluene are a l l greater than 1.0. These activity coefficients cannot, therefore, satisfy a consistency test,and this method of calculating activity coefficients for the benzene-toluene system o at 200 C is invalid. METHANOL - TOLUENE SYSTEM 54. Previous Measurements 54 Benedict, Johnson, Solomon and Rubin have reported vapor-liquid equilibrium measurements for the methanol-toluene system at a pressure of 76O mm. of mercury. The measurements were made with an atmospheric-pressure, constant-volume type apparatus. Temperatures were measured, in this apparatus, with a thermocouple located in a well in the liquid phase. A barometer reading was taken with each temperature reading and was used to correct the measured temperature to that corresponding to a pressure of "j60 mm. of mercury. Samples were taken of both the liquid and vapor phases in order to determine the composition of each. Methanol (CP. grade) was purified for the measurements by fraction-ation. The d i s t i l l a t e from this fractionation was dehydrated by refluxing i t over magnesium methylate and was then r e d i s t i l l e d . Special high-purity toluene supplied by the Barrett Company was further purified by fractionation before i t was used. Purities of both toluene and methanol were specified, and the composition of mixtures of the two were measured by means of their refractive index. The composition of both the liquid and vapor phases was measured by 54 Benedict et a l for 10 temperatures^and the values obtained are given in Table 10 (page VI-l). An azeotrope boiling at 63.6° C. was found at a composition of 88 mole percent methanol. The composition of the azeotrope boiling at 76O mm. of mercury pressure 70 has also been measured by Berg and Harrison . These workers used Baker's "Analyzed" toluene, and methanol purified by d i s t i l l a t i o n for the measurements. The type of equipment used for the azeotrope determination was not described ex p l i c i t l y but was probably a glass Othmer s t i l l . An azeotrope containing 86 mole percent methanol and boiling at a temperature of 63.8 C. was found. 55-T l Robinson, Wright, and Bennett have also measured azeotropic compositions for the methanol-toluene system at several temperatures. "Eastman"-grade chemicals and laboratory-grade reagents were purified for the measurements. The azeotropes were determined by measuring the composition of '"constant-evaporation" mixtures. Dried air was drawn through a prepared mixture of toluene and methanol and a' sample of the liquid removed periodically for analysis. The pressure at which the measurements were made was not specified. Results of the measurements are given in Table 17-Results The vapor-liquid equilibria of the methanol-toluene system have been measured at 125°, 150°, 175°. 200°, 225°, and 250° C The pressure and the composition of the vapor were measured over the entire liquid composition range at each temperature. The results of the measurements are given in Tables 11 to 16 (pages VT-2 to VI-7)- Reported pressures have'been corrected for the presence of mercury in the vapor phase assuming Dalton's law. Independent samples are represented in these tables by different run numbers. Isothermal vapor-liquid equilibrium "x-y" composition diagrams and isothermal pressure-composition diagrams of the measured data are shown in Figures 9 to 20. Equilibrium ratios have been calculated for both methanol and toluene from the experimental data. The calculated values are included in Tables 11 to 16 and shown plotted as a function of the logarithm of the pressure'in Figures 21 and 22. An azeotrope was found for this system at temperatures of 125°, 150°, and 175° C The composition of the azeotrope at each temperature was determined from a large scale "x-y" composition diagram of the measured data and is given as a function of temperature and pressure in Table 17- The 56 0 10 20 30 40 50 60 70 80 • 90 100 MOLE PERCENT METHANOL IN LIQUID PHASE Figure 9 7000 _ VAPOR-LIQUID EQUILIBRIUM PRESSURE COMPOSITION DIAGRAM FOR METHANOL - TOLUENE SYSTEM AT 125 ° C. 6 0 0 C -4 5 0 0 -2 0 0 C _ • liquid phase O vapor phase 10 20 3 0 4 0 5 0 6 0 7 0 80" 90" 100 I MOLE P E R C E N T METHANOL F i g u r e 10 58 MOLE PERCENT METHANOL IN LIQUID PHASE Figure 11 VAPOR-LIQUID EQUILIBRIUM PRESSURE COMPOSIT ION DIAGRAM FOR M E T H A N O L - T O L U E N E S Y S T E M A T 1 5 0 ° C. 10000 — 1000 — # liquid phase O vapor phase 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT METHANOL Figure 1 2 60 MOLE P E R C E N T M E T H A N O L IN LIQUID P H A S E Figure 13 1 3 5 0 0 -12500 — 1 0 5 0 0 -950C — VAPOR-LIQUID EQUILIBRIUM PRESSURE COMPOSITION DIAGRAM FOR M E T H A N O L - T O L U E N E SYSTEM A T 1 7 5 ° C. 7 5 0 0 -550C — # liquid phase O vapor phase 4 9 0 TOO 10 20 3 0 4 0 50 6 0 7 0 MOLE P E R C E N T METHANOL Figure ih 62 MOLE PERCENT METHANOL IN LIQUID PHASE Figure 15 3500C-3 3 0 0 0 -31000. 29000_ 2 7 0 0 C -2 5 0 0 0 -2 3 0 0 C -2 1 0 0 0 -19000 17000 15000 -13000 11000 -9 0 0 0 7 0 0 0 VAPOR-LIQUID EQUILIBRIUM PRESSURE COMPOSITION DIAGRAM FOR M E T H A N O L - TOLUENE S Y S T E M A T 2 0 0 6 C. • liquid phase O vapor phase 50001 1 1 10 20 30 4 0 50 6 0 70 80 90 100 M O L E P E R C E N T METHANOL Figure l6 6k 0 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT METHANOL IN LIQUID PHASE Figure 17 60000r 56000 52000-48000• 44000-40000 36000 32000 28000 24000 20000 /6000 12000 8000 4000 VAPOR-LIQUID EQUILIBRIUM PRESSURE COMPOSITION DIAGRAM FOR METHANOL- TOLUENE SYSTEM AT 225° C. • liquid phase O vapor phase 1 0 2 0 30 40 50 60 70" MOLE PERCENT METHANOL Figure 18 "90—roo 66 0 10 20 30 40 50 60 70 80 90 100 M O L E P E R C E N T M E T H A N O L IN LIQUID P H A S E Figure 19 67 6 0 0 0 0 5 6 0 0 0 5 2 0 0 0 4 8 0 0 0 4 4 0 0 C -4 0 0 0 0 £ 3 6 0 0 0 £ 3 2 0 0 0 E E UJ cr => CO CO 2 8 0 0 0 £ 2 4 0 0 0 2 0 0 0 0 T T T VAPOR-LIQUID EQUILIBRIUM PRESSURE COMPOSITION DIAGRAM FOR M E T H A N O L - T O L U E N E S Y S T E M A T 2 5 0 ° C. 16000 12000 w -8 0 0 0 4 0 0 0 -# liquid phase O vapor phase 10 20 30 4 0 50 6 0 70 80 9 0 100 MOLE P E R C E N T M E T H A N O L Figure 20 1000 2 0 0 0 3 0 0 0 4 0 0 0 5 0 0 0 6 0 0 0 8 0 0 0 10000 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0 PRESSURE - mm. of mercury Figure 21 OD Figure 22 TABLE 17 70. CALCULATED AND MEASURED VALUES FOR THE METHANOL-TOLUENE SYSTEM AZEOTROPE Pressure - mm. of mercury Composition - mole percent methanol Worker Temperature Measured Values Calculated Value s ° C Composition Pressure Composition Pressure This work 175-0 150.0 125.0 98-3 96.O 93-7 18350 10530 5585 98-3 96.O 93-7 I856O IO58O 5520 Benedict et elf 63-6 88 760 88.0 770 Berg and ^ . 70 Harrison 1 63-8 86 760 88.0 769 Rohinson et a l ? 1 62.5 50.0 25-0 0.5 89.6 89.I 88.6 87.8 — 88.0 86.8 8U.5 82.2 --71-relationship between the temperature and the composition can be expressed to the nearest 0.1 percent by the equation x = 8 2 . 2 + .092t ( 2 5 ) where x = the mole percent methanol t = temperature in °C. The pressure of the azeotrope can be expressed as a function of temperature to within 1 - 5 percent by the relationship log P = 8 . 4 5 6 2 - I877.O T ( 2 6 ) where P = pressure in mm. of mercury T = temperature in °K. A comparison of the measured values with values calculated from equations ( 2 5 ) and ( 2 6 ) i s given in Table 1 7 - A temperature-composition diagram for the azeotrope is shown in Figure 2 3 , and a pressure-composition diagram in Figure 2k. Calculation of Activity Coefficients Activity coefficients have been calculated for both components of the methanol-toluene system according to the definition activity coefficients from equation ( l 6 ) requires a knowledge of both the fugacity of component rtk" in the mixture, and fk°, the fugacity of pure component " k " in i t s standard state (defined as pure liquid "k" at the temperature and pressure of the mixture). Since insufficient data are 72 200, 150 o o I *I00 +-O i_ Q> CL E 0) 50 1 1 AZEOTROPE 1 1 1 1 COMPOSITION / / / / / / / / / / O This work © ^ 3 Benedict et al - 3 © Berg and Harrison — 3 Robinson et al 3 - 3 1 1 1 1 1 1 1 — 86 88 90 92 94 96 MOLE PERCENT METHANOL Figure 23 98 1 0 0 200 175 150 125 100 75 Temperature = °C. Figure 2k are available for the methanol-toluene system to allow the calculation of these fugacities by the exact relationships In f \ S J - ^ . ^ j j p 4 | M t P ( 1 5 ) l«*A = ' " f ^ p + £ d P (13) lo f ^ v o p - £ P ^ ( V *- + In P^vop (12), fugacities of both components have been calculated with the aid of several solution theories. Approximate methods based on these solution theories are in common use by Chemical Engineers, and from them the fugacity of a component in a mixture can be calculated using the properties of the pure materials forming the mixture. Fugacities determined in this way were used with the fugacities of the pure components to calculate activity coefficients and the coefficients from each method have been compared and checked for thermodynamic consistency. The c r i t i c a l properties and the saturated li q u i d volume at the temperature of the mixture were required for both methanol and toluene in most of the calculation methods used. The c r i t i c a l temperature and pressure of methanol were obtained from the compilation of the properties of organic 56 solvents of Reddich and Toops (Tc = 2^0.0° C , Pc = 78.6 atm.); the c r i t i c a l volume from the tabulation of Reid and Sherwood1^ (Vc = 3-673 ml/gr); and the saturated li q u i d density from the compilation of the thermodynamic properties 72 of methanol of Smith . The c r i t i c a l temperature, pressure, and volume of toluene were obtained from the A.P. I. Project kh^ (Tc = 320.8° C , Pc = hO.O atm., Vc = 3-12 ml/gr.). The saturated liquid density of toluene was 73 calculated from the equation proposed by Guggenheim for reduced saturated liq u i d densities 9 r --|+C(|-T0 ' / 3 ^d( l-Tr) (27) 75-The constants in this equation were evaluated using data quoted by 69 Driesbach . Activity coefficients have been calculated from each of the approximate methods described below using an Alwac III E d i g i t a l computer. Since the computer was used only as a tool for solving analytical expressions, a description of the programs used has not been included. 1. Activity coefficients in terms of deviations from Raoult's law If a vapor mixture behaves as a perfect gas, then the fugacity of a component in that mixture is given by f k = %? (28) and equation (12) reduces to ^AVOP = il^op (29) If, in addition, the effect of pressure on the fugacity of a liquid can be neglected, then - f^vop = ^vap (30) and the activity coefficient is expressed in terms of deviations from Raoult's law (equation 17) Activity coefficients have been calculated for methanol and toluene from the measured data according to equation (17) for temperatures of 125° , 150° , and 175° C The results of these calculations are given in Tables 18 to 23 (pages VI-8 to VI-13). 76. 2 . Activity coefficients by the Lewis and Randall rule. k According to the Lewis and Randall rule , the fugacity of a component in a vapor mixture i s equal to the product of the mole fraction'of that component in the mixture and the fugacity of the pure component in the vapor state at the temperature and pressure of the mixture. The activity coefficient i s then defined, as given earlier, by V = % ( 1 9 ) Fugacities of pure methanol and pure toluene vapors have been calculated for use in equation ( 1 9 ) by two methods; a) Generalized fugacity chart of Gamson and Watson 2 k Gamson and Watson have presented a generalized chart of fugacity coefficients (ratio of fugacity to pressure) for gases as a function of the reduced pressure and temperature of the gas. From this chart, fugacities have been calculated for pure methanol and for pure toluene at each of the measured pressures at temperatures of 1 2 5 ° and 1 5 0 ° C. The fugacity of each pure liquid in i t s standard state was then calculated using the value of the saturated fugacity and the assumption that the liquid volume was independent of pressure and equal to the saturated value. The fugacities of the pure liquid and of the pure vapor calculated as described above were used with equation ( l 9 ) t o evaluate the activity coefficients for each material. The results of these calculations are given in Tables l 8 to 2 1 . (More accurate values of the fugacities of 'methanol - - and toluene would probably have been obtained i f 25 the improved generalized correlation of Lyderson et a l could have been used; however, 'this correlation, which i s presented in terms of tables of values, contains too few values near the saturation conditions for accurate interpolation.) 77-b) Generalized equation of state of Black Black^ has proposed a generalized equation of state of the form V=&? +fe- | * r (3D where b= § ^ (32) a- 27 6RTc/8 (33) ! = A +B/7r - C / V f D/Tr 3 + m/l7T™ +!° (34) and |VG R^T/ t H P r / V + KfP/ftop)3 + Wr/fr (35) The values of "A", nB", "Cn, "D", "F", "Grt, "H", and "K" are constants independent of the material for which the equation is used, and the values of "E n and "m" are functions of the material considered, nE" being equal to zero for non-polar substances. An expression for the fugacity coefficients of a pure substance has been calculated from equation (3l) in an exact manner by Black to give (36) From equation (36), fugacities have been calculated for pure methanol and for pure toluene at each of the 'measured pressures at temperatures of 125°, I5O0, and 175° C. (The value of P/Pvap was set equal to 1.0 in the case of toluene to allow extrapolation of equation (36) to the a r t i f i c i a l state of a pure vapor at a pressure greater than i t s vapor pressure.) The fugacity of 78. each pure li q u i d in i t s standard state was then calculated using the value of the saturated fugacity and the assumption that the liquid volume was independent of pressure and equal to the saturated value. The fugacities of the pure liq u i d and of the pure vapor calculated as described above were used with equation (19) to calculate activity coefficients for each material. The results of these calculations are given in Tables 18 to 23-3- Generalized equation of state of Hirschfelder et a l and fugacity equation of Joffe 29 Hirschfelder, Buehler, McGee and Sutton have proposed a general-ized equation of state for the gas region in which the reduced pressure is expressed as a function of the c r i t i c a l compressibility, the reduced tempera-ture and volume, and the slope of the vapor pressure curve at the c r i t i c a l point. The equation can, with the aid of experimental saturated li q u i d density data for two temperatures, be extended to describe the liquid region as well. The authors have derived from their equation of state, formulas for the thermodynamic excess functions and the fugacity coefficient suitable for solution by a d i g i t a l computer. calculated from the c r i t i c a l properties of the pure components by a Mnear combination with mole fraction. The properties of the mixtures of methanol and 29 toluene were then calculated from Hirschfelder et al's equations at each of the measured pressures for temperatures of 125°, 150°> and 175° C . The mixture properties calculated by this method were used with the equation derived by Joffe- 3 for the fugacity of a component in a mixture Psuedo-critical properties for use in this equation of state were T c 4 AH°-B\ em (21) ,to evaluate the fugacities of both methanol and toluene. 79-The fugacities of pure li q u i d methanol and pure li q u i d toluene were 29 also evaluated at each of the measured pressures using Hirschfelder et al's y equation and saturated liquid density data. Mixture and pure component fugacities calculated in this manner were then used with equation'(l6) to calculate activity coefficients at temperatures of 125°> 150°> and 175° C. The results of these calculations are given in Tables 18 to 23-4. Generalized equation of state of Redlich and Kwong 28 Redlich and Kwong have proposed an emperical equation of state of the form P = RT/(V-±0 - a / r ' / L V ( V+ b ) (37) where <* O. 42 78 Rl Tc2 r / Pc ( 3 8 ) b = 0 0867 RTc/Pc. (39) For convenience in the use of the equation, the terms R-_ o / f ? v T z r (ho) B- b/RT (4i) have been defined. The authors have also derived an expression for the fugacity coefficient of a gas, from their equation of state, in the form /n4 = -lW2-BP)- A* B ln( l f BP/2)+2-1 (h2) 80. 28 Redlich and Kwong have suggested that their equation can be applied to a vapor mixture by combining "A" and "B" for the components of the mixture to give Bm,* = I ^ B t (44) In this case, the expression for the fugacity of a component of the mixture becomes /n4~-° 4 3 ^ 3 ( 2 - / ) ^ -gV - l n ( 2 + BP)nM, flm\» ft m i x The fugacities of methanol and of toluene in each of the mixtures studied at temperatures of 125°, 150°, and 175° C. have been calculated from equation (45)- Fugacities of pure saturated methanol and toluene were calculated from equation (42) and used with the assumption that the pure liquid volume is independent of pressure to calculate the fugacity of the pure materials in their standard states. Activity coefficients were then evaluated from equation (l6) for both toluene and methanol using the calculated fugacities. The values of the coefficients determined in this manner are given in Tables 18 to 23-5- Generalized equation of state of Black Black^- has proposed that his equation of state (equation 3l) applied to vapor mixtures by combining the values of "a", "b", and n £ B for the pure components to give values for the mixture. A linear combination of the values of "b" to give 81. and a separate geometric combination of the polar and non-polar parts of the values of "a" to give and {Q-lf'L =Z(*>\S* (U8) -4-where § is the non-polar part of g and ^ i s the polar part of | was proposed. Using these combinations the equation of state for a mixture becomes v ^ + r M j - f c f a - l f l i f - \£(a^)0'%f (U9) From this equation Black has derived an expression for the fugacity of each component in an n n n component mixture to give where H T R T R T (50) (51) (52) (53) •In the derivation leading to the last two terms of (50)> Black has assumed that the terms " ^ i " are independent of pressure and equal to the value of He has estimated that the errors arising from this assumption are small and negligible except in the c r i t i c a l region. 82. The fugacities of methanol and toluene in each of the mixtures studied at temperatures of 125°, 150°, 175°, 200°, and 225° C have been calculated from equation (50). (The value of P/Pvap for toluene was set equal to 1.0 in the evaluation of the term y k P/Pk v a p - n Q r d e r t Q a l l o w the use of the equation in the region where the total pressure was significantly greater than that of the vapor pressure of toluene.) The fugacities of saturated methanol and toluene were calculated from equation (36) and used with the assumption that the pure liquid volume of each i s independent of pressure, to calculate the fugacities of the pure materials in their standard state. Activity coefficients were then evaluated from equation (l6) for both toluene and methanol. The values of the coefficients determined in this manner are given in Tables 18 to 23, and Tables 33 and- 34 (pages VI-21 and VI-22). Thermodynamic Consistency Check of the Experimental Data Van Ness"^ has shown that isothermal activity coefficients for a binary mixture are thermodynamically related by the expression (54) The right hand side of equation (54) is frequently very small, and the equation may be written, as noted earlier, as In A = 0 (10) Values of (n X^j^t. determined from the activity coefficients calculated as described earlier are given in Tables 2k to 27 (pages VI-lU to VI-17)- The logarithms of the ratio of the activity coefficients obtained by using Black's equation (50) ,are shown plotted as a function of the mole percent 83-methanol in the liquid phase for temperatures of 1 2 5 ° , 1 5 0 ° , 1 7 5 ° > 2 0 0 ° , and 2 2 5 ° C in Figures 25 to 29- Similar diagrams, showing only the "best f i t t i n g line (drawn by eye) for values from each of the calculationrmethods used at temperatures of 1 2 5 ° , 1 5 0 ° , and 1 7 5 ° C , are given in Figures 30 to 32. Graphs similar to those shown in Figures 30 to 32 have been used to determine the value of the integral J I r\ c^/7^ * - / 4 ) r p = ( K 7 7 n . £ ) ( 5 5 ) and therefore / H = R T I n + /x£ ( 5 6 ) The difference between the actual chemical potential of the component in the mixture and the chemical potential i t would have in an ideal solution i s given by /4 = R T l " * A ^ + f l - R T U x - - R T I A V * ( 5 7 ) 0 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT METHANOL Figure 25 50 00 50 50 00 50 RATIO OF ACTIVITY COEFFICIENTS IN THE SYSTEM METHANOL-TOLUENE AT 150° C. O Experimental data and Black's Equation 1_— Correlation 00. 10 20 30 40 50 60 70 MOLE PERCENT METHANOL 80 90 100 Figure 26 0 10 2 0 3 0 4 0 5 0 6 0 70 8 0 9 0 100 MOLE PERCENT METHANOL Figure 27 1.50 1.00 .50 - . 5 0 U -1.00 -1.50 -2.0CL 1 RATIO OF ACTIVITY COEFFICIENTS IN THE SYSTEM M E T H A N O L - T O L U E N E AT 2 0 0 ° C. — O Experimental data and Black's Equation — Correlation 1 10 20 30 4 0 50 6 0 70 MOLE PERCENT METHANOL Figure 28 80 90 100 1.50 1.00 .50 -.50 -1.00 •1.50 -2.oa 1 1 1 1 1 1 1 1 1 RATIO OF ACTIVITY COEFFICIENTS IN THE SYSTEM METHANOL-TOLUENE AT 225 ° C. O Experimental data and - — Correlation Black's Equation 1 1 1 i i i i i i 20 30 40 50 60 70 MOLE PERCENT METHANOL 80 90 100 Figure 29 co co MOLE PERCENT METHANOL Figure 30 0 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT METHANOL Figure 31 o MOLE PERCENT METHANOL Figure 32 H TABLE 28 92. YALUES OF THE INTEGRAL | /n — dz, o Calculation methods are numbered.to correspond with methods of calculation described in text. Method of Calculation Positive Negative Difference Total Difference in Temperature of Activity, Area Area in • Area Area Area Percent °C. Coefficients of Total 1 •345 •435 -. 090 .780 -12 125 3 •373 .405 .032 .778 -4.1 125 5 .405 •399 .006 .804 0.8 125 4 • 403 •376 .027 •779 3-5 125 2A .414 • 346 .068 .760 8.9 125 2B .440 •336 , .104 •776 14 125 1 •255 •435 .180 .690 -26 150 3 .296 • 386 .090 .682 -13 150 5 •358 •345 .013 •703 1.8 .150 4 .361 .317 .44 .678 6.6 150 2A •373 .265 .108 .638 •17 150 2B •390 .267 .123 .657 19 150 1 •175 •439 .264 .614 -43 175 3 .231 .382 .151 .613 -25 175 5 • 304 .281 .023 .585 3-4 175 4 • 322 • 243 .079 •565 14 175 2B .383 • .185 .198 .568 35 175 5 .282 •251 .031 •533 5-8 200 5 .280 • 223 .057 .503 11 . 225 . 93-The term ^u.^ is. known as. the excess chemical .potential. The excess free energy of a one molar solution may be written^ from equation ('57) ^ F E - l * t / u * ZzrfTUXi (58) Redlich and Kister^ have proposed relating .the - excess free energy of a binary solution.to experimental data by an equation of the form ( A F E ^ R T x , x t r B + C ( x r ^ + D ( ^ r ^ t + I : ( ^ - ^ J 3 + — J (59) Since 4 F E- RK^.IAV, I f J (58) i t can easily be shown that in - ^ ^ (60) and therefore In B(2 v - x , ) + C f 6^ , - 1 ) i 0 (^r^)^Xa^V£(zr^(iox^r\) ( 6 l ) The constants'"B n, nCn, "D n ;."E n, ... can.also be used to calculate the individual activity coefficients l« y, = * i [B + C f3*,-x J + D (*i-^Ks.V*0+ £ ( * > - * ^ V 7*rXt2 (62) Activity coefficients for temperatures of 125°, 150°,.175°, 200°, and 225° C , calculated using the equation of Black^, have been correlated using a four-constant form of equation (59)- The equation was rearranged in the form * . , / n r - S+C(*rV> + D ( x r i J l + B ( W (59) 9h. for convenience of calculation. The coefficients "B", "C", "D", and "E" were calculated to give a "least-square f i t " of the data assuming b * X 1 n t o be the dependent variable. The activity coefficients for values of the mole fraction of methanol less than 0.10 and greater than 0.95 were not.included in the calculation of the constants since values of the composition, and thus of the activity coefficients at the two extreme:, ends of the composition range, were subject to a high degree of experimental error. Values of the coefficients nB", "C", "D", and "E" from the correlation calculations are given in Table 29. The values of these coefficients have been used with equations (62) and (63) to calculate activity coefficients for both methanol and toluene at temperatures of 125°, 150°, 175°, 200°, and 225° C The values calculated from these two equations are presented in Tables 30 to 3^ for comparison with values calculated directly from Black's^ equation. The percent differences given in these tables are based on those from Black's equation. The values of the coefficients "B", nC", "Bn, and "E" have also been used to evaluate equation (6l), the expression for the logarithm of ratio of the activity coefficients. The resulting equations have been plotted in Figures 25 to 29 for comparison with the values calculated directly from 6 Black's equation. Discussion Presentation of experimental data The vapor-liquid equilibrium "x-y n composition and pressure-composition diagrams (Figures 9 to 20) and the equilibrium ratio-pressure diagrams (Figures 21 and 22) have been plotted using measured values of the pressure, temperature, and composition; however, not a l l of the measured TABLE 29 .95-COEFFICIENTS OF REDLICH AND KISTER CORRELATION FOR METHANOL-TOIUENE SYSTEM Temperature C. B C D E 125 1.58^ 4 . 0 8 9 4 4 . 1 8 6 2 - . 0 6 3 2 150 1-4273 . 0 8 1 0 7 . 1 6 9 2 - . 0 0 8 2 175 1 . 2 5 7 0 .O765O •1475 +.0426 200 1 . 0 9 8 5 • . 0 3 7 1 8 .2194 - . 0 0 3 5 225 . 9 8 7 5 . 0 2 4 8 4 . 1915 - . 0 3 8 3 96. points have been included. When more than one measurement was made for a solution of a given composition, or when measurements were made for two or more solutions very close to one another in composition, only one point is shown. Azeotrope The composition of the azeotrope was found to follow a linear relationship with temperature for the three temperatures 125°, 150°, and 175° C. The fact that Rao et al obtained a similar result for, the azeotrope in the closely related methanol-benzene system serves to confirm.the validity of a relationship of this form. If equation (25), the equation relating temperature and azeotropic composition, is extrapolated to a composition of pure methanol, i t i s found that the azeotrope disappears at 193-4 C No azeotrope was found at a temperature of 200° c., indicating that a value of about 193° C. is not unreasonable. If equation (25) is extrapolated.in the opposite•direction to a temperature of 63.6° c., an azeotropic composition of 88.0 mole percent methanol is calculated. This value gives excellent agreement with the value of 88 percent reported.by Benedict et a l . Much poorer agreement is found between the calculated azeotropic compositions and the values reported Joy Berg and Harrison^ and by Robinson et a l ^ for temperatures of 63.8° and 62-5° C respectively. It appears lik e l y , therefore, that the-values of these two workers are in error. The calculated azeotropic compositions and 71 r o those reported by Robinson et al for temperatures, below 62.5 0 . show very poor agreement;. however, extrapolation of equation(25) to the lower temperatures may not be valid. The azeotropic pressures calculated from equation (26) agree reasonably well with both the values measured in this work-and.the value 97-o reported at 63-6 C. by Benedict et a l , particularly when the very simple nature of the equation and the relatively large pressure range' over which i t is applied are considered. Activity coefficient calculations Activity coefficients calculated from the same experimental data hut using different approximation methods have quite different values. The activity coefficients, for example, of toluene for solutions containing 99-1 mole percent methanol are, when calculated in terms of deviations from Raoult*s law, 6-55 at 125° C , 5-68 at 150° C , and 5.78 at 175° C. The 6 activity coefficients for the same solutions as calculated using Black's k equation of state 'for pure toluene, and the Lewis and Randall fugacity rule, are 5-26 at 125° c., 3-95 at 150° C , aid 3-23 at 175° c., or a difference of 20, 30, and hk percent respectively (percentages based on Raoult's law value). Although these percentage differences are the maximum that occur in the activity coefficients for these solutions, each approximate method gives a significantly different value from the others. It i s obvious, "therefore, that at most, only one of the methods tested can give correct values. methods have been tested, as was described earlier, for thermodynamic 5 consistency using the Redlich and Kister test At 125 C ., as shown in Table 28, the value of the l e f t hand side of equation (10) for activity coefficients calculated from the experimental data using Black's^ equation i s only .006 units. This value represents only 0.8 percent of the "t o t a l " value of the integral (the total value being Activity coefficients calculated from each of the approximate (10) o 98. defined as the sura of the positive area and the absolute value of the negative area under the curve of In ffi- versus x), and is probably less TT than the uncertainty with which the integral was evaluated. At the same temperature, the activity coefficients calculated in terms of deviations from•Raoult1s law resulted in a value for the integral of 0.09 units, or 12 percent of the-"total" value, and the coefficients calculated from Black's^ 4 equation for the pure components, and the Lewis and Randall fugacity rule gave a value of .104 units, or 14 percent of the "total" value of the integral. At 150° O., the value of the integral in equation (10) for activity coeffic-ients calculated by Black's method i s .013 units or 1.8 percent of the "total" value, while at the same temperature the other methods resulted in values of the • integral of up to 26 percent of the "total". At 175° C , the value of the integral in equation (10) for activity coefficients calculated by Black's^ equation was .023 or ^ .h percent of the "total" value while the next best 28 method, that of Redlich and Kwong , gave a value of .079, almost four times as large. For the same temperature, values of up to 4-3 percent of the "total" were found for activity coefficients calculated by the other methods. From the results in Table 28, i t i s obvious that each of the 6 methods of calculation used, except that of Black , results in values of activity coefficients which are thermodynamically inconsistent. While the cause of the failure of each method i s not always definitely known, in many cases at least part of the reason i s f a i r l y obvious. The accuracy of the 28 method proposed by Redlich and Kwong depends in part upon the accuracy with which their generalized equation of state, depending only on the c r i t i c a l temperature and pressure, can be used to describe the properties of the pure components used to form the mixture. Since it"has been shown that generalized equations based only on the c r i t i c a l temperature and 99-pressure are of relatively poor accuracy, particularly for polar substances, 28 the method of Redlich and Kwong must be subject to the same limitation. Activity coefficients calculated from the generalized equation of on oQ Rirschfelder et a l and the fugacity equation of Joffe- 3 are in error for 29 two reasons. Psuedo-critical properties for use with Hirschfelder et al's equation were calculated by a linear combination with composition of the properties of the pure components. It is unlikely that such a simple method would be satisfactory for a system containing a polar substance such as was •3D considered in this work. Joffe's fugacity equation is also restricted in accuracy by the assumption that the c r i t i c a l temperature and pressure of the mixture are given by Kay's^ rule, and this restriction would probably result in additional error in the activity coefficients. It i s of interest to note here that Storvick and Smith have attempted to calculate vapor phase enthalpies for several polar-non-polar systems, including the methanol-benzene system, using a generalized correlation. Psuedo-critical constants were calculated using Kay's^ method, and using these constants, mixture enthalpies were evaluated from the generalized 75 correlation of Hougen and Watson . When the calculated values for the methanol-benzene system were compared with the experimental values of McCrackei and Smith'^, errors of up to 35 percent were found. Similar errors were found for the other systems. In an effort to improve the.accuracy of the calculations, true c r i t i c a l mixture properties were used with the generalized correlations. Only a relatively small improvement was found. It was the conclusion of the authors that generalized procedures give unreliable results for mixtures containing both a hydro-carbon and an associating substance. Activity coefficients for a mixture, calculated from the properties k of the components used, and the Lewis and Randall fugacity 1 0 0 . rule, are incorrect because of the failure of the fugacity rule. This rule is based on the assumption of no volume change on mixing for the vapor phase, a condition not met with non-ideal solutions, particularly at elevated pressures. A striking demonstration of the failure of the fugacity rule i s given by the 6 fact that Black's equation of state for pure components can be generalized to give a satisfactory equation of state for mixtures using only.the properties of the pure components. If, however, the equation of state for the pure components is used directly with the fugacity rule to calculate mixture fugacities, large errors result. Activity coefficients calculated in terms of deviations from Raoult's law are invalid because of the assumptions made in the derivation of the Raoult's law expression. The vapor phase of a non-ideal solution is not a perfect gas mixture, and the effect of pressure on the liquid phase is not negligible. Cause of apparent inconsistencies in activity coefficients calculated by Black's method. i While the value of the integral J In Y,/Xu d^i was small for the 6 activity coefficients calculated from the experimental data by Black's method, i t was not equal to zero as required by the Redlich and Kister^ test, nor even negligibly small at the higher temperatures. Part of the inconsistency i s , of course, due to errors in the activity coefficients: errors caused by inaccuracies 6 in the experimental data and errors caused by the i n a b i l i t y of Black's equation of state to represent the system studied exactly. (Black^ has noted that for some mixtures in which special "chemical" effects are present, e.g. strong interassociation or compound formation between unlike molecules, one or more of which is nonassociated in the pure vapor state, individual binary coeffip-ients may be required. These binary coefficients are not readily predicted from the data for the pure components.) Part of the inconsistency, however, is 101. almost certainly caused by the use of a constant-pressure, constant-temperature form of the consistency test equation. The complete form of the equation used to check the thermodynamic consistency of isothermal data is f In A d%, = f 4£ dP (64) In order to estimate the importance of the right hand side of equation (6k), the quantity A v avg w a s defined where The value of £ i V avg required in order to make equation (6k) f i t the experimental data exactly, was calculated, and i t was found that for each of the temperat-ures for which activity coefficients were determined (125°, 150°, 175°, 200°, and 225° C.) the value was a constant fraction (constant within + k percent) of the average of the saturated volumes of methanol and toluene at the same temperature. While the author makes no claim for the accuracy of the numerical value obtained for avg by this method, the constancy.of the value (in terms of the average volume of the pure components) does indicate that a change in volume .of mixing which results in a negligible value of the right hand side of equation (6k) at 125° C (.006 units) can result in a substantial value for o , . the same quantity at 225 C. (-057 units). Activity Coefficient Correlation Activity coefficients calculated from the four-constant Redlich and Kister'' correlation equation agree reasonably well with the experimental o o values at temperatures of 125 and 150 C At these two temperatures, the maximum difference between the experimental and the calculated values is 7 RT dp (65) 102. percent, and most differences are considerably less. At 175°C the maximum difference increases to 10 percent but most differences are less than 5 percent. At 200° and 225°C, the correlation results in values considerably higher than those calculated from the experimental data. The values agree in the 10 to 90 percent composition range to within 10 percent but differences of hO percent occur for the dilute solutions. The cause of the relatively poor agreement for the values of the experimental activity coefficients and those calculated from the Redlich and Kister^ correlation at elevated temperatures i s the fact that this equation applies exactly only under conditions of constant temperature; and pressure. Considering that the restriction of constant pressure is not valid for the activity coefficients reported in this work, the agreement between experimental and calculated activity coefficient i s not unreasonable, except, perhaps, for dilute solutions at the highest temperatures. Correlation of equilibrium ratios with data reported by Rao et a l 80 An attempt has been made by Shemilt to correlate the equilibrium ratios measured in this work for the methanol-toluene system with those 52 obtained by Rao et a l for the closely related methanol benzene system. To date a satisfactory correlation has not been obtained. CONCLUSIONS AND RECOMMENDATIONS IO3. Vapor Pressure Measurements Excellent agreement is found between the vapor pressure of benzene measured in this work and the value calculated from the vapor pressure equation 57 given by Bender et a l at each temperature for which a value was measured. 2 The present values also agree well with those reported by v. Huhn . The 58 59 values of Gornowski et a l and of Young , which agree well with one 57 another, are lower than those measured in this work; however,•Bender et al 59 have noted that Young's values may be shown to be consistently, low. Vapor pressures calculated from the Antoine equation proposed in 2 this work for toluene agree well with values reported by v. Huhn and by 63 Zmaczynski . Since vapor pressures calculated from this equation differ 62 widely from values proposed by Krase and Goodman f and since the accuracy 2 63 of the equation is supported by the values ,of v. Huhn and Zmaczynski , i t 62 appears that Krase and Goodman's values are' in error. At least one of the 3 values reported for the vapor pressure of toluene by Griswold et a l i s also almost certainly•invalid. Good agreement is found between the vapor pressure of methanol measured in this work and the vapor pressure interpolated from the data of 65 Kay and Donham for each temperature at which measurements were made, over the common temperature region. Since the accuracy of the values measured in 65 this work is supported by the modern measurements of Kay and Donham the 59 methanol vapor pressures reported by Young appear to be slightly low, and 52 those of Rao et a l very low. Benzene-toluene System The vapor-liquid equilibrium values measured in this work for the o benzene-toluene system at 200 C. show only f a i r agreement with those reported 2 3 for the same temperature by v. Huhn and by Griswold et a l . The equilibrium 104. 3 compositions of this work agree reasonably well with those of Griswold et al , but the agreement between the two sets of pressure measurements is poor. It was found, however, that at some temperatures the vapor pressure measurements of Griswold et a l for benzene and for toluene did not agree well with those of other workers, and therefore their pressure measurements are probably in 2 error. The pressure measurements of v. Huhn showed good agreement with those obtained in this work, but the equilibrium composition measurements for the two sets of values were significantly different. The cause of this difference is not known. Although the agreement between the values obtained in this work and 3 2 those obtained by Griswold et a l and v. Huhn is only f a i r , the measurements of these last two sets of workers do not agree with one another as well as they agree with the values from this work. The fact that an excellent thermodynamic consistency check was obtained for the present data gives further indication of accuracy of the measurements. Methanol-toluene System An azeotrope was found for the methanol-toluene system at temperatures of 125°, 150°, and 175° 0. with i t s composition a linear function of temperature for these three temperatures. Extrapolation of this linear relationship to calculate the azeotropic composition at 63.6° C. gives a 54 value of composition identical to that measured by Benedict et al . The 70 71 azeotropic composition reported by Berg and Harrison and by Robinson et al would therefore appear to be in error. A consistency check of activity coefficients calculated from the experimental data for this system by six approximate methods shows that five of the methods lead to incorrect values. Of the methods of calculation used, 105-6 only that proposed by Black gives thermodynamically consistent activity coefficients, even at temperatures as low as 125°C The method of testing 5 consistency, that of Redlich and Kister , indicated the most accurate activity o coefficients at 125 C , with the accuracy decreasing as the temperature increased. The Redlich and Kister^ method is s t r i c t l y valid, however, only at constant temperature and pressure, and much of apparent inconsistency at the higher temperatures is probably due to the fact that the restriction of constant pressure does not apply. Since the experimental data were shown by Black^'s method to be thermodynomically consistent, the thermodynamic consistency test has been used to evaluate several approximate methods of calculating activity coefficients and thus of the solution theory on which the approximate methods are based. On the basis of this test, the Raoult's law expression, the Lewis and Randall fugacity rule, the fugacity equation of Joffe , and the Redlich and Kurony mixture equation are inadequate to describe the methanol-toluene system at elevated pressures. The four-constant Redlich and Kister-' equation used to correlate activity coefficients from Black's method allows, at the lower temperatures, the calculation of coefficients in good agreement with those on which the equation is based. With increasing temperatures the agreement becomes poorer, particularly for dilute solutions. This lack of agreement at the high temperatures i s due to the fact that the assumption of constant pressure, on which the equation is based, becomes less and less valid as the temperature increases. Recommendations for Future Work The approximate methods of calculating activity coefficients used in this work were checked by an indirect method, i.e., by testing the thermody-namic consistency of the calculated coefficients. If the activity coefficients io6. had f i r s t been calculated by an exact method, a direct check would have been possible. It i s therefore suggested that in future work of the type presented in this project, the pressure-vblume-temperature properties (the properties necessary for the calculation of activity coefficients by exact methods) of the mixtures studied be measured as well as the phase equilibrium properties. The assumption of constant pressure in the thermodynamic consistency check of activity coefficients introduces an error of unknown magnitude. If data on the isothermal volume change on mixing as a function, of composition were available for the li q u i d phase, this assumption could be eliminated, and a more accurate measure of the thermodynamic consistency of the experimental data could be obtained. These data could be obtained from the volume measurements suggested above, but i f in future work complete pressure-volume-temperature data are not measured, i t is suggested that at least the volume change on mixing for the li q u i d phase be determined. TABLE OF SYMBOLS 107. A constant in Black's equation of state A terra in Redlich and Kister's fugacity equation a activity a attraction constant in Black's equation of state a constant in Redlich and Kister's equation of state B constant in Black's equation of state B coefficient in Redlich and Kister's correlations B term in Redlich and Kister's fugacity equation .b covolume constant in Black's equation of state b constant in Redlich and Kister's equation of state C constant in Black's equation of state C coefficient in Redlich and Kister's correlations c number of components c constant in reduced density equation D constant in Black's equation of state D coefficient in Redlich and Kister's correlations D degrees of freedom d constant in reduced density equation E constant in Black's equation of state E coefficient in Redlich and Kister's correlations F free energy F constant in Black's equation of, state &F excess free energy of mixing f fugac ity f° standard state fugacity f fugacity of a component in a mixture f y a p fugacity of saturated vapor G constant in Black's equation of state G° non-polar term in Black's fugacity equation G° polar term in Black's fugacity equation H enthalpy H° enthalpy of ideal gas £H change of enthalpy on mixing H constant in Black's equation of state K constant in Black's equation of state m constant in Black's equation of state N number of phases P pressure Pc c r i t i c a l pressure Pr reduced pressure P y ap saturated vapor pressure R gas constant S entropy T temperature - °K. Tc c r i t i c a l temperature Tr reduced temperature t temperature - °C. 109-V ,volume Vc c r i t i c a l volume V. . liquid volume 'liq u AV volume change on mixing V partial molal volume •x mole fraction in liq u i d y mole fraction in vapor Z compressibility Zc c r i t i c a l compressibility )g activity coefficient !• attraction coefficient in Black's equation of state °^ attraction coefficient at zero pressure |" non-polar part of attraction coefficient § polar part of attraction coefficient u term in Black's fugacity equation chemical potential /i° chemical potential in standard state p.E excess chemical potential Jr reduced density Subscripts 1 state 1 1 component 1 2 state 2 110. 2 component 2 i component i j component j k component'k P constant pressure T constant temperature avg average mix mixture LITERATURE CITED 111. 1. Sage, B.H. and Lacey, W.N. Trans. A.I.M.E., 174:102, 1948, 2. von Huhn, W. Forsctu Gebiete Ingenieurw., A2:129, 1931* 3. Griswold, J., Andres, D., and Klein,V.A. Trans A.I.Ch.E., 39:223, 1943. 4. Lewis, G-N. and Randall, M. Thermo dynamics and the Free Energy of Chemical Substances , New York, McGraw-Hill Book Company Inc., 1923. 5. Redlich, 0 and Kister, A.T. Ind. Eng. Chem., 40:34l, 345, 19^8. 6. Black, C Ind. Eng. Chem., 50:391, 1958. 7. Robinson, CS. and( G i l l i l a n d , E.R. Elements of Fractional D i s t i l l a t i o n . 4th Ed., New York', McGraw-Hill Book Co. Inc., 1950. 8. Aroyan, H.J. and Katz, D.L. Ind. Eng. Chem., 43:185, 1951. 9. Bahlke, W.H. and Kay, W.B. Ind. Eng. Chem., 24:291, 19.38. 10. Kay, W.B. Ind. Eng. Chem., 28:1015, 1936. 11. Kay, W.B. and Rambosek, G.M. Ind. Eng. Chem., 54:221, 1953-12. Sage, B.H. and Lacey, W.N. Trans A.I.M.E., 136:136, 1940. 13. Whittle, D.J. M-A.Sc Thesis in Chemical Engineering. U.B.C, 1958. 14. Reid, R.C and Sherwood, T.K. The Properties of Gases and Liquids. New York, McGraw-Hill Book Co. Inc., 1958. 15. Guggenheim, E.A. Thermodynamics. 4th Ed., Amsterdam, North-Holland Publishing Co., 1959-16. Van Ness, H.C Chem. Eng. Sci., 10:225, 1959-17- Carlson, R.C, and Colburn, A.P. Ind. Eng. Chem.., 34:58l, 1942. 18. E l l i s , S.R.M. and Bourne, J.R. Brighton International Symposium on Dis t i l l a t i o n , p-62, i960, preprint. 19- Dodge,-B.F., Chemical Engineering Thermodynamics. New York, McGraw-H i l l Book Co. Inc., 1944. 20. Herington, E.F.C Nature, 160:610, 1947. 21. Coulson, E.A., Hales, J.F., and Herington, E.F.G. Trans. Faraday Soc., 44:629, 19^8. 22. Tunell, G. J. Phys. Chem., 35:2885, 1931. 23. Newton, R.H. Ind. Eng. Chem., 27:302, 1935. 24. Gamson, B.W. and Watson, K.M. Natl. Pet. NPWS, Tech. Sec, 36:R623, 1944-112. 25- Lyderson, A.L., Greenkorn, R.A., and Hougen, O.A. Generalized Thermodynamic Properties of Pure Fluids. Univ. Wisconsin Eng. Exp. Sta., Report 4, Oct., 1955-26. Pitzer, K.S., Lippman, D.Z., Curl, R.F., Higgins, CM., and Peterson, D.E. J. Am. Chem. Soc, 77:3433, 1955-27- . Biedel, L. Chem. Ing. Tech., 26:83, 257, 679, 1954. 28. Redlich, 0. and Kwong, J.N.S. Chem. Revs., 39:333, 1946. 29. Hirschfelder,.J.O., Buehler, R.J., McGee, H.A., and Sutton, J.R. Ind. Eng. Chem., 50:375, 386, 1958. 30. van der Waals, J.D. Kon. Ak. Amsterdam, 1880. 31. Keyes, F.G. Proc. Nat. Acad. Sci., 3:323^ 1917 . 32. Beattie, J.A. and Bridgeman, 0.C J. Am. Chem. Soc, 49:1665 1927". 33. Benedict, M., Webb, G.W., and Rubin, L.C. J. Chem. Phys.., 8:334, 1940. 34. Hirschfelder, J.O., Curtiss, CF., and Bird, R.B. Molecular Theory of Gases and Liquids. New York, John Wiley and Sons Inc., 1954. 35. Beattie, J.A. Chem. Revs., 44:l4l, 1949-36. Hougen, CA., Watson, K.M., and Ragatz, R.A. Chemical Process Principles- Part II, 2nd Ed., New York, John Wiley and Sons Inc., 1959-37- Lewis, G.N., and Randall, M. Thermodynamics. 2nd Ed.7 New York, McGraw-Hill Book Co. Inc., 1961T 38. Joffe, J. Ind. Eng. Chem., 40:1738, 1948. 39- Ju Chin Chu, Getty, R.J., Brennecke, L.F., and Paul, R. Distillation Equilibrium Data. New York.7 Reinhold Publishing Co., 1950-40. Hala, E-, Pick, J., Fried, V., and Vilim, 0. Vapor Liquid Equilibrium. New York, Pergamon Press, 1958. 41. Timmermans, J. The Physcio-Chemical Constants of Binary Systems in Concentrated Solutions. I-IV, New York, Interscienic Publishers, Inc., 1959-42. Comings, E.W. Ind. Eng. Chem., 39=948, 1947. 43. Smith, J.M. Ind. Eng. Chem., 45:963, 1953. 44. Newitt, D.M. High Pressure Plants and Fluids at High Pressure. New York, Oxford University Press, 1940. 45-46. Carol, M.M. B.A.Sc. Thesis in Chemical Engineering. U.B.C, 1952. Drymose, L. B.A.Sc Thesis in Chemical Engineering. U.B.C. 1957-113-hi. Emerson, H-L. and C u n d i l l , T.G. B.A.Sc. Thesis i n Chemical Engineering. U.B.C, 1951. 48. Hamburg, A. B.A.Sc. Thesis i n Chemical Engineering. U B . C , 1952. 49. Howey, G.R. M.A.Sc. Thesis i n Chemical Engineering. U.B.C, 1951. 50. Waldichuck, M. M.A. Thesis i n A r t s and Science. U.B.C, 1950. 51. McCracken, P.G., and Smith, J.M. A.I.Ch.E J o u r n a l , 2:498, 1956-52. Rao, V.N.K., Sarma, K.J.R., Swami, D.R., and Rao, M.N. J . S c i . Ind. Res., l6B:4, 1957-53. K r i c h e v s k i , I.R., Khazanova, N.E., and L i n s h i t s , L.R. Zur. F i z . Khim., 31:2711, 1957-54. Benedict, M., Johnson, CA., Solomon, E., and Rubin, L.C Trans A-I.Ch.E., 41:371, 1945. 55. C r o i l , T. M.A.Sc. Thesis i n Chemical Engineering. U.B.C, 1959-56. R i d d i c k , J.A., and Toops, E.E. Organic Solvents. 2nd Ed v New York, I n t e r s c i e n c e P u b l i s h e r s , Inc., 1955-57- Bender, P., Furukawa, G.T., and Hyndman, J.R. Ind. Eng. Chem., 44:387, 1952. 58. Gornowski, E.J., Amick, E.H., and Hixson, A.N. Ind. Eng. Chem., 39:1348, 1942. 59- Young, S- Proc. Roy. Soc. Dublin,.12:374, 1910. 60. Poynting, J.H. P h i l . Mag., 12:32, 1881. 61. Jepson, W.B. and Rowlinson, J.S. J . Chem. Phys., 23:1599, 1955-62. Krase, N.W. and Goodman, J.B. Ind. Eng. Chem., 22:13, 1930-63. Zmaczynski, M.A. J . Chim. Phys., 27:503, 1930. 64. American Petroleum I n s t i t u t e . S e l e c t e d Values of P h y s i c a l and Thermodynamic P r o p e r t i e s of Hydrocarbons and Related Compounds. P r o j e c t 44, Carnegie Press, P i t t s b u r g , 1953-65. Kay, W.B. and Donham, W.E. Chem. Eng. S c i . , 4:1, 1955-66. Kay, W.B. Ind. Eng. Chem., 28:10l4, 1936. 67. Ambrose, D. and Grant, D.C Trans. Faraday Soc. , 53:771, 1957-68. Organick, E.I. and S t u d h a l t e r , W.R. Chem. Eng. Prog., 44:847(1948). 69. Driesbach, R.P. P h y s i c a l P r o p e r t i e s of Chemical Compounds. Advances i n Chemistry S e r i e s , 15, 1955-114. 70. Berg, L. and Harrison, J.M. Chem. Eng. Progr., 43:487, 1947-71. Robinson, E., Wright, W.A., and Bennett, G-W. J. Phys. Chem., 36:658, 1932. 72. Smith, J.M. Chem. Eng. Progr., 44:521, 1948. 73. Guggenheim, E.A. J. Chem. Phys., 13:253, 1945-74. Storvick, T.S., and Smith, J.M. J. Chem. Eng. Data., 6:28, 1961. 75- Hougen, 0-A. and Watson, K.M. Chemical Process Principles. Part 2., New York, McGraw-Hill Book Co. Inc., 1943. 76. American Instrument Co. Inc. Superpressure Catalog 407- Silver Spring, Maryland, 1961. 77- Perry, J.H. Chemical Engineers Handbook. 3 r (l Ed., New York, McGraw-Hill Book Co. Inc., 1950. 78. Sanderson, R.T. Vacuum Manipulation of Volatile Substances. New York, John Wiley and Sons Inc., 1948. 79- Hodgeman, C.D. Handbook of Chemistry and Physics. 31st Ed., Cleveland, Chemical Rubber Publishing Co., 1949-80. Shemilt, L.W. Letter to the writer, February 6, 1962. APPENDIX I - APPARATUS I - l Equilibrium Cell The equilibrium c e l l (part number l ) * was machined from 304 stainless steel bar stock. The c e l l chamber was 2 in. in diameter and 9 3/8-in. deep and therefore had a volume of about 480 c c The'top of the c e l l was sealed with a l i d (3), gasket (l4), and cap (2). An assembly, drawing ©f this seal i s shown in Figure 1-8. The gasket was a F l e x i t a l l i c Gaske,t Company "Special Style R" stainless steel-and-/ > teflon gasket, formed of alternate rings of V-shaped stainless steel and teflon. It was f i t t e d into a groove :ipi the top of the bomb and trapped there by a spiggot on the l i d . The lid'was forced against the gasket by six flat-bottom / set screws (32) which were held by the cap and acted through a 304 stainless i , steel protecting ring (4). • The cap i t s e l f was f i t t e d to the bomb by means of a thread which matched a similar one on the top outside surface of the bomb. When the set screws were tightened (torque-750 inch-pounds), sufficient pressure was exerted through the protecting ring on the l i d and gasket to seal the bomb against an internal pressure of 5000 p.s.i.a. Volume Compensating C e l l The volume compensating c e l l (43) was also machined from 304 stainless steel bar stock. The c e l l cavity was 7 3/^-in. high and-\ in. in diameter and therefore had a volume of about 25 cc. *The number which follows each part refers to the part number on the assembly and detail drawings shown in Figures 1-8 to 1^ 27 included at the end of the text. 1-2 Relative Location of the Cells and Method of Clamping The two cells were mounted so that the equilibrium c e l l was directly above the volume compensating one and were oriented so that the packing gland in the end of the equilibrium c e l l and the one in the end of the volume compensating c e l l were facing each other. The top bomb (or equilibrium c e l l ) was clamped on a cradle (53) located inside a constant temperature bath (104). This cradle was formed from two lengths of 2-in. channel iron held together with cross braces so as to l i e with the webs facing each other. The top web of each channel was milled so as to give a f l a t surface. The cradle was bolted at each end to the bath and was positioned to l i e above the bath center. A ^--in. thick mild steel ground plate (9) was bolted to the milled top of the cradle. This plate had a hole machined through i t s center, and the hole was of such a size that when the bottom two sections of the bomb were lowered through i t , the base of the top section was caught and rested on the plate. The bomb was clamped to the mild steel plate in the manner shown in Figure 1-8. A 5-in. wide s p l i t ring clamp (lO) was f i t t e d into a -5-in. by £-in. groove machined into the outer surface of the bomb. Since the ring was wider than the groove was deep, i t extended i^-in. beyond the outside surface of the bomb. A ring cover was slipped over the ring, holding i t firmly in placejand was bolted to the ground steel plate. When these bolts were tightened, the bomb was held firmly in place and could not be moved either upwards or downwards. It was held from rotating by two dog-point set screws (3*0-The height of the channel-iron cradle above the bottom of the bath was fixed so that the upper edge of the 2^-in. outside diameter section of the bomb was level with the top of a packing gland (8) bolted to the bath bottom. This gland was sealed to the bath with a lead gasket. Considerable care was 1-3 taken when l o c a t i n g the bath gland to ensure t h a t i t s center l i n e and that of the bomb support p l a t e were t r u l y c o - i n c i d e n t a l and thus t h a t no s t r e s s was pl a c e d on the t h i n neck of the bomb when i t was clamped i n p l a c e . The constant temperature bath was h e l d on a frame formed from lengths of 3 - i n - angle i r o n welded together. Since the bath and frame were b o l t e d together, the upper bomb, bath, and frame formed a s i n g l e u n i t . Also b o l t e d to the frame, and thus forming p a r t of t h i s u n i t , was a support assembly (ho) f o r the bottom bomb. This assembly was l o c a t e d so th a t the bottom bomb was h e l d v e r t i c a l w i t h i t s center l i n e c o - i n c i d e n t w i t h t h a t of the top bomb. The bottom bomb was h e l d to the support assembly w i t h a r i n g and cover shown assembled i n Figure I - l l . Measuring Head A drawing of the assembled measuring head i s shown i n Figure 1-10. A 22-gauge copper w i r e , i n s u l a t e d w i t h a g l a s s wool sleeve, entered at the bottom of the 3/8-in. by l / 8 - i n . t u b i n g (l6) c a r r y i n g the measuring head and was pushed up the tube t o the base of the head. At t h i s p o i n t , the wire was hard soldered t o the two i n t e r n a l w i r e s of a Conax Corporation "Con-O-Clad" s t a i n l e s s s t e e l -sheathed, m i n e r a l - i n s u l a t e d , chrome1-alumel thermocouple ( T l ) - This thermocouple, which had a sheath diameter of .062 i n . and contained two 30-gSUge w i r e s , passed through a s e a l (92) i n the head and then out of the top of the measuring head i n t o the e q u i l i b r i u m c e l l . The design of the s e a l was based on that used by Conax Corporation f o r t h e i r " T G - l U - A l M thermocouple gland. Leakage from around the outside of the thermocouple was prevented by compressing the sealant p l u g (92), machined from n a t u r a l magnesium s i l i c a t e , between a ceramic i n s u l a t o r (66) set i n t o the head below the s e a l , and a gland f o l l o w e r (67), c o n t a i n i n g another p o r c e l a i n i n s u l a t o r (68), set i n t o the head above the s e a l . Pressure was a p p l i e d t o the gland f o l l o w e r by t i g h t e n i n g a cap threaded on t o the end of the head. i-h The follower was held from rotating, as the cap was turned, hy a small key (73 )-The insulators, sealant, and gland follower were standard units supplied 'by Conax Corporation for their nTG- l4-Al" thermocouple gland. With the use of these standard pieces, a seal was made which was found to be satisfactory for pressure varying from high vacuum to 5000 p.s.i.a. Since the head, when in use, was often completely immersed in mercury, the p o s s i b i l i t y of a mercury short between the wire passing through the eenter of the head and the enclosing walls had to be eliminated. For this reason, a sheathed conductor rather than a solid one- was used. A single sheathed wire would have, of course, been quite as satisfactory,as the double one that was used, but sheathed wire pairs, designed for use as thermocouples, were more readily available. The .005-in. diameter nichrome detecting wire was held below the measuring head by three pins ( 1 7 , 1 8 , 19) set into the cap. The f i r s t pin ( 17 ) was insulated from the cap by a teflon sleeve ( 23 ) and was held in place by a nut ( 2 l ) screwed on to i t s upper end, and a small screw threaded into the side of the cap so as to pinch the teflon sleeve. The two leads from the sheathed' thermocouple were attached to the upper end of this pin and" held there by a second nut ( 2 l ) , thus making this pin part of the c i r c u i t formed by the wire coming up the inside of the measuring head tube. The nichrome wire was connected to the pin at a point . 05 in. from i t s lower end. The wire was slipped through a small hole d r i l l e d through the pin and the end tied around the pin to give a strong connection. The loose end was then passed through a small hook at the bottom of a nichrome spring (70) held by the second pin ( l 8 ) . This pin was also insulated from the head with a teflon sleeve and was held in place in the same manner as the f i r s t one. The wire was then connected to the base of the third pin ( 19 ) which was threaded into the measuring head cap. The wire was connected by passing i t through a hole in the base of the pin and tying the free end. 1-5 Since this third pin was not insulated from the head but in good ele c t r i c a l contact with i t , the measuring head tube could be used as a second lead from the detecting wire, and the problems associated with bringing two wires up the measuring head tube were eliminated. After the detecting wire was connected to the three pins,, the vertical position of each pin was adjusted so that the nichrome wire was level. The pin carrying the c o i l spring was then rotated u n t i l the wire was under a slight tension. When the wire was heated,,either by the current being carried or by heating the c e l l as a whole, the wire expanded slightly, but this expansion was taken up by the spring, and the wire remained horizontal. (Originally a . 0 0 3-in. diameter platinum wire was used as the detecting element in accordance with the recommendations of Sage and Lacy^.l However, this wire had such a short operating l i f e that i t proved to be quite unsatisfactory. Because of either solvent or mercury attack, the wire quickly formed a spongy layer on i t s surface^and this layer continued to grow u n t i l the wire-broke. When the broken wire was removed and heated in an open flame, the spongy layer was very much reduced in thickness. Examination under a microscope of a length of wire which had been previously heated, -showed the presence of wide "fins" lying along what appeared to be die marks on the wire.) Measuring Head Bridge The levels of the interfaces in the equilibrium c e l l were detected, as was described earlier, by noting a sudden change in resistance of the detecting wire as i t was raised or lowered. This change of resistance was determined by making the nichrome wire and connecting leads one arm of a Wheatstone Bridge. Since the absolute value of the resistance' change was not important, the bridge was not calibrated. A diagram of the bridge is given in Figure I - l . It was designed so Figure I - l H i that each arm had approximately the same resistance, and so that while two arms were fixed a third would he varied in order to achieve a balance. Coarse adjustment of the bridge was made with a 5-ohm variable resistor and fine adjustment with the 50-ohm one connected in parallel with the smaller resistor. When the instrument was in use, i t was found that one setting of the coarse resistor:].I was satisfactory for a l l operating conditions. A 50-0-50 micro-ammeter was used as the sensing device for the bridge. The voltage source for the bridge was a 6-volt lead storage battery. This battery was connected through a variable resistor by means of which the voltage supply to the bridge could be varied. A second fixed resistor in the connecting line was normally l e f t in the c i r c u i t when determining the mercury interface and was shorted by means of a shorting switch, for additional sensitivity, when determining the liquid-vapor interface. Measuring Head Rod Assembly The measuring head rod assembly, carrying the level detector at i t s upper end, extended from one bomb into %he other. This assembly was composed of six sections, the tubing (l6) which supported the measuring head, a universal joint (shown assembled in Figure I - l l ) joining the f i r s t and third sections, a 5/8-i n- diameter splined shaft, a l/Q-ln. diameter, 6-TPI, square-threaded screw, and a length of 3/8-in. diameter stainless steel rod. The last three sections are shown as a unit (4l). Measuring Head Rod Seal Both bombs were isolated from the' atmosphere by glands which were tight to 5000 p.s.i.a. An assembly drawing of the seal on the top bomb is given in Figure 1-9 and. one showing the seal on the bottom bomb in Figure I - l l . The seal on the lower bomb was made using a linen-laminated bakelite plastic 1-8 packing stop (75) and several turns of •5--in. square-braided, teflon-fibre packing. Pressure was applied to the packing with a gland nut (I+5) and follower (hk). This seal proved-satisfactory and originally the gland in the top bomb was designed for the same packing. However, i t was later found that the long length of packing required for a seal caused large f r i c t i o n loads inside the gland^and d i f f i c u l t y was experienced in raising and lowering the upper rod. (The corresponding d i f f i c u l t y was not experienced with the lower rod, perhaps because of the great r i g i d i t y of the lower bomb support.) For this reason, the gland in the upper bomb was shortened by inserting a monel-metal bushing (36) and was then packed with a tj--in. thick,solid teflon ring. The packing was compressed, as in the lower bomb, with a gland (6) and follower (5). ( \ Measuring Head Rod Drive An assembly drawing of the measuring head rod drive i s shown in Figure 1-12 and individual assembly drawings of the two major components of the drive (reduction gear and universal joint) in Figure 1-11. The square-threaded lead screw, which formed part of the measuring head rod assembly, passed through a threaded bronze disk which was bolted to the frame (ho) carrying the lower bomb. Thus,,if the lead screw-were rotated, i t also moved vertically, the direction of movement depending upon the direction of rotation of the screw. The lead screw could be rotated by means of a worm gear (93) and worm (92) combination having a 30 to 1 ratio. The worm gear was f i t t e d over the splined part of the measuring head rod and was held in a gear case which in turn was bolted to the lower bomb support assembly. The gear was keyed to the shaft with a relatively loose-fitting trapped key (9*0- The worm was also f i t t e d into the gear case. This gear was held by two journal bearings (90 and 87) and was keyed to a drive shaft (86). The worm assembly (shaft, 1 - 9 worm, spacers, and bearings) was designed so that i t could be installed or removed as a unit. The worm shaft was driven by a l/6-horsepower A.C reversing induction motor through a 5 0 to 1 commercial reduction gear unit. Since one turn of the rod assembly raised the rod l / 6 in., and since the 1 7 2 5 r.p.m. motor was connected to the rod through a 5 0 to 1 reduction gear and 3 0 to 1 worm and worm-gear set, the rate of vertical movement of the rod was . l 8 0 in. per minute. As was described earlier, the copper lead from the nichrome detecting wire passed down the center of the measuring head tube and out from the tube to.a Wheatstone Bridge. In order that the tube, and thus the copper lead, would not rotate along with the lower part of the measuring head rod assembly, the two sections were joined by means of a universal joint (Figure I - l l ) . An SKF " 2 3 0 5 " self-aligning b a l l bearing ( 8 l ) wasl'held in a mild steel case ( 5 1 ) which in turn was threaded on to the upper end of the splined lead screw. The case was locked in position with a -j-in. flat-headed screw ( 8 2 ) . The bearing was trapped in the case by a mild steel l i d ( 5 3 ) whose outside diameter was such that i t extended \ in. beyond the outside diameter of the case. The center of the l i d was d r i l l e d out to an inside diameter of 1 ^ in. Since the diameter of the center shaft ( 7 9 ) carrying the bearing was only 1 . 2 6 in. at the point where i t passed through the l i d , ample freedom of movement was available to allow the self-aligning bearing to help compensate for any slight misalignment of the two shafts. The 3 / 8 - i n . by l / 8 - i n . diameter tubing which formed the upper end of the rod was threaded into the bearing center shaft ( 5 0 ) a n d was pinned in place by a £-in. socket-head pin ( 7 9 ) . The thread for the tubing was at the base of a carefully ground hole and was machined slightly oversize so that the tubing center line was fixed by the ground sides of the hole,and not by the thread i t s e l f . This I-10 design was used to align the two shafts as i t was found to be impossible to make the tubing and shaft centerlines co-incident when the thread i t s e l f was used as the locating device. Measuring Head Rod Drive Circuit A diagram of the measuring head rod drive c i r c u i t is given in Figure 1-2. A four-pole, double-throw switch was used to control the direction of rotation of the motor. When this switch was reversed, the polarity of the starting c o i l with that of the f i e l d c o i l was reversed, and therefore the direction in which the motor would start was also reversed. Four microBwitches, which were actuated by the l i p on the universal joint assembly, were used to -limit the travel of the measuring head. The four switches acted in two pairs and, since the pairs were symmetrical, only the operation of the upper limiting switches w i l l be described. A Minneapolis Honeywell Regulator Company'"ISLI" two-circuit microswitch was wired in combination with a double-pole, double-throw relay in such a manner that the ^ two were equivalent to one double-pole, double-throw switch. When the rod assembly was raised to such a position that this switch was opened, the lines to both the f i e l d and starting coils of the motor were broken, and a pi l o t light was turned on. The rod could then be raised no further, but reversing the position of the four-pole switch took the actuated microswitch out of the cir c u i t , and the rod could be lowered. A second microswitch, a "BZ-2RW82" single-pole, double-throw switch, was mounted a few hundredths of an inch above the f i r s t to guard against the possibility of failure of the f i r s t switch. This second switch, when actuated, opened the voltage supply line. When this switch was opened, no power could be supplied to the motor regardless of the position of the four-pole motor switch, and the motor had to be turned by hand u n t i l the switch was closed again. MEASURING HEAD Figure 1-2 1-12 Bomb Tubing Connection The details of the ^--in. diameter tubing connections on the upper and lower bombs i s given in Figure 1-10. These connections were based on the standard design used by the American Instrument Company^ and were made so as to meet-Perry's^ recommendation of having at least four f u l l threads engage on the gland nut (60) when sealing against pressure of up to 2000 atmospheres. A l l other tubing connections shown in Figure I (i.e. connections on valves, , tees, cmsses, couplings, etc.) were made according to the specification of the manufacturer of each item. In every case the unit and connections were specified as being suitable for pressures of 10,000 p.s.i.a. or greater. Equilibrium C e l l Stirrer A magnetically driven stir r e r was used to agitate the contents of the equilibrium c e l l . An assembly drawing of the stir r e r i s given in Figure I-10. A small well was machined in the underside of the l i d used to seal the c e l l . A General Electric "22U18B" Alnico II magnet (28) was placed in this well and supported there on the end of a shaft passing through two stainless steel b a l l journals (48). The journals were f i t t e d into a stainless steel bar (26) held on the underside of the l i d by two screws (30 and 38)- When the supporting bar and magnet were installed, the magnet's position was adjusted by sliding the magnet shaft through the two bearings, u n t i l i t was as high in the bomb as was possible without touching the under-surface of the l i d . A stirrer ( 2 5 ) , consisting of a l/8-in. thick bar carrying two long l4-gauge nichrome wire vanes, was held on the magnet shaft below the support bar by two nuts (29 and 3 l ) - The length of the nichrome vanes was such that under operating conditions they extended into the liquid solvent but not the mercury phase. A ring formed from 20-gauge nichrome wire was welded to the vanes near the bottom to act as a stiffener. 1 - 1 3 When the stirr e r was allowed to rotate freely, i t would follow the driving magnet up to a limiting speed and would then drop out. No external indication was available to show whether or not the stirrer was operating. The c e l l s t i r r e r was therefore designed so that one of the screws ( 3 0 ) holding the support bar ( 4 4 ) extended down below the l i d and prevented the stir r e r from making a complete revolution. With this design, a distinct "ping" could be heard through a rod placed against any part of the equilibrium c e l l every time the stirrer h i t the screw. The oscillating motion of the stirrer, resulting from the stirrer hitting the screw and bouncing back, seemed, in a test made outside of the bomb, to s t i r very effectively. The stirrer was driven with an externally mounted " U 5 3 0 4 " horseshoe type Alnico V General Electric magnet ( 1 5 ) f i t t e d into the inside of the protecting ring ( 4 ) on the upper bomb. A small brass nut ( 1 2 ) threaded into the bomb cap ( 2 ) acted as a bearing for the shaft used to rotate the magnet. This nut also acted as a stop for a collar ( 1 3 ) which was f i t t e d over the magnet shaft and which was used to adjust the vertical position of the magnet. The driving magnet was rotated by a variable-speed stirrer motor. This gear-driven unit, in which the speed could be varied by changing the supply voltage, was coupled to the magnet shaft through a very loose f i t t i n g two-pin coupling and was mounted on one of the plates forming the l i d of the bath. The upper bomb, i t s l i d , and i t s protecting ring were a l l made of 3 0 4 stainless steel in order to keep flux losses between the two magnets to a minimum. It was found that when the thickness of the l i d at the bottom of the well (i.e. the minimum distance by which the two magnets were separated) was 0 . 5 in., as required for a safe design for 5 0 0 0 p.s.i.a. internal pressure, the maximum velocity at which the stirr e r would respond to the driving magnet was very low. Since the maximum pressure at which i t was planned to use the equipment i n i t i a l l y was not 5 0 0 0 p.s-i.a., but 1 5 0 0 p.s.i.a., the thickness of the material at the bottom of the l i d was reduced to 0 . 2 5 i n- With this I-Ik thickness separating the 'two magnets, the highest velocity at which the stirrer would follow the driving magnet was raised to approximately 200 r.p.m.? with the range 100 to 150 r.p.m. being the most satisfactory. This change in thickness of the l i d meant that the equipment as discussed above i s suitable only to 1500 p.s.i.a. internal pressure; however^if this l i d were replaced by one whose minimum thickness were \ in., as was done for testing purposes, the equipment could be used safely at pressures up to 5000 p.s.i.a. Sampling Tube Assembly The sampling tube assembly was constructed from 316 stainless steel pressure tubing 9/l6 in. in outside diameter and 5/l6 in. in inside diameter-. The assembly was divided in two sections, joined by a 9/16-in. by 9/16-in. by •5--in. tubing tee, and the two ends of the joined sections terminated in 9/l6-in. by -5-in. by.^-in. tubing tees. The assembly had an overall length of 19 in. and was connected to the equilibrium c e l l through valves at three points. One length of -5-in. tubing connected the top of the sampling assembly to a point on the c e l l 1-5- in. below the l i d . Since the bomb cap extended below this point on the bomb, a hole was d r i l l e d through the cap to allow the tubing and gland nut to pass through. A second length of ij-in. tubing connected the middle tee (located 12 in. from the top of the assembly) to a point on the c e l l 5/8 in- from the top. A third length of tubing connected the bottom of the assembly to,a point 3/8 in. above the bottom of the equilibrium c e l l . The valve in each line was positioned so that i t lay. below the corresponding bomb connection, and therefore, i f the tubing connecting the valve to the bomb was once f i l l e d with mercury, i t would remain f u l l as long as the valve was not opened. High Pressure Equipment Company Model "30-11HFV' teflon packed, non-rotating-stem valves were used in the three £-in. tubing lines connecting the 1-15 c e l l to the bath. Since the valves were located in the bath, they were opened or closed by means of extensions on the valve stems. Constant Temperature Bath The constant temperature bath in which the equilibrium c e l l was immersed was 28 in. in diameter, 30-in. high and constructed from 12-gauge, 3l6 stainless steel sheet metal. In order to reduce heat losses from the bath, the top, bottom, and sides were a l l insulated. The sides were covered with a 4-in. thick layer of glass wool which in turn was covered with a ^ - i n . thick layer of a powdered asbestos and portland cement mixture. A f i n a l layer of cotton canvas was glued to the asbestos and then painted to give an o i l resistant fin i s h . The bottom of the bath was also covered with a k-ln. thick layer of glass wool, and.this layer was held in place with a plywood frame. At those-locations in the sides and bottom where tubing passed through the bath wall,to the equilibrium c e l l , small sections were cut from the insulation and packing glands were f i t t e d into the wall. The top of the bath was covered with an insulated l i d . This l i d was used both to support the insulation and to prevent fumes from the hot o i l escaping into the room. It was made from lengths of 1-in. mild steel tee sections welded together at right angles, base down, so as to form a web with the same overall shape as the bath. The ends of the tee sections extended out over the insulated sides of the bath to act as supports. A 1-in. deep ring, 7/8 in. less in diameter than the inside diameter of the bath, was welded to the underside of the tee sections. This ring was used.to give r i g i d i t y to the web and, when wrapped with two lengths of ^ --in. braided jute rope, acted as a splash seal for the l i d . The spaces between the uprights of,the tee sections were f i t t e d with 12-gauge mild steel plates. These plates were bolted to the base of the tees and then covered with shaped sections of glass wool insulation 1-16 h in. thick. The use of a l i d of this type offered a considerable advantage over one formed from a single piece. Access to any particular part of the bath was possible simply by removing the appropriate insulation block and plate. The sub-divisions also permitted many different pieces of equipment to be individually f i t t e d and withdrawn. The bath, when in use, was f i l l e d with a high temperature l u b r i -cating o i l . When this o i l was heated above 200°C?very unpleasant fumes were produced, and i t was found necessary to vent the bath. A vent of 2-in. galvanized steel drain pipe ran from the bath l i d to the roof of the building. An a i r injector in the vent was found to provide ample draft to remove the fumes. Heat could be supplied to the bath vby means of three heaters or removed from i t with a cooling c o i l . Two of the heaters, both rated at 230 volts and 2500 watts, were threaded into the bath from the bottom. The third heater, rated 110 volts and 1000 watts, was hung from one of the plates in the top of the bath. The cooling c o i l was formed from a single•length of 3/8-in. diameter copper tubing which was coiled so as to l i e along the side of the bath over about 2/5 of i t s circumference. The c o i l was located in this manner in order that the bomb support could be installed or removed without removihg^,the copper tubing. The two ends of the tubing passed out of the bath bottom through packing glands and then to the water supply and drain. The c o i l was purposely formed from a single length of tubing so that there would be no possibility of a water leak into the hot o i l with the resultant rapid generation of steam. A ^--horsepower Greey "Lightnin Model S-2" stirr e r driving a shaft f i t t e d with two 3-bladed stirrers provided adequate stirring. 1-17 Temperature Control of the Bath The temperature of the hath was regulated with a control c i r c u i t (shown in Figure 1-3) in which the sensing element was a Leeds and Northrup 25-ohm platinum "thermohm". The thermohm was connected to a Leeds and Northrup •"Model No. 8067" Mueller Bridge, and the resistance that this sensing element would "have had i f the bath were -at the desired temperature was set on the bridge dials. The out-of-balance signal resulting from any difference between the actual and the desired resistance of the thermohm was fed from the bridge to a Leeds and'Northrup D.C. null.detector. In this c i r c u i t , the null•indicator was used only as a D.C pre-amplifier and the amplified signal was sent to,a Leeds and Northrup•"model 6000 Speedomax" recorder-controller. The control c i r c u i t on the Speedomax was set so that the control point was zero voltage input. The control switch was thus opened or closed depending upon whether the input signal was positive or negative. A lead from a 110-volt line was fed through this switch to a 110-volt AC, 6-volt DC r e c t i f i e r which in turn supplied the activating current for a "heavy duty mercury relay. The relay was used to open or close the line from an autotransformer to one of the 220-volt,,2500-watt heaters. The other two heaters were suppliedwith power from autotransformers as well, but only the f i r s t one was operated through the relay. When the control c i r c u i t was operating, one of the '"steady-on" heaters was adjusted through, the autotransformer so as to supply most of the-heat input, and the "off-on" heater so as to•supply only a small fraction of the total,lead. The autotransformers used to adjust the power supplied to the three heaters were mounted on a common panel board. Included as well on this panel board were either a voltmeter or an ammeter or"both for each heater, in order that an estimate could be made of the power being supplied, and a panel light TEMPERATURE CONTROL CIRCUIT Figure 1-3 1-19 .to indicate whether or not the control heater was on. It was found possible, with the ci r c u i t described above, to control the bath temperature.at a point near the equilibrium c e l l and about 12 in. from the sensing thermohm to within +.01°C. of the set point. However, the temperature variation could be kept within these limits only with very careful adjustment of the bath heaters, and therefore variations of up to +.02°C were allowed. The accuracy of the temperature control was also affected by changes in room temperature. One of the units in the control c i r c u i t was sensitive to temperature changes, and over a period of six or seven hours, a set point d r i f t of up to .1°C. would sometimes occur. Because of the possi b i l i t y of this d r i f t , no phase equilibria measurements were made unless a frequent check chowed that the set point temperature had remained within +.03°C of the desired value for at least one hour. Temperature Measurement The temperature of the bath was measured with a "Model 8l62", serial number 16931^ Leeds and Northrup platinum resistance thermometer suspended from the bath l i d at a point about 2 in. away from the equilibrium c e l l . A calibration report, issued by the National Research Council of Canada on October 26, 1959, w a s used to convert the thermometer resistance, as measured on a Leeds and Northrup "Model 8067", serial number 16931^ Mueller Bridge, to a temperature reading in degrees Centigrade. (The thermometer was issued with a report rather than a certificate as the purity of the platinum wire was found, by the National Research Council, to be less than that specified in the International Temperature Scale of 19^8.) The value of the-ice point resistance, used in the conversion of the thermometer resistance reading to a temperature one, was obtained using d i s t i l l e d water ice. Since the temperature in the equilibrium c e l l was determined by 1-20 measuring the bath temperature and not that of the c e l l directly, a check was made to see how closely the two agreed. Eight iron-constantan thermocouples were strapped to the bomb at the 'locations shown in Figure I-kA, the bath heated to approximately. 150°C., and the temperature of the bomb measured using the thermocouples. The bath was then cooled, some of the thermocouple positions changed to those shown in Figure I-hB, and the bath reheated to approximately 150 C Once the thermocouple temperatures lhad been read, the bath was heated to approximately 200°C, and the readings repeated. Further readings were made .at 250°C. The four sets of readings obtained are given in Table I - l . The values given in Table I - l are not quoted to show the range of temperature variation over a control cycle, as the control device used when they were measured was an earlier design much inferior to that f i n a l l y used, but to give the temperature variation over that part of the bomb occupied by the equilibrium c e l l , and to compare the c e l l temperature to the temperature at the point where the thermometer i s located. While the thermocouples were not calibrated, they were a l l cut from the same length of wire, and the junctions were a l l similarly made. On this basis i t is unlikely that the calibration of the eight thermocouples would be very different. TABLE I - l RESULTS OF TEMPERATURE SURVEY OF EQUILIBRIUM CELL Bath Temperature approximately 150°C Thermocouples position A Thermocouple Number Reading (millivolts) 1 7.8905 - 7-8935 2 7.8886 - 7-8916 3 7.8960 - 7-8976 k 7.8959 - 7.8976 5 7.8960 - 7.8981 6 7.8959 - 7-8987 7 7.8965 - 7.8980 8 7.8970 - 7-8995 8*-i_jr8 3C6 1 - 2 1 ! I 3X ><4 3>C X4 POSITION A POSITION B THERMOCOUPLE POSITION FOR TEMPERATURE SURVEY Figure I-k 1-22 Bath Temperature approximately 150 C. Thermocouples position B Thermocouple Number Reading (millivolts) 1 7.8855 - 7.886O 2 7.8840 - 7-8850 3 7.89OO - 7.8912 4 7.8895 - 7.892O 5 7.8905 - 7-8915 6 7.8905 - 7.892O 7 '7.8910 - 7.8916 8 7.8905 - 7-8915 ?emperature approximately 200°C Thermocouples po Thermocouple Number Reading (millivolts) 1 IO.969O•- 10.9720 2 IO.968O IO.97OO 3 .10.9795 - IO.98IO 4 IO.979O - IO.9815 5 IO.976O - 10.9770 6 IO.9765 - IO.9785 7 IO.976O - IO.9775 8 10.9765 - 10-9775 Bath Temperature 250°C. Thermocouple position B Thermocouple Number Reading (millivolts) 5 13.5221 (avg.) 6 13.5230 7 13.5220 8 13-5155 Over the temperature range considered in Table I - l , .054 mill i v o l t s is equivalent to about 1°C As can be seen from the tables, the surface temperature of the bomb atl50°C. was essentially uniform from the bottom of the equilibrium cell,and was equal to that of the o i l bath where the platinum thermometer was located. At 250°C, a temperature gradient of about 0.1°C existed between the top of the cap and a point just below the top of the equilibrium c e l l , and the temperature was uniform to within 0.02°C. over the t - 2 3 rest of the cell. As a result of this test, i t was felt that i t was unlikely that measuring the bath rather than the bomb temperature would result in a significant error when reporting the temperature to 0 .1° C., as long as the top of the bomb was covered with bath fluid. Measurement of Pressure The pressure in the equilibrium cell was measured with one of two Heise Bourdon tube pressure gauges. The fi r s t gauge, serial number H21870, was used for the pressure range 0 to>150 p.s.i.g. It had a 12-in. face, with 0.2 p.s.i- divisions, and was stated by the manufacturer as being accurate to within +.15 p.s.i. A calibration obtained from the manufacturer for this gauge is given in Table 1-2 and a similar one obtainedby the author near the completion of this research,using a Barnett dead weight tester model number MKT serial number 0131^ is given in Table 1-3. The second gauge, serial number H21869, was used for the pressure range 150 t o l 5 0 0 p.s.i.g. It was the same size as the first gauge, and the face was divided into 2 p.^s.i. intervals. The accuracy was stated by the manufacturer as being + 1.5 p.s.i. and calibrations similar to those for the other gauge are given in Tables I-h and 1-5-TABLE 1-2 CALIBRATION SUPPLIED BY MANUFACTURER; FOR HEISE PRESSURE GAUGE N0.2187O Dead Weight or Mercury Column Reading Gauge Deviation from Dead Weight or Mercury Column 7.5 p.j.t. 15 — 22-5 — 30 --37-5 --45 --52.5 - -60 - -67.5 — 75 --82.5 - -90 — 97-5 --105 112.5 — 120 - -127-5 - -135 — 142. 5 - -150 Maximum Hysteresis 1-5 p.s.i. Corrections are noted where error is 1-5 p.s.i. or more 1-25 TABLE 1-3 CALIBRATION' DETERMINED USING A BARNETT DEAD WEIGHT TESTER MODEL NUMBER MKT SERIAL NUMBER 0131 Dead Weight Reading Gauge Deviation from Dead-Weight Reading 11 P.5.U Increasing - Decreasing , 21 . 31 • hi 0 - 0.2 - 51 0.05 - -3 61 0.0 - 0.2 71 0.0 - 0.2 81 0.0 - 0.3 91 .0.15 - 0.3 101 0.2 - 0.3 111 0 - 0.2 121 0.05 - 0.2 131 ,0.3 l U l 0.3 IL9 Corrections are indicated where error i s more than .15 p.s.i. 1-26 TABLE 1-4 CALIBRATIONVSUPPLIED BY MANUFACTURER?". FOR HEISE PRESSURE GAUGE No. 21869 Dead Weight or Merc=ury Column Reading . Gauge Reading (Deviation from Dead Weight or' Mercury Column) 75 p.s.i. 150 — 225 --300 --375 --450 — 525 — 600 --675 — 750 --825 --,900 — 975 --1050 --1125 \ --1200 — 1275 --1350 — 1425 — 1500 Maximum Hysteresis 1.5 p.s.i: Corrections are noted where error is 1.5 p.s.i. or more 1-27 TABLE 1-5 CALIBRATION DETERMINED USING A BARNETT DEAD WEIGHT TESTER MODEL NUMBER MKT SERIAL NUMBER 0131 Dead Weight Reading Gauge Deviation from Dead .Weight Reading 200 P-S.L. 300 --Uoo — 500 --600 --700 800 --900 1000 — 1100 1200 — 1300 1U00 — 1500 - — Corrections are indicated vhere error is more-than 1.5 p.s.i. Maximum Hysteresis 1.5 p.s.i. 1-28 Associated Glassware Associated with the equipment described above were five auxilliary glassware unite. The f i r s t of these units was essentially a vacuum manifold which was connected, through a cold trap, to a mercury diffusion pump. The pressure in the manifold could be measured with a mercury-filled McLeod gauge which was connected to the manifold through a high vacuum stopcock. The second of the glassware units, which was based on apparatus described by Kay and Donharn^ .^ , 'was designed to.remove dissolved gas from •the solvents to be studied'in the equilibrium c e l l . A diagram of the unit i s given in Figure 1-5- It consisted of seven glass vessels (F 1 to 7), each k.5 cm. in diameter and 33 cm. long, and each connected•through a-high vacuum stopcock to an 8-mm. tube. This tube was joined in turn to ,the vacuum manifold. The pairs of vessels F 2 and F 3, F k and F 5, and F 6 and F 7 are functionally associated with the 1-litre round bottom flasks F 10, F 9, and F 8 respectively. Each functional group was used for one material. Vessel F 1 was used in common for a l l materials. Solvent was i n i t i a l l y added to a 1-litre flask (for example, F 10) and the flask was then connected .through a Quickfit and Quartz "CCB19" ground mercury sealed cup cone and socket joint to one of the vessels (for example, F 2). The line joining ,the two pieces contained an 8-turn c o i l , 8 cm. in diameter for flexibility^and a high vacuum stopcock. The third glassware unit was used to store deaerated material from the transfer apparatus before, i t was introduced to the equilibrium c e l l . A diagram of this flask i s given in Figure 1-6. It consisted of a 30-cm. high • double-walled flask whose outer and inner walls were formed from tubing 7U-mm. and 36-mm. in diameter respectively. The annular space was connected at MERCURY SEALED UNGREASED STOPCOCK FLASK MERCURY STORAGE FLASK Figure 1-6 1 - 3 1 the top, through an ungreased mercury-sealed stopcock S ' 1 3 , to the-common transfer vessel F 1 , and, at the bottom, through a mercury-cut-off valve S I , to a glass-to-metal kovar seal. This seal was used to join the glass line from the storage chamber to a length of -j^-in. metal tubing and the high pressure valve V 4 connected to .the top of the equilibrium c e l l . The fourth of the glass auxilliary units, shown in Figure 1 - 6 , was used as a mercury storage reservoir for the High Pressure Equipment Company Incorporated "model 6 2 - 6 " mercury pressure generator. It consisted of a 17-cm. high glass vessel J made'from 5 1 - ™ i - tubing, connected at the top to the high pressure side of the mercury diffusion pump and at the bottom through a glass c o i l to a glass-to-metal kovar seal. This seal was in turn connected to a high.pressure valve V 4 which could be used to isolate the reservoir from the pressure generator. The glass c o i l was included to prevent any stress being placed on the kovar seal as the volume, and thus weight, of mercury, in the reservoir was changed. Mercury could be added to the reservoir through a glass side arm connected near the top of the reservoir. The side arm was joined through a-barometric leg to the bottom of a short length of 22-mm. glass tubing H. Mercury could be poured into this tubing, through a funnel formed from a ground glass socket, and then transferred through the barometric leg to the mercury reservoir by atmospheric pressure. The f i n a l glassware unit, shown in Figure 1 - 7 , was used to collect samples from the pressure sampling tube assembly D in the bath. It consisted of three small flasks connected in parallel. The line joining the flasks was connected through a kovar glass-to-metal seal and valve V 1 3 to the sampling tube assembly and, through a mercury cut-off valve S 1 2 , to the vacuum manifold. Each of the flasks was joined to the common line through a Quickfit and Quartz mercury sealed " C C B 1 9 " cup cone and B19 socket joint identical to those used on the one-litre solvent flasks described earlier. The nature of these greaseless 1-32 M MERCURY-SEALED GROUND GLASS JOINT LIQUID SAMPLE COLLECTION TUBE VAPOR SAMPLE COLLECTION T U B E SAMPLE COLLECTION TUBE S 12 ACUUMI MERCURY C U T - O F F VALVE TO PRESSURE SAMPLING TUBE rh GLASS TO L-H METAL SEAL K MERCURY TRAP Figure 1-7 1-33 vacuum tight joints allowed each flask to be easily removed or replaced and, since the joint was not lubricated, there was no possibility of the samples becoming contaminated with stopcock grease. The f i r s t of the-three flasks, Flask K, was used as a mercury trap. It was made from 17-mm. tubing approximately 5 in- long and was located directly below the kovar seal. The second of the flasks, Flask L, was used to collect vapor samples. It was approximately 10§- in. long and was formed from two sizes of tubing. The top k in. were of 19-mm. tubing and the'bottom 5? in- of either 6-mm.,^ 8-irkrrr,; or 10-mm. tubing, depending upon the size of vapor sample to be collected. The third flask, Flask M, was used to collect l i q u i d samples. It was similar in shape but longer than the second. The top portion was of 19-mm. tubing 5 in- long and the bottom of 10-mm. tubing 8 in. long. 1-3^ Figure 1-8 1-35 Figure 1-9 Figure I-10 Figure 1-11 1-38 Figure 1-12 1-39 Figure 1-13 MILD STEEL - ONE REQ D PART NUMBER 7 H-i-h DETAIL DRAWING NUMBER 2 PARTS NUMBER b, 6, a 7 i C A L E » r i . ' L L S I Z E DW 8/6/61 N F THREAD \ —5-° 0025,- 0037 J air - 3 7 7 " + 0 0 2 + 004 •FLATS FOR WRENCH 304 STAINLESS STEEL ONE REQ'D PART NUMBER 5 Figure I-lk i R-4 _l£ , !2 TPt - « 4 -MILD STEEL - ONE REQ'D PART NUMBER 6 A L L D IMENS IONS IN I N C H F S I o MILD S T E E L - ONE REQ ' D PART NUMBER 9 DETA I L DRAWING N U M B E R 3 P A R T S N U M B E R 8 8 9 i - t i - E - FJI.L SIZE D W 9/6/61 J- N C T A P - 6 HOLES E Q U A L L Y S P A C E D ON 4 - J f l N C H CIRCLE N C T A P - 2 H O L E S E Q U A L L Y S P A C E D ON 3 INCH D I A M E T E R C I R C L E C E N T R E L I N E O F T H E S E H O L E S M U S T P A S S T H R O U G H C E N T R E L I N E O F L O C A T I N G H O L E S ON P A R T N U M B E R I i _ T \--Z3ia.-y 0 0 3 , - 0 0 0 —^ / - 2 5 0 " . + 0 0 3 , - 0 0 0 -Y77ZZZA 3 N F T H R E A D MILD S T E E L - ONE R E Q 0 P A R T N U M B E R 8 Figure 1-15 ! t—1 C U T WIDTH L E S S T H A N ^ - C L E A R A N C E HOLE FOR ^-INCH B O L T - 6 H O L E S E Q U A L L Y S P A C E D ON 4 ^ P D C i , r- 7 — 4 1 1 1 J 1* t 7 ,. 1 J BREAK BOTH INNER A N D O U T E R C O R N E R S TO SPL IT RING C L A M P MILO S T E E L - ONE R E Q ' D P A R T NUMBER 10 . B R E A K C O R N E R T O ^ MILD S T E E L - ONE REQ 'D P A R T NUMBER II DETAIL DRAWING NUMBER 5 PARTS NUMBER 10, II, 12, 8 13 S C A L E • F U L L S IZE D W 22/6/61 I S E C T I O N A H | » C T H R E A D STAINLESS STEEL ONE REQ'D PART NUMBER 16 TEFLON - ONE REQ'D PART NUMBER 23 l ° -oso! •J V S OVT-STAINLESS STEEL ONE REQ'D PART NUMBER 17 A 0 7 5 -J U STAINLESS STEEL ONE REQ'D PART NUMBER 19 k—2~H 8 u 12I* ' T • a LU 4° t - J 1 1 A L L DIMENSIONS IN INCHES P A R T S 17,18, 19, 2 0 , 2 1 , 2 2 , 23 - T W I C E F U L L S I Z E PART 16 - F U L L S IZE STAINLESS STEEL ONE REQ'D PART NUMBER 20 1 ° -^1 STAINLESS STEEL ONE REQ'D PART NUMBER 18 JJ.iL 1-TEFLON-ONE REQ'D PART NUMBER 22 4 - 4 8 T H R E A D STAINLESS STEEL - THREE REQ D PART NUMBER 21 DETAIL DRAWING NUMBER 6 PARTS NUMBER 16 - 23 S C A L E AS N O T E D D W 12/6/61 Figure 1-18 o V ^ L 7 I-U-16 WIRES SPOT WELDED TO VANES AND TOGETHER PART NUMBER 25 STAINLESS STEEL - ONE REQ'O STIRRER VANES ALL DIMENSIONS IN INCHES GROOVE FOR 020° NICHROME RING CLAMP TWO GROOVES 310 INCHES APART STAINLESS STEEL ONE REQ'O PART NUMBER 26 F37S. g STAINLESS STEEL ONE REQ'O PART NUMBER 27 G E ALNICO MAGNEL NO 2 2UI8B -FORCE FIT ran" SHAFT OF STAINLESS STEEL - ONE REQ'D PART NUMBER 28 STAINLESS STEEL ONE REQ'O PART NUMBER 29 PARTS 25, 26, 28 - FULL SIZE PARTS 27, 29, 30, 31 - TWICE FULL SIZE -8 - 32 THREAD -HFE i _ STAINLESS STEEL ONE REQ'D PART NUMBER 30 STAINLESS STEEL ONE REQ'D PART NUMBER 31 DETAIL DRAWING NUMBER 7 PARTS NUMBER 25 - 31 SCALE AS NOTED DW 13/6/61 Figure I-19 Si Pt-I ft HIGH TENSILE STRENGTH ALLOY STEEL ONE REQ 'O PART NUMBER 62 — (! -0016, -0025 STAINLESS STEEL ONE REQ 'D PART NUMBER 23 I — 37 - 0025 - 0037 S3 L_isaJ j^TAP- 2 HOLES FROM BOTOM J l 1 , 5 ^ STAINLESS STEEL ONE REQ 10 PART NUMBER 68 ALL DIMENSIONS IN INCHES PARTS 23, 62 • FULL SIZE PARTS 36, 68 - TWICE FULL SIZE MONEL METAL ONE REQ1D PART NUMBER 36 DETAIL DRAWING NUMBER 8 PARTS NUMBER 23, 36, 62, 68 SCALE AS NOTED D W 22/6 SI Figure 1-20 H i Figure 1-21 1-49 Figure 1-23 -10-24 TAP - 6 HOLES -|- DEEP 13 D EQUALLY SPACED ON 2-j§- CIRCLE , 13 2 4409 2 4421 -0591" — 2 165 Ha -CLEARANCE HOLE FOR 10-24 SCREW 6 REQ'D DN 2 j | P D CIRCLE "^f NC TAP COUNTER SINK FOR FLATHEAD SCREW PART NUMBER 51 MILO STEEL - ONE REQ 'D -Hi 2 165 -2 441 rjzzz PART NUMBER 53 MILO STEEL - ONE REQ'D T T 32 TPI PART NUMBER 52 MILD STEEL - ONE REQ'D DETAIL DRAWING NUMBER 12 PARTS NUMBER 50, 51, 52. a 53 SCALE • FULL SIZE D W 16/6/61 32 • " TPI PART NUMBER 50 MILD STEEL - ONE REQ D PARTS 8 4 . 88 , 9 4 - T W I C E F U L L S I Z E PART 9 3 - F U L L S I ZE A L L D I M E N S I O N S IN I N C H E S r-1 f t ONE REQ'D - MILD STEEL PART NUMBER 94 M A C H I N E K E Y W A Y IN WORM G E A WORM GEAR PART NUMBER 93 DETAIL DRAWING NUMBER 13 PARTS NUMBER 84, 88, 93, 94 S C A L E AS N O T E D D W I9/6/6I H I H 3936 MIN _ 3939 MAX PART NUMBER 86 MILD STEEL - ONE REQ ' D CLEARANCE HOLE FOR 6-32 TAP 4 HOLES ' i — l i t PART NUMBER 85 MILD STEEL - ONE REQ'D -CLEARANCE HOLE FOR 8-32 TAP \ 4 - HOLES X^ I 850 -— I 614 I—L -1*1-PART NUMBER 89 MILD STEEL - ONE REQ'D PARTS 86, 91 - TWICE FULL SIZE PARTS 85, 89 - FULL SIZE 9056 6 69 — 1 1 t - — i — • PART NUMBER 91 MILD STEEL - ONE REQ'D DETAIL DRAWING NUMBER 15 PARTS NUMBER 65, 86, 89,8 91 SCALE AS NOTED D W 21/6/61 Figure 1-26 I vn ro P A R T N U M B E R 7 5 P A R T N U M B E R 7 7 P A R T N U M B E R 7 8 S P L I N E D S H A F T L A M I N A T E D B A K E L I T E P L A S T I C M I L D S T E E L - O N E R E Q ' D M I L D S T E E L - O N E R E Q ' D P A R T N U M B E R 4 9 B R O N Z E - O N E R E Q ' D D E T A I L D R A W I N G N U M B E R 14 P A R T S N U M B E R 4 9 , 7 5 , 7 7 , 8 7 8 S C A L E • F U L L S IZE D W 2C/6/6I M I VJl V>J Figure 1-27 APPENDIX II - DETAILS OF PROCEDURE I l - i Cleaning of Equipment The apparatus was thoroughly cleaned after i t was assembled and before i t was used. I n i t i a l l y , at least six ^--litre volumes of hot water in which detergent had been dissolved were forced through each line and f i t t i n g (with the exception of the associated glassware). These detergent washes were followed by a thorough tap water rinse. The rinse water was removed with several technical-grade acetone washes, and the equipment was then washed with three lots of technical-grade benzene, and f i n a l l y with three lots of once-distilled, technical-grade acetone. Following each of the acetone washes, dry nitrogen was blown through the lines. After the f i n a l acetone wash, the entire apparatus was purged with nitrogen, and f i n a l traces of solvent were removed by evacuating the equipment for several hours using the diffusion pump. The glassware was cleaned before assembly. Each component was placed in a bath of boiling water and detergent for several minutes and then thoroughly rinsed with tap and d i s t i l l e d water. The rinse water-was removed by oven drying, and any moisture introduced when the components were joined together was removed by prolonged evacuation of the completed unit. Introduction of Mercury and Definition of Pressure Gauge Zero Once the assembled equipment had.been cleaned, i t was evacuated for 2k hours. The mercury storage flask J (Figure l ) was then f i l l e d with mercury from the funnel H. During the transfer of mercury from the funnel to the flask, the top of the flask was connected to the vacuum pump so that any air which had been dissolved in the mercury would be pumped off. When the flask was f u l l , valve V 11 was opened and the mercury was allowed to run into the system. This procedure was repeated u n t i l the entire apparatus, with the exception of the equilibrium c e l l and the glass units, was f i l l e d . II-2 Since the pressure gauges F and G were supplied"hy the manufacturer with an undefined zero, i t was necessary to choose ar b i t r a r i l y a value for zero pressure. This arbitrary zero value was defined in the following manner. Mercury was introduced to the equilibrium c e l l A u n t i l the level was at a point 1 cm. above the lowest tubing connection to the bomb. The measuring head rod was then positioned so that the detecting wire was at the level of the interface, and the height of a graduation on the measuring head rod was read on ; a scale located beside the rod. This scale reading was found to be 2.4 cm. The c e l l was next opened to atmospheric pressure and the pressure gauge scales set to read zero. Zero pressure for the gauges was therefore defined as that pressure exerted by a column of mercury in the equilibrium c e l l which corresponded to a reading of 2.4 cm. on the measuring head rod scale when the pressure in the c e l l above the mercury was the same as that on the outside of the gauge. Since this method of locating the gauge zero required that the c e l l be open to the atmosphere, and since i t was impossible to allow air into the c e l l when using the equipment, a secondary calibration point was defined which could be used to check the position of the gauge scale. The mercury level in the mercury storage flask J was adjusted to an arbitrary reproducible level ?and the flask was connected to the gauges through valves V 6, V 7, V 9, V 11. The gauge pressure corresponding to this level of mercury in the flask was read and recorded. The gauge zero could then be checked at any time simply by adjusting the mercury level in the storage flask to the 'correct position and comparing the gauge reading with the value i n i t i a l l y obtained. A correction had to be made, of course, for any change in atmospheric pressure from that at the time when the i n i t i a l value was obtained. Conversion of Gauge Reading to Absolute Pressure in the Equilibrium Cell Three corrections must be applied to the value read on the pressure H - 3 gauges before the pressure in the equilibrium c e l l can be obtained. 1. Correction for atmospheric pressure The pressure as read on the gauges was the difference between the gauge internal and external pressures. Atmospheric pressure therefore had to be added to the gauge reading to convert i t to an absolute reading. Atmospheric pressure was read from a Precision Thermometer and Instrument Co. "Princo Fortin Type" barometer and was corrected.for temperature and gravity, using correction tablesr:supplied by the manufacturer. 2. Correction for the variation of the mercury level inside the equilibrium c e l l The gauge reading was a function of the mercury level in the equilibrium c e l l , and therefore a correction .had to be made for any variation of this level from i t s "zero position". If the mercury level was above the zero value of 2-k cm., the pressure corresponding to the extra height of mercury had to be subtracted from the pressure reading, and i f the level was below-the 2.k cm. level, the correction term;had to be added. 3- Correction for change of mercury density with temperature The line connecting the gauges to the equilibrium c e l l could be considered to be a mercury U-tube, and since bothtarms of the U-tube were at the same temperature when the gauge zero was defined, a correction term had to be applied i f the temperature, and thus mercury density, were not the same in both arms when a pressure reading was made. For the purpose" of calculating this correction, the arm on the c e l l side could be divided into three sections: one section 36 cm. long, located inside the bath and therefore at:the bath temperature; one section 11. h cm. long, enclosed by the bath insulation, and therefore at a temperature intermediate between that of the bath and the room (this intermediate temperature was assumed to be the average of the room and II-4 the hath temperature); and a f i n a l section at room temperature. Since the f i r s t two of these sections were always hotter than the arm connected to the gauge, the correction for the decrease in mercury, density was added to the gauge reading. The resultant correction was the sum of the three discussed above. Let P a ^ s = absolute pressure in equilibrium c e l l in mm. of mercuty P gauge = gauge reading in mm. of mercury P a -k m , = atmospheric pressure" in mm. of mercury 5o = density of mercury at 0° Centigrade 5 room = ^ ^ s ^ y °^ mercury at room temperature -^bath = ( ^ e n s ^ t y °^ mercury at the bath temperature 51 o v e r = density of mercury at the average value of avg. i the room and bath temperature a = height of mercury -level in mm. as measured on the measuring head rod scale t h e n P abs. = P gauge + P atm. + 0 5>room / 5*0 J>room/ Introduction of Materials Materials to be studied were introduced to the system through the one-litre solvent flasks shown in Figure 1 and Figure 1-5 • One flask was used for each material. Since a common procedure was followed with each substance, only.that used for methanol w i l l be described. A few grams of "Drierite" anhydrous calcium sulphate were added to ,flask F 10 and the flask was flushed with dry nitrogen. A quantity of methanol in.excess of that required was II-5 added, and the flask was connected.through the ungreased,mercury-sealed cone and socket joint to the line from transfer flask F 2. The solvent was stored in the flask for at least 2k hours in order that any moisture picked up during the pouring would be removed. (Sodium-lead alloy was used as the drying agent for benzene and toluene and the standing time was usually reduced to a few hours.) The methanol was transferred to flask F 2 by vacuum d i s t i l l a t i o n . Stopcocks S 3 and S 'U were opened, and the ai r and a small percentage of the methanol pumped out of F 10. When the air had been removed, stopcock S k was closed and a Dewar flask containing liquid nitrogen was placed around flask F 2. If a large amount of methanol were to be transferred, a heating mantle could be used to warm the methanol flask. After the desired amount of methanol had been transferred (enough methanol was always l e f t in F 10 to cover completely the drying agent), stopcock S 3 was closed and stopcock S k opened. The space above the frozen methanol was then evacuated u n t i l the pressure, as read at the vacuum manifold, was of the order of 10 mm. of mercury. When the evacuation was completed, the stopcock connecting the transfer flasks to the vacuum manifold was shut, and the liquid nitrogen.Dewar flask was placed around flask F 3- The solvent in F 10 was allowed to melt and then d i s t i l l e d through stopcocks S k and S 5 "to Flask F 3- The frozen solvent in F 3 w a s again evacuated to remove any air that might have been trapped in the solvent during the f i r s t evacuation and was then allowed to melt and vacuum d i s t i l l e d through stopcocks S 5 and S 2 to flask F 1. Once the methanol was present in F 1, toluene was d i s t i l l e d from F 5 to give a mixture of approximately the desired concentration and with a volume of between 100" and 150 cc. The frozen solvent mixture was evacuated through -k S 2 u n t i l the pressure, as measured at the manifold, was less than 10 mm. of mercury and stopcock S 2 was then shut. After the mixture had melted, i t was d i s t i l l e d through the ungreased stopcock S '13 to the intermediate storage flask T. II-6 This d i s t i l l a t i o n was accomplished by keeping flask F 1 at room temperature and f i l l i n g the centre tube of the intermediate storage flask with liquid nitrogen. When the d i s t i l l a t i o n was almost completed, stopcock S 13 was closed, stopcock S lh opened, and the last small percentage of the mixture rejected. Stopcock S 13 was then opened and any last traces of air removed from above the solvent mixture. The mixture was transferred'from the intermediate storage flask to the equilibrium c e l l using the mixture vapor pressure as the driving force. Stopcock S 13 was shut and the centre tube of the storage flask f i l l e d with hot water. When the mercury cut-off valve S 1 and valve V k were opened, the solvent mixture was forced, by the pressure of the vapor above the liquid, into the previously evacuated equilibrium c e l l . The design of the solvent transfer apparatus allowed neither the composition nor the weight of the solvent added to the c e l l to be known. If the equipment were to be used for specific volume measurements, the design would have to be modified. However, the design used was found effective for transferring an approximate amount of solvent in an air-free condition. Preparation for Measurements Once a solvent mixture was present in the equilibrium c e l l , valve V k was shut and valves V 6, V 7> a n d V 8 opened. (Valve V 5 was always open except when used for testing^purposes.) F u l l power was supplied to each of the three bath heaters u n t i l the bath temperature had reached the desired value and the 1000-watt heater was then turned off. At this point the off-on heater was switched off and the power supplied to the steady-on heater reduced u n t i l the bath temperature slowly f e l l . The off-on heater was then turned on and adjusted so that the temperature slowly rose again. The control c i r c u i t was then turned on. II-7 The pressure in the equilibrium c e l l was noted next, valve V 8 was shut, and the sampling tube assembly was carefully evacuated through valve V 13 and stopcock S 12 u n t i l the pressure as read on the McLeod gauge -1+ was less then 10 mm. of mercury. Valve V 13 was then closed, valves V 9 and V 10 opened, and the assembly f i l l e d with mercury from the pressure generator IJto a pressure of 100 p.s.i. greater than that in the c e l l . Valve V ' l was then opened and 5 cc. of mercury were pumped through the line between V, 1 and the c e l l , i n order to clear i t of solvent and leave the line f u l l of mercury. Valve V 1 was then closed again, the pressure in the assembly raised to i t s previous value, and V 2 opened. The line from valve V 2 to the c e l l was purged in the same manner as the f i r s t one and V '2 closed. After both lines had been purged, valve V 13 was opened to connect the assembly to the vacuum pump, and any solvent which had escaped from the c e l l was removed. At the conclusion of the pumping, valve V.13 was closed and the assembly f i l l e d with mercury to the same pressure as existed in the c e l l . Valve V 10 was then closed, valve V 3 opened, and the sampling tube, assembly l e f t f u l l of mercury and connected to the c e l l through V 3 u n t i l a sample was taken. The level of the mercury interface in the equilibrium c e l l was next adjusted, by the removal or addition of mercury through Valves V 8 and V 9> un t i l i t lay at least 1 cm. below the liq u i d sampling port, and at least 1 cm. above the lowest port. In practice, a level 1 cm. below the sample port was generally chosen. When this adjustment was completed, a check was made to see that the vapor-liquid interface lay between the liq u i d and the vapor sampling ports. If i t was found that this interface did not.lie between these two ports, then either the mercury level was readjusted or solvent was removed from the c e l l . II-8 When both i n t e r f a c e s were at the d e s i r e d l e v e l , the magnetic s t i r r e r was turned on, and a check was made to see t h a t the s t i r r e r was operating. The c e l l contents were then allowed at l e a s t one hour to come to phase e q u i l i b r i a . Frequent checks were made during t h i s time to see that the bath temperature d i d not d r i f t from the d e s i r e d value. Measurements and Taking of Samples A f t e r the contents of the e q u i l i b r i u m c e l l had been h e l d at a constant temperature f o r one hour, the pressure and temperature were determined, and samples were taken of both the l i q u i d and vapor phases. The r e s i s t a n c e of the platinum thermometer was measured f i r s t . The gauge pressure, e q u i l i b r i u m c e l l mercury l e v e l , atmosphere pressure, and room temperature were then recorded. When the above measurements had been made, a vapor phase sample was taken. In order to take t h i s sample, the sample c o l l e c t i o n tubes were f i r s t evacuated through the mercury c u t - o f f valve S 1 2 , and t h i s valve was then c l o s e d . Valve V 1 was opened, a l l o w i n g the mercury i n the sampling tube assembly t o be re p l a c e d by an equ i v a l e n t volume of vapor from the c e l l , and valves V.1 and V 3 were then both closed,. i s o l a t i n g the sample i n the sampling tube assembly. A Dewar f l a s k of l i q u i d n i t r o g e n was p l a c e d around c o l l e c t i o n f l a s k L and val v e V 13 w a s opened. When the sample had d i s t i l l e d from the sampling tube assembly and been c o l l e c t e d i n f l a s k L, the l i q u i d was removed and the sample allowed to run down i n t o the end of the c o l l e c t i o n tube. I t was then r e f r o z e n and the tube sealed o f f , while s t i l l under vacuum, and stored f o r l a t e r a n a l y s i s . The c o l l e c t i o n f l a s k s and sampling tube were re-evacuated before a l i q u i d sample was taken. Valve V 13 was then closed, and valve V 2 opened f o r a f r a c t i o n of a second,allowing a l i t t l e of the l i q u i d phase to b l e e d i n t o I I - 9 the sampling tube assembly. Once 'the liquid phase sample had been isolated in ,the assembly, i t was collected:in flask M, sealed off under vacuum, and stored for later analysis. Further Measurements When one set of equilibrium measurements had been completed, preparations were made for the next. These preparations f e l l into one of three categories. If noLfurther measurements were desired.from the material that was in the cell, or i f insufficient material remained, the bomb was evacuated through the sampling tube assembly and the bath was cooled. If, however, additional measurements were required at compositions near to that of the material in the cell, the•composition could be changed by bleeding off some of the vapor phase. The amount of change that could Ibe achieved by •this method was a function of the shape of the equilibrium.curve for'the materials present and the amount of material in the cell, but in the case of the toluene-benzene system,,changes of up to 15 percent' were possible. When equilibrium values were required at more than one temperature from material of the same composition, as was the case with the toluene-methanol system, the first set of readings were taken at the lowest temperature at which measurements were to be made, and further sets taken simply by raising the temperature. It was found to be possible, with this latter system, to introduce initially- enough material to the cell to take two sets of readings at each,25° interval from 125 to 250° C. Temperature Limits for the Use of the Equipment * The equipment was found to be satisfactory, when used with either the benzene-toluene or toluene-methanol system for continuous use at temperatures up to 200° C; however, i t could not be used, at least with the 11-10 toluene-methanol system, at temperatures higher than this for more than a few hours. When equilibrium measurements were made at 25° intervals up to 250° C and the equipment coaled immediately after the last reading, no problem was encountered. However, at the conclusion of thisJproject, an attempt was made .to obtain several measurements from one f i l l i n g at 250° C. After the bath had been at this temperature for a few hours and a sample of the vapor phase was taken, i t was found that the sample contained enough permanent gas (a gas which was not condensible in a liquid nitrogen trap) to prevent i t s collection by vacuum d i s t i l l a t i o n . Sampling of the liquid phase resulted in the same d i f f i c u l t y . Whether this gas was produced by the •decomposition of the solvent material being studied, or of the teflon seals in the equilibrium c e l l , i s not known. However, the fact that i t was produced, presented a definite upper limit for the temperature at which measurements could be made, and severely limited the number of readings which could be made from one f i l l i n g at temperatures just below this limit. APPENDIX I I I I I I - l CHROMATOGRAPHIC ANALYSIS OF PURIFIED METHANOL Methanol, which had been d i s t i l l e d and d r i e d as described e a r l i e r , was analysed f o r i m p u r i t i e s w i t h a Beckman ."GC2" gas chromatograph. The a l c o h o l was t e s t e d u s i n g a s i x - f o o t column of "8N8 F l e x o l P l a s t i c i z e r " £2-2 -(2-ethylhexanamido) - d i e t h y l d i 2-ethylhexoate_^] on f i r e - b r i c k at 70° C. No i m p u r i t i e s , w i t h the exception of one p o s s i b l e t r a c e , were found. This column, however, was not completely s a t i s f a c t o r y f o r determining methanol p u r i t y as i t would not separate methanol from water, a very l i k e l y impurity. A s i x - f o o t column of 2,2,methoxy, ethoxy-ethyl ether on f i r e - b r i c k o was then t e s t e d at 70 C^and i t was found t h a t t h i s column would separate methanol and water. When a sample of p u r i f i e d methanol was used, one impu r i t y , water, was found, but the water-methanol separation was not adequate to estimate the amount of water present. A s i x - f o o t , column of "Carbowax 1000" (polyethylene g l y c o l ) on t e f l o n was used f i n a l l y . This column gave e x c e l l e n t s e p a r a t i o n , at 70° C , of the components i n prepared mixtures of water, and methanol, and of water, methanol, and formaldehyde. Two t e s t s were made w i t h p u r i f i e d methanol u s i n g t h i s column, and i n each case the only detectable - i m p u r i t y was water; no t r a c e of formaldehyde was found. In t h e ' f i r s t t e s t the methanol a n d ' " D r i e r i t e " over which i t was afoored were separated, not too s u c c e s s f u l l y , by decanting, and a water content of about 0.15 mole percent was found. In the second t e s t the methanol and " D r i e r i t e " were separated by vacuum d i s t i l l a t i o n and a water content of 0.08 mole percent was found. Since methanol used i n the phase e q u i l i b r i u m studies was separated from the " D r i e r i t e " by vacuum d i s t i l l a t i o n , the water content i n the methanol used - was \'0 l e s s than 0.1 mole percent. APPENDIX IV IV-1 ANALYSIS OF BENZENE-TOLUENE AND METHANOL-TOLUENE MIXTURES Benzene-Toluene Mixtures Benzene-toluene mixtures obtained during the phase equilibrium studies were analysed using a Beckman "G02" gas chromatograph. The chromatograph was f i t t e d with a six-foot column of "8N8 Flexol Plasticizer" ^2-2 -(2-ethylhexan-amido)-diethyl di 2-ethylhexoate^ oh fire-brick, which had previously been shown to give excellent separation of these two substances. Mixtures of known composition were used to calibrate the chromatograph and column. These mixtures were prepared in four-ml. teflon stoppered glass weighing bottles. A clean dry stoppered bottle was f i r s t weighed on an F. Mettler "Gram-atic" electrical balance, then removed from the balance,,unstoppered, and partly f i l l e d with benzene from a hypodermic syringe. After the stopper was replaced in the bottle, the weight of the bottle and benzene was measured. Toluene was added to the bottle in a similar manner,'using a clean syringe, and the weight of the mixture was determined. At least three samples from each standard mixture were used to calibrate the chromatograph. The height of the benzene and of the toluene peaks were measured on the chromatographic traces obtained from each sample}and .the ratio of the peak heights was calculated. The averaged values of the results obtained in this manner for each sample were plotted on two graphs. For mixtures of more than 50 mole percent benzene, the log of the ratio of moles toluene to moles benzene was plotted versus the log of the ratio of the peak height toluene to peak height benzene, and for mixtures of less than 50 mole percent benzene, the log of the ratio of moles benzene to moles toluene was plotted versus the log of the ratio of peak height benzene to peak height toluene. It was found that the relationship between the peak height ratio and IV-2 the concentration ratio did not remain constant'but changed significantly over a period of a few weeks. For this reason, a check was made of the calibration with a freshly prepared standard solution before each group of samples was analysed, and i f a significant change-had occurred, a new calibration was prepared. The results from a typical calibration are given in Table IV-1 and •shown plotted in Figures IV-1 and IV-2. An estimate of the accuracy of the chromatographic analysis i s given in Table IV-2. The results of the • individual peak height measurements are given for two solutions, solutions lhA, for -which the variation in individual peak height measurements was larger than with any other standard solutioniased, and solution hA for which the variation in individual measurements was larger than with any other standard solution in the kO to 60 percent composition range. It can be seen from this table that for dilute solutions, no significant error was introduced by the analysis when the concentration was expressed to the nearest 0.1 mole percent, and that for concentrated solutions a maximum error of +0.1 mole percent was introduced by the analysis when the concentration was expressed to the nearest 0.1 percent. (The actual error was probably less than +0.1 mole percent for concentrated solutions, as at least two samples of every mixture to be analysed were used with the chromatograph and the results obtained averaged.) Toluene-methanol Mixtures Toluene-methanol mixtures obtained during the phase equilibrium studies were analysed using a Bausch and Lomb "Abbe-3L" refractometer. The temperature of the refractometer prisms was controlled with water circulated from a Colora Messtechneck Co. Ltd. "Ultra-thermostat". The range of temperature variation of the water in the thermostat was +0.02° C. The refractometer was calibrated with standard solutions prepared in TABLE IV-1 IV-3 CHROMATOGRAPH CALIBRATION FOR BENZENE-TOLUENE SYSTEM tfole Percent Log Log Log Log Benzene Benzene-toluene Toluene-benzene Benzene-toluene Toluene-benzene Mole Ratio Mole Ratio Peak Height Peak Height Ratio Ratio 2.7k -1.5502 -r..2644 4.16 -I.3629 -I.O836 10.33 -.9385 -.6694 21-33 -.5569 -•3110 3h.lk -.2853 -.0520 41.07 -.1567 •1567 •0637 - . O 6 3 6 56.55 .1145 -.1145 .3094 -•3094 57-80 .1366 -.1366 •3261 -•3261 58.51 .1492 -.1492 •3369 -•3369 66.73 -.3032 .4740 -.4740 68.40 -•3353 -•5042 68.47 -.3369 -.5060 75.01 -.4774 -•6315 82.7^ -.6807 -•8179 89.13 -.91U0 -1.0374 97-02 -1-5133 -I.6189 98.14 -1.7246 -I.8265 TABLE IV-2 I ACCURACY OF CALIBRATION OF CHROMATOGRAPH IV-4 Mixture Mole Height ' Height ?oluene -benzene Mole Percent No. Percent Benzene Toluene 3eak Height Ratio Benzene Benzene Peak Peak from Calibration Curve 14A i 97-02 221.5 in. 5-3^ in. .0241 97-02 i i 97-02 222-5 in. 5-37 in .0239 97-04 i i i 97.02 220.5 in. 5-27 in • 0239 97-04 kA i 57-80 134.4in. 63.2 in .470 57-9 i i 57-80 I36.6 in. 6k.5 in .472 57-8 i i i 57-80 I37.8 in. 65.O in .472 57-8 iv 57-80 139-2 in. 65.9 in •473 57-7 IV-5 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 LOG m n l t t a Toluana moles Benzene Figure IV-1 iv-6 -1.8 -1.6 -1.4 -1.2 -1.0 -0JB -0.6 -0.4 -0.2 0 0.2 LOG mglw Benzene moles Toluene Figure IV-2 IV-7 the same manner as was used for the benzene-toluene calibration described above. The refractometer prisms were cleaned before each set of readings with benzene and petroleum ether and between each reading with petroleum ether. A period of several minutes was allowed after the cleaning before the refractometer was used. A hypodermic syringe was used to introduce samples to the refract-ometer. The f i r s t part of the contents of the syringe was rejected and then just enough material added to the refractometer to f i l l the space between the two prisms. Two independent determinations were made for each mixture, and i f 'these determinations differed by more than one figure in the fourth decimal place, the readings were rejected. For some unknown reason the calibration of the rrefractometer did not remain constant, but changed by three or four units in the fourth decimal place periodically. For this reason, i t was checked with pure toluene and pure methanol before each set of readings was made, and i f a change had occurred in the refractive index of'these substances, the calibration was redetermined. The results of a typical calibration are given in Table IV-3 and shown plotted in Figure IV-3. The refractometer could be read to 0 . 0 0 0 1 units and the readings for a given substance were reproducible to +0 .0001 units. Composition of mixtures which were almost pure toluene could be determined, therefore, to ,+0.2 mole percent, and of mixtures which were almost pure methanol to +0 .05 mole percent. Two independent determinations of refractive index were made on each unknown sample to be analysed (except for a very few vapor samples where insufficient material was obtained for more than one determination). The values obtained were averaged and the average value used with refractive index calibration curves to obtain the sample concentration. TABLE IV-3 IV-8 TYPICAL REFRACTOMETER CALIBRATION FOR METHANOL-TOLUENE SYSTEM AT 20° C. Mole Percent Refractive Methanol Index 100 1-3272 95-79 1.3445 90.69 1.3627 87.52 1.3728 80.56 1-3923 74.46 1.4068 69.76 1.4168 60.86 1.4329 53-19 1.4446 42.02 1.4594 35-75 1.4663 31-81 1.4705 24.61 1.4772 24.49 l.4774 15-90 1-4847 6-73 1.4916 0 l .'4960 APPENDIX V V-1 PURIFICATION OF MERCURY Technical-grade mercury was purified for use in the experimental 78 apparatus using procedures recommended by Sanderson and the Handbook of 79 Chemistry and Physics . Surface•impurities were removed from the mercury by f i l t e r i n g , i t through a chamois leather sealed funnel, and the clean metal was then air-bubbled for. 2k hours. The surface of the mercury was kept covered with a frequently changed 1 percent solution of n i t r i c acid during the bubbling in order to aid in the removal of impurities such as dissolved a l k a l i metals, zinc, copper,.or lead. The mercury,was next washed three times in a 110-cm. high, 10 percent sodium hydroxide scrubber, three times in a similar 10 percent n i t r i c acid scrubber, and f i n a l l y three times in a d i s t i l l e d water scrubber. The mercury was introduced to each scrubber in the form of very fine droplets obtained by forcing i t through a section of chamois leather. Mercury from the f i n a l water wash was dried with clean blotting paper and then vacuum d i s t i l l e d to remove traces of the noble metals or t i n . TABLE 10 VI-1 VAPOR-LIQUID EQUILIBRIUM DATA FOR THE METHANOL-TOLUENE SYSTEM AT A PRESSURE OF ONE ATMOSPHERE (BENEDICT ET AL 5^ ) Temp. °C. Mole Methanol Liquid Vapor 100.61 0.0 0.0 70.25 13-0 74.2 66.44 26.6 78.2 65.58 40.7 8O.3 64.47 59-3 8I.9 64.10 69.2 82.9 63-79 77-9 84.5 63-67 84.3 86.9 63.58 88.2 88.3 63.62 92.7 91.1 63.95 96.9 95-0 64.51 100.0 100.0 TABLE 11 VI-2 VAPOR-LIQUID EQUILIBRIUM DATA FOR THE METHANOL-TOLUENE SYSTEM AT 125° C Pressure - mm. of mercury Composition - mole percent methanol Run Pressure Liquid Vapor Equilibrium Equilibrium Number Composition Composition Ratio Methanol Ratio Toluene - 5530 100 100 1.00 — 35-125 5550 99-1 98.8 •997 1-33 36-125 5575 97-h 96.9 •995 1-19 37-125 5585 95-3 94.9 •996 1.09 38-125 5585 93-3 1.00 .985 39-125 5575 90.2 91-5 1.01 .867 40-125 5520 83-6 88.6 1.06 .695 17-125 5510 82.9 88.6 1.07 .666 42-125 5420 74.7 86.2 1.15 .545 18-125A 5320 67.8 84.8 1-25 .472 18-125B 5320 67.8 84.7 1.25 .475 43-125 5105 53-7 82.4 1-53 .380 23-125A 5030 49.5 81.9 1.66 .358 23-125B 4040 41.7 80.2 ' 1.92 • 3^ 0 22-125 4440 29-0 11-1 2.68 • 31^ 27-125 3735 16.8 71.6 4.26 •3^ 1 41-125A" 3710 16.2 71.2 4.40 •343 41-125B 3040 9-9 63-7 6.43 .403 21-125B 2730 8.2 58.9 7.18 .448 41-125C 2270 •^9 50.7 10.4 .518 41-125D 1625 2.0 30.4 15-2 .710 1130 0 0 — _ 1.00 TABLE 12 VI-3 VAPOR-LIQUID EQUILIBRIUM .DATA FOR THE METHANOL-TOLUENE SYSTEM AT 150°.c. Pressure - mm. of Mercury Composition - mole fraction Methanol Run Pressure Liquid Vapor Equilibrium Equilibrium Number Composition Composition Ratio Methanol Ratio Toluene 10465 100 100 1.00 35-150 10520 99-1 99-0 •999 1.11 31-150 10480 99-1 99-0 -.999 1.11 25-150 10530 97-7 97-5 .998 1.09 36-150 10530 91-h 97-2 .998 1.08 37-150 10500 95-3 95-4 1.00 •979 24-150 10520 95-2 95-4 1.00 .958 38-I5O 10500 93-2 94.0 1.01 .882 16-150B 10400 90.5 92.4 1.02 .800 16-150A 10450 90.3 92.4 1.02 •783 39-150 10450 90.1 92.4 1.03 .768 40-150 10290 83-5 89-3 1.07 .648 17-150 10260 82.6 88.5 1.07 .661 42-150 10090 74.7 87.1 1-17 .510 18-150 9770 66 .8 84.9 1.27 .455 28-150 9400 55-6 82.7 1.49 •390 43-150 9260 52.6 82.1 I.56 •378 23-150A 9080 48.9 80.1 1.64 •389 23-150B 8640 4 i . o 79-3 1-93 •357 23-150C 8590 40.5 79-3 1.96 • 3^8 22-150 766O 28.1 75-1 2.67 • 346 27-150 6155 16.5 67-6 4.10 • 388 41-150A 6035 15-3 66.6 4-35 •394 41-150B 4845 9-1 57-7 6.29 .465 21-150 7-7 52:4 6.81 .516 41-150C 3595 k.l 42.2 8.98 .607 41-150D 2720 1.8 23'.6 13.I •778 34-150B 2450 1.0 15.1 15-1 .858 2065 0 0 — — 1.00 TABLE 13 Vl-4 VAPOR-LIQUID EQUILIBRIUM DATA FOR THE METHANOL-TOLUENE SYSTEM AT 175° C Pressure --•mm.'of Mercury Composition - mole"percent-Methanol Run Pressure Liquid-' Vapor Equilibrium Equilibrium Number Composition Composition Ratio Ratio Methanol Toluene I83OO 100 100 1.00 35-175 I836O 99-1 99-0 •999 1.11 25-175 183OO 97-6 97-7 1.00 .958 • 24-175 18280 95-2 95.8 1.01 .875 16-175B 18080 90.3 92.7 1.03 •753 16.175A 18080 90.3 92.7 . 1.03 •753 40-175 17720 83-3 89.6 1.08 .623 17-175 17660 82.2 89.2 1.09 .607 42-175 17170 74.5 86.9 1.17 . 514 18-175 16470 64.9 84.2 1-30 .450 28-175 15610 55-0 .81.8 1.49 .404 3^-175 15120 49-1 80.3 1.64 •387 23-175 13970 39-2 77-1 1.97 •376 22-175 .I2O5O 27-5 72.0 2.62 .386 27-175 9250 15.0 62.0 4.13 .447 21-175 6495 7-4 44.3 5-99 .602 41-175 ' 4290 1-7 17.2 10.1 • 843 34-175 3985 •1.0 H-3 11.3 .896 3525 0 0 — 1.00 TABLE 14 VI-5 VAPOR-LIQUID EQUILIBRIUM .DATA FOR THE METHANOL-TOLUENE SYSTEM AT 200° C. Pressure- - -mm. of mercury Composition - mole percent Methanol Run Pressure Liquid Vapor Equilibrium Equilibrium. Number Composition Composition Ratio Methanol Ratio Toluene 30220 100 100 1.00 — 35-200 30250 99-1 99-2 1.00 .889 25-200 30140 97-6 97-9 1.00 .875 24-200 30010 95-1 96.0 1.01 .816 16-200A '29550 90.0 92.8 1.03 • 720 16-200B 29530 90.0 92.8 1.03 ".720 17-200 28460 81.4 88.7 1.09 .608 42-200 27480 74.2 86.2 1.16 •535 18-200 25610 61.7 82.1 1-33 .467 28-200 24280 54.0 79-8 1.48 .439 43-200 22770 46.7 77-3 1.66 .426 23-200 20730 37-9 73-8 1.95 .422 22-200 17440 26.3 . 66.6 2-53 •453 27-200 13150 15.1 55-4 3-67 •525 21-200 9350 6.8 37-7 5-54 .668 41-200 6515 •1-3 12.6 9-69 .885 34-200 6195 1.0 8.0 8.00 •929 5650 0 0 1.00 TABLE 15 VI-6 VAPOR-LIQUID EQUILIBRIUM RATIO FOR,THE METHANOL-TOLUENE SYSTEM'AT 225° C. Pressure r- mm. of Mercury Composition - .mole percent Methanol Run Pressure Liquid Vapor Equilibrium Equilibrium Number Composition Composition Ratio Ratio Methanol Toluene 47540 100 100 1.00 _ _ 35-225 47390 99-1 99-2 1.00 .889 25-225 47290 97-5 97-8 1.00 .880 24-225 4683O 95-1 95-9 1.01 .837 16-225A 4567O 89.5 92.1 1.03 •752 16-225B 4567O 89.2 91.9 1.02 .750 17-225 43290 80.0 86.9 1.09 •655 42-225A 41450 73-8 84.2 1.14 .603 42-225C 38390 63-3 ,80.1 1.27 • 542 18-225 3638O 56.6 77-6 1-37 .516 28-225 35150 52.9 76.2 1.44 .505 43-225 32320 45.2 73-0 1.62 .492 23-225 28790 36.3 68.7 I.89 .491 22-225 23810 25-7 60.7 2.36 •529 27-225 17730 14.6 48.2 3-30 .606 21-225 12980 6.2 30.4 4.90 • 7^ 2 41-225 9570 1-3 9-1 7.00 • 921 34-225 9260 .8 5.8 7-25 •950 8600 0 0 — 1.00 TABLE 16 VI-7 VAPOR-LIQUID EQUILIBRIUM RATIO FOR THE METHANOL-TOLUENE SYSTEM AT 250° rj. Pressure - mm. of mercury Composition - mole-percent methanol Run Pressure Liquid Vapor Equilibrium Equilibrium Number Composition Composition Ratio Ratio Methanol Toluene 42-250B 57680 71.8 •75-7 1.05 :862 42-250D 55^ 50 67-0 74.8 1.12 .764 44-250B 54550 65-3 74.2 1.14 .744 44-250C 52980 62.1 73-2 1.18 .707 28-250A 4765O 52.1 69.7 1.34 •633 43-250 43040 44.0 66.1 1.50 .605 28-250B 41840 42.1 65.1 1.55 .603 23-250 37820 35-4 61.5 -1.74 .596 22-250 31200 24.9 55-6 2.23 .591 " 27-250 23370 14.2 40.6 2.86 .692 21-250 175TO 6.0 24-2 4.03 .806 41-250 13700 1.1 C6 1.1 5-55 .949 34-250 13310 .6 4.2 7.00 .964 30-250 12600 0 0 — — 1.000 VI-8 TABLE 18 ACTIVITY COEFFICIENTS FOR METHANOL AT 125° C Columns are numbered to correspond with method of calculation discussed in test. Mole percent Methanol in Calculation Method Liquid 1 2A 2B 3 4 5 ' 100 1.000 1.000 1.000 1.000 1.000 1.000 99-1 1.001 1.001 1.000 1.000 1.000 1.000 97-4 1.003 1.003 1.002 1.000 1.002 1-003 95-3 1.006 1.005 1.004 1.002 1.005 1.006 93-3 1.011 1.010 1.009 1.008 1.010 1.012 90.2 1.023 1.022 1.021 1.020 1.022 1.025 83.6 1.058 1.059 1.058 1-053 1.059 1-063 82.9 1.065 1.066 1.060 1.060 1.066 1.071 74.7 1-13 1U3. 1.13' 1-13- 1-13 1.14 67.8 1.20 1.21 1.21 H19 1.21 1.22 67.8 1.20 1.21 1.21 1.19 1.21 1.22 55-9 I.38 1-39 1-39 1-37 1.39 1.40 53-7 1.42 1-43 1-43 l . 4 l 1-43 1.44 49-5 1.51 1-52 1-53 1.50 1-52 1.54 41.7 1.68 1.70 1.72 1.69 1.70 1-73 29.0 2.15 2.19 2.22 2.16 2.19 • 2.24 16.8 2.88 2-97 3-02 2.91 2.96 3-06 16.2 2-95 3-04 3.10 2.98 3.03 3-14 9-9 3-54 3-63 3.78 3.61 3-67 3-84 8.2 3-55 3.71 3-82 3-63 3.70 3-88 4-9 4.25 4.47 4.66 4^35 4.46 4.72 2.0 4.47 4.76 4.96 4.63 4.74 5.07 TABLE 19 Vl-9 ACTIVITY COEFFICIENTS FOR TOLUENE AT 125° C Columns are numbered to correspond with methods of calculation discussed in text. Ylole'percent 1 Methanol in Liquid Calculation Method 1 2A 2B 3 4 5 99-1 6-55 5.38 5.26 6.24 5-67 , 5-99 97-4 5-88 4.83 4.72 5.58 5.08 5-36 95-3 ,5.36 4.40 4.30 5.08 4.62 4.87 93-3 4.87 4.00 3.90 4.61 4.19 4.40 90v2 4.28 3-52 3-^ 3 4.05 3-68 3-86 83.6 3-40 2.79 2-73 3-20 2.92 3-05 82.9 3-25 2.68 2.62 3.06 2.80 2.92 74.7 . 2.62 2.16 2.12 2.46 2.25 2.3^ 67.8 2.22 1.85 1.81 2.09 1.92 1.99 67.8 2.24 1.86 1.82 2.11 1-93 2.01 55-9 1-77 1.49 -1.46 1.67 1-5^ 1.59 53-7 1.72 1.44 1.41 1.62 1.49 1.54 49-5 1.60 1-34 1.32 1.51 1.39 1.44 41-7 1.46 1.24 1.21 1.38 1.28 1.31 29.0 1.23 1.066 1.049 1.18 1.10 1-13 16.8 1-13 1.006 •99^ 1.086 l . 028 1.047 16.2 1.13 1.008 •995 1.089 1.029 -1.047 9-9 1.08 ?997 .988 1.062 1.011 1.024 8.2 1.08 1.007 1.001 1.056 1.020 1.030 4-9 -1.04 .991 .986 1.009 •999 I.005 2.0 1.023 1.000 •999 1.012 1.002 1.005 0 1.000 1.000 1.000 1.000 1.000 1.000 TABLE 20 VI-10 ACTIVITY COEFFICIENTS FOR METHANOL AT 150° C Columns are numbered to correspond with methods of calculation described in text. Mole Percent Methanol in Liquid Calculation Method 1 2A 2B 3 4 5 100 1.000 1.000 1.000 1.000 1.000 099H 1.004 1.004 1.003 1.000 1.004 1.004 99-1 .999 •999 •999 -1. .999 1.000 1.001 97-7 •998 .998 •996 •996 1.003 1.004 97-4 1-003 1-003 1.002 •999 1-003 1.004 95-3 1.001 1.002 1.001 •998 1.004 1.007 95-2 1.002 1.003 1.002 •999 1.007 1.009' 93-2 1.009 1.010 .1009 1.006 1.012 -1.015 90.5 1.011 1.014 1.014 1.004 1.016 •1.021 90.3 1.018 1.020 1.020 1.010 1.022 1.027 90.1 1.021 1.022 1.022 1.013 1.025 1.029 83.5 1-049 •1.053 1.054 1.038 1.051 1.063 82.6 1.046 1.050 1.052 1.039 1.049 1.063 74.7 1.12 1-13 1-13 1.11 1.13 1.14 ,--66.8 1.18 1-19 1.20 1.17 1.20 1.22 55-6 1-33 1-35 1.36 1.32 1.36 1-38 52.6 1.38 1.40 1.41 1-37 i.4i 1-43 48.9 1.42 1-45 1.46 l . 4 l 1.45 -1.48 4i.o 1-59 1.64 1.66 l.6l 1.64 1.68 40.5 1.60 1.65 1.67 1.61 1.65 1.70 28.1 •1-95 2.03 2.07 1.97 2.03 2.11 16.5 2.40 2-55 2.63 2.45 2-55 2.69 15-3 2.50 2.67 2-75 2.58 2.66 2.8l 9-1 2-93 3.17 3-29 3-04 3-16 3-38 7-7 2.82 3.07 3-20 2.95 3-07 3-30 4.7 3-07 3-39 3-55 3-25 3.38 3-67 1.8 3-40 3-79 3-99 3.61 3-78 4.15 1.0 3-52 3-95 4.16 3-84 3-94 4-33 TABLE 21 VI-11 ACTIVITY COEFFICIENTS FOR TOLUENE AT 150° C Columns are numbered to correspond with methods of calculation described in text. Mole Percent Methanol, in Calculation Method Liquid 1 , 2A 2B 3 4 5 99-1 5.68 4.03 3-95 5-20 4-43 4.81 99-1 5.65 4.06 3-92 5.18 4.42 4.79 97-7 5-53 3-97 .3-84 5-06 4.32 4.69 97-4 5.50 3-90 3-83 5.02 4.28 4 .64 95.3 U.98 3-57 3-45 4.54 3-87 4.18 95-2 -U.87 3-50 3-38 4.45 3.80 4.10 93.2 4.49 3-22 3-11 4.09 3-49 3.76 ' 90.5 4.03 2.91 2.82 3-66 3-13 3-37 90.3 3-97 2.86 2.76 3-60 3-08 3.31 90.1 3-89 2.79 2.71 3-53 3-02 3-24 83.5 3-24 2.34 2.27 2.92 2-59 2.69 82.6 3.28 • • 2.38 2.31 2-97 . 2.67 2-73 7U.7 2.49 1.82 1.77 2.25 1.94 2.07 66.8 2.15 1-59 1-55 1.95 I.69 1.80 55-6 1.77 1-33 •1.30 1.61 l . 4 i 1.49 52.6 1.69 1.28 1-25 1.54 1-35 1.42 1+8.9 1.72 1.30 1.27 1.56 1.37 1.44 4i . o 1.47 1.13 l . l l 1-35 1.19 1-25 1+0.5 1-45 1.12 1.099 1-33 1.18 1.23 28.1 1.29 I.030 1.016 1.19 1.076 1.12 16.5 1.16 .985 •977 1.094 1.015 1.041 15-3 1-15 .987 • 978 1.095 1.015 1.040 9-1 1.002 • • ; 98 i •975 1.050 •998 1.014 7.7 1.085 •993 .989 1.054 1.008 1.021 4-7 1.056 •994 •993 1-035 1.004 1.013 1.8 1.025 1.000 •998 1.015 1.002 1.006 1.0 1.017 1.002 1.002 1-025 1.004 1.006 0 1.000 1.000 1.000 1.000 1.000 1.000 TABLE 22 VI-12 ACTIVITY COEFFICIENTS FOR METHANOL AT 175° C. Columns are numbered to correspond with methods of calculation described in text. Mole Percent Calculation Method Methanol in 1 Liquid 1 2B 3 4 5 100 1.000 1.000 1.000 1.000 1.000 99-1 1.002 1.001 •997 •999 1.002 97-6 1.001 1.001 •999 •999 1-003 95-2 1.006 1.006 •1.002 1.003 llOlO 90.3 1.015 1.019 1.006 1.015 1.025 90.3 1.015 1.019 1.006 1.015 1.025 83-3 1.040 1.052 1.027 • 1.047 I.O63 82.2 1.047 •1.059 -1.032 1.054 1.070 74.5 1.093 1.12 1.080 1.108 1-13 64.9 1.17 1.21 1.18 1.19 1.22 55.O 1.28 1-34 1.28 1-32 1-36 49.I 1-35 1.43 1-35 1.40 1.46 39-2 1.50 1.62 1.51 I.58 1.65 27-5 1.72 1.92 1-77 1.86 1-97 15.0 2.09 2.44 2.19 2-33 2.51 7-4 2.13 2-59 2.28 2.44 2.70 1-7 2-37 2-99 2.61 2.81 3-15 1.0 2.46 3-12 2.70 2.92 3-29 TABLE 23 Vl-13 ACTIVITY COEFFICIENTS FOR TOLUENE AT 175° C Columns are numbered to correspond with methods of calculation described in text. Mole Percent Methanol in Calculation Method Liquid 1 2B 3 4 5 99-1 5.78 3.23 5-04 •3-91 4.42 97-6 .4:98 2-79 •^33 3-56 3.78 95-2 4.54 2.56 3-9U 3.05 3-^ 3 90.3 3.86 2.19 3-33 2-59 2.90 90.3 3.86 2.19 3-33 2.59 2.90 83-3 • 3.13 1.80 2.68 2.10 2.34 82.2 3.04 1-75 2.60 2.05 2.28 74.5 2.50 1.48 2.14 1.70 1.88 64.9 2.10 1.28 1.82 1.45 1.59 55-0 1.80 1-13 1.56 1.27 1.38 49.I 1.66 1.069 1-^ 5 1.19 1.29 -39-2 1.49 1.009 1-31 l . l l . 1.18 27.5 1.32 •965 1.20 1.033 ,1.088 15-0 1.17 .956 1.098 •99^ 1.027 7-^ 1.11 •999 1.070 1.015 1.032 1-7 • 1.025 .998 1.016 1.001 1.006 1.0 1.013 •997 1.002 •999 1.002 0 1.000 1.000 1.000 1.000 1.000 TABLE 24 VI-14 NATURAL LOGARITHM OF THE,RATIO -OF THE' ACTIVITY COEFFICIENTS FOR METHANOL AND TOLUENE' AT'125° C Columns are numbered-to correspond with methods-of calculation described in text. Mole Percent • Methanol in Calculation Method Liquid l 2A 2B 3 4 5 99-1 -1.879 -1.682 -II66O -I.828 -1-735 -1.790 97-k -1.769 -1-573 -1.550 -1.719 -1.623 -I.676 95-3 -1.674 -1.479 -1.455 -1.623 -1.526 -l."576 93-3 -1-572 -1-375 -1.352 -1.519 -1.423 -1.470 90.2 -1.431 -1.236 -1.212 , -I.378 -1.281 -1-325 83-6 -1.166 -.970 -.948 -1.112 -1.015 -1.053 .82.9 -1.116 -.921 -.898 -1.06l -.965 -1.QO3 74.7 -.839 -.647 -.623 -.784 --.687 -.720 67.8 -.614 -.425 -.401 -.56I' -.463 -.492 67.8 -.621 -•433 -.408 -•569 -.471 -499 55-9 -.254 -.069 < • -.045 -.200 -.106 -.129 •53-7 -•193 -.010 .016 -.139 -.046 -.067 L9.5 -.055 .123 .148 -.007 .087 . .068 - 41.7 .146 • 322 •348 .200 .288 - 275' 29.0 .556 • 720 .748 .604 .690 .687 16.8 •937 ,1.081 1.112 .988 1.057 ,1.072 16.2 .961 1.104 1-135 1.008 ,1.081 1-097 9-9 -I.183 .1.307 1-342 1.223 1.290 I.322 8.2 1.187 - 1-304 1-340 1-235 1.288 1.328 k.9 1.451 1.508 •1.548 1.243 1.497 1-547 2.0 1.476 1.561 1.603 ..1.520 1-5,54 1.619 TABLE 25 vi-15 NATURAL LOGARITHM; OF THE RATIO OF THE ACTIVITY COEFFICIENTS FOR METHANOL AND TOLUENE AT 150° C. Columns are numbered to correspond with methods of.calculation described in text. Mole Percent Methanol in Calculation Method Liquid l 2A 2B 3 4 5 99-1 -1-733 -1-399 -1-370 -1.649 -1.484 -1.566 99-1 -1-733 -1.401 -I.367 -1.649 -1.485 -1.566 97-7 " -1.712 -1.381 -1.346 -1.626 -1.461 -1.540 97-4 -I.702 -1.360 -1-340 -1.615 -1.451 -1-530 95-3 -1.604 -1.271 -1.234 -1.516 -1.350 -1.425 95-2 -I.582 -1.249 -1.216 -1.494 -1.328 -1.403 93-2 -1.493 -1.160 -1.127 -1.403 • -1.237 -1-309 90.5 -1.383 -1.054 -1.-22 -1.294 -1.126 -1.194 90.3 -1-359 -1.029 --997 -1.270 -1.102 -1.170 90.1 -1-337 -1.006 -•975 -1.245 -1.079 -1.148 83.5 -1.126 -•779 -.768 -1-036 -.903 -•927 82.6 -1.143 -.818 -.786 -1.050 -.933 -.942 74.7 -•799 -.476 -.446 -.708 -.541 -.594 66.8 -.599 -.281+ -.254 -.508 -•344 >-389 55-6 -.287 .018 .050 -.196 -.038 -•073 52.6 -.207 •093 .126 -.118 •039 .008 48.9 • -.190 .106 •139 -.099 .056 • 030 41.0 .080 .367 •399 •171 •317 •299 1+0.5 .102 •359 .419 .191 •337 •319 28.1 .1+17 .680 .712 .502 .636 •637 I6.5 •731 •953 •992 .810 •923 • 950 15-3 •775 •994 1.033 •855 .965 •995 9.1 .986 •1.174 1.218 1.062 1.154 1.205 ' 7-7 •954 1.129 1-175 1.029 1.113 1-172 1 4.7 1.069 1.225 1.274 1.143 1.215 I.287 1.8 1.198 1-333 1-385 I.269 1.329 1.142 1.0 1.242 1-371 1.424 1-320 1.368 1.460 TABLE 26 VI-16 NATURAL LOGARITHM OF THE RATIO•OF THE ACTIVITY COEFFICIENTS FOR METHANOL AND TOLUENE AT 175° C Columns are numbered to correspond with methods of calculation described in text. Mole Percent Methanol in Calculation Method Liquid 1 • 2B 3 4 5 99-1 -1-753 -1-173 -1.620 -1-364 -1.483 97-6 -1-603 -1.024 -1.467 -1.212 -1-327 95-2 -1-507 -.928 -1.369 -1-111 -1.222 90-3 -1-337 -.763 -I.196 -•935 -1-038 90.3 -1-337 -.763 -1.196 --935 -1.038 83-3 -1.000 -•538 -•959 -.698 -.790 82.2 -1.066 -.504 -•295 -.663 -•754 7U.5 -.827 -.279 -.686 -.428 -.509 -64.9 -•589 -.060 -.436 -.198 -.264 55-0 -•345 .165 -.197 .035 -.017 49.1 -.202 .289 -.069 .165 .122 39-2 .006 .472 1138 •358 •335 27-5 .267 .688 •••395 •587 •592 15.O •577 .936 •693 .850 .895 7-4 .651 •953 •758 •879 .961 1-7 •839 1.098 -942 1.030 1.142 1.0 .888 1.141 -990 1.074 1.190 TABLE 27 Vl-17 NATURAL LOGARITHM, OF THE RATIO OF THE ACTIVITY COEFFICIENTS FOR METHANOL AND TOLUENE CALCULATED BY THE METHOD OF BLACK FOR TEMPERATURES OF 200° AND 225° C 200° C 225° Mole Percent Logarithm Mole Percent Logarithm Methanol in Ratio Methanol in Ratio Liquid Liquid 99-1 -I.I36 99.1 -.963 97-6 -1.111 97-5 -.941 95-1 -1.024 95-1 -873 90.0 -.861 89-5 -•725 90.0 -.861 89.2 -.719 '8l.4 -.619 80.0 -.511 74.2 -.421 73-8 -.376 61.7 — 145 63.3 -.165 54.0 .022 •56.6 -.038 46.7 .164 52.9 .029 37-9 .330 45-2 •159 26.3 •512 36.3 .306 15.1 •715 25.7 .432 6.8 .868 14.6 .482 1-3 1.134 6.2 •765 1.0 .894 1-3 .888 0.8 •893 TABLE 30 Vl-18 COMPARISON OF ACTIVITY COEFFICIENTS CALCULATED FROM EQUATION OF BLACK AND FROM REDLICH AND KISTER CORRELATION FOR METHANOL, AND TOLUENE--AT 125° C Methanol Toluene Mole Percent Methanol in Black's Correlation ^Difference Black's Correlation ^Difference Liquid Equation Equation 99-1 1.000 1.000 0 5-99 6.00 .2 97-4 1.003 1.002 .1 5.36 5.49 2.4 95-3 1.006 1.006 •"0 4.87 4.94 1.4 93-3 1.012 1.012 0 4.40 4.50 2.2 90.2 1.025 1.024 .1 3-86 3-93 1.8 83.6 I.O63 I.O63 0 3-05 3-02 1.0 82.9 1.071 1.068 •3 2.92 •3.00 '2-7 74.7 ,1.14 1.14 0 2-35 2-35 .0 67.8 1.22 1.22 0 1-99 -2.00 • 5 67.8 1.22 -1.22 0 2.01 2.00 .5 55-9 1.40 1.39 •7 , 1-59 1.60 .6 53-7 , 1.44 1.42 •7 1.54 1.55 .6 49.5 1-54 1.52 1-3 1.44 - 1.46 :ilk 41.7 1-73 1-71 112: I.32 1-32 0 29-0 2.24 2.17 3-1 1-13 1.16 2.6 16.8 3-06 2.96 3-2 1.047 1-055 .8 • 16.2 3-14 3-01 4.1 1.047 1.051 .4 9-9 3-84 3-69 3.9 I.023 1.020 •3 8.2 3.89 3-92 .8 1.030 1.014 1.6 4.9 4.72 4.44 6.0 1.005 1.005 0 2.0 5.07 4.99 1.6 I.005 1.001 .4 TABLE 31 vi-19 COMPARISON OF ACTIVITY COEFFICIENTS CALCULATED FROM EQUATION OF BLACK AND FROM REDLICH AND KISTER CORRELATION FOR METHANOL AND TOLUENE AT 150° C Mole Percent Methanol Toluene Methanol in Black's Black's Liquid Equation;- "Correlation ^Difference Equation Correlation ^Difference IOO 99-1 1.004 1.000 .4 4.81 5.O8 5-3 99-1 1.001 1.000 .1 4-79 5.O8 5-7 97-7 1.004 1.001 •3 4.69 4.76 1-5 97-4 1.004 1.002 .2 4.64 4.69 1.1 95-3 1.007 1.005 .2 4.l8 4.27 2.2 95-2 1.009 1.005 .4 4.10 4.25 3-5 93-2 1.015 1.011 .4 3-76 3-91 4.0 90.5 1.021 1.020 .1 3-37 3-52 4.4 90.3 1.027 1.021 .6 3-31 3-k9 5-U 90.1 1.029 1.022 •7 3-24 3-46 ' 6.8 83-5 I.O63 I.O57 .6 2.69 2.76 2-5 " 82.6 , I.O63 I.O63 0 2-73 2.69 1-5 7U.7 1.14 1-13 •9 2.07 2.17 4.8 66.8 1.22 1.21 .8 1.80 1.82 l . l 55-6 1.38 1.36 l.U 1.49 1-51 1-3 52.6 1-43 1.41 i:k\ 1-43 I.45 1.4 48. 9 1.48 1.48 0 1.44 I.38 4.2" 41.0 1.68 1.65 1:8 1.25 1.26 .8 40.5 1.70 1.66 2.4 1.23 1.26 2.4 28.1 2.11 2.05 2.8 1.12 1-13 •9 16.5 2.69 2.65 1-5 1.041 1.047 .6 15-3 2.81 2.74 2-5 1.040 1.042 .2 9-1 3-38 3-29 2-7 1.014 1.016 .2 7-7 3-30 3-kk 4.2 1.021 1.011 1.0 k.l 3-67 3-82 4.0 1.013 1.004 •9 1.8 4.15 4.26 2.6 1.006 1.001 •5 1.0 4-33 4.40 1.6 1.006 1.000 .6 TABLE 32 VI-20 COMPARISON OF ACTIVITY COEFFICIENTS CALCULATED FROM EQUATION OF BLACK AND FROM REDLICH AND KISTER CORRELATION FOR METHANOL AND TOLUENE AT 175° C Mole Percent Methanol Toluene . Methanol in Black's Black's Liquid Equation Correlation ^Difference Equation Correlation ^Difference 99-1- 1.002 1.000 .2 4.42 4.39 •7 97-6 1.003 1.001 .2 3.78 3-78 8.2 95-2 1.010 1.005 •5 3-43 3-67 6-5 90.3 1.025 1.021 .4 2.90 3-03 4.4 90.3 1.025 1.021 .4 2.90 3-03 4.4 83-3 .1.063 1.055 .8 2.34 2.42 3-4 82.2 1.070 1.062 .8 2.28 2.35 3-1 74.5 1-13 .1.12 •9 1.88 1.95 3-6 64.9 1.22 1.21 .8 1-59 1.64 3-0 55-0 1.36 1-32 2.9 I.38 1.42 2-9 49.1 1.46 1.41 3-4 1.29 1-33 3-1 39-2 1.65 1-59 3-6 1.18 1.21 2-5 27-5 1-97 1-91 330 1.088 1.102 1.4 15.0 2.51 2.45 2.4 1.027 1.031 .4 7-4 2.70 2-95 9-2 1.032 1.008 2.4 1-7 3-15 3-44 8.4 1.006 1.000 .6 1.0 3-29 3-51 6.6 1.001 1.000 .1 TABLE 33 VI-21 COMPARISON OF ACTIVITY COEFFICIENTS CALCULATED FROM EQUATION OF BLACK AND FROM REDLICH AND KISTER CORRELATION FOR METHANOL, AND TOLUENE AT 200° C. Mole Percent Methanol Toluene Methanol in Liquid Black's Equation Correlation ^Difference Black's Equation Correlation ^.Difference 99-1 97-6 95-1 90.0 90.0 81.4 74.2 61.7 54.0 46.7 37-9 26.3 15-1 6.8 1-3 .1.0 0 1.003 1.004 1.011 1.029 1.028 1.070 1.12 " 1.24 1-33 1.43 1.58 I.83 2.13 2.44 3-12 2.46 1.000 1.001 1.005 1.020 • 1.020 1.06l 1.11 1.21 1.28 1-37 1.50 1.76 ,2.21 2.80 3-42 3-U6 •3 •3 .6 •9 •9 .8 .8 2.4 3.8 4.2 5-1 3.8 3-8 14': 7 9-6 40.7 3.12 3.O5 2.82 2.43 2.43 1.99 1.71 1.43 1.30 1.22 1.14 1.094 1.043 1.022 1-003 1.007 3-71 3-48 3-14 2.62 2.62 2.06 1-77 1.47 1-35'. , 1.27 ' 1.19 1.101 1-038 1.009 1.000 1.000 I8.9 14.0 11-3 7.8 7.8 3-5 3-5 2.8 3-8 4.1 4.4 .6 •5 1-3 •3 •7 TABLE 3!+ VI-22 COMPARISON OF-ACTIVITY COEFFICIENTS CALCULATED FROM EQUATION OF BLACK AND FROM REDLICH AND KISTER CORRELATION FOR METHANOL, AND. TOLUENE AT,225° C Mole Percent Methanol Toluene Methanol in Blackss 'Black's Liquid Equation Correlation ^Difference Equation Correlation ^Difference 99-1 1.001 1.000 .1 2.62 3-11 I8.7 97-5 1.005 1.001 .4 ,2.57 •2.95 14.7 95-1 1.010 1.004 .6 2.42 2.73 12-9 89.5 1.028 1.017 • l . l 2.12 • 2.33 970 89.2 1.030 1.018 l . l 2.11 2.31 9-4 80.0 1.071 I.O58 1.2 1-79 I.87 4'. 5 •73.8 .1.11 1.094 1.4 •1.6l • 1.67 3-7 63-3 • 1.19 1.17 1.7 i.4o 1.45 3-5 56.6 1.26 •1.22 '3-1 1.31 1-35 3-1 52.9 1.30 1.26 3-1 1.26 1.30 3-1 45.2 1.39 1.34 3-6 •1.19 1-23 3-4 36.3 •1.52 • 1.46 4.0 1.12 1.16 3-6 25.7 1.68 1.67 .6 1.093 I.090 - ~^3 14.6 1.90 2.06 8.4 :1.048 1-035 1-3 6.2 2.21 2.6l 18.1 I.028 •1.007 2.0 1.3 2.45 3-12 27-3 1.006 •1.000 .6 0.8 2.46 3-19 29.6 I.009 1.000 •9 APPENDIX VII VII-1 PROCEDURES USED TO PURIFY BENZENE, TOLUENE, AND METHANOL FOR VAPOR PRESSURE MEASUREMENTS Benzene 57 Bender, Furukawa, and HjHLcbman Three samples of benzene were used. The f i r s t , reagent-grade benzene, was treated with sulfuric acid u n t i l i t gave a negative test for thiophene, then washed with water, and next dried over calcium chloride and sodium wire. It was f i n a l l y d i s t i l l e d over fresh sodium wire. The second sample, P h i l l i p s Petroleum Company research-grade benzene, was c e r t i f i e d to be 99-93+.03 percent benzene. The third sample, thiophene-free, reagent-grade benzene, was dried over sodium wire and then d i s t i l l e d . The three samples gave consistent results. 58 Gornowski, Amick, and Hixson Baker's CP. thiophene-free'benzene was agitated with concentrated sulfuric acid and then washed twice with d i s t i l l e d water. The sample was next treated twice with 0.1 normal sodium hydroxide and again washed twice with d i s t i l l e d water. The benzene was f i n a l l y twice purified by fractional crystallization. Y o u n g 5 9 Commercial benzene was f i r s t d i s t i l l e d , then twice fractionally crystallized, and next shaken repeatedly with concentrated sulfuric acid. The f i n a l purification was by fractional d i s t i l l a t i o n . \ 3 Griswold, Andres, and Klein Thiophene-free benzene was purified by fractional d i s t i l l a t i o n . VII-2 2 v. Huhn The best grade of benzene commercially available at the time of the measurements was dried by d i s t i l l a t i o n and then stored over sodium wire u n t i l used. Toluene 62 Krase and Goodman Toluene of an unstated grade was once d i s t i l l e d . 2 v. Huhn The best grade of toluene commercially available at the time of the measurements was dried by d i s t i l l a t i o n and then stored over sodium wire u n t i l used. 63 Zmaczynski Toluene -supplied by the International Bureau of Physico-chemical Standards was used. This material was considered to have been purified as carefully as possible. 3 Griswold, Andres, and Klein Nitration-grade toluene was purified by d i s t i l l a t i o n . Methanol 65 Kay and Donham A high-purity commercial sample of methanol was d i s t i l l e d over sodium; 200 ml. of each 1500 ml. lot were retained for the measurements. VII-3 Young Methanol was prepared by d i s t i l l i n g crystalline methyl oxalate and ammonia. The d i s t i l l a t e was re d i s t i l l e d , then d i s t i l l e d over quioklime, and f i n a l l y over barium oxide. It was then allowed to stand over anhydrous calcium sulphate for a number of weeks, after which i t was d i s t i l l e d six times to a constant boiling point. 52 Rao, Sarma, Swami, and Rao Methanol free from acetone was d i s t i l l e d before use. 200, AZEOTROPE COMPOSITION 150 o o I 0> 100 0) Q. E K 50 (I O This work © Benedict et al © Berg and Harrison 3 Robinson et al 86 88 90 92 94 96 MOLE PERCENT METHANOL 98 100 3V Xi KJ A> ^ 2.0d 1 I I I i I I I I 0 10 20 30 4 0 50 60 70 80 90 100 MOLE PERCENT METHANOL -2.0d I I 1 I i 1 I I I I 0 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT METHANOL ' .teh%7 VZ RATIO OF ACTIVITY COEFFICIENTS O 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT METHANOL I tea _ , _ ; - - f r f c ^ j o c o £ c Q> C a> 3 1.50 RATIO OF ACTIVITY COEFFICIENTS IN THE SYSTEM METHANOL-TOLUENE AT 2 0 0 ° C. -.50-O Experimental data and Black's Equation — Correlation -1.50 -2.0CL 0 10 20 30 40 50 60 70 MOLE PERCENT METHANOL 80 90 100 fa, 2 8 ' MOLE PERCENT METHANOL * Mc SL 4 / 1.00 -.50 -1.00 -1.50 0 I I I i i RATIO OF ACTIVITY COEFFICIENTS N THE SYSTEM-METHANOL-TOLUENE AT 125° C. Curves are numbered to correspond with methods of calculation described in text 10 20 30 40 50 60 70 .MOLE PERCENT METHANOL /£3 C o 3 C 30 40 50 60 70 MOLE PERCENT METHANOL 100 /, W L 4 V Z i o c a JZ a> E C a> o c _J 1.50 1.00 .50 -.50 -1.00 -1.50 -2.001 — I I I I I I . RATIO OF ACTIVITY COEFFICIENTS IN THE SYSTEM METHANOL - TOLUENE A T 175° C. — O Experimental data and Black's Method ^ ~ ^ Q — — Correlation i i i 1 I I 1 I I 10 20 30 40 50 60 70 MOLE PERCENT METHANOL 80 90 100 Ac I J>c§ 3>C >C4 X )C2 8 * 7 X . )(6 5^< ! J 3>C )(4 IX ±2. POSITION A POSITION B THERMOCOUPLE POSITION FOR T E M P E R A T U R E SURVEY 7J1 -5"-^ r 5^ ^ H \ ^ ^ f |so .fe o& ^ -^j M V J MERCURY-SEALED GROUND GLASS JOINT \ J LIQUID SAMPLE COLLECTION TUBE VAPOR SAMPLE COLLECTION TUBE SAMPLE COLLECTION TUBE S 12 MERCURY CUT:-OFF VALVE TO ACUUMl TO PRESSURE SAMPLING TUBE GLASS TO METAL SEAL K MERCURY TRAP