Citation
A study of the synthesis of some aluminum and phosphorus derivatives of alkyl hydrazines

Material Information

Title:
A study of the synthesis of some aluminum and phosphorus derivatives of alkyl hydrazines
Creator:
Nielsen, Robert Peter, 1937-
Place of Publication:
Gainesville
Publisher:
[s.n.]
Publication Date:
Copyright Date:
1962
Language:
English
Physical Description:
vii, 153 l. : illus. (part fold.) ; 28 cm.

Subjects

Subjects / Keywords:
Atoms ( jstor )
Chlorides ( jstor )
Hydrazines ( jstor )
Hydrolysis ( jstor )
Infrared spectrum ( jstor )
Nitrogen ( jstor )
Oxides ( jstor )
Phosphorus ( jstor )
Solvents ( jstor )
Sulfides ( jstor )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Hydrazines ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: l. 149-152.
Additional Physical Form:
Also available on World Wide Web
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030111313 ( AlephBibNum )
11025334 ( OCLC )
ACH2310 ( NOTIS )

Downloads

This item has the following downloads:


Full Text











A STUDY OF THE SYNTHESIS OF SOME

ALUMINUM AND PHOSPHORUS

DERIVATIVES OF ALKYL

HYDRAZINE










By

ROBERT PETER NIELSEN


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY











UNIVERSITY OF FLORIDA
June, 1962













ACKNOWLEDG ES


The author wishes to acknowledge the friendly cooperation of his committee and the faculty of the ChAmistry Department of the University of Florida. It is considered both an honor and an opportunity to have worked under the direction of Dr. Harry I. Sisler, who, although entrusted with the momentous task of administration of this Department, is able to give generously of his time and abilities to direct meaningful scientific inquiry. Uis friendship and wide

scope of interests have proved inspiring.

The time and effort spent by Comiander N. L. Smith (Ret. USIM) in acquainting the author with synthetic technique and phosphorus-nitrogen chemistry is sincerely appreciated. Besides offering an open door and a sympathetic ear, he provided encouragement when it was needed and advice when it was desired.

Thanks go also to Dr. Wallace Brey and Mr. Ken Lawson for nuclear magnetic resonance mesureents, and Mr. Howard Latz and Mr. Leo Pijanovaki for assistance in obtaining infrared spectra. The fellow members of the Inorganic Section deserve thanks for friendly competitive spirit and helpful discussion.

The author is indebted to the Petroleum Research Fund for a

fLinancial grant which made this work possible.













TABLE OF COUTEUS


ACMIMMEfDGlTS LIST OF TABLES LIST OF FIGURES Chapters

I THE IrNEUACTIO11 OF TWO ALK7YL HYDRAZINES WITH
TRIETHYLALURMII I

Introduction

Experimental a-nd Results

Discuzsion Conclusion

II REACTIONS OF ARYIIALOIl-OSPRITES AND DERIVATIVES
OF ARYUIIALOPHOSPI-IINES ITH SEVERAL ALML HYDRAZIIES

Introduction

Experimental and Results

Discussion Conclusion

BIBLIOGrPIY BIOGRAPHICAL SKETCH


Page ii iv vi











LIST OF TOWS


Table P"g.

1 Amine Alenes 5

2 Bia-Aine AImn 6

3 Bis-(Miid)amine Alares 6

4 Inner Complex Amine Almos 6

5 ninoalmae 7

6 Fusmedd Nmbwyrk Aluminum-Nitrogen Polymers 8

7 1tdreeine Alanss 8

8 Hydruainoelenes 8

9 1Wdrazilane-Hydvum e Adducts 9

10 Polymeric SPOc14 9

11 Alkyl mid Aryl ~osphorohydracirk.aw 39

12 Alkyl md Aryl Pkeo orekrnidothiiptoz 40

13 Alkyl md Ayl PiosphorophanylhydrLmidothioates 42

14 Ph*Wlhydrazidoalkyl po phonates 43

15 Bis (hdrazino) plIny lplIia OKides 44

16 Ary pheoph rod, dwtwidehlee.es 45

17 Tr"(hydrazino)p pbiir 0widac and Sulfides 47

18 Arylphosphorochl oidolidrezidothi seer 48

19 Al1qylphoGeroi-ydwwodoic Aci4s 49

20 ryl pbe. phoroaidehydrzidothiestes 50

21 Amino-bi (hydr-azino)phosphiR, OtidM 54

22 ni~no-bis(hydraeino)phoahaine Sulfide 55

23 Salts of Ph whydEat-*s 56








LIST OF TABLES Continued


Table Page

24 Benzylidene Derivatives of Aikylphosphorohydrazidates 58

25 Hydrazones of Bis(hydrazino)pHcnylphosphine
Orides-lHydrazones of Bis(hydrazino)phenylphosphine
Sulfides 59

26 Principal Infrared Asorption Frequencies for
Some Phosphorus Hydrazine Derivatives 124

27 Infrared Abeorption Frevuencies and Assignments 125

28 Nuclear IaGnetic Resonance Data Srmnaiy 128

29 Shale of Chemical Shift Values from N. 1. R.
Data 133












LIST OF FIGURES


Figure Page

1 Mini-Lab Distillation Apparatus 12

2 Reaction Apparatus 13

3 Molecular Weight Apparatus (Moadified for AtmosphereSensitive Compounds) 14

4 Infrared Spectrum of 2,2-Dimethylhydrainodiethylalone (Melt) 18

5 Infrared Spectrum of 2,2-Dimethyihydrazinodiethylalone (Dilute Iloerane Solution) 19

6 Infrared Spectrum of Condensation Product,
(EtAlNk*B2)x (Heicane Solution) 22

7 Mini-Lab Reaction Apparatus 63

8 Dry Box (Three-eighths Inch Lucite Construction) 64

9 Infrared Spectrum of 2,2-Dimethylhydrazzinodiphenylphosphine (Nujol Mall) 67

10 Proton Nuclear NAgnatic Resonnece Spectrum of 2,2Dimethylhydrazinodiphenylphosphine 68

11 Infrared Spectrum of 2,2-DizEthylhydrazinodiphenylphosphine Oxide (ujol Mull) 71

12 Infrared Spectrum of 2,2-Dimethylhydrazinodiphenylphosphine Sulfide (Nujol Mull) 73

13 Infrared Spectrum of 2,2-Dimethylhydrazinomethyldiphenylphosphonium Iodide (Nujol Mull) 75

14 Infrared Spectrum of l,1-Bis(diphenylphosphino)2,2-dimathylhydrazine (Nujol Mull) 81

15 Infrared Spectrum of the Product of the Reaction of
2,2-Dimethylhydrazinodiphenylphosphine and Carbon
Disulfide (Nujol Mull) 85

16 Infrared Spectrum of Diplhenylphosphine (Cell) 90








LIST OF FIGURES Continued


Figure Page

17 Infrared Spectrum of 1,l,2-Tris(diphenylphosphino)methyihydrazine (Nujol Mull) 93

18 Infrared Spectrum of 1,2,2-Trimethylhydrazinodiphenylphosphino Oxide (Nujol Mull) 96

19 Infrared Spectrum of l-Ethyl-2,2-dimethylhydrnzinodiphenylphosphina Sulfide (Nujol Mull) 101

20 Infrared Spectrum of l-Ethyl-2,2-dinthrylhydrazinodihe.ylphosphine Oxide (ujol 1ull) 104

21 Infrared Spectrum of 1,2-Bis(diphenylphosphino)hyd.-zine (Rujol Mull) 107

22 Infrared Spectrum of Bis (2,2-dimethylhydrazino)phenylphosphine (Nujol h-uall) 110

23 Infrared Spectrum of Bis(2,2-dimethylhydrazino)phenylphozphine Oxide (Hujol Null) 113

24 Infrared Spectrum of Bis (2,2-dinethylhydrazino)phenylphosphine Sulfide (Nujol Mull) 116

25 Infrared Spectrum of Phosphoryl Tri(2,2-dirtethylhydrnzide) (,Njol Mull) 119

26 Infrared Spebtrum of Thiophosphoryl Tri (2,2dimethylhydrazide) (Nujol Mull) 121













CIIt PrER I

TIM INTERACTION OF MO ALKYL ILYDPAZIIES WITH TRIETHILALUIM

Introduction


The highly interesting and challenging field of organometallic

chemistry has undergone tremendous growth in recent years, and among other valuable contributions to the science of chemistry there have been discovered new modes of chemical combination, entirely new types of chemical compounds, new useful materials at lower costs than previously-used synthetics, and new routes to lmoun materials.

Evidence of great interest in the field of organometallics

manifests itself in the recent rppearance of several texts devoted to

this unique combination of organic and inorganic chemistry ( 1-4 ).

One of the areas of organometallic cheraistry which has interested investigators for as long as metal alkyls have been known is the preparation and study of molecular addition complexes between the metal alkyls and various Leuis bases (electron pair donor

compounds). The metals in which most interest has been shown are those of Group III of the Periodic System. Review articles covering various such molecular addition complexes have appeared over the years (5-1 ), but none of these articles is concerned exclusively with nitrogen complexes of aluminum, although nltninun-nitrogcn complexes have been mentioned (1.








The reason that the many nitrogen-base complexes of aluminum alkyls and the various other almnes have not been the subject of an extended review article is that work on this group of compounds was begun in earnest only within the past few years. The importance of such compounds at this point seems to lie in their intrinsic properties as well as the fact that they are intermediates in condensation reactions. The aluminum-nitrogen co pounds in which this great interest has been demonstrated may someday find application as high temperature polymeric materials and perhaps as propulsion fuels. As intermediates to new and exotic compounds they are virtually unexplored.

Well-established chemical end physical evidence concerning the nature of various alanes, and specifically the aluminum alkyls, indicate that they act as electron pair acceptors, or Lewis acids. The acceptor tendency is so great as to preclude their existence in the unassociated state (1,2). This fact provides a basis for the observed reluctance of investigators to use the Grignard method for preparing aluminum alkyls, since the compounds prepared can with only very great effort be freed of the ether used in the synthesis (3. The diethyl etherate

of triethylaluminum has been studied in detail ( 13-15 ). The dimethyl etherate of trimethylaluminum has been found to be an extremely stable compound, so stable in fact that no dissociation data can be obtained in the gas phase (6; ruch data have been obtained for many other Group III addition complexes, however.

Various methods of preparation of molecular addition complexes of aluminum alkyls have been used, but the most common include direct combination in the vapor phase (both diluted and undiluted), direct combination in the liquid phase, and direct combination in a suitable








solvent. Some experiments have been carried out in which one (or both) of the reactants were generated in situ.

Since this work is concerned with nitrogen derivatives of aluminum it is appropriate to discuss those reactions and compounds

reported in the literature which include aluminum-nitrogen molecular addition complexes and the aluminum-nitrogen covalent bond.

At this writing there have been reported in the literature

three main types of covalently bonded aluminum-nitrogen compounds: the mine-alanes, the aminoalanes, and polymeric materials which contain aluminum-nitrogen bonds. These three types of compounds are

intimately related and may be arranged in a reaction sequence where the initial molecular addition complex is the parent compound in a pyrolytic series, as in the following example: (17,18)

1. Adduct Formation of an Amine Alane.

A12(CH3)6 + 211H3 = 21131:AI(C'13)3 m.p. 56.7�C.

2. Condensation to an Aminoalane.

700C.
H31-:Al (ClI3)3 H 2-Al(CIH3)2 + CR4

m.p. 134.20C.

3. Further Condensation to Polymeric Material.

200�C.
Il2I-Al(CH3)2 ('aIRICH3) x + CH4

not fully

characterized

4. High Temperature Pyrolysis to Extended Polymer.

y(HhlAlCl3)x = (NAl)x + CI4







A systematic investigation of systems of this nature has recently been carried out and the scope of the reaction has been extended to include many anines and several alnes (15). Exaimples of compounds of this type (excluding those shown above) are given in Tables 1-10, pages 5-9.

Reactions of these compounds other than by hydrolysis and

condensation are largely unknom, and to date there is no mention in the literature of a useful polymer in the aluminum-nitrogen system. Various theoretical aspects of aluminum chemistry have been demonstrated, however, and no tendency for pi-bond formation between aluminum and nitrogen has been found.

Logical extensions to this work with lanes seem to lie in two directions: 1) Lewis bases can be used which contain functional groups other than and in addition to an amino group, and 2) polynuclear Lewis bases can be used.

The first of these alternatives has been explored in a very perfunctory manner and involves the reactions of alanes, with olefinic mines, ethers, and thioethers to give various ring compounds and organic derivatives (19). Obvious possible uses for these reactions a-Ld intermediates lie in the field of polymerization catalysts.

The use of polynuclear Lewis bases with alanes was, until very recently, totally une.plored. Within the last year, however, there have appeared two articles concerning the interactions of altainum alkyls with hydrazine and alkyl hydrazt-n:: (29, 21 ). It is interesting to note that the experimental results obtained by these workers are in most aspects in accord with the results we have recently obtained






TA5WE 1
Mine Alanes

Comound m.p. b.p. Raferenees

(C13)3N:A1H3 760C. 22-24
(CU3) 3N:A1D3 77-780C. 22
(C2H5) 311: A1113 1s-190C. 22
(C3117)3N; AlH3 80-810C. ---- 22
I3CN (C2115) 2: All3 ---- 22
"2C=CIICU2N (CH3)2: AlH3 120C. ---- 22
(CH3) 31: A1 (CH3) H2 -350C. 250/1 m. 25
(C'13) 3N:A1 (C2s)12 ---- 390/1 mu. 25
(C13) 311:A1 (C'13) 211 420/1 m. 25
(C13)3M:AI(C2H5)2t1 -280C. 630/1 -. 25
(C13)2UH:A1 (CH3)3 510C. 186C. 16
(C113) 311:A1 (CH3) 3 1050C. 1770C. 16
(CIt3) 311:Al (CI23) C 1 124�C. ---- 16
(CI13) 3N:Al (C25) 2c1 -50C. 17
H3CNIt2:A1(C215)2C1 -11.50C. 17
H3CIIH2:AI(C2H5)C12 150C. 26

H3N:Al( / CI )3 --- 26


l1311:A1(- \- )3 ---- 26








TABLE 2

Bis-Amine Alanes


Compound m.p. b.p. References


[(CH3)3N]2:AlH3 950C. d. ---- 23,27,28

I (C2Is5)3N]2:AlH3 ---- 28

E (C2H5) 2NCU3]2 :AlH3 ---- 28
[ (C 3H7) 3N ]2 : AH3 ------- 28

[CH2=CHCH2N(CH3)2]2 :A1113 ---- ---- 28




TABLE 3

Bis- (Mixed)amine Alanes


Compound m.p. b.p. References


(CH3)311(C3H7)3N:A1H3 ---- 28

(CH3)3N(C2H5)3U:AIH3 ---- --- 28

H3C (C2H5)2N(H3C)3N:A1H3 .---- 28




TABLE 4

Inner Complex Amine Alanes


Compound m.p. b.p. References


/ CH2- CH2\

(A1C2)2N --C2 115-116�C./1.5 ai. 19

Al
i (C4H9) i- (C4R9)








TABLE 5 Aminoalanes Compound M.p. b.p. References


[ (cI3)2N-A (CR3)2] 2
((CH3)2-A1H2] 3 [(C215)2N-A1H2 ]2

([(CH3)2 C112N-A1H'32.16 [cl3lTH-A1 (C219) Cl]

[ (c2u5)2N-Al (C2H5)2]n [(CH3) 2NA1 (C215)C1J n

[ (CH3)UII-AIC123n

[ (Cl3) 2UAIC12] 2.4 BI.S

{[ (CH3) 2N ]2AIHJ2.5 11 (CU3) 2I]2A1ClJ2 [[(Cu3) 2C1l 12VAI I (CH3) 2]2 TRIS

[(CH3)2N3A1 32.2 {[(CH3) 2CH]2N} 3A1


154-1560C. 98- 900C.

420C. 130-1310C.

910C.



80- 81�C. 90-1040C.


141-1450C./14 m.


630C. 55- 570C.





87- 890C. 58- 590C.


16 22 22 29

17 19 17 17 17,29



24,29

29 29



24,29

29








TABLE 6

Extended Iletork Aluminum-Nitrogen Polymers


Compound m.p. b.p. References


[(cH3)NAIC] n. ---- 17

(AIN), above 2200C. 18,30






TABLE 7

Hydrazi no Almes


Compound m.p. b.p. References


(CH3)2N-N(CH3)2:AI(CH3)3 80 -83 �C. ---- 21

(CH3) 211-NI-IC3: A1 (C13) 3 65.5-66.0�C. ---- 21






TABLE, 8

Hydrazinoalanes


Compound m.p. b.p. References


[(CH3)2N- *I-Al (CH3)212 77.0-78.50C. 21

[(CH3) 211-1 (CeU3) -Al (CI13) 212 125.0-126.5�C. 21

L(CH3) 2AI-NH"'U-A' (CfI3) 21 shock sensitive ---- 20








TABLE 9

Hydra inoa l e-Wdrazine Adducts Compound M.p. b.p. References


H2-"(CH3)2: (CH3)2A1-NH-N (CC3) 81.0-R.0Cc ---. 21








TABLE 10

Polymeric Species


Compound M.p. b.p. References


(I{3CANI!M3H3)n 2000C. d. ---- 21

[Al (NC3-'ICH3) 3] n -- 21


independently at the Uiversity of Florida. been no duplication of effort since our werk triethylalumnum and the studies reTorted in with other aluminum allyl-.


Fortunately, there has

is concerned with the literature were made


The following procedure war found to be very satisfactory for

studying the intewction of triethylalninum with l,l-divethylhydrazine and wm~e thylhydracine.








Iiteripls. l,l-Dimethylhydrazine is cormrcially available and was distilled prior to using. The material used had a boiling range of 62.2-63.0�C./753 mn. Distillation was carried out over calcium hydride to remve any traces of water which the cormercial

material may have contained. Care was taken to observe completely anhydrous conditions during the transfer and handling of this and the other reagents and solvents used.

Treithylaluminum was purchased in 250 gram quantities in steel lecture bottles which were fitted with Teflon-packed needle valves. The purity of this material was not high enough to permit use without purification. Distillation at 569C./0.50 n. provided samples which analyzed 99.6 per cent triethylaluminum, as determined by measuring the ethane evolved upon aqueous hydrolysis.

Analytical Reagent grade solvents were distilled over calcium hydride prior to using end were then stored in air-tight glass containers with metal foil or polyethylene closure liners. All transfers were performed either in a dry box or by pipette with a nitrogen flush.

High purity nitrogen was used in the dry bon and wherever a constant flush seemed necessary to avoid contact with the atmosphere.

Equipment. A dry box is essential in carrying out synthetic work with triethylluminum and this piece of equipment is described in detail in Chapter I! of this dissertation, under the Equipment section.

Distillations were carried out in 1lini-Lab Standard Taper 14/20 apparatus as shown in Figure 1. The actual reactions were carried out in a Standard Taper 14/20 100 ml. round bottom flask fitted with two 10 mn. diameter by 30 am. length side arms. Figure 2








illustrates this flask and the complementary equipment, which includes a pressure-equalizing addition funnel of 50 ml. capacity and a Precision Wet Test Meter. The Wet Test 1ater'can be read directly to the nearest 3 ml.

A freezing point depression type apparatus was used for

molecular weight determinations. Its design was modified slightly in order to provide a constant, slou nitrogen flush so that contact of

the solution with atmospheric oxygen could be avoided, as shown in Figure 3.

Elemental analyses ere performed by Galbraith Nicroanalytical Laboratories, Knoxville, Tennessee. Carbon, hydrogen, and nitrogen were estimated by combustion methods and special precautions were observed to maintain sample integrity before combustion.

Aluminum analyses were performed by the method of Schwarzenbach (31) in which the aluminu-containing solution is treated with excess disodium ethylenediaminetetraacetate and the excess titrated with standard zinc sulfate solution to the pinh Eriochrome Black T end point.

The interaction of triethylalumninum with ll-dimethylhydrazine

In a typical experiment 8.0321 grams (0.0704 mole) triethylaluminum was vacuum distilled intj a 100 ml. Standard Taper 14/20 round

bottom flask fitted with two side arms and which contained a small, Teflon-encapsulated, magnetic stirring bar. All ground glass joints

were lubricated with Kel-F fluorocarbon grease, which was found to be more resistant to attach by triethylaluminum then are other, more

conventional stopcock greases, including silicone grease.







































vacuum inlet


0


Figure 1. Mini-Lab Distillation Apparatus









Figure 2. Reaction Apparatus


Laboratory Jack


C old( T ra ps


wVet -rest -Meter





14









If





























II









1F'Cfe it, oi . t cp. 10 .pp~L at- .soI , ,t P-",h~i

([or a ati, dr, it.Ao~, 1) Moleiwlav Wi_,t Apoa,'att. Mo ioird bo Ai, oA picre-i> n2t1e Co- poif,








The triethylaluminum was frozen by im-ersing the lower half of the flask in a dry ice-acetone slurry. While a nitrogen flush was maintained above and around the flask, the stoppers which capped the

side arms during distillation were removed and through one side arm was inserted a low range (to -1000C.) pentane thermometer. The other side arm was connected via two dry ice-acetone traps to the Wet Test

Hater.

The addition funnel, containing 4.44 grams (0.0739 mole)

1,1-dimethylhydrazine, was fitted to the flask and with the lower half of the flask at -780C. the Wet Test Meter was set to zero.

The contents of the addition funnel was added in very small increments (one drop or less) to the frozen triethylaluminum, and the system was warmed end melted for mixing between additions. Ethane (identified by vapor phase chromatography) was the only gaseous product and the evolution was slow and controllable.

Evolution of gas was evident at temperatures as low as -600C. for the first half of the addition, but during the second half higher temperatures (near 0�C.) were needed. No liquid or solid collected in the cold traps. The addition took four hours, after which the system was warned to 250C. and stirred overnight (12 hours) to assure complete

etha-ne evolution.

The observed, corrected volume of ethane was 1.596 liters; theory calls for 1.577 liters for the reaction

A12(C215)6 + 2(C13)21-N2 = f(CH3)2tij-1Ir-Al(C2115)2]2 + 2C2H6.

The 5 per cent excess l,l-dimethylhydrazine was removed by

pumping briefly at 450C./0.15 rmn. The resulting product was a colorless,








crystalline solid which weighed 10.11 grams (99.6 per cent yield, based on triethylaluinum used) and which melted at 43-440C. (sealed tube, uncorrected).

H-ydrolysis of a 2.7071 gram s-ample of the product gave 0.845 liter of ethane. Calculated for [(C3)2 Al(212154 2: 0.841 liter. [(CII321A(C25 + 8 O+ + 4H20 = 4C2 11 + 2(CH3)2UNU3 + 2A! (U 0)63

Analysis. Found: C, 50.05; 1, 11.61; N, 19.70; Al, 18.57;

40.73. Calcd. for [(C3)2ITNNIAI(C2H5) 2]2: C, 49.93; H, 11.38; N, 19.43; Al, 13.71; C2H5, 40.31. 1blecular weight, found: 231 (cryoscopic in benzene). Calcd. for [(CUI3)2NiHE' l(C2H5)212: 238.4.

The product is soluble in most cotnnon, inert solvents. The infrared spectra obtained on a melt and 10 per cent in n-hexane show bands in the regions expected for C-H stretch and bend, N-11, N-N, C-11-1, and N-CH3. Figures 4 and 5 shout the infrared spectra.

The observed molecular weight indicates a dimeric structure. Three possible structures, two of which are geometric isomers may be considered.
N5C2 C215
I\Z
Al

11-N (11)


(113C) 21l -H



115C2 CA


1. Si:- Nembered Ring








H5C2 C25 115C2 H1C2
11 >Al~ NI \13)1 "Al 11
S \ 1/ /1(1C
- 1 /. N 113>2 1,/



(113C)2N/ Al 11 (C113)2 Al 11 (C113)2

115C2 C2115 H509 02115


2. CiL-Four IkTmbered Ring 3. Trans-Four lumbered Ring


The proton nuclear magnetic resonance spectrum of 2,2-dimethylhydrazinodiethylalane indicates that only one molecular species is

present and shows the Al-C112-C113 and N-CH3 structure peaks. Because of the uncertainty in position associated with the 11-11 group, and in addition some overlapping of peak areas, the data obtained were insufficient to uc. in assigning a definite structure on the basis of the spectrum alone. It is felt however, that if the structure involves

the four nembcred ring, there might be expected two forms (cis- and trans-) in a mixture, and the n.m.r. spectrum would certainly indicate

this.

Infrared data give circumstratial evidence for the six membered ring. The position of the 1N-11 stretch in 2,2-dimethylhydrazinodiethylalane corresponds closely to that observed for tricovalent nitrogen

compounds which contain the IZ-11 group. The four membered ring does not contain this particular arramgement since it is the nitrogen atom involved in the 11-11 group which coordinates the aluminum atom and thus becomes tetracovalent. Little importance can be placed upon these observations, however, since hydrogen bonding in the solid sample would tend to equalize the environents of the nitrogen atoms in question.













C ).
200111 50i
- __ _ _ - --- _______


Kj
It /
j r FI ;


FiLure 4. Infrared Spectrum of 2. 2-Dimethlhydrazinodieth3l1alane (MIelt)


I )O(o I-


110


-\

1'*


\/ \, '\!'
















*11)O 30(0o 'ho



V i


1 5u OI

V -


i. i re . .u o ' 2. '_-I)L, L thvlh\,lrjtah oitthvlt]n)Inc (r Ohui ll,'xhte (hlhltiu


1 t)(1)


V









As a possible solution to the problem of analyzing the infrared data, a sample of 2,2-dimethylhydrazinodiethylalane was run as a 10 per

cent solution in n-hexane, where H bonding would be minimiized. The resulting spectrum (Figure 5) shows little, if any, shift in the N-Il stretching frequency at approximately 3150 cm.-!. A new peak was observed at 1590 cm.' I which is in the region usually associated with the N-11 deformation frequency. This observation 'ias not explained.

Support for the four-membered ring structure might lie in the observed peak at 1400 cm.-I which is generally found in compounds containing an ammonium or substituted arnonium ion. This too, however, could be a result of hydrogen bonding in either the four or six member ring structure. The prospects for elucidating the structure of this dimer b infrared meos appear rather bleak. Suggestions for further tweic along these lines might include a study of frequencies associated tith various ring sizes which contain atoms of size similar to aluminum and nitrogen, perhaps cyclic silylhydrazines.

Pyrolytic condensation of 2,2-dinethvlhydrazinodiethyla lane.

In an experiment designed to test the degree of lability of the N-H bond in 2,2-dimethylhydrasinodiethylalane, a pyrolytic condensation tas performed.

In the first experiment in this series the saiple was heated to 1950C. at which point rapid evolution of gases caused the reaction vessel to be blown apart. The product of the reaction was a light browm solid which was very brittle and apparently highly polymeric in nature. Analysis of this solid residue for carbon, hydrogen, and nitrogen gave a ratio which corresponds to C 1.113.11 1.0 which my be








formulated as N-N(CH3)2, thus indicating that even though the material was subjected to temperatures in excess of 350�C. the dinethylhydrazino group remained intact.

This material was highly opaque to infrared and had no melting point; thus, little characterization was possible.

In a second experiment the condensation was run at 1500C. over a five hour period. Ethane evolution was slow and easily controlled in the temperature range 110�C. (where evolution first is observed) to 1600C. The product in this case was a dark, viscous tar which was easily soluble in benzene and hexanc. The material was analyzed for active ethyl content by aqueous hydrolysis and gave the following results.

Found: C2H5, 25.52. Calcd. for [(CH3)2H"nAlC2H15]y 25.46.

It thus appears that the condensation occurs according to the reaction 1500
nf(CH3)2WNNHAl(C2H5)2] 2 = 2nC2H6 + x f(Cl3)2 IAiC2 11 Y and the hydrolysis may be represented by the equation


[(C113)2NNA1C2H5]y + 2yl1O + 4yH304= YC2116 +(C%)2iM5 + yAl(H0)63+


Other evidence which serves to substantiate this reaction

sequence is demonstrated by the conspicous absence of the N-l absorption bands in the infrared spectrum (Figure 6). The presence of the l,l-dimethylhydrazinium ion in the hydrolyzed sample was shown by adding aqueous sodium hydroxide until the pH was 10, and testing the vapor above the solution for a volatile free base (1,1-dimethylhydrazine) with moist red litmus, and for a reducing substance with a
















2000


K

I/
1<, 1-

I.,



I


In U
- _____ .1


F"


I


, I ! ,/
I .

I (,







i-:lr .Infr-ared Z, titnn 01 C'on, cns~i on 'r'odu( ' Ut INN\1(.,2'i (IlexNane ,SoLo ion)


E- r-








drop of KhiO4 solution on a strip of filter paper. Both tests were positive.

A third experiment was designed to carry out the condensation as slowly as possible at a lower temperature in order to see if ethane could be removed without opening the ring or changing the degree of polymerization from a dimeric species. A sample of 2,2-dimethylhydrazinodiethylalane was heated at 750C./0.20 um. for a period of 144 hours. During this time the colorless, mobile liquid gradually

became dark in color and grew increasingly viscous. When _he reaction was complete, as evidenced by the disappearance of the N-il band in the infrared, the material had become a dark, viscous tar.

A molecular weight determination was performed cryoscopically in benzene. Molecular weight found: 646. The formula weight of a

[(CII3)21wAIC2 11 unit is 114.13, thus the polymer formed in this experiment has an average degree of association, y = 5.66.

Possible structures for this polymer include 1) rings linked together, 2) rings larger than six member, and 3) chains with end groups of en undetermined nature. In any event, the polymer appears to have the repeating unit


C2'5 MY_ 2

y


The information gained in this experiment indicates that the hydrogen atom in the II-1 group in 2,2-dimethylhydrazinodiethylalane is less labile than that in l,l-dimethylhydrazine itself (or in some intermediate molecular addition complex formed prior to 2,2-dimethylhydrazinodiethylalane).









These additional data are better interpreted on the basis of a six meaibered ring rather than on a four mnebered ring. In the four membered ring the aluminum atom is coordinated by the nitrogen atom which contains the IT-H group

H

AlIlI-Al



N13C CHU3


The effect to be expected if the 11-U nitrogen atom acts as a Lewis base is labilization of the N-H bond, which is not the experimental result.

Attented dc-dinerizntion of 2,2-dimethylhydrazinodiethyl1lane by adduct formation. A sample of 2,2-dirmthylhydrazinodiethylalane was mixed with a twofold excess of unsym. -dinethylhydrazine and heated to 650C. for five hours. The product of the reaction was a light straw liquid after the excess unsv .-dimethylhydrazine had been removed by pumpingat 250C./O.20 ma. for one hour.

The desired reaction may be represented by the following equation:


[(C113)2umTLI (C21i5)2] 2 + (C'3) 2111H2 = 2 (CI3)lrIAl (C2H5)2 :N(CH3) 2 2 A possible reaction that was thought unlikely yould proceed via ethane elimination


- (2'C 1-1 + 2 (C= 2 [(C3) ]AIC211 -.- 2C 11
k(C7YinmiTkl '25212 (C3)21% ) 111M CI 2 m 2 5 2 6








Analysis of the product of the reaction disclosed, however, that the only process that had occurred was that or partial condensation to [(C"3) 2U11AIC21g] Y

Found: N, 20.05, 20.39. Calcd. for [(Cu3)2 maIAl(C2115)212: 19.43. Calcd. for [(C3)2IIAC2H5 : 24.56. A change in the peak height of the 1-H band in the infrared spectrum served to confirm the course of the reaction.

These data give an indication of the strength of the dative

Al-IT bond in the dimer and also seems to shot that l,l-dimethylhydrazine is capable of catalyzing the pyrolytic condensation.

Attempted de-dimerization using phosphines. Experiments in

which tributylphosphine and triphenylphosphine were heated with 2,2dimethylhydrazinodiethylalane resulted in dark, uncharacterizable, viscous liquids.

The interaction of triethylaluminum with methvlhydrazine

In an experiment identical in detail with that performed in the synthesis of 2,2-dimethylhydrazinodiethylalane, 1.569 grams (0.0341 mole) methylhydrazine was reacted with 3.8G7 grams (0.0341 male) of triethylaluninum. Very slow dropwrise addition was carried out at -780C. with warming for melting and mixing between additions.

It was immediately apparent that the reaction was very

exothermic (more so than with 1,1-dimethylhydrazine) and that a somewhat different product was being produced than was found with 1,1dimethylhydrazine. The amount of ethane evolved approached three times the theoretical and the reaction product was a brown, inhoxaogeneous solid.








The total volume of ethane evolved was 2.029 liters. The amount calculated for the reaction A12 (C2115) 6 + 2CHy3IlT1 = 2 [H 2(C13)A1(C21Iy)2] + 2C2I1


was 0.763 liters. For the reaction


Al2(C25)6 + 2C1%TIiIH2 = 2[AI(C13)11] + 6C25


however, the calculated volume of ethane was 2.2S3 liters. The quantity observed was 87.5 per cent of that required for the elimination of 3 moles of ethane per mole reaction unit.

The solid was collected and any excess triethylaluminum was removed by pumping at 25�C./0.20 mms. for five hours. An attempt to prepare a Nujol mull for infrared analysis resulted in spontaneous combustion of the solid upon contact with air. Whether this was due to the material itself or unremoved triethylaluminum is not 1nmoTun.

The infrared spectrum was obtained on a mull prepared in the

dry box and was very opaque, which is characteristic of highly polymeric materials. A very weak N-i1 peak showed that the reaction had not gone to completion.

Analysis. Found: Al, 37.15 per cent. Caled. for CH32AI: Al, 33.55 per cent.

The material was found to be insoluble in all the solvents

tried, and dissolved only with very great reluctance in hot 1511 nitric acid. The nitric acid solution was highly colored, but was decolorized with hydrogen peroxide in order for aluminum analysis to be run on the solution.








The solid has no melting point, but chars wzhen heated to 3600C.



Discussion


Although therewere not a large amount of data collected in our study of the interactions of l,l-dimethylhydrazine and methylhydrazine with triethylaluminum, certain interesting differences were noted in the behavior of the two different hydrazines and in the

behavior of hydrazines and amines with respect to their interactions with aluminum alkyls. One pleasing result of this work is the synthesis and characterization of the new compound, 2,2-dimethylhydrazinodiethylalame. Although a recent publication (2!) reports methylaluminum derivative of hydrazines, this compound is the sole Irnowm ethylaluminum derivative of a hydrazine.

An idea which was a motivating factor in the initial phases of this work concerned the possibility of preparing a solid adduct

between triethylaluminum and 1,1-dimethylhydrazine. Both the alane and the hydrazine are mobile liquids and both have been used as liquid propellants in the missile industry and in the U. S. space effort. For many reasons solid fuels are preferred to liquids in certain types of applications. Thus, a solid adduct formed from the two liquids might well be an interesting fuel.

From the preceding description of the observations made of

the interaction of triethylaluminum with l,l-dimethylhydrazine it can be seen that no adduct was found in the temperature range studied. It is possible, of course, that work at very low temperatures would








produce evidence indicating the formation of such an adduct, but in ony case no such compound exists above -60�C. Similar results were encountered in the reaction of triethylaluminum with monomethylhydrazine.

Of the various nitrogen donor-alane systems which are known, those systems containing a hydrazine are certainly the most reactive. CT the reported amine-alanes, where the amine may or ray not contain the I2 group, it may be generally said that the adduct formed between the amine and the alane is stable at room temperature. In the case of hydrazine-alanes, however, none have been reported hich contain the %2 group e::cept (CH3)21:.!IAI(CI3)2:NH2N(CH3)2 which is not a simple hydrazine-alane. (21)

kn approach to a clearer understanding of the nature of the

hydra zine-alanes, the hydrazinoalanes, and their condensation products may lie in a stepwise consideration of the interactions involved in the fornation of these compounds.

,Any acceptable mechanism must be consistent with the following known facts: 1) the reaction was carried out in such a way as to have an excess of trithylaluminuzm present in the reaction mixture until addition of l,l-dimethylhydrazine was essentially complete, 2) ethane evolution occurred at -600C. during the first half of the reaction, but was less rapid during the second half and required higher temperatures, -hoP., 3) the product is a dimer, and 4) the NT-I group in the product is less active than that in l,l-dinethylhydrazine.

Fetter and Bartocha (21) have recently made statements concerning proposed mechanisms of interaction and the structure of hydrazinoalanes. They believe that a plausible first step in the interaction









of 1, 1-dimethylhydrazine with an a luminum alkyl is the coordination of the -11 end of the molecule with the AIR3 unit,


113AI + 11211-1(C13)2 - R3tl:TU21(CU3)2


rather than coordination of the -14(C113)2 end, as shown below:


R3Al-:- (CI3)2111k2 = R3AI:N(CII 3)2112


The reason for their choice is the fact that their preparations of dimethylhydrazinoalanes produce as a gaseous product, r-H, rather than P-C13. From a consideration of the bond energies alone, N-1 (92 kcal./mole), and 11-C (66 kcal./mole), one might favor such an interpretation. However, if there are kinetic effects which favor a different reaction, there is no reason to dismiss coordination by the

-11(CI3)2 end of 1,1-dimethylhydrazine.

If we first accept the premise that the-](C113)2 end of the H211(C13)2 molecule is more basic than the -I112 end because of the inductive effect of the methyl groups, II 0113
\ /


/ \
11 CH3


then in the presence of the strong Lewis acid, AlEt3, we should c:pect the initial coordination as shot in equation I on page 30.

In the presence of a large excess of Lewis acid, AEt3, we ould expect a second step (equation 2) to occur very quickly, to form the bis- complex, Et3Al:1T12-1(CH3)2:AlEt3.







A Proposed lechanimn for the Interaction of Unsvarm.-Dimethylhydrazine with Triethylaluminum


1I CI3
1. : C11:
/ I
II 0113


Et
I
- Al-Et
I
Zt


H 7it3 I I

II CH3


.olculnr Addition


Ut

2. ft -Al
I
gtL


H C113
I I
4- :N- IT: A!Et3 =
I I
11 C13


Et II CiT3
I I 3
Et-Al :11-11:AIEt3
I I U
Ut 11 01-I3


loleculr Addition
of
Second A1It3


Et 11 0'13
I I I
3. Et-AlI1-i:AIEt3
I4 IT:
Ut H C13



4. :11- 1:
/ \
II e113


= C21


Et I 113%
+ t-A-11- -:AlEt3

0113


Et II 113
I I I
+ Et-Al- 1-Ih AIEt3

L1-13


Ethne Eliminotion


Et 11 C113 Ut
I I I /
E t- AI- N- Et:l- Et

(113C) 21' CI-13 Et I11.I
II


Iblecular Addition










Et Et


Al




112N: N (C'13)2

Al-Et I \Et
Et


Et Et


A


6 (113C)211 I-11
I I
IL-11 IT1 (Ci 3) 2

11 Al-Et
/ "Et
Et


EL Et






(11 I 3 C)02 1 12 1- 1I ' 1 (C '3 ) 2

H 'Al-Et
I \Et
Et


Cyclization by Intr~molecular
Addition


Et Et


Al


c 216 + (130 2N
2N


Second Ethane Elimination to

11(0il3)2 Form Six nered
32 rin , Diiner


Et Et








The decomposition of the bis- couple: to ethanc cnd the AIEt3 adduct of the hydrazinoalane, Et2, lhIE(C3: AIEt3, would follow according to equation 3. At this point the stoichiometry is a 2:1 ratio of alane to hydrazine, and it is at this point in the preparation here the tenperaturc at which ethane will evolve suddenly changes to a higher value. In fact it has been obsc-vcd in the laboratory that the remaining addition of l,l-diraethylhydrasine may be made quickly at -250C. and that no appreciable gas evolves until the mi::ture is warmed to 00C.

In the presence of additional free 1,1-dnethylhydrazine addition would be expected according to equation 4, followed by cyclization as shoin in equation 5, and ethane elimination to the final cyclic six membered ring dimer as in equation 6.

A suggestion for further work in this system appears quite obvious at this point. If the mechanism proposed above is correct there should exist the possibility of preparing Et2A -UIl(CHl3)2:AlEt3 by the reaction of the alane with the hydrazine in a 2:1 mole ratio. The prospects for preparing any of the other intermediates do not appear promising, howaoer. It is quite possible that Et2Al-iRT(C13)2:AlEt3 would not be isolated as a monomer since it could

very well associate to form diners or chain polymers; these species would probably react in much the sane manner as the monomer with 1,1dimethylhydrazine, although less vigorously. An interesting point is

that if this material did exist as a tionomer, it vould indicate that the need of aluminum for an electron pair was being satisfied by the adjacent nitrogen atom uhich has an unshared pair of electrons available.








This observation would provide evidence for the existence of pibonding between aluminum and nitrogen, which has boon discounted by Laubengayer (30).

Condenmation of 2.2-dLnethylhvdrazinodiethvlrlne. Since it has not been observed that ethane is intramolecularly eliminated, a process which would result in the formtion of [(Cu3)21uuAIC2u5J2, with a molecular weight of about 223, it appears reasonable to assume that the mechanism of the condensation involves attack of an 11-11 group on the AI(C2115)2 group in 2,2-dimethylhydrazinodiethylalane. This condition could only be fulfilled if the 1-H group has an unshared pair of electrons and cM act as a Lewis base; thus it does not appear that the four mebered ring structure, pege 17 , can participate in such a condensation reaction.

In the presence of excess l,l-dimethylhydrazine it was found that the condensation proceeds at a lower temperature than is observed when pure 2,2-dimethylhydrazinodiethylalane is condensed by heating. Thus it appears that the presence of N-11 groups (with free electron pairs) is necessary for the condensation reactxton.

These observations leave doubt that a dimeric condensation product will be observed since intramolecular elimination of ethane in the ring apparently does not occur, as evidenced by the observed high molecular weight of the condensation product.

Condensation of triethylalninum with methylhydrazine. The observations made on this system tend to corroborate the mechanism

suggested for the interaction of triethylaluminum with 1,1-dimethylhydrazine.









in methyihydrazine te should expect, from considerations of the inductive effect, that the most acidic TI-1 bond is on the

-Nil end of the molecule and that the stronger Lewis base is the

-U,(C113)Ui end.

II HI
\ /

/\
II C113


Once a methyihydrazine molecule donates an electron pair, however, there is cone question as to which 11-1 bond is most acidic.

An initial attach of methylbydrazine on triethylaltinum

would produce the molecular complex Et3Al:' 'IC3U., which could either eliminate ethane intra- or intermolecularly. The inhoogeneity of the product may serve to indicate that both these processes occur, or that at any rate, molecular speciec are produced in various degrees

of association.

Considering the abundance of active IT-H groups in the initial adduct, it is not surprising that complete condensation to ['%"'(C1'3)I]y is the final product.

Although it is not implicit in the formulation of this

polymeric species, it is expected that alirainum achieves a covalency of four, either by accepting a pair of electrons from a nearby nitrogen atom or by accepting a pair of electrons from an adjacent nitrogen atom. The fact that the polymer appears to be a highly extended netvor 7ould lend support to the former suggestion.








Conclusion


The results of our study of the interactions of triethylaluminum with two different alkyl hydrazines has resulted in the synthesis of three new species, [(CII3)2DiUi'Al(C2H5)212, [(Cd3)21;IAlC2115J, nAlN(CH3)IT1V and has given various data through observation of reactivites, stoichiometries, and reaction conditions. Some physical characterization was possible, and all available tools were employed in an attempt to elucidate the structure of 2,2-dimethylhydrazinodiethylnlcne, which is thought to have a sin membered ring structure.

A mechanism for the formation of these species consistent

with the observed data is suggested and applied to both this work and sone additional work reported in the literature.

Although no evidence for pi-bonding between aluainum and

nitrogen has been found, it is certainly ithin the realm of possibility that such bonding may exist. Preparation of the intermediate, Et2AII"I(CH3)2:AlEt3, in a monomeric form would lend more credence to

the possibility of aluminum forming pi-bonds with nitrogen.

Further study is indicated.














CIA2TER II

REACTIONS Or ARYLIALOPHOSPHITEZ AND DERIVATIVES OF
ARYLIIALOPHOSPHIHES WITH SEVERAL ALKYL HYDP&ZnIES Introduction


The field of phosphorus-nitrogen chemistry has been periodically reviewed and there are two major reference texts (32,33) and an excellent review article (34) which although not chiefly devoted to phosphorus-nitrogen chemistry certainly provide a foundation for the worker in this field. Ho up-to-date listing of phosphorus-nitrogen compounds is available, however. Particularly lacking are reference summaries in the area of the hydrazine derivatives of phosphines and

phosphorus acids.

A thorough compilation of all known hydrazine derivatives of

the phosphines and related compounds vas prepared in order to determine

the extent to hich these materials have been studied and to what use the information has been put. As a result of this survey it has been found that hydrazinophosphine, '2PI.2, has never been reported, nor have any of its organic derivatives. Most of the work uith hydrazinophosphorus compounds has been in the area of hydrazine derivatives of the esters of phosphorus acids.

There exist, on the basis of the more common substituents for phosphorus, a great variety of possible classes of compounds containing

the hydrazinophosphorus group, -P-ti-N-, but the survey shows that work








has been done on only a few of the possible series of such compounds. Many of the compounds reported in this dissertation are the only known members of their series of compounds.

The general method of preparation of hydrazine derivatives of organophosphorus compounds is analogous to the methods used for the syntheses of aminophosphines and aminophosphorus compounds (35,36 ). The hydrazine is usually dissolved in an anhydrous solvent and the resulting solution is added to a halophosphorus compound dissolved in the same solvent. Excess hydrazine ckan be used to absorb the hydrogen halide produced in the reaction, or a tertiary aine such as triethylamine or pyridine may be used for this purpose.

Example (37):


0 0
I I
(PhO)2PCl + 221k1 - (Ph0)2P"-UII{2 + (12115)Cl


The monohydrazinophosphorus compounds which have been reported in the literature are listed in Tables 11-14 along with pertinent data; bis- and tris-hydrazinophosphdrus compounds comprise Tables 15-17.

Recent work has shown that stepwise substitution of chlorine

can be obtained in some arylphosphorodichloridothioates by partial solvolysis (38):


S S CI
I 1/
PhO-PCl2 + 2214- PhO-? +~ (,'12 1y)C

IUR1112


The remaining chlorine is found to be less labile as a result of the substitution of the less electronegative hydrazino group and is








thus resistant to further solvolysis. In some cases the compounds can be mater-washed without appreciable hydrolysis. Arylphosphorochloridohydrazidothioates are listed in Table 13.

The second chlorine in these molecules can be made to undergo solvolysis at higher temperatures, and in the presence of water, form arylphosphorohydrazidothioic acids, wh-iile in the presence of amines the various arylphosphoronmidohydrazidothioates are formed (see

Tables 19 md 20).

Example:


SS I 10000. 1
PhO-P- II- 2 + 2R -2I PhOPUUII + (R' 2 IdCl
I I
Cl I12


An effect of changing the reactant ratio has been observed in

the formation (in small yield) of a cyclic compound (39):



S S 1111- +T6I S
2PhOPCI2 + 6N2H4 P P

Pho N-H-NH OTn

+ 4(112115)CI

m.p. 1830C.


A scheme devised by A. lichaelis is responsible for the

synthesis of nino-bis(hydrazino)phosphine oxides (40). He found that emidophosphonic dichlorides can be prepared by refluxing a mixture of phosphoryl chloride and a secondary amine hydrochloride until 1C1 evolution ceases:






TABLE 11
Alkyl and Aryl Phosphorohydrazidates

Product Yield M.p. Reference

0
(C2H5 O)2pITImlm n.a. 1l3-I14o 41

9
(PhO) 2 PINHNH2 n.a. 1120 42
0
(PhCI2)2tp'RNH2 95% 730 43
9
(CH2=CHCI20) 2PMU1.iPh 76% 85-870 42
0
(C'30) 2 kNNIiPh 93% 132-1340 42

CI{30 (C2II5 0) 4 h 82% 77.5-80.50 42


n.a. - not available.








TABLE 12

Alkyl and Aryl Phosphorohydrazidothioates Product Yield ra.p./b.p.


s
(PhO) 2k. MH


(1hO) (CH30) PNrMh2


I
(02N \0) (CH30) PNHN12 (cH3o / \ ) (CH3o) hkUH2 ( \0) 113o)FMm






(CH3o) P"-I2





Cl

113 C )

( / o) (CII30) TMM2

Cu3

Br

(tert.-C4O9 0) (CH30) RINMM2


m. 630


78% no distillate
at 0.01 ina. H 68% i. 1030






m. 76-770


Reference


39 44 45



45, 46


m. 85-870 46









-- 46


m. 92.5-93.50








TABLE 12 Continued


Product Yield m.p./b.p. Reference


Cl C S
( 0) (CH3o) k NL2

Cl.


( / \0) (C2H50) ItIII2 (021, -o) (C2I{5O)P(S) MIH2

C1 S
C - 0)(c250) N H2

C1


s


(I3C) 2N N '(CH3)2 --ii
Nfl 2


m. 88-88.50


b. 147-1500


M. 800


no distillate at 0.01 . It








m. 117-180








TABLE 13

Alkyl and Aryl Phosphorophenylhydrazidothioates Product Yield m.p.


S
(C21'5o) 2bNmHPh

s
(Pho) (C"3o) k*WHPh

s
(Po) (C2H50) i'TU1HPh


(02N 0) (CH30) PklH Ph


(021T N 0) (C2150) N mIIPh

Cl

(Cl o0) (c2H5o) k-URI
S


Cl
S

(Cl -/ ) (C2H50) PNIRlhPh

Cl


S


NIC
(113c)211 NH N (cI13) 2 --


68-690


42% 800


35% 570


125-1270


680




108-110�


146-147.50 47


Reference
















TABLE 14

Phenylhydrazidoall-yphosphonates Product Yield m.p. Reference


9
C2H50-P(CC13)INIHHPh -154-156.50 42


/ -&CH2/-e --- 173-1740 48,49








TABLE 15

Bis (hydrazino)phenylphosphine Ohides


Product Yield m.p. Reference



Ph?(MHIH2)2 --- 1310 50

0
Ph (NIMMh) 2 --- 1750 48
Cl'3



13C H ) 2 --- 2080 51


H3c


H3C b (NHEIl ) 2 --- 1710 49
.----------- .------------------------------------------------A Bis hydrazino)phenylphosphine Sulfide



Product Yield m.p. Reference


S
Phi(11---2) 1150 50








TABLE 16

Arylphosphorodihydrazidothioates Product Yield m.p.


PhO-P(NIP12)2

s
Po-P (HI'INHh)2











2-- -P

Cl

/ -P(NmeI1) 2

Cl

S
C1 o00-P(NUNI12)2


C1

Cl





C1
c 16 -P (HN P) 2



C 1
ci / N 0-P(NINHmPh)2


C 1


90%


87%



93.7%


950(1030)



1360 1060 1420 1760



145-1470


93.5% 152-1530








96.7% 156-1570


158-1590


Reference


43,52 53


52 45 45 38




38, 54








38, 54






38






46

TABLE 16 Continued Product Yield M.p. Reference


Cl

(H3C)3 C O- P (IWRINPh)2 92.3% 151153o 38


Cl



Cl 2 93.5% 123-1250 38,54

Cl








TABLE 17 Tri(hydrazino)phosphine Oxides and Sulfides


Product Yield M.p. Reference


OXIDES

0

P (NH2) 3 75% -- 43

0
Pi ( l--h)3 1960 55

0

0? (NI�NBH C13)3 1890 55


SULFIDES

S
i
P( NIn2) 3 unstable 44

S
1
P (N'mIalmiN 3 154' 55

S

/(ni \C CH3)3 --unstable 55








TABLE 18 Arylphosphorochloridohydrazidothioates


Product Yield m.p. Reference


S I
PhO- P-t.H(CH3)2 97.1% 38

Cl

C1 s

/ - N .-UI(C113)2 95.8% --38

C1

Cl s
I
Cl

Cl
Cl o-P-NHN(CH3)2 99.0% 74-75 38
-08
Cl Cl Cl 0-P-UNH 79(CI3)2 -- 38








Cl Cl Cl

C1





c1
cGI








TABLE 19 Alkylphosphorohydrazidoic Acids* Product Yield m.p. Reference


0

C2O-P-rn Cl3 - - 1950 (d.) 52
011-ZRl- \C1


0

C 2 H_0-P-ITHNH - Br 1- 870 52

Oil 0
oI


C21150-P-IIIHPI -- 192o (d.) 52

OH

0
I
C2o50-P-'UHW2 -- 1000 52

OH



*See TABLE 3 for examples of salts of acids of this type.








TABLE 20

A-yIphosphoroamdohydrazidothioates


Product Yield m.p. Reference


Cl




c 1



C1I
Cl I O-P(TBH2)1NHh2







ClI Cc




Cl1 ClI


C 1-/ 0-'(01=3) T (Ch1 )2




Cl

CI

Cl


86-870 137-139o


90.5% 104-1050


100%


131-132o


44% 112-113�








0 0
1 1
PC13 + (R2Ifl2)Cl - R2NPC12 + 211C1

The amino-bishydrazino)phosphine oxide is the product obtained upon hydrazinolysis of the amidophosphonic dichloride:


0 0
1 1
R2IIPCl2 - 4II2U Y"h- - - R27, IIhn-)2 + (Thl11113)Cl


Tables 21 and 22 list the reported mnino-bis(hydrazino)phosphine oxides and sulfides.

An example of another reaction which produces mixed aminohydrazino derivatives of arylphosphonic acid involves a transamination reaction (56):


0 0 l,

Ph )2 112114 - P fhP + 'r'3




Ilydrazinolysis of the phosphonitrilic chloride trimer is also known (57, 58):


11PU2 1T21{3

(C1211)3 + 1211214 - "P% + 6(1125)C1
113112 . II 1/1413 11"2 /\ N N, 12"3








Reactions

The reactions of the hydrazine derivatives of organophosphorus compounds are interesting in that they may help to determine the structure of the compounds and in many cases lead to entirely new classes of compounds. The reactions are frequently troublesome and may occur as side reactions during the preparation of the desired compound and thus Iower the yield.

The P-N bond is susceptible to hydrolysis and the degree to which this occurs depends in large part on the nature of the substituents on the phosphorus atoma and whether or not the phosphorus is in either of the oxidized states, the oxide or the sulfide. In some cases, especially when the compound contains an ester group, it is not the P-Il bond which undergoes hydrolysis initially, but the ester group (37,43):


0 0
I I
(Pho)2P1'ln.lhI - N-aO (aq.) >- Na [ PoPNI] - + PhOI0


Further hydrolysis will yield a salt of the phosphorus acid


o 0

lie Emoh- P-111u212]- NeOH (eq.) -11 a2 12'1112

0


and the free acid can be obtained by metathesis:


0 0
4 + I
Na2 [o2nnniu1'12j 2H1 -* (1 )p'UIRII + 214a








Complete hydrolysis is obtained by prolonged boiling in

aqueous sodium hydroxide and yields free hydrazine and the phosphate ion:


0
I

I[o2rMUrm2 + on - P04 + 112H14


Table 23 lists some hydrolysis products obtained in this manner.

Hydrolysis does not always occur in the manner described above, but may immediately attack the P-U bond as in the illustration below

(50):


0 0
1 1
Ph-P(NHI7I2)2 + 2H20 > PhP(O12 + 2172114


In this case hydrolysis occurs so readily as to preclude the existence of the hydrazine derivative in the presence of imter.

The former behavior is typical of esters of phosphoric acid and the latter is generally observed for derivatives of phosphonic acid.

Hydrolysis is a competing reaction when a hydrazine derivative of an orgmophosphorus compound is treated with a chloroester, consequently low yields are to be expected for a reaction of this nature (50):

S 0 S 0
h 11 1+ 2 -aOH0 1 II
IP(IHN742)2 + 2C1-0-00211.5 - 211C1 PIhP(NIHHCOC2H5)2


m.p. 1330C. (46 per cemt)







TABLE 21

Amino-bis (hydr azino)phosphine Oxides

Product Yield m.p. Reference


Primary

C21I5N1J (1T1NPh)2 -- 1530 40
0
a-C3 17UH-b(kRUPh)2 -- 1510 40

0
is�-C4H9NMI-"(kN}IPh)2 -- 1410 40
0
n-C5 HI I Im- P (1H0,ZPh) 2 -- 1220 40

Secondary

(CH3) 2 (11 llRTh) 2 -- 194-1950 40

0
(C2H5) 2"T- P (IMHPh) 2 -- 184-1850 40

(--C3H7) 2'" (NITHPh)2 -- 1640 40

Wivso- C4l) 21- (N IIIPh) 2 - 1680 40

P, (C 13) ? (NMUH') 2 -- 1480 40









TABLE 22

Amino-bis (hydrazino)phosphine Sulfides


Product Yield M.p. Reference


Prisry

S
iCo-C4HqI41H,- P(IH~h)2



Secondary

S
(C21"5) 2"- P (ImNiPh) 2

S
(n-C 317)2- (N-Ph) 2


1290


1960






TABLE 23
Salts of Phosphorohydrazidates

Product M .p. Reference

Na (PhOPO2NINH2) . 37, 43
Na2 (OPO2Nt1l2) -- 37
Na2 (OP02NIA2). -i20 43
Na (HOPO2NI112) -- 37
K(Ph0PO2MHUTIN2) -- 43
K(IOP2NM m2) -- 37
NI-I4 (PhOPO2NUH2) 37
Ba (PhOPO2NnH2) 2 -- 37
Ba (oP02NIRI2) -- 37
Pb (PhOPO2NIU2) 2 -- 37
Pb (OP02T11thI2) --37


0
PhCI20TPkRf'%l2 43
OH

Ia (PhCH20PO2NHI2) --43
K (I,CI12PO2NHNH2) . 43








Hydrazone formation has been observed in compounds where the water produced in the reaction does not appreciably hydrolyze the reactants or products ( 43 ):


0 0 0 c113
I II 1
(PhCN2O)2PNULTH2 + CII3CCH3 > (PhCH20)2II:z'C + 110

Cn3

m.p. 109 C.


Other such knoxn hydrazones are listed in Tables 24 and 25. Both aldehydes and ketones have been used to prepare such hydrazones and the reaction is apparently general for hydrazine derivatives which contain the -H2 group.

When an arylphosphorohydrazidate is treated with rnhydrous hydrogen chloride salt formation is observed (43):


0 0
1 r1+1
(Fho)2pu11,2 + c -I [(FhO)21.111'.131Cl

m.p. 1500C. (dec.)


This reaction is analogous to the formation of hydrazinium salts and the proton attack invariably occurs on the most nucleophilie nitrogen atom (59).

Quarternization reactions using methyl iodide have been

reported, but the reactions described are not always similar to the reaction described above with hydrogen chloride. Instead, it is found that some nucleophilic centers will quarternize in preference to the nitrogen atoms contained in the hydrazino group (47).








TABLE 24

Benzylidene Derivatives of Alkylphosphorohydrazidates


Product m.p. Reference


9 /(CH30)2PN(CH3)N=CH Fl C1 69-71o 60

Cl
9 Cl1
(CI'I3�)2'IM'CH/ 123-4� 60

Cl


(C2AIO)Alamocll 52-53� 60


Cl


(C215o) 2A (CH3)=CH- ICl 54-55o 60


Cl


(C2!50) 2 ,IRM=Cl / 122-1230 60

Cl Cl








TABLE 25

lHydra&ones of Bis(hydratzino)phanylphosphine OxidesHydrazones of Bis(hydrazino)phenylphosphine Sulfides


Product m.p. References


0
h- P ("'=C (CH3) 2) 170 0 58

0
Ph-P(N.-C(CI13) 1)2 2010 58

0
Ph- P (1 iNl,(CH ( Ci) ) 2 1710 58




S
Ph-P (UIUa C (CUI3) 2)2 1550 58

S I
Ph-P(BU C(CH3) 4O C1)2 1620 58
S

PI-~oc(H)r'2133 0 58








S I


(11C) T 0P 11(CI 3) +2C131 2112








In2



m.p. 156-1580C. (dec.) The final reaction to be mentioned here is that of condensation. This type of reaction is potentially very promising as a

preparative method and occurs eith the intermolecular elimination of hydrazine at elevated temperatures (37):


0 0 0
1 15o0c1 5 1
2(PO)2PIIN12 - > (PhO)2PF11-fIIIP(OPh)2 + 12114

m.p. 1400C.


The product of this reaction was synthesized by another route in order to confirm its identity (43).

With respect to the practical applications of hydrazinophosphorus compounds, several patents have been granted which relate to the use of these materials as insecticides, fungicides, nermtodicides, and fertilizers.








Experinental and Results


lMateriois

HvTdrazines. l,l-Dimethylhydrazine and methylhydrazine are both corercially available. The samples used in this work were purified prior to use by distillation from calcium hydride. The reagents thus purified possessed very narrow boiling ranges: 1,1dimethylhydrezine, 62.2-63.00C./758 rm., methyihydrazine, 87.388.00C./761 cm_. All hydrazines were stored in air-tight glass containers in a cool, dark location.

1,1,2-Trimothylhydrazine was prepared from ,l-dinethylhydrazine by the method of Class, et al. ( 61 ); similarly, 1-ethyl2,2-dimethylhydrazine was prepared as reported by Klages, et al. (62). Both were distilled from calcium hydride or lithium aluminum hydride and boiled in the ranges 50-62�C. and 92-93�C., respectively. Identity in each case was confirmed by comparison of the infrared spectrum

with that reported in the literature (63).

Triecthylamine was purchased in the highest available purity and then refluxed over calcium hydride and distilled; the fraction

collected was in the range .8.5-39.5�C./759 Em.

phosphorus compounds were obtained commercially and in most cases were used as received. Many oxygen- and moisture-sensitive

phosphines deteriorated with age once the container had been opened and these were vacuum distilled prior to using.

Solvents. Regent grade solvents were used throughout. Where

drying was required the usual drying procedures were used. The dry solvents were then distilled. Every effort was made to avoid absorption of moisture by these solvents.








Nitrogen. Nitrogen was used to provide a dry, inert atmosphere wherever it was called for. Water-pumped nitrogen was purified by passing the gas over metallic copper turnings at 400�C. to remove oxygen and then through anhydrous magnesium perchlorate to remove moisture.

Equipment

In addition to the usual laboratory glassware, several pieces designed to carry out small scale reactions in the absence of air and moisture were used. Ace Mini-Lab apparatus (Figure 7) is an example of such special equipment.

Transfers and handling under anhydrous conditions were

facilitated by the use of a Lucite dry box in which was maintained a dry, nitrogen atmosphere, Figure 8. Manipulations were performed through a pair of Neoprene gloves and every effort was made to avoid unnecessary opening of the box. Several dishes of phosphorus (V) oxide were placed in each stage of the box to absorb moisture.

Most elemental analyses were performed by the Glbraith Microanalytical Laboratories, Knosville, Tennessee. Some nitrogen analyses, however, were performed with a Coleman nitrogen analyzer, Model 33.

Melting points were determined in sealed capillary tubes in a Thomas-Hoover melting point apparatus. Infrared spectra were obtained on a Perkin-Elmer Infracord Nodel 137 infrared spectrophotometer. Nuclear magnetic resonance spectra were obtained on a Varman High-Resolution Nuclear Magnetic Resonance Spectrometer, Model V-4300-2.





Nitrogen Inlet


I I9


Thermometer


Filtrate Receiver


FPigure 7. Mini-Lab Reaction Apparatus


Stirrer


Addition Fumnel








Nit co,"el Ine . Nitrogen Outlet Inet lI(B










' ransfur Ports


Inner Stage Outer Stage


Figure M. Dr'v Pox (Three-Eighths Inch Lucitc Construction)








fjaj1MMs yi a.ntl~ ra. d h. .



diphonylphosphie ws synthoined by tle hydree:nolysis ol chlorodiphonylphoophine as shown below:


FPhtP + 2H2NWb2 -Ph2HKe2 + Ns 2 . HCl


Similar solvolytic rejection ae well known and have been used to prepare ainophoophinos ad other coounds In which it was desired to fora a phoophorus-nitrogen covalent bond. However, this method had not been used previously to qrntheeim hydraeimpeoprhine. The nature of the reactants are such that moisture and owmi must be avoid ed. It is also desirable to use a solvnt in wkich the d~oired product is ealuble, but from which the l,l-di~mthyhydrecinium salt will precipitate.

Fifty-five and o e tenth g. (0.25 nole) chlorodiphmiylphosphine wee dissolvd in 25 ml. dry bensene sad added, with stirring and cooling, to t solution of 33 S. (0.55 mole) l,l-dimath'lWde ine in 25 ml. beebeae. The addition took four hous siter which time the miucure wao allowed to slowly vem to room temperature. The mixture vms then stirred for an additional hour at room t eerture to allow the ppcipituted crystals ef l,l-diathylhydrazinium chloride bo aeew sufficient size for eeaW filtration.

Filtr4*ion yielded, miaec washiag succesivelW withL beeeene and ether, 23.99 g. of a %We. crysalline solid, m.p. 79-810C. (literature value for l,l-d4ylhydYAAinium chleea, 81-820C.

(64)). This swunt is 99.3 per cent of theory.








Evaporation of the filtrate at room temperature and reduced pressure gave 59.97 g. of white solid, m.p. 62-66oC. This solid was dissolved in 175 ml. dry hexane at 700C. and the resulting solution was filtered. Upon cooling the solution to room temperature, crystals formed; these were collected and found to weigh 55.0 g. and melted at 65-67�C.

Sublimation of this product at 600C./0.20 mia. gave long,

prismatic crystals, m.p. 68.5-69.50C. The overall yield was 51.5 g.

(34.5 per cent of thcory bared on the equation presented above).

Analysis. Found: C, 68.65; H, 7.17; N, 11.29; P, 12.66. Calcd. for C14H17tl2P: C, 68.83; H, 7.02; N, 11.47; P, 12.68.

The infrared spectrum (Figure 9 ) and nuclear magnetic resonance spectrum (Figure 10) are consistent with the following structure:


\ /
Ph 11
Ell P - 11 CH3



2,2-Dimethylhydrazlnodiphenylphosphine oxide. In order to

further characterize 2,2-dimethylhydrazinodiphenylphosphine its oxide was prepared by three alternate routes: 1) atmospheric oxidation of the hydrazinophosphine, 2) oxidation of the hydrazinophosphine with activated manganese dioxide, and 3) the reaction of diphenylphosphinic

chloride with 1,l-dimetlylhydrazine.

1. Atmospheric oxidation.


0

PhP$M"'2 + 1/2 02 - Ph2PI-R2n12












-1
Cm.


I . . .) 0 1. . . .


N

K


10/


Figure 9. nft]'e Cd Spectrum of 2. "-liiethylhydrailiphenylphostl}inc (Nnirl Mull)


I m10J












Solvent: CDC13 Frequency: 56. - mc.


Penk Area Position
(ppil )

A U. 2 -n. ss

B 1. ( 3.229

C (.2 -1. 07


A B C


Figure 10. Proton Nuclear Magnetic Iesonmnce Spectrum

of 2.2-Dimethvlhydir-azi inuliplcn\ i)hosl)hine


68


Type



-C6115

-NII

-Cl3








A solution of 2.94 g. (0.012 mole) 2,2-dimethylhydrazinodipbhaylphosphine in 50 ml. benzene iws heated overnight while a stream of dry air was passed over the solution. Upon complete evaporation of the benzene there tms obtained a white, crystalline rmss and a dark oil. The cry.tals wxre collected mnd recrystallized from warm benzene and then sublimed at 1600C./0.32 rm.; 0.94 g. (30 per cent of theoretical, based on 2,2-dimetlhylhydra-inodiphenylphosphine) of a white, crystalline solid, m.p. 166.5-160.00C. resulted.

2. Oxidation with activated MnO 2.


0
I
Ph21 ,E'02 + M102 - Ph2PI,.2 + MMO

Two md ninety-four hundredths grms (0.012 mole) 2,2-dimethylhydrazinodiphenylphosphine in 50 ml. beoiaene was heated at 500C. with 4.1 g. (0.72 mole) activated mmganese dioxide (65) for 12 hours. The mixture was then filtered and upon evaporation of the benzene there was obtained a crop of white crystals. This solid was recrystalliaed from benzene &nd then sublimed at 1600C./0.30 mt. to give 1.60 g. of a white, crystalline solid, m.p. 166.5-168.50C. (45 per cent yield, based on 2,2-dimathylhiydrazinodiphenylphosphine).

3. election of diphw1ithoohinic chloride with 1,1-dimethylhydraauat.


0 0
I 1
I2P1+ 2H11e2 >~' P112PXIRCN2 + N2tI2IK~l








Tenty-three and seven tenths grams (0.10 mole) diphenylphosphinic chloride in 35 ml. benzene was added, with stirring and cooling, to a solution of 13.0 S. (0.21 mole) 1,l-dicthylhydrazine in 20 ml. of dry benzene.

When the addition was complete the mixture was heated to 700C., stirred for one-half hour and filtered hot. Upon cooling, the filtrate deposited a whiite, crystalline solid, m.p. 155-164�C. The ,ldimethylhydrazinium chloride on the filter was extracted with hot benzene and the washings combined with the filtrate. Reduction of the volume of the resulting solution gave additional solid.

Recrjstallization of the solid from 1:3 n-hexane:benzene

solution followed by sublimation at 1400C./0.20 am. gave 21.6 g. of product, m.p. 167.0-168.50C. (02.5 per cent yield, based on diphenylphosphinic chloride).

Analysis. Found: C, 64.39; H, 6.38; N, 10.73; P, 11.92. Calcd. for C14H117N2PO: C, 64.60; R, 6.58; N, 10.77; P, 11.90.

The infrared spectrum of this product is identical with those obtained from the products of atmospheric and Wi02 oxidation of 2,2-dimethylhydrazinodiphonylphosphine. lixed melting point determinations melted at 166-1680C.

The infrared spectrum (Figure 11) and n.m.r. spectra (both I1 and 31P) are consistant with this structural formula:


ph 0 H

cF-
p 1 ,IAH3

Ph U 1














CI


2(100


I-I






Iv V













Filnirc I I InlIa-co Spcc-trui i : 2. 2-D)ilncth\ Iii vdraincxdiphcII Iphospi tic (I ~'e (Nujol IuL I


3110((( I


] OI)t)
t





72


2,2-D-rtlvlhvdrazinod iphenvlphosphine sulfide.




Ph2P11111Thb2 + 1/8 S8 - Ph2 PNI"Ia2


Threa and forty-seven hundredths grams (0.0142 mole) of 2,2dinnthylhydrazinodiphenylphosphine was dissolved in 50 ml. dry benzene and added to 0.48 3. (0.015 mole) of finely divided sulfur in a small flask. The sulfur dissolved easily as the solution was warmed to 60 C. The solution was heated at 600C. for 30 minutes and then cooled to room temperature; no solid appealed on cooling.

Upon evaporation of the solvent a white solid, m.p. 87-96 C.,

waas obtained. Recrystallization from 1:1 benzene:n-hesane gave

3.59 g. (92 per cent of theory, based on the above reaction) of white crystals, m.p. 95.5-97.0'C.

knalysis. Found: C, yO.67; H, 6.20; N, 10.14; P, 11.21; S, 11.60. Calad. for C141117N2PS: C, 60.85; H, 6.41; N, 10.26; P, 11.45; S, 11.41.

The n.m.r. and infrared (F igure 12) spectra iere consistent with the structural formula below: Ph S C113


Ph CUi3

2,2-Dinethvlhydrazinomathyldiphenylphosphonium iodide.


Me
I+
Ph2 PNIERhIe-2 +- Yal-I [Ph2 PHI-lneb2 I I












Cm-I

.t OW '(Fl 21u, 1501) W 9 ''











" -I I













2'g~i( 1'-1. lii[rted Sped rIumlil ot _._ iiL c, iltt\ ' l liaoo'fl)l CI \ Ij i; stplin uli , i fFt, (\utlo llIF








One and seventy-wo hundredths g. (0.00704 mole) 2,2dimethylhydrazinodiphenylphosphine and 1.0 g. (0.00704 mole) methyl iodide were dissolved in 25 ml. dry ether and the solution was stirred at 250C. overnight. At the end of this time a solid was filtered from the solution and dried at room temperature and reduced pressure. The white solid weighed 2.71 g. (100 per cent yield, based on the

above equation) and melted at 156-1580C. An attempt to sublime this material resulted in thermal decomposition at 1600C. The salt is soluble in absolute ethanol.

Analysis. Found: C, 46.85; II, 5.47; N, 7.10; P, 7.85. Calcd. for C15H20U2PI: C, 46.65; H, 5.22; N, 7.25; P, 8.02.

A water-alcohol solution of this solid gives a positive iodide ion test, and iodine is liberated by the addition of nitric acid.

The infrared spectrum (Figure 13) is consistent with the

structural formula:






Ph CH3


The structure was further confirmed by basic aqueous hydrolysis to

l,l-dimethylhydrazine and methyldiphenylphosphine oxide. The oxide was identified by conversion to methyldiphenylphosphlinic hydrogen carbonate.

















, Ir /--.






'--I











I. izure i;. Infll . i'I , ;)t'tll r 0l of . 2-f)im tIdia{,i'{lznonwthl~IcIilJh~icn Iipthsplloniumn I{xtLIe (Ntulol M ull)








Hydrolysis of 2.2-dimethylhydrazinomethyldiphenylphosphonium iodide.


C13 "lC1.3

Ph^P 2l + Oil - -->- Fh2P-O II2NUIDI2


Two grams of sodium hydroxide was added to 3 g. of 2,2dimethylhydrazinomcthyldiphenylphosphonium iodide in 25 ml. of 1:1 ethanol:water solution and the mixture was boiled for one hour. As the alcohol evaporated it was replaced with water. The vapor above the solution was tested for the presence of free base (1,1-dimethylhydrazine) with damp red litmus paper and for the presence of a reducing substance with a drop of potassium permanganate solution on a strip of filter paper. Both tests were positive.

An oil separated from the aqueous solution. The amount of oil was too small for distillation, but an infrared spectrum consistent with methyldiphenylphosphine oxide was obtained. The oil was treated with a solution of sodium carbonate at 900C. for two hours and upon evaporation of the water a white solid residue was left. This was extracted with hot benzene and filtered. Evaporation of the filtrate gave one gram of a white solid, m.p. 107-1090C. This solid evolved carbon dioxide upon contact with a drop of hydrochloric acid. The literature value for the melting point of methyldiphenylphosphinic hydrogen carbonate, is 109-1110C. (66), and it is reported to liberate carbon dioxide upon contact with hydrochloric acid.








CH3 0 CH3

P.2 P- 0- C-0-PPh2
I I



From the foregoing experiswntal evidence we can conclude that the alkylation of 2,2-disethylhydrazinodiphenylphosphine with methyl iodide produces the hydrazinophosphonium salt rather than the hydrazinium salt indicated below:

+
[ Ph2FPT (CU3)3]I -.


Attested alklaU.Lz of 2.2-dinathvlbydraulamthyvldiiphenvlphospion i.s iodide with excess nhyl iodd. Treatment of 2,2dimethylhydrazinomethyldiphenylphosphonium iodide with excess methyl iodide in ether or toluene (heterogeneous reaction) g&"s quantitative recovery of starting wtorials. kt is clear, tleretAft, tiat, under the conditions cited here, alIkyletion of 2,2-dimethylhykrazinomethyldiphenylphosphonium iodide doac not occur.

Otber hYdzaxiuophosphonljPA mats. Sales of 2,2-dinethylhydrazinodiphenylphosphine were treated with various organic halides in an attempt to obtain additional informtion relevant to the ease of alkylation of the phosphine.

Reaction with be ul chloride. Four and eight hundredths grams (0.0168 mole) 2,2-dinethylhydrazinodiphenylphosphine end 2.12 grams (0.0168 mole) bensyl chloride were dissolved in 50 ml. dry toluene and the mixture was refluxed at 1100C. for 12 hours.

Upon cooling to room te erature two liquid layers were

observed. The toluene was removed smd attempts were mode to initiate








crystallization by cooling and by adding ether to the layer containing the desired product. No crystallization occurred and the product could not be purified by crystallization from absolute ethanol. The clear, yellow, viscous liquid gave a poLtive Cl test, however, and although the compound was not obtained pure, its infrared spectrum does indicate a salt-like structure which contains the bonds expected for 2,2-dimethylhydrazinobenzyldiphenylpbosphonium chloride


Ph

Cu2
[%12 P -_ IT(CH3) 2]C 1
+


Reaction with carbon tetrachloride. Upon dissolving 2,2dimethylhydrazinodiphenylphosphine in reagent grade carbon tetrachloride there forms in the yellow solution a faint precipitate which gradually disappears upon standiiC. Although no compound was isolated, there is the possibility that alkylation occurs according to the following equation:

Cl3

Ph21 UM (CH3)2 + CCl4 4- [h2P--uu (CI3)2J C1



Reaction with phenyl iodide. One gram 2,2-dimethylhydrazinodiphenylphosphine was mixed with excess phenyl iodide in dry ether and heated for one hour on the steam bath while the sample was protected from moisture with a drying tube. Several small crystals








formed in the liquid and these were washed with ether and dried in the air. The melting point was 162-1750C., and a nitric acid solution gave a positive I test.

No suitable method of purification was found. Sublimation attempts resulted in thermal decomposition. The compound is thought to be 2,2-dimethylhydrazinotriphenylphophonium iodide,

+
[ph3PNN(CII3)2] i

Reactipn with a . I'dibromoethyl ether. Two and forty-three ! I
hundredths g. (0.01 mole) 2,2-dimethylhydrazinodiphenylphosphine was reacted with a threefold excess of / , /'dibromoethyl ether in toluene at 600C. for 5 hours. A semi-solid %ich was not purified was the only observed product.

SrntbesIs of 1,l-bis(diphaylphosphino)-22-dimwthylhydrgzjne.

An experiment designed to test whether chlorodiphenylphosphine would undergo hydrazinolysis by 2,2-dimethylhydracinodiphenylphosphine resulted in the synthesis of l,1-bis(diphanylphosphino)-2,2-dimethylhydrazine according to the following equation:



Ph21MHMM*2 + Ph2PCI + Et3N 500C. N-N m

Ph2P me


+ TEt3 I]Cl


Three and fifty-four hundredths g. (0.0161 mole) chlorodiphenylphosphine and 3.62 g. (0.0358 mole) triethylamine were dissolved in 50 ml. dry toluene and to this wes added quickly at room









temperature a solution of 3.84 g. (0.0161 mole) 2,2-dimethylhydrazinophenylphosphine in 50 ml. toluene. There was no immediate evidence of reaction.

The temperature was slowly increased and at 500C. a solid appeared in the solution. Above 500C. the precipitation was copious. The mixture was stirred at 1100C. for one hour and filtered; the precipitate melted at 251-2530C. (literature value for triethylammronium chloride is 2540C.). Yield: 2.12 g. (96 per cent of theory, based on the above equation).

Evaporation of the filtrate gave 6.51 g. of a white solid, m.p. 126-1330C. An attempt at sublimation resulted in decomposition at 1350C. Recrystallization from dry n-heptane gave fine, white crystals, m.p. 129.5-132.50C., in 76 per cent yield, based on the equation above.

Analysis. Found: C, 72.65; H, 5.98; N, 6.39; P, 14.56. Calcd. for C26H26I2P2: C, 72.88; H, 6.12; N, 6.54; P, 14.46.

The n.m.r. spectra and infrared spectrum (Figure 14) are

consistent with the proposed structure, but a small absorption peak at 1176 cm.-I in the infrared shows some oxygen (as P=O) as an impurity.

Chlorophosphination of triethylsmine. Since, as in the synthesis described above, it has in several instances proved convenient to use triethylamine rather than an excess of the hydrazine as a hydrogen chloride acceptor, it was desirable to determine whether or not chlorodiphenylphosphine reacts directly with triethylamine.














_ _ __( _ _ _ _ __) ((I) ,h ( I ,9)( I(�




!-'



I'







V _Fi~ui'e 1. Inrared Speccti r o{ 1l1 -Eis(diphen~ 1phospi~ mm-2. 2-di~Lelh3 1: .3d ainc (Nuioi ,Mull)








In view of the fact that chloramine has been shown to react with tertiary phosphines in accordance with the equation (67)

+
CN2+ R3P > [R3 j Cl


it might be expected that chlorophosphines such as (C6115)2PCI would react with tertiary mines according to the following equation:

+
Ph2PCI + R3N [ [R3NPh2]c"


In an experiment designed to test whether or not chlorophosphination of triethylamine occurs under the conditions usually employed in the hydrazinolysis of chlorodiphenylphosphline 5.46 g. (0.054 mole) triethylamine was dissolved in 50 ml. anhydrous ether and added to 11.90 g. (0.054 mole) chlorodiphenylphosphine in 50 ml. ether. An immediate cloudiness appeared in the solution which persisted throughout a 30 minute reflux at 400C.

Filtration gave 0.63 g. of a white solid, m.p. 254-2550C., which was completely water soluble. A mixed melting point determination with an authentic sample of triethylammonium chloride melted at 2532540C.; the infrared spectrum of this solid is identical with that of triethylmnnonium chloride.

Evaporation of the dher and triethlyamine from the filtrate at reduced pressure gave a yellow viscous liquid which was distilled at 105-1070C./0.17 rm. and shown to be ehlorodiphenylphosphine by its infrared spectrum.








It may, therefore, be concluded that chlorophosphination of triethylamine with chlorodiphenylphosphine does not occur under the conditions employed here for the hydrazinolysis of chlorodiphenylphosphine.

It should be noted that a little triethylammonium chloride

resulted from the reaction mixture, showing that although all reagents had been previously distilled and dried, some hydrolysis had occurred. It was later found to be possible to avoid the formation of triethylamnoniumn chloride upon mixing chlorodiphenylphosphine and triethylamine by performing all transfers in the dry box. For the urual bulk reaction, however, it is unnecessary to take the extra care to avoid this small amount of hydrolysis as it lowers the yield by only a fraction of a per cent.

Reaction of 2,2-dimethvlhvdrzinodiphenylphosphine with carbon ditulfide. Five ml. reagent grade carbon disulfide was added to 1.97 g. (0.0807 mole) 2,2-dinethylhydrazinodiphenylphosphine in 10 ml. anhydrous ether. A deep red color developed immediately and slowly faded to yellow as the solution was evaporated at 400C. ever a five hour period.

Upon standing overnight, large, white crystals appeared in the solution. These were collected and washed with he-ane and melted at 110.5-141;50C. The yield was 82 per cent of theory, assuming the 2,2dimethylhydrazinodiphenylphosphine reacted with the carbon disulfide in a 1:1 ratio. The analysis corresponds to (C6115)2PHMN(CH3)2-CS2: Found: C, 56.09; 11, 5.55; N, 8.69; P, 9.46; S, 20.17. Called. for C15H17N2PS2: C, 56.23; 11, 5.35; N, 8.74; P, 9.67; S, 20.01.

It was found that if the white, crystalline product of this reaction is redissolved in carbon disulfide, the red color appears once more.








The literature describes the interaction of tertiary phosphines with carbon disulfide, and there are reported compounds of the type


S
+ /
R P- C




which are red, crystalline solids (55,68 ). The structures of these compounds have been confirmed by X-ray diffraction analysis (69) and

it is well-known that the compounds contain a P-C bond and no formal P-S bonds. The red color is thought to arise as a result of the Zwitterion-type structure.

Since a red color develops in the interaction of 2,2-dimethylhydrazinodiphenylphosphine with excess carbon disulfide, it appears that a Zwitterion complex is the initial product of the reaction: S" S
C


P - P-NI(C"3)2


As the reaction proceeds, however, the color fades as the amount of carbon disulfide is decreased. The final product is a

white solid and has an infrared spectrum (Figure 15) which is complex, but shows no absorption in the region assigned to the N-11 bond. The N-H absorption has been shifted to a higher frequency which indicates a change in one of the substituents on the nitrogen atom attached to the phosphorus; the -1(C113)2 group appears to be intact. A weah, absorption is evident in the S-H region. The monosubstituted phenyl














Cm


1)00 9(1


I-----~1~~~'~


K


Fi4ure 15. Infrared Specl rum of the Product of the Rea .t ion o1 2.2 -Diiocth% lh y razinodiphen lpl, osp) ine with Carbon ) izAl ide (Nuiol Alull)


3o( )O 2(100 15 00





86


group peaks are unchanged and the P-phenyl absorption has not been shifted at all. The structure best fitting the infrared data is

drawn below:
S S

Ph ", C C113
P - I- NPh zCH3

The proton nuclear magnetic resonance spectrum is in general agreement with this structure. Four peaks are observed, none of which is the characteristic N-H close doublet which is observed in 2,2dimethylhydrazinodiphenylphosphine. The peaks assigned as phenyl protons and methyl protons agree with the areas expected for the structure given above. The remaining two peaks are of unequal area and are presumed to arise from 1) the S-H proton and 2) the proximity of the

S-11 proton to the 31p atom, which is an odd nucleon and with which 1H will interact by spin-spin coupling.

The fact that a red color develops on dissolving this material in carbon disulfide may indicate that the phosphorus atom is still available for loose coordination in excess carbon disulfide, and gives a molecular complex as shown below.

S S~SE
C C-S
I /

N(C11)2


Treatment of 2,2-dimethylhydrazinodiphenylphosphine oxide with carbon disulfide resulted in no reaction.

PYrolytic condensation of 2.2-dinethvlhydrazinodiphenylphosphine. It has been observed that when a sample of 2,2-dimethylhydrazinodiphenylphosphine is purified by sublimation, there sometimes is








left behind in the pot of the sublimation apparatus a yellow, rosious material. On theme occasions there is also found in the cold trap used to protect the vacu~n pump a wall amount of a volatile

liquid. Infrared analysis and vapor phase chromatography data indicate that this liquid contains dimethylei2me and l,l-dimethylhydrasine.

Diqnthylsmine can be produced by a thermal decompoition which produces the phosphonitrilic system,

xPh2PM]11(CH3)2 - - (P12 )Mx + x(C13)2WH

and 1,1-dinathyihydrazine can be one of the products when 1,1-bis(diphenylphosphino)-2,2-dimethyliydrazine is also a pyrolytic condensat ion product:


2 Ph2P tEI(CH3)2 -- (P-(2P)2NN(C13)2 + (CH3)2NMH2


It should be noted rhat only a small amount of this solid,

resinous material is observed after sublimation of 2,2-dimethylhydrazinodiphenylphosphine, but a suggestion for further work would be to investigate these reactions on a larger scale ad identify with certainty the reaction products.

Hydrolysis of 2.2-d . vhlhydreziaod:Lphewylpl-osphine. A sample of 2,2-dimethylhydracinodiphenylphosphine was treated with 0.1 N 1101 with the result that both 1,1-dimethylhydrazine and diphenylphosphinic acid (m.p. 191-1930C.) were isolated in high yield from the product mixture. The diphenylphosphinic acid was filtered from the solution and the l,l-dimethyhydrazine was distilled from the filtrate which was made basic by the addition of NaOH solution.








Hydrolysis by atmospheric moisture vas found to be a minor problem with 2,2-dimethylhydrazinodiphenylphosphine. EMeriments with methvlhydrazine and chlorodiphenylphosphine

Reaction of methylhydrazine with chlorodiphenilphosphine. Ten and four tenths g. (0.0472 mole) chlorodiphenylphosphine was dissolved in 35 ml. dry ether and added to 4.35 g. (0.0945 mole) methyihydrazine (redistilled and dried over calcium hydride) in 40 ml. dry other. The addition took four hours and was performed with cooling and stirring under a nitrogen atmosphere. No precipitation occurred at O�C., but upon warming the mixture to 250C. solid appeared and the mixture was stirred at 250C. to allow the reaction to proceed to completion.

Filtration gave 3.71 g. methyihydrazinium chloride (theoretical is 3.88 g., based on chlorodiphenylphosphine), and 9.57 g. of a yellow, viscous liquid which did not contain chlorine, as evidenced by a silver nitrate test on a small portion dissolved in dilute nitric acid. No crystallization could be induced in the liquid and no way was found to effect purification.

Analysis. C, 67.55; 11, 6.46; N, 11.96; P, 14.34. Caled. for C13U115N2P: C, 67.81; H, 6.57; IT, 12.17; P, 13.45.

The reaction product may have either of wo forms.

CuI3

Ph2 1 tI'IICI3 or Ph2N

Nl2


The n.m.r. spectra indicate that a mixture of both these species is present in die product, but the data are too complex to indicate the








relative percentages of the constituents. The infrared spectrum contains all the expected absorption frequencies, but is of little value in determining per cent composition.

An attempt to vacuum distill the product resulted in thermal decomposition. A liquid fraction was collected in the range 98-990C./0.36 ma.; however, the bulk of the product remained in the distilling flask as a resinous, amber-colored solid.

Extrapolation of the boiling point at reduced pressure to the normal boiling point of the liquid on a temperature-vapor pressure noaograph gave ca. 1750C./760 nm. (diphenylphosphine boils at 2800C. per 760 a. (32).

taglsis. Found: C, 77.42; 11, 6.22; N, 0.25; P, 15.88. Cnlcd. for (C6H5)2PH: C, 77.41; I, 5.95; N, 0.00; P, 16.64.

The molecular weight (cryoscopically, in benzene) is 185. Calcd. for diphcnylphosphine: 186.2.

A methyl iodide derivative, prepared according to the equation
+
Ph2PH + 2CH31 - [Ph2P(CH3)2] i - + HI


has a m.p. 2410C.

The infrared spectrum (Figure 16) is consistent with the

structure for diphenylphosphine and contains a very prominent absorption peaki at 2295 cm.'l, which is characteristic for the P-}1 bond.

The thermal decomposition of the product mixture apparently is according to the following equation:


12P' 'CI3 - Ph2PH + condensed species
Ph2 PNCH3NH2J












Cm.

Ioo0 300U 201")( 1500 10( ( 901 "1.0



I l / J









it " /




17-







-;_ Iure 1I. InfrareI Specfllirn 01 Dit)hen lp;osplj i,,C (Cc11









An attempt was made to sepemete the two componei-ts from each other by reaction with benzaldshyde, but the only isolated product ims met lfbe&ylidenehydrazrvie, C611CIN-NiMH3, m.p. 178-1790C. (literature value 1790C.). Apperitly the woter produced in hydrazone formetion hydrolyued the P-M bond in the product.

N10 separation was me4c on the residue from the thermal decoaosition ard the analytical results did not aree with any single coiyound. The infrared spectrum was highly opoque, which is characteristic of polymuric materials.

PA~e.ion of m ylhydr= -odiphanvphosphine with sulfur. The mixture of methylhydrazinodiphenylphosphines produced in the reaction

of chlorodiphenylphotphine with methylhydrazine was reacted with a smll amount of finely divided sulfur in bena*ne solution with the result that hydrogen sulfide evolved and a dark gum wes produced from which no product was isolated.

Rj~i Qf thvlh yrasinodiphe*YlphQsitine viA.h carbon

disulfide. Upon dissolving the zethylhydrazinodipymiylphosphine mixture in carbon disulfide there was no immediate evidence of reaction, however, over a 12 hour period hydrogen sulfide evolved, and upon removal of the solvent a dark gum remained from which no pure material was isolated.

Synthesip of 1,1 ,2-tris (diphenvlphosphino)vthvlhydrasine.


/P Ph2
3Fh2PCI + CH3"% + 3(C21)3P -- (Ph2P)2M"\ + 3 (C2115)311 Cl
CH3








Eleven and ninety hundredths g. (0.054 mole) chlorodiphenylphosphine and 7.24 g. (0.0732 mole) triothyla.ine were dissolved in 50 ml. dry

toluene (b.p. 109.1-109.2�C.) and to this was added slowly with stirring and cooling 0.83 g. (0.018 mole) methylhydrazine in 35 ml. dry toluene.

When the addition was complete the mixture was heated to 100�C. and stirred for one hour and filtered hot. The weight of

triethylamaoniun chloride on the filter was slightly more than theory based on the above equation; apparently some of the toluenesoluble product was occluded in the salt. Upon evaporation of the solv.nt at room temperature and reduced pressure, there was produced a yellow gum. Three hundred ml. of dry ether was added to the gum and a white solid separated from the resulting solution. This solid was washed with additional ether) collected, and dried under vacuum. The amount collected was 2.55 g. (21.8 per cent yield, based on chlorodiphenylphosphine), m.p. 151�C. doe.

Purification was affected by recrystallization from acetone in the drj box. The resulting crystals melted at 152.3-153.0�C.(dec.).

Analysis. Found: C, 74.33; 11, 5.71; N, 4.75; P, 15.41. Calcd. for C371133172P3: C, 74.24; H, 5.56; N, 4.68; P, 15.52.

The infrared spectrum (Figure 17) of this material is consistent with the structure


Ph2P PPI2
P /2

iP/P li














-1


Figure 1. Inlared Spectrum of 1. 1 2 -Tiis(diphenylphospl, ino)methy Ihvc razine (Nujol Mulil




Full Text

PAGE 1

A STUDY OF THE SYNTHESIS OF SOME ALUMINUM AND PHOSPHORUS DERIVATIVES OF ALKYL HYDRAZINES By ROBERT PETER NIELSEN A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA June, 1962

PAGE 2

ACKNOI-JLEDGMENTS Tlae author wishes to acknowledge the friendly cooperation of his conmittee and the faculty of the Chemistry Department of the Iftiiversity of Florida. It is considered both an honor and an opportunity to have worked under the direction of Dr. Harry U. Sisler, who, although entrusted with the taomentous task of administration of this Department, is able to give generously of his time and abilities to direct meaningful scientific inquiry. His friendship and wide scope of interests have proved inspiring. The time and effort spent by Commander N. L. Smith (Ret. USNR) in acquainting the author with synthetic technique and phosphorus-nitrogen chemistry is sincerely appreciated. Besides offering an open door and a sympathetic ear, he provided encouragement when it was needed and advice when it was desired, Thanlcs go also to Dr. Wallace Brey and Mr. Ken Lawson for nuclear magnetic resonance measurements, and Mr, Ifoward Latz and Mr. Leo Pijanowski for assistance in obtaining infrared spectra. The fellow members of the Inorganic Section deserve thanks for friendly con?>etitive spirit and helpful discussion. The author is indebted to the Petroleum Research Fund for a financial grant which made this work possible. ii

PAGE 3

XABLE OF CCNTEOT3 Page ACKHOiJLEDGMEHTS n LIST OF TABLES jy LIST CF FIGURES ^i Chapters I THE IWTERACTION CF TWO ALKXL HYDRAZINES WITH TRIETIKLALmilNUM 1 Introduction 1 Es^erimcntal and Results 9 Discussion 27 Conclusion 35 tl REACTIONS CF ARYLHALOHiOSPHINES AND DERIVATIVES OF ARYLIIALOPHOSPHINES IJITH SEVERAL ALKYL HYDRAZINES 36 Introduction 35 EKperinental and Results 61 Discussion 123 Conclusion 147 BIBLIOGEAFJY U9 BIOGRAPHICAL SKETCH I53 iii

PAGE 4

LIST CF TABLES Table Page 1 Amine Manes 5 2 Bis-Amine Alanes g $ Bis(Mixed) amine Alanes ^ A Inner Coioplex Amine Alanes 6 5 Aminoalanes 7 6 Extended Network Alumimim-Nitrogen Polymers 8 7 Itydrazine Alanes % B Hydrazinoalanes 8 9 Hydrazinoalane-Hydrazine Adducts 9 10 Polymeric Species 9 11 Alkyl and Aryl Hiosphorohydrazidates 39 12 Alkyl and Aryl Pliosphorohydrazidothioates 40 13 Alkyl and Aryl Hiosphorophenylhydrazidothioates 42 14 Phenylhydrazidoallcylphosphcnates 43 15 Bis (hydrazine) phenylphosphine Oxides iA 16 Arylphosphorodihydrazidothioates 45 17 Tr is (hydrazine) phosphine Oxides and Sulfides 47 18 Arylphosphorochloridohydrazidothioates 48 19 All
PAGE 5

LIST OF TABLES Continued Table ^age 24 Benzylidene Derivatives of Alley Iphosphorohydrftzidates 58 25 Hjrdrazones of BisOiydrazino)pKenylphosphine Oxides-IIydrazones of Bis(hydraaino)pheiiylphosphino Sulfides 59 26 Principal Infrared Absorption Frequencies for Some Hiosphorus Hydrr.zine Derivatives 124 27 Infrared Absorption Prectuencies and Assignments 125 28 Nuclear Magnet5-c Resonance Data Sunanary 128 29 Scale of Chemical Shift Values from H, M. R. Data 133

PAGE 6

LIST OF FIGURES Figure Page 1 Mini-Lab Distillation Apparatus 12 2 Reaction Apparatus 13 3 Molecular Weight Apparatus (Modified for AtraocphereSensitive Conq>ounds) 14 4 Infrared Spectrum of 2,2-Diinethylhydrazinodiethylalane (Melt) 18 5 Infrared Spectrum of 2,2~Dimethylhydrazinodiethylalane (Dilute Hexane Solution) 19 6 Infrared Spectrum of Condensation Product, (EtAlNNMe2)j5 (Hexane Solution) 22 7 Mini-Lab Reaction Apparatus 63 8 Dry Box (Three-eighths Inch Lucite Construction) 64 9 Infrared Spectrum of 2,2-Dimethylhydra2inodiphenylphosphine (Nujol >fejll) 67 10 Proton Nuclear Magnetic Resonance Spectrum of 2,2Dimethylhydrazinodiphenylphosphine 68 11 Infrared Spectrum of 2,2-Diraethylhydrazinodiphenylphosphine Oxide (Nujol Mull) 71 12 Infrared Spectrum of 2,2-Dljnethylhydrazinodiphenylphosphine Sulfide (Nujol Mull) 73 13 Infrared Spectrum of 2,2-Dimethylhydrazinomethyldiphenylphosphonivtm Iodide (Nujol Mull) 75 14 Infrared Spectrum of I,l-Bis(diphenylpho6phino)2,2-dimethylhydrazine (Nujol Mull) 81 15 Infrared Spectrum of the Product of the Reaction of 2,2-Dimethylhydrazinodiphenylphosphine and Carbon Disulfide (Nujol liull) 85 16 Infrared Spectrian of Diphenylphosphine (Cell) 90

PAGE 7

LIST OF FIGURES Continued Figure Page 17 Infrared Spectrum of l,l,2-Tris(diphenylphosphino)laethylhydrazine (Nujol Mull) 93 18 Infrared Spectrun of l,2,2-Trixaethylhydra3inodiphenylphosphine Oxide (Nujol liill) 96 19 Infrared Spectrum of l-Ethyl-2,2-diiaethylhydrazinodiphenylphosphine Sulfide (Nujol Mull) 101 20 Infrared Spectrum of l-Ethyl-2,2-diinethylhydrazinodipheiiylphosphine Oxide (llujol Hull) 104 2i Infrared Spectrum of l,2-Bis(diphei:Qrlphoephino)hydrazine C*JJol Mull) 107 Z2 Infrared bpectrtna of Bis(2,2-dimethylhydrazino)phenylphosphine (Nujol Mull) 110 23 Infrared Spectrum of Bis(2,2-dimethylhydrazino)phenylphosphine Oxide (Nujol Mull) 113 24 Infrared Spectrum of Bis ^2, 2-dimethylhydrazino) phenylphosphine Sulfide (l^^ujol Mull) 116 25 Infrared Spectriim of Hiosphoryl Tri(2,2-dinethylhydr?)zide) (l^jol nill) 119 26 Infrared Spectrum of Thiophospboryl Tri(2,2dimethylhydrazide) (Nujol t&ill) 121 vii

PAGE 8

CHAPTER I THE INTERACTION (ff T.IO ALKYL HYDRAZINES IJITH TRIETKYLALUIIINUM Introduction The highly interesting and challenging field of organometallic chemistry has undergone troaendous grot/th in recent yer.rs, and among other valuable contributione to the science of chemistry there loave been discovered new modes of chemical coirfaination, entirely new types of chemical compounds, new useful materials at lower costs than previously-used synthetics, and new routes to Icnosm materials. Evidence of great interest in the field of organometallics manifests itself in the recent appearance of several te;:ts devoted to this unique combination of organic and inorganic chemistry (1-4 ). One of the areas of organometallic chemistry which has interested investigators for as long as metal alhyls have been knoim is the preparation and study of molecular addition complexes betv^een the metal alkyls and various Lev;is bases (electron pair donor compounds) The metals in i^ich most interest has been shoi
PAGE 9

Tlie reason that the rnany nitrogen-bace complexes of aliraiinum allqrls and the various other alanec have not beeii the cubject of an extended review article is that \rork on this group of compounds was begun in earnest only within the past few years. The importance of such compounds at this point seems to lie in their intrinsic properties as well as the fact that they are intermediates in condensation reactions. The aluminum-nitrogen compounds in iTiiich this great interest has been demonstrated may someday find application as high temperature poljnaeric materials and perhaps as propulsion fuels. As intermediates to new and exotic compounds they are virtually unexplored. Well-established chemical and physical evidence concerning the nature of various alanes, and specifically the aluminum alkyls, indicate tliat they act as electron pair acceptors, or Lewis acids. The acceptor tendency is so great as to preclude their existence in the unassociated state (1,2), This fact provides a basis for the observed reluctance of investigators to use the Grignard method for preparing aluminum alkyls, since the compounds prepared can with only very great effort be freed of the ether used in the synthesis (13). The diethyl etherate of triethylaluminum has been studied in detail ( 13-15 ). The dimethyl etherate of trimethylaluminum has bceia found to be an extremely stable compound, so stable in fact that no dissociation data can be obtained in the gas phase (l^; fuch data have been obtained for many other Group III addition con5)lexes, hot^ever. Various methods of preparation of molecular addition complexes of aluminum alkyls have been used, but the most common include direct combination in the vapor phase (both diluted and undiluted) direct combination in the liquid phase, and direct combination in a suitable

PAGE 10

solvent. Some experiments have been crrried out in vjtiich one (or both) of the reactants were generated in situ Since this work is concerned with nitrogen derivatives of aluminum it is appropriate to discuss those reactions and compounds reported in the literature xdiich include aluminum-nitrogen molecular addition complexes and the aluminum-nitrogen covalent bond. At this \vrriting there have been reported in the literature three mnin types of covalcntly bonded aluminum— nitrogen compounds: the amine-alanes the aminoalanes, and polymeric materials which contain aluminum-nitrogen bonds. These three types of compounds are intimately related and may be arranged in a reaction sequence where the initial molecular addition complex is the parent compound in a pyro lytic series, as in the following example: (17,18; 1, Adduct Formation of an Amine Alane. Al2(CH3)(3 -;21^3 = 2H3N:A1 (0113)3 m.p. 56.7C. 2* Condensation to en Aminorlane. 70C. H3N:A1(CH3)3 = HjN-Al (0^13)2 •: CH^ m.p. 134. 2C. 3. Further Condensation to Polymeric Material. 200C. H2N-A1(CH3)2 = (mLMCH3) ^ v CH^ not fully characterized 4. High Temperature ]^ro lysis to Extended Polymer. y(HNAlCIl3)j^ = C^M)^ + CH^

PAGE 11

A systematic investigation of systemc o£ this nature has recently been carried out and the scope of the reaction has been extended to include many amines and several alanes (TO Exaraples of confounds of this type (excluding those shown above) arc given in Tables 1-10 j pages 5-9. Reactions of these compounds other than by hydrolysis and condensation are largely vinlcnoim, and to date there is no mention in the literature of a useful polymer in the aluminiau-nitrogen system. Various theoretical aspects of aluminum chemistry have been denonstrated, however, and no tendency for pi-bond formation between aluminLim and nitrogen has been found. Logical extensions to this work with alanes seem to lie in two directions: 1) Levjis bases can be used uhich contain functional groups other than and in addition to an amino group, and 2) polynuclear Let7is bases can be used. The first of these alternatives has been explored in a very perfunctory manner and involves the reactions of alanes with olef inic amines, ethers, and thioethers to give various ring compounds and organic derivatives (19). Obvious possible uses for these reactions and intermediates lie in the field of polymerization catalysts. Tiie use of polynuclear Lewis bases with alanes was, tmtil very recently, totally unexplored. iJithin the last year, hoi/ever, there have appeared tiTO articles concerning the interactions of aluiainura alkyls with hydrazine and alkyl hydrazin:.( 2), 21 ) It is interesting to note that the experimental results obtained by these workers are in most aspects in accord with the results we have recently obtained

PAGE 12

TABLE 1 Amixie Alanes Conipound m.p. b.p. References (CH3)3N:A1H3 (CH3)3N:A1D3
PAGE 13

TABLE 2 BisAmine Alanes Compound m.p, b.p.

PAGE 14

TABLE 5 Amlnoalanes Compound

PAGE 15

TABLE 6 Extended Nctxrork Aluminura-Nitrogen Polymers Conqpound m.p. b.p. Referencee r(CH-)NAlCl] L 3 n above 220"C, 17 18,30 TABLE 7 ifydrasine Alanes Confound n.p. b.p. References (CH3) 2K-N (CH3) 2 : Al (CH3) 3 (CH3) 2N-IJHCH3 : Al (CH3) 3 80 -83 C, 65.5-66. 0C. 21 21 TABLE 8 HTdrazinoalanes Con5>ound m.p. b.p. References [(CH3)2N-m.Al (CH3)2]2 [(CH3)2K-N(CH3)-A1 (CH3)2]2 [(0113) 2AI-KH-NH-AI (CH3) 2] 77.0-78.5C. 125.0126. 5C. Ghock sensitive 21 21 20

PAGE 16

TABLE 9 Hydrazinoalane-Ifydraziiie Adducts Compotmd

PAGE 17

10 Materials l.l-Dimethylhydrazine ic coianercially available and was dictilled prior to using. The material used had a boiling range of 62.2-63.0C./75G nan. Distillation t/as carried out over calcium hydride to resaove any traces of water which the consaercial material may have contained. Care was taken to observe completely anhydrous conditions during the transfer and liandling of this and the other reagents and solvents used. Treithylaluninuni was purchased in 250 grata quantities in steel lecture bottles which were fitted with Teflon-packed needle valves. The purity of this niaterial was not high enough to permit use without purification. Distillation at 56C,/0.50 nsn, provided samples v;hich analyzed 99,6 per cent triethylaluminum, as determined by measuring the ethane evolved upon aqueous hydrolysis. Analytical Reagent grade solvents were distilled over calcium hydride prior to using and were then stored in air-tight glass containers with metal foil or polyethylene closure linerc. All transfers v/ere perforraed either in a dry bos or by pipette with a nitrogen flush. High purity nitrogen was used in the dry box and wherever a constant flush seemed necessary to avoid contact with the atmosphere. Equipment. A dry box is essential in carrying out synthetic tTork with triethylaluminum and this piece of equipment is described in detail in Chapter II of this dissertation, under the Equipment section. Distillations vjere carried out in Mni-Lab Standard Taper 14/20 apparatus as shown in Figure 1. Tlie actual reactions were carried out in a Standard Taper 14/20 100 ml. round bottom flask fitted with two 10 ran. diaroeter by 30 mm. length side arms. Figure 2

PAGE 18

11 illustrates this flask and the con5>lenientary equipment, which includes a pressure-equalizing addition funnel of 50 ml. capacity and a Precision Met Test Meter. The Wet Test lleter can be read directly to the nearest 3 ml. A freezing point depression type apparatus was used for molecular weight determinations. Its design was modified slightly in order to provide a constant, slow nitrogen flush so that contact of the solution with atmospheric oitygen could be avoided, as shown in Figure 3, Elemental analyses were perfoirmed by Galbraith Microanalytical Laboratories, Knoxville, Tennessee. Carbon, hydrogen, and nitrogen were estimated by combustion methods and special precautions were observed to maintain sample integrity before coaabustion. Aluminum analyses vxcre performed by the method of Schwarzenbach (31 ) in vrfiich the a luminvimcontaining solution is treated with excess disodium ethylonedianinetetraacetate and the excess titrated with standard zinc sulfate solution to the pink Eriochrome Black T end point. Tlie interaction of triethylaliJgainum with 1.1-dimethylhydrazine In a typical experiment 8.0321 grams (0.0704 mole) triethylaluminim was vacuum distilled inLc a 100 ml. Standard Taper 14/20 round bottom flask fitted with tt70 side arms and \^ich contained a small. Teflon-encapsulated, magnetic stirring bar. All ground glass joints were lubricated with Kel-F fluorocarbon grease, vThich was found to be more resistant to attach by triethylalianinum than are other, more conventional stopcock greases, including silicone grease.

PAGE 19

12 vacuum inlet Figure 1. Mini-Lab Distillation Apparatus

PAGE 20

13 ihiM.i',Mi:i znr) ) u

PAGE 21

14 Freo/.ii..^ oii t j fcp/ 3 . lo .ppaatis bol -or.t P';bbler (Lor sat ratii'.iv dry )dtro^en) Fif^. 3. Molecuiai' Woi-ht Apparat'i-; [Moaiaod ior .tiuosphore-Sensitive CopipoiCid

PAGE 22

Tlie triethylaluminum was frozen by inner sing the lower half of the flask in a dry ice-acetone slurry. I.liile a nitrogen flush was maintained above and around the flask, the stoppers which capped the side arras during distillation were rooioved and through one side arm was inserted a low range (to -100C.) pentane thejnnoiaeter The other side ana was connected via two dry ice-acetone traps to the Wet Test Meter. The addition funnel, containing 4.44 grams (0.0739 mole) 1,1-dimethylhydrazine, was fitted to the flask and with the lower half of the flask at -7oC. the iJet Test Meter was set to zero. The contents of the addition funnel was added in very email increments (one drop or less) to the frozen triethylaluminum, and the system was warmed and melted for mixing between additions. Ethane (identified by vapor phase chromatography) was the only gaseous product and the evolution was slow and controllable. Evolution of gas was evident at temperatures as lov; as -60C. for the first half of the addition, but during the second half higher temperatures (near 0C.) were needed. No liquid or solid collected in the cold traps. The addition took four hours, after v/hich the system was warmed to 25C. and stirred ovemijihi: (12 hours) to assure complete ethane evolution. Tlie observed, corrected volume of ethane was 1.59& liters; theory calls for 1.577 liters for the reaction AI2 (02115)5 + 2(CH3)2K-NH2 f(CH3)2K-WH-Al (€2115)2] 2 • 2C2l% Tlie 5 per cent excess 1,1-dimethylhydrazine was removed by pumping briefly at 45C./0.15 mm. The resulting product was a colorless.

PAGE 23

16 crystalline solid which weighed 10.11 graras (99.6 per cunt yield, based on 'criethylaluminum used) and ijhich melted at 43-44C. (sealed tube, uncorrected). Hydrolysis of a 2.7071 gran sample of the product gave 0.345 liter of ethane. Calculated for [(Cll^) 2imiAl (C^ll^) ^ 2' O'^^^i liter. [(CiygNWH/^l (02^13)21 2 -JSH3O'' :4H2O 4C2% -r 2(CIl3)2NNH3'' 42Al(Ii20)/"'Analysis. Found: C, 50.05; II, 11.61; N, 19.70; M, 18.57; C2H5, 40.73. Calcd. for [(CH3)2lMiAl (02113)2] 2: C, 49.93; H, 11.38; N, 19.43; n, 10.71; C2H3, 40.31. Jtolecular weight, found: 2G1 (cryoscopic in benzene) Calcd. for [(Cn3)2NWaM(C2H3)2]2: 233.4. The product is soluble in raost coiainon, inert solvents. The infrared spectra obtained on a melt and 10 per cent in n-hexane show bands in the regions expected for C-II stretch and bend, H-H, N-W, C-W-N, and N-CH^. Figures 4 and 5 shov7 the infrared spectra. The observed molecular weight indicates a dineric structure. Tliree possible structures, two of V7hich are geometric isomers may be considered. II5C2 Colic: \ / n ii-w ^11(0113)2 (1130)211 N^H Al IIsCs C2K5 1. Six I'leiabered Ring

PAGE 24

17 "5^2 C2H3 H5C2 H5C2 H ^M^ U (HoC),N ^Al-^ II \ /\ ^ \ /\ /^ /\/\ /\/^\ (1130211 Ai iKCHo), ir /I N(CHo)H5C2 C2II5 H5C2 C2II5 2. Cis-Four Ilembered Ring 3. Trans-Four Mecibered Ring Tlie proton nuclear magnetic resonance spectrum of 2,2-dinicthylhydrazinodiethylalane indicates that only one molecular species is present and shoxjs the AI-CH2-CLI2 and H-CHo structure peaks. Because of the uncertainty in position associated with the N-Il group, and in addition some overlapping of peak areas, the data obtained were insufficient to ul,c in assigning a definite structure on the basis of the spectruia alone. It is felt however, that if the structure involves the four iiierabered ring, there night be expected txjo fonac (cisand trans-) in a mixture, and the n.m.r, spectrum v/ould certainly indicate this. Infrared data give circumstantial evidence for the six menfcered ring. The position of the W-II stretch in 2,2-diraethylhydrazlnodiethylalone corresponds closely to that observed for tricovalent nitrogen confounds which contain the N-II group. The four membered ring does not contain this particular arrangement since it is the nitrogen atom involved in the N-H group which coordinates the aluminum atom nnd thus becomes tetracovalent. Little importance can be placed upon these observations, however, since hydrogen bonding in the solid sample vzould tend to equalize the environments of the nitrogen atorae in question.

PAGE 25

? I 0) u ^_ / ^~f h s_;l„l. r ^ L_ oouij-|}]uisunaj, jj

PAGE 26

V, > "Z ^ 19 n 2^1 lV_L 115 „ L L o.iut;)iims'iiu.ix J L

PAGE 27

20 As a possible solution to the problem of analyzias the infrared datn, a ssmple of 2,2-diinethylhydrazinodiethylalane was run as a 10 per cent solution in n-he:iane, where II bonding t/ould be minimized. The resulting spectrum (Figure 5) shows little, if any, shift in the N-H stretching frequency at approximately 3150 era.""'-. A new peak was observed at 1590 cm. which is in the region usually associated with the N-H defonnation frequency. This observation was noc explained. Support for the four-raembered ring structure might lie in the observed peal: at 1400 cm,"^ t/hich is generally found in compounds containing an aramoniian or substituted amnonium ion. Tliis too, however, could be a result of hydrogen bonding in either the four or six member ring structure. The prospects for elucidating the structure of this dimer b infrared means nppear rather bleak. Suggestions for fiirther woik along these lines might include a study of frequencies associated with various ring sizes which contain atoms of size similar to aluminum and nitrogen, perhaps cyclic silylhydrazines. Pyrolvtic condensation of 2T2-diinethvlhydra2inodiethvlalane In an experiment designed to test the degree of lability of the N-II bond in 2,2-dimethylhydra2inodiethylalanc, a pyrolytic condensation was performed. In the first experiment in this series the sacgjle was heated to 195^^0. at which point rapid evolution of gases caused the reaction vessel to be blown apart. The product of the reaction was a light brotm solid which was very brittle and apparently highly polymeric in nature. Analysis of this solid residue for carbon, hydrogen, and nitrogen gave a ratio which corresponds to C, ^H ^^ ^ „hich may be 1.0 3.1 1,0 "^

PAGE 28

21 formulated as N-N(CH ) thus indicating that even though the material was subjected to temperatures in excess of 350C. the dimethylhydrazino group remained intact. This material was highly opaque to infrared and had no melting point; thus, little characterization was possible.. In a second experiment the condensation was run at 150 C. over a five hour period. Ethane evolution was slow and easily controlled in the temperature range 110 C. (where evolution first is observed) to 160 C. The product in this case was a dark, viscous tar which was easily soluble in benzene and hexane. The material was analyzed for active ethyl content by aqueous hydrolysis and gave the following results. Found: C2H5, 25.52. Calcd. for [(CH3)2miAlC2U3] 25.46. It thus .ippears that the condensation occurs according to the reaction 150 n[(CH3)2NKIlAl(C2H3)2]2 = ^"^^2^^^ '^ ^^^^-^^ 2!-^^l^'^^2^^^ y and the hydrolysis may be represented by the equation [(CH3)2NNAlC2H5]y :2yll20 -: 4yH30'^" = yC2Hg -: y (CH3) 2NIIH3"'' + yAKHgO)^^'" Other evidence which ser\'es to substantiate this reaction sequence is demonstrated by the conspicoue absence of the II-II absorption bands in the infrared spectrum (Figure 6) Tlie presence of the 1,1-dimethylhydrasinium ion in the hydro lysed sample was shoti/n by adding aqueous sodium hydroxide until the pH was 10, and testing the vapor above the solution for a volatile free base (1,1-dimethylhydrazine) with moist red litmus, and for a reducing substance with a

PAGE 29

22 ^.

PAGE 30

23 drop of I\MnO, solutlcm on p. strip of filter paper. Both tests were positive, A third experiment was designed to carry out the condensation as slowly ss possible at a lower temperature in order to see if ethane could be removed without opening the ring or changing the degree of polymerization from a dimeric species. A saniple of 2,2-dimethylhydrazinodiethylalane was heated at 75 C./0.20 ram. for a period of 144 hours. During this time the colorless ; mobile liquid gradually became dark in color and grev; increasingly viscous. Iflien he reaction was complete, as evidenced by the disappearance of the N-H band in the infrared, the material had become a dark, viscous tar. A molecular weight determination was performed cryoscopically in benzene, itolecular weight found: 646. The formula weight of a [(CH2)2lWIj^lC_Hcl unit is 114.13, thus the polymer formed in this experiment has an average degree of association, y 5.66. Possible structures for this polymer include 1) rings linked together, 2) rings larger than six member, and 3) chains with end groups of an undetermined nature. In any event, the polymer appears to have the repeating unit I I Al N The information gained in this experiment indicates that the hydrogen atom in the M-H group in 2,2-dimethylhydrazinodiethylalane is less labile than that in 1,1-dimethylhydrasine itself (or in some intermediate molecular addition cowplex formed prior to 2,2-dimethylhydrazinodiethylalane)

PAGE 31

24 These -idditional data are better interpreted oia the basis of a six mevabered ring rather than on a four merabered ring. In the four laembered ring the aluminum atom is coordinated by the nitrogen atom which contains the N-H group H ^Al— N-Al N /\ The effect to be expected if the IT-I-I nitrogen atom acts as a Lewis base is labilization of the N-H bond, which is not the experimental result. Attempted de-diracrization of 2.2-dlmethylhydrazinodiethylalane by adduct formation A saiaple of 2,2-dimethylhydrazinodiethylalane was mixed with a twofold excess of unsymm -dimethyl* hydrazine and heated to 65C. for five hours. The product of the reaction v/as a light stravj liquid after the excess unsymm -dimethylhydrazine had been removed by pumping at 25C./0.20 lam. for one hour. The desired reaction may be represented by the following equation: [(CIl3)2MIL^l(C2H2)2l2 "' (CH3)2m^H2 2(Cll^)l7JimiC2li^)2'^](Cll^)^m^ A possible reaction that was thought unliicely vjould proceed via ethane elimination [(CIl3)2mri]/l(C2il3)2]2 •' 2 (011^)21111112 = 2 [(CH )2mni]2A.lC H^ ^^^K

PAGE 32

25 Analysis of the product of the resction disclosed, however, that the only process that had occurred was that or partial condensation to [(CH3)2MMC2H5] Found: N, 20.05, 20.39. Calcd. for Ucn^y^imU'liC^ll)]: 19.43. Calcd. for [(CH2)2NNA1C2H5] : 24.56. A change in the peak height of the K-H band in the infrared spectrum seirved to confirm the course of the reaction. These data give an indication of the strength of the dative Al-N bond in the diraer and also seems to show that 1,1-dimethylhydrazine is capable of catalyzing the pyrolytic condensation. Attempted de-dimerization usinn phosphines Experiments in which tributylphosphine and triphenylphosphine were heated with 2,2dimethylhydrazinodiethylalane resulted in dark, uncharacterizable, viscous liquids. T he interaction of tricthylaluminum with methylhydrazine In an experiment identical in detail with that performed in the synthesis of 2,2-dijniethylhydrazinodiethylalane, 1.569 grmis (0.0341 mole) methylhydrazine was reacted with 3.087 grams (0.0341 mole) of triethylaluminum. Very sIot/ dropwise addition was carried out at -7GC. with warming for melting and mixing betv/een additions. It was imm ediately apparent that the reaction was very exothermic (more so than V7ith 1,1-dimfithylhydrazine) and that a somewhat different product was being produced than was found with 1,1diiiKthylhydrazine. Tlie amount of ethane evolved approached three times the theoretical and the reaction product was a brovm, inhomogeneous solid.

PAGE 33

26 The total volmne of ethane evolved v/as 2,029 liters. Tlie amount cnlculsted for the reaction M^ic^n^)^ -r 2CH2KIMI2 = 2\R^mHcii^)M(C2ii^)^'^ ;202% was 0.763 liters. For the reaction AI2 (021-15)^ -!2CH2in-mH2 = 2[A1N(CH2)H] V 6C2H5 however, the calculated volume of ethane was 2.283 liters. The quantity observed was 37.5 per cent of that required for the elimination of 3 moles of ethane per nole reaction unit. The solid wgs collected and any excess triethylaluminum was removed by pumping at 25 C./0.20 mm. for five hours. An attempt to prepare a Nujol mull for infrared analysis resulted in spontaneous combustion of the solid upon contact with air. iiHiether this was due to the material itself or unremoved triethylaluminutm is not Imo^m. The infrared spectrum was obtained on a mull prepared in the dry box and was very opaque, vrtiich is characteristic of highly polymeric materials. A very weak W-H peal: showed that the reaction had not gone to completion. Analysis Found: Al, 37.15 per cent. Calcd. for CH^H2A1: Al, 33.55 per cent. The material was found to be insoluble in all the solvents tried, and dissolved only with very great reluctance in hot 151! nitric acid. Tlie nitric acid solution was highly colored, but was decolorised with hydrogen peroxide in order for aluminum analysis to be run on the solution.

PAGE 34

360"C. 27 The solid has no melting point, but chars i';hen heated to Discussion Although there were not a large amount of data collected in our study of the interactions of Ijl-ditaethylhydrazine and methylhydrazine with triethylaluminun, certain interesting differences were noted in the behavior of the ttro different hydrazines and in the behavior of hydrazines and amines with respect to their interactions with aluminum alky Is. One pleasing result of thit, work i£> the synthesis and characterization of the new compound, 2,2-dimQthylhydrazinodietliylalane. Althoxjgh a recent publication (21) reports methylaluminum derivative of hydrazines, this compound is tlie sole known cthylaluminum derivative of a hydrazine. An idea xThich xras a motivating factor in the initial phases of this X7ork concerned the possibility of preparing a solid adduct between triethylaluminum and 1,1-dimethylhydrazine. Both the alane and the hydrazine arc mobile liquids and both have been used as liquid propellents in the missile industry and in the U. £. space effort. For many reasons solid fuels are preferred to liquids in certain types of applications. Tlius, a solid adduct formed from the two liquids might well be an interesting fuel. From the preceding description of the observations made of the interaction of triethylaluminum with 1,1-dimethylhydrazine it can be seen that no adduct was found in the temperature range studied. It is possible, of course, that work at very low temperatures would

PAGE 35

28 produce evidence indicatinc the foinnation of such an ndduct, but in sny case no such compound exists above -60C. Similar results were encountered in the reaction of triethylaluminun v/ith Ejonomethylhydrazine. Of the various nitrogen donor-slane systems which are knovm, those systems containing a hydrazine are certainly the most reactive. Of the reported amine-alanes where the amine may or raay not contain the NH2 group, it may be generally said that the adduct formed between the amine and the alpjie is stable at room temperature. In the case of hydrazine-alanes however, none have been reported which contain the NH2 group e:xept (CH3)2lMIAl(CH3)2:KH2N(CH3)2 which is not a simple hydrazine-alane, (21) An approach to a clearer understanding of the nature of the hydrazine-alanes, the hydrazinoalanes, and their condensation products may lie in a stepwise consideration of the interactions involved in the formation of these compounds. My acceptable mechanism must be consistent with the following knoim facts: 1) the reaction was carried out In such a way as to have an excess of triethylalummum present in the reaction mixture until addition of l,l-dinethylhydra2ine was essentially complete, 2) ethane evolution occurred at -60C. during the first half of the reaction, but was less rapid during the second half and required higher temperatures, 0-inOp.^ 3) the product is a dimer, and 4) the N-Il group a the product is less active than that in 1,1-diinethylhydrazine. Fatter and Bartocha (21) have recently made statements concerning proposed mechanisms of interaction and the structure of hydrazinoalanes. They believe that a plausible first step in the interaction

PAGE 36

29 of Ijl-dimethylhydrazine with an c liiraimaii alkyl is the coordination of tliG -nil2 end of the nolecule with the AIR3 imit, R3AI + II2N-H (01-13)2 = Rg^lrNHgN (0113)2 rather than coordination of the -11(0113)2 end, as shotm below: R3AI -1(CIl3)2MNH2 = R3Al:N(CH3)2mi2 Tlie reason for their choice is the fact that their preparations of dimethylhydrazinoalanes produce as a gaseous product, R-H, rather than R-OII3. From a consideration of the bond energies alone, N-H (92 kcal./mole) and N-C (66 kcal./mole), one might favor such an interpretation. However, if there are kinetic effects v;hich favor a different reaction, there is no reason to disraiss coordination by the -N(CU3)2 end of 1,1-dimethylhydrazine. If we first accept the premise that the-N(CH3)2 end of the 112^(0113)2 molecule is more basic than the -NH2 end because of the inductive effect of the methyl groups, H OIL, \ / ^ N II / \ H CHo then in the presence of the strong Lewis acid, AlEt3, we should expect the initial coordination a6 shown in equation 1 on page 30. In the presence of a large excess of Lewis acid, AlEt3, we would expect a second step (equation 2) to occur very quickly, to form the biscon^lex, Et3Al:mi2-N(0H3)2:AlEt3.

PAGE 37

30 I B •a

PAGE 38

31 >. u

PAGE 39

32 The decoit5)ositioti of the biacomplex to ethane end the AlEt^ adduct of the hydrazinoalane, Et^-^II-JIIN (0112)2 J AlEto, would follow according to eqi.vntion 3. At this point the ctoichionetry ic a 2:1 ratio of alane to hydrazine, and it ic at this point in the preparation ?7here the temperature at which ethane will evolve suddenly changes to a higher value. In fact it has been observed in the laboratory that the remaining addition of 1,1-dimethylhydrazine may be made quickly at -25C. and that no appreciable gas evolves until the mixture ic trarmed to C. lu the presence of additional free Ijl-diiticthylhydrazine addition would be expected according to equation 4, followed by cyclization as shotm in equation 5, and ethane eliaination to the final cyclic six ruecibered ring diner as in equation C. A suggestion for further work in this system appears quite obvious at this point. If the mechanism proposed above ic correct there should exist the possibility of preparing Et2Ai-IIIEI(CIl3)2:AlEt2 by the reaction of the alane with the hydrazine in a 2:1 mole ratio. Tlie prospects for preparing any of the other intermediates do not appear promising, howevtjr. It is quite possible that Et2Al-NHK(CH)2:AlEtvrould not be isolated as a monomer since it could very well associate to form dimers or chain poljmiers; these species would probably react in much the same manner as the monomer i/ith 1,1dimethylhydrazine, although less vigorously. An interesting point is that if this material did exist as a monomer, it ^rould indicate that the need of aluminum for an electron pair was being satisfied by the adjacent nitrogen atom which has an unshared pair of electrons available.

PAGE 40

33 This observation would provide evidence for the existence of piIbonding between aluminum and nitrogen, which has been discounted by Laubengayer (30). Condensation of 2.2-dijnethylhydraginodiethylalane Since it has not been observed that ethane is intraiaolecularly eliminated, a process which would result in the formation of [(CU2)2^^^^^2^5^2* with o molecular v/eight of about 223, it appears reasonable to assume that the mechanism of the condensation involves attack of an IT-H group on the Al(C2Hr)2 group in 2,2-dimethylhydrazinodiethylalane. This condition could only be fulfilled if the II-H group has an unshared pair of electrons and can act as a Lewis base; thus it does not appear that the four necijcred ring structure, page 17 can participate in such a condensation reaction. In the presence of excess 1,1-dimethylhydrazine it was found that the condensation proceeds at a lov;er temperature than is observed XThen pure 2,2-dimethylhydrazinodiethylalane is condensed by heating. Thus it appears that the presence of N-H groups (with free electron pairs) is necessary for the condensation reaction. These observations leave doubf. that a dimeric condensation product will be observed since intramolecular elimination of ethane in the ring apparently does not occur, as evidenced by the observed high molecular weight of the condensation product. Condensation of tricthylaluminun with methylhydrazine The observations made on this system tend to corroborate the mechanism suggested for the interaction of triethylalianinum with 1,1-dimethylhydrazine.

PAGE 41

34 In njethylhydraziiXG we should expect, froa cons iderst ions of the inductive effect, that the nost acidic N-II bond is on the -NIL end of the molecule and that the stronger Lewis base is the -KCCIIg)!! end. II H \ / ili-U: / \ H Clio Cteice a methylhydraaine laolecule donates an electron paii-, hot/ever, there is cone question as to which IIII bond is most acidic. y*a initial attacr. of nethylhydrazine on triethylalurainuni ^TOuld produce the ciolecular coraplcx Et2M:IIIK;il2lIIl2, njhich could either eliminate ethane intraor intenaolecularly. Hie inhomoseneity of the product may serve to indicate that both these processes occur, or that at any rate, molecular species are produced in various degrees of association. Considering the abundance of active N-H groups in the initial adduct, it is not surprising that coi!5>lete condensation to [AllKCIIg)!]] is the final product. Although it is not implicit in the forraulation of this polyiaeric species, it is expected that altrainum achieves a covalency of four, either by accepting a pair of electrons from a nearby nitrogen atom or by accepting a pair of electrons from an adjacent nitrogen atom. The fact that the polymer appears to be a highly extended network would lend support to the former suggestion.

PAGE 42

35 Concrasion The results of our study of the interactions of triethylaluminua with two different alkyl hydrazines has resulted in the synthesis of three new species, [(CH3)2NNHA1(C2H5)2]2 [(CH3)2NKAlC2ll5]j^ [A1N(CH3)hI and has given various data through observation of react ivites, stoichioaetrieSj and reaction ccmditions. Sone physical characterisation was possible, and all available tools were enqployed in an attempt to elucidate the structure of 2,2-ditaethylhydrazinodietliylalane, v/hich is thought to have a six member ed ring structure. A mechanism for the formation of tliese species consistent with the observed data is suggested and applied to both this ^x)rk and some additional work reported in the literature. Although no evidence for pi-bonding between aluminum and nitrogen has been found, it is certainly within the realm of possibility that such bonding may exist. Preparation of the intermediate, Et2AlNHN(CH3)2:AlEt35 in a monomer ic form i-TOuld lend more credence to the possibility of aluminum forming pi-bonds with nitrogen. Further study is indicated.

PAGE 43

CHAPTER II REACTIONS OF ARYLHALOPHOSPHINEL; AIJD DERIVATIVES OP ARYLHALOPHOSPHINES WITH SEVERAL ALKYL HYDRAZINES Introduction TliG field of phosphoruE-niCrogen chetaistry has been periodically reviewed and there are two oajor reference texts (32, 33) and an excellent review article (34) which although not chiefly devoted to phosphorus-nitrogen chemistry certainly provide a foundation for the trorker in this field, ifo up-to-date listing of phosphorus-nitrogen compounds is available, however. Particularly lacking are reference summarieE in the area of the hydras:ine derivatives of phosphinoc and phosphorus acids. A thorough coiapilation of all kno^m hydrazine derivatives of the phosphines and related compounds was prepared in order to detsrraine the extent to which these materials have been studied and to what use the informe;ion has been put. As a result of this survey it has been found that hydrazinophosphine, H2PNHNK2J ^^^^ r^ever been reported, nor have any of its organic derivatives, liost of the v7ork with hydrazinophosphorus compounds has been in the area of hydrazine derivatives of the esters of phosphorus acids. There exist, on the basis of the nwre common substituente for phosphorus, a great variety of possible classes of compounds containing the hydrazinophosphorus group, -P-N-N-, but the ctirvey shows that work 36

PAGE 44

37 has been done on only a fev; of the possible series of such conq>ounds. Many of the compounds reported in this dissertation arc the only known meiabers of their series of corapounds. The general method of preparation of hydrazine derivatives of organophosphorus compounds is analogous to the methods used for the syntheses of aminophosphines and aminophosphorus compounds (35,36 ). The hydrazine is usually dissolved in an anhydrous solvent and the resulting solution is added to a halophosphorus compound dissolved in the same solvent. Excess hydrazine can be used to absorb the hydrogen halide produced in the reaction, or a tertiary amine such as triethylamine or pyridine may be used for this purpose. Example (37): 1 1 (Pho)2Pci -' 2N2H4 > (Hio)2Rnra2 -^ Gi2%)ci The monohydrazinophosphorus compounds which have been reported in the literature are listed in Tables 11-14 along with pertinent data; bisand tris-hydrazinophosphorus compounds comprise Tables 15-17. Recent ^rork has shovm that stepwise substitution of chlorine can be obtained in some arylphosphorodichloridothioates by partial Bolvolysis (38) : S S CI 1 1/ PhO-PClg + 2N2H^ FaO-P ^ (N2H5)C1 mmii2 The remaining chlorine is found to be less labile as a result of the substitution of the less electronegative hydrazino group and is

PAGE 45

38 thus resistant to further so Ivo lysis. In some cases the compounds can be water-washed withouc appreciable hydrolysis, Arylphosphorochloridohydrazidothioates are listed in Table IG, The second chlorine in these molecules can be made to undergo solvoiysis at higher temperatures and in the presence of water, form arylphocphorohydrazidothioic acids, t;hile in the presence of amines the various arylphosphororjaidohydrazidothioates are fomed (see Tables 19 and 20). Example: S S ^ 100C. ^ HiO-P-NHNIi^ -: 2R2NH > PhOHIHNil2 "!(R2Mi2)Ci CI 1.1R2 An effect of changing the reactant ratio has been observed in the fomation (in small yield) of a cyclic compound (39): S Nilmi S \ / \ / 2PhOPCl-16N^H, ^ P P KiO NH-NH OPh + 4(H2l%)Cl m.p. 1G3 C. A scheme devised by A. llichaelis is responsible for the Ejnathesis of amino-b is (hydra sino)phosphine oxides (40). lie found that amidophosphonic dichlorides can be prepared by ref luxing a mixture of phosphoryl chloride and a secondary amine hydrochloride until HCl evolution ceases:

PAGE 46

39 TABLE 11 Alkyl and Aryl Phosphorohydrazidatee Product Yield ra.p. Reference (CgH^Oj^HNHPh n.a. 113-114 41 9 (PhO)2PNHNH2 n.a. 112 42 (PhCH20)2PNHNH2 95% 73 43 9 (CH2=CHCH20)2PtminiHi 767, 85-87 42 (CH30)2^NHiniPh 937. 132-1340 42 CH30(C2H50)PNHNHHi 82% 77.5-80.5 42 n.a. not available.

PAGE 47

40 XABLE 12 Alkyl and Aryl Ehosphorohydrazidothioates Product rield m.p./b.p. Reference (PhO)2Pl>IHm2 (PhO)(CH30)PNHNH2 (Ozl}^ ^0)(CH30)RqHIJH2 (CH30 /^ xVo) (CH30) liraNHg 0)(CH30)PNmiH2 0)(CH30)PIIHmL HoC. ( <^ Vo)(CH30)PNHNH2 ^CH, Br //^ 68% m. 63^ 3^44 78% no distillate 45 at 0.01 mm. I^ m. 103 45,46 46 m. 76-77 m. 85-87 46 46 46 (tert.-C4Hc)-<^ H0) (CH3O) PNHl^Hg — m. 92.5-93.5 46

PAGE 48

41 TABLE 12 Continued Product Yield m.p./b.p. Reference Cl^ ^Cl ( // NVOXCHgOkmiHj — m. 88-88.5 46 CI (// \\.0)(C2H50)PNHNH2 85% b. 147-150 45 (OjN^ y-0)(C2H50)P(S)l>raNH2 63% m. 80 45 Cl^ c iy/\^0)(C2ll^O)hmm2 91% no distillate 45 \—-/ at O.Oi mm. Bg CI ai3C)2N A=i/ NH ^=^ N(CH3)2 m. 117-118 47 NH2

PAGE 49

42 TABLE 13 Alkyl and Aryl Riosphorophenylhydrazidothioates Product Yield m.p. Reference (C2H50)2^NHNHPh — 68-69 41 S (PhOXCHgOkHNHPh 42% 80 45 S (PhO) (C9ILO) fe^INHPh 35% 570 45 CI Gl '{J S '<3 —

PAGE 50

43 TABLE 14 Phenylhydrazidoallcylpliosphonates Product Yield m.p. Reference 9 C2H50-P(CCl3)Nm'mPh — 154-156.5 42 O y-0?(CH2-<^ y)NHNH^ y 173-1740 48,49

PAGE 51

44 TABLE 15 Bis (hydrazmo)phenylphosphine Oxides Product Yield m.p. Reference PhP(NHNtl2)2 PhP(NHNHPh)2 ClU H3C-<^ y^(IKINH-<^ \\)2 H3C H3C-/ y-^(NHWH-<^ %)^ 131 175 208 171 50 48 51 49 A Bis(hydrazino)phenylphosphine Sulfide Product Yield m.p. Reference PhP(mEfll2)2 115' 50

PAGE 52

TABLE 16 Arylphosphorodihydrazidothioates 45 Product Yield m.p. Reference PhO-P(NHNH2)2 24% 950(103) 43,52 PhO-P(miNHPh)2 OP (miNH2)2 02N\. /0-P(NHNH2)2 OgN-C _^ )-O.P(HHNHPh)2 CI //\yo-p(miiiH2)2 ^^^ ci ClT/ \yo-P(NHNH2)2 CI CI Cl/^ y-0-P(l^HNHPh)2 90X 87X 1360 106O lo 176 93.7% 145-1470 93.5% 152-153 96.7% 156-157 53 52 45 45 38 38,54 38,54 CI "^'""" CI 73X 158-159 38

PAGE 53

46 TABLE 16 Continued Product

PAGE 54

47 TABLE 17 Tr is (hydrazine) phosphine Oxides and Sulfides Product Yield m.p. Reference OKIDES \ 757. 43 P(NHNHPh)3 P(NHNH-/ V SULFIDES CH3). 196 189' 55 55 I (NHim2)3 unstable 44 P(NHNHPli). 15Ar 55 P(imNH-<^ ^yCH3)3 unstable 55

PAGE 55

TABLE 18 Arylphosphorochloridohydrazidothioates Product Yield ra.p. Reference PhO-P-lfflK(CH3)2 CI CI s (/ yo-P-NHI^ (0113)2 Ci ^^ S ci^ y-o-p-miN(CH,) CI 97.1% 95.8% 3)2 99.0% 74-750 38 38 38 CI 0.^ 0-P-iraiJ (0113)2 84% CI 38 CI Cl// ^0-PNHN(CH3)2 100% 76-78 Cl Cl 38 Cl Cl// \)-C-P-lJHNHPh Cl / Cl 79% 120-122 38

PAGE 56

49 TABLE 19 Alkylphosphorohydrazidoic Acids* Product Yield m.p. Reference C2li50-P-IEra-<^ y-CH3 -• 195 (d.) 52 OH C2H50-P-traNH-<^ y-Br -• 187 52 OH 1 Cgl^O-P-m-EIHPh -* 192 (d.) 52 OH 1 C2H5O-P-KHNH2 — 100 52 OH *Eee TA5LE 3 for examples of salts of acids of this type.

PAGE 57

50 TABLE 20 Arylphosphoroffinidohydrazidothioates Product CI C 1 0-P(mi,)lM>IHHi CI CI Cl-<^ y-0-P(NHCil2)tJHHH2 CI CI ci-K^ y0-? (NHCH3) mm (CH3) 2 ci Cl c 1 -<^ y 0p (NHCH3) ffiMiPh Cl Yield ra.p. 75% 86-87 Reference 38 37% 137-139 38 90.5% 104-105 38 100% 131-132 38 44% 112-113 38

PAGE 58

51 1 1 PCI3 -'' (R2lJH2)Cl > R2NECI2 -;2HC1 The aiaino-biG(lQrdrazino)phosphine oxide ie the product obtained upon hydrazinolycic of the araidophosphonic dichloride: ^ 1 R2NPCI2 -i4ri2NNliPh R2KP02IHi)2 + (PhlII:El3)Cl Tables 21 and 22 list the reported amino-b is (hydrazine) phosphinc oxides and sulfides. M exaiaple of another reaction which produces mixed aiainohydrazino derivatives of arylphosphonic acid involves a transamination reaction (56) : IHL, 1 1/ ^ in£m2 Ifydrazino lysis of the phosphonitrilic chloride trimer is also knoxm ^7,58): H3N2 N2H3 \ / (Cl2PN)3 + 12N2H^ ^ /P^ -i60^2h'^^'^ H3K2 ll 1/hh ^P P %% ^1/ "^^2113

PAGE 59

52 Reactions The reactions of the hydrazine derivatives of organophosphorus compounds arc interesting in that they may help to determine the structure of the compounds and in many cases lead to entirely new classes of compounds. The reactions are frequently troublesome and laay occur as cide reactions during the preparation of the desired compound ajid thus lower the yield. The P-H bond is susceptible to hydrolysis and the degree to which this occurs depends in large part on the nature of the substituents on the phosphorus atom and whether or not the phosphorus is in either of the oxidized states, the oxide or the sulfide. In some cases, especially when the compound contains an ester group, it is not the P-U bond v;hich undergoes hydrolysis initially, but the ester group (37,43): ^ r ^ n (HiO)2H-imni2 !NaOII (aq.) > Ka LH1OKJIMI2] -!PhOH Ir Further hydrolysis v/ill yield a salt of the phosphorus acid r 1 1 Na[PhO-P-lIffini2j -:• HaOH (aq.) >Na2[02RnMi2] r and the free acid can be obtained by nffitathesis: r ^ "I Na2 [02HnitIH2] -r 2ii • -^ (110)2 PIIIEIII2 -iZNa"*'

PAGE 60

53 Con5>lete hydrolysis is obtained by prolonged boiling in aqueous sodium hydroxide and yields free hydrazine and the phosphate ion: 1 [OjPNmraj] -r OH >PO4 + N2H4 Table 23 lists some hydrolysis products obtained in this manner. Ifydrolysie does not always occur in the manner described above, but may innediately attack the P-N bond as in the illustration below (50): 1 1 Ph-P(NHNH2)2 -!" 2H2O > PhP(0H)2 ''• 2N2H4 In this case hydrolysis occurs so readily as to preclude the existence of the hydrazine derivative in the precence of water. "Bxe former behavior is typical of esters of phosphoric acid and the latter is generally observed for derivatives of phosphonic acid. Hydrolysis is a con5)eting reaction when a hydrazine derivative of an organophosphorus ccsapound is treated with a chloroester, consequently low yields are to be expected for a reaction of this nature (50) : S 8 I II 1 M NaOH EliPC^tm2)2 -f2CI-C-OC2H5 2HC1 ^ PhP(NHITIffiOC2H3)2 m.p. 133C. (46 per cent)

PAGE 61

54 T-ABLE 21 Amino-bis (hydrazine) phosphine Oxides Product Yield m.p. Reference Primary C2H5lIH-P(WHNHPh)2 n-C3l-l7 MJ (WIE-mPh) j isoC/,Hr.NH-^(NHNHHi) 2 II-C5 Hj jNHP (IJHNHPh) j Secondary (CH3)2H-P(NHNHEh)2 — 194-195 40
PAGE 62

55 TABLE 22 Araino-b ic (hydrassino) phosphine Sulfides Product Yield n.p. Reference Primary iBO''C f^ HoHHP (NHNHPh) 2 — 129 40 Secondary (C2ll5)2N-P(NHNHPh)2 •• — 40 S (n-C3H7)2N-P(in^iraPli)2 •196<* 40 /E\N-P(IfflNHPh)2 *• 158 40

PAGE 63

56 TABLE 23 Salts of HiosphorohydrazidateE Product n.p. Reference Ka (PhOP02NIINH2) Na2(0P02NHKH2) Na2 (OPOjIJHNHj) • H2O Na(HOP02miNIl2) K(HlOP02NHim2) K(HOP02NHNH2) NH4(PhOP02NHKH2) Ba(PhOP02MMH2)2 Ba(OP02NHNH2) Pb(HiOP02NHNH2)2 Pb (0P02NffiIH2) 37, 43 37 43 37 43 37 37 37 37 37 37 1 PhCH20HrtEra2 OH Na (PhCH20P02NHli2) K (PhCH20?02NHNH2) 43 43 43

PAGE 64

57 Hydrazone fornation has been obeerved in compounds where the water produced in the reaction does not appreciably hydrolyze die reactants or products (43 ) : CH1 II 1 / ^ (PhCH20)2HraKH2 ^Cn3CCH3 > (PhCH20)2PNIttI=C + ILO 0% Q-p. 109C. Other such hnovm hydrnzones are listed in Tables 24 and 25. Both aldehydes and ketones have been used to prepare such hydrazones and the reaction is apparently ceneral for hydrazine derivatives which contain the -Mi^ group. iJIien an arylphosphorohydrazidate is treated with anhydrous hydrogen chloride salt fornation is observed (43) : 1 I "^ 1 (Pho)2Hniini2 •;IlCl ^ [(PiiO)2HnEm3jCl 1 ia.p. 150C. (dec.) This reaction is analogous to the fornation of hydrazinium salts and the proton attack invariably occurs on the most nucleophilic nitrogen atom (59). Quartemization reactions using methyl iodide have been reported, but the reactions described are not always similar to the reaction described above with hydrogen chloride. Instead, it is found that some nucleophilic centers v/ill quartemize in preference to the nitrogen atoms contained in the hydrazine group (47).

PAGE 65

58 TABLE 24 Benzylidene Derivatives of Alkylphosphorohydrazidates Product m.p. Reference (CH30)2KJ(CH3)N=CH<^ "^Cl 69-71 (C2a50)2PN(CH3)K=CH^ yCl CI (C2H50)2KIHN=CH- 122-123 CI CI 60 CI (CH30)2PNHN-CH -<^ y 123-124 60 CI (C2H50)2PMHN=CHV_y 52-53 60 CI 54-55 60 60

PAGE 66

59 TABLE 25 Itydrazones of Bis (hydrazine) phenylphosphine QxidesHydrazones of Bis (hydrazine) phenylphosphine Sulfides Product n.p. References Ph-P(HHN=C(CH3)2)2 170 58 Ph-P(Wm^=C(CH3)-CH3); Ph-P(roE^-CHV >Cn3)2 171" 58 s PhP (NHN=C (CH3) 2) 2 155 r\^ Ph-P(NHN=C(CH3)^ 7C1)2 162 ^^ S 1 Hi-P(iraN=C(CH2) 40112)2 133 53

PAGE 67

60 (H„C)^K 7; ^^-^-^-7^ ^"^'^^"'5>2 "' 2CH3I -I2H2O > s + 1 ig-aHgO m.p. 156-150G. (dec.) The final reaction to be mentioned here is that of condensation. This type of reaction is potentially very prouicing as a preparative method and occurc with the intermoiecular elimination of hydrazine at elevated temperatures (37) : 1 2(PhO)2Hn^l2 150"C. > (HiO) 2 KTIILIIIP (OPh) 2 m.p. 14G C. Noll 2"4 The product of this reaction was synthesized by another route in order to confirm its identity (43). With respect to the practical applications of hydrazinophoephoruG compounds, several patents have been granted t^ich relate to the use of these materials as insecticides, fimgicides, ncmatodicides, and fertilisers.

PAGE 68

61 Experic^ntal and Results llaterirls Ilydrazines I l-Diraethylhydrazine and methylhydrazine are both comnercially available. The Epjaples used in this work uere purified prior to use by distillation from calcium hydride. Tlie reagents thus purified possessed very narrow boilins ranges: 1,1dinethylhydrazine, 62.2-G3.0C./753 laa. methylhydrazine, 37,388.0C./761 lii-i. All hydrazines were stored in air-tight glass containers in a cool, dark location. 1,1,2-Trimethylhydrazine vras prepared from 1,1-diiaethylhydrazinc by the method of Class, et al ( 61 ); siiailarly, 1-cthyl2,2-diinethylhydrasine was prepared as reported by Klages, ejt al (62 ). Both were distilled from calcium hydride or lithium aluminum hydride and boiled in the ranges 50-62 C. and 92-93 C, respectively. Identity in each case was confirmed by comparison of the infrared spectrum with that reported in the literature ( 63 ) Triethylamine v/as purchased in tlie highest available purity and then refluxed over calciun hydride and distilled; the fraction collected was in the range 03.5-09.5C./759 am. Eliosphorus compounds were obtained conmercially and in most cases were used as received, llnny oxygenand laoisture-sensitive phosphines deteriorated v/ith age once the contaiLner load been opened and these were vacuum distilled prior to using. Solvents Reagent grade solvents were used tliroughout. IJhere drying was required the usual drying procedures were used. Tiic dry solvents were then distilled. Every effort was made to avoid absorption of moisture by these solvents.

PAGE 69

62 Nitrogen. Hitrogen was used to provide a dry, inert atmosphere wherever it was called for, Water-pun?)ed nitrogen was purified by passing the gas over metallic copper turnings at 400C. to reoove oxygen and then through anhydrous magnesiucj perchlorate to remove moisture. Equipment In addition to the usual laboratory glassware, several pieces designed to carry out small scale reactions in the absence of air and moisture xfere used. Ace Mini-Lab apparatus (Figure 7) is an example of such special equipment. Transfers and handling under anhydrous conditions were facilitated by the use of a Lucite dry box in v;hich was maintained a dry, nitrogen atmosphere. Figure 8. Manipulations were performed through a pair of Neoprene gloves and every effort X7as made to avoid unnecessary opening of the box. Several dishes of phosphorus (V) oxide were placed in each stage of the box to absorb laoisture. Most elemental analyses were performed by the Galbraith Microanalytical Laboratories, Knosville, Tennessee. Some nitrogen analyses, however, were performed with a Coleman nitrog^i analyzer. Model 33. tfelting points were determined in sealed capillary tubes in a Thomas-Hoover melting point apparatus. Infrared spectra were obtained on a Perkin-Elmer Infracord tfodel 137 infrared spectrophotometer. Nuclear magnetic resonance spectra were obtained on a Varian High-Resolution Nuclear l
PAGE 70

63 Nitrogen Inlet stirrer K.y Addition Funnel Filtrate Receiver Thermometer Hsu re 7. Mini-Lab Reaction Apparatus

PAGE 71

64 J5 Q 2 o

PAGE 72

65 Experiiaents with 1 1-dioethylhydrazine and chlorodlphenylphosphine 2 2*Diraethy Ihydraz inodipheny Iphosphine 2,2-Diiaethylhydraziiiodiphenylphosphine was synthesized by the hydrazine lysis of chlorodiphenylphosphine as shown below: Ph2PCl + 2H2NNHe2 ^ Ph2R^INMe2 -Me2NNH2*HCl Similar solvolytic reactions are well known and have been used to prepare aminophosphines and other compounds in which it was desired to form a phosphorus-nitrogen covalent bond. However, this method had not been used previously to synthesize hydrazinophosphines The nature of the reactants are such that moisture and oxygen must be avoided. It is also desirable to use a solvent in which the desired product is soluble, but from which the 1,1-diinethylhydrazinium salt x^ill precipitate. Fifty-five and one tenth g, (0.25 mole) chlorodipheny Iphosphine was dissolved in 25 ml. dry benzene and added, with stirring and cooling, to a solution of 33 g. (0.55 raole) 1,1-dimethylhydrazine in 25 ml. benzene. The addition took four hours after which time the mixture was allowed to slowly warm to room temperature. The mixture was then stirred for an additional hour at room temperature to allow the precipitated crystals of 1 1-dimethylhydrazinlum chloride to assvime sufficient size for easy filtration. Filtration yielded, after washing successively with benzene and ether, 23.99 g. of a white, crystalline solid, m.p. 79-81 C. (literature value for 1,1-dimethylhydrazinium chloride, 81-32 C. (64)). This amount is 99.3 per cent of theory.

PAGE 73

66 Evaporation of the filtrate at roora temperature and reduced pressure gave 59.97 g. of white solid, m.p, 62-66C, This solid was dissolved in 175 ml. dry hexane at 70C. and the resulting solution was filtered. Upon cooling the solution to room ten5>erature, crystals formed; these were collected and found to weigh 55.0 g. and melted at 65-67*^C. Sublimation of this product at 60C./0.20 mm. gave long, prismatic crystals, m.p. 68.5-69,5C. The overall yield was 51.5 g. (34.5 per cent of theory bared on the equation presented above) Analysis Found: C, 68.65; H, 7.17; N, 11.29; P, 12.66. Calcd. for C14H17H2P: C, 68.33; H, 7.02; N, 11.47; P, 12.68. The infrared spectrum (Figure 9 ) and nuclear magnetic resonance spectrum (Figure lo) are consistent with the followiitg structure: Sh

PAGE 74

67 o E ooinui! 'iur..i L '„,

PAGE 75

68 Solvent: CDCl.^ Frequency: 56.4 mc. eak

PAGE 76

69 A solution of 2.94 g, (0.012 mole) 2,2-diiftetliylhydrazinodiphenylphosphlne in 50 nl. benzene xms heated CTvemight xAile a stream of dry air was passed over the solution. Upon ccmplete evaporation of the benzene there was obtained a white, crystalline mass and a dark oil. The crystals trere collected and recrystallized from warn benzene and then sublimed at 160C./0.32 lan. ; 0.94 g. (30 per cent of theoretical, based on 22-diinethylhydrazinodlphenylphosp^ine) of a wiiite, crystalline solid, m.p. 166,5-163 .0C. resulted. 2. Oxidation with activated M1O2. 1 Ph2PtniNMe2 -iM:i02 "^ Ph2PNIMIC2 + WO Two and ninety-four hundredths graos (0.012 mole) 2,2-dimethylhydrazinodiphenylphosphine in 50 ml. benzene was heated at 50C. with 4.1 g. (0.72 laole) activated manganese dioxide (65) for 12 hours. The mixture was then filtered and upon evaporation of the benzene there was obtained a crop of white crystals. This solid was recrystallized from benzene and then sublimed at 160C./0.30 oa. to give 1.60 g. of a ^ite, crystalline solid, n.p. 166. 5-168. 5C. (45 per cent yield, based on 2 ,2-dimctliylhydrazinodiphenylpIiosphine) 3. Reaction of diphenylphosphinic chloride with 1.1-dimethylhydraaine 1 i Ph2PCl + 2H2KNlie2 ^ Ph2RJHNMe2 "* lIe2l^NH2HCl

PAGE 77

70 Twenty-three and seven tenths grams (0,10 nole) diphenylphoe* phinic chloride in 35 ml, benzene was added, with stirring and cooling, to a solution of 13,0 g. (0,21 mole) l,l"diiaethylhydrazine in 20 ml. of dry benzene. When the addition was coE?>lete the miKture was heated to 70 C, stirred for one-half hour and filtered hot. l^on cooling, the filtrate deposited a white, crystalline solid, m.p, 155-164C. The 1,1dime thy Ihydrazinium chloride on the filter was extracted with hot benzene and the washings coabined with the filtrate, Eduction of the volume of the resultirsg solution gave additional solid, Recrystallization of the solid from 1:3 n-hexane: benzene solution followed by subliiaation at 140C,/0,20 lan. gave 21,6 g. of product, m.p, 167,0-168,5C, (G2.5 per cent yield, based on diphenylphosphinic chloride). Analysis Found: C, 64,39; H, 6,38; K, 10.73; P, 11,92. Calcd. for Ci4Hi7K2^= ^> 64.60; H, 6.58; H, 10.77; P, 11.90. The infrared spectrum of this product is identical with those obtained froai the products of atarospheric and IfaG2 oxidation of 2,2-diiaetlTylhydrazlnodiphenylphosphine. Mixed melting point determinations melted at 166-16SC, The infrared spectrum (Figure 11) and n,m.r, spectra (both 1 31 H and P) are consistant with this structural formula: fh H .1 .-X \ Eh' H \ CE P— W. yCH^

PAGE 78

71 o_ O.lUinil'HSUlMX

PAGE 79

72 2 .2-Diae t:hylhvdraaiaodiphcnylphoGphiae sulfide \ PhgHflESfeg -J1/8 Sq 5. Ph2PNffiC'le2 Three and forty-seven himdredthc grsaas (0.0142 nole) of 2,2dimefchylliydrazinodiphenyiphosphine was dissolved in 50 ml. dry benzene and added to 0.43 g. (0.015 mole) of finely divided sulfur in a small flask. The sulfur dissolved easily as the solution was warmed to GOC. The solution was heated at 60C. for 30 minutes and then cooled to roOTi ten5>erature; no solid appealed on cooling, l^n evaporation of the solvent a tAita solid, ta.p. 37-96C., was obtained. Recrystallisation from 1:1 benzene :n-hexane gave 3.59 g. (92 per cent of theory, based on the above reaction) of i^ite crystals, m.p. 95.5-97.0C, Analysis. Found: C, yO.67; H, 6.20; K, 10.14; P, 11.21; S, 11.60. Calcd. for C14H17N2PS: C, 60.35; H, 6.41; N, 10.26; P, 11.45; S, 11,41. Tlie n.Ei.r. and iafrared (Figure 12) spectra were consistent with the structural formula below: Hi S Cllq Ph CH3 2,^2-DimethvlhvdrazinomethvldiphenvlphosphoniuEi iodide. Me Fa2HniHife2 -iMel ^ [mi2mim^y]i

PAGE 80

73 i.)ui;ii!uiKui:.i.

PAGE 81

74 One and seventy-two hundredths g. (0,00704 nole) 232dtmethylhydrasinodiphenylphosphine and 1.0 g, (0.00704 mole) nethyl iodide were dissolved in 25 ml. dry ether and the solution was stirred at 25C. overnight. At the end of this time a solid was filtered from the solution and dried at room teioperature and reduced pressure. The white solid ireighed 2.71 g. (100 per cent yield, based on the above equation) and melted at 15G-158C, M attempt to subline tiiis material resulted in tloermal decomposition at 160C. The salt is solxible in absolute ethanol. /snalvsis Found: C, 46.85; H, 5.47; N, 7.10; P, 7.05. Calcd. for C15H20N2PI: C, 46.65; H, 5.22; II, 7.25; P, 3.02. A water-alcohol solution of this solid gives a positive iodide ion test, and iodine is liberated by the addition of nitric acid. The infrared spectrum (Figure 13) is consistent with the structural formula: Hi *f^3 /CH3 Ph ^CH3 The structure was further confirmed by basic aqueous hydrolysis to 1,1-dinethylhydrazine and methyldiphenylphosphine oxide. The oxide was identified by conversion to nethyldiphenylphosphinic hydrogen carbonate.

PAGE 82

75 9-xl

PAGE 83

76 Hydrolysis o£ 2.2-diiaethylhydra2inoinethyldiphenylphosphoniiHa iodide CH3 I -JOH ^ Hi. NHKMe2-' H.2P^ OH ^ Hi2P-*0 -!H2NHMe2 Two grams of sodium Irydroxide was added to 3 g. of 2,2dintethylhydrazinoiaethyldiphenylphosphoiiium iodide in 25 ral, of 1:1 ethanol: water solution and the mixture was boiled for one hour. As the alcohol evaporated it %jas replaced with water. The vapor above the solution was tested for the presence of free base (1 l-dimethy 1hydrazine) with danq> red litmus paper and for the presence of a reducing substance with a drop of potassium peinaanganate solution on a strip of filter paper. Both tests were positive. An oil separated from the aqueous solution. The amount of oil was too small for distillation, but an infrared spectrum consistent with methyldiphenylphosphine oxide was obtained. The oil was treated x^ith a solution of sodium carbonate at 90 C. for two hours and upon evaporation of the water a white solid residue was left. Tliis was extracted with hot benzene and filtered. Evaporation of the filtrate gave one gram of a white solid, m.p, 107-109C, This solid evolved carbon dioxide upon contact t/ith a drop of hydrochloric acid. The literature value for the melting point of methyldiphenylphosphinic hydrogen carbonate, is 109-111C. (66), and it is reported to liberate carbon dioxide upon contact with hydrochloric acid.

PAGE 84

n CH, CHc, I II I Ph,P-0-C-0-PPh. I I OH OH From the foregoing experimental evidence we can conclude that the alley lat ion of 2,2-dimethylhydrazinodiphenylphosphine with methyl iodide produces the hydrazinophosphonium salt rather than the hydrazinium salt indicated below: + [Hi2PIIHtKCH3)3ll '. Attempted alkylation of 2.2"diiaethylhydrasinomethyldiphenyl~ phosphonium iodide with excess methyl iodide Treatment of 2,2dimethylhydrazinomethyldiphenylphosphonium iodide with excess methyl iodide in ether or toluene (heterogeneous reaction) gives quantitative recovei^r of starting materials. It is clear, therefore, that, under the conditions cited here, alley lation of 2,2-ditaethylhydrazinomethyldlphenylphosphonium iodide does not occur. Other hydrazinophosphonium salts Saiiq>les of 2,2-dimethylhydrazinodiphenylphosphine were treated with various organic halides in an attempt to obtain additional information relevant to the ease of alley lat ion of the phosphlne. lleaction with benzyl chloride Four and eight hundredths grams (0.0168 mole) 2,2-dimethylhydrazinodiphenylpho6phine and 2,12 grams (0.0168 mole) benzyl chloride were dissolved in 50 ml, dry toluene and the mixture was refluxed at llO^C. for 12 hours. Upon cooling to room ten5>erature two liquid layers xiexQ observed. The toluene was removed and attempts vtoxe. made to initiate

PAGE 85

78 crystal Xizat ion by cooling and by adding ether to tlio layer containing the desired product. Jk> crystallization occttrred and the product could not be purified by crystallisation from absolute ethanol. The clear, yellow, viscous liquid gave a positive CI* test, hOTrever, and although the con5)otjnd wac not obtained pure, its infrared spectrum does indicate a salt-like structure ^ich contains the bonds expected for 2,2-diiaetl^lhydrazin6benzyldiphenylphosphonivmi chloride Ph I CH2 [Hi2-P-NHN(CH3)2]c1 Reaction with carbon tetrachloride Upon dissolving 2,2dimethylhydrazinodiphenylphosphine in reagent grade carbon tetrachloride there forms in the yellow solution a faint precipitate which gradually disappears upon standing. Although no corrqxjund was isolated, there is the possibility that alkylation occurs according to the following equation: CCI.5 I Ph2PNHN(CH2)2 -1CCl^^ ^ [ 1^2? -11^(0113)2] CI Reaction with phenyl iodide Oae gram 2,2-dimethylhydrazinodipheirylphosphine was mixed with excess phenyl iodide in dry ether and heated for one hour on the steam bath while the sample was protected from moisture with a drying tube. Several small crystals

PAGE 86

79 formed in the liquid and these were washed with ether and dried in the air. The melting point was 162-175C., and a nitric acid solution gave a positive l' test. No suitable method of purification was found. Sublimation attempts resulted in thermal decoiq>osition. The compound is thought to be 2,2-dlmethylhydrazinotriphenylphosphonium iodide [Hi3PNHN(CH3)2] I Reaction with b ^ lb 'dibromoethyl ether Two and fortythree hundredths g. (0.01 mole) 2,2-diinethylhydrazinodiphenylpho6phine was reacted with a threefold excess of P /3' dibromoethyl ether in toluene at 60C. for 5 hours. A seml-solld which was not purified was the only observed product. Synthesis of l.l-bis(diphenylphosphino)-2,2dimethylhydrazine An experiment designed to test x^ether chlorodiphenylphosphine would undergo hydrazinolysis by 2,2-dlmethylhydrazinodlphenylphosphine resulted in the synthesis of l,l-bls(diphenylphosphlno)-2,2-dlmethylhydrazine according to the following equation: Ph2P Mb PhoPNHlttie, + PhoPCl + EtoN ^^^ — ^ N — N ^ 2 Z :i / \ Fh2P Me -i[Et NH]C1 Three and fifty-four hundredths g. (0.0161 mole) chlorodiphenylphosphine and 3.62 g. (0.0353 mole) triethylamine were dissolved in 50 ml. dry toluene and to this was added quickly at room

PAGE 87

80 temperature a solution of 3.84 g. (0.0161 mole) 2,2-diiaethylhydrazinophenylphosphine in 50 ml. toluene. There was no immediate evidence of reaction. The temperature was slowly increased and at 50C. a solid appeared in the solution. Above 50C. the precipitation was copious. The mixture was stirred at 110C. for one hour and filtered; the precipitate melted at 251-253C. (literature value for triethylamoonium chloride is 254C.). Yield: 2.12 g. (96 per cent of theory, based on the above equation) Evaporation of the filtrate gave 6.51 g. of a white solid, m.p. 126-133 C. An attempt at siiblimation resulted in decomposition at 135C. Recrystal ligation from dry n-heptane gave fine, white crystals, m.p. 129. 5-132. 5C. in 76 per cent yield, based on the equation above. Analysis. Found: C, 72.65; H, 5.98; N, 6.39; P, 14.56. Calcd. for C2gH2^N2P2: C. 72.88; H, 6.12; N, 6.54; P, 14.46. The n.m.r. spectra and infrared spectrum (Figure 14) are consistent with the proposed structure, but a small absorption pealc at 1176 cm. in the infrared shows some oxygen (as ]^=0) as an impurity. Chlorophosphination of triethvlamine Since, as in the synthesis described above, it has in several instances proved convenient to use triethylamine rather than an excess of the hydrazine as a hydrogen chloride acceptor, it was desirable to determine whether or not chlorodiphenylphosphine reacts directly with triethylamine.

PAGE 88

81 o a. o.-juciiiuisin;.!.

PAGE 89

82 In view of the fact that chloramine has been shcnm to react with tertiary phosphines in accordance with the equation (67) ClNHg "IR3P > [R3PNH2]c1 it might be expected that chlorophosphines such as (€5115)2^01 would react with tertiary amines according to the following equation: + PhgPCl '1R3N > [R3NPPh2]Cl" In an experiment designed to test whether or not chlorophosphination of triethylamine occurs under the conditions usually employed in the hydrazinolysis of chlorodiphenylphosphine 5.46 g. (0.054 mole) triethylamine was dissolved in 50 ml. anhydrous ether and added to 11.90 g. (0.054 mole) chlorodiphenylphosphine in 50 ml. ether. An immediate cloudiness appeared in the solution wiiich persisted througho out a 30 minute ref Ivix at 40 C. Filtration gave 0.63 g. of a white solid, m.p. 254-255C., which was conq>letely water soluble. A mixed melting point deteirmination with an authentic saxiq>le of triethylammoniiim chloride melted at 253254C.; the infrared spectrum of this solid is identical with that of triethylamraonium chloride. Evaporation of the ether and triethlyamine fr
PAGE 90

It may, therefore, be conclioded that chlorophosphination of triethylamine with chlorodiphenylphosphine does not occur under the conditions eEf)loyed here for the hydrazinolysis of chlorodiphenylphosphine. It should be noted that a little triethylamaonium chloride resulted from the reaction mixture, showing that although all reagents had been previously distilled and dried, some hydrolysis had occurred. It Xi7as later found to be possible to avoid the formation of triethylammonium chloride upon mixing chlorodiphenylphosphine and triethylamine by performing all transfers in the dry box. For the usual bulk reaction, howev-ir^ it is unnecessary to tal;e the extra care to avoid this small amount of hydrolysis as it lowers the yield by only a fraction of a per cent. Reaction of 2.2-dimetIivlhydrazinodiphenvlphosphine with carbon disulfide Five ml. reagent grade carbon disulfide was added to 1.97 g. (0.0307 mole) 2,2-dimethylhydrazinodiphenylphosphine in 10 ml. anhydrous ether. A deep red color developed immediately and slowly faded to yellow as the solution was evaporated at 40^*0. over a five hour period. Upon standing overnight, large, white crystals appeared in the solution. These were collected and washed with hexane and melted at 140.5-141i5C. The yield was 82 per cent of theory, assuming the 2,2dimethylhydrazinodiphenylphosphine reacted with the carbon disulfide in a 1:1 ratio. The analysis corresponds to (C5H5)2KIHN(CH3)2-CS2: Found: C, 56.09; H, 5.55; N, 8.69; P, 9.46; S, 20.17, Calcd. for C15H17N2PS2: C, 56.23; H, 5.35; K, 8.74; P, 9.67; S, 20.01. It was foimd that if the white, crystalline product of this reaction is redissolved in carbon disulfide, the red color appears once more.

PAGE 91

84 The literature describes the interaction of tertiary phosphines with carbon disulfide, and there are reported compounds of the type S s which are red, crystalline solids (55,68 ). Tlie structures o£ these compounds have been confirmed by X-ray diffraction analysis (69 ) and it is well-known that the conpounds contain a P-C bond and no formal P-S bonds. The red color is thought to arise as a result of tlie Zwitterion-type structure. Since a red color develops in the interaction of 2,2-diinethylhydrazinodiphenylphosphine witli excess carbon disulfide, it appears that a Zwitterion coraplex is the initial product of the reaction: s.S +1 Ph2P-NHN(CH3)2 As the reaction proceeds, however, the color fades as the amount of carbon disulfide is decreased. Tlie final product is a white solid and has an infrared spectrum figure 15) which it; complex, but shows no absorption in the region assigned to the K-II bond, Tlie N-N absorption has been shifted to a higher frequency which indicates a change in one of the substituents on the nitrogen atom attached to the phosphorus; the -tKCH^)^ group appears to be intact. A weaic absorption is evident in the S-H region. The monosubstifcuted phenyl

PAGE 92

85 OOUUJI!^ JSUTJ.i I 'i

PAGE 93

86 group peaks are unchanged and the P-phenyl absorption has not been shifted at all. The structure best fitting the infrared data is dratra beloxi?: s s \ I / P N — N Ph CH3 The proton nuclear magnetic resonance spectriim is in general agreement with tliis structure. Four pealcs are observed, none of which is the characteristic N-H close doublet which is observed in 2,2dimethylhydrazinodiphenylphosphine. The pealcs assigned as phenyl protons and methyl protons agree with the areas expected for the structure given above. The remaining two peaks are of unequal area and are presumed to arise from 1) the S-H proton and 2) the proximity of the S-H proton to the ^^P atom, which is an odd nucleon and with which ^H will interact by spinspin coupling. The fact that a red color develops on dissolving this material in carbon disulfide may indicate that the phosphorus atom is still available for loose coordination in excess carbon disulfide, and gives a molecular coE5>lex as shown below. c c — s 1 / Ph2P — N K (0113) 2 Treatment of 2,2-dimethylhydrazinodiphenylphosphine oxide with carbon disulfide resulted in no reaction. Pyrolytic condonsntion of ^?-H-; xieehvlhvdr..^.-;nodiDhenv1p hnc£hine. It has been observed that t^hen a sample of 2,2-dimethylhydrazinodiphenylphosphine is purified by sublimation, there sometimes is

PAGE 94

87 lefc behind in the pot or the Bublimation apparatus a yellow, resinous material. On these occasions there is also found in the cold trap used to protect the vacuum pvnnp a small amount of a volatile liquid. Infrared analysis and vapor phase chromatography data indicate that this liquid contains dimethylamine and 1,1-dimethylhydrazine. Dimethylamine can be produced by a thermal decon5)osition which produces the pliosphonitrilic system, xPh2PNHN (0113)2 ^ (Pla2H0x "l" x(CH3)2NH and 1,1-dimethylhydrazine can be one of the products when 1,1-bis(diphenylphosphino)-2,2-dimethylhydrazine is also a pyro lytic condensation product: 2Eh2PNHN(CH3)2 3(Ph2P) 2^^(0113) 2 -! (CH3)2NNH2 It should be noted that only a small amount of this solid, resinous material is observed after sublimation of 2,2-dimethylhydrazinodiphenylphosphine, but a siiggestion for further work would be to investigate these reactions on a larger scale and identify with certainty the reaction products. Hydrolysis of 2,2-diiaethylhydrazinodiphenylphosphine A saii?le of 2,2-dlmethylhydrazinodiphenylphosphine was treated with 0.1 N KCl with the result that both 1,1-dimethylhydrazine and diphenylphosphinic acid (m.p. 191-193C.) were isolated in high yield from the product mixture. The diphenylphosphinic acid was filtered from the solution and the 1 l-dimethylhydrazine was distilled from the filtrate which was made basic by the additi(m of NaOH solution.

PAGE 95

88 Hydrolysis by atmospheric moisture was found to be a minor problem with 2,2-diiBethylhydrazinodiphenylphosphine. Experiiaentc with aethvlhvdrazine and chlorodiphenvlphosphin^ Reaction of oethvl hydrazine with chlorodjphenvlphosphine Ten and four tenths g. (0.0472 mole) chlorodiphenylphosphine was dissolved in 35 ml. dry ether and added to 4.35 g. (0.0945 mole) methylhydrazine (redistilled and dried over calcium hydride) in 40 ml. dry ether. The addition took four hours and vms performed with cooling and stirring under a nitrogen atmosphere. No precipitation occurred at 0C., but upon warming the mixture to 25C. solid appeared and the mixture was stirred at 25 C. to allow the reaction to proceed to completion. Filtration gave 3.71 g. methylhydrazinium chloride (theoretical is 3.88 g.. based on chlorodiphenylphosphine), and 9.57 g. of a yellow, viscous liquid which did not contain chlorine, as evidenced fay a silver nitrate test on a small portion dissolved in dilute nitric acid. No crystallization could be induced in the liquid and no way was found to effect purification. M,nlysis C, 67.55; H, 6.46; N, 11.96; P, 14.34. Calcd. for Cl3Hl5%PC, 67.81; II, 6.57; H, 12.17; P, 13.45. Tlie reaction product m^ have either of two forms / ^ Ph2PNHNHCIl3 or PhgN '^ Nlij The n.m.r. spectra indicate that a mixture of both these species is present in (he product, but the data are too complex to indicate the

PAGE 96

89 relative percentages of the constituents. The infrared spectrum contains all the expected absorption frequencies, but is of little value in determining per cent ccm^Kssition. An attempt to vacuum distill the prodiict resulted in thermal decomposition. A liquid fraction was collected in the range 93-99 C./0.36 mm.; hov^ever, the bulk of the product remained in the distilling flask as a resinous, amber-colored solid. Extrapolation of the boiling point at reduced pressure to the normal boiling point of the liquid On a temperature-vapor pressure nomograph gave ca. 175C./760 ma. (diphenylphosphine boils at 280C. per 760 laa. (32). Analysis Found: C, 77.42; H, 6.22; N, 0.25; P, 15.88. Calcd. for (CgH5)2PH: C, 77.41; H, 5.95; N, 0.00; P, 16.64. The molecular weight (cryoscopically in benzene) is 1S5. Calcd. for diphenylphosphine: 186.2. A methyl iodide derivative, prepared according to the equation -r Ph2PH + 2CH3I > [Ph2P(CH3)2]l -; HI has a m.p. 241C. The infrared spectrum (Figure 16) is consistent with the structure for diphenylphosphine and contains a very prominent absorption peal; at 2295 cm."^, which is characteristic for the P-H bond. The thermal deconq>o&ition of the product mixture apparently is according to the following equation: PI12PNHNHCH3 ? • Ph2PH Hcondensed species Ph2pNCH3NH2 J

PAGE 97

90 OOUK-nUUSUTI.IX

PAGE 98

93. An attec^t was laade to separate the two conponents iron each other by reaction with benzaldehyde, but the only isolated product wac nffithylbenzylidenehydrazine, C(^H5CH=K-NHCH3 n.p. 178-179 C. (literature value 179C.). Apparently the water produced in hydrazone formation hydrolyzed the P-IT bond in the product. Ito separation was made on the residue from the thermal decomposition and the analytical results did not agree with any single coEqjound. Tlae infrared spcctnjn was highly opaque, which is characteristic of polyneric materials. Reaction of methylhydrazinodiphGnylphosphine with sulfur The mixture of methylhydrazinodiphenylphosphines produced in the reaction of chlorodiphenylphosphine with methylhydrazine was reacted with a email amount of finely divided sulfur in benzene solution with the result that hydrogen sulfide evolved and a dark gum was produced from which no product vras isolated. R.eaction of octhvlhydrasinodiphcnvlphosphine with carbon disulfide l^n dissolving the methylhydrazinodipyenylphosphine mixture in carbon disulfide there was no immediate evidence of reaction, however, over a 12 hour period hydrogen sulfide evolved, and upon removal of the solvent a dark gum remained from which no pure material was isolated. Synthesis of l,1.2-tris(diphenylphosphino)aethvlhvdrazine PRio SHigPCl !CH3NIINH2 -!3(C2H5)3lI ^ (^2^)2^^'. "'" ^ (02115)3^1 CI CH3

PAGE 99

92 Eleven and ninety hundredths g, (0.054 raole) chlorodiphenylphosphine and 7.24 g. (0.0732 mole) triethylaiaine were dissolved in 50 jal, dry toluene (b.p. 109 1-109. 2C,) and to this was added slocrly with stirring aad cooling 0.83 g. (0.018 mole) methylhydrazine in 35 ul. dry toluene. Wien the addition was complete the mixture was heated to 100C. and stirred for one hour and filtered hot. The tjight of tricthylamaoniua chloride on the filter was slightly laore than theory based on the above equation; apparently soiae of tlie toluenesoluble product was occluded in the salt. Upon evaporation of the solv.Tit at room temperature and reduced pressure, there was produced a yellow gum. Tlirce hundred ml. of dry ether was added to the gvna and a white solid separated from the resulting solution. This solid was washed with additional ether, collected, and dried under vacuum. The axaount collected was 2.55 g. (21.8 per cent yield, based on chlorodiphenylphosphine), m.p. 151C, dec. Purification was effected by recrystallization from acetone in the dry box. The resulting crystals melted at 152. 3-153. 0C. (dec). Analysis Found: C, 74.33; H, 5.71; N, 4.75; P, 15.41. Calcd. for C37II33II2P3: C, 74.24; H, 5.56; N, 4.68; P, 15.52. The infrared spectrum (Figure 17) of this material is consistent with the structure PhoP PPh, Pb-P ^CH3

PAGE 100

SB ooiarjiiuisuu.ij.

PAGE 101

94 No difficulty mjub experienced in handling this material in contact with the atmosphere, and a saaqple left out in the air for several hours underwent no detectable change. Experiments with trioethylhydrazine and chlorqdiphenylphosphine The synthesis of l,2.2"t:riraGthylhvdrazinodiphenylphosphine CH3 / R^jPCl -J2CH3NI1N (0113)2 > H12PN CH3 CH3NHH(CH3)2*HC1 N ^CH3 Eleven g. (0,05 mole) chlorodiphenylphosphine was dissolved in 20 ml, dry benzene and added slotjrly with stirring and cooling (to O^^C) to 7.5 g, (0,1 mole) 1,1,2-triiaethylhydrazine in 20 ml, dry benzene. The addition took 2-1/2 hours and at the end of this period little precipitate was observed r'.n the reaction mixture. Additional preo cipitate appeared when the mixture was heated to 50 C, and stirred for one hour. Filtration of the mixture gave S.'M) g, (98 per cent of theory based on chlorodiphenylphosphine) 1,1,2-trimethylhydraziniuEi chloride, m,p, 54-56C, Evaporation of the solvent Jai vacuo at room tojperature gave a clear, viscous liquid. No crystallization could be induced and vacuum distillation resulted in decomposition. No characterization was attengited. Instead, the oxide was prepared as a derivative.

PAGE 102

95 Atmospheric oxidation of 1.1,2-trijaethylhydrazinodiphenylphosphine CH3 CH3 PhgPN CH3 -!1/2 O2 *HigPU CII3 CH3 ^Cll3 Four and three tenths g, o£ the liquid product of the above reaction was dissolved in 40 nl. benzene and escposed to dry air for four days. At the end of this time much of the benzene had evaporated and the flask contained long, white crystals inbedded in c yellov/, viscous liquid. The crystals were collected and washed with a 1:1 benzene: n-hexane nixture. The n.p. was 139-162C,; recrystallization from hot n-hexane gave a cleaner product, n.p. 159-163C., and sublimation at 160C./0,17 mta. gave 2.9 g, of a product which melted at 164.5166.5*^C. (63 per cent of theory). Analysis Found: C, 65.52; H, 6.92; N, 10.08; P, 11.03. CalcJ. for C^^ll-^gr^^O: C, 65.63; H, 6.98; N, 10,22; P, 11.29. Tlie infrared (Figure 13) and n.in.r. spectra were consistent with the strtLCtural formula Ph 9 CH3 y \ y ^ Ph IV \ CH3 Tliis experiment serves to further confirm the identity of the product of the reaction between chlorodiphenylphosphine and 1,1,2tr ime tliy Ihydrazine •

PAGE 103

96 u 0.niU]]KUSllL'JX ij

PAGE 104

97 Reaction of ll,2-triiaethylhydra2ine yyjth diphenylphosphlnic chloride. CH ^ 1 / PhgPCl -i2m]H3N (0113)2 "Ph2^ "^ McNNHMeHCl N(CIl3)2 Eight and fifteen himdredtlis g, (0,11 mole) 1,1,2-triEiethylhydrazine was dissolved in 25 lal, dry benzene and added to 11,83 g. (0,05 mole) chlorodiphenylphosphine in 25 ml. benzene at 0C, with stirring. There was no evidence of reaction at 0C,, but the reaction proceeded rapidly at 45 C, as evidenced by precipitation. The mixture was warmed to 60 C, for one hour and filtered to give 13.4 g, of white solid on the filter (theoretical is 5,55 g,, based on the above reaction). Evaporation of the filtrate gave 6.15 g, of a gunny solid. After repeated attempts at purification by rccry stall izat ion and sublimation, attempts at purification were abandoned; most of the material decomposed on heating in solution or in the sublimation apparatus. The precipitate from the filtration was heated in cyclohexane in an attempt to recover some of the product which liad evidently been only partly solxible in the benzene and x<^ich had remained with the 1,1,2-trimethylhydraziniun chloride on the filter. The resulting mixture was filtered to give 5,49 g. I,l,2trimethylhydra2iniiaa chloride, m,p. 54-56 C, and upon cooling the filtrate there was obtained 6,2 g. of fine, white crystals, m.p, 165, 0-166, 5C, (45.2 per cent yield, based on the above equation). The infrared spectrum of this material was identical with that of the atmospheric oxidation product of the liquid obtained from the

PAGE 105

98 reaction of chlorodiphenylphosphine and 1,1,2-trliaethylhydrazine. A mixed melting point determination of these products melted at 165-166C. This experiment serves to confirm the identity of 1,2,2trimethylhydrazinodiphenylphosphine oxide. Experiments witli l-cthvl-2.2-diaethylhydra2ine and chlorodiphenylphosphine 1-Ethy 1-2 ,2-dimethylhydra2inodiphcnylphosphine HigPCl -h C2%NIEI (0113)2 -!(62115)3^ > I?li2^^^ N (0113)2 + [(02115)3^1 01 Six and seventeen hundredths g* (0.07 mole) l-ethyl-2,2-dimethylhydrazine in 20 ml, dry benzene was added to a solution of 15.44 g. (0.07 mole) chlorodiphenylphosphine and 7.23 g. (0,0715 mole) triethylanine in 50 ml. benzene. There was no evidence of reaction at 5 0., but precipitation occurred as the solution was warmed to 35-400. The mixture uas stirred at 400. for 3-1/2 hours and filtered to give 9.15 g. (95 per cent of theory, based on the equation above) trlethylamoonium chloride, m.p. 253-255C. A viscous, strax^colored liquid was obtained upon evaporating the solution at room tenq?erature and reduced pressure, llo purification was attempted, but a negative test for chloride ion was obtained on a small portion dissolved in dilute nitric acid.

PAGE 106

99 Analysis Fomd: C, 70.38; H, 7.63; N, 10.14; P, 11.44. Calcd. for C^ehlV' ^* 70.57; II, 7.77; K, 10.29; P, 11.37. The product was too viscous for an n.m.r. spectrum to be run on an undiluted sample, and poor results were obtained using a benzene soluticm. The proposed structtire is given below. Pli C2H5 P_N CHPh N \ CH3 Further cliaracterization is described below. Oxidation of l-3thyI-2,2''ditaethylhydrazinodiphcnvlphosphine A benzene solution of the product of the above reaction was exposed to undiluted o:qrgen at room temperature for three hours. Oxidation resulted in con5)lete deconqwDcition of the sample and none of the desired l-ethyl-2,2-diiaetlTylhydrazinodiphenylphosphine oxide was obtained. Reaction of l-Qthvl-2.2-dii3ethylhydrazinodiphenylphosphinc with sulfur S 1 PhgPNEtHMeg -:1/8 Sg ^ Ph2H^tNIIe2 Four and twenty-four hxmdredths g. (0.0156 mole) l-ethyl-2,2-dimethyl* hydrazinodiphenylphosphine, from the reaction of chlorodiphenylphosphine with l-othyl2,2-dimethylhydrazlne, was dissolved in 25 ml. dry benzene and added to 0.513 g. (0.0156 inole) sulfur in a 123 ml.

PAGE 107

100 erleruaeyer flask. The mixture was warmed to oOC. and allov/ed to boil for three hoxurs after which the benzene was allowed to evaporate, yielding 3.37 g. (31.4 per cent of theory) of white crystals, m.p. 124.0-126.5C. Recrystallization from acetone gave long, highly refractive crystals, a.p. 126,5-127 .5C, No loss was suffered handling this material in air. Analysis Found: C, G3.30; H, 7.09; N, 9.10; P, 10.34; S, 10.62. Calcd. forCj^gligiNjPS: C, 63.13; H, 6.95; N, 9.20; P, 10.18; S, 10.53. The n.m.r. spectra, run in CiXJl^, and the infrared spectrum (Figure 19) were both consistent with this structural formula: Ph •;; Col-Ie \^ X HI /\ Ph N(CH3)2 This e3q)eriment serves to further characterize l-ethyl-2,2dicffithylhydrazinodiphenylphosphine. Reaction of l-ethyl-2 ,2-diaethyruydrazinodiphenylphoGphine with methyl iodide A threefold excess of methyl iodide was added to a benzene solution of 2.49 g. (0.0089 mole) l-ethyl-2,2-dimethylhydrazinodiphenylphosphine. The solution rapidly became cloudy mid a yellow oil separated within a matter of minutes. Addition of ether to the gum produced some crystallization within a semisolid matrix. The crystals were separated from the gum and washed with benzene; they rapidly discolored from yellow to a dark brown upon exposure to air, but turned white upon heating to the melting point, 182-190OC.

PAGE 108

101 0.>UB^JTlUSin!.IJ, "j

PAGE 109

102 No purification method was found. The infrared epectruia was very poor, but indicated a salt-like structure by comparison with other spectra of phosphonium salts. The suggested structural formula is given below: PhoP— N \ N(CH3)2 Pveaction of l-'ethyl-2 2-dimethylhydr32ine with diphenyl* phosphinic chloride PhgPCl -V C^l\^mm(
PAGE 110

103 Filtration gcve 3.97 g. (93.6 per cent of theory) triethylamaonitfln chloride, m.p. 253-254 C, and 5.36 g. of wliite solid, m.p. 123-134 C, upon evaporation of the solvent in the filtrate. Stiblimation of this solid gave a clear, highly refractive, crystalline solid which melted at 140,0141. 0C, Analysis Found: C, 66.44; H, 7.24; N, 9.55; P, 10.89. Calcd. for Cj^gHgj^NjPO: C, 66.65; H, 7.34; N, 9.72; P, 10.74, The infrared (Figure 20) and n.m.r, spectra of the product are consistent with the structure Ph C2H5 P— N^ rn y \ /CH3 Ph N^ CH3 Reaction of l-ethyl-2.2-dijnethylhydrazinodiphenylphosphine oxide with methyl iodide One g. samples of l-ethyl-2,2-dimethylhydrazinodiphenylphosphine oxide were dissolved in benzene and toluene and heated with excess metliyl iodide to 50C. and 90C., respectively, and allowed to evaporate to dryness. The product of both reactions was unreacted l-ethyl-2,2-diiaethylhydrazinodiphenylphoEphine oxide and the absence of iodide ion was demonstrated by the addition of a drop of nitric acid which gave no reaction, It is therefore concluded that alkylation of l-ethyl-2,2dimcthylhydrazinodiphenylphospliine oxide with methyl iodide does not occur.

PAGE 111

104 aouB-i^nusuujj,

PAGE 112

10 5 Expcrinents with hydrazine and ehlorodiphenylphosphines 1 .2-Bis (diphenylphosphino) hydrazine In an attempt to prepare tetrskis (diphenylphosphino) hydrazine [(C^i^) 2^1 2^*^ ^^(^51^5) 2] 2 l,2-biE(diphenylphoephino)hydrazine was isolated instead: 2Hi2PCl -iNjH^ + 2(C2H5)3N aPh2PNHiraPPh2 -!2 [(C2H5)3m]Cl One and eight hjindredths g. (0,0337 mole) anhydrous hydrazine (97 per cent) was added slowly to a solution of 29.73 g. (0,1348 mole) chlorodiphenylphosphine and 13.2 g. (0,1796 mole) trietliylamine in 150 ml. chloroform, A deep yellow color developed immediately, but no precipitate formed. Heating the mixture to 65C. for one hour had no apparent effect. The solution was evaporated at room temperature under reduced presEvirc to give a white solid, which was immediately extracted with 300 ml. dry benzene and filtered to give 16,64 g, (86 per cent of theory) of triethylansnonium chloride, m,p, 251-253C. The filtrate was evaporated to a thick paste and upon adding 300 ml, dry ether a white solid precipitated. This solid was recrystal* lized from acetone to give 3.60 g. (26.7 per cent of theory, based on the above equation) of a white, crystalline solid, m.p, 129,0129, 5C, Analysis Found: C, 72.02; H, 5,57; P, 15.06, Calcd. for C24H22N2P2: C, 71,97; H, 5.54; P, 15.48. Inasmuch as this product was recxystallized from acetone, it is unlikely that it contains the -NH2 group since a hydrazine which contains the -NHgroup will react with acetone to produce a hydrazone

PAGE 113

106 with an entirely different elemental analysis from that given above, thus the correct structure is probably tliat given belw^: P-NHITH-P Ph Ha The infrared spectrum (Figure 21) is in good agreement with this structure, except that the N-H stretching absorption is not observed in its normal position, 3500-3200 cm."^, which may indicate some interaction of the phosphorus atcms xrlzh tha N-H protons. In view of the high yield of trie thy lammonium chloride and loi^ yield of 1, 2-b is (diphenylphosphino) hydrazine, it is possible that other, unobserved products may exist in the crude product. This system warrants further investigation. E^eriments r.:ith phe nvlhvdrazine and chlorodiphenvlphofip h-inp Phenylhydrazinodiphenvlphosphine. Tventy-tw g. (O.IO mole) cI^lorodiphenylplKJsphine in 75 ml. dry benzene was added to 21.6 g. (0.20 mole) phanylhydrazine in 75 ml. benzene. The addition was performed slowly to allow the heat to dissipate and the mixture was stirred at room temperature for one hour after addition was complete. l^on filtration 14.5 g. (100 per cent of theory) of phenylhydrazinium chloride, m.p. 238C. (literature value, m.p. 2'?tOC. (70)) was obtained. Evaporation of the benzene from the filtrate left an oily liquid from which no crystals were obtained. The infrared spectrum of the liquid is consistent with the structural formula PhgPNHNHPh

PAGE 114

107 U o.)m;;i!uisur;.ij^ ',

PAGE 115

108 Vacuum diBtillation of the liquid resulted in thermal decomposition, giving a volatile fraction, b.p. 73-74C./0.5 mm., and 17.5 g. of a resinous solid, m.p, 79-82 C, Extrapolation to the normal boiling point of the volatile fraction on a vapor pressureo temperature nomograph gives ca 135 C,/760 mm. Aniline has a b.p. o 184.5 C, and it is a possible product of the reaction given below: The material balance is in accord with this suggestion, however, the solid residue is apparently a mixture of condensed species in which some oxidation has occurred, as evidenced by infrared absorption at 1175 cm.*^, which ie characteristic of the P=0 bond. The absorption frequency characteristic for the N-N bond is not present in the Infrared spectrum of this material, however, it is frequently difficult to find. A weak absorption at 1250 cm," is evident, 2nd is a characteristic absorption of the N-P^N group in phosphonitrilics. Experiments with dichlorophenylphosphine Bis (2 .2-dimethylhydrazino)phenylphosphine PhPCl2 + 4H2KN(CH3)2 ^ PhP[Nmi (0113)2] 2 ^ 2l
PAGE 116

109 o 1 l-dimethylhydrazinivim chloride, m.p. 79-82 C, and a yellow filtrate frcxn which a yellow, viscous liquid was obtained upon removal of solvent at room temperature under reduced pressxire. The viscous liquid was dissolved in hot hexane and upon cooling a crop of white crystals formed, m.p, 58-60 C. Sublimation of this solid gave long, needlelike crystals, m.p. 61-63C,, which were very sensitive to the atmosphere and absorbed moisture rapidly. The yield on sublimation was very low, 12 per cent of theory, and the bulk of the material remained in the pot of the siiblimation apparatus as an amber, resinous material, which pulverized readily and which became tacky upon e:q>o&ure to the air. Appreciable quantities of 1,1-dimethylhydrazine were recovered from the cold trap used to protect the vacuim pump in the sublimation system. The white, crystalline solid gave the following analysis: Found: C, 51.79; H, 8.15; N, 22.01; P, 14.10. Calcd. for C-^Qllig\?'. C, 53.08; H, 8.46; N, 24.77; P, 13.69. The poor agreement of the analytical results with the suggested fonmila is attributed to the extreme sensitivity of the product to moisture and ojqrgen. The infrared spectrum (Figure 22) of this material is in good agreement with the proposed structure, but indicates that o:^gen (as P=0) is an impurity. <^~\-P[nHN(CH3) 2J2 The resinous material produced during sublimation has an analysis which Is in fair agreement with that expected for a polymeric material with the repeating unit shown below:

PAGE 117

110
PAGE 118

Ill 9 P N Analysis Found: C, 56.79; H, 5.54; N, 15.34; P, 21.12. Calcd. for CgHiiN2P: C, 57.83; H, 6.67; N, 16.86; P, 18.64. The molecular weight, determined cryoscopically in benzene, was 2076 for the reaction described here, wfaich indicates an average degree of polymerization, x = 12.5. No attempt was made to alter the molecular weight by changing reaction conditions. The fact that l,l-dimethylhydra2ine is produced in the thermal decon;>osition along with the polymer indicates that the reaction probably proceeds according to the equation below: xPhP(NHNMe2)2 ^ xMegNlfflg '^' nCPhPNNNegJy Oxidation of bis(2.2-diinethvlhydrazino)phen\lphosphine 1 PhP(NHNMe2)2 ^' 1/2 O2 >PhP(tmNMe2)2 A solution of 3.6 g. bis (2,2-dimethylhydrazino)phenylphosphine in 100 ml. benzene was boiled overnight in contact with dry air. Long, needlelike crystals appeared in the solution upon cooling. Tliese were sublimed at 160C./0.20 mm. gave white crystals, m.p, 161.0-163.0C., in 40 per cent yield, based on the equatitm above. Analysis Found: C, 49.74; H, 8.03; N, 23.01; P, 12.91. Calcd. for CioHi9N4^C. 49.58; li, 7.90; N, 23.13; P, 12.79.

PAGE 119

112 Infrared spectral data (Figure 23) are consistent with the structural formula 1 Hi2P(NHNMe2)2 This experiment serves to further characterize the product of the reaction of dichlorophenylphosphine and 1,1-diiaethylhydrazine as b is (2 2-diinethy Ihydrazino) pheny Iphosphine Reaction of l.l-dimethylhydrazine with phenvlphosphonic dichlorlde 1 1 Ph2PCl2 + 4(CH3)2NNH2 5^ PhP[Nim(CH3)2l2 ^2(CH3)2NNH2'HC1 Nineteen and five tenths g. pheny Iphosphonic dichloride was dissolved in 20 ml. chloroform and this was added slowly, under a nitrogen atmosphere, to a solution of 25.0 g. (0.42 mole) 1,1-dimethylhydrazine in 20 ml. chloroform. The addition took five hours; after which time the reaction mixture was warmed to 60C,, stirred for one hour, and filtered hot. The yield of 1,1-dimethylhydrazinium chloride was not quantitative; apparently this salt is somewhat soluble in chloroform. The chloroform was evaporat ad at rocMi temperature under reduced pressure to give a white solid, x^iich was extracted twice with boiling benzene. The benzene solution deposited 17.0 g. of finely crystalline solid, m.p. 154-160C., upon cooling. Sublimation at 16OC./0,18 ram. gave 14.5 g. (61 per cent of theory based on the equation above) of white crystals, m.p. 161, 0-164. 0C,

PAGE 120

113 o.)U1JV,ivusui:jj.

PAGE 121

114 A mixed melting point determination of this product with the product of the air oxidation of bis(2,2-dimethylhydrazino)phenylphosphine gave 160-162 C, and a con^ariscn of infrared spectra of the two materials showed conclusively that they were the same compoimd. This experiment serves to confirm the identity of the air oxidation product of bis(2,2-dimethylhydrazino)phenylphosphine as its oxide. Bis(2,2-dimethylhydrazino)phQnylphosphine sulfide S S PhPClg + 4Me2tMl2 PhP(NIES-fe2)2 "^ 2Me2NNH2'HCl Sixteen and eighty-nine hundredths g. (0.08 mole) pherqrlphosphonothioic dichloride was dissolved in 25 ml* dry benzene and added to 19.6 g. (0.326 mole) 1,1-dimethylhydrazine in 25 ml. benzene. The addition was carried out at 0C,, but no evidence of reaction was observed until the reaction mixture was warmed to 10C., at wiaich point precipitation began. Tlie mixture was warmed to 25 C. and stirred for one hour and then filtered to yield 14.70 g. (94.3 per cent of theory) 1,1-dimethylhydrazinium chloride. Removal of the solvent from the filtrate gave 20.10 g. of a white solid, ra.p. 91-102C, "Ehis material was recrystallized from 50 ml. of 1:1 benzene zn-hexane to give 12.8 g. of white, crystalline solid. Subliination gave a product which melted at 106.0-106. 5C. Analysis Found: C, 46.43; li, 7.59; N, 21.80; P, 12.07; S, 12.25. Calcd. for Cj^q%9N4PS: C, 46.50; H, 7.41; N, 21.69; P, 11.99; S, 12.41.

PAGE 122

U5 The infrared (Figure 24) and n.ia.r. spectra are consistent with the structural forniula below. S Ph-P[NHN(CH3)2], Partial hydrazinolysis of dichlorophenylphosphine PhPClo + MeoNNH, ^ EhPC ^ Me-,NNHHC1 Seventeen and nine tenths g. (0.1 loole) dichlorophenylphosphine in 20 ml. dry benzene was added to a solution of 12.0 g. (0.20 mole) Ijl-dimethylhydrazine in 20 ml. benzene. The reaction was exothermic and required cooling. Vlhen addition was complete the mixture was stirred at room temperature for one hour and filtered to yield 9.2 g. (95 per cent of theory, based on the above equation) of 1,1-dimethylo hydrazinlum chloride, m.p. 79-83 C. Removal of the solvent at room temperature and reduced pressure gave the product: a clear, viscous liquid. The cold trap in which the solvent condensed did not contain 1,1-dimethylhydrazine, as shown by testing an aqueous extract of the solvent with KltoO^. An atte3:i?)t to distill the product resulted in evolution of hydrogen chloride and gave an aober, resinous material. Inasmuch as little was found useful in the way of, purification, this product was not characterized other than obtaining a chlorine analysis. Analysis Found: CI, 24.97. Calcd. for Cglii2^2^124.70,

PAGE 123

116 O L^ .,-.i=. i_Li J \i I I J i. ;').)m7;i]ursin:.i 1, \

PAGE 124

U7 This loaterial is extremely sensitive to oxygen and moisture and fumes in moist air. EKperiments with phosphorus trichloride Hxosphorus tri (2 2-'dinethylhydrazide) PCI3 + 6tfe2NNH2 > P(NHNMe2)3 + SMsjNNHj'HCl Six and eighty-seven hundredths g. (0.05 mole) phosphorus trichloride in 25 ml, dry benzene was added slowly, under a nitrogen atmosphere, to 24.06 g. (0.4 mole) 1,1-dimetl^lhydrazine at C. Precipitation began as the solution was warmed to room temperatinre and continued as the solution was warned to 35 C. The mixture was stirred at 30 C. for one hour and filtered to give 13,8 g. (95 per cent of theory) 1,1dimethylhydrazinium chloride. Upon removal of the solvent from the filtrate the product was obtained as a yellow, viscous liquid, which contained no chlorine, as determined by testing a small aquecub extract with silver nitrate solution. The product could not be distilled and thus no characterization was attempted; rapid reaction with moist air was noted, however, and an e:qposed sample quickly absorbs moisture froa tlie air and becomes tacky. Pliosphoryl tri (2 ,2-diiaethylhydrazide) 1 \ PCI3 + 6Me2NNH2 ^ P(HHNlle2)3 -i3>le2inJH2*HCl Eight and twenty-one hundredths g. (0.0535 mole) phosphoryl chloride in 25 ml, chloroform was added over a two hour period to 20.0 g.

PAGE 125

118 (0.33 mole) 1,1-dintethylhydrazine in 25 ml. chloroform at 0C. The ten^erature was raised to 30C. and stirred for one hour and filtered to give 9.97 g. (64 per cent of theory, based on the above equation) 1,1-diniethylhydraziniuni chloride. The filtrate was evaporated at rocaa temperature under reduced pressure and then extracted with hot benzene and filtered. Upon cooling, the benzene solution deposited a white crystalline solid which was then sublimed at 105C./0.10 xm, to give 3.79 g. (76 per cent of theory) of white, crystalline product which melted at 193. 5-194. 0C. Analysis. Found: C, 31.67; H, 9.57; N, 37.03; P, 13.56. Calcd. for CQK2iNgP0: C, 32.14; H, 9.44; N, 37.43; P, 13.01. The infrared (Figure 25) and n.ra.r. spectra are consistent with the structural formula given below. 1 p[min (0113)2]. Thiophosphorvl tri(2.2-diiaetlTvlhvdrazide) S S 1 i PCI3 -i6(CH3)2HHH2 ^ p[miN(CH3)2]3 ->'3(CH3)2mcl2-I-ICl Thirty-seven g. (0.62 mole) l,l-diraethylhydra2ine was dissolved in 20 ml. dry benzene and to this was added with stirring and cooling a solution of 10.64 g. (0.063 mole) thiophosphoryl chloride in 20 ml. benzene. \Jhen addition was complete the mixture v/as warned to 45C. and stirred for 30 minutes and filtered.

PAGE 126

119 o ^ /^ o.JUUDUusuL'ax \j

PAGE 127

12 Eighteen and six tenths g. Ijl-dimethylhydrazinii3in chloride was collected on the filter, and 15.0 g, (99 per cent of theory) white solid was obtained upon evaporation of the solvent at room temperature and reduced pressure. Sublimation of the crude product gave 7,5 g. (50 per cent yield) white, crystalline solid, n.p. 75.5-78.0C, Analysis Pound: C, 29.74; H, 8.52; N, 35.07; P, 13.07; S, 13.49. Calcd. for CqU21^c?B: C, 29.99; H, 8.81; N, 34.97; P, 12.89; S, 13.34. The infrared (Figure 26) and n.m.r. spectra are consistent with this structure: S P[lIHH(CH3)2] Thiophosphoryl tri(l 2 ,2-triiaethylhydrazide) S S PCI3 + CH3NHN(CH3)2 ^ P[nCH3N(CH3)2] -!3CH3NIIN(CH3)2'HC1 Five and eight hundredths g. (0.03 taole) thiophosphoryl chloride was reacted with 14.83 g. (0.20 mole) 1,1,2-triiaethylhydrazine in 50 ml. benzene at 0-5C., under a nitrogen atmosphere. The solution was warmed to room ten^jerature after addition mid the precipitated salt was filtered from the mixture. Wine and eighty himdredths g, 1,1,2triraethylhydrazinium chloride, m.p. 54-56C., was collected on the filter (98.6 per cent yield). Upon evaporation of the solvent at room temperature and reduced pressure there was obtained a yellow gum. Crystallization was

PAGE 128

121

PAGE 129

122 attenpted by cooling the gum, but neither this treatment nor the addition of cyclohexane or ether caused formation of solid product. Rapid deconiposition occurred when an attempt was made to purify the product by vacuum distillation. The decomposition tenqperature was 165C./0.4 on. No further characterization was attempted. Thiophosphoryl tri (lethy 1-2 ^ 2-dImethy Ihydrazida) S S 1 1 PCI3 -i6Me2NNH2 ^ PCraNMe2)3 + Mc2KNH2*HCl Four g. (0.0236 mole) thiophosphoryl chloride was dissolved in 25 ml. dry benzene and added to 13.2 g. (0.15 mole) l-ethyl-2,2-dimethylhydrazine over a one hour period at C. The mixture was then warmed to 30 C, and stirred for one hour and filtered to give 5.5 g. (62 per cent of theory) l-ethyl-2,2-Jiaiethylhydrazinium chloride, m.p. 5657C. An additional lot of l-ethyl-2,2*diraeti^lhydrazinivjm chloride o was obtained by stirring the filtrate for 15 hours at 30 C. The aielting point of this lot was 57-60C,, and from infrared spectral data it was determined that the salt was contaminated with a phosphoruscontaining c(apound. Successive lots of precipitate, obtained by allowing the solvent to evaporate slowly, had higher melting points and the intensity of the infrared absorption frequency associated with the P=S bond at 7 2 2 cm. increased in intensity, A nttmber of solvents were tried In order to find one which will separate the desired product from l-ethyl-2,2-dimethylhydrazinium

PAGE 130

123 chloride, but apparently this salt is soluble in the solvents one ;/ould e:q)ect to dissolve thiophosphoryl tri(2,2>diia&tliylhydrazide) thus no scparaticm was loade. An alternate route to the synthesis o£ this coc^uv^ would be to use a tertiary amine as an hydrogen chloride acceptor so that the salt fonoed would not be soluble in benzene or chlorofomn. This method is still untried in tiiis particular system. Discussion Infrared analysis An intrared spectrum was obtained on each hydrazinophosphine and derivative prepared in the course of this work and the absorption frequencies found in the more significant co!iq>ounds ^jere tabulated (Table 26), Assignments are found in Table 27, Itost of the spectra were obtained from Nujol mulls since most of the compounds studied are crystalline solids; diphcnylphosphine is an exception and its spectrum was obtained in a 0.0295 ram, cell without solvent. The assijpments made are on the basis of loiown absorption freqtiencies (71,72) and by comparison of the various spectra in which a frequency is observed which can be correlated with a particular group. The infrared spectra are in excellent agreement with those e3q>ected on the basis of the anticipated molecular structures. No unusual shifts in the absorption frequencies are observed except in the case of the N-IT absorption band which is ticually found in the vicinity of 950 cm," and which is often very weak and difficult

PAGE 131

m4 W 14 S TOO

PAGE 133

PttJKll lirfnr^ tbtotftUa F 1

PAGE 134

U5 TABLE 27 Infrared Absorption Frequencies and Assignments Group Range (Cm, ) N-H (free) N-H (associated) C-H (stretch, aromatic) C-H (stretch, aliphatic) P-H phenyl, loono-siibstituted H-H (deformation) phenyl, skeletal in-plane vibrations C-H (deformation, aromatic) C-H (deformation, aliphatic) P-phenyl -NRg, N-CH3 C-CHg phenyl P=0 phenyl (near P=0 or P=S), monosubstituted N-CH3 N-N N-H P-N C-H chains, P=S phenyl (C-H, out-of-plane deformations) 3450-3590 (w) 3175-3320 (w-m) 3030-3100 (s) 2880-3000 (s), 2610-2800 (w-m) 2280 (m) 1900-1975 1835-1900 (w-triplet) 1790-1880 1625-1700 (w) 1565-1590 (w) 1475-1487 (w-m) 1451-1470 (m) 1370-1380 (w) (doublet) 1430-1440 (s) 1390-1430 (w-m) 1373-1380 (w-m) 1318-1358 (w) 1175-1188 (pi-s) 1109-1220 (io-s) 1040-1050 (w-m) 950-996 (w) 880-917 (m-s) 743-773 (m-s) 700-727 (m-s) 690-697 (ra-s), 671-688 (m)

PAGE 135

126 to locate. In most hydraziaophosphorus compounds, as in alkyl hydrazines, the N-N absorption frequency is found in the range, 943-960 cm. In some of the hydraziaophosphorus compounds, the absorption is v/ealc and difficult to find, Tlie various N-II absorption frequencies are observed in the spectra of those hydrazinophosphorus ccmpounds tjhich are estpected to contain the N-H bond, and a new, intense band in the range 880917 cm."^ is obseived in all the N-H containing hydrazinophosphorus con?>ounds, but not in the triaikylhydrazine derivatives, which do not contain the N-H bond. llany characteristic bands are observed which are distinct and easily located: P=0, 1175-1188 cm."^; P-phenyl, 1430-1440 cm.~^; and P-H, 2280 cia."^. Other bands \Aich are more difficult to find and which are frequently obscured include the N-CHo (13901430 cm.*^), the P-N (743-773 cm."^, and the P.s' (700-727 cm."^) bands. The absorption frequencies associated with the C-C and C-H bonds, as well as those assigned to aromatic ring vibrations, are all found in the ranges described in the literature (71), but the monosubstitutod phenyl ring band at 1109-1220 cm."-^ appears to be intensified by the proximity of the P=0 and P=S groups. This band is not found in hydrazinophosphines, but is found in hydrazinophosphine oxides and hydrazinophosphine sulfides. The observation that P=0 and P=S groups intensify monosubstituted phenyl absorption in the infrared leads one to speculate

PAGE 136

127 ae to whether the absorption frequency range consnonly assigned to P=0, 1175-1188 cn."^ (71), is actually a P=0 absorption or an attenuated aromatic ring vibration in diphenylphosphinocompoundE which contain the P=0 bond. A band in this region is observed in all diphenylphosphinoconq>ounds, but unless the P=0 group is present the peak is extr^oiely weak, and is comaonly regarded as being due to the presence of oxygen (in the form of P=0) as an in5>urity. There is some question as to the actual bond order oC the PN bond in hydrazinophosphines and other cocipounds whicli contain the ?-K bond. The infrared spectnaa is a potential tool for solving the problem, but to date no one has collected extensive data on the P-N absorption frequencies, and it has been conmon practice to include a rather wide range of infrared values in the characterization of the P-N bond, 750-870 cm." (72). This wide range leaves room for speculation as to the actual absorption frequency associated with a P-N single bond. In our work we have narrowed this range to 743773 cm. wliich seems to include all our examples. Nuclear magnetic resonance analysis Tlie n.n.r. spectra of many of the Ir/drazinophosphorus compounds synthesized in this study were obtained and the data are tabulated in Table 23. The chemical shift is expressed in tau values for protons and in parts per million (ppn.) for the ^^P nucleus. The pattern and splitting observed for nailtiplets are given and assignments are made for the various signals found. The relative areas of the peaks are not given; however, these are all in very good agreement with the proposed structures of the coiiq>ounds stvidied.

PAGE 137

128

PAGE 138

U9 I O 53 g 0) S3 O f 31 g| Si i n 9 u •§.5 9 3 i 5 s CQ

PAGE 139

130 ^ 1

PAGE 140

131 f a o c3 I •§ •§ 33 3 Of) O O W M O rt ^ O i-l -a o o e. CJ o is' ^1 B O 3

PAGE 141

132 The infonaatiOQ sought in this segment of the \70rk, other than confirmation of the assigned stmctioral formulae, conceims the 31 relative electron densities around the nucleus studied, P. Since the chemical shift is a phencsaenon directly associated xttth the diamagnetic shielding around the nucleus, these quantities can be correlated and it may then be said that, in general, the chemical shift is a function of the electron density around the nucleus; low chemical shift values (downf ield from an arbitrary reference) are associated with low electron density, and high chemical shift values indicate high electron density around the nucleus. The '^P nucleus has a larger chemical shift than does H because the presence of electrons in higher energy orbitals in the phosphorus atom allows for greater polarizability of the phosphorus atom than of the hydrogen atom* Table 29 shows a scale of the chemical shift values observed for some hydraziiiophosphorus c-ocrpounds. This scale shov/s that, among the compounds considered, the electron density about the phosphorus atom is greatest for hydrazinophosphine oxides and least for hydrazino* phosphine sulfides. Although it iTOuld be desirable to have additional information upon which to base generalizations, an adequate argument for rationalizing the positions of these compounds on the scale can be raade, Tlie stereochemistry of a phosphine oxide or sulfide may be described as a distorted tetrahedral structure such as that indicated below.

PAGE 142

TABLE 29 Scale of Chemical Shift Values frcsn N.M,R. Data 133 Chemical Shift (ppm.) Con^KJund -12.5 -23.2 -47.4 -57.3 P(KmiMe2)3 1 -22.0 Ph2PNHNMe 1 -26.2 Ph2RIEtIIM32 1 PhoPNtieNlIe' (Ph2P)2ffife2 S 1 Ph2H'IHNMe -61.1 RiP(NHniiQ2)2 S 1 -64.4 m2^(Wtl^'^2'^

PAGE 143

13-4 ;0: -S In the oxide, back-donation of electrons from the o:cygen atom to the phosphorus atcra can occur in p^ -d^ bonding, thus delocalizing a portion of the electron doasity about the oxygen atom towards the phosphorus atcsa. In the sulfide, such back-donation is unlikely because of the larger intemuclear distance between the phosphorus and sulfur atoms and the fact that higher energy orbitals would have to be used by the sulfur than are favorable for such back-donation. Tlie electron density about the phosphorus atom is thus greater in phosphine oxides than it is in phosphine sulfides. Tlte same effect, back-donation of electron charge, is responsible for the ICKjex reactivity of phosphine oxides coiapaxed with phosphine sulfides and unoxidized phosphines since those chemical reactions which depend on nucleophilic attack of a basic species on the phosphorus atom are favored by the phosphorus atom which has least electron density about it and tjhich has available vacant d-orbitals for such attack. Interpretation of synthetic results Hydrazinolysis of the phosphoruschlorine bond It has been amply demonstrated by the synthetic results reported in this study that hydrazine lysis of the phosphorus-chlorine bond is a general reaction and may be applied to mono-, di-, and trichlorophosphorus compounds, with phosphorus in the three-covalent or four-covalent state.

PAGE 144

135 Observations of the reaction conditions show that different chlorophosphorus compounds and hydrazines vary in reactivity. The tenq)erature of a reaction mixture containing a hydrazine and a chlorophosphorus conipound does not rise appreciably until the hydrochloride salt of the hydrazine (or tertiary amine) begins to precipitate; thus, the temperatures associated with the first appearance of precipitate are characteristic of individual reactions. Most hydrazinolyses will occur below 25C.; however, a few of the reactloiS observed required elevated temperatures. Qualitative observations of the reactions of several chlorophosphorus compounds with 1,1-dimethylhydrazine show that the reactionb become increasingly exothermic in the order Ph2P(0)Cl < Ph2PCl < PhP(S)Cl2 < P(S)Cl3< HiP(0)Cl2
PAGE 145

136 It is possible to resaavc all the protons from a hydrazine (except hydrazine itself) using chlorodiphenylphosphinej and removal of the first proton can be accomplished at low temperatures, seldom above 35C, Hotjev'er, as the second and third proton is removed higher temperatures are required (up to 110 C), In an attempt to substitute all four protons on hydrazine, disubstitution and the formation of l5 2-bis(diphenylphosphino)hydrazine, Hi2PtniNHPPh2, resulted and no tetrasubstituted product was isolated. Since the substitution of the second proton in 2,2-dimethylhydrazinodiphenylphosphine \rill proceed only at higher temperatures than that required for the first hydrazino lysis, HigPCl -iH2l?Me2 '-^' Ph2pHIINM32 the proton in the N-H group in 2,2-dimethylhydrazinodiphenylphosphine is apparently less active than that in 1,1-diiaetliylhydrazine. In order to deactivate this proton it is necessary to increase the electron density about the nitrogen atom in the H-H group, which leads to the postulate that the diphenylphosphinogroup is electrondonating and deactivates the second proton towards substitution. Ph CH, p-^ m-i-ii Ph ^CHj Ifechanism of hydrazino lysis The mechanism of hydrazinolysis of chlorophosphorus compounds probably involves nucleophilic attack

PAGE 146

137 of the hydrazine on the phosphorus as an initial step v^ich can occur in either of two w^s. Examples: ft ="3 CI I, H-N-NMe, -r CI-P:^ ^ Ii,N — K : P — Hi "* ^ ^Ph ^ I I CH, Ph Vh H CI /^ I I 2. m.imiy -iCl— P^ ^ lieoN — N : P_Ph 2 ^ \ph ^ I I H Ph The aiergetics of this step would depend on the ^lergy released by the formation of the P-N coordinate bond and the energy required to change the bond angles to that in the intermediate adduct. It is expected that this step xTOuld be reversible at low temperatures and that no further reaction would result from the adduct Gho;i in equation 1. The product from equation 2, liowever, could undergo intramolecular elimination of hydrogen chloride by a concerted electron transfer process as sho^m in equation 3: He

PAGE 147

138 The energetics of this step include breaking an N-H bond (H72 kcal./raole) and a P-Cl bond (-177 kcal./mole) and forming an H-Cl bond (-102 kcal./fflole) This step is not favored by the thermodynamics of the process, but additional energy is released by the formation of salt produced when the evolved hydrogen chloride reacts with the free base provided for this purpose: EtgN: -f HCl (Et3NH)Cl + energy (+102 kcal./mole-llO kcal./raole^lattice, salt) An alternate view, which is energetically identical, would involve the reaction of the labilized proton in the N-H bond of the adduct with the free base (+72 kcal./mole110 kcal./mole), followed by elimination of a chloride ion from the phosphorus intermediate (+77 kcal./mole) and subsequent salt formation ("Ej^gt-t-j^^e salt^ • The overall process, regardless of route, is described by the following thermodynamic equation: ^^reaction^ <-Eadduct formation) (-72 kcal.) + (+77 kcal.) + (-110 kcal.) + (-Eiattice, salt> ^ ^^IP^EA^Ejp g^ includes the ionization potential and electron affinity corrections for the bond energies involved. Inasmuch as the reaction is known to be spontaneous at room ten?)erature, and since it is not thought that the adduct formation is of major inqjortance energetically, the proton affinity for the amine and the lattice energy of the salt produced in the reaction make the largest contributions to the driving force of the reaction. The fact that hydi-azinolysis of a chlorophosphine oxide or sulfide is less vigorous than hydrazinolysis oi a chlorophosphine may

PAGE 148

139 be e:q)lained by considering 1) the availability of d-orbitals on the phosphorus (for adduct formation) 2) the size of the phosphorus atom and the space available for coordination of an attacking hydrazine molecule, and 3) the effect of an oxygen atom or sulfur atom as a substituent on the phosphorus atom. The availability of d-orbitals on the phosphorus atom would be greater in phosphine sulfides than in phosphine oxides because of the higher electron density around the phosphorus atom in the oxides, which presuonably results from the back-donation of electrons as described previously. Tlius, the initial step in the mechanism, adduct formation, would be favored in phosphine sulfides over phosphine oxides on the basis of d-orbital availability. The size of the phosphorus atom is sufficient to allow a coordination nionber of six in the hexaf luorophosphate ion, EF^ and five in the vapor phase form of phosphorus pentachloride and many organic derivatives which contain bulky groups. Thus, it is not expected that adduct formation of hydrazines with phosphines, phosphine oxides, or phosphine sulfides would be severely restricted by steric requirements. An order of reactivity, based on ease of access of a hydrazine molecule to the pliosphorus atom, should be observed in which reactivity decreases in the series phosphinesN phosphine oxides/ phosphine sulfides. High electron density is apparently the major effect favoring reactivity of chlorophosphorus compounds towards the solvolytic reaction. N.m.r. data indicate that the electron density about the phosphorus atom is greater in phosphine oxides than in phosphine

PAGE 149

140 sulfides. Thus, by comparison with unoxidized phosphines, tetracovalent chlorophosphorus compounds with oaygen as a substituent on the phosphorus atom should be deactivated towards hydrazinolysis less than would those with sulfur as a substituent. In sumaarizing the various considerations affecting the hydrazino lysis reaction, it is suggested that d-orbital availability is not greatly affected by substituents on the phosphorus atom and that such orbitals are available whether or not the phosphorus is in the oxidized form, since the orbitals which participate in d^-p^rback-donation are not those which participate in adduct formation. Although steric considerations are in agreement with the observed reactivity, they are probably of minor importance. The major factor, high electron density about the phosphorus atom, is expected to labilize the chloride towards nucleophilic displacement or anionic elimination, which, in this particular system, appears to be an irreversible step. Oxidation of hydrazinophosphines Both phosphines and hydrazines are well-lcnown for their reducing properties. Phosphines will, in general, react with o^qrgen to produce phosphine oxides, whereas hydrazines may be oxidized to a ntimber of nitrogen-containing coii?)ounds, depending on the extent of oxidation and the system in which oxidation occurs. A great deal is yet to be learned concerning the direct oxidation of alkyl hydrazines. In the hydrazinophosphines it has been found that oxidation occurs preferentially on the phosphorus, and in this way some hydrazinophosphine oxides were prepared. Attack on the hydrazinogroup also occurs, however, and in the case of l-ethyl-2,2-dimethylhydrazinodiphenylphosphine complete decomposition x^as observed upon

PAGE 150

141 prolonged exposure to dry, undiluted o^gen. Ifader milder oxidizing conditions it V7as found possible to prepare hydrazinophosphine sulfides by direct reaction of the hydrazinophosphine with elemental sulfur with undetectable effect on the hydrazinegroup. There is a basic difference in the mechanism of oxidation of phosphines and hydrazines: phosphorus may simply add an oi^gen atom upon treatment with molecular o^qrgen; however, some N-N bonds are usually broken in the complex oxidation of hydxazines. The speed and sin^licity of the phosphine oxidation is apparently the predominant reason for preferential reaction with oxygen and sulfur. Hydrolysis of hydrazinophosphines Hydrolysis of hydrazinophosphlnes occurs upon contact with moisture. It can be detected in samples by inspection of the infrared spectrum in the regions where the 0-H and B"0 absorption frequencies are found. In the absence of oxygen, hydrolysis of 2,2-dimethylhydrazinodlphenylphosphlne proceeds according tc the equation below. Ph2PKHNMe2 + HgO = Hig^"^^ "^ H2NNMe2 The diphenylphosphinous acid formed in this reaction undergoes a slow disproportionation as shoxm by the equation (73) 2Hi2P-OH = Ph2PH v Ph2P(0)0H, and it is for this reason that the hydrolysis can be detected by the presence of the intense ^0 absorption frequency in the infrared. Atmospheric hydrolysis is acconq)anied by fast oxidation and the only observed product is diphenylphosphinic acid, Ph2P(0)0H,

PAGE 151

142 It was observed that bis(2,2-dimethylhydra2ino)phenylphosphine is extremely susceptible to hydrolysis. The ease of hydrolysis is touch greater than that observed £or any monohydrasinophosphine. The effect is apparently due to the presence of tiro electronegative hydrazinogroups on the molecule. The mechanism of hydrolysis follows the basic solvolytic process and presumably involves an initial attack of a water molecule on the phosphorus atom in the bis (hydrazino) phosphine as in the equation, H f H ^ \ \ I 0. { P-HH-Klfeo = — P-m-KMeo H ir NHime.2 NHmcg followed by the elimination of free 1,1-dimethylhydrazinG as shown in the equation below: H Eh Ih 0— P-NmJMe2 = H0-P-iraNMe2 + H2lMyfe2. H Nffi^Me2 It is generally observed that as the number of electron-withdrawing groups on phosphorus is increased, the ease of solve lysis becomes greater; thus, it is not anomalous that bis (hydrazino) phosphines, with two electronegative hydrazino-groups attached to the phosphorus atom should exhibit greater reactivity towards nucleophilic reagents that do laonohydrazinophosphines. Hydrazinophosphine oxides and sulfides are, in general, more resistant towards hydrolysis than are the hydrazinophosphines. in

PAGE 152

143 these instances, especially in the case of the oxides, the increased electron density about the phosphorus atom does not facilitate attack of nucleophilic reagents, such as water. The expected difference is noticed between Hi2P(0)HHNMe2 and HiP(O) (NHNlle2)2; the latter is less resistant towards hydrolysis, Disproportionation of hydrazinophosphines Inasmuch as diphenylphosphinous acid is kno^m to disproportionate to diphenylphosphine and diphenylphosphinic acid (as described in the hydrolysis section) it might be anticipated that hydrazinophosphines would display similar behavior. Infrared and n.m.r, data have shown no evidence of disproportionation at room ten5)erature, however, and only one hydrazinophosphine has shown such behavior at elevated temperatures: Eh2PNmraCH3 Q '^^ ^ :2^ Eh2PH -r condensed t.pecies. Ph2FNCH3NH2 (mixture) In this reaction, the condensed species may be idealized as Ph2P = N-NMe2 NH-NMe2 However, it was not possible to isolate this cosqK>und and it is probable that this species would undergo further condensation at the teii5>erature required for the disproportionation. Since it is the N-H bond which is broken when a hydrazine participates in hydrazine lysis a hydrazinophosphine which contains

PAGE 153

144 N-H bonds is subject to further reaction, with the N-H group acting as a. nucleophilic reagent. It is observed that the reactivity of hydrazinophosphincs increases with the nunfaer of N-H bonds in the molecule, thus while 2,2-dimethylhydrazinodiphenylphosphine (Hi2K^JHNMe2) can be siiblimed at 75 C. without serious loss by disproportiOTXation, the njethylhydrazinodiphenylphosphine mixture obtained by the reaction of methylhydrazine with chlorodiphenylphosphine is not separable by distillation because disproportionation occurs below the boiliiig point at 0.10 mm. Condensation of hydrazinophosphincs Condensation of hydrazinophosphines by intcrraolccular elimination of hydrazines hab Lean shown to occur at elevated temperatures. However, this behavior is much more evident in bis (hydrazino) phosphines than in the other hydrazino* phosphines studied and occtnrs according to the following equation: / NH NMeo / 7CPh-P NH-NMe. Ph NMd,' I I .p ij — XH2NNMe2 Tlie evidence for other hydrazinophosphincs undergoing a similar reaction is not conclusive enough to permit generalization at this point. It is expected that condensation, rather than disproportionation, is the predominate high temperature reaction of most hydrazinophosphincs. This effect should be most noticeable in those coapounds in lAich the phosphorus atom possesses low electron density, which xjould facilitate nucleophilic attack of N-H groups

PAGE 154

145 resulting in coupling and perh^s extensive polymer isaticsx o£ the hydrazinophosphines Alkylation of hydrazinophosphines Alky lat ion of substituted hydrazinophosphines with netliyl iodide has been shown to produce hydrazinophosphonium salts, rather than pliosphinohydraziniun salts. The site of alkylation was detemined by hydrolysis of the alkylation product to loioim phosphine oxide derivatives and the free hydrazine. This result is particularly interesting in view of the fact that a study comparing the base strengths of phosphines and amines by titration with acids in noninteracting solvents indicates that, with respect to certain r.cids, anines are stronger baees than are similarly substituted phosphines ("^4). In addition, phenyl phosphines are aiaong the least basic of the phosphines stiulied. The low basicity of phenyl phosphines is attributed to 1) steric effects and 2) pibond formation between the ring and the phosphorus atom. The alkylation of 2,2-dlmethylhydrazinodiphenylpho6phine with methyl iodide proceeds as given, in the following equation: + -NH-NMei HijP-NH-NMej + I'iel = [ph2P-NH-NMe2] I In 2,2-dijnethylhydrazinodiphenylphosphine there are three possible sites capable of nucleophilic attack on methyl iodide by (presumably) the Sjj2 mechanism. Although n.m.r. data vrfaich would indicate quantitatively the electron density about the phosphorus atom in 2,2-dimethylhydrazinodiphenylphosphine are not available.

PAGE 155

146 it is thought that the electron density at the phosphorus atom in this molecule is greater than that in hydrazinophosphine oxides and sulfides. Since alkylation occurs at the phosphorus atcm, this is apparently the most electron-rich site in the molecule. The question arises regarding possible contributions of electron density from substituents on the phosphorus atom in hydrasinophosphines. The diphenylphosphinogroup has been shown to be an electron-donating group towards the hydrazinegroup in hydrazinophosphines. Thus, there is some polarization of the P-H bond in hydrazinophosphines, which may be compensated for by back-donation of electron density from the non-bonding electron pair on the nitrogen atom to the available d-orbitals on the phosphorus atom. Further contribution to the electron density about the phosphorus atom may come from the phenyl rings which are in conjugation with each other and with the P-N bond: P — ilH— iJMeg. Either or both of these contributions would allow the phosphorus atom to achieve the basicity necessary for successful attack on methyl

PAGE 156

iodide. The nitrogen atoms in this molecule are apparently unable to increase their basicity by such means and must rely on the inductive effects of their siistituents, which appears insxjfficicnt to permit allcylation in the presence of tricovalent phosphorus. Alky lat ion was atteii?>ted with methyl iodide on several hydrazinophosphine oxides and sulfides, but no reaction was observed. The oxidized forme of hydrazinophosphines apparently do not possess the basicity at either of the nitrogen atoms necessary for nucleophilic attack on methyl iodide. This would serve to indicate that the diphenylphosphonoand diphenylphosphorotiiionogroups are electronwithdrawing groups with respect to the hydrazinogroups, since alkyl t^drazines are quite easy to alkylate with methyl iodide. No study was made of the effect of the solvent in these reactions and further work is indicated in this area. Conclusion The work reported here describes a series of new chemical species: substituted hydrazinophosphines, hydrazinophosphine oxides, hydrazinophosphine sulfides, and various derivatives of these compoxmds. These materials were characterized by elemental analysis, infrared spectral studies, nuclear magnetic resonance data, and observations of the reaction conditi
PAGE 157

148 The reactions of the new cong>oundG were studied, and special eii5>hasis was placed on oxidation, hydrolysis, condensation, dis proper tionat ion, and alkylation of the new hydrazinophosphines. Extensive data were collected which provide information concerning the electronic arrangements about the phosphorus atoms in these compounds, and general statements concerning the e3q>ected behavior of hydrazinophosphines can now be made. Tlie diphenylphosphino group was found to act as an electron-donating group in hydrazinophosphines, iTliereas the diphenylphosphonoand diphenylphosphorothionogroups are apparently electron-withdrawing in these con5)ounds. The hydrazinophosphines are more reactive towards nucleophilic reagents than are hydraainophosphine sulfides and hydrazinophosphine oxides, which are less reactive in similarly substituted compounds. Thus, the hydrazinophosphine oxides are resistant to hydrolysis, condensation, and attack by electron-rich reagents, whereas more reactivity is to be expected for hydrazinophosphines and hydrazinophosphine sulfides. The bis(hydrazino)phei^lphosphines are found to be highly reactive, presumably as a result of the presence of the two electronegative hydrazinogroups on the phosphorus atom. Consequently, these caapounds are difficult to obtain in high purity and are extremely sensitive to atmospheric moisture and oxygen. The results of this study have been correlated with current hypotheses concerning the relationships between molecular structure and chemical properties.

PAGE 158

BIBLIOGRAPHY I* Rochow, Hurd, and Lewis, The Chemistry of Organoiaetallic Compounds J. Wiley 6e Sons, Inc., New York (1957). 2* G. E, Coates, OrRano-Metallic Coinpounds J, Wiley & Sons, Inc., New York (1960). 3, H. Zeiss, Editor, Or?xanometal 1 ic Chemistry (A. C. _S. Monograph N o 147) Reinhold Publishing Co., New York (1960). 4 Advances in Chemistry Series, tletal-Organic Compounds American Chemical Society, Washington, D. C. (1959). 5. D. R. Martin, Chem Revs .. 42, 581 (1943). 6. D. R. Martin, ibid .. 34, 461 (1944). 7. II. S. Booth and D. R, Martin, Boron Trifluoride and Its Derivatives J. Wiley & Sons, Inc., New York (1949). 8* N. N. Greenwood and R. L. Martin, Quart Revs., 8, 1 (1954). 9. A. B. Burg, Record Chem Prog C^resge-Ilooker Sci Library) 15, 159 (1954). 10. F. G. A. Stone, Quart Revs .. 9., 174 (1955). 11. M. F. Lappert, Chem. Revs .. ^, 959 (1956). 12. F. G. A. Stone, ibid . 53, lOi (1958). 13. E. Krause and B. Went, Ber .. 56, 466 (1923). 14. E. B. Balcer and H. H. Sisler, J. Am. Chem Soc 75, 4828 (1953). 15. E. B. Baker and H. H. Sisler, ibid .. 25, 5193 (1953). 16. N. Davidson and H. C. Brown, ibid .. 64, 316 (1942). 17. A. W. Laubengayer, J. D, Smith, and G. G. Ehrlich, ibid .. 83 542 (1961). 18. G. Bahr, Fiat Review of German Science 1939-1946 24, Inorganic Chemistry, 155. 149

PAGE 159

150 19. I. Zakharkin and L, A, Savina, Bull Acad Sci U.S.S.Px. . 420 (March, 1957). 20. W. G. Paterson and M. Onyczchulv, Canad J. Chera 39, 2324 (1961) 21. N. R. Fetter and B. Bartocha, ibid .. ^, 2001 (1961). 22. J. K. Ruff and M. F. Hawthorn, J. M' Chem Soc., 82, 2141 (1960) 23. E. Wiberg, H. Graf and R. Uson, 2. anpXR u. allgem Chem . 272 221 (1953). 24. E. Wibert and A. May, Z. Maturfurschg 10b 234 (1955). 25. F. M. Peters and B, Bartocha, Chem and Ind. 1272 (1961). 26. E. Krause and P. Dittman, Ber., 63, 2401 (1930). 27. G. Schoaburg and E. G. Itoffiaan, Z. Elektrochem Ber Bunsenees Phvsik Chem 61 1110 (1957) 28. J. K. Ruff and M. F. Hawthorn, J. M, Chem Soc 33, 535 (1961), 29. J. K. Ruff, ibid .. ^, 2825 (1961). 30. A. W. Laubengayer, Inorganic Polymers (Special Publication No* IS)* P* 78, The Chemical Society (London) (1961). 31. G. Schwarzenbach, Titrations with Complexones Uctikon Chemical Co., Uetikon, Switzerland (1953). 32. G. M.Kosolapoff Organopho sphorus Compound s J. Wiley & Sons, Net7 York (1950) 33. J. R. Van Wazer, Phosphorus and Its Compounds Vol. I, Interscience Publishers, Inc., New York (1958). 34. A. W. Franlc, Chem Revs .. 61, 339 (1961). 35. E. G. Rochow, Editor, Inorganic Syntheses Vol. VI, p. 108, McGraw-Hill Book Co., New York (1960). 36. A. B. Burg and P. J. Slota, J. Am. Chem Soc 80, 1107 (1958). 37. Fritz Ephraim and M. Sackheim, Ber .. 4^, 3416 (1911). 38. E. H. Blair and H. Tolkmith, J. Org Chem .. 25, 1620 (1960). 39. W. Autenrieth and W. Meyer, Ber .. 58, 848 (1925).

PAGE 160

151 40. A. Michaells, ^m., 326, 129 (1903). 41. T. W. Mastin, G. R. Nonrmn and E. A. Weilimienster, J. Am Chem Soc., 67, 1662 (1945). 42. Shou Tung and Shyh-Tsong Chem, Ihia Hsueh Hsueh Pao 24, 30 (1958). (C. A. 52, 199038; 53, 31131) 43. R. Klement and K. Knollimller, Ber 93, 834 (1960). 44. R. Kleoent and K. Knollnwller, Ber .. 93, 1088 (1960). 49* A. G. Zenkevich, P. G. Zaks, Ya. A. Mandelbaixa and N. N. Melnikov, J. Gen Chem U.S.S.R .. 30, 2298 (1960) (Eng.). 46. Etcyl Blair, U. S. 2,855,423, Oct. 7, 1958. 47. Howard M. Tltch, U. S. 2,759,961, Aug. 21, 1956. 48. A. Mlchaelis, Aim., Z21 193 (1896). 49. A. Michaells, Arm., ^93, 261 (1896). 50
PAGE 161

152 63. R. H, Pierson, A, N. Fletcher and E. St. Clair Gantz, Anal Cheia 2Q. 1218 (1956). 64. Backer, Rec. trav chim .. 31, 150 (1912). 65. Beacon Chemical Industries, Inc., 33 Rlchdale Avenue, Canbxtdge ^, I^ssachusetts. 66. D. D. Coffman and S. S. Marvel, J. Ag. Cheia Soc. 5 1 3496 (1929) 67. H. H. Sisler, A. Sarkis, H. S. Ahuja, R. J. Itarago and N. L. Smith, J. M, Chem Soc , ^, 2982 (1959). 68. K. A. Jensen, J. prakt Ghent .. 148 107 (1937). 69. T. W. Garbulis and D, H. TGn?)leton, J. ^. Chem Soc 83 995 (1961). 70. Heilbron, Dictionary of OrRanic Coiapounds Vol, IV, p. 139, Oxford University Press, New York (1953). 71. L. J. Bellany, The Infra-red Spectra of Cowplex Ifolecules John Wiley & Sons, Inc., New York (1960). 72. R. A. tfclvor and C. E. Hubley, Canad J. Chau. 37, 869, (1959) 73. Victor Chemical Vforks (Chicago), Technical Bulletin: Diphenvlphosphinous Chloride (1952) 74. W. A. Henderson, Jr., and C. A. Streuli, J, An. Chem Soc 82, 5791 (1960).

PAGE 162

BIOGRAPHICAL SKETCH Robert Peter Nielsen was bom March 14, 1937, at New Brxmswick, New Jersey, He received his elementary education in the public school system in Highland Park, New Jersey, and was graduated from Highland Park High School in 1954. In Jane, 1958, he received the degree of Bachelor of Science from Rutgers, The State Uiiversity. In 1958, he enrolled in the Graduate School of the Uiiversity of Florida. He worked as a graduate assistant in the Department of Chemistry until June, 1959, and as a teaching assistant until June, 1960. From June, 1960, until the present time he has pursued his work toward the degree Doctor of Philosophy as a research fellow under a grant from the Petroleum Research Fimd administered by the American Chemical Society. Robert Peter Nielsen is married to the former Linda R^ Galbraith. He is a mecaber of the American Chemical Society, The Chemical Society (London), Hii Lambda I^silon, and Sigma Xi. 153

PAGE 163

This dissertation was prepared under the direction of the chairman of the candidate's supervisory committee and has been approved by all members of that committee. It xms stibmitted to the Dean of the College of Arts and Sciences and to the Graduate Council and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy. June 11, 1962 Dean, College of Arts artu Sciences Dean, Graduate School Supervisory Consnittee: ^.. .^'ya iiM'^/V ^ Jii^r^ C^xX yK)^'^A.,,K\A U A C.

PAGE 164

J S3 40 9 <