A STUDY OF THE TERNARY SYSTEM DIOXANE- TETRAHYDROPYRAN-DINITROGEEN TETR OXIDE

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By BETTY JANE GIBBINS,\t 9 B.S. 9, M.S. The Ohio State University 1953

Approved by: 1

ACKNOWLEDGMENTS

The author expresses her appreciation to Dr, Harry H. Sisler for suggesting this research and for his constant interest and guidance throughout the course of the investigation. Appreciation is also extended to Dr. Gunther L. Eichhorn and Dr. Bernard Rubin for data for the binary systems. The author also expresses her appreciation to Dr. Fairfax E. Watkins for constructing Figure 28 and to Dr. William J. Taylor for contributing to the discussion of the structure of the dioxane— dinitrogen tetroxide complex. That part of this research was conducted with the assistance of the Ordnance Corps, U.S. Army, through a contract with The Ohio State University Research Foundation, is also gratefully acknowledged.

\ 12:\'7'7fy ii

TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ...... i LIST OF FIGURES ...... iv LIST OF TABLES ...... vl I. INTRODUCTION ...... 1 II. STRUCTURE OF DINITROGENTETROXIDE ...... 2 III. MOLECULAR ADDITION COMPOUNDS OF DINITROGEN TETROXIDE ...... 5 IV. TERNARY PHASE DIAGRAMS ...... 11 a. The Phase Rule ...... 11 b. The Right Triangular Prism ...... 12 c. Plotting Concentrations ...... 13 d. System with no Compounds or Solid Solutions ...... lb 1. Experimental Behavior ...... 15 2. Space Diagram ...... 17 3. Projection of the Space Diagram on the Base ...... 20 e. Systems with Compounds ...... 22 V. CRYOSCOPIC STUDY OF THE SYSTEM,DIOXANE— TETRAHYDROPYRAN-DINITROGEN TETROXIDE .. 24 a. Plan of Investigation ...... 24 Ill

Page b. Purification of Materials ...... 26 c. Experimental Procedure ...... 31 d. Results ...... 38 e. Discussion of Results and Conclusions .. 78 f. Summary ...... 88 BIBLIOGRAPHY ...... 92 AUTOBIOGRAPHY ...... 95

4 iv

LIST OF FIGURES

Figure Page 1. A Concentration Triangle ...... 14 2. System with no Compounds or Solid Solutions ...... 16 3. Projection of the Diagram on the Base ...... 21 4. Dinitrogen Tetroxide Purification Apparatus ...... 29 b. Dinitrogen Tetroxide Transfer Cell ...... 30 6 . Ether Storage Flasic ...... 32 7. Freezing Point Cell ...... 33 8. Transfer of Dinitrogen Tetroxide to Freezing Point Cell ...... 34 9. System: 10/90 mole per cent CeHlo0/0(CH2CH2)20-Na04 42

tt 11 tt it »t tf 10. 2 0 / 8 0 4 4

tt If n tt tt tt 11. 3 0 / 7 0 4 6

M tt tt M ft tt 1 2 . 4 0 / 6 0 4 8 bO/bO !1 *t ti tt tt tt bO

11 tt tt tt tl tt 6 0 / 4 0 b2

It tl tt tl 11 ft 7 0 / 3 0 0 4

tt tf it t1 tt tt 6 0 / 2 0 b6

It ft it tt ft It 17. 8 8 / 1 2 0 8

11 tt tt tt tt tt 18. 9 0 / 1 0 6 0

tt tt it It ft tt 19. 9 2 / 8 6 2

tt tt tt tt tt tt 20. 9 6 / 4 6 4 Figure Page

21. System: 94.1/5.9 mole % CgHioO/NgO^-OlCH2CH2)fiO.. 66 22. " 91.8/8.2 " " " H " ..67 25. " 86.6/13.4 " « « " « ..69 24. " 7 7.2/22.8 ” ,f " H " ..70 2b. " 60.1/34.9 " " n " " ..72 26. " 04.2/40.8 •* " " " " ..73 27. Projection of the Diagram on the Base ...... 76 28. System: Dioxane-Tetrahydropyran-Dinitrogen Tetroxide ...... 77 v i

LIST OF TABLES

Tab la Page

1. System 10/90 mole % C6Hio0/0{CH8CHS )s 0-N80. . 41

H tl tt H tt 2. 20/80 ” m 43

tf ft tt tt tt 3. 30/70 " e 4b

tf tf 4. 40/60 " tt tt ft • 47

ft tt tl tt ft b. 50/50 " • 49

tf If tt tt It 6. 60/40 " * bl

M tf tt tt ft 7. 70/30 " • 53

tt tf tt t* tt 8. 80/20 ” • 55

tl ft y . 88/12 " tt tl tt • 57

If 10. tt 90/10 " if tl tf • 59

tt n n tt ft n . 92/8 " • 61

tt H n tv tt 12. 96/4 ” • 63 13. System 94.1/b.9 mole % cbh 10o/nso.- 0(CH8CH2 )20 65

14. tt 91.8/8.2 tt tt n tt tt • 65 lb. n 66.6/13.4 tt tt tt rt M • 68

it tf 16. 77.2/22.8 tt It tl It • 68

17. tt 6b.1/34.9 it tt II tt ft * 71

H tt tt ft tt tt 18. b4.2/4b.8 • 71 19. Maxima from Figures 9 tnrough 26 74 20. Binary Eutectic Mixtures ...... 75 21. Ternary Eutectic Mixtures ...... 79 A STUDY OF THE TERNARY SYSTEM DIOXANE-TETRAHYDROFYRAN-DINITROGEN TETROXIDE

I• INTRODUCTION.

During the course of an investigation of molecular addition compounds of dinitrogen tetroxide with some ethers Rubin, Sisler, and Shechter^ found that 1,4-dioxane forms an addition compound with properties strikingly different from those of the addition compounds of the other ethers. It was noted that 1,4-dioxane and tetrahydropyran form dinitrogen tetroxide addition compounds with different mole ratios and strikingly different melting points. Dinitrogen tetroxide-dioxane melts at 45.2*C.; dinitrogen tetroxide*2 tetrahydropyran, at -56.8°C. A bicyclic monomer, a dimer, and a polymeric aggregation were proposed as possible structures for the dioxane-dinitrogen tetroxide addition compound, A study of the ternary system dioxane-tetrahydro- pyran-dinitrogen tetroxide was proposed (1) to investigate the possibility of a ternary compound, (2) to compare the stability of the crystalline dinitrogen tetroxide-dioxane compound with that of the crystalline dinitrogen tetroxide- tetrahydropyran compound over a large range In concentra­ tion, and (3) to perhaps more fully substantiate or 2 eliminate some of the structures proposed for the binary compounds. Although compos!tion-temporature phase diagrams have been widely used to detect compounds and solid solutions in equilibrium mixtures of two components, phase diagrams for non metallic, non aqueous, ternary systems are believed to be rare. Study in the field of ternary diagrams is compli­ cated by the relative complexity of the systems and by the necessity or representing the data in three dimensions. The system dioxane-tetrahydropyran-dinitrogen tetroxide thus offered an opportunity to construct a relatively rare type of diagram and to develop technique in determining and representing ternary systems.

IX. STRUCTURE OF DINITROGEN TETROXIDE

The structure of dinitrogen tetroxide has been a matter of much discussion. The following three structures have received considerable support:

I ii III

+ -0 N -0-N o - n v ',0sn'm-o

./ \ . a 0 0 3

Structure I formerly was favored on the basis of chemical evidence, but now has been discarded on the basis of physical evidence. It was used to explain such reactions as the reaction with olefins to form nitroso, nitro, nitrite, and nitrate derivatives. More recently Levy and 2 Scaife studied this reaction and reported that, unless dinitrogen trioxide is present as an Impurity, no nitroso compounds are formed. The formation of the dinitroalkanes and nitronitrites has been explained by Ingold3 , using structure II and an electrophilic attack. The N0e+ attaches Itself first, and the N0fi” is free to attach either as a nitro or nitrite group. The slight dissociation of dinitrogen tetroxide Into N0+ and N03” , which was indicated by Addison and Thompson,4 can be explained with structure II or III. With structure II, a N-0 bond would be broken and an oxygen atom would be transferred across the weak N-N bond. With structure III, only an electron transfer is necessary. Thus, chemical evidence does not clearly differentiate among the three structures. Structures II and III are supported by both chemical and physical evidence. Infrared absorption and Raman scattering spectra indicate that dinitrogen tetroxide is symmetrical and non- 4

linear. Sutherland'5 Interpreted the low, very intense Raman frequency as evidence for a weak N-N bond and evidence for structure II. Longuet-Higgins6 proposed structure III and stated that the low, very intense Raman frequency could be caused by angular deformation of the ring. Thus, either structure seems to be consistent with spectroscopic data. Longuet-Higgins calculated the entropy value from spectroscopic data, and found that, in order to obtain agreement between his entropy value and the experimental value of Giaque and Kemp7 , he had to assume no free relative rotation of the ends of the molecule. This seemed to give additional support to structure III. More recently, how­ ever, Bernstein and Burns® calculated the entropy, by using structure II and the following bond distances and angles: angle 0N0 = 120°, N-0 - 1.15A°, and N-N * 1.66A®. Finding that they had to assume internal rotation to make their calculated value agree with that of Giaque and Kemp7 , they discarded structure III. The most conclusive evidence seems to be the X-ray data of Broadley and Robertson®. Their study of single crystals of the solid Indicates that the molecule is planar, and the bond distances and angles are as follows: N-N, 1.64 ± 0.03A*; N-0, 1.17 * 0.03*; 0-0 (both on one N 5 atom), 2.09 ± 0.03A*; and the angle 0N0, 126 * 1*. Structure II and the above measurements will be assumed in subsequent discussions.

III. MOLECULAR ADDITION COMPOUNDS OF DINITROGEN TETROXIDE.

Spath10 prepared crystalline U0 s (N03 )2*2N0s, by adding a mixture of dinitrogen pentoxide and dlnitrogen tetroxide to a solution which had been prepared from partiall dehydrated U02(NOs )2 »6H20 and fuming nitric acid. Composition-temperature phase diagrams constructed by Pascal and Gamier-*--*- indicated the formation of two com­ pounds: 5 dinitrogen tetroxide*4 camphor, with a melting point of -52*C., and 2 dlnitrogen tetroxide*3 camphor, with a melting point of -45.5°C. n 2 Reihlen and Halce*1- prepared the compounds 2 SnCl4 * 3N204 and 2 TiCl**3N204 by adding dinitrogen tetroxide to a carbon disulfide solution of the chlorides. Addition compounds of dinitrogen tetroxide and various polycyclic ethers were reported by Lukin and Dachevskaya. The existence of a compound with the empirical formula, N80A •2EteNN0, end a melting point of 37.5*C. was detected by the cryoscopic measurements of Addison, Conduit, and Thompson. 6

The first addition compound of dinitrogen tetroxide with an ether was reported by Crowder.^-5 He isolated a white crystalline powder with the empirical formula NgO^* 0(CH£CHe)20 from a mixture of the two components in petroleum ether. In studying the reaction of dlnitrogen tetroxide 2 with olefins, Levy and Scaife noted that the use of certain ethers as the solvent reduces the oxidation of the olefin and greatly increases the yield of addition products. It was suggested that the effect Is produced by the ether forming a molecular addition compound with the dinitrogen tetroxide. This intermediate compound might react with the olefin to form addition, not oxidation, products. To detect such molecular addition compounds, Rubin, Sisler, and Shechter^ constructed temperature—composition phase diagrams of systems containing dlnitrogen tetroxide and each of the following ethers: diethyl ether, tetra­ hydropyran, tetrahydrofuran, 1,4-dioxane, and B,B-dichloro— diethyl ether. Compounds with the following empirical formulas and melting points were detected: Ng04-2(CfiH8)gO, -74.8WC•; N204 '2C6Hlo0, -56.6°C.; NgO*-C^HgO, -20.b°C.; NB04*2C4H80(7), incongruent melting point, and NgO** 0(CHgCHg)gO, 4b.2°C. No compound formation was detected In the system, B, B* dichlorodiethyl ether-dinitrogen tetroxide. It was

A 7 suggested that the chlorine atoms would both decrease the electron density around the ether oxygen atom and would also cause steric interference with the NgO^ molecule. Magnetic and spectroscopic data indicated that the dinitrogen tetroxide is present as NgO^ units, not NOg , N02+ , NOg” , N0+ or N03“. The sharpness of the maxima in their composition- temperature curves indicated that tetrahydrofuran and tetrahydropyran form dinitrogen tetroxide addition com­ pounds which are more stable than that of diethyl ether. Brown and Adamsfound that tetrahydrofuran is a stronger base than diethyl ether with respect to boron trifluoride also. They attributed the difference in basicity to the fact that the terminal methyl groups of diethyl ether offer more steric hindrance tnan a rigid five-membered ring. In the lowest energy state, diethyl ether has the following configuration:

Since, in this configuration, the methyl groups would interfere with tne addition of a bulky acid molecule to the ether oxygen, Brown and Adams proposed that the ether 8 molecule must form a new configuration and reach a higher energy state. The stability of addition compounds thus formed would be lessened to some extent because of the strain produced by the rearrangement of the ether molecule. The difference in stability of the dinitrogen tetroxide addition compounds with diethyl ether and with the cyclic ethers tetrahydrofuran and tetrahydropyran could be explained in the same way. The bicyclic structure proposed for the compound, dioxane•dinitrogen tetroxide, is shown on page 8 5 . If polymeric aggregates are formed, liquid mixtures of the two components should have considerably higher viscosities than the pure components. By measuring viscosities over a concentration range from 0 to ,b mole fraction dinitrogen tetroxide Ling and Sisler^7 showed that no high polymers are present at a temperature slightly above the melting point. To differentiate between the dimeric and monomeric structures, Ling and Sisler^7 measured the depression of the freezing point of dioxane produced by a .known weight of the dinitrogen tetroxide-dioxane compound. Their results indicate only one Ng0^ unit per molecule. To determine the efrect of longer and more bulky alkyl groups on the ether, to show the effect of electro­ positive and electronegative groups placed on an ether 9 ring, and to compare the stability of the 1:1 dinitrogen tetroxide -1, 4-dioxane compound with that of the compounds formed by dinitrogen tetroxide and other "dibasic" and "tri basic" etners, Whanger and Sislerl® constructed phase diagrams of systems containing dinitrogen tetroxide and the following ethers: di-n-propyl ether, di-iso-propyl etner, di-n-butyl ether, di-tert-butyl ether, perfluoro- tetrahydrofuran,o<6-methyltetrahydrofuran, ethylene glycol diethyl ether, 1,3-dioxane, and trioxane. The dinitrogen tetroxide-diethyl ether compound melts congruently at -74.b®C. The di-n-propyl ether and di-n-butyl ether compounds with dinitrogen tetroxide have incongruent melting points of -77.£>°C. and -79,&°C. Thus, an increase in the length of the alicyl straight change seems to cause some lowering of the tendency for compound formation. The formation of a compound between di-iso-propyl ether and dinitrogen tetroxide, with a probable mole ratio of 1 N804/2 ether and an Incongruent melting point of -60°C., indicates that the presence of two methyl groups on the«£ carbon atoms does not eliminate compound formation. The presence of three methyl groups on the oC carbon atom, as in dl-tert-butyl ether, does prevent compound formation under the conditions studied. The replacement of the hydrogen atoms in tetrahydro- 10 furan with fluorine atoms decreases the electron density about the ether oxygen sufficiently to prevent compound formation. The presence of the methyl group in ^-methyl tetrahydrofuran apparently does not cause serious steric hindrance. A compound with an ether/dinitrogen tetroxide mole ratio of S/1, and a melting point of -50.5°C. is formed. 1,3-dioxane and dlnitrogen tetroxide form a 1:1 addition compound, which melts at 2.0°C. and is approx­ imately as stable as the 1 ,4-dioxane-dinitrogen tetroxide compound. Ethylene glycol ether forms two dlnitrogen tetrox­ ide addition compounds which are much less stable than the 1,4-dioxane compound. The possibility of steric hindrance is much less with the rigid six—membered dioxane ring than with the ethyl groups in the compounds Ng04 •CeH60CHgCHs0CeH5 and N20 4 -2 C2H50 CH2 CH20 C2Hb . The composition—temperature phase diagram indicates that trioxane forms a compound with a probable 1:1 mole ratio and an incongruent melting point of -11*C. The steep slope of the curve approaching the incongruent melting point indicates that its stability is perhaps comparable to that of the 1,4-dioxane compound. 11

IV. TERNARY PHASE DIAGRAMS a. The Phase Rule.

The behavior of a heterogeneous system at equilibrium is described by the phase rule, which is expressed by the formula:

P + F = C + 2 (1).

There P represents the number of phases, F represents the number of degrees of freedom, and C represents the number ofcomponents in the system. Phases are the homogeneous portions of a heterogeneous system, which are separated from one another by definite surfaces and differ from one another in properties. The number of degrees of freedom is the number of factors (pressure, temperature, concentration) which cat be varied inde­ pendently, within limits, without charging the number of phases present. Components are the constituents whose concentration can be changed independently of the other constituents of the system. If the system is under a pressure greater than the vapor pressure of the substance, the vapor phase becomes negligible and the pressure is constant. For such a "condensed” system, one degree of freedom is lost and the 12 phase rule takes the form shown below.

P + 3T - c + 1 (2).

Equation is used for a three component "condensed” system.

P + F - 4 (3). b. The Right Triangular Prism.

The phase rule itself merely indicates the relation snip between the number of phases that exist in equilibrium and the number of independently variable conditions for a heterogeneous system of £ components. To indicate the nature of the phases present at specific conditions, a phase diagram must be experimentally determined. Equation (3) indicates that the greatest number of independent variables which can exist in a ternary "condensed” system at constant pressure is three. It is therefore possible to represent the system in a three dimensional space diagram. A simple, frequently-used model is shown in Figure 2. It is a right triangular prism with an equilateral triangle as the base. The concentra­ tions of the three components are plotted on the base. The temperature is plotted perpendicularly to the base. Thus, all points representing the state of a system of 13 composition X at different temperatures lie on a line perpendicular to the base at X. c. Plotting Concentrations.

In Figure l t points A, B, and £ represent the corresponding pure components. Points on lines AB, B C , and CA represent binary systems of A and B, B and C, and C and A, respectively. All points within the triangle, such as M, represent mixtures of the three components. Using a simple geometric law for an equilateral triangle, Masing and Rogers^-*- show that, for a mixture of total concentration M, the ratios concentration of A/total concentration, concentration of B/total concentration, and concentration of C/total concentration are given by the ratios yp , yp , and yp , where Ma, Mb, and Me are the perpendicular distances from the point M to the side opposite the component whose concentration is being de­ termined. In practice, it is unnecessary to draw and measure the perpendicular distances. If line pq is parallel to side AB. then - Ac E"~ IS • To obtain ratio concentration C in M it ls BC total concentration necessary to consider the line (pq) through the point M and parallel to the side opposite C (AB). The ratio V

14

A

m

C

Fig. I. A Concentration Triangle or can be measured. It; is customary to use graph paper BC with one hundred divisions per side and to read the molar percentages directly from the paper. Similarly, the molar percentages of A and B in mixture M are given by the ratios

or and X§ °r * resPectively- a line is

drawn from one corner (£) to the opposite side (AB), the points on the line represent systems with varying amounts of that component (C) and a fixed ratio (DB) of the other two — AD components (A to B).

d. System with No Compounds or Solid Solutions.

1. Experimental Behavior.

A solution of composition and temperature repre­ sented by X in Figure 2 is cooled. At elevated temperatures, only one phase, the homogeneous melt, exists. If a con­ stant temperature gradient is used, a plot of time versus temperature of the homogeneous melt gives a straight line. At a certain temperature, crystals of B begin to form, and the slope of the cooling curve changes. As the crystallization of B continues, the concentration of A and C in the melt increases, and the freezing point continues to drop. 16

X

E

A B

C

Fig. 3. System with No Compounds or Solid Solutions 17

When the solution becomes saturated with A as well as B, a binary eutectic mixture of A and B begins to crystallize. At this temperature, the cooling curve shows a second slope change. During the binary eutectic crystallization the freezing point continues to drop, as the concentration of C in the melt increases. Finally, the melt becomes saturated with respect to C. Then all three components crystallize as a ternary eutectic mixture. The temperature remains constant until all the melt has solidified. A plateau on the cooling curve indicates the freezing point of the ternary eutectic mixture.

2. Space Diagram.

The behavior just described is represented on the space diagram in Figure 2. X represents the composition of the system and the initial temperature. As the system is cooled, the point representing the system moves down the perpendicular line XY. So long as the system remains homogeneous, according to equation (3), it posses three degrees of freedom. Thus, all points representing a homo­ geneous system lie within a given space. Temperature and concentration can be varied at will, within the limits of this space, without causing a second phase to appear. 18

Point Y represents tne system of composition X at the temperature at which B beings to crystallize. If one were to make up other compositions and determine the temperatures at which the first crystals (A, B, or C) appear, a series of such points would be obtained. Since the system now contains two phases, melt and solid, and two degrees of freedom, such points form, not a space, but a surface called the "surface of primary crystallization". As the identity of the first crystals changes, the shape of the surface changes. Thus, each substance has its in­ dividual surface of primary crystallization. In Figure 2 surfaces EKLF. GFLH. and TKXH represent the surfaces of primary crystallization for A, B, and C, respectively. As the solid is cooled below Y, the point repre­ senting the total system moves to 2, a point within the space of primary crystallization. This space is bounded above by the surface of primary crystallization and below by a surface representing the concentrations and temperatures where binary eutectic crystallization begins. To obtain the composition of the melt in the total system represented by Z, it is necessary to draw a horizontal line through to the edge (G-B) representing the pure component which has begun to crystallize (B). Since the melt is more dilute in B, the line is extended in a direction away from _B until it intersects the surface of primary crystallization at 0. 19

When the total system has cooled to Z, it has separated into pure B and a melt of composition Q, Constructing a series of such tie lines or conodes indicates that during the crystallization of B, the composition of the melt moves along curve YOM. M represents the composition of the melt and the temperature at which binary eutectic crystallization of A and B begins* ~ Since there are now three phases, melt, B, and A, the system has only one degree of freedom. Points such as M, representing systems saturated with respect to two components, thus lie on space curves called curves of binary eutectic crystallization. They are the intersections of the various surfaces of primary crystallization. In

Figure 2 , curves K L , FL and HL represent the curves for binary eutectic mixtures of A and C, A and B, and B and C, respectively. During the binary eutectic crystallization of A and B, as the freezing point falls and the melt becomes richer in C, the composition of the melt follows the curve ML. L represents the composition of the ternary eutectic mixture and its constant freezing point. Since there are now four phases present, the system is invariant. In summary, as X is cooled, the composition moves down XY until B begins to crystallize at Y. The compo­ sition of the melt and its freezing point move down YQM 20 until binary eutectic crystallization of A and B begins at M. During the binary eutectic crystallization, the composition of the melt and its freezing point move down ML. At L tiie ternary eutectic crystallization occurs at constant temperature.

3. Projection of the Space Diagram on the Base.

For convenience in plotting curves and in obtaining their intersections, it is customary to construct a pro­ jection of the space diagram on the triangular base. In Figure 3, Af-£jc, K^&hC, and represent the projections of the surfaces of primary crystallization of A, C, and B, respectively. The projections of the curves of binary eutectic crystallization for A and B, B and C, and C and A are represented by f JL-, ^ h . and , respectively. ^ 1 3 the projection of L in Figure 2. Temperature is not repre­ sented in Figure 3, but it is simple to follow the compo­ sition of the melt x during cooling. Since x lies in fBh^, B begins to crystallize first. If a line is drawn through x and B, the coiaposition of the melt during the crystalliza­ tion of B follows the line xB in a direction away from B, i.e., toward x *. When this line intersects the curve iJL, the binary eutectic crystallization begins. Point M represents the composition of the melt at that time. During 21

A

w'x

m

C

Fig. 3. Projection of the Diagram on the Base

I 22

tne binary eutectic crystallization, the composition of the melt follows l U L - The composition of the ternary eutectic mixture is represented by J2/ .

e. Systems with Compounds.

A ternary system may contain ternary or binary com­ pounds. The presence of a ternary compound with a congruent melting point is indicated by a peak on the surface of primary crystallization within the concentration triangle. Such a peak would be surrounded by a surface of primary crystallization for the ternary compound. This surface would intersect the other surfaces of primary crystalliza­ tion forming at least one extra curve of binary eutectic crystallization. The presence of a binary compound which originates in the binary system and participates in the ternary system is indicated by a maximum on the binary curve on the face of the prism. Figure 26 and Figure 27 show the space diagram and its projection on the base for a system containing two such binary compounds. The diagram is discussed in Section Ve. Such systems may be regarded as composed of two, or in this case, three simple systems placed side by side. There are no new terms involved. Only the number of surfaces of primary crystallization, the number of binary eutectic curves and the number of ternary eutectic points increases. Two additional principles arise. Since the ternary eutectic points are each the lowest points in tneir immediate sur­ faces, tne binary eutectic line connecting two ternary points must rise from one ternary, pass through a maximum temperature and fall to the second ternary point. In Figure 28, both curve CB and curve CO- pass through a maxi­ mum. Each ternary point is formed by the intersection of three curves of binary eutectic crystallization. No more than three ourves can intersect, and they must intersect in such a way that only three different solid phases exist in equilibrium with the melt. According to the pliase rule, only four phases in equilibrium give such an invariant point. Thus, in Figure 27, points B and C! cannot coincide. To follow the composition of the melt as a mixture in Figure 27 is cooled, one uses the same procedure used in tne simple system without compounds or solid solutions. A line is drawn from the point to the phase crystallizing first. The concentration follows this line away from the crystalline constituent until it crosses one of the binary eutectic curves. It then follows this curve down to the nearest ternary eutectic point. Compositions to the left of tne maximum in CG change to the ternary point C_; those to the right of the maximum, to G. Systems which inter­ sect CG on the maximum solidify completely at the maximum 24 temperature. Since they behave like binary systems, they are called "quasi-binary sections" of the system.

V. CRYOSCOFIC STUDY OF THE SYSTEM, DIOXANE- TETRARYDROPYRAN-DINITROGEN TETROXIDE a. Plan or Investigation.

This method is only one of several possible approaches to the problem. The amount of data taken de­ pends upon how complete a determination of the surface of primary crystallization is desired. The minimum data requirement is enough to construct the projections in Figure 27. To determine which sections of the diagram need close study, it is customary to study sections of the diagram represented by lines from one corner and inter­ secting the opposite side at regular intervals. Since mixtures with a fixed tetrahydropyran/dioxane mole ratio could easily be prepared and stored, it was convenient to study systems containing a fixed dioxane/tetrahydropyran mole ratio and varying amounts ol‘ dlnitrogen tetroxide. These systems are represented by lines from the dlnitrogen tetroxide corner to the opposite side of the ternary 25 diagram. Composit ion-temperature phase diagrams were con­ structed for systems with fixed tetrahydropyran/dioxane mole ratios at ten per cent intervals from 10/90 to 90/10. The concentrations were plotted on the triangular base. They were marked off on a separate rectangular graph paper, and the curve was plotted on the rectangular paper, cut out and mounted perpendicularly to the base. Since neither of the two binary eutectic curves obtained had yet reached a minimum temperature, it was indicated that sill three ternary eutectic points were probably between the planes of the systems 90/10 mole per cent tetranydropyran/dioxane- dinitrogen tetroxide and tetrahydropyran-dinitrogen tetroxide. When this had been substantiated by studying the system 8 8 / 1 2 mole per cent tetrahydropyran/dioxane- dinitrogen tetroxide, the systems 92/8 and 96/4 mole per cent tetrahydropyran/dioxane-dinitrogen tetroxide were studied. After the compositions of the binary eutectics were plotted on triangular paper, the curves FE and IH in Figure 27 and the approximate location of the ternary eutectics were established. The location of the binary eutectic A and the approximate locations of the ternary eutectics indicated that there were at least two approximately horizontal curves near TTJ in Figure 27. Such curves could be established by studying portions of 26 systems which would cross these curves, i.e. systems con­ taining fixed tetrahydropyran/dinitrogen tetroxide mole ratios and varying amounts of dioxane. Composition- teiaperature phase diagrams were therefore constructed for systems containing the following fixed tetrahydropyran/ dinitrogen tetroxide mole ratios: 94.1/5.9, 91.8/8.2, 86.6/15.4, 77.2/22.8, 65.1/34.9, and 54.2/45.8. b. Purification of Materials.

Dioxane (Carbide and Carbon Chemical Corporation) was purified by Feiser’s method. 22 Approximately one part by volume ol concentrated hydrochloric acid, 7.25 parts water, and 72.5 parts dioxane were refluxed for twelve hours, while a stream of nitrogen was passed through the solution to entrain acetaldehyde impurities. The cool solution was neutralized by adding potassium hydroxide pellets with shaking, until the solution was saturated and a second layer had separated. The dioxane layer was de­ canted, dried with fresh potassium hydroxide pellets, and refluxed for twelve hours with sodium. The dioxane was then distilled from sodium under a nitrogen atmosphere. The fraction boiling at approximately 100.5°C. (748 mm.) was collected. The value reported in the literature is ioi.i^c.20 27

The freezing point or the puriried dioxane, ll.b®C., agreed with the literature value of 11.7®C. , 2 4 and with the

1 2 ® value obtained most frequently by Dr. G. L. Eichhorn, wording in this laboratory. Dr. Eichhorn also obtained a second freezing point, 10.1®C., which conforms to the o value, 10.b°C., reported by Rubin, Sisler and Shechter.~ It is oelieved tnat botn tetrahydropyran and dioxane exist in two crystalline forms. Tetrahydropyran (E. I. du Pont de Nemours Co.) was dried over sodium hydroxide pellets for two days and then fractionally dis­ tilled from sodium under a nitrogen atmosphere. The fraction boiling at approximately 161.0°C. (747 mm.) was collected. The boiling point agreed with that reported by

Rubin, Sisler and Sheehter.2 Both of the ethers were distilled at a reflux ratio of between 2 0 and bO to 1 , in a three-foot column packed with glass helices. Ttie freezing points of samples obtained by separate distillations of different portions of tetrahydropyran, -49.b®C. and -4Q.7b°C., agreed, within the limits of accuracy of the apparatus, with tne previously reported value of -49.0®C.J-y This value is b® lower than that reported by Rubin, Sisler and Shecnter.^ Dr. G. L. Eichhorn, working in this laboratory, obtained both melting points on the same sample. 28

To prevent tne formation of peroxides, botn etners and tneir mixtures were stored over sodium and iron wire, under a nitrogen atmospnere, and in a dark place. Dinitrogen tetroxide (Matneson Company) was purl- fled by a method similar to that of Griaque and Kemp* The all-glass apparatus is shown in Figure 4. The gaseous dinitrogen tetroxide from a metal cylinder was condensed In the first trap, whicn was surrounded by a mixture of dry ice, carbon tetrachloride, and cnloroform. The blue solid was then melted and kept at 0 *0 . by an ice-water bath, while dry oxygen was bubbled through the liquid until the dark green color had changed to orange-red. The dry ice cooling bath was then placed around the dinitrogen tetroxide transfer cell (Figure 5), which was protected from moisture in the air by a drying tube filled with phosphoric anhydride and sand. The purified liquid was then distilled, either at room temperature or with slight warming, from the first trap, through two 16-inch glass tubes filled with phosphoric anhydride and sand, into the transfer cell. The dry dinitrogen tetroxide formed a white crystalline solid, which was stored in a dry ice-carbon tetrachloride-chlorofora bath. Its freezing point, -11.5*C.,

agreed with the literature value.7 In all equipment containing dinitrogen tetroxide, the ground glass joints were lubricated with graphite and a small amount of high P4O10 and sond Impure N2O4

P4OK) and sand

. 4. DINITROGEN TETROXIDE PURIFICATION APPARATUS. 30

FIG. 5. DINITROGEN TETROXIDE TRANSFER CELL 31 vacuum stopcock grease. They were held together by rubber bands. c. Experimental Procedure.

The mixtures containing a fixed tetrahydropyran/ dioxane mole ratio were prepared by combining the calculated weights and storing them in ether storage 1‘lasks like the one shown in Figure 6 . The freezing points of the various mixtures were determined in an all-glass cell equipped with two standard

taper ground glass joints (Figure 7 ). Known amounts of ether were admitted, with minimum exposure to the air, through a Lunge pipette ground to fit the 12/30 outer joint on the freezing point cell. During the addition, tne other inlet (a 14/35 S.T. inner joint) was attached to a drying tube filled with phosphoric anhydride and sand. Dry nitrogen was used to force the ether from the ether storage flask into the Lunge pipette. The amount of ether added to the freezing point cell was determined by weighing the Lunge pipette before and after the addition* The dinitrogen tetroxide transfer cell was con­

nected, as shown in Figure 8 , to the freezing point cell by means of a glass bridge equipped with ground glass ends and an oblique—bore stopcock at the center. After the 32

FIG. 6. ETHER STORAGE FLASK 33

FIG. 7. FREEZING POINT CELL 03 •ft

\L/

FIG. 8. TRANSFER OF DINITROGEN TETROXIDE TO FREEZING POINT CELL. 35 dinitrogen tetroxide transfer cell had reached room tempera­ ture and the freezing point cell had been placed in either liquid nitrogen or a dry ice-carbon tetrachloride-chloroform bath, the stopcocks on the bridge and transfer cell were opened. The dinitrogen tetroxide distilled into the freezing point cell. When approximately the desired weight had distilled (This was judged by the lowering of the liouid level in the transfer cell.) the stopcock on the transfer bridge was closed. The transfer c'^ll was cooled with liquid nitrogen until all the dinitrc, tetroxide vapor beyond the bridge stopcock had condensed in the transfer cell. The stopcock on the transfer cell was closed, and the transfer cell was removed from the bridge and stoppered. At first the amount of dlnitrogen tetroxide added was determined by the difference in weight of the freezing point cell and bridge before and after the addition. This was advantageous when leaks in the bridge or dinitrogen tetroxide transfer cell occurred. As systems with higher ratios of tetrahydropyran were studied, some oxidation occurred in the mixture as the freezing point cell was warmed to room temperature before weighing. To avoid warming the freezing point cell and to minimize the reaction, the method was changed. The amount of dinitrogen tetroxide added was determined by weighing the dinitrogen tetroxide transfer cell before and after the addition. 36

The mixtures were stirred by a glass stirrer, equipped with an enclosed iron slug and actuated by an intermittent magnetic field actuated by an air-cooled solenoid placed around the neck of the freezing point cell* Usually the cooling rate was 1° to 2® per minute. The rate was controlled by the choice of jacket and of coolant. Different sizes of single glass tubes, two con­ centric glass tubes insulated with glass wool, and a vacuum jacket were used in various instances. Various cooling baths were used, depending upon the freezing point. Liquid nitrogen was used below approximately -50*C.; carbon tetrachloride-chloroform-dry ice mixtures, from approximately -50° to -12°C.; ice and sodium chloride eutectic mixture, from approximately - 1 2 to +1 0 ®C.; ice baths or the air, above 10®C. To avoid supercooling, the mixtures were frozen, then gradually warmed by hand. Just as the last crystals were melting, the cell was immersed in the cooling jacket. By this method the temperature of the mixture was not allowed to rise very much above the freezing point. Supercooling seemed to be reduced by the use of cells which had been scratched by the stirrer, and, in some cases, by placing a few tiny pieces of platinum wire inside the cell. Time-temperature cooling curves were recorded

A 37 with a recording potentiometer (Leeds and Northrup Micromax), using a copper-constantan thermocouple immersed in the well of the rreezing point cell. The heat transfer medium in the well was either a carbon tetrachloride- chloroform mixture or petroleum ether, depending upon the temperature• The apparatus was calibrated at the freezing points of distilled water, 0°C., chloroform, -63.b°C., carbon tetrachloride, -22.6°C. , and diethyl ether, -116.3° and -123.3°C, The data recorded in the following tables have been corrected in accordance with the calibration. All freezing points were determined at least twice, and are believed to be accurate within -1.5°C. 38

d. Results.

Concentration-temperature phase diagrams were con­ structed for the following systems:

10/90 mole % tetrahydropyran/di oxane--dinitrogen tetroxide 20/80 n « n tt n it 30/70 N « tt n ti tt 40/60 tl it ii •t n n b0/50 tt H n n tt n 60/40 M n •t n it tt

n it H it 70/30 m n

tt tt It « 80/20 H t«

8 8 / 1 2 tf tt n tf tt it 90/10 n ft •t it tt tt 92/8 if If tt •i n it 96/4 tf ft it »t tt n 39

Portions oi the following systems were studied:

94.1/5*9 mole % tetrahydropyran/dinitrogen tetroxide-dioxane

91.8/8,2 n n m tt tt M

86.6/13.4 " M " "

7 V. 2/22.8 M ” " " « « 60.1/34.9 « " * « « n 54.2/43.8 « « « „ «

Data for these systems are listed, in the above order, in Tables 1 through 18.

The compos iti on-1 emperatur e curves are plotted in Figures 9 through 26, respectively.

The freezing points and compositions of the maxima in the curves in Figures 9 through 26 are listed in Table 19. Table 20 contains the compositions and freezing points of the binary eutectic points determined from the data repre­ sented by Figures 9 through 26. The binary eutectic compositions from Table 20 are plotted on the concentration triangle. The resulting projection on the base of the space diagram is shown in Figure 27. The data from Tables

1 through 18, the data for the systems tetrahydropyran- dinitrogen tetroxide and dioxane—dinitrogen tetroxide reported by Rubin, Sisler and Shechter,^ and the data for the system tetrahydropyran-dioxane studied by Dr. G. L. Eichhorn working in this laboratory, were combined to make a model or the three dimensional phase diagram. A sketch of this model is shown in Figure 28. 41 TABLE 1. SYSTEM: 10/90 MOLE PER CENT TETRAHYDROPYRAN/DIOXANE-Ng0A Mole^ Freezing iSutecti o * Ng0^______point;, *C. temp., v/. 0 7.0 4.4 3.5 7.7 2.0 -0.5 10.2 1.0 -1.0 10.3 0.5 11.4 0.5 12.1 2.0 - .05 14.1 6.3 -1.0 13.0 6.5 -1.3 16.9 13.73 19.2 17.0 21,3 21.3 24.3 23.73 28.7 30.3 31.1 33.0 32.5 36.5 32.6 36.5 36.6 39.25 39.4 40.75 41.6 42.25 43.9 42.75 48.4 43.75 50.9 43.0 34.1 42.5 36.1 42.25 39.1 41.0 39.3 41.0 64.4 37.0 68.0 33.0 70.3 30.5 73.4 26.0 77.9 17.5 80.6 11.0 83.3 3.5 86.6 - 7.5 91.1 -16.0 -17.5 93.8 -14.5 -17.75 96.0 -13.75 100. -11.5 i. . ytm Dntoe Ttoie- O Tet ­ a tr e T IO% - Tetroxide Dinitrogen System: 9. Fig. 20 d Temperatureo -IO -20 yrprn Dioxane. % 0 —9 hydropyran 50 0 4 0 3 20 oe N04 N20 % Mole 0 4 60 80 IOO 43 TABLE 2.

SYSTEM: 2 0 / 8 0 MOLE PER CENT TETRAHYDROFYRAN/DIOXANE-N 8O^ Mole % Freezing Eutectic N8Q4______point, *C. temp., *C.

0 1.0 7.2 - 1.5 10.4 - 3.5 12.6 0.75 - 4.25 18.9 15.25 23.8 22.5 26.4 25.75 29.8 29.75 33.9 33.5 38.2 36.25 41.8 38.25 4D.5 39.5 48.7 39.25 51.8 39.75 53.9 39.5 56.2 39.25 61.2 36.0 63.7 34.5 68.6 29.0 72.6 22.75 75.0 19.75 77.9 13.5 81.5 2.5 84.4 - 4.0 87.4 -16.25 -18.25 89.4 -16.2 -18.25 92.4 -15.25 -18.25 95.3 -13.0 100 -11.5 44 50

4 0 -

3 0

o 2 0 e 0>

oX— o. E

-IO

-20

20 40 60 80 IOO Mole % N 2O 4 Fig. 10. System: Dinitrogen Tetroxide — 20% Tetra- hydropyran — 80% Dioxane. 45

TABUS 3

SYSTEM: 30/70 MOLE PER CENT TETRAHYDROPYRAN/DIOXANE-N,0*

Mole % Freezing Suteotle point, *C. temp., *C. 0 - 3*25 2*8 - 4.0 10*5 - 7.75 13*0 - 2.5 - 9.0 18*7 10.75 24*3 19.0 29*5 25.5 33*8 31.25 35*6 31.5 39.9 34.5 41.8 33.75 46.8 36.75 49.3 37.0 51*4 37.0 54.1 36.25 36*8 34.5 61*3 32.0 62.8 31.5 64.6 29.5 69.2 24.0 75.0 14.5 78.9 5.5 82*4 - 4.0 85*9 -15.5 -19.0 88.1 -17.75 -19.0 90.8 -15.25 -18.75 100.0 -11.5 46 50

40

30

om 20 o \o a> c l e

- i o

-20

20 40 60 80 IOO Mole % NzO a

Fig. II. System: Oinitrogen Tetroxide — 30% Tetrahydropyron — 70% Dioxone. 47

TABLE 4

SYSTEM: 40/60 MOLE FEB CENT TETRAHYDROFYRAN/DIOXANE-N^O*

Mole Freezing Eutectic NB04 point, *C. temp,, *C.

O - 6.25 6.1 -10.75 10.9 -11.5 -14.25 12.5 - 4.75 -14.0 15.8 3.0 22.3 14.5 28.0 22.25 34.4 27.5 39.5 30.5 43.4 31.75 48.0 31.75 50.6 31.75 53.6 31.25 57.0 29.5 59.9 27.75 62.2 25.25 65.0 22.75 69.6 17.25 75.2 8.5 79.9 - 3.0 82.8 -11.75 -13.75 86.3 -17.25 88.2 -18.0 91.7 -15.0 100.0 -11.5 F i g .1 Sy 2 s t . e mD * i n i t r o g eT n e t r o x i d e — 40% Temperature °C. -20 50 -IO 40 20 30 Tetra hydropyran Dioxane % 0 6 — 0 2 46 o e N204 Mole % 40 60 80 IOO 49 TABLE 5 SYSTEM: 50/bO MOLE PER CENT TETRAHYDROFYRAN/DIOXANE-N.O. w.T . '* ------m _ 1 _ _------a-I — »— » Mole ‘jt freezing Euteetic n 2o a point, *C. temp., *C. 0 “12*5 1.4 “13.0 3.e “13.5 6.4 -14.2 10.8 -17.5 16.8 0*5 -20.0 13.9 3.0 20.2 8.5 20.5 12*0 21.0 11.0 22.4 14.6 27.4 19.2 52.5 21.0 33.6 26 .2 34.8 25.5 37.4 26.0 39.6 27.7 42.2 28.0 44.1 27.5 47.1 28.0 47.4 28.5 43.8 28*5 52.6 25.0 53 .5 27.5 56.5 25.0 61.0 20.5 64.2 20.5 66.7 14.0 74.0 2.0 74.5 5.0 76.0 - 2.5 76.3 - 2.0 80.4 - 7.0 -20, -22 32.5 -13.5 -20 86.5 -18.0 -19.8 87.7 -18.5 94.0 -13.9 -22 100.0 -11.9 so 40

3 0

20

o

a>

-20

20 4 0 6 0 8 0 IOO Mole % N2 0 4 Fig- 13. System: Dinitrogen Tetroxide — 50 % Tetrahydropyron - 50% Dioxane 51

TABLE 6

SYSTEM: 60/40 MOLE PER CENT TETRAHYDROFYRAN/DIOXANE-N^O*

Mole i > Ereezing Euteotic n 8o * point, *C. temp., *C. 0 -19.75 4.7 -21.0 8.9 -23.75 -24.5 12.1 -17.25 -24.5 14.5 -9.25 15.1 -8.0 21.1 4.5 25.0 10.0 27.7 14.0 55.2 18.5 35.1 20.25 35.2 20.0 41.2 22.5 43.8 22.75 46.8 22.75 50.5 22.25 52.1 21.5 53.9 21.0 56.2 19.5 59.9 16.5 64.8 11.5 70.6 4.0 74.6 - 4.5 78.2 -12.75 -22.25 80.0 -17.5 -22.0 82.4 -21.25 85.5 -18.5 -22.25 88.8 -16.25 92.3 -13.5 100.0 -11.5 52 40

3 0

20

o o a> o k_ a> c l E -IO

-20

- 3 0

20 4 0 6 0 8 0 IOO Mole % N2 O4

Fig. 14. System: Dinitrogen Tetroxide — 60% Tetrahydropyran — 40% Dioxane 53 TABLE 7

SYSTEM: 70/30 MOLE PER CENT TETRAHYDROFYRAN/DIOXANE-Nz04

Mole i o Freezing Eutectic N204 point, *C. temp., *C. O -27.5 4.7 -28.5 9.3 -30.5 -32.0 11.9 -26.75 -32.5 18.0 -11.0 21.4 - 2.5 24.2 + 2.0 27.4 5.5 32.6 11.75 35.9 14.0 41.0 16.0 44.1 16.5 46.4 16.25 48.9 15.75 51.3 15.25 54.7 13.75 58.5 11.25 60.9 9.0 63.7 3.75 67.2 .25 69.5 - 4.5 72.2 - 7.75 75.5 -15.5 77.3 -20.0 -24.75 80.1 -24.25 81.1 -21.75 83.8 -19.25 87.7 -16.75 92.6 -14.5 100.0 -11.5 F i g .Sy 1 s 5 t e m ;D i n i t r o g eTe n t r o x i d e — 70% Temperoture #C. -JO -20 30 -3 40 h 0 3 20 0 - 10 Tetrahydropyron — Dioxane 0% 3 20 54 oe NgO % Mole 0 6 4 0 6 IOO 55 TABLE 6

SYSTEM: 80/20 MOLE PER CENT TETRAHYDROFYRAN/DIOXANE-NeO*

Mole % Freezing Euteetie N20* point, *C. temp, *C. 0 -38.5 2.9 -39.25 4.3 -40.25 6.6 -41.0 7.3 -41.25 9.2 -42.25 10.1 -43.0 10.1 -43.25 12.2 -33.75 -48.75 16.8 -17.25 22.9 - 5.75 26.1 - 1.75 32.0 + 3.5 34.3 4.25 39.6 7.25 42.1 6.25 44.9 8.25 48.9 6.75 50.9 6.0 53.2 4.75 57.7 1.25 61.6 - 3.25 66.3 -10.25 70.4 -17.0 70.5 -16.75 -28.75 73.8 -25.25 75.4 -24.0 76.9 -22.5 79.2 -21.5 83.0 -18.75 88.8 -15.0 92.9 -13.25 100.0 -11.5 F i g .1 Sy 6 s . t e mDi ; n i t r o g e nT e t r o x i d e — 80% Temperature °C. 0 5 - -20 0 4 - 0 3 - 20 to o - o - Tetrahydropyran Dioxone.— % 0 2 20 56 oe N % Mole 0 4 0 6 2 0 4 0 8 IOO 57 T A B U 9 SYSTEM: 88/12 MOLE PEE CENT TETRAHYDROPYRAN/DXQXANE-N*O* Mole % Freezing Binary eutectic Ternary eutectle N20* point, *C. temperature, *C. temperature, #C. 0 -50.75 5.1 -55.25 5.6 -54.75 8.2 -56.5 8.3 -56.75 10.0 -56.25 -89.0 11.9 -49.5 -66.0 -87.5, -90 15.0 -37.0 -67.0 -89.0 18.5 -26.25 -87.5 22.1 -20.25 25.7 -13.0 29.2 - 9.0 33.3 - 5.25 37.5 - 3.25 39.5 - 3.0 41.1 - 2.75 43.5 - 3.5 46.4 - 3.5 47.9 - 3.75 50.3 - 5.0 53.2 - 7.25 56.8 -12.25 58.1 -13.0 60.6 -16.5 -65.75, -67.25 63.3 -21.5 66.0 -23.0 66.7 -27.5 66.6 -30.0 70.6 -29.0 71.5 -26.0 74.9 -22.5 79.9 -19.25 84.8 -17.0 89.1 -15.5 92.2 -14.25 100.0 -11.5

4 Fig. System: 17 Dinitrogen Tetroxide % 8 8 — Temperature 0 8 - 0 6 - -20 0 3 - 0 1 - Tetrahydropyron 12% — Dioxane. 20 58 oe N % Mole 4 080 8 046 0 2 0 4 IOO TABLE 10 SYSTEM: 90/10 MOLE PER CENT TETRAHYDROPYRAN/PIOXANE-NeO* Mole % Freezing Ternary/Eutectic point, *C. temperature, *C. 0 —49*5 1.8 -52.5 3.4 -58.75 3.3 -59.25 4.7 -61.75 5.2 -63.5 6.9 -64.0 7.1 -65.0 9.5 -86.0 11.2 -60.05 12.3 -58.5 -89.5 12.6 -57.0 14.6 -44.0 15.8 -41.5 -89.5 18.0 -33.75 -89.0 20.4 -28.0 23.6 -22.25 24.9 -18.75 28.3 -16.0 28.9 -14.0 30.9 -13.0 34.6 - 9.5 39.3 - 6.5 41.3 - 6.75 43.7 - 6.75 46.4 - 7.25 49.2 - 8.75 51.5 -12.25 54.7 -14.5 60.6 -21.75 64.8 -29.0 65.9 -32.25 -65 69.0 -27.5 -65 74.0 -23.75 78.3 -20.25 83.7 -17.75 88.5 -15.0 100.0 -11.5 o Fig. 18. System: Dinitrogen Tetroxide % 0 9 — Temperature -40 0 3 - -10 0 8 - 2D -2 0 6 - 0 9 - 0 7 - T e t r a h y d r o p y r alO n %D — i o x a n e . 0 2 60 Mole Mole 0 4 % g0 N 0 6 4 oo 0 8 IOO TABUS 11

SYSTEM: 92/8 MOUE PER CENT TETRAJTH)ROPYRAN/DIOXANE-Ne 04 Mole i a Freezing Ternary eutectic NcO* point, *C. temperature, *C. 0 —49 .0 5 -63.25 8.6 -76.5 10.0 -86.5 -90.0 12.1 -64.5 -88.5 to -90.5 12.5 -64.0 16.4 -48.5 -88.75 18.6 -43.0 -88.5 20.9 -35.5 25.1 -27.0 -89.0 31.0 -21.5 33.8 -16.75 36.3 -14.25 39.2 -13.0 41.6 -13.25 43.0 -13.0 45.8 -14.0 49.3 -15.75 53.5 -18.0 56.9 -21.0 -67.5 59.9 -25.5 -67.5 61.8 -27.5 -67.5 64.5 -32.25 -67.5 65.3 -34.25 -67.5 67.3 -34.25 -67.5 69.6 -31.3 71.8 -28.5 -68.0 79.5 -22.75 35.3 -17.75 90.9 -15.5 100.0 -11.5 Fig. 19. System: Dinitrogen Tetroxide — 92% Temperature 0 6 - 0 -5 0 7 - 0 8 - 0 3 - -20 -10 Tetrahydropyran Dioxane. — % 8 20 62 oe N % Mole 0 4 0 6 2 O 4 0 8 oo to 63 TABLE 12 SYSTEM: 96/4 MOLE PER CENT TETRAHYDROPYRAN/DIOXANE-NgO* Mole f > Freezing Ternary eutectic NgO^______point, *0. temperature, *C. 0 -49.25 4.3 -50.5 7,3 -74.0 10,2 -88.0 -90.5 12.4 -84.0 -90.0 13.5 -74.0 -90.0 16.6 -64.5 -92.0 16.9 -58.0 19.2 -58.0 -90.0 27.5 -42.0 30.9 -36.25 33.2 -33.5 36.4 -29.25 38.6 -29.25 40.9 -29.25 42.8 -30.5 46.6 -32.0 49.7 -33.0 -67.5 33.1 -36.0 -67.5 33.9 -39.75 -68.5 38.2 -42.5 -68.5 39.9 -44.0 -67.5 61.9 -40.5 68.9 -31.0 -67.5 100.0 -11.5 F i g .2Sy 0 s . t e mDi ; n i t r o g e nT e t r o x i d e — 96% 0 4 - o Temperature 0 9 - 0 8 - 0 6 - 50 -5 0 3 - -20 - 10 T e t r a h y d r o p y r a 4% nD — i o x a n e . 20 64 oe N % Mole 0 8 0 6 0 4 2 O 4 oo to M 6b T A B U 13 SYSTEM: 94.1/3.9 MOLE FEB CENT TETRAHYDROFYRAN/DIOXANE-N80* Mole ^ Freezing Euteotlc N804 point, *C. temp., *C. 0 -67.0 2.1 -67.0 4.1 -67.5 b.6 -67.0 6.6 -67.0 8.3 -67.2b 9.8 -61.5 11.6 -5b.b -67.0 lb.6 -46.0 -67.0

TABLE 14 SYSTEM: 91.8/8.2 MOLE PER CENT TETRAHYDR0PYRAN/DI0XANE-N80*

Mole % Freezing Eutectic N80* point, *C. temp., *C. 0 -74.0 1.0 -74.0 2.2 -74.0 3.6 -74.25 5.4 -74.5 7.1 -73.75 8.5 -66.5 -74.5 11.5 -54.5 -74.5 14.7 -46.75 F i g . 2 1Sy s . t e m :D i n i t r o g e n T e t r o x i d e— 9 4 . 1 % Temperature 0 6 - 0 8 - - 0 3 - 0 5 - 0 1 - 0 4 - “20 0 7 - T e t r a h y d r o p y r a n— IO 66 oe N % Mole 20 5.9% 0 4 0 3 2 0 4 Dioxane 0 5 A 0 5 - p F i g . 2Sy 2 . s t e mDi - n i t r o g eTe n t r o x i d e - Temperature 0 3 - 0 7 - 0 9 - 0 4 - 0 8 - -20 0 6 - T e t r a h y d r o p y r a n o o 'o 67 oe N % Mole 20 -82% 0 3 2 0 4 Dioxane. 0IO 5 0 4 91.8% 68 TABLE 15

SYSTEM; 86,6/15.4 MOLE PER CENT TETRAHYDROFYBAN/DIOXANE-Ne04 Mole % Freezing Eutectlo n«o4 point, •C. temp., #C.

0 . -82.5 0*5 -82.5 1.3 -83.0 1.9 -83.5 2*6 -84.0 3.4 -75.0 -84.0 4*7 -65.5 6*8 -58.0

TABLE 16

SYSTEM: 77.2/22.8 MOLE PER CENT TETRAHYDR0FYEAN/DI0XANE-Ne04 Mole % Freezing Euteetic N 804 point, *C. temp., *C.

0 . -67.0 0.5 -67.5 0.9 -67.5 1.3 -67.5 1.9 -67.5

2.2 -67.75 2.4 -57.0 -67.5 3.0 -50.0 3.6 -43.5 F i g .2 3Sy . s t e mDi : n i t r o g e n T e t r o x i d e — Temperature °C. 0 3 - 70 -7 0 6 - 0 9 - 0 8 - Tetrahydropyran — 2 69 oe N % Mole 4 13.4% 8 6 2 O 4 Dioxane 86.6% IO A 70

-20

- 3 0

- 4 0

o & E - 6 0 4>

- 7 0

- 8 0

2 4 6 8 O Mole % N20 4 Fig. 24. System: Dinitrogen Tetroxide — 77 2% Tetrahydropyran - 22.8% Dioxane 71 TABUS 17

SYSTEM: 6b. 1/34.9 MOLE PER CENT TETRAHYT)ROFYRAN/DIOXANE-Nb Oa Mole i(t Freezing n 2o * point, »C. 0 -57.0 1.1 -57.0 1.4 -50.5 1.6 -42.5 2.1 -36.75 2.7 -30.0

TABLE 18 SYSTEM: 54.2/45.8 MOLE PER CENT TETRAHYDR0PTRAN/DI0XANE-NB04 Mole % Freezing n £o * point, *C. 0 -64.0 0.4 -64.5 0.6 -64.5 0.9 -57.0 1.3 -43.0 1.5 -35.5 2.2 -31.0 F i g . 2 5 .S y s t eDi m : n i t r o g e nT e f r o x i d e- 6 5 . 1 % o e Temperature 0 4 - 0 3 - 0 7 - 0 6 - -to 0 5 - -20 Tetrahydropyran — 2 72 oe N % Mote 4 34.9% 6 2 O 4 Dioxane 8 to i. 6 Sse; iirgn erxd — 2% .2 4 5 — Tetroxide Dinitrogen System; 26. Fig. Temperature 0 7 - 60 -6 50 -5 40 -4 -IO erhdoya — 8% Dioxane. % .8 5 4 — Tetrahydropyran 73 oe N2 « 20 N % Mole

74 TABLE 19

MAXIMA FROM FIGURES 9 THROUGH 26

Mole *fo Maximum System temp., *C.

10/90 CgH^oO/C4HeO# 50.0 4 3 .5

2 0 /8 0 " * 5 0 .0 4 0 .0

3 0 /7 0 " ** 5 0 .0 3 7 .0

4 0 /6 0 " " 5 0 .0 3 1 .7 5

5 0 /5 0 " " 5 0 .0 2 6 .5

6 0 /4 0 * " 4 6 . 2 2 .7 5

7 0 /3 0 " * 45 . 1 6 .5

8 0 /2 0 " * 4 4 . 8 .2 5

8 8 /1 2 " " 4 2 .5 - 2 .7 5

9 0 /1 0 " " 4 2 .0 - 6 .2 5

9 2 /8 " " 4 1 .0 -1 3 .0

9 6 /4 " " 4 1 .0 -2 9 .2 5 75 TABUS 20

BINARY EUTECTIC MIXTURES Mole"* Freezing System n £ 0 * point, *C.

10/90 CgH^ o0 /C4 H e0 E-N8 0 A 1 1 . 0 - 0.5 ft M tl " ■" 89.0 -17.5 20/80 " " " 10.5 - 4.25 MW «t " " 8 8 . 0 -18.25 30/70 " " * 1 1 . 0 - 9.0 n n m " " 8 6 . 0 -19.0 40/60 " “ " 10.5 -14.0 M M N " " 83.0 -18.75 50/50 " * * 12.5 -2 0 . 0 M M It " " 84.0 - 2 0 . 0 60/40 " " " 10.5 -24.5 M M ft " " 81.5 -2 2 . 0 70/30 " *• " 1 1 . 0 -32.0 M It N * " 79.0 -24.25 80/20 " " " 1 1 . 0 -48.75 M M M " " 73.0 -28.75 8 8 / 1 2 " " * 10.5 -66.0 to -67.0 tl tt M * * 68.3 -31.0 90/10 " " " 9.5 -86.0 to -90.0 i« « tt " * 66.3 -33.0 92/8 " M 1 0 . 0 —86*5 to -90.0 « « it " * 66.5 -35.0 96/4 " * " 1 0 . 0 -88.0 to -90.0 tl M H " " 59.5 -44.0

System Mole * Freezing Dioxane point, °C. 94.1/5.9 C5Hlo0/Ne04 -C4 H 8 08 8 . 8 -67.5 91.8/8.2 « H it ? # 0 -74.5 86.6/13.4 " it tt 2 . 8 -84.0 77.2/22.8 " " 1.9 -67.5 65.1/34.9 « “ " 1 . 2 -57.0 54.2/43.8 it tt tt .7 -64.5 27. Projection of the Diagram on the Base. System: Dioxane - Tetrahydropyran - Dinitrogen Tetroxide. T itrahydropyran r<3 ->8 78 e. Discussion of Results and Conclusions.

A model of the composition-temperature phase dia­ gram has been constructed. In the sketch of the model (Figure 28), ABCET, DFBA. FIGC. INHG. and CGHE represent tne surfaces of primary crystallization for tetrahydropyran dioxane, dinitrogen tetroxide-dioxane, dinitrogen tetroxide and dinitrogen tetroxide-2 tetrahydropyran, respectively. Curves AB, FB, BC, CE. CG, HG, and GI represent the curves of binary eutectic crystallization for mixtures of tetra­ hydropyran and dioxane, dioxane and dinitrogen tetroxide* dioxane, tetrahydropyran and dinitrogen tetroxide•dioxane, tetrahydropyran and dinitrogen tetroxide*2 tetrahydropyran, dinitrogen tetroxide-dioxane and dinitrogen tetroxide*

2 tetrahydropyran, dinitrogen tetroxide*2 tetrahydropyran and dinitrogen tetroxide, and dinitrogen tetroxide and dinitrogen tetroxide•dioxane respectively. Points B, C 9 and G represent the ternary eutectic mixtures of dioxane, tetrahydropyran, and dinitrogen tetroxide•dioxane, of tetrahydropyran, dinltrogen tetroxide*2 tetrahydropyran, and dinitrogen tetroxide-dioxane, and of dinitrogen tetrox- ide*dioxane, dinltrogen tetroxide-2 tetrahydropyran, and dinitrogen tetroxide respectively. The compositions and freezing points of the ternary eutectic mixtures are listed in Table 21. 79

TABLE 21 TERNARY EUTECTIC MIXTURES

Mole % Mole % Mole Freezing point, Dioxane Tetrahydropyran Dinitrogen • C. Tetroxide

b 8b io -87.5 to -90.0 3 87 10 -67.5 to -90.0 approximately 0.2b 47.7b 52 -65.0 to -67.0

The isolated points in Figures 17, 18, 19 and 20 represent the freezing points of the ternary eutectic mixtures• The projection of the space diagram on the base is shown in Figure 27, The letters have the same significance as those in Figure 28, When combined to form the ternary phase diagram, the data appear to be consistent. There are no noticeable dis­ continuities in the surfaces, and the binary eutectic points lie on quite smooth curves. In the cases where two intersecting systems were studied, the two curves inter­ sect within the limits of experimental error. Mixtures very rich in tetrahydropyran reacted slightly with the dinitrogen tetroxide to produce a light 80 green solution. The color is probably caused by the formation or a small amount of dinitrogen trioxide by an oxidation-reduction reaction. This oxidation was kept to a minimum by taking all points on a given sample within a short period of time, and by not warming the solids any more than necessary to melt them. Fortunately, the melting points in this section are the lowest ones on the diagram. As a result of this oxidation reaction, the freezing points may be less accurate for approximately the portions from 0 to lb per cent dinitrogen tetroxide of the following systems: 88/12, 90/10, 92/8, and 96/4 mole per cent tetrahydropyran/dioxane-dinitrogen tetroxide. The same inaccuracies occur in the systems: 94.1/0.9, 91.8/8.2, and 86.6/13.4 mole per cent tetra- hydropyran/dinltrogen tetroxide-dioxane. Let us consider the effect of these small temper­ ature inaccuracies upon the corresponding binary eutectics in Figures 1? through 23, upon the curves FB, BC, AB, and CO (Figure 27) and upon the ternary eutectic points B and Since the mixtures differ so little in composition, the freezing points just before and just after the binary eutectic point in any given curve (Figures 17 through 23) probably are affected to about the same degree. The major effect would be to shift the curve a little on the temperature scale, without appreciably affecting the 81 composition of the intersdctioa. Since the curves have such steep slopes, differences in temperatures would not appreciably affect the indicated compositions of the binary eutectics. Thus, the compositions represented by curves FB, B C , AB, and CG would not be inaccurate, and the compositions given by their intersections B and £ would not be seriously affected. Tables 20 and 21 contain a temperature range rather than one definite freezing point for some eutectic mixtures. In such cases, numerous cooling curves for the same sample at different cooling rates indicated different temperatures within the given range. It is believed that the inconsistency was caused by supercooling and by in­ efficient stirring due to high viscosity. In the region of points B and £ in Figure 28, obtaining an accurate eutectic temperature was rurtner complicated by the similarity in the compositions and freezing points of the two ternary eutectic mixtures and the binary eutectic mixtures in that region. In some systems, it is possible to obtain an accurate freezing point for a eutectic mixture by plotting the temperature-composition diagram and obtaining the intersection of the curves of primary crystallization of the two components of the mixture. The steep slopes of the curves in Figures 17 through 23 make It Impossible

4 82 to obtain accurate temperature readings for the inter­ sections by this method.

Figures 27 and £ 8 indicate that there is no ternary compound formation. The existence of a ternary compound would have been indicated by a peals or slope change on the surface of primary crystallization. So many points on the surface were determined tnat it is extremely unlikely that such a peak or slope change and an additional curve of binary eutectic crystallization could have been missed. If a peak were to exist between the points determined, the compound so indicated would have too small a surface of primary crystallization to be important. Figures 27 and 28 also indicate that the field of primary crystallization for the compound dinitrogen tetroxide*dioxane is by far the largest on the diagram. That of dinitrogen tetroxide-2 tetrahydropyran is extremely small• The literature contains no information on the relative basicities of tetrahydropyran and dioxane with respect to a common acid other than dinitrogen tetroxide. However, the following racts may be considered: (1) The molecular structures of 1,4-dioxane and tetrahydropyran differ only by the substitution of one

oxygen atom (16 atomic weight units) for one CH2 group (14 atomic weight units). 83

(2) It is believed that the two oxygen atoms in the 1,4-dioxane molecule are separated too far to very much affect each other in a saturated carbon ring.

(3) Infrared absorption spectra2 5 and dipole moment data2 6 indicate that both ethers are non planar. Considering the similarity of their molecular structures, one would not expect the ethers to differ mucn in their tendency to form addition compounds with dinltrogen tetroxide, unless the structures of the two compounds differ greatly. Thus, it is believed that the structures of the two compounds are considerably different, that of the dioxane compound yielding a higher stability. The temperature-composition curves for the binary etner-dinitrogen tetroxide systems are plotted on two faces of the prism (Figure 27). They indicate that the only tetrahydropyran-dinitrogen tetroxide addition com­ pound formed has an ether/dinitrogen tetroxide mole ratio of 2/1. Since the molecular structure of 1,4-dioxane is so similar to that of tetrahydropyran, It should also be sterically possible to coordinate two dioxane molecules about one dinitrogen tetroxide molecule. However, 1,4- dioxane does not form such a compound. Its temperature- composition curve indicates only that dioxane forms a stable compound with the empirical formula NgO^•0(CHgCHg)gO.

4 84

Viscosity and cryoscopic molecular weight determinations by Ling and Sisler1 7 indicate that the 1,4-dioxane compound is not polymeric at temperatures just above its melting point. Any structure proposed for tne compound dinitrogen tetroxide-1,4-dioxane skould be compatible with tne tacts previously mentioned: (1) Whereas 1-4-dioxane forms a stable dinitrogen tetroxide addition compound with the

ether/dinitrogen tetroxide mole ratio 1 /1 , tetrahydropyran does not form such a compound. (2) Whereas tetrahydro­ pyran forms a stable dinitrogen tetroxide addition compound with the etner/dinitrogen tetroxide mole ratio 2/1, 1,4- dioxane does not form such a compound. (3) No stable ternary compound was formed at the conditions of this ex­ periment. It is believed that the formation of the compound dinitrogen tetroxide•1,4-dioxane is a typical reaction between a Lewis acid, dinitrogen tetroxide, and a Lewis base, 1,4-dioxane. Each of tne oxygen atoms in the 1,4- dioxane molecule shares one of its unbonded pairs or "lone pairs" of electrons wxtn one of the nitrogen atoms of the dinitrogen tetroxide molecule. It is also possible that the bond may be partially or wholly an electrostatic bond between the positive center of the dinitrogen

a tetroxide molecule and the negative centers of tne ether oxygen atoms with their unshared electrons. Even if the bonds were to have considerable electrostatic character, the following discussion of configuration may be important in bringing the positive and negative centers as close together as possible. If the 1,4-dioxane molecule exists in the "boat” form, a lateral projection of the carbon-oxygen skeleton will be approximately as shown below.

• 2 superimposed carbon atoms

0 - 1 oxygen atom

o • 1 nitrogen atom

© 2 superimposed oxygen atoms

According to quantum theory studies, tne two "lone pairs" of electrons on each of the oxygen atoms in the dioxane molecule occupy directed orbitals represented 86 by a, a* , b., b_* , which will make approximately tetrahedral angles with the two bonding orbitals on each oxygen. The "erfactive" oxygen to oxygen distance, i.e., the distance between the orbitals a and a*, and the orientation or the orbitals is such that it appears reasonable that the orbitals a and a* can form bonds with two "pi" orbitals of the nitrogen atoms in the dinltro'gen tetroxide molecule. Any discrepancy in N-N and 0-0 distances can be reduced and the "overlap" can be increased if some rehybridization occurs at the nitrogen atoms as is indicated below.

However, in the "chair" form of 1,4-dioxane the

0 - 0 distance is so large and the orientation of the orbitals is such that it is difficult to see how the "chair" form could act as a bldentate coordinating group with dinitrogen tetroxide. It therefore appears that the 1,4-dioxane in the compound must exist in the "boat" form. However, electron diffraction data^^* 20, 29 indicate that isolated

A 87

1 ,4 -dioxane exists in the "chair" form. If it is assumed that the difference in energy in the "chair" and "boat" forms and the activation energy for the conversion are of the same order of magnitude for 1,4-dioxane as for cyclo- hexane, it seems very probable that during the formation of the addition compound the "chair" form is converted to the "boat" form. The lack of a stable 1/1 tetrahydropyran-dinitrogen tetroxide compound may be due to the fact that tetrahydro­ pyran can not act as a bidentate molecule and form such a bicyclic structure. Since the two pairs of electrons from the two oxygen atoms fill all the available bonding orbitals on the nitrogen atoms, a second ether molecule can not be coordinated with a molecule of this bicyclic structure. The fact that neither a stable dinitrogen tetroxide-2 dioxane compound nor a stable ternary compound was formed at the conditions of this experiment may be due to the existence of a stable bicyclic structure for the compound dinitrogen tetroxide-dioxane. When this research was begun, it seemed possible that the compound with the empirical formula

N 8 CV 0

A 8 6 molecules might replace some dioxane molecules and stop the chains. Thus one might obtain ternary compounds. The lack of ternary compound formation is compatible with results of recent viscosity and molecular weight determinations^-7 showing that the compound is not poly­ meric at temperatures just above its melting point.

f• Summary•

The composition-temperature phase diagram was constructed for the ternary system, dioxane-tetrahydro- pyran-dinitrogen tetroxide. It is a right triangular prism with an equilateral triangle as the base. (Figure 28). The concentration is plotted on the triangular base, and the temperature is plotted perpendicularly to the plane of the triangle. ABCET, DFBA, FIGCB, IKHCr, and CGHE represent the surfaces of primary crystallization for tetrahydropyran, dioxane, dinitrogen tetroxide-dioxane, dinitrogen tetroxide, and dinitrogen tetroxide-2 tetra­ hydropyran, respectively. Curves AB, FB, BC, CE, CQ, HQ. and G1 represent the curves of binary eutectic crystal­ lization for mixtures of tetrahydropyran and dioxane, dioxane and dinitrogen tetroxide•dioxane, tetrahydropyran and dinitrogen tetroxide-dioxane, tetrahydropyran and dinitrogen tetroxide-2 tetrahydropyran, dinitrogen tetrox- 89

Id© 'dioxane and dinitrogen tetroxide *2 tetrahydropyran,

dinitrogen tetroxide*2 tetrahydropyran and dinitrogen tetroxide, and dinitrogen tetroxide and dinitrogen tetroxide-dioxane respectively. Points B, C! and G represent ternary mixtures of tetrahydropyran, dioxane and dinitrogen tetroxide•dioiane,

of tetrahydropyran, dinitrogen tetroxide*2 tetrahydropyran, and dinitrogen tetroxide-dioxane, and of dinitrogen

tetroxide-dioxane, dinitrogen tetroxide-2 tetrahydropyran, and dinitrogen tetroxide, respectively. The composition and freezing points of these ternary eutectic mixtures are listed in Table 21. The projection of the phase diagram on the triangular base was drawn. (Figure 27). Figures 27 and 26 indicate that (1) There is no peak nor any slope change characteristic of a compound with an incongruent melting point. (2) The surface of primary crystallization for dinitrogen tetroxide•dioxane is by far the largest on the diagram, while that for dinitrogen tetroxide-2 tetrahydropyran is extremely small. (3) There is a maximum at approximately 33 mole per cent dinitrogen tetroxide in the curve for the system tetra- hydropyran-dinitrogen tetroxide and a maximum at bO mole per cent dinitrogen tetroxide in the curve for the system dioxane-dinitrogen tetroxide. 90

Those characteristics indicate the following facte: (1) There is no ternary compound stable at the conditions of this experiment, unless its surface of primary crystal­ lization is so small that it is unimportant. (2) Over a wide concentration range, the tendency of liquid dioxane to form a solid dinitrogen tetroxide addition compound is much greater than that of liquid tetrahydropyran. (3) At the conditions of these experiments there are compounds with the empirical formulas, Ns0^-2CBHlo0 and

* 0 ( CH g CHg ) g 0 • It w&i s no tod t t v 13 o o 9 one! molecular weight determinations17 indicate that the latter compound exists as a monomer at temperatures just above its melting point. It is believed that each of the oxygen atoms in the 1,4-dioxane molecule shares one of its "lone pairs" of electrons with one of the nitrogen atoms of the dinitrogen tetroxide molecule. If the 1,4-dioxane molecule exists In the "boat" form, the "effective" oxygen to oxygen distance and the orientation of the orbitals a and a* occupied by a "lone pair" of electrons on each of the ether oxygen atoms Is such that it appears reasonable that the orbitals a and a/ can form bonds with two "pi" orbitals of the nitrogen atoms in the dinitrogen tetroxide molecule. If some re­ hybridisation occurs at the nitrogen atoms, the "effective" 0-0 and N-N distances and the orientations of the bonding 91 orbitals are even more favorable for the formation of such a bicyclic structure. It is, however, difficult to see how the "chair** form of the 1,4-dioxane molecule could form such a structure with dinitrogen tetroxide. Although electron diffraction data27* 28» 29 indicate that isolated 1,4-dioxane exists in the "chair** form, it seems probable that the activation energy for the conversion is of about tne same magnitude for 1,4-dioxane as for cyclohexane, and that tne "chair" form of the isolated 1,4-dioxane molecule is converted to the "boat" form during compound formation. It was suggested that the fact that tetrahydro­ pyran does not form a stable addition compound with a tetrahydropyran/dinitrogen tetroxide mole ratio of l/l may be due to the fact tnat tetrahydropyran cannot act as a bidentate group and form such a bicyclic structure with dinitrogen tetroxide. It also was suggested that the fact that neither a stable dinitrogen tetroxide-2 dioxane nor a stable ternary compound was formed at the conditions of this experiment may be due to the existence of a stable bicyclic structure for the compound ainitrogen tetroxide- dioxane. In such a structure there are no bonding orbitals on the nitrogen atoms available for coordinating a second ether molecule. 92 BIBLIOGRAPHY

1. Rubin, B. f Slsler, H. H. , and Shechter, H . , J. Am. Chem. SOC. 74, 877 (1952). 2. Levy, N., and Scaife, C. W., J. Chem. Soc. 1093 (1946). S. Ingold, C. K . , and Ingold, E. H . , Nature 159, 743 (1947). 4. Addison, C. C . , and Thompson, R . , £. Chem. Soc. 5211 (1949). 5. Sutherland, G. B. B* M . , Proc. Roy. Soc. (London) A141, 342 (1933). 6. Longuet-Higgins, H. C., Nature 153, 408 (1944).

7. Giauque, W. F . , and Kemp, J. D . , £. Chem. Phys. 6 , 40 (1938). 8. Bernstein, H. J., and Burns, W. G., Nature 166, 1039 (1950). ». Broadley, J. S., and Robertson, J. N . , Nature 164, 915 (1949). 10. Spath, E., Monatsh. 33, 853 (1912). 1 1 . Pascal, P., and Garnier, P., Compt. rend. 176, 450 (1923). 1 2 . Reihlen, H. , and Ha ice, A., Ann. 452, 47 (1927). 13. Lukin, A.M., and Dachevskaya, L.D., Compt. rend, acad. sci. U-R.S.S. 55, 825 (1947). 14. Addison, C.C., Conduit, C. P., and Thompson, R., J. Chem. Soc. 1303 (1951). 93 lb. Crowder, J. A., U. S. 2, 402, 31b, June 18, 1946. 16. Brown, H. C. , and Adams, R. M. , _J. Am. Chem. Soc. 64, 2bb7 (1942). 17. Ling, H. W. , and Sisler, H. H. , J. Am. chem. Soc. 7b, bl91 (1953). 16. Whanger, F. G. , and Sisler, H. H., Am. Chem. Soc.

7b, bl86 (1953). 19. Sisler, H. H. , Batey, H. H . , Pfahler, B., and Mattair, R-» J* Am. Chem. Soc. 70, 3821 (1948). SO. Sisler, H. H . , Schilling, E. E., and Groves, W. 0., J. Am. Chem. Soc. 73, 426 (1951). 2 1 . Masing, G. , Ternary Systems. Reinhold Publishing Corporation, New York, 1944, p. 9. 2 8 . Feiser, L. F . , Experiments in Organic Chemistry. D. C. Heath and Company, New York, 1941, Part II, p. 368.

2b. Lange, N. A., Handbook of Chemistry. Handbook Publishers Incorporated, Sandusky, Ohio, 7th Edition, 1949, p. 467. 24. Hodgman, C. D., Handbook of Chemistry and Physics. Chemical Rubber Publishing Company, Cleveland, Ohio, 34th Edition, 1952, p. 867.

2b. Burket, S., and Badger, R., J_. Am. Chem. Soc. 72, 4397 (1950). 26. Robles, H. D., Rec. trav. chim. 59, 184 (1940). 94

27. Broclcway, I.. 0. , Rer. Modern Phys. 8, 231 (1936). 28. Haasel, 0., and Vlervoll, H., Acta. chlr. Scand. 1^ 149 (1947). 29. Shand, California Institute of Technology unpublished data. 95

AUTOBIOGRAPHY

I, Betty Jane Gibbins, was born in Canton, Ohio, May 7, 1923. I received my secondary school education in the public schools of the city of Canton, Ohio. My undergraduate training was obtained at Mount Union College, Alliance, Ohio, from which I received the degree Bachelor of Science in June, 1945. In September, 1946, I accepted a university scholarship at The Ohio State University. In December, 1947, I received the degree Master of Science, From February, 1948 to June, 1951, I served as an instructor in the Department of Chemistry of the College of Wooster, Wooster, Ohio. In September, 1951, I received an appoint­ ment as an assistant in the Department of Chemistry at The Ohio State University. I held this position until January, 1952. At this time I accepted a position at The Ohio State University as a Research Fellow, which I held for one year and a half while completing the requirements for the degree Doctor of Philosophy.