Citation
Reactions of amine boranes and related compounds: (I) Mechanism of dehydrogenation of dimethylamine borane (II) Synthesis of trimethylamine chloroboranes

Material Information

Title:
Reactions of amine boranes and related compounds: (I) Mechanism of dehydrogenation of dimethylamine borane (II) Synthesis of trimethylamine chloroboranes
Added title page title:
Amine boranes and related compounds, Reactions of
Creator:
Wiggins, James William, 1940-
Place of Publication:
Gainesville
Publisher:
[s.n.]
Publication Date:
Copyright Date:
1966
Language:
English
Physical Description:
xi, 178 l. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Adducts ( jstor )
Amines ( jstor )
Atoms ( jstor )
Boranes ( jstor )
Chlorides ( jstor )
Ethers ( jstor )
Flasks ( jstor )
Heating ( jstor )
Hydrogen ( jstor )
Infrared spectrum ( jstor )
Amines ( lcsh )
Borane ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis - University of Florida.
Bibliography:
Bibliography: l. 175-177.
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:
030112562 ( AlephBibNum )
11062888 ( OCLC )
ACH2405 ( NOTIS )

Downloads

This item has the following downloads:


Full Text












REACTIONS OF AMINE BORANES AND
RELATED COMPOUNDS:
(I) MECHANISM OF DEHYDROGENATION
OF DIMETHYLAMINE BORANE
(II) SYNTHESIS OF TRIMETHYLAMINE
CHLOROBORANES







By
JAMES WILLIAM WIGGINS


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


December, 1966











ACIKNOWLEDGMENTS


I acknowledge with sincere gratitude the assistance

given by the Chairman of my supervisory committee, Dr. G. E.

Ryschkewitsch, during the preparation of this work. His

enthusiasm and patient instruction during the course of the

research made the work a pleasure. The decisive influence

and interest in my professional career of Dr. Ryschkewitsch

has been deeply appreciated.

I sincerely thank the members of my supervisory com-

mittee and the many other faculty members who have expressed

an interest in my growth as a chemist.

I express my thanks for financial support to the

National Science Foundation Grant G19738 and the Chemistry

Department for support on the Science Development Grant.

I thank Mr. D. D. Davis and Dr. Alan Hagopian for

obtaining the mass spectra. I thank Dr. Wallace S. Brey,

Jr., and Dr. K. N. Scott for obtaining the B1 nuclear

magnetic resonance spectra.

A special thanks is extended to Mr. R. G. Logsdon

for building and repairing the glass vacuum system used in

the work.

I thank Mrs. Thyra Johnston for typing the final copy

of the dissertation.










The many enlightening discussions and endeavors

with Dr. Gerhard M. Schmid have made my work toward the

Ph.D. degree a real joy.


iii










TABLE OF CONTENTS

Page
ACKNOWLEDGMENTS . . . . . . . . ii

LIST OF TABLES .. . . . . . . . . . viii

LIST OF FIGURES . . . . . . . . .. x
PART I. MECHANISM OF DEHYDROGENATION OF
DIMETHYLAMINE BORANE

INTRODUCTION. . . . . . . .. . . 1

EXPERIMENTAL .. ............ . . 3

Nomenclature . . a . . . . . 3
Origin of reagents . .. . . ... . 3
Purification of reagents . . . . 3
Instruments. . . . . . . ... 4
General method for the analysis of the
dimethylamine boranes ............ 5
Pyrolysis of B2D6. . . . . . . .. 6
Infrared spectral analysis .......... 8
Mass spectral analysis ... ..... 19
Preparation of (CH ) NBD3 from (CH )3NBHk in an
acidic Dp2solution .. . . . . . . 26
Preparation of B2D6 from (CH )3NBD3 and BF3(g) . 29
Variation in the per cent reaction of (CH3)3NBH3
and B F3(g) to yield B26 . . . . . .. 32
Preparation of B2D6 from NaBD4 and BF3(g) in
diglyme . . . .6 . . . . 36
Preparation of B2H6 from NaBH and BF3(g) in
diglyme . . . . . . . . . . 38
General procedure for preparation of (CH )2ND. . 39
General method for the preparation of (CH )2HNBH3
from B2H6 and (CH)2NH . . . . . 42
Preparation of (CH3)2DNBH3 from (CH )2ND2Cl and
LiBH4 d C *. * *. 46
Hydrolysis of (CH )2DNBD3 in 0.1 M hydrochloric
acid. . . . . . . . 52








Page


Determination of reaction conditions for the
hydrogen elimination reactions. . . . . 53
Hydrogen elimination on.heating dimethylamine
boranes . 55
Experiments to eliminate possibility of isotopic
interchange during the elimination reactions. 58

I. Heating of D2 and H2 in the presence of
mercury vapor . . . . . 58
II. Heating D2 with (CH )2HNBH3 and
(CH )3 NBH3 to determine if exchange
occurred. . . ..... . . 58
III. Heating of (CH )2DNBH3 to determine if
exchange occurred between ND and BH
within the molecule . . . . . 63
IV. Heating of (CHE)2DNBH3 and (CH )2HNBD to
determine if amine exchange occurred. 65

DISCUSSION OF RESULTS . . . . . . . . 68

Possibility of hydrogen-deuterium exchange . . 68
Heating mixtures of dimethylamine boranes
containing various distributions of hydrogen
isotopes for one hour . . . . . 76
Heating mixtures of dimethylamine boranes
containing various distributions of hydrogen
isotopes for twenty-four hours. . . . . 82

CONCLUSION. . . . . . . . . .. . 88

SUMMARY . . . . . . . . .... 90

PART II. SYNTHESIS OF TRIMETHYLAMINE CHLORO-
BORANES

INTRODUCTION. . . . .. . . . .. 92

EXPERIMENTAL. . . . . . . . . . . 95

Nomenclature . . . . . . . . . 95
Reagents and purification. . . . . . . 95
Instruments . . . . . . . . 96
Extraction of BN compounds from the reaction
mixture . . . . . . . . . 97
Infrared spectral analysis . . . . . . 99
Nuclear magnetic resonance spectra . . . . 101







Page


Reaction of (CH3)3NBH3 and HgC12 . .. . . 109
Reaction of (CH3)3NBH3 and HgC12 on a large scale. 110
Reaction of (CH 3) BH2C1 and HgC12 . . ... 114
Attempted reaction of (CH 3)3NBHC12 with HgC12. . 115
Relative rates of reaction of (CH3)3NBH3 with
HgC12 and HC1 in ether at 0. . . . . 116
Reaction of (CH3)3NBH3 and HgC12 in the autoclave. 118
Reaction of (CH3) 3NBH3 and HgC12 in presence of
acetic acid in benzene. . . . . . 119
Reaction of (CH3)3NBH3 and Hg012 in water and in
potassium chloride solutions. . . . . 120
Reaction of (CH )3NBH3 and HgC12in water--the
change in pH with time. . . . . . . 121
Reaction of (CH3)3NBH2C1 and (CH3)3NBHCi2 with
HgC12 in water--the change in pH with time. . 125
Reaction of (CH3)3NBH3 and excess HCl(g) . . 126
Reaction of (CH )3-NBH3 and HCl(g) in benzene . 128
Reaction of (CH-) NBH and concentrated HCl(aq) in
water . . . . . . 129
Reaction of (CH ) cNITBH3 and concentrated HCl(aq) in
benzene . 150
Reaction of (CH3) NBH3 and concentrated HC1(aq) in
carbon tetrachloride. . . .. . . . 131
Reaction of (CH 3)NBH3, (CH ) NBH2Cl and
(CH )3 BHC12 with (CH ) iHCl1. . . . . 131
Reaction of (CH3)3TBH2C1 with SbCl5. . . 135
Reaction of (CH3) NBH3, (CH3)3NBHI2C1,
(CH3)3NBHC12 and (CH3)3NBC13 with SbCl . . 137
Reaction of (CH3)3NBH3 with SOC12. . . . . 138
Reaction of (CH)NBH3 and S02C12 . . . .. 139
Reaction of (CH)3 NBH3 and ZnC12 in glacial acetic
acid. . . . . . . . 140
Reaction of (CH);NBH3 and (iCHO)I NBC1I in an
autoclave and the stability of the mono- and
dichloroborane adducts under these conditions 143












DISCUSSION OF RESULTS AND CONCLUSIONS .

Reactions of (CH3)3NB3H3 and HgC12. .
Reaction of (CH 3)NB3H3 and (CH3)3NHC1.
Reaction of (CH 3)NBHH and SbC15 .
Reaction of (CH3) NBH3 and SbC13 . .
Reaction of (CH3)3NBH3 and S02C12. .
Reaction of (CH 3)NBH3 and SOC12 . .
Conclusions from the reactions . .
Thermal stability of the adducts . .

SUMMARY . . . . .. . .

BIBLIOGRAPHY. . . . . .

BIOGRAPHICAL SKETCH . . . .


* * *

* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *

* * *

* * *

* * *


vii


Page

146

146
158
161
162
164
165
166
169

173

175

178










LIST OF TABLES


Table Page
1. Analysis of Dimethylamine Boranes . . . 7

2. Pyrolysis of B2D6 . . . . . 8

3. Calculation of Per Cent Deuterium in the
Deuterated Dimethylamine Boranes from Infra-
red Spectra Using the CH Deformation Peak as
the Internal Reference. . . . . 14

4. Sensitivity Coefficient of Mass Spectrometer
for H2, HD and D2 . . . . .. 21

5. Tendency for Parent Ion to Lose a Hydrogen
or Deuterium Atom . . . . * *. 25

6. Reaction of (CH3)3NBD3 and BF3(g) . . . 30

7. Reaction of (CH3)31BH3 and B35(g) . . .. 33

8. Preparations of (CH3)3ND. . . . .. 41

9. Preparation of (CH3)2HNBH3 Containing Various
Distributions of Hydrogen Isotopes. . . 44

10. Results of Heating (CH3)2HNiBH3 Containing
Various Distributions of Hydrogen Isotopes. 45

11. Per Cent BD and NH Bonds in (CH3)2DDNBH3
Prepared from (CH3)21ND2Cl and LiBH4 . . 50

12. Reaction Conditions for H2 Elimination
Reactions . . . . . . . .54

13. Results of Hydrogen Elimination by
Dimethylamine Boranes . . . . . . 56

14. Reaction of (CH3)2HIfBH3 and (CH3)3NBH3 with
D2 . . . . . . . . . . 64


viii










Table Page

15. Infrared Spectra of (CH)3 53BH and the
Trimethylamine Chloroboranes. . . . . 100
16. B Nuclear Magnetic Resonance Spectral
Results . . . . . . . . 106
17. Reactions of (CH3)3NBH3 and HgC12 ... . 111
18. Reaction of (CH3) 3NBH3 and HgC12 in Water and
KC1 Solutions . . . . . . .. 122

19. Reactions of (CH3)3NBH3 and Trimethylamine
Chloroboranes with (CH3)NHC1 . . . . 133
20. Reaction of (CH3)3T BH3 and Trimethylamine
Chloroboranes on Heating. . . . . . 144










LIST OF FIGURES


Figure

1. Infrared spectrum of (CH3)2HNBH3 . .

2. Infrared spectrum of (CH3)2DNBH7 .

3. Infrared spectrum of (CH3)2DNBH3 after
heating . . . . . . . .

4. Infrared spectrum of (CH3)2HNBD3 . . .

5. Infrared spectrum of (CH )2DNBD3 . . .
6. Infrared spectrum of (CH3)2HNBH3 after
heating. . . . . . . . . .

7. Mass spectrometer sensitivity coefficient
H2, HD and D2 as a function of the total
pressure . . . . . . . . .

8. Mass spectrum of (CH3)2HIHBH3 at 70 ev. .

9. Mass spectrum of (CH3)2DN3BH at 70 ev. .

10. Mass spectrum of (CH3)2HNBD3 at 70 ev. .

11. Mass spectrum of (CH3)2DNBD3 at 70 ev. .

12. Infrared spectrum of (CH3)3NBD. . .

13. Infrared spectrum of B2D6. . . .

14. Change in total pressure with time for the
reaction of excess (CH3)3NBH3 and BF3(g) )

15. Change in total pressure with time for the
reaction of (CH )3 BH and excess BF3(g) .

16. Infrared spectrum of (CH3)2DNBH3 prepared
from (CH3)2l-ND2C1 and LiBH4 . . .

17. Infrared spectrum of (CH3)3BH3 after
heating with D2 for one hour . . .


Page

S 9
S 10

. 11

S 12

S 15


20

23

23
24

24

28

31

34

34


49


60








Figure Page
18. Infrared spectrum of (CH)2HNBH3 after
heating with D2 for one hour . . . . 60
19. Infrared spectrum of (CH3)2HNBH3 after
heating with D2 for twenty-four hours. . 61

20. Infrared spectrum of (CH )3NBH3 after
heating with D2 for twenty-four hours. . 62

21. Mass spectrum at 70 ev after heating
(CH3)2DNBH3 and (CH3)2HNBD3 for twenty-four
hours. . . . . ..... . . 66

22. Infrared spectrum of (CH3)3BH2C1. . . 102

23. Infrared spectrum of (CH3)3NBHC12. .... 103
24. Infrared spectrum in 300-600 cm1 region of
(CH3) NBH3 and the trimethylamine chloro-
boranes. ...... .. . . . ... 104
25. B1 Nuclear magnetic resonance spectrum of
(CH3)3NBH2C1 . . . . .. . .... 107
26. B1 Nuclear magnetic resonance spectrum of
(CH3)3NBHC12 .............. 108
27. Change in pH with time during the reaction
of (CH3)3NBH3 and HgC12 in water . ... 124
28. Change in pH with time during the reaction
of (CH3)3NBH2C1 and (CH3)3NBHC12 . .. 127
29. Infrared spectrum of the reaction product of
(CH3)3NBH3 and ZnC12 in glacial acetic acid. 142
30. Comparison of product on heating
(CH3)3NBHC12 with and without (CH3)3N
present. .. . . ... . . .. 145












PART I. MECHANISM OF DEHYDROGENATION OF
DIMETHYLAMINE BORANE


INTRODUCTION


When dimethylamine borane is heated, hydrogen is

eliminated and dimethylaminoborane is formed. The mechanism

of this reaction should be the same as for the first step

in the production of borazenes, for example, N-trimethyl-

borazene, by heating monomethylamine borane. Thus, the

reaction mechanism, or molecularity, would be worthy of

investigation.

The reaction of dimethylamine borane to yield hydro-

gen and dimethylaminoborane does not lend itself readily to

common methods of kinetic determination such as measuring

the increase in total pressure or the concentration of any

single species. This is due, in the first case, to such

reactions as the dimerization of dimethylaminoborane or the

disproportionation of the dimethylaminoborane (7) which

occur at significant rates in the temperature range at which

hydrogen elimination can be conveniently measured. In the

second case, the separation of unreacted dimethylamine

borane, or the separation of dimethylaminoborane from the

reaction mixture, would be difficult because the reaction

would be occurring while the separation was being carried out.








2

Since conventional kinetic studies were impractical

for the most part, the following method was used to yield

data which could be used to determine the reaction

molecularity. N-deuterodimethylamine borane-d3 and di-

methylamine borane in a 1:1 molar ratio were heated and the

non-condensible products analyzed in a mass spectrometer.

The ratio of H2:HD:D2 found was compared to that expected

for either a unimolecular or bimolecular reaction. The

results indicated a bimolecular reaction and a kinetic

isotope effect. The isotope effect was investigated using

various isotopic distributions of hydrogen in the dimethyl-

amine borane and analyzing in the mass spectrometer the

gaseous products eliminated on heating. The established

isotope effect was that hydrogen atoms were eliminated more

readily than deuterium atoms.

The possibility of hydrogen-deuterium exchange re-

actions occurring during the elimination reaction was

investigated thoroughly.











EXPERIMENTAL


Nomenclature

The compounds formed by the reaction of an amine and

diborane wera named as amine adducts of borane. The follow-

ing is the list of amine boranes in Part (I):
dimethylamine borane, (CH3)2HNBH ;

dimethylamine borane-d3, (CH3)2HN3D3 ;

N-deuterodimethylamine borane, (CH )2DNBH ;

N-deuterodimethylamine borane-d3, (CH3)2DNBD3 .

Origin of reagents


D20:

(CH3) 3NH3

(CH3)2NSH:

S02C12:

BF3 :

C4H9Li:

(CH3)2HITBH3:

NaBD 4:

LiBH 4:

Purification


Tracerlab, 99.7% D20
Callery Chemical Co.

Matheson Co., Inc.

Matheson, Coleman and Bell Div.

Matheson Co., Inc.

Foote Mineral Co.
Chemical Procurement Laboratories

Alfa-Inorganics

Metal Hydrides, Inc.

of reagents

used without further purification--being
handled in a N2 atmosphere.










(CH3)3NBH

S02C12

(CHi3)2NH


BF3


C4H9Li


(CH3)2HNBH

(CH) 3NBD

NaBD4

LiBH4


sublimed once, then resublimed into the re-
action flask (or tube).

bp 68-700C, was used without further purifi-
cation--transferred in a N2 atmosphere.

stored over Na for over 24 hours in freezer
compartment of refrigerator, then distilled
into reaction flask (or tube).

distilled from a -780 trap through a -1190
trap into a -1960 trap, then distilled into
reaction tube.

used without further purification--trans-
ferred with a syringe under a flowing stream
of N2.

sublimed once, then resublimed into the re-
action flask (or tube).

sublimed after preparation and then resublimed
into reaction tubes.

used without further purification.

used without further purification.


Instruments

The vacuum system used in the experimental work was

similar to the vacuum line described in Synthetic Inorganic

Chemistry by W. L. Jolly (21). Apiezon N grease was used

on all ground joints in the system.

A Bendix Time-of-Flight mass spectrometer was used

to obtain the mass spectra.

A Beckman IR-lO or a Perkin-Elmer 21 spectrophoto-

meter was used to obtain the infrared spectra in either

the gas phase or in a carbon tetrachloride solution.










General method for the analysis of the dimethylamineboranes

Dimethylamineborane was sublimed from a weighed

storage flask into a 50 ml round bottom flask. After the

sublimation, the storage flask was weighed, the difference

in weight being the amount of sample to be analyzed.

Analysis was based on the equation


(CH )2N-IBH3 + 2H20 + H 130+ (CH )2NH2+ + B(OH)3

+ 3H2


[1]


The compound was first hydrolyzed by condensing 20

ml of 0.1 N HC1 (Acculute) into the flask containing the

dimethylamine borane. Hydrolysis was allowed to continue

overnight at room temperature. The contents of the re-

action flask were condensed in a liquid N2 bath and the

non-condensible gas was transferred into a calibrated bulb

by a Toepler pump. The amount of hydrolyzable hydrogen was

thus obtained.

The acid solution from the hydrolysis of dimethyl-

amine borane was transferred into a 400 ml, boron-free glass

beaker. The boron was determined as boric acid using the

mannitol titration method.

The amount of nitrogen, as dimethylammonium ion,

could be determined since a known amount of strong acid had

been used to hydrolyze the sample. The difference in the

equivalents of strong acid added initially and the equiva-











lents of base necessary to neutralize the strong acid

remaining after the hydrolysis reaction was the amount of

strong acid neutralized by the dimethylamine.

Using this method, the following data in Table 1

were determined.

Pyrolysis of B2D6


Diborane-d6 was pyrolyzed by passing the gas through

a 9 mm Vycor tube, 37.5 cm long, surrounded by a 0.75 inch

stainless steel pipe and heated by two Meeker burners. The

gas was allowed to pass slowly into the tube. Attempts to

make more than one pass of material through the hot tube

did not increase the amount of non-condensible gas; evi-

dently no condensible material passed through the hot tube.

The non-condensible products were Toepler pumped into a
/-
flask and analyzed in the mass spectrometer. Results of

these experiments were as follows, in Table 2.






















OL 0
00 0



00 0
00 0
* *
HH H



00 0



00 0









NH H\
(\0 H '-\
* *


LC\E'- LCUM\


C00M CUC0
JC\j OJC\j

coo3 cOO
r-i,-\ i \ J




rd rd



O CO'd
H H


S0 0
o do



M4 OC




0 0


O


O
0




0
0

r-t






0
**
O'-.O'


LSN' N-\ r-N
0 r- u\
ir UM S


00 0 Lr\ 0000
00 N-\CO cc O
oj r(\j o rC^


Lr\0 0 t rc\
z-" 000 KlOJ
-- Oco t%" CMO
-t r-. r- OJ Od


rd rc rd
C .) 4O
d d d



do do do
cdl crl c0

S-i 04-i
0 0 0 d 0









v v v


CO






0








H


O

H
-i:
1 Cl <


(a)



Cd
) 4-I
r-t
*H


o c

0 0
h p


Td









TABLE 2

PYROLYSIS OF B2D6


Prepared by B2D6 Per Cent Per Cent
mmoles Reaction D2 HD H2
mmoles 2 2

(CH3) 3NBD3 +

BF3(g) 0.098 105.8 81.5 18.5 -

NaBD4 + BF (g) 0.12 88.9 91.0 8.4 0.5


Infrared spectral analysis

The spectra of the dimethylamine boranes (Figures 1,

2, 3, 4, 5) were taken on a Beckman IR-10 spectrophotometer,

using matched cells 0.2mm thick and the slow scan speed.

The compounds were dissolved in Fischer spectroanalyzed

grade carbon tetrachloride and the same solvent was used as

a blank in the reference beam. The absorbency, A, expressed
Intensity blank
as the log ntnityion was calculated from the
Intensity solution
infrared spectra in which transmittancy was plotted as a

function of the wave number.

The absorbency of the CH deformation peak at 1475 cm-1

was used as the internal reference and all calculations were

made with respect to this peak. The results are given in

Table 3.




















Clj



0




rl~
0
H



0








0
>-$
















o N
CC






NZ


W4





00
N
C E
O 0
























0
C) C
cCj
o I
N~t
H c












-~ 1 0






o~~ 0
'.0 4
G~Ue~w~ue~~4.UG .C&










































































aouBiTulssuea Juao Jad























C)





r-



rl


ad(





CD
1 0,
r4

(N









0
Co M
rl a

OZ
r-





:J
>
-Q






h

1-4







0
C)
r(N
















C\





C)










aN -~~lsv! uO ladN










12








o
0



0





















4,
o
O









-o N
CC
H -



o 0







0
> Th





CO


0I
*
o











0




o









I I & a

0 0 0 0 0
CO 0 N










13









0
0









0
_0





o N



0
r-I



Q






I 0






00 ) 0
o -
0 s





0- .0 a














C:)
cO c?


N( I


































C)
Cl"
C




S I I

cot























0;u 0; 0Gd
"0 -O N












0



r


FW4





HE-
p C
H











S0
H



E M
H





CO

pq
H







0 P








Hm
^i

00


a.

Q






-P









bO










0
0 *


H H
60
o


N\J -
t. 0
o0 L0
0 O
00


S C0
lu Lrl 0CO
H 00C


cO


0
CO






0-

0
O


1N H
K\ r-I
iLr co
Lc CO
0 -
00


C\J
0 0D
L\ H
* O
H 0






0D KI 0
ON CO rci
o 00




















z~X FI


M A
Oj N p fi p
4.^ 2 pq pq0
r^
Uf
0


0 0(7N r-
0 0 0'i
J co KI
N NcON.~C


co 0
r- ^--
C C\j


NCH-
H- c
r CO
r- L0
r (


O pq U
0


M
X




rc\
U
v
















p-4


co-P
43 0
0 0





l 0


0
4,*


0

cu 44,


00


00
LO

00
Lf OJ
* C


etM

O

t^
R
0
^^>


CO



ON 0
r-


\ C- cOJ
r- cO 0
r-i


o
0
*rl

Pi
CO r-
M 0


HH

b0
0








rd









0
P4
ri
0
0


00

H ,'
00


C;


-p



4P


"d
a)






0
0




0
0O

r*
rd



r.


0



,a





0
rd















43
rd





4
0






4,
0
0



0,
1-1
,
r0


r-


00
; o


LN 0
O D
N I'


M


Fl
~I

M
X
v








The ratio of the absorptivities of the BH:CH and the

NH:CH in the spectrum of the completely undeuterated sample

were found to be 1.90 and 0.992, respectively. The per cent

NH and ND were calculated as follows:

ANH aNH CNH
CH aCH CH

ANH/ACH calculated from spectra

aNH/aCH absorptivity ratio from spectrum of undeutera-

ted compound

CNH/CCH ratio of concentrations

Assume in partially deuterated compounds that

CNH + CND = CCH

CBH + CBD = CCH

This assumption allowed C D to be determined without the

actual absorbency being known. This was necessary since

the ND and BH stretching frequencies both occur between

2300-2500 cm-l It was not possible to separate the ab-

sorbency due to each vibration. Since (CH3)2DNBD3 was known
not to be completely deuterated from mass spectral data of

the B2D6 pyrolysis product, the absorptivity ratio of BD:CH

could not be obtained for its infrared spectrum. Therefore,

the ratio of the absorptivities for BD:CH was assumed to be

the same as that for BH:CH.









The absorptivity ratio of BH:CH for the volatile re-

action product after heating (CH3)2HNBH3 for thirty-four

hours at 1000 was 1.50 (Figure 6). Thus, one of the BH

containing products which was formed when dimethylamine

borane was heated, did not absorb as strongly in the BH

stretching vibration region as did the original starting

material. This could cause a lower estimation of the amount

of BH-containing material after heating than would actually

be present and therefore introduce an error in any calcula-

tions made using the BH stretching absorption.

The absorption in the 1700-1800 cm1 region which

was attributed to the BD stretching vibration in the mole-

cule (CH )2HNBX3 [X=H and/or D] did not occur at exactly

the same wave number in each spectra. The absorption varied
-1
from 1755 to 1785 cm Qualitatively, this variation

appeared to be concentration dependent. The greater the

concentration of BD bonds in the molecule the larger the

wave number at which the absorption occurred. The absorption

for the BD stretching vibration in (CH3)2DNBD3 occurred at

1785 cm-1 and for (CH3)2DNBE3 containing 5 per cent BD bonds,
at 1735 cm" An explanation for this variation could be

that a shift in the stretching vibration occurred in the

-BH3 group as the hydrogen atoms in the -BH3 group were re-

placed by deuterium atoms. The largest shift was implied by

the spectra at low D2 percentages when mostly BH2D groups

should have been present.
























0 o
O











--o
0









0
> >











C)






0


O







0
o












0













o. 0
0










0C






















0- 00 0 0
r

^=

*r ---
^*^,

-- -- -------- = ~ --- .
., __________ ---*-- >.
/-- -***

.1^ ,, :
_^-t --- g ------------ ---- ^ ----- ^ ------ ^ ----- j ------ ---






0 0 0 0
00 ^0-^- (
souB~usuej. q~uo js









Mass soectral analysis

A Bendix Time-of-Flight Mass Spectrometer was used for

the analysis. The sensitivity coefficient of the spectro-

meter to H2 and D2 was determined before each set of analyses,

using commercial samples of hydrogen and deuterium gas. The

sensitivity coefficients used in calculating the data were

obtained at the same total pressure as that in the sample.

In calculating the results of the hydrogen elimination re-

action the sensitivity coefficient for HD was assumed to be

intermediate between that of H2 and D2. This assumption

proved to be a valid one as may be seen by the graphs of

pressure versus sensitivity coefficient in Figure 7.

Three sets of sensitivity coefficients were determined

and listed in Table 4. A sensitivity coefficient for HD

was determined from a gas sample prepared by the hydrolysis

of CaH2 with D20. This gas sample contained small amounts

of D2 and H2, but it was predominantly HD. The spectra

were corrected for the presence of D2 and H2 and then the

sensitivity coefficient of HD was calculated. The variation

in the sensitivity coefficients with total pressure was

determined and plotted in Figure 7. In samples containing

D2, the H2 content was determined from the peak at m/e 2

after subtracting the portion due to D+. The necessary

data were obtained from the intensity ratio of m/e 4 and

m/e 2 peaks in the mass spectrum of pure D2.













O





0O
O O


1.60 -




1.40




1.20 -


0


0


o
^^~~~o


I 6
60


1 I
140


I 1
180


220


Pressure, microns
Fig. 7.--Mass spectrometer sensitivity coefficient of H2,
HD and D2 as a function of total pressure.
(0) D2; (G) HD; (o) H2.


2.20 -




2.00-




1.80-


1.00




.80


- -


















0
Cm Cto 0


oJ a 0 o *


/ -pq -r* 0 4+
N^ HQ o a *
SC OH Od 0 .
c 00 r-X l 4- '0 0
Sr 0 0 >-NP 0

d 0 a -4 0-p
coj o + o j o 0
o rl E 0- a o -l
SO .l a- b
o 't 0 d d ( o- I
O 0 4- CO 0 *r*

E1 P 0 I
SO -H O
0 P A C ca O

0 0W


E- C 1 0
EO 0
O OJ 0



o H m 1-e-
dH rd ( C
0 r, 0)




E- 1 4- rd ,C r
CdLCo 0 %
H C!S 0 0) 0P
0 r P *HH 0 a Y




W P: P 0
o o "- o oo
O C;H L N H rd

H 0 rH 0C
E-1 0 0 *H rP

H 0 0 0C
E- P 4-> F-'d H
H *i *j-4 00) 0a
M > g-P 0
S 4 OJ CM
m o1
0 0
CO 0 0 d









The mass spectra of the solid dimethylamineboranes

were obtained at 70 ev by placing the sample on the end of

a probe which extended into the ionization chamber. The

probe was at room temperature during the measurement of the

spectra. The vapor pressure of dimethylamineborane at room

temperature was sufficient to produce good spectra. For

the results of these analyses see Figures 8, 9, 10 and 11.

The mass spectra of the solid compounds showed a low

intensity peak at the m/e corresponding to the mass of the

parent compound. The mass peak of m/e one less than the

parent peak, and the parent peak had a ratio of 8.4 in the

cases where the hydrogen was bonded to the boron atom and

a deuterium bonded to the nitrogen atom compared to 7.9 for

the reverse case where hydrogen was bonded to nitrogen and

deuterium to boron. This suggested that the BH bonds were

lost more readily than NH bonds. The spectra of the com-

pletely deuterated and undeuterated compounds also showed

the same trend as indicated in Table 5. The ratio of the

m/e peak of mass two less than the parent peak, to the parent

peak, implied that a BD bond was lost much more readily than

an ND bond.

The most intense m/e peak in each spectrum corres-

ponded to the loss of a deuterium atom when the compound

contained BD bonds and to the loss of a hydrogen atom when

the compound contained BH bonds. But the numerous peaks, in












S\0


--1


Lo





- rD


-o a,
\o
10












N0
o- oi
cU




0)
--u'h















-0 C
o o
C\
o



o






1-
- 0


~ -H


XI.TSuauI


-4


-


Q)
0
C'-





4-




o
cr














-l
.ON
0

0*)



U,






*-
*l
h0
*>-
&I


j3rTsueGuI


_1









-I
3


.XCsuaquI


--1



1




I









~j


o
"o o
$0






-I
-o o















-p
-0 4















Cl
ho







- Y.H


















0
I
r



S- t










-0
O






OO
a













-0 0
0 l








I
-0 r-"

*
--p


























0 D
aH 5H
0 -




S m N 0 0r3

C) *\ 000*
0 0 tC Cx L LCa'


S 0 0 0 0
B m*
o p H*)* ** * LA




H CH 0 C\ r-i C
o d
4 i 0HKH CO 0-

0 02-P LA


>0 H 0 00
a 0 0 0 0




O
E- 1 -P
.al 0 K 0 tx\ Cx 0













0
E-100 3 rN










E-ia 1 L
H







0







E-l0 Cx Cxi C CM
a rc'\ C tC\
u u o o o
<^ l u u -' <-









each group of peaks differing only by one m/e unit, suggested

that the CH bonds in the methyl groups were also being broken

under the conditions at which the spectra were obtained.

This would make any quantitative use of the relative peak

intensities open to doubt as to whether the hydrogen atom

lost had been originally bonded to a carbon, nitrogen or

boron atom.

Preparation of (CH )3IBD3 from (CH ) NBH3 in an acidic D20

solution (11)

Sulfuryl chloride (0.5 ml) was pipetted into a flask

containing deuterium oxide (20 ml) in a dry nitrogen atmos-

phere and stirred magnetically for twenty minutes. Tri-

methylamineborane (15.7 mmoles) was dissolved into 50 ml of

anhydrous diethyl ether in a 200 ml round bottom flask. The

deuterium oxide solution was poured into the ether solution

and immediately the flask was fitted with an adapter for the

vacuum system, attached to the system, and submerged in

liquid N2, and evacuated.

The reaction flask was warmed to room temperature

(240) and stirred magnetically. After six hours, the re-

action mixture was condensed in liquid N2 and, in approxi-

mately a 400 ml volume, there was 95.5 mm of non-condensible

gas. The D20-ether mixture was transferred to a separatory

funnel and the reaction flask washed with a 20 ml portion of

ether which was then added to the reaction mixture. The









layers were separated, keeping the ether layer in the

separatory funnel. Excess anhydrous potassium carbonate was

added to the ether solution and the mixture was set aside

for forty-five minutes. The ether solution was transferred

into a 200 ml round bottom flask, washing the K2003 with two

20 ml portions of anhydrous ether and the wash solutions

added to the ether solution.

The ether was removed by distilling under vacuum from

room temperature into a liquid N2 trap. When liquid ether

was no longer visible, a -780 bath was placed about the

flask and the last of the ether removed into the liquid N2

trap. This was to prevent loss of product by sublimation.

A white solid residue remained. A white product was sub-

limed from this residue to give a 74.2 per cent yield.

The infrared spectrum (Figure 12) of the sublimed

product in CC14 solution agreed with the spectrum of tri-

methylamineborane with the peaks attributed to a BD stretch-

ing vibration shifted to longer wavelengths. There was a

peak in the region of the BH stretching vibration, but it

was less intense than the BD peak. Assuming the absorptivity

coefficient, a, to be the same for both the BD and BH con-

taining compounds, from Beer's law, A = abc, the concentra-

tion of the BH compound was calculated to be 2.7 per cent of

the concentration of the BD compound. No other analyses

were made of this compound.











80






70










o

250







-pI
C,




30 -






20






10 -



2500 2200 2000 1800

Wave number, cm-1
Fig. 12.--Infrared spectrum of (CH3)3NBD3.










Preparation of B2D6 from (CH )3NBD5 and BF (g)

Trimethylamine borane-d3 (4.0 mmoles) was sublimed

into a reaction tube fitted with a stopcock and a side arm

filled with mercury such that if the tube were inverted the

mercury sealed the stopcock from the contents of the tube.

Boron trifluoride gas was entered into the vacuum system

from the storage tank. The gas was purified by distilling

from a CC14-CHC13-CO2(s) trap (-780), through an ethylbromide

slush (-1190), into a liquid N2 bath (-1960), before con-

densing into the reaction tube. The reaction tube was then

warmed to room temperature and set aside for an extended

period of time.

A liquid phase was present in the reaction tube after

the tube warmed to room temperature, but the liquid phase

slowly disappeared.a After eighty to ninety hours at room

temperature, the reaction tube was attached to the vacuum

system, the products condensed in liquid N2)and any non-

condensible gas removed. The reaction tube was then warmed
to -780 and the volatile fraction was removed and condensed

onto excess anhydrous diethylether. Any unreacted BF3(g)
would form the etherate and the B2D6 could be separated from

it. The ether flask was warmed to -780 and a product, B2D6,


aUsually this occurred overnight. Care must be taken
to prevent this liquid phase from holding the mercury next
to the stopcock on solidifying and thus sealing all gaseous
product in the tube.









was distilled from a -780 bath, through a -119 bath, into
a -1960 bath. The distillation was done rapidly to prevent

contamination of the B2D6 with ether vapor. The material
in the -1960 trap was diborane-d6.
The results of the experiments were as follows in
Table 6.
TABLE 6
REACTION OF (CH3)3NBD3 AN D BF3(g)


Compound Mmoles Yield Yield Vapor Pres-
mmoles Per Cent sure at CS2
Slusha
(CH)3NBD3 4.00 1.64 82.0 239.0 mm

BF3(g) 5.71
(CH)3NBD3 2.11 0.90 84.9 238.5 mm

BF3(g) 3.78

aLiterature value is 238.3 mm (6).

The infrared spectrum of the B2D6 was in agreement
with the reported spectrum (43). The resolution of the
spectrum (Figure 13) was poor but it did show a low intensity
at 2500 cm-1 which was due to a BH stretching vibration, and
the intense BD stretching vibrations at 1810-1840 cm1 and

1954 cm-1. A rough estimate of the concentration of BH to
BD from Beer's law gave a ratio for CBH/CBD of 0.15. The
absorptivity of BH and BD were assumed to be equal in this
calculation.






























C)




0 ,

00-






30-





20-







2500 2200 2000 1800

Wave number, cm-1


Fig. 13.--Infrared spectrum of B2D6.









A sample of the diborane-d6 was pyrolyzed and a mass

spectrum run on the products. For the results of this

experiment see Table 2.

Variation in the per cent reaction of (CH,) NBH, and BF (g)

to yield B2H6


Boron trifluoride will displace diborane from tri-

methylamine borane according to the equation (28):

1
(CH 3)NBH3 + BF(g () -(CH ) BF3 + B2H [2]

The extent of this reaction in mercury-sealed bulbs

was studied as a function of time in order to determine the

optimum time for the practical synthesis of diborane at

room temperature. The following variations, listed in Table

7, in the per cent reaction were determined.

The extent of reaction was followed by measuring the

total pressure in the reaction flask as a function of time.

For complete reaction, the total pressure should be one

half of the initial pressure according to equation [2]. The

graphs (Figures 14 and 15) showing the variation in total

pressure with time indicated that after fifty hours the re-

action, for practical purposes, was complete since the addi-

tional amount of diborane produced in the next thirty hours

did not warrant the extra time spent. The plots indicated

little difference in the rate of decrease in total pressure























LPl


^- N Cd- d- N N K1 C C


LR rWc L>\ Nt< r(\






L\ CNJ tO 0 -

CD -d- t 00- 0
V. Dz DC'cO


LD OJ Lr\





D 0o

C- co co


),\ c"- c"- co WN ti Z- 0
t"- ;- tD cO oO .D 0 .O O
OIO
00000 C~lrHO0


( JL OJ O -










(L' Lt\ L-> L0
N 0 0 0
\ C\J C\j nC\j


C'-~





0


O4

0
H
E-


)
0 ) MS
AP40) N
- P-


CO









p
0)





0r

F-e
-p




H
0










0)0
rC








10-N
co









rd
m *






00

0 C
CO








-P a
c0



0)
'
*r


a
H 0





0) d








oi
! C
aw




I P


CO
0)

H


0 \ i CO
,, to 0- 0-
*\ %'- %

qc







O 0
r^ C 0 i-
-s *

r^ O --O


H -)
WD 0



+


rcd




00
o o
O v















































0 0
\O CN
C^\ C\


nuau 'ajnssaja


o -P
.1-4 ca


O m


0) C


00 M
o C oh
-o ~ -i-' 0

r 0 0)


(1) Ot)
-- *C 0


-4 OX




*
U. C 0








0 4
CQ


*,


















3 cl


.0 0

-P




,) an



E- 0 C
S-H
H 00





C c $-





M




r-q b.0
*H -H
rx 0 ^
1 -P

?-
-i- o~


0 0
C0 c\


CO
co










with time whether BF3(g) or (CH3)3NBH3 were in excess, al-

though the sample containing excess BF3(g) had a 10 per

cent greater extent of reaction.

The data implied that at room temperature the re-
action was only 85 per cent complete in eighty to ninety
hours. It could be possible for the reaction mixture to

reach an equilibrium state in which B2H6 was displacing

BF3(g) in the reverse reaction according to the equation:
1
(CH3) NBH + BF3(g) = (CH ) NBF + 2 B2H [3]

An equilibrium such as this could explain the small change
in total pressure after sixty hours. The attainment of an
equilibrium state is supported by the work of Miller and
Onyszchuk (28)2who found in forty-five minutes at 1300-1400.

an average displacement of 23.4 per cent BF3(g) in

(CH3)3NBF3 by B2H6 and an average displacement of 83.3 per
cent B2H6 from (CH3) 3NBH3 by BF3(g). However, Graham and
Stone (17) reported that after heating B2H6 and (CH )3NBF3
for twelve hours at 800 the gas did not show any evidence
of BF3 in the infrared spectrum. They concluded that no
reaction had occurred under these conditions. The implica-
tions appear to be that for the equilibrium to be established,
the temperature must be greater than 800 or the time must be

longer than twelve hours. But the data show that the rate
of displacement depends greatly on temperature. Miller and










Onyszchuk (27) achieved the same per cent displacement in

forty-five minutes at 1300-141O that we obtained in approxi-

mately ninety hours at room temperature (230-250).

The displacement reaction proved to be an impractical

method for preparing diborane. It was used in this work to

prepare deuterated diborane. R. E. Davis (11) had reported

the exchange of boron hydrogens in trimethylamine borane

with acidic D20 to be rapid and quantitative. In this

manner, (CH )3NBD3 could be prepared and then B2D6 displaced

from the adduct by BF3(g). Thus, B2D6 could be prepared

from readily available and inexpensive starting materials

without the use of deuterium gas to deuterate the diborane,

or without the use of a borodeuteride salt.

The displacement reaction did not give diborane of

sufficiently high deuterium content and B2D6 was prepared

afterwards with HTaBD4 as the source of deuterium.

Preparation of B2D6 from NaBD4 and BF3(g) in diglyme (4)

Sodium borodeuteride (1.0 gram) was placed in a 100

ml round bottom reaction flask in the Dri-Lab controlled

atmosphere box.a A stopcock adapter for attaching the flask

to the vacuum system was added to the reaction flask. The

flask was then attached to the vacuum system and was evacua-

ted.

aA static charge on the powdered NTaBD4 prevented a
quantitative transfer of the material from the glassine
weighing paper into the flask.










The diglyme (ethylene glycol dimethyl ether) to be

used as the solvent was refluxed and distilled from sodium

metal, again distilled from LiAlH4, and finally transferred

from LiAlH under vacuum into the reaction flask at -1960C.

Boron trifluoride was condensed into the vacuum

system directly from the storage tank. The BF3 was then

purified by distilling it from a CC14-CHC13-C02 trap (-780)

through an ethyl bromide slush trap (-ll19) into a liquid N2

trap (-1960). The gas was then condensed into the reaction

flask submerged in a liquid N2 bath. A total of 53.14

mmoles of BF3 was condensed into the reaction flask. The

molar ratio of BF (g) : NaBD~ was 2.2:1.

The reaction flask was allowed to warm to room tempera-

ture (220). After thirty minutes, the reaction was cooled

to -780 and the volatile fraction removed into a -1960 trap.

This procedure was repeated twice with reaction times at

220 of thirty minutes and sixty minutes. The total amount

of volatile material removed from the reaction flask was

10.76 mmoles. The material had a vapor pressure of 254.0 mm

at carbon disulfide slush temperature (-111.90). The

literature value (6) for the vapor pressure of B2D6 at this

temperature was 238.5 mm.

To remove any possible BPF impurity in the B2D6, the

gas was condensed onto anhydrous diethyl ether. The gas-

diethyl ether mixture was warmed to -780 and the volatile










portion was transferred into a liquid N2 trap. After thirty

minutes, 10.55 mmoles of volatile material had transferred

from the ether flask. This material had a vapor pressure

of 258.5 mm in a carbon disulfide slush bath, in agreement

with the previously cited literature value.

An infrared spectrum was run on the material before

it was reacted with ether. The spectrum was in agreement

with that reported for B2D6 (45). No spectral evidence was

noted for the impurity which was removed by the diethyl

ether. There was a low intensity peak at 2510 cm-. A

rough estimate using Beer's law showed, that according to
this BH peak, the ratio of CBH : CBD was 0.058 assuming that

the absorptivity for BH and BD are equal. The assumption

was correct for the absorptions in the infrared spectra of

(CH ) NBH5 and (CH3) 3NBD3.
A sample of this B2D6 was pyrolyzed and a mass

spectrum run on the non-condensible products. For the

results of this experiment see Table 2.

Preparation of B2H6 from NaBH4 and BF3(g) in diglyme

This preparation was done in the same manner as the
preparation of B2D6 from NaBDL and BF (g) in diglyme. The
amounts of reagents used were 0.4 g (10.58 mmoles) NaBH4 and

22.25 mmoles BF3(g). The molar ratio of BF3(g) : NaBH,

was 2.1:1.










A product was isolated in 78.7 per cent yield (7.05

mmoles) which had a vapor pressure of 225 mm at carbon di-

sulfide slush temperature. The reported value (6) at this

temperature is 225 mm. An infrared spectrum of this compound

was identical with that reported for B2H6 in the literature

(43).

General procedure for preparation of (CH3)2ND

A solution of n-butyl lithium in n-hexane (1.6 M)

was syringed into a 100 ml round bottom flask under a stream

of nitrogen. An adapter to the vacuum system was inserted

immediately into the flask, the contents were condensed in

a liquid N2 bath, and the flask evacuated. Excess dimethyl-

amine which had been stored over sodium metal was condensed

into the flask containing the hexane solution.

The reaction flask was then warmed to room tempera-

ture and immediately a white precipitate appeared. After

thirty minutes at room temperature, the reaction flask was

cooled to 0 for thirty minutes, and then a volatile fraction

was removed into a liquid N2 trap. The transfer was done

slowly to prevent excessive spattering of the white solid

as the liquid phase was removed. The remaining excess di-

methylamine, n-butane, and n-hexane were removed with the

flask at room temperature. A white solid residue, LiN(CH )2,

remained in the flask, according to the equation:










(CH3)2NH + n-qH9Li LiN(CH3)2 + n-C4H10 [43

A vial equipped with a capillary break-off tip con-

taining deuterium oxide (1 ml) was attached to the vacuum

system, the tip of the vial was broken and the D20 con-

densed onto the amide salt in a liquid N2 bath. The re-

action flask was slowly warmed to room temperature and an

immediate increase in pressure was noted. The white solid

had turned dark brown after fifteen minutes at room

temperature. Part of the flask was cooled to 0 and kept

at this temperature for one hour and forty-five minutes.

All the volatile material in the reaction flask was trans-

ferred into a flask containing excess anhydrous potassium

carbonate in a liquid N2 bath. The K2CO3 flask was warmed

to 00 in an ice bath and kept at this temperature for two

hours. The K2C03 mixture was then cooled to -780 and a

volatile fraction removed into a liquid N2 trap, requiring

approximately forty-five minutes. This material was

deuterated dimethylamine. A possible impurity in this ma-

terial would be monodeuterated n-butane due to incomplete

reaction of the n-butyl lithium and the dimethylamine

according to equation [4].

The results of the experiments are as shown in


Table 8.
























So
o
0O
0 42
Cd -i cd
hp-i
>>(




kdr
0 0oH




0 ,

rd



o4
0 Q
0 0
om





N 0










-p


M

0

U,
0
H



E-


Ll\ Lu\





C- --t -




CO r-
* *






















c~ o
* (*
00 0








oj u\ ><
Oj Oj O






H H 8


M
C\j


0
0rd

0 0o
o
*H 0 -p

0 0

o Cd
0o o

o N
0: 0
0 e-




co p





\ o rI



*- 0 0




X p
0 4H- 0
-P 00


c 0 0)

H O *0H
O C I O
0 0

-p-
*H CH *H0
0 (0 -0



d 1 0.



i E-iC )
dHriCt









Gas phase infrared spectra of these different prepa-

rations were identical. The spectra were similar to the

spectrum of undeuterated dimethylamine (34) except for some

shifting of peaks in the 1300 cm-1 to 1000 cm-1 region.

The deuterated amine which had a vapor pressure at 00 great-
11
er than 760mm did have an extra peak at 2150 cm-1. An

attempt to purify this sample by distilling a fraction from

an ethylbromide slush (-1190) into a liquid N2 trap resulted

in an increase in intensity in the infrared spectrum of the

peak at 2150 cm1 in the fraction which transferred into

the liquid N2 trap. The infrared absorption peaks of n-

butane (20) were not detectable in the spectrum. The peak

at 2150 cm-1 could be attributed to a C-D stretching fre-

quency in monodeuterated n-butane. The C-D stretching

frequency in the deuterated methanes varies from 2085 cm1

in CD4 to 2205 cm1 in DCH3 (29). This material was more

volatile than-the amine, which was consistent with the

relative vapor pressures of dimethylamine and n-butane. n-

Butane has a higher vapor pressure than dimethylamine (8).

General method for the preparation of (CH )2HNBH3 from B2H6
and (CH3)2NH

Diborane and an excess of dimethylamine were condensed

into a 50 ml round bottom flask in a liquid nitrogen bath.

A CC14-CHC1 -CO2(s) bath (-780) was then placed about the










flask and it remained at this temperature for an extended

period of time (for the exact reaction times see Table 9).

The excess amine was removed from the flask at 0 into a

liquid N2 trap, and the liquid product remaining in the

flask slowly solidified on storage at room temperature.

Mass spectra of the boranes showed a mass peak of

low intensity corresponding to the mass to charge ratio of

the parent ion, and a high intensity peak at a mass to

charge ratio corresponding to the loss of a hydrogen or

deuterium atom.

The results of the preparations of the variously

deuterated dimethylamineboranes are given in Table 9. For

the method of analysis see page 5. The infrared spectra

are given in Figures 2, 4 and 5.

Each of the variably deuterated dimethylamine

boranes was heated at 100-1020 and the non-condensible re-

action product analyzed in the mass spectrometer for the

percentages of D2, HD, and H2. The results of these elimi-

nation reactions are given in Table 10.

The infrared spectrum of the compounds in carbon

tetrachloride solution, in each case, contained peaks where

the NH, ND, BH, and BD stretching vibrations occur. Beer's

law was used to calculate the relative percentages of each

compound using the CH deformation absorption as the internal

reference. For the results see Table 3.




















pt4 Lr N
0 ON ON ON
o r-O r rd
0 -o rll r*- rlcl rcr
G-o H Hi r-1 rH
1-H *


H
P4
E- erd rd rd rd



o 0o op a aU a C.


p H. -P
0 O O O k PO .N ,-
0 D0 Hrd 0 0 H
o D O a Ot O O O 0 L 0
ZO 4j + 0 *
aN H E I C
0 d o 0
Pq E- C/ -i 0 -4
CZH C -. v *P

E N. 'r ) 0 0 0

C 10 co- 0F-
0 0 4a 0 w



c Erl co C I O G

+m ,M
0cU) 0- W H


H Ur 4 M O0 ri 0 N 0 KH aC A,
% gl O 0O o a U)r .
od
S4-' o- o
lr cd rd
K om oI pi O

Poo COM L)
00 t., 0 C 0 n C
.- 0 OJ O C\ O\ %9 0 N 0






























CO
0



Sco
OR


HO
SH
EH
0 0



M





O H
0
m
F










0
0H
CO


-) *H0
0-
Co

a P












0
a
E-l













rd
0
l-







0






o
0


e- C-
I I *
H- d-


Hl LC\
* *



D 0^
r01 1

cO *
0 0











(2 (0


o o

o4


0 *


C-O uL

o0 0
C -Z


0
*


d-
o

ox


Lr\ C-I 0 0











Inn a P
O t H C'-











CM CM\ CM CM

0 0 0 0
oC or o)
KIN


- I'llJ










Preparation of (CH3)2DNBH3 from (CH3)2ND2C1 and LiBH4

N-Deuterodimethylammonium chloride was prepared by

condensing N-deuterodimethyl amine (7.67 mmoles) into a

tared 50 ml round bottom flask containing deuterium oxide

(1 ml) and thionyl chloride (15 mmoles).a The reaction

flask was warmed to 00. After one and one-half hours, all

volatile material was removed from the flask. The increase

in weight of the reaction flask implied that only 0.86

mmoles of product was formed. The volatile material was

transferred back to the reaction flask.

A gas phase infrared spectrum of the most volatile

materials in the reaction flask was identical to that of

S02(g) (3 ) and mono-deuterated n-butane which was known
to be a contaminant in the (CH3)2ND used. The n-butane
and some of the S02(g) was transferred from the flask at
-780 into a -1960 trap in thirty minutes.

More deuterium oxide (1 ml, making a total of 2 ml)
was distilled into the reaction flask and dimethylamine

(6.3 mmoles) was condensed into the flask. After forty-five
minutes at 00 and thirty minutes at 250, all the volatile
material was removed; the weight gain by the reaction flask
implied 6.48 mmoles dimethylammonium chloride had formed.


aThe hydrogen chloride impurity in the thionyl
chloride was removed by warming the thionyl chloride to -780
and exposing it to a -1960 trap for twenty-five minutes.










The ion should have been almost completely deuterated since

(CH3)2NH2+ is known (38) to exchange rapidly with the solvent

in acidic solution and a large excess of heavy water had

been used. The infrared spectrum in a Nujol mull contained

absorptions in the 1900-2400 cm- region and none greater

than 3000 cm-1, indicating the absence of NH absorption.

The completely deuterated ammonium ion has absorptions at

2214- and 2546 cm1 (30). Therefore, the product should be

primarily the deuterium-containing material.

In the Dri-Lab controlled atmosphere box, lithium

borohydride (approximately 11.5 mmoles) was added to the

flask containing the (CH3)2ND2C1. The flask was attached to

a vacuum system and approximately 25 ml of diethylether

(stored over CaH2) was distilled into the flask. After

forty-five minutes at 0 and fifteen minutes at room tempera-

ture, no evidence for reaction was noted. The reaction

flask was returned to the Dri-Lab and LiBH4 from another

bottle added to the solution. Immediate gas evolution was

noticed. Excess LiBH4 was added and the solution was

magnetically stirred. After one hour when no more gas

evolution was noticed, the reaction mixture was filtered

and the residue washed with approximately 10 ml of ether.

After the ether was removed by vacuum distillation, a liquid

phase containing a white solid remained in the flask.










A small amount of the liquid product was vacuum dis-

tilled from the reaction flask at room temperature into a

-1960 trap. An infrared spectrum of this material showed

absorptions at 3210 cm-1 (NH), 2300-2400 cm-1 (ND,BH), and

at 1750 cm-1 (BD). Since the material distilled so slowly,

it was recrystallized from cold carbon tetrachloride and n-

hexane. The recrystallized product was a solid at room

temperature. It was placed in a vacuum sublimation apparatus,

and the most volatile fraction was removed by pumping on the

sublimator at room temperature and collecting a product in

a -1960 trap. After twenty-five minutes, the cold finger

in the sublimator was cooled to -780 with CO2(s) and the

material was collected for nine hours. The initial material

removed from the sublimator into the -196 trap was a liquid

at room temperature and the compound collected on the cold

finger was a solid at room temperature. The infrared spectra

(Figure 16) of the recrystallized compound and on the

fractions obtained by sublimation were identical and showed

absorptions in the same regions as the material initially

transferred from the reaction flask. The percentages of NH

and BD bond in the compound were calculated from the infrared

spectra using Beer's law. The percentages were calculated

relative to the CH deformation at 1475 cm-1 as the internal

reference. The results were given in Table 11.











49









7 ol








Cdl
o


N
cn





u


0
0
Cd







0



ra
E n




C N
Cd C
cz
0












C0 0
C)
0 S






C-C
cna
C CCC
oZ
4-4~
Na





0 c
0
0 b




0 0 0 0 0
NE


u;ue3~Tsue; ;ua~o aa










TABLE 11

PER CENT BD AND I1N BONDS IN (CH3)2DNBH3 PREPARED
FROM (CH3)2ND2C1 AND LiBH4


Compound: (CH3)2DNBH3 Per Cent Per Cent
BD Nh BD NH

Fraction Absorbency
Sublimed from 0.342 0.201 18.0 20.2
Reaction Flask

Recrystallized
CC14-n-hexane 0.0492 0.0492 14.7 28.0

Recrystallized
First Fraction
from Sublima-
tion 0.0453 0.0414 13.3 23.3

Recrystallized
and Collected
on Cold Finger 0.155 0.167 13.2 27.2










Any dimethylaminoborane impurity should be contained

in the initial fraction sublimed from the reaction flaska

A small impurity of this compound in the spectrum would

cause a low percentage of NE relative to CH in the calcula-

tion, and the absorption at 1750 cm-1 in the spectrum of

dimethylaminoborane would cause a high percentage of BD

relative to CH to be calculated. Another source of error

would be in using the absorptivity ratio of BH:CH in

(CH )2HNBH3 to calculate the per cent BD, the assumption

made here has been shown previously to be a questionable

one. The first fraction sublimed from the recrystallized

compound could also contain an aminoborane impurity. The

percentage of BD and hTH containing compounds show that some

fractionation was accomplished by the method of purification,

but still the recrystallized material before and after

sublimation were essentially the same.

The larger percentage of NH bonds compared to BD

bonds would indicate that hydrogen-deuterium exchange

occurred before the formation of the aminoborane. Thus,

this method of preparation, under the experimental condi-

tions used, did not give a pure product containing deuterium


aDimethylaminoborane has a vapor pressure of 10mm at
230 (39).
See page 17.









only on the nitrogen atom, due to the exchange between ND

and BD bonds prior to the reaction to form the amine

borane.

Hydrolysis of (CH3)2DNBD3 in 0.1 M hydrochloric acid

N-Deuterodimethylamine borane-d3 (0.914 mmoles) was

sublimed into a 50 ml round bottom reaction flask and hydro-

chloric acid (20 ml of 0.1 M) was distilled into the flask.

After nine hours at room temperature, the reaction product

was condensed and the non-condensible gas removed with a

Toepler pump. The amount of non-condensible gas (2.70.

mmoles) corresponded to complete reaction according to the

equation:

(CH )2DNBD + H30 + 2H20 (CH)2NDH+ + B(OH)3

+ 3HD [53

The mass spectrum of the non-condensible gas gave 14 per

cent HD and 86 per cent H2. The 14 per cent HD in the non-

condensible hydrolysis product indicated that the rate of

exchange of BD with the solvent was not so much more rapid

than the rate of solvolysis that all of the deuterium bonded
to boron exchanged before the solvolysis reaction was com-

plete.

R. E. Davis (11) reported only H2 gas produced in the

acid hydrolysis of (CH)3 NBD3 due to the rapid acid catalyzed

exchange of the BD with the solvent. However, the rate of









acid hydrolysis of (CH )2HNBH3 is greater than that of

(CH3) NBH3 (24). Therefore, the BD in (CH 3)NBD3 would

have had more time to exchange before solvolysis than in

(CH )2HNBD .

H. C. Kelly (23) reported that for p-toluidine borane-

d in a 50/50 mixture of dioxane and water with no acid

present, the rate of exchange of BD with solvent was negli-

gible relative to the rate of solvolysis; but that at high

acid concentrations the rate of exchange increased. Kelly

found no primary hydrogen isotope effects in the solvolysis

reaction.

Determination of reaction conditions for the hydrogen
elimination reactions

Dimethylamine borane was heated in sealed glass tubes

for various periods of time and the amount of hydrogen

eliminated measured by transferring the hydrogen into a

calibrated bulb with a Toepler pump. The temperature of

1000 was used for the reactions because it gave a reasonable

rate of hydrogen evolution. In general, the extent of re-

action at low percentages was not very reproducible, since

the small amounts of hydrogen being measured (usually 0.4

mmoles ot 0.03 mmoles in a volume of 105.8 ml) were subject

to experimental error.

The results of the experiments were given in Table

12. These experiments led to the selection of reaction











TABLE 12


REACTION CONDITIONS FOR


H2 ELIMINATION REACTIONS


(CH )pHNBH Time Temperatures Per Cent
mmoles (hour) Reaction


1.54 1 1000 13.5
1.47 1 700 2.32
0.99 10 1000 42.8
2.40 3 1000 15.8
2.20 1 1050 16.4
2.45 0.33 1000 1.02
2.80 0.66 1000 1.54
2.57 0.85 1010 12.8
2.22 0.66 1000 2.1
2.18 0.75 1000 2.5
2.07 + trace 0.75 1000 2.2
(CH )NH
2.45 0.83 1000 2.2


~pp"









times of twenty-four hours in the initial experiment of

heating (CH3)2DNBD3 with (CH3)2HNBH3 and of a time of one

hour in the experiments where just the initial reaction

products were desired in an attempt to ascertain a kinetic

isotope effect.

A trace of free dimethylamine added to one of the

reaction tubes did not significantly affect the extent of

reaction.

Hydrogen elimination on heating dimethylamine boranes

Dimethylamine borane was sublimed from a storage

flask into a reaction tube equipped with a capillary break-

off tip. The amount of compound sublimed into the reaction

tube was determined by weighing the storage bulb before and

after the sublimation. Then a second dimethylamine borane,

containing a different isotopic distribution of hydrogen

atoms on boron and nitrogen was sublimed into the reaction

tube, and the tube sealed off with a torch. In a typical

experiment, approximately one mmole of each compound was

used. The reaction tubes were heated at 100 + 20C for the

desired reaction time. After rapid cooling the non-con-

densible gas was transferred into a bulb by using the

Toepler pump. The gas samples were stored at room tempera-

ture until analysis in the mass spectrometer.

The results of these experiments were given in Table


13.













































,-4
0









Hff







O
SH




0





o





O

01


SON
0 0




*


.













r- ;K





0 OC
S *

















02 0














;CM









rC' cO
*












o









M


A- 0 0


o


'D H
*






L 0r\







A I
*!







00

_- O-






CO






















A L\O 0
I








fI0 0









-O
^-/







P1 L\ 0


ID OJ
(U 0
.















dLr ('
OC






*0 o
*
Lf\O
















*












SOO

0 HO
O C








a
O O
r^







^-^


cO 00

*


-t r
CO CO












I I
I I









COC



-4 r-*


N oP











m 00
N-

p 0 0
O
CO
/-l r- r-{





KN
+




- i
Q 00

OJ


do
a1) -H
0 -P
O
00



















0
* r





















CO




o








































0










0)






















o

5 9


0
O
PC0


S
H-1 ON


p








CM


r1f
C)


u
^^


E- kD







HO
i\ LT











CMC










HCM










H H


C-

00








DO0
* *
00
o > 0"
<^cr


KM

pq



CM









N
C%
O
V
r^










Experiments to eliminate possibility of isotopic inter-
change during the elimination reactions

I. Heating of D2 and H2 in the presence of mercury

vapor.--A bulb was attached to the vacuum system, evacuated,

and submerged in a liquid N2 bath. After condensing mercury

vapor into the bulb for ten hours, approximately equal moles

of hydrogen and deuterium were placed in the bulb. The

bulb was closed and heated at 1000 for twenty-four hours.

The gas was analyzed by the mass spectrometer and shown to

be only hydrogen and deuterium. No exchange of the H2 and

D2 occurred under these conditions. Therefore, it was con-
cluded that the reaction gases did not exchange among

themselves, even in the presence of mercury vapor.

II. Heating D2 with (CH )2HHNBH3 and (CH )NBH3 to

determine if exchange occurred.--The amine borane was placed

in a tube with a capillary break-off tip, condensed in

liquid N2 and evacuated. Deuterium (150 mm) was placed in

the tube and the tube glass sealed. The tube was heated at

1000 for one hour and then a sample of the gas removed from

the tube by breaking the tip and allowing the gas to expand

into a bulb. The gas was analyzed in the mass spectrometer

to determine if any HD had been produced. The gaseous

product did not contain any material with a m/e of 3

according to the mass spectral analysis. The solid materials









were dissolved in spectral grade carbon tetrachloride and

the infrared spectra (Figures 17 and 18) determined. The
-1
spectra showed no absorption whatever at 1750-1800 cm ,

where BD absorbs intensely, but were identical to the

spectra of (CH3)2HNBH3 and (CH3)3NBH .

The experiments were repeated heating the deuterium-

amineborane mixtures for twenty-four hours. The infrared

spectra in both cases contained absorptions in the 1700-

1800 cm-1 range. The spectrum (Figure 19) of (CH3)2HNBH3

and D2 after heating had weak absorptions at 1725 and 1850

cm and an intense absorption at 1785 cm which is where
the BD stretching vibration occurs. The spectrum (Figure

20) of (CH3)3NBE3 and D2 after heating had a medium absorp-
tion at 1740-1750 cm1.
An infrared spectrum (Figure 6) was taken of

(CH3)2HNBH3 after thirty-four hours at 1000 and the spectrum
-1
had a medium absorption at 1750 cm This absorption must

be due to some reaction product and not to a BD vibration

since there was no deuterium in the molecule or in contact

with it. The absorption would cause an error in any esti-

mation of the absorption due to BD in this region. There-

fore, any calculations by Beer's law of the BD percentage,

after heating which considered the absorption in the 1750
-1
cm- region would be in error. The calculation would imply

a larger percentage BD than actually existed.










60



4)




\0 P- 0
0C\) 0


>C




co 7
C) 4-4


rc) E
10 Ic
E ~-.-



>) V) C
cz r4+
C~ cz a

o 5 ac)









cr







CH









_ _ _ _ _ _4-)
_0









cz)
4-I
r44
C)~
























OZ0 \0 +C.
quaoO
c c
k

4-4




C'-C





-~v N





















0
0
CN







0
0








O






0

O FO






o




N
C)






0














CO,
0
0
0
Cu\










0
0
OC








0
0
0
c\


a~~e3~rnsueJ~ ~ua~ Ja6
















O
O
O




1-I


0









O(




-O




0
\o E
0l o
O




0



0c >
i- (C








O













C\
(M


aou emsuej3. ;uqo ja









The infrared spectrum of (CH3)3 NBH3 was unaffected

by heating for twenty-four hours at 1000.

The data showed that exchange between dimethyl or

trimethylamine borane and deuterium did not occur in one

hour at 1000 but that exchange did occur after twenty-four

hours at 1000.

The data are summarized in Table 14.

III. Heating of (CH )2DNBH to determine if exchange

occurred between ND and BH within the molecule.--N-deutero-

dimethylamine borane was heated at 1000 for eleven hours in

a sealed glass tube.' The material was handled in the

vacuum system or in the Dri-Lab controlled atmosphere box.

The solid material was dissolved in spectral grade carbon

tetrachloride and an infrared spectrum (Figure 3) obtained.

This spectrum was compared to a spectrum of an unheated

sample of (CH3)2DNBH3 (Figure 2).

Beer's law was used to calculate the concentration

of compound containing NH, ND, BH, and BD using the CH de-

formation peak as the internal reference (see Table 3).

The spectra (Figures 2 and 3) showed an enrichment in the

percentage of deuterium contained in the unreacted material,

and a decrease in the percentage of hydrogen-containing ma-

terial. The concentration of NH containing compounds de-

creased from 22 per cent to 15 per cent, ND increased from





















cnO
CO 0
M )







04


F-P






o



CQ)

0





E-p



0







0





rd
0
0

















*


NI




















0


0


0
H
c)
v>



rc

ff

(\


CM


+ +


c(I


N


v>


UP
z



0


a)
bO
to
0

pq 0
F9 >S


0
0
0























0
0d
r4






LTt






rlX


r-i

0d
o 0



OH

p *aO4*
pQ

rl







4-






r\.
CM







Cj


0


H


COi
CM -


0


z

0
r<\


rC\
X
0
^/


I









78 per cent to 85 per cent, BH decreased from 97 per cent

to 92 per cent and BD increased from 3 per cent to 8 per

cent when the compound was heated. During heating, di-

methylaminoborane was formed, which had an absorption peak

in the same region (1750 cm-1) as the BD absorption. There-

fore, the absorption peak at 1750 cm- in the spectrum of

(CH3)2DNlBH3, after heating, could not be attributed solely

to BD containing compounds, and the calculation of 8 per

cent BD was an over-estimate. Any calculations or quanti-

tative considerations of this 8 per cent BD would be

questionable.

IV. Heating of (CH3)2DNBH3 and (CH )2HNBD to determine

if amine exchange occurred.--Dimethylamine borane-d3 and N-

deuterodimethylamine borane were sublimed into a tube equip-

ped with a capillary break-off tip, and the glass was sealed.

Duplicate tubes were heated at 1000 for twenty-four hours.

The non-condensible product was removed and analyzed

in the mass spectrometer (see Table 15). The solid products

also were analyzed in the mass spectrometer (Figure 21).

The experiment was dpne in duplicate. In neither

case did the mass spectra show a peak at the mass to charge

ratio of 65. The peak at this m/e would occur only if the

completely deuterated compound were present in the solid

material. If amine exchange occurred, then this peak would















--


o
Co











-o



-D
O^


-0


cn





-!
hO



--0 r
0
C) -






;fo
CH ;C


Q)
C.)






C) C
C:










cC
-0
c- Z
r^

c\


S r-i
b .





C vC

C)







O
4- <







cz

C)





C)







I



--
r0


==r6


3
3





--I

_I
-r


C)





x




C.)
-4

*r-1









be present; otherwise, the highest mass to charge ratio

would be 62, the parent ion peak for (CH3)2HNBD3. There-

fore, amine exchange does not occur under the reaction

conditions.

In the mass spectrum for (CH3)2 HNBD3, the ratio of

the intensities of peak 61:62 was 7.9. The mass spectra,

taken on the solid material after heating, contained a ratio

of the peak intensities of 61:62 of 5.4 and 5.8 for the two

experiments. In the recorded spectra, the peak at m/e of

62 was a shoulder on the peak at m/e of 61. The same base

line was used to get both peak heights. This may not be

the actual height of peak 62, but it may actually be less

intense than measured, which would tend to bring the ratio

of 61:62 more into line with that for the spectrum of

(C 3)2HNBD .











DISCUSSION OF RESULTS


Possibility of hydrogen-deuterium exchange

Every isotope study must be thoroughly checked to
make sure that an exchange process does not vitiate the

conclusions. Exchange between hydrogen and deuterium,

either in the gas phase or when bonded to other atoms with-

in a molecule, could invalidate the experiment. Therefore,

it becomes of primary importance to ascertain if any ex-

change reactions could occur under the experimental

conditions. Even a kinetic isotope effect in the hydrogen

elimination reaction would cause enrichment in the unre-

acted compounds of the less reactive isotope, and thus
affect the measured ratio of H2:HD:D2 from a long term

reaction. Therefore, the critical experiments to determine
the isotope effect were run for only 1 to 2 per cent hydro-
gen elimination to avoid this possibility.

There are five processes by which isotopic inter-
change could occur:

(a) H2+ D2 = 2 HD [6]
(b) (CH3)2HNBH3 + D2 = (CH3)2HNBH2D + HD [7]

(c) (CH3)2DNBH3 = (CH3)2HNBH2D [8]
(d) (CH 3)2HNBH + (CH )2DNBD =- (CH )2DNBH3

+ (CH3)2HNBD3 [9]









(e) Exchange between reactants prior to amineborane
formation. Each of these possibilities was
examined experimentally.

The mass spectrum of a mixture of D2 and H2 with

mercury vapor, heated for twenty-four hours at 1000, did
not contain a peak at m/e of 3. Therefore, exchange re-

action [6], even over mercury metal, did not occur.
Dimethylamine borane and trimethylamine borane were

heated with deuterium gas to determine if exchange occurred.
After one hour at 1000, neither (CH3)2HNBH3 or (CH 3)NBH3
had exchanged with the D2 gas. A mass spectrum of the gas

from the reaction tube showed no HD, and the infrared
spectrum of the solid materials showed no BD absorptions

in the 1750-1800 cm-1 region.
However, after twenty-four hours at 1000, both

(CH3)2HNBH3 and (CH3) NBH3 had exchanged to some extent with
the D2 gas, since their infrared spectra showed BD absorp-

tions in both instances.

The reaction products when dimethylamine borane was

heated were hydrogen gas and dimethylaminoborane, according

to the equation:

(CH3)2HNBH heat ) H2 + (CH3)2NBH2 [10]

At 1000, the dimethylaminoborane will disproportionate (7)

according to the equation:


5(CH3)2NBH2 = [(CH3)2N]2BH + (CH3)2NB2H5


[11l]









Noeth (31) has reported that (CH )4N2'2BH3 will give

(CH )2NBH2 and (CH )2HNBH3 when heated, and at 1000 the
principal product was (CH )NBH2 and H2 with some

[(CH3)2N]2BH and (CH3)2NB2H5 being produced. Therefore,
under the experimental conditions, the reaction products

would not be just hydrogen and the aminoborane, but a more

complicated mixture of compounds. Since aminodiborane is

known (6) to exchange with deuterium gas, the argument could

be made that it was this species or possibly the aminoborane
which was exchanging with the deuterium gas and not the

amine borane. But, the BH bonds in trimethylamine borane

exchanged with deuterium gas under the same conditions.

Therefore, it is not unlikely that exchange also occurred

between dimethylamine borane and D2 gas.

Deuterium gas has been reported to exchange also with

diborane (6,12,57) and with the BH bonds in borazine (10).

From the experimental data, it can be concluded that

exchange between deuterium gas and dimethyl or trimethyl-

amine borane did not occur to a measurable extent in one

hour at 1000, but that it did occur in twenty-four hours

at 1000. This result must be considered in the interpreta-
tion of the experimental results.

The intramolecular exchange (equation [8]) did not

occur when (CH3)2DNBH3 was heated. Since the starting

material contained some NH and BD bonds, the change in









relative concentrations in the infrared spectra, on heating,

had to be calculated using Beer's law. The change in NH

containing compound gave the more accurate resultsa and

indicated an enrichment of the deuterium containing compound

in the unreacted material. The enrichment in ND bonds, in

the material remaining after heating, implied that the NH

bonds were lost much more readily. A relative change in the

moles of NH to the moles of ND can be calculated from the

spectral data. If one considered the total moles of di-

methylamineborane, a, to be the sum of the moles of NH and

ND, then before heating there were 0.22a moles of NH and

0.78a moles of ND. Assuming that the reaction was 30 per

cent completeb according to equation [10], then 0.30a total

moles of both NH and ND reacted. After heating, the moles

of NH were (0.15) (1.00a-0.30a) which was 0.105a and the

moles of ND were (0.85) (1.00a-0.30a) which was 0.60a. This

corresponded to a decrease in moles of NH of 0.115a or 52.3

per cent and in moles of ND of 0.18a or 23.1 per cent.

These calculations indicate the following:

(1) A hydrogen atom was eliminated 2.3 times more
readily than a deuterium atom from the nitrogen.

(2) There was an increase in concentration of the

deuterium-containing compound in the unreacted material.


aSee page 16.

This figure should be reasonably accurate considering
the data in Table 12.










(3) Exchange between the ND and BH was not occurring

to any significant extent during the reaction time. Other-

wise, a greater decrease in ND would be expected and less

of a decrease or even an increase in NH would be expected

since there were 3 BH bonds per ND bond available for ex-

change in the original molecule.

The analogous calculation using the BD or BH absorp-

tions would contain too large an error to be meaningful.

Since ND and BH absorb in the same region, the BH absorption

could not be used for the calculation. The BD absorption

could be used, but to do so the following assumptions must

be made:

(1) The absorptivities ratio of BH:CH is the same as

that for BD:CH.

(2) The absorptivities of BH:CH in (CH3)2HNBH3 is the

same as that for BH:CH in (CH3)2NBH2.

The second assumption was checked by heating

(CH3)2HNBH3 for thirty-four hours at 1000 and then subliming
out the most volatile portion of the reaction mixture. The

infrared spectrum gave a ratio of absorptivity of BH:CH of

1.30 compared to that of 1.90 for (CH3)2HNBH3. The lower

value for the absorptivity ratio would cause an under-

estimation of the amount of BH compound actually present if

the value of 1.90 were used in the calculation. But, the








most significant point in this spectrum was an absorption

at 1750 cm-1. This absorption meant that a reaction product

also absorbed in the same region as the BD vibration. Thus,

an over-estimation would be made of the amount of BD con-

taining material in an infrared spectrum after the compound

was heated; the 8 per cent BD calculated from the spectrum

after heating was probably much greater than the actual

amount of BD.

Therefore, the experiment indicated that any intra-

molecular exchange between hydrogen and deuterium on heating

was insignificant.

To determine whether amine exchange occurred between

the amine borane molecules, N-deuterodimethylamine borane

and dimethylamine borane-d3 were heated at 1000 for twenty-

four hours. A mass spectrum of the solid materials did

not contain a peak at m/e 63 which would be present if amine

exchange had occurred to form N-deuterodimethylamine borane-

d3. The mass spectrum did show an intense peak at m/e 61

which had a shoulder at m/e 62. If (CH3)2DNBD3 had been

present a low intensity peak at m/e 63 and a more intense

peak at m/e 62 would have been expected. The intensity of

the peak at m/e 62 did not increase, therefore no amine ex-

change took place. Moreover, its intensity could be fully

accounted for by the presence of unreacted starting material.

Therefore it was concluded on this basis that amine exchange

did not occur.










However, the possibility of exchange between the

gaseous elimination products and the dimethylamine boranes

could account for the absence of the completely deuterated

amine borane. Even if amine exchange occurred to give

(CH5)2 DBDJ a subsequent reaction with H2 could conceivably

have reduced the parent mass peak to an undetectable level,

Thus, the above conclusion is open to some doubt.

The failure to prepare (CH )2DNBH3 and (CH )2HNBD3

in which there was not also an impurity of BD or ND bonds

implied that an exchange reaction was occurring before

adduct formation. The only other method to produce the

impurity would be an intramolecular exchange in the adduct,

but experimental evidence discounted this possibility even

when the adduct was heated.

The compounds were prepared by condensing diborane

and excess dimethylamine together and warming to -780 to

form the adduct. An exchange reaction must have occurred

at a rate comparable to the rate of adduct formations at

this temperature. Dahl and Schaeffer (10) have reported

that N-deuterodiethylamine exchanged with the BH bonds in

borazine at -300 within three minutes. Since the ND bond

in diethylamine exchanges with the BH bond in borazine, it

would be reasonable to expect that a NH bond in a secondary

amine could exchange with a BH bond in diborane. The

infrared analysis of'the dimethylamine boranes, prepared










from the amine and diborane, seemed to indicate that this

exchange did occur to some extent.

The preparation of (CH3)2DNBH3 from (CH )2ND2C1 and

excess LiBH4 contained even more BD bonds according to the

infrared spectrum than were present in the compound pre-

pared from amine and diborane. The infrared spectrum of

(CH3)2DNBH3 implied that 72.8 per cent ND bonds and 13.2

per cent BD bonds were present in the compound. The calcu-

lation was made from Beer's law using the CH deformation at

1475 cm-1 as the internal standard. The comparison of 13.2
per cent BD bonds to 27.2 per cent Na bonds indicated that

an exchange reaction occurred prior to the reaction to pro-

duce the amine borane. Otherwise, the per cent BD bonds

should be equal to the per cent NH bonds if the exchange

occurred after the amine borane was formed. If the exchange

reaction did occur before amine borane formation, as the

data implied, then the unreacted LiBH4 should have contained

some BD bonds.

N-deuterodimethylamine borane, prepared from the

amine and diborane, contained 3 per cent BD as compared to

13.2 per cent in the compound prepared from (CH3)2ND2C1 and
LiBH4. These data indicated that the exchange reaction be-

tween CH3)2TND2 and BH-4 occurred to a greater extent be-

fore adduct formation than was observed for the amine

diborane reaction.










Heating mixtures of dimethylamine boranes containing various
distributions of hydrogen isotopes for one hour

In order to determine if the kinetic isotope effect

in the elimination reaction could be attributed to the

nitrogen or boron-bonded hydrogen isotope, mixtures of di-

methylamine boranes were heated at 1000 for one hour. A

reaction time of one hour was used to minimize the reaction

extent so that the observed product compositions could be

related unequivocally to a known isotopic distribution in

the reactants. In each mixture, either the boron or nitro-

gen bond was one hydrogen isotope and the other hydrogen

bonds in that molecule and in the other molecules were the

other hydrogen isotope; for example, one mixture was

(CH )HNBD3 and (CH3)DNBD3. The isotopic distribution was

varied to determine if a NH or a ND bond was eliminated

more readily or whether a BH bond reacted more readily than

a BD bond.

The results in Table 15 show the following: (1) If

one molecule contained a NH bond, and the rest of the

hydrogen in the system was the deuterium isotope, then the

principal product was HD, the ratio of H2:HD:D2 being

0.15:1.00:0.39. (2) If one molecule contained ND bonds,
and the rest being the hydrogen isotope, then the principal

product was H2, the ratio of H2:HD:D2 being 4.75:1.00:0.0.

(3) If one molecule contained BH bonds, and the rest being









the deuterium isotope, then the principal product was HD,

the ratio of H2:HD:D2 being 0.52:1.00:0.67. (4) If one

molecule contained BD bonds, and the rest being the hydrogen

isotope, then the principal product was H2, the ratio of

H2:HD:D2 being 1.32:1.00:0.05.

The data implied that a hydrogen atom was eliminated

more readily than a deuterium atom since in each instance

the primary product was either H2 or HD and not D2, even

in the cases where there was only one NH or BH bond in the

system. The data were consistent with a BD bond being
eliminated more readily than a ND bond. In the experiments
which compare these two bonds, the H2:HD ratio was 4.75:1.00
for the ND case and 1.32:1.00 for the BD case. This implied
that HD was eliminated between a BD bond and a NH bond

approximately 3.6 times faster than between a ND bond and
a BH bond. The HD:D2 ratio of 2.56, when (CH3)2DNBD3 and

(CH3)2pHNBD3 were heated, compared to the HD:D2 ratio of

1.39, when (CH )2DNBD3 and (CH3)2DNBH3 were heated, indi-
cated that HD was eliminated between a BH bond and a ND

bond 1.84 times faster than between a NH bond and a BD bond.
The HD:D2 ratio of 2.56, when (CH3)2DNBD3 and

(CH )2EHNBD were heated, and the H2:HD ratio of 4.75, when

(CH3)2DNBH 3 and (CH )2HL\TBH3 were heated, indicated a large
kinetic isotope effect to be occurring when the hydrogen

isotope bonded to the nitrogen atom was varied. Thus, the








ratio of the rate constants for the nitrogen atom eliminating

Hk N
a hydrogen or deuterium atom experimentally determined,

varied between 2.7 and 4.8. Edwards (13) has predicted a

ratio of the rate constants for the bond breaking process in-

volving NH and ND bonds of 8.5. The average experi-
k \
mental isotope effect for the ) of 3.8 was less than the

predicted ratio of 8.5. This implied that there was con-
siderable, but not complete, loss of the NH stretching
vibration in the activated complex.
The HD:D2 ratio of 1.39, when (CH3)2DNBD3 and

(CH3)2DNBH3 were heated, and the H2:HD ratio of 1.32, when

(CH3)2HNBD3 and (CH3)2HINBH3 were heated, indicated a small
kinetic isotope effect to be present when the hydrogen
isotope bonded to boron was varied. Thus, the ratio of the
rate constants for the boron atom eliminating a hydrogen

or deuterium atom BH, experimentally determined, was
(BD /
1.3 to 1.4. Hawthorne and Lewis (17) calculated the ratio

of B )to be 4.2 for the expected effect of isotopic

substitution on boron from the loss of the BH stretching
vibration at 2300 cm-1. The observed isotope effect was
much less than that predicted, implying that there was only
a small loss of the BH stretching vibration in the activated
complex.









The maximum isotope effect would be obtained when

the bond to hydrogen or to deuterium was essentially com-

pletely cleaved in the activated complex, and the isotope

effect would decrease with increasing bonding in the

activated complex (44). Therefore, the isotope effects

observed should allow some predictions about a possible

configuration of the activated complex in the hydrogen

elimination reaction. The larger effect when ND was substi-

tuted for NH than when BD was substituted for BH predicts

that the NH(ND) bond was more affected in the activated

complex than was the BH(BD) bond. Hawthorne (17) has re-

ported a similar situation in the hydrolysis of pyridine

diphenylborane with water or deuterium oxide in acetonitrile

solution. The reaction was found to be first order in both

pyridine diphenylborane and water, and a primary kinetic

isotope effect was determined. The ratio of the rate

constants( -H of 6.90, when deuterium oxide was substi-
OD
tuted for water, was nearly as large as that predicted for

a complete loss of the OH stretching vibration in the

activated complex of 9.9. The ratio of the rate constants

H- of 1.52 observed on isotopic substitution on boron was
kBD B

much less than the predicted ~- )of 4.2 for complete loss
of the BH stretching vibration in the activated complex.
of the BH stretching vibration in the activated complex.









Hawthorne (17) proposed the transition state (I)

for the BH bond hydrolysis and suggested that this type of

non-linear transition state may be general for hydride

transfer.



O H
H- o '- -B- Py

C6H5 06H I


The similarity between the isotope effect observed

by HawthorneJand that observed here in the hydrogen elimina-

tion reaction of dimethylamine borane, suggests that the

reactions might be occurring through a similar activated

complex. Namely, that the activated complex was not a

linear configuration involving NH and BH bonds but that the

NH bond was stretched more along the bond axis than was the

BH bond, in a manner analogous to that for the OH and BH

bonds in I.

These data must be interpreted in the light of

possible exchange reactions occurring faster than the

elimination reaction. Since hydrogen and deuterium gas did

not exchange when heated for 24 hours at 1000, this possi-

bility could be discounted. If amine exchange between

boranes occurred, there would be no net change in the

systems, so this would not cause any difficulties. Deuterium









gas when heated at 1000 for one hour with dimethylamine

borane did not affect the infrared spectrum of the solid

material. Therefore, it may be concluded that no isotope

exchange reactions occurred during one hour at 1000, and
that the data were not subject to any uncertainties due to
exchange.
The starting materials were not as pure in their

isotopic distribution as would be necessary to determine a
precise kinetic isotope effect in the hydrogen elimination
reaction. The infrared spectra showed the following:

(1) (CH3)2DNBH3 contained 3 per cent BD bonds
relative to the per cent CH bonds.

(2) (CH3)2HNBD3 contained 7 per cent ND bonds
relative to the per cent CH bonds.

(3) (CH3)2DNBH3 contained 13 per cent BD bonds
relative to the per cent CH bonds.a (Prepared from

(CH3)2ND2+ and BH4-).
Compounds (1) and (2) were prepared by condensing
dimethylamine and diborane together and forming the adduct
at -78. Compound (3) was prepared from (CH3)2ND2+ and BH,-.
In each instance some exchange must have occurred before
the reaction producing the amine borane adduct.


aThis compound was not used in any of the hydrogen
elimination experiments.









The presence of isotopic impurities introduced

during synthesis would account for the D2 produced when

(CH3)2HNBD3 and (CH3)2HIHBH3 were heated and for the H2
produced on heating (CH3)2HNBD3 with (CH3)2DNBD3 and

(CH3)2DNBH3 with (CH )2DNBD3.
The data from the one-hour heating experiments did
not unequivocally distinguish between a unimolecular and

a bimolecular reaction mechanism. But, the analogy to

Hawthorne's work with the hydrolysis of pyridine diphenyl-

borane (17) and the large percentages of HD obtained in

each case does favor a bimolecular reaction. However, the

molecularity of the reaction was resolved by the experiment
in which (CH3)2DNBD3 and (CH3)2HNBH3 were heated.

Heating mixtures of dimethylamine boranes containing various
distributions of hydrogen isotopes for twenty-four hours

Dimethylamine borane and N-deuterodimethylamine

borane-d3, in 1:1 molar ratio, were heated at 1000 for

twenty-four hours and the non-condensible products were
analyzed in a mass spectrometer. If the hydrogen elimina-
tion reaction were unimolecular, then the gaseous product
should contain H2, D2 and perhaps some HD due to incomplete
deuteration. If the reaction were bimolecular, the gaseous
products should be H2, HD and D2 in the ratio of 1:2:1,
respectively, neglecting isotope effects.










The experimentally determined ratio of H2:HD:D2 was

3.8:4.3:1.0. The data implied that the reaction was bi-

molecular and that there was an isotope effect favoring H2

eliminations in the reaction.

Dimethylamine borane-d3 and N-deuterodimethylamine

borane, in a 1:1 molar ratio, were heated for twenty-four

hours at 1000 and the non-condensible reaction products were

analyzed in the mass spectrometer.a If the reaction were
unimolecular the gaseous product should be HD with the D2

and H2 due to incomplete deuteration. If the reaction were

bimolecular, the gaseous products should be H2, HD and D2

in ratio of 1:2:1 neglecting isotope effects.

The gaseous products had a H2:HD:D2 ratio of 3.1:3.6:

1.0. Since the gaseous products of the reaction contained

more H2 and D2 than could be accounted for by incomplete

deuteration, the reaction must have been bimolecular. The

data were in close agreement with the previous results and

indicated that it made no difference whether the deuterium

atoms were all in one molecule or partly on the nitrogen

in one molecule and partly on the boron in the other molecule.

The results in both cases gave the same percentages of H2,

HD and D2 in the gaseous product.

The total deuterium percentage in each of the re-

action systems, (CH3)2HNBH3 (CH )2DNBD3 and (CH,)pDNBH3 -

aFor mass spectral analysis of the solid reaction
products see Figure 21.









(CH3)2HNBD3, was almost the same. From the infrared spectra,
the system (CH )2DNBH3 (CH3)2HNBD3 contained 47.4 per

cent deuterium bonds and from the pyrolysis products of the

diborane-d6, the system (CH3)2HNBH3 (CH )2DNBD3 contained

a minimum of 45.4 per cent deuterium bonds. Therefore, it

was not surprising that under the same conditions, if the

reaction were bimolecular or if an equilibrium reaction

were established, for these two systems to give the same

ratio of H2:HD:D2 in the gaseous elimination products.

However, the interpretation of these results was made

questionable by the possibility that exchange occurred during

the reaction between deuterium and the reactants or other

reaction products. Some exchange did occur between di-

methylamine borane and deuterium gas on heating for twenty-

four hours at 1000, the same conditions as in these experi-

ments. In order for the two experiments to have had the

same ratio of H2:HD:D2 in the gas phase with exchange

occurring between the gases and other compounds in the

system, the reaction mixtures in each experiment must have
reached the same equilibrium.

Equilibrium would have been established according to
the following equations for the boron containing reaction

products (7).

[(CH )2NBH2]2 2(CH)2NBH2 [12]









3(CH3)2NBH2 = [(CH )2N32 BH + (CH3)2NIB2H5 [133

Burg (6) reported that the aminodiborane in the

presence of excess D2 after seventy-four hours at 1000 was
found to be 65 per cent deuterated. In experiments using

D2 gas to deuterate a compound, a large excess of D2 gas

and long reaction times were used to be sure an equilibrium

was established. In the hydrogen elimination reaction of

(CH )2DNBD3 and (CH3)2HNBH3, the per cent reaction was

31.3 per cent. Therefore, of the original 1.59 mmoles of
amine borane, there remained 1.09 mmoles unreacted amine

borane, with 0.50 mmoles of gas and 0.50 mmoles of amino-

borane produced. The 0.50 mmoles of gas contained 0.06
mmoles D2, 0.22 mmoles HD and 0.20 mmoles H2. If an exchange

reaction were occurring between the gaseous hydrogen iso-

topes and the boron-nitrogen compounds, then the ratio of

H2:D2 in the gaseous product should be the same as the ratio

of H2:D2 over any other catalytic system. Essentially the
boron-nitrogen compounds would be serving as a catalyst for
the hydrogen-deuterium exchange reaction, H2 + D2 = 2HD.
Considering the total moles of gas to be A, then the amount
of hydrogen gas available in the (CH3)2HNBH3 (CH3)2DNBD3
system would be 0.26A moles and of deuterium would be
0.24A moles. Using the equilibrium constant,3.48 at 4000K,
calculated by Urey (42) for the deuterium exchange, the

moles of HD were calculated to be 0.20A; and thus, the









concentration at equilibrium of H2 would be 0.08A and of

D2 was 0.06A. The ratio of H2:D2 at equilibrium would

therefore be 1.33. The observed ratio of H2:D2 was 3.53,

implying that the gaseous products were not at equilibrium
for the exchange reaction.

If an exchange reaction were occurring between the
gaseous hydrogen isotopes and the boron-nitrogen compounds,

then a consideration of the difference in zero-point energy

between a BH and a BD bond compared to the difference in H2
and D2 bonds should give an indication of the thermo-
dynamically favored reaction.

For two isotopes in an otherwise identical bond, the
difference between the two zero-points of energy is given
by E = h(--Vi), where V and )/ refer to bonds containing
the lighter and heavier isotopes X and X', respectively.
The difference inVUBH and 1BD calculated from only the
stretching vibrations at 2550 cm-1 (BH) and 1875 cm-1 (BD)

was 1.36 kcal/mole, and the difference in HH and~)DD
calculated from the frequencies for the fundamental vibra-
tion transitions (1) of 4159 cm-1 (HH) and 2990 cm-1 (DD)
was 3.34 kcal/mole. Thus, a comparison of the differences
in zero-point energies indicated that the reaction to pro-
duce D2 would be thermodynamically the most favored and the

gas phase should be enriched in D2. This was contrary to
the observed kinetic isotope effect and the observed ratio










of 41.3 per cent H2:47.7 per cent HD:11.0 per cent D2 in

the gaseous product.

Therefore, it does not appear likely that the ex-

change reaction between H2, HD or D2 and the amine boranes

or aminoboranes had reached equilibrium in twenty-four

hours at 1000. The exchange reaction, thus, has only a

secondary effect on the eliminated gases and the reaction

must be bimolecular as was inferred by the experiments on

heating the dimethylamine boranes for one hour.












CONCLUSION


The reaction of dimethylamine borane to eliminate

hydrogen was bimolecular and a kinetic isotope effect

occurred during the elimination of hydrogen. The data

showed that a hydrogen atom was eliminated faster than a

deuterium atom and that a BD bond reacted more readily than

a ND bond.

The ease and speed of deuterium and hydrogen atoms

exchange limits many of the possible experiments which

might be used to elucidate the behavior of the dimethylamine

borane system on heating, and in studying the kinetic iso-

tope effect in the reactions of the amine boranes. More

work needs to be done to establish the conditions and

possibly the rates of the deuterium-hydrogen exchange re-

action in the BN containing compounds. A case in point

being R. E. Davis' report (11) that the rate of deuterium

exchange in acidic D20 with trimethylamine borane was much

more rapid than the hydrolysis reaction. In this work,

evidence was presented that for N-deuterodimethylamine

borane-d3, the rate of exchange with acidic H20 was approxi-

mately equal or only slightly faster than the rate of

hydrolysis. This inferred a possible order of magnitude for

the rate of the exchange reaction.




Full Text

PAGE 1

REACTIONS OF AMINE BORANES AND RELATED COMPOUNDS: (I) MECHANISM OF DEHYDROGENATION OF DIMETHYLAMINE BORANE (II) SYNTHESIS OF TRIMETHYLAMINE CHLOROBORANES By JAMES WILLIAM WIGGINS 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 December, 1966

PAGE 2

f_ . . -, I i -, i : . :i

PAGE 3

ACKHOWLEDGMENTS I acknowledge witli sincere gratitude tlie assistance given by the Ciiairman of my supervisory committee, Dr. G. E. Rysclikewitscli, during the preparation of this work. His enthusiasm and patient instruction during the course of the research made the work a pleasure. Q?he decisive influence and interest in my professional career of Dr. Ryschkewitsch has been deeply appreciated, I sincerely thank the members of my supervisory committee and the many other faculty members who have expressed an interest in my growth as a chemist. I express my thanks for financial support to the National Science Foundation Grant G19738 and the Chemistry Department for support on the Science Development Grant. I thank I-Ir. D. D. Davis and Dr. Alan Hagopian for obtaining the mass spectra. I thank Dr. Wallace S. Brey, Jr. , and Dr. K. N. Scott for obtaining the B nuclear magnetic resonance spectra. A special thanks is extended to Mr. R, G. Logsdon for building and repairing the glass vacuum system used in the work. I thank Mrs. Thyra Johnston for typing the final copy of the dissertation. ii

PAGE 4

The many enlightening discussions and endeavors with Dr. Gerhard M. Schmid have made my work toward the Ph.D. degree a real joy. XXX

PAGE 5

TABLE OP CONTENTS Page ACKNOWLEDGMENTS , ' . . . ii LIST OE TABLES viii LIST OE FIGURES . X PAET I. MECHANISM OE DEHYDROGENATION OP DIMETHYLAMINE BORAJIJE INTRODUCTION 1 EXPERIMENTAL 5 Nomenclature 3 Origin of reagents 5 Purification of reagents 3 Instruments. ^ General method for the analysis of the dimethyl amine boranes 5 Pyrolysis of '^oPf,' • • • ^ Infrared spectral analysis . 8 Mass spectral analysis 19 Preparation of (CH^);2N3D^ from CCH^)^NBH^ in an acidic DoO solution 26 Preparation of B2Dg from (CH^)^NBD^ and BP^Cg) . . 29 Variation in the per cent reaction of (CH,):zNBH, and BP^(g) to yield 'S>^^ 52 Preparation of B^Dg from NaBD^ and BP^(g) in diglyme .?.....« 36 Preparation of BoHrfrom NaBHy, and BP^Cg) in diglyme . . ., 7 . . . ? 38 General procedure for preparation of (CH:z)2ND. . . 39 General method for the preparation of (CH:,)2HNBH, from B2Hg and CCH^)2NE A-2 Preparation of (CH2)oDNBH^ from (.OYi^^^l^^Gl and LiBH^ ^,, ,<.,,.<,. ^6 Hydrolysis of (CH;,)5DNBD, in 0.1 M hydrochloric acid ? T . .^. . . . 7 52 iv

PAGE 6

Page Determination of reaction conditions for tlie , . hydrogen elimination reactions • . 53 Hydrogen elimination on heating dimethyl amine horanes 55 Experiments to eliminate possihility of isotopic interchange during the elimination reactions. . 58 I, Heating of Do and Ho in the presence of mercury vapor 58 II, Heating D^ with (CH^)2HNBH^ and (GE-;r) y•BE-;r to dctermino if exchange occurred , , 58 III. Heating of (GE^) ^^^'^.^^ to determine if exchange occurred betv;een ND and BH within the molecule 65 IV. Heating of CCH^)2DNBH^ and (CH^^gHKBD^ to determine if amine exchange occurred. . 65 DISCUSSION OF RESULTS . 68 Possibility of hydrogen-deuterium exchange .... 68 Heating mixtures of dimethylamine boranes containing various distributions of hydrogen isotopes for one hour 76 Heating mixtures of dimethylamine boranes containing various distributions of hydrogen isotopes for twenty-four hours. 82 CONCLUSION 88 SUMMARY 90 PART II. SYNTHESIS OF TRIMETHYLAMINE CHLOROBORANES INTRODUCTION 92 EXPERIMENTAL 95 Nomenclature . 95 Reagents and purification 95 Instruments 96 Extraction of BN compounds from the reaction mixture 97 Infrared spectral analysis 99 Nuclear magnetic resonance spectra . 101 V

PAGE 7

Page Reaction of (CH^),NBH^ and HgCl2 109 Reaction of (CH;z),HBH;, and HgClp on a large scale, 110 Reaction of (CK^)^l^iBH2Cl and HgClg 11^ Attempted reaction of (CH;z)^KBHCl2 with HgGl2. . . 115 Relative rates of reaction of (CH,);,NBH, with. HgCl2 and HCl in ether at 0° 116 Reaction of (CH^)2NBH^ and HgCl2 in the autoclave. 118 Reaction of (CH,),NBH2 and HgCl2 in presence of acetic acid in benzene, ,,,,,,,,,,,, 119 Reaction of ( CH;, ) :zNBH;, and HgCl2 in water and in potassium chloride solutions 120 Reaction of (CH,);,NBH;z and HgCl2in water — the change in pH v;ith time 121 Reaction of (GH,)^NBH2C1 and (CH,)^NBHCl2 with HgCl2 in water — the change in pH with time. , . 125 Reaction of (CH^)^NBH^ and excess HCl(g) 126 Reaction of (CH^);,N3H^ and KCl(g) in benzene . . . 128 Reaction of (CH^;)^!^!!^; and concentrated HCl(aq) in water ...??.? 129 Reaction of {GEy.)^l^'d^ and concentrated HCl(aq) in benzene ..;..? 150 Reaction of (CH:,):,NBH-, and concentrated HCl(aq) in carbon tetrachloride 151 Reaction of (CH^);,K3H;,, (CH^),1CBH2C1 and (CH^)^NBHCl2 v;ith (CH^)^mCl 151 Reaction of (CH^)5JIBH2C1 with SbClc 155 Reaction of (CE^)^KBE^, (CH^)^NBH2C1, (CHj)^lTBHCl2 and (CH^)^NBCl^ with SbCl^ .... 157 Reaction of (CH,)^NBH^ with SOCI2 158 Reaction of (CHj)^NBH^ and SO2CI2 159 Reaction of (CH;,)^N3H;, and ZnCl^ in glacial acetic acid. ...??.?... .7 140 Reaction of (CH^)^NBK^ and (CH,)UTOC1^ in an autoclave and the stability of the monoand dichloroborane adducts under these conditions . 1^5 vi

PAGE 8

Page DISCUSSION OF RESULTS AND CONCLUSIONS 146 Reactions of (CH;,)^i\BH^ and HgCl2 1^5 Reaction of (CH^)^NBH^ and (CH^)^NHCl 158 Reaction of (CH^):^NBH^ and SbCl^ 161 Reaction of (CH^)^I^Hj and SbCl^ 162 Reaction of (CH^)^NBH^ and SO2CI2 164 Reaction of (CH,)^NBH, and SOCI2 165 Conclusions from the reactions 166 Thermal stability of the adducts . . , 169 sumARY 173 BIBLIOGRAPHI 175 BIOGRAPHICAL SKETCH 178 Vll

PAGE 9

LIST OF TABLES Table Page 1, Analysis of Dimethyl amine Boranes 7 2. Pyrolysis of 62!)^ 8 5« Calculation of Per Cent Deuterium in the Deuterated Dimethyl amine Boranes from Infrared Spectra Using the CH Deformation Peak as the Internal Reference, ....*.««•.• 1^ ^. Sensitivity Coefficient of Mass Spectrometer for H2, HD and D2 21 5. Tendency for Parent Ion to Lose a Hydrogen or Deuterium Atom 25 6. Reaction of {CE^^^KBB^ and BP,(g) 50 7. Reaction of (CH^)UT3H, and 3?,(g) 33 8. Preparations of (CK^)^KD 41 9. Preparation of (CH;,)2KNBE;, Containing Various Distributions of Hydrogen Isotopes 44 10. Results of Heating (CH^)2HNBH^ Containing Various Distributions of Hydrogen Isotopes, . 45 11. Per Cent BD and Wd Bonds in (CH^)2DNBH, Prepared from (CH;z)2ND2Cl and LiBH^ 50 12. Reaction Conditions for H2 Elimination Reactions 54 13. Results of Hydrogen Elimination by , . Dimethylamine Boranes 56 14. Reaction of (CH^)oHI^H:, and (CH,):,NBH^ with D2 ^, . , r . . . .^,^, .^, . . . 64 vxxx

PAGE 10

Table Page 15. Infrared Spectra of (CH2),NBH, and the Trimethyl amine Cliloroboranes 100 16. B Nuclear Magnetic Resonance Spectral Results 106 17. Reactions of (CH^)^EBH^ and HgCl2 Ill 18. Reaction of (CH,)^NBH^ and HgCl2 in Water and KCl Solutions 122 19. Reactions of (CH^);,i\iBH;, and Trime thy 1 amine Cliloroboranes v/ith (CH^)^KHCl 133 20. Reaction of (GH^)^EBH, and Trime tbyl amine Cbloroboranes on Heating 1^^ IX

PAGE 11

LIST OF FIGURES Figure Page 1, Infrared spectrum of (CH^)2HEBH^ ' 9 2, Infrared spectrum of (GE^) 2^'^'S>'B.y 10 5. Infrared spectrum of ( CH,, ) oDNBH^ SLfter heating ^. . . ? 11 ^i Infrared spectrum of (CH^)2HNBD^ 12 5. Infrared spectrum of (CH;,)pDNBD;, 15 6. Infrared spectrum of (GE:r) ^EKBEy, sifter heating ^. . . < 18 7. Mass spectrometer sensitivity coefficient of Ho, HD and Do as a function of the total pressure 20 8. Mass spectrum of (CH^)2Hi\^BH^ at 70 ev. ... 23 9. Mass spectrum of (OE^) ^^'SBE^r at 70 ev. ... 23 10. Mass spectrum of (CH^)2H1TBD^ at 70 ev. ... 2^ 11. Mass spectr\im of (CH^)2D^^BD^ at 70 ev. . . . 2^ 12. Infrared spectrum of (CH^)^KBD;, 28 13. Infrared spectrum of 'S>2^f^ 51 14. Change in total pressure vri.th time for the reaction of excess (.CE:.) ^WBE-^ and BF;,(g) . . 3^ 15. Change in total pressure with time for the reaction of (CH^^^NBH^ and excess BF^Cg) . . 5^ 15, Infrared spectrum of (CH^)2DNBH, prepared from (CHj) 21^1)2^1 and LiBH^ ^9 17. Infrared spectrum of (CH^)-,NBH;2 after heating with ©2 for one hour 60 X

PAGE 12

Figxire Page 18. Infrared spectrum of (CH^)2HNBH2 after heating with D^ for one hour 50 19. Infrared spectrum of (CK^)2HNBH, after heating with Do for twenty-four hours. ... 51 20. Infrared spectrum of (CH^)^imH^ after heating with Dp for twenty-four hours. ... 52 21. Mass spectrum at 70 ev after heating (CH^)2DNBHj and (GE^)^^!^!)^ for twenty-four hours • 56 22. Infrared spectrum of (CH^)^NBH2C1 102 23. Infrared spectrum of (CH^)^lfflHGl2 105 2^. Infrared spectrum in 300-600 cm" region of (CH,)^N]BH^ and the trimethyl amine chloroboranes lOA25. B Nuclear magnetic resonance spectrum of (CH,),KBH2C1 107 25. B Nuclear magnetic resonance spectrum of (CH^),l\[BHCl2 108 27* Change in pH with time dxiring the reaction of (CH^)^NBH^ and HgGl2 in water 12428. Change in pH with time during the reaction of (CHj)j2TBE2Cl and (GH^)^N3HCl2 127 29. Infrared spectrum of the reaction product of (CH2);,NBH;,''and ZnCl2 in glacial acetic acid. 1^2 30. Comparison of product on heating (CH^),l\BHCl2 with and without (CH^),1T present. ......... 1^5 XI

PAGE 13

PART I. MECHAI^ISM OP DEHYDROGENATION OF DIKSTHILAMIES BORANE INTRODUCTION Vlien dimethyl amine "borane is heated, hydrogen is eliminated and dimethyl aminob or ane is formed. The mechanism of this reaction should be the same as for the first step in the production of borazenes, for example, N-trimethylborazene, by heating monome thy 1 amine borane. Thus, the reaction mechanism, or molecularity, would be worthy of investigation. The reaction of dimethylamine borane to yield hydrogen and dimethyl aminob or ane does not lend itself readily to common methods of kinetic determination such as measuring the increase in total pressure or the concentration of any single species. This is due, in the first case, to such reactions as the dimerization of dimethylaminoborane or the disproportionation of the dimethylaminoborane (7) which occur at significant rates in the temperature range at which hydrogen elimination can be conveniently measured. In the second case, the separation of unreacted dimethylamine borane, or the separation of dimethylaminoborane from the reaction mixture, would be difficult because the reaction would be occurring while the separation was being carried out.

PAGE 14

Since conventional kinetic studies were impractical for the most part, the following method was used to yield data which could be used to determine the reaction molecularity, N-deuterodimethylamine borane-d, and dimethylamine borane in a 1:1 molar ratio v;ere heated and the non-condensible products analyzed in a mass spectrometer. The ratio of H2:HD:D2 found was compared to that expected for either a unimolecular or bimolecular reaction. The results indicated a bimolecular reaction and a kinetic isotope effect. The isotope effect was investigated using various isotopic distributions of hydrogen in the dimethylamine borane and analyzing in the mass spectrometer the gaseous products eliminated on heating. The established isotope effect was that hydrogen atoms were eliminated more readily than deuterium atoms. The possibility of hydrogen-deuterium exchange reactions occurring during the elimination reaction was investigated thoroughly.

PAGE 15

EXPERIMENTAL Nomenclature The compounds formed by the reaction of an amine and diborane v/ere named as amine adducts of borane. The following is the list of amine boranes in Part (I): dimethylamine borane, (CH^)2KN3H^ ; dimethylamine borane-d^, (CH;,)2HlT3D;j ; N-deuterodimethylamine borane, (GE-,) ^^^^^t. » N-deuterodimethyl amine borane-d,, CCE-z)2^'^'^-z • Orip;in of reap;ents Tracerlab, 99.7% D2O Gallery Chemical Co. Matheson Co., Inc. Matheson, Coleman and Bell Div. Matheson Co., Inc. Foote Mineral Co. Chemical Procurement Laboratories Alfa-Inorganics Metal Hydrides, Inc. Purification of reap;ents DpO used v:ithout further purification — being handled in a Np atmosphere. D20:

PAGE 16

(CH5)3NBH, SO2CI2 CCH,)2NH BFC^HgLi iGE^)2HimE. NaBD; LiBH, sublimed once, then resublimed into the reaction flask (or tube), bp 68-70°C, V7as used without further purification — transferred in a ITp atmosphere, stored over Na for over 24hours in freezer compartment of refrigerator, then distilled into reaction flask (or tube), distilled from a -78° trap through a -119° trap into a -195° trap, then distilled into reaction tube, used without further purification — transferred with a syringe under a flowing stream of N2. sublimed once, then resublimed into the reaction flask (or tube). sublimed after preparation and then resublimed into reaction tubes. used without further purification. used v/ithout further purification. Instruments The vacuum system used in the experimental work was similar to the vacuum line described in Synthetic Inorganic Chemistry by W. L. Jolly (21), Apiezon N grease was used on all ground joints in the system. A Bendix Time-of -Flight mass spectrometer was used to obtain the mass spectra. A Beckman IR-10 or a Perkin-Elmer 21 spectrophotometer was used to obtain the infrared spectra in either the gas phase or in a carbon tetrachloride solution.

PAGE 17

General method for the analysis of the dimethyl amineborajies Dimethyl amineh or ane was suhlimed from a weighed storage flask into a 50 ml round bottom flask. After the sublimation, the storage flask was weighed, the difference in weight being the amount of sample to be analyzed. Analysis was based on the equation (CH^)2HiraH^ + 2K2O + H^0+ -* (CE^)^^'^ + B(OH)^ + 3H2 Cl] The compound was first hydrolyzed by condensing 20 ml of 0.1 N HCl (Acculute) into the flask containing the dimethylamine borane. Hydrolysis was allowed to continue overnight at room temperature, The contents of the reaction flask were condensed in a liquid Np bath and the non-condensible gas was transferred into a calibrated bulb by a Toepler p\imp. The amount of hydrolyzable hydrogen was thus obtained. The acid solution from the hydrolysis of dimethylamine borane was transferred into a A-00 ml, boron-free glass beaker. The boron v;as determined as boric acid using the mannitol titration method. The amount of nitrogen, as dimethyl ammonium ion, could be determined since a knovm amount of strong acid had been used to hydrolyze the sample. The difference in the equivalents of strong acid added initially and the equiva-

PAGE 18

lents of "base necessary to neutralize the strong acid remaining after the hydrolysis reaction was the amount of strong acid neutralized by the dimethyl amine. Using this method, the following data in Table 1 were determined. Pyrolysis of '^'^^ Diborane-dg \jb.s pyrolyzed by passing the gas through a 9 nm Vycor tube, 37.5 cm long, surrounded by a 0.75 inch stainless steel pipe and heated by two Meeker burners. The gas was allov;ed to pass slov;ly into the tube. Attempts to make more than one pass of material through the hot tube did not increase the amount of non-condensible gas ; evidently no condensible material passed through the hot tube. The non-condensible products were Toepler pumped into a flask and analyzed in the mass spectrometer. Results of these experiments were as follows, in Table 2.

PAGE 19

I U5 C\J « OJ o •H oo oo HH O O O O O o o o CO pq H Ipq d W OO OO • • 1-1 iH [>-cO lAO O o O O o o op CJ^CTN OMJ^ O^CT^ lA O O « * LA LA EH EH H O CO H CO •P o 0) pq I o o o LAO• « C\J H C\JC\J rH LA 'd -P erf rH o rt H ;3 rf o rA o LAtA CMOJ CMOJ CUCU 00 o fHC\J oo OJ AJ COCU LA CO ^DLA rAco * • CUrA ci-O OOO •xi O •P rH o rt O O «H -d rA Pi P C\] O iH rH •d 0) is H pi-d ri o 0-o « « ^A*JD CM CM ^ rA rACM COO rHCM -d O -P 05 H O 'd P! o fA CM hA o rH m 05 • iH N H O fH W •H CO ^, rH O -d -d

PAGE 20

TABLE 2 PYROLYSIS OP B^Dg 8 Prepared by mmoles Per Cent Reaction D. Per Cent HD H, (CH3)^NBDj + BPjCs) NaBD^ + BP^(g) 0.098 105.8 81.5 18.5 0.12 88.9 91.0 8.^ 0.5 Infrared spectral analysis The spectra of the dimethylamine "boranes (Figures 1, 2, 3» ^, 5) were taken on a Beckman IR-10 spectrophotometer, using matched cells 0.2mm thick and the slow scan speed. The compounds were dissolved in Pischer spectroanalyzed grade carbon tetrachloride and the same solvent was used as a blank in the reference beam. The absorbency, A, expressed ^^ ^'^^ 1°S lllllliZ r^lS^lon «^^ calculated from the infrared spectra in which transmittancy was plotted as a function of the v/ave number. The absorbency of the CH deformation peak at 1475 cm' was used as the internal reference and all calculations were made with respect to this peak. The results are given in Table 5. -1

PAGE 21

o o o o o o o o CO O O O I z m z: a: CM C^ o o o a, (0 n « c M I I 00 o o o o o SOUB-^-^-piUSUBJ^ -^USO JOJ

PAGE 22

10 o o o

PAGE 23

11 o o cvj

PAGE 24

12 o o o o o o I e o o o 00 ^ 3 C > O o o CM Q O o (-1 O o CM O o o o CO o o o souBr^.q.xwsuBj':^ ^uao aej

PAGE 25

13 o o cvj iH o

PAGE 26

1^ EH U P H cj -P -p fl O O EH O 1 H 0) WW o ri o o -p o o u > CD -H ^ +3 5! o H O •H -P PJ rH O CO H o 1-^ fH •H o ft a o o H 00 ON O OON H OJ C0v£) rcs lA lA • • LPs lAH 00 H oocr> OJ

PAGE 27

15 u c6 -P -P ti O
PAGE 28

16 The ratio of the absorptivities of the BH:CH and the NH:CH in the spectmim of the completely undeuterated sample were foiind to be 1.90 eind 0,992, respectively. The per cent M and ND were calculated as follows: ' hm ^ ^NH I — '^ a — S — ^CH ^CH ^CH Aj^/A«jj calculated from spectra ajjrr/aQjT absorptivity ratio from spectrum of undeuterated compound C^^/Cqtt ratio of concentrations Assume in pairtially deuterated compounds that This assumption allowed C^^ to be determined without the actual absorbency being known. This was necessary since the KD and BH stretching frequencies both occur between 23OO-25OO cm" . It was not possible to separate the absorbency due to each vibration. Since (CH;,)2DNBD, was known not to be completely deuterated from mass spectral data of the BoDg pyrolysis product, the absorptivity ratio of BD:CH could not be obtained for its infrared spectrum. Therefore, the ratio of the absorptivities for BD:CH was assumed to be the same as that for BH:CH,

PAGE 29

17 The absorptivity ratio of BH:CH for tlie volatile reaction product after heating (CH;z)2HNBH;, for thirty-four hours at 100* v;as 1.30 (?igure 6). Thus, one of the BH containing products v;hich v;as formed when dimethyl amine borane was heated, did not absorb as strongly in the BH stretching vibration region as did the original starting material. This could cause a lower estimation of the amount of BH-containing material after heating than would actually be present and therefore introduce an error in any calculations made using the BH stretching absorption. The absorption in the 1700-1800 cm" region which was attributed to the BD stretching vibration in the molecule CCH^)2HNBX^ [X=H and/or D] did not occur at exactly the same wave number in each spectra. The absorption varied from 1735 to 1785 cm" . Qualitatively, this variation appeared to be concentration dependent. The greater the * concentration of BD bonds in the molecule the larger the wave number at which the absorption occiu?red. The absorption for the BD stretching vibration in (0Ey,')2^^^'^7^ occurred at 1785 cm" and for (CH,)2DNBH, containing 3 per cent BD bonds, at 1735 cm" . An explanation for this variation could be that a shift in the stretching vibration occurred in the -BH, group as the hydrogen atoms in the -BH^, group were replaced by deuterium atoms. The largest shift was implied by the spectra at low Do percentages when mostly BHpD groups should have been present.

PAGE 30

18 o o ^

PAGE 31

19 Mass spectral analysis A Bendix Time-of-Flight Mass Spectrometer was used for tlie analysis. The sensitivity coefficient of tlie spectrometer to Hp and Dp v;as determined "before each, set of analyses, using commercial samples of hydrogen and deuterium gas. The sensitivity coefficients used in calculating the data were obtained at the same total pressure as that in the sample. In calculating the results of the hydrogen elimination reaction the sensitivity coefficient for HD v/as assumed to be intermediate between that of Ho and D2. This assumption proved to be a valid one as may be seen by the graphs of pressure versus sensitivity coefficient in Figure 7. Three sets of sensitivity coefficients were determined and listed in Table ^, A sensitivity coefficient for HD was determined from a gas sample prepared by the hydrolysis of CaHp with DpO. This gas sample contained small amounts of Dp and Hp, but it was predominantly HD. The spectra were corrected for the presence of Dp and Hp and then the sensitivity coefficient of HD v;as calculated. The variation in the sensitivity coefficients with total pressure was determined and plotted in Figure 7. losamples containing Dp, the Hp content v;as determined from the peak at m/e 2 after subtracting the portion due to D"*", The necessary data were obtained from the intensity ratio of m/e 4 and m/e 2 peaks in the mass spectrum of p\ire Dp.

PAGE 32

20 c •H +J •H to C (1> cn 2.20 2.001.801.60l.-IfO 1.20 1.00 .80 T 100 1^0 Pressure, microns Fig. 7. — Mass spectrometer sensitivity coefficient of H21 HD and D2 as a function of total pressure. (O) D2; (©) HD; (©) H 2*

PAGE 33

21 w ^ 9 CM OJ w w o w o « EH O CQ o EH H o H pR (in O O !h EH H > H EH H CQ ^ O w m d 3 o «H O o -p O W C\J •H Ch «H Q O O !>, -P •H > •H -P •H CO O CQ CM CQ -P •H O C3N • o =is a

PAGE 34

22 Th.e mass spectra of tlie solid dimethyl amineb or anes were obtained at 70 ev by placing the sample on the end of a probe which extended into the ionization chamber. The probe was at room temperature during the measurement of the spectra. The vapor pressure of dimethylamineborane at room temperature was sufficient to produce good spectra. For the results of these analyses see Figures 8, 9, 10 and 11.' The mass spectra of the solid compounds showed a low intensity peak at the m/e corresponding to the mass of the parent compound. The mass peak of m/e one less than the parent peak, and the parent peak had a ratio of 8.4 in the cases where the hydrogen was bonded to the boron atom and a deuterixim bonded to the nitrogen atom compared to 7»9 for the reverse case v/here hydrogen was bonded to nitrogen and deuterium to boron. This suggested that the BH bonds were lost more readily than EE bonds. The spectra of the completely deuterated and undeuterated compounds also showed the same trend as indicated in Table 5. The ratio of the m/e peak of mass two less than the parent peak, to the parent peak, implied that a BD bond was lost much more readily than an ND bond. The most intense m/e peak in each spectrum corresponded to the loss of a deuteriiim atom when the compound contained BD bonds and to the loss of a hydrogen atom when the compound contained BH bonds. But the numerous peaks, in

PAGE 35

23 o

PAGE 36

24

PAGE 37

25 Hi 9 EH H W Eh g o cii o « w CO o •H -P •H H W .06 S o o •H -P O fcD O O >s o •H O J4 O M -P m o o o \ ctf S J4 -c? O o o LPs

PAGE 38

26 each group of peaks differing only by one m/e unit, suggested that the CH bonds in the methyl groups were also being broken under the conditions at v/hich the spectra were obtained. This would make any quantitative use of the relative peak intensities open to doubt as to v/hether the hydrogen atom lost had been originally bonded to a carbon, nitrogen or boron atom. Preparation of (CH;^)a^D^ from (CH^)^NBH, in an acidic D2O solution (11) Sulfuryl chloride (0,5 ml) was pipetted into a flask containing deuterium oxide (20 ml) in a dry nitrogen atmosphere and stirred magnetically for twenty minutes, Trimethylamineborane (15.7 mmoles) was dissolved into 50 ml of anhydrous diethyl ether in a 200 ml round bottom flask. The deuteriiim oxide solution was poured into the ether solution and immediately the flask was fitted with an adapter for the vacuum system, attached to the system, and submerged in liquid Np* and evacuated. The reaction flask was warmed to room temperature (2^°) and stirred magnetically. After six hours, the reaction mixture was condensed in liquid No and, in approximately a ^00 ml volume, there was 95.5 cua of non-condensible gas. The DoO-ether mixture was transferred to a separatory funnel and the reaction flask washed v/ith a 20 ml portion of ether which was then added to the reaction mixture. The

PAGE 39

27 layers were separated, keeping the ether layer in the separatory fxinnel. Excess anhydrous potassiiim carbonate was added to the ether solution and the mixture was set aside for forty-five minutes. The ether solution was transferred into a 200 ml round bottom flask, washing the KoCO, with two 20 ml portions of anhydrous ether and the wash solutions added to the ether solution. The ether v:as removed by distilling under vacuum from room temperature into a liquid Np trap. When liquid ether was no longer visible, a -78° bath was placed about the flask and the last of the ether removed into the liquid N^ trap. This was to prevent loss of product by sublimation, A white solid residue remained. A white product was sublimed from this residue to give a 7*^.2 per cent yield. The infrared spectrum (Figure 12) of the sublimed product in CCl^ solution agreed with the spectrum of trimethylamineborane v/ith. the peaks attributed to a BD stretching vibration shifted to longer v;avelengths. There was a peak in the region of the BH stretching vibration, but it was less intense than the 3D peak. Assuming the absorptivity coefficient, a, to be the same for both the BD and BH containing compounds, from Beer's law, A = abc, the concentration of the BH compound was calculated to be 2,7 per cent of the concentration of the BD compound. No other analyses were made of this compovmd.

PAGE 40

80 70 60 50 40 30 20 10 28 2000 Wave number, cra--^ Fig. 12.— Infrared spectrum of (CH^)^NBD-j.

PAGE 41

29 Preparation of '^2^^ ^^^^ (^^'^5)3^^5 ^^^ BF;,(g) Trimethyl amine borane-d^ (-^.0 mmoles) was sublimed into a reaction tube fitted with a stopcock and a side arm filled with mercurj such, that if the tube were inverted the mercury sealed the stopcock from the contents of the tube. Boron trifluoride gas v:as entered into the vacuum system from the storage tank. The gas v;as purified by distilling from a CG1^-CHC1^-C02(s) trap (-78°), through an ethylbromide slush (-119°), into a liquid Np bath (-196°), before condensing into the reaction tube. The reaction tube was then warmed to room temperature and set aside for an extended period of time. A liquid phase v/as present in the reaction tube after the tube warmed to room temperature, but the liquid phase slowly disappeaLred. After eighty to ninety hours at room temperature, the reaction tube v;as attached to the vacuum system, the products condensed in liquid Np.and any noncondensible gas removed. The reaction tube was then warmed to -78° and the volatile fraction ^^^as removed and condensed onto excess anhydrous diethylether. Any unreacted 3F^(g) would form the etherate and the BpD^ could be separated from it. The ether flask was warmed to -78° and a product, BpDg, Usually this occurred overnight. Care must be taken to prevent this liquid phase from holding the mercury next to the stopcock on solidifying and thus sealing all gaseous product in the tube.

PAGE 42

30 was distilled from a -78° bath, through a -119° hath, into a -195° hath. The distillation V7as done rapidly to prevent contamination of the BoD^ with ether vapor. The material in the -196° trap was dihorane-dg.' The results of the experiments were as follows in Table 6. TABLE 6 EEACTION OF CCH-,)-,E[BD-, AITD BF-,(g) Compound Mmoles Yield Yield Vapor Presmmoles Per Cent sure at CS Slush^ 2 (CHj)^NBD, i^.OO 1.6^^ 82.0 239.0 mm BFjCs) 5.71 CCH,)^NBDj 2.11 0.90 84.9 258.5 mm BPjCs) 5.78 literature value is 258,5 mm (6). The infrared spectrum of the BpDg was in agreement with the reported spectrum (^5). The resolution of the spectrum (Figure 15) v;as poor but it did show a low intensity at 2500 cm" which was due to a BH stretching vibration, and the intense BD stretching vibrations at 1810-184-0 cm"* and 1954 cm" . A rough estimate of the concentration of BH to BD from Beer's law gave a ratio for C-ntr/Cg^ oi" O.I5. The absorptivity of BH and BD were assumed to be equal in this calculation.

PAGE 43

31 2000 .-1 2500 2200 Wave number, cm Fig. 13. — Infrared spectrum of '^^p^^ 1800

PAGE 44

32 A sample of tlie diborane-dg was pyrolyzed and a mass spectrum run on the products. For the results of this experiment see Table 2, Variation in the per cent reaction of (CH:,),NBH, and BF:.(g) to yield BpHg Boron trifluoride will displace diborane from trimethylamine borane according to the equation (28): (CH^)jNBHj + BF^(s) (CEj)^!^?^ + \ ^^^ [2] The extent of this reaction in mercury-sealed bulbs was studied as a function of time in order to determine the optimum time for the practical synthesis of diborane at room temperature. The folloxving variations, listed in Table 7, in the per cent reaction v/ere determined. The extent of reaction v;as followed by measuring the total pressure in the reaction flask as a function of time. For complete reaction, the total pressure should be one half of the initial pressure according to equation [2], The graphs (Figures l-^and 15) showing the variation in total pressure v/ith time indicated that after fifty hours the reaction, for practical purposes, was complete since the additional amoxint of diborane produced in the next thirty hours did not warrant the extra time spent. The plots indicated little difference in the rate of decrease in total pressure

PAGE 45

53 m EH W pq o O O H EH O W o cnix! cti f^pq 02 a ^ •H O &H W Fh Q) P4tH CO W O CM a pq s --^ to w a rOi o F-4 e FQ B -d pq w o^^ o ft r^ a SK a o o ^ OJ 4d^ ^ OJ CM CM OJ OJ CM CM OU CM C\J LTN KN LIA rCs K\ * « • • • C7N KN O CJ^ iH pH <^ [>OITS CM VD 1) O« « • • • CM CM dO O VD -d*X» [>CO CJN OOCO KN O^ kD CO CO % * « ' • « o o o o o CM LA ^ O OLfN LfN Lf\ lTn *i) H CM O O O % « « % • to, CM CM ^<^ CM lA

PAGE 47

35 with time V7h.etlier BFv(s) or (CH;,);,EBH;, were in excess, altiiough. tlie sample containing excess BF^(q) had a 10 per cent greater extent of reaction. The data implied that at room temperature the reaction was only 85 per cent complete in eighty to ninety hours. It could be possible for the reaction mixture to reach an equilibrium state in which BpHg was displacing B?,(g) in the reverse reaction according to the equation: (CH^)3NBH^ + BF^(g) ^ (CH^)^NBP^ + | B^E^ [3] An equilibrium such as this could explain the small change in total pressure after sixty hours. The attainment of an equilibrium state is supported by the work of Miller and Onyszchuk (28) , who found in forty-five minutes at 130^-1^0° • an average displacement of 23. ^Jper cent BF,(g) in (CH^),1T3F, by ^2^^ and an average displacement of 85,3 per cent ^2^^ from (CH^),N3H^ by B?;,(g). However, Graham and Stone (17) reported that after heating BpHg and (CH;.),NBP, for twelve hours at 80° the gas did not show any evidence of B?^ in the infrared spectrum. They concluded that no reaction had occurred \inder these conditions. The implications appear to be that for the equilibrium to be established, the temperature must be greater than 80° or the time must be longer than tv;elve hours. But the data show that the rate of displacement depends greatly on temperatirre. Miller and

PAGE 48

36 Onyszchuk (27) achieved the same per cent displacement in forty-five minutes at IJC-l-^-O" that we obtained in approximately ninety hoiirs at room temperature (23°-25°). The displacement reaction proved to be an impractical method for preparing diborane. It was used in this work to prepare deuterated diborane. S, S. Davis (11) had reported the exchange of boron hydrogens in trimethyl amine borane with acidic DoO to be rapid and quantitative. In this manner, (CH^)^E3D^ could be prepared and then BoDg displaced from the adduct by BP;^(g). Thus, BpD^ could be prepared from readily available and inexpensive starting materials, without the use of deuterium gas to deuterate the diborane, or without the use of a borodeuteride salt. The displacement reaction did not give diborane of sufficiently high deuterium content and BpD^ was prepared afterwards with E'aBD^ as the source of deuterium. Preparation of B^D^ from ITaBD^ and 3F^(g) in diglyme (4) Sodium borodeuteride (1.0 gram) v/as placed in a 100 ml round bottom reaction flask in the Dri-Lab controlled atmosphere box.^ A stopcock adapter for attaching the flask to the vacuum system was added to the reaction flask. The flask was then attached to the vacuiom system and was evacuated. A static charge on the pov;dered ITaBD^ prevented a quantitative transfer of the material from the gl as sine weighing paper into the flask.

PAGE 49

37 The diglyme (ethylene glycol dimethyl ether) to be used as the solvent was refluxed and distilled from sodium metal, again distilled from LiAlE^, and finally transferred from LiAlH^ under vacuum into the reaction flask at -196*'C. Boron trifluoride was condensed into the vacuiim system directly from the storage tank. The B'F-^ was then purified by distilling it from a CCl^-CKCl^-COp trap (-78") through an ethyl bromide slush trap (-llS**) into a liquid Np trap (-196°). The gas was then condensed into the reaction flask submerged in a liquid STo bath. A total of 53.1^ mmoles of BF^, was condensed into the reaction flask. The molar ratio of B?;.(g) : K'aBD^ was 2.2:1. The reaction flask was allowed to v;arm to room temperature (22°), After thirty minutes, the reaction was cooled to -78° and the volatile fraction removed into a -196° trap. This procedure was repeated twice with reaction times at 22° of thirty minutes and sixty minutes. The total amount of volatile material removed from the reaction flask was 10.76 mmoles. The material had a vapor pressure of 25^,0 mm at carbon disulfide slush temperature (-111.9**). The literature value (6) for the vapor pressure of 'S>2^f^ at this temperature was 238,5 mm. To remove any possible BP^ impurity in the B2Dg, the gas was condensed onto anhydrous diethyl ether. The gasdiethyl ether mixture was warmed to -78° and the volatile

PAGE 50

38 portion was transferred into a liquid No trap. After thirty minutes, 10.55 nnaoles of volatile material liad transferred from the ether flask. This material had a vapor pressure of 258.5 mm in a carbon disulfide slush bath, in agreement v/ith the previously cited literature value. An infrared spectrum was run on the material before it X'/as reacted with ether. The spectrum was in agreement with that reported for '^2^c^ C^5). No spectral evidence was noted for the impiirity which vj-as removed by the diethyl ether. There was a lov: intensity peak at 2510 cm" , A rough estimate using Beer's law showed, that according to this BH peak, the ratio of C-n-r : C^-q was 0.058 assuming that the absorptivity for BH and BD are equal. The assumption was correct for the absorptions in the infrared spectra of (CHj)^NBHj and (CHj)^NBD^. A sample of this BpDg was pyrolyzed and a mass spectrum run on the non-condensible products. For the results of this experiment see Table 2. Preparation of BgHg from NaBH^ and BP;,(g) in diglyme This preparation v;as done in the same manner as the preparation of BpDg from NaBD^ and BP^^Cg) in digljrme. The amounts of reagents used were 0,^ g (10.58 mmoles) NaBH^ and 22.25 mmoles BF^(g). The molar ratio of BF,(g) : NaBH^ was 2.1:1.

PAGE 51

39 A product was isolated in 78.7 per cent yield (7.05 mmoles) which had a vapor pressiire of 225 nim at carbon disxilfide slush temperature. The reported value (6) at this temperature is 225 in^i. An infrared spectrum of this compound was identical with that reported for BpHg in the literature (^3). General procedure for preparation of (CH^)2NI> A solution of n-hutyl lithiiim in n-hexane (1,6 M) was syringed into a 100 ml round bottom flask under a stream of nitrogen. An adapter to the vacuum system was inserted immediately into the flask, the contents were condensed in a liquid No bath, and the flask evacuated. Excess dimethylamine which had been stored over sodium metal was condensed into the flask containing the hexane solution. The reaction flask was then warmed to room temperature and immediately a white precipitate appeared. After thirty minutes at room temperature, the reaction flask was cooled to 0° for thirty minutes, and then a volatile fraction was removed into a liquid N^ trap. The transfer was done slowly to prevent excessive spattering of the white solid as the liquid phase was removed. The remaining excess dimethylamine, n-butane, and n-hexane were removed with the flask at room temperature. A white solid residue, LiN(CH,)2t remained in the flask, according to the equation:

PAGE 52

40 (CH^)2NH + n-C^H^Li LiN(CH^)2 "^ ^"^li^lO ' '"^^ A vial equipped v/itii a capillary break-off tip containing deuteriiim oxide (1 ml) was attached to tlie vacuum system, the tip of the vial was broken and the D^O condensed onto the amide salt in a liquid No "bath. The reaction flask was slov/ly warmed to room temperature and an immediate increase in pressure was noted. The white solid had turned dark brown after fifteen minutes at room temperature. Part of the flask was cooled to 0° and kept at this temperature for one hour and forty-five minutes. All the volatile material in the reaction flask was transferred into a flask containing excess anhydrous potassium carbonate in a liquid ^2 bath. The K2CO, flask was warmed to 0° in an ice bath and kept at this temperature for two hours. The K2CO;, mixture v;as then cooled to -78" and a volatile fraction removed into a liquid Uo trap, requiring approximately forty-five minutes. This material was deuterated dimethyl amine. A possible impurity in this material would be monodeuterated n-butane due to incomplete reaction of the n-butyl lithium and the dimethylamine according to equation [4] . The results of the experiments are as shown in Table 8.

PAGE 53

41

PAGE 54

42 Gas phase infrared spectra of these different preparations were identical. The spectra were similar to the spectrum of undeuterated dimethylamine (34) except for some shifting of peaks in the I3OO cm" to 1000 cm" region. The deuterated amine v;hich had a vapor pressiore at 0° greater than 760mm did have an extra peak at 2150 cm" , An attempt to purify this sample by distilling a fraction from an ethylhromide slush (-119°) into a liq.uid Np trap resulted in an increase in intensity in the infrared spectrum of the peak at 2150 cm" in the fraction which transferred into the liquid N^ trap. The infrared absorption peaks of nbutane (20) were not detectable in the spectrum. The peak at 2150 cm" could be attributed to a C-D stretching frequency in monodeuterated n-butane. The C-D stretching frequency in the deuterated methanes varies from 2085 cm" in CD^ to 2205 cm" in DCH^^ (29). This material was more volatile than, the amine, v/hich was consistent with the relative vapor pressures of dimethylamine and n-butane, nButane has a higher vapor pressure than dimethylamine ( 8 ). General method for the preparation of (CH,)2KNBH;z from "22^6 and (CH^)2ira Diborane and an excess of dimethylamine were condensed into a 50 ml round bottom flask in a liquid nitrogen bath, A CCl^-CHCl,-G02(s) bath (-78°) was then placed about the

PAGE 55

^3 flask and it remained at this temperature for an extended period of time (for the exact reaction times see Table 9). The excess amine was removed from the flask at 0° into a liquid No trap, and the liquid product remaining in the flask slowly solidified on storage at room temperature. Mass spectra of the boranes shov;ed a mass peak of low intensity corresponding to the mass to charge ratio of the parent ion, and a high intensity peak at a mass to charge ratio corresponding to the loss of a hydrogen or deuterium atom. The results of the preparations of the variously deuterated dimethylamineboranes are given in Table 9. For the method of analysis see page 5. The infrared spectra are given in Figures 2, 4 and 5. Each of the variably deuterated dimethylamine boranes was heated at 100-102° and the non-condensible reaction product analyzed in the mass spectrometer for the percentages of D2, HD, and H2. The results of these elimination reactions are given in Table 10, The infrared spectrum of the compounds in carbon tetrachloride solution, in each case, contained peaks where the NH, KD, BH, and BD stretching vibrations occur. Beer's law was used to calculate the relative percentages of each compound using the CK deformation absorption as the internal reference. For the results see Table 5»

PAGE 56

HAON o en o H PQ H « EH CQ H P 03 O H s o H EH o5 O EH CO ^ H O o 12; txj O pq w o •H -p « © o o O -P o I t O fH o «Hn5 CO •d +3 o •H i:i +» -P o •H +5 +3 o o erf ri -P to o ^ 2 •H O O •H g tfs N^CM • • •• • • •« -d I CO o rt -d CM I iH H • • •• •d 0) p a iH,Q d -d o d rt o O
PAGE 57

^5 B CO B g H O E^ H O S O M W « pq w w < OJ KNO w O CQ O CU H O H EH Es pq W « W EH CO O O CO EH CO C\J pi -P o Ota p ?H C\J PI a o
PAGE 58

A-6 Preparation of (CE^)2^1IBE^ from (CH^)2ND2C1 and LiBH^ N-Deuterodimethyl ammonium chloride was prepared "by condensing K-deuterodimethyl amine (7.6? mmoles) into a tared 50 ml round bottom flask containing deuterium oxide (1 ml) and thionyl chloride (15 mmoles).^ The reaction flask was warmed to 0°, After one and one-half hours, all volatile material was removed from the flask. The increase in weight of the reaction flask implied that only 0.86 mmoles of product was formed. The volatile material was transferred back to the reaction flask. A gas phase infrared spectrum of the most volatile materials in the reaction flask v/as identical to that of SOpCg) (3 ) and mono-deuterated n-butane which was known to be a contaminant in the (CH^)^^) used. The n-butane and some of the SOpCg) was transferred from the flask at -78° into a -196° trap in thirty minutes. More deuterium oxide (1 ml, making a total of 2 ml) was distilled into the reaction flask and dimethylamine (6.3 mmoles) was condensed into the flask. After forty-five minutes at 0° and thirty minutes at 25°, all the volatile material was removed; the weight gain by the reaction flask implied 6,^8 mmoles dimethyl ammonium chloride had formed. The hydrogen chloride impurity in the thionyl chloride was removed by warming the thionyl chloride to -78° and exposing it to a -196° trap for twentyfive minutes.

PAGE 59

^7 The ion should have been almost completely deuterated since (CH^)pNH2'** is kno\^m (38) to exchange rapidly with the solvent in acidic solution and a large excess of heavy water had "been used. The infrared spectrum in a Nujol mull contained absorptions in the 1900-2^00 cm" region and none greater than 3000 cm" , indicating the absence of NH absorption. The completely deuterared ammonium ion has absorptions at 221^ and 23^6 cm" (30). Therefore, the product should be primarily the deuterium-containing material. In the Dri-Lab controlled atmosphere box, lithium borohydride (approximately 11,5 mmoles) was added to the flask containing the (011^)2^2001, The flask was attached to a vacuum system and approximately 25 ml of diethylether (stored over OaHp) was distilled into the flask. After forty-five minutes at 0® and fifteen minutes at room temperature, no evidence for reaction v;as noted. The reaction flask was returned to the Dri-Lab and Li3H^ from another bottle added to the solution. Immediate gas evolution was noticed. Excess LiBH^ v;as added and the solution was magnetically stirred. After one hour when no more gas evolution was noticed, the reaction mixture was filtered and the residue washed with approximately 10 ml of ether. After the ether was removed by vacuum distillation, a liquid phase containing a white solid remained in the flask.

PAGE 60

^8 A small amount of the liquid product was vacuum distilled from the reaction flask at room temperature into a -196" trap. An infrared spectrum of this material showed absorptions at $210 cm""-^ (NH) , 2300-2-4-00 cm""'" (m},BH), and at 1750 cm"-^ (BD). Since the material distilled so slowly, it was recryst alii zed from cold caroon tetrachloride and nhexane. The recrystallized product was a solid at room temperature. It was placed in a vacuum sublimation apparatus, and the most volatile fraction was removed by pumping on the sublimator at room temperature and collecting a product in a -196° trap. After tv;enty-five minutes, the cold finger in the sublimator was cooled to -78*^ with COoCs) and the material was collected for nine hours. The initial material removed from the sublimator into the -196° trap was a liquid at room temperature and the compound collected on the cold finger was a solid at room temperature. The infrared spectra (Figure 16) of the recrystallized compound and on the fractions obtained by sublimation were identical and showed absorptions in the same regions as the material initially transferred from the reaction flask. The percentages of NH and BD bond in the compound were calculated from the infrared spectra using Beer's lax^r. The percentages were calculated relative to the OH deformation at 1475 cm as the internal reference. The results were given in Table 11,

PAGE 61

49 o

PAGE 62

50 a?ABLE 11 PER CENT BD AND mi BONDS IN CCH^)2DNBHj PREPARED PROM (CH^)2ND2C1 AND LiBH^ Compound:

PAGE 63

51 Any dimethylaiairLoborane impurity should be contained in the initial fraction sublimed from the reaction flask, A small impurity of this compound in the spectrum would cause a lov; percentage of MI relative to GH in the calculation, and the absorption at 1750 cm" in the spectrum of dimethylaminoborane would cause a high percentage of BD relative to CH to be calculated. Another source of error would be in using the absorptivity ratio of BH:CH in (CEy.) ^iKBE-;r to calculate the per cent BD, the assximption made here has been shovm previously to be a questionable one. The first fraction sublimed from the recrystallized compound could also contain an aninoborane impurity. The percentage of BD and NH containing compounds show that some fractionation was accomplished by the method of purification, but still the recrystallized material before and after sublimation were essentially the same. The larger percentage of NH bonds compared to BD bonds would indicate that hydrogen-deuterium exchange occurred before the formation of the aminoborane. Thus, this method of preparation, under the experimental conditions used, did not give a pure product containing deuterium 23'' (39), b ^Dimethylaminoborane has a vapor pressure of 10mm at See page 17.

PAGE 64

52 » only on ttie nitrogen atom, due to tlie exchange between KD and BD bonds prior to tbe reaction to form tbe amine borane. v Hydrolysis of (CH^)2DNBD^ in 0.1 M hydrocbloric acid N-Deuterodimethylamine borane-d, (0.91^ mmoles) was sublimed into a 50 ml round bottom reaction flask and bydroctaoric acid (20 ml of 0.1 M) was distilled into the flask. After nine bours at room temperature, tbe reaction product was condensed and tbe non-condensible gas removed with a Toepler pump, Tbe amount of non-condensible gas (2,70 mmoles) corresponded to complete reaction according to tbe equation: (CH,)2DNBD^ + H^O"^ + 2H2O (CH^)2NDH'^ + B(OH), + 3HD [5] Tbe mass spectrum of tbe non-condensible gas gave 1^ per cent HD and 86 per cent H2. The lAper cent HD in tbe noncondensible hydrolysis product indicated that tbe rate of exchange of BD with the solvent was not so much more rapid than the rate of solvolysis that all of the deuterium bonded to boron exchanged before the solvolysis reaction was complete, R. E. Davis (11) reported only H2 gas produced in the acid hydrolysis of (CHv):zNBD^ due to the rapid acid catalyzed exchange of the BD with the solvent. However, the rate of

PAGE 65

53 acid hydrolysis of (CH,)2KNBH, is greater than that of (CH,),KBHj (24). Therefore, the BD in (CHj)^NBD, would have had more time to exchange before solvolysis than in (CH,)2HimD,. H. C. Kelly (2$) reported that for £-toluidine horaned;z in a 50/50 mixtiire of dioxane and water with no acid present, the rate of exchange of BD with solvent was negligible relative to the rate of solvolysis; but that at high acid concentrations the rate of exchange increased. Kelly found no primary hydrogen isotope effects in the solvolysis reaction. Determination of reaction conditions for the hydrogen elimination reactions" Dimethylamine borane was heated in sealed glass tubes for various periods of time and the amount of hydrogen eliminated measured by transferring the hydrogen into a calibrated bulb with a Toepler pump. The temperature of 100** was used for the reactions because it gave a reasonable rate of hydrogen evolution. In general, the extent of reaction at low percentages was not very reproducible, since the small amounts of hydrogen being measured (usually 0,4 mmoles ot 0.03 mmoles in a volume of 105,8 ml) were subject to experimental error. The results of the experiments were given in Table 12, These experiments led to the selection of reaction

PAGE 66

5^ TABLE 12 REACTION CONDITIONS FOR E^ ELIMINATION REACTIONS (CH^)2HNBH, mmoles

PAGE 67

55 times of twenty-four lioiirs in the initial experiment of heating (CH,)2DNBD, with (CH^)2HKBH, and of a time of one hour in the experiments where just the initial reaction products were desired in an attempt to ascertain a kinetic isotope effect, A trace of free dimethylsmine added to one of the reaction tubes did not significantly affect the extent of reaction. Hydrogen elimination on heating dimethylamine boranes Dimethylamine borane was sublimed from a storage flask into a reaction tube equipped with a capillary breakoff tip. The amount of compound sublimed into the reaction tube was determined by weighing the storage bulb before and after the sublimation. Then a second dimethylamine borane, containing a different isotopic distribution of hydrogen atoms on boron and nitrogen was sublimed into the reaction tube, and the tube sealed off with a torch. In a typical experiment, approximately one mmole of each compound was used. The reaction tubes were heated at 100 + 2°C for the desired reaction time. After rapid cooling the non-condensible gas was transferred into a bulb by using the Toepler pump. The gas samples were stored at room temperature until analysis in the mass spectrometer. The results of these experiments were given in Table 15.

PAGE 68

56 H EH EH M f=l pq o H EH M s H o W o CO EH CQ CVJ P O u CVJ d o O -P o U cd © o 0) •H EH CO xi ti CQ OrH PhO o i o KN O IN ON O CvJ ro, CM « • kD LTN LA 0^ ^ 00 » I CM I vD CM (JN O i^ GD m m

PAGE 69

57 CVJ p ti o o u o If f^ O-P o O (1> •H EH o O K> <1> EH W ci to p o OH p«o s a o s o • • H O CM KN • • ITS lA ON ID • « 5 5 rCN

PAGE 70

58 Experiments to eliminate possibility of isotopic interclianp^e during^ the elimination reactions I, Heating of Do and H^ in the presence of mercury vapor , — A bulb was attached to the vacuum system, evacuated, and submerged in a liquid Np bath. After condensing mercury vapor into the bulb for ten hoiirs, approximately equal moles of hydrogen and deuterium were placed in the bulb. The bulb was closed and heated at 100° for twenty-four hours. The gas was analyzed by the mass spectrometer and shown to be only hydrogen and deuterium. No exchange of the Hp and Dp occurred under these conditions. Therefore, it was concluded that the reaction gases did not exchange among themselves, even in the presence of mercury vapor. II. Heating Dg with CCH^)2H1CBH^ and CCH^),NBH^ to determine if exchanp;e occirrred . — The amine borane was placed in a tube with a capillary break-off tip, condensed in liquid Np and evacuated. Deuterium (150 mm) was placed in the tube and the tube glass sealed. The tube was heated at 100° for one hour and then a sample of the gas removed from the tube by breaking the tip and allowing the gas to expand into a bulb. The gas was analyzed in the mass spectrometer to determine if any HD had been produced. The gaseous product did not contain any material with a m/e of 3 according to the mass spectral analysis. The solid materials

PAGE 71

59 were dissolved in spectral grade carbon tetrachloride and the infrared spectra (Figures 1? and 18) determined. Thie spectra showed no absorption whatever at 1750-1800 cm" , where BD absorbs intensely, but were identical to the spectra of (CH,)2HNBH^ and (CH^)^NBH^. The experiments were repeated heating the deuteriumamineborane mixtures for twenty-four hours. The infrared spectra in both cases contained absorptions in the 17001800 cm""'" range. The spectrum (?ig\ire 19) of (CH,)2H1TBH, and Dp after heating had weak absorptions at 1725 and 1850 cm" , and an intense absorption at 1785 cm" which is where the BD stretching vibration occurs. The spectrum (Figure 20) of (CH;,),1TBH;, and D2 after heating had a medium absorption at 17^0-1750 cm"-^. An infrared spectrum (Figure 6) was taken of (CH,)2HNBH^ after thirty-four hours at 100° and the spectrum had a medixim absorption at 1750 cm" , This absorption must be due to some reaction product and not to a BD vibration since there was no deuteriiim in the molecule or in contact with it. The absorption would cause an error in any estimation of the absorption due to BD in this region. Therefore, any calculations by Beer's law of the BD percentage, after heating which considered the absorption in the 1750 cm" region would be in error. The calculation would imply a larger percentage BD than actually existed.

PAGE 72

60 o o MO r-i O O CO I E c o o o CM 0) e pi c > CO o o o o I E o o t^ O 0) 00 X! rH e C 0) > o o o C\i o o CM CQ O CM a) '-^ c 5-1 O tH CM Q -P c •H -P o J3 CJ) o E d -P o o a, to Ti u c 00 r-i 0) -p CO • u a: o CQ X ^-^ C ^-^ o O CM a f-, -p -P .H O > 0) a, bo CO c
PAGE 73

61 o o o o o o e o o u O (U CO X> ^^ e c > CO o o o o o o o o CO u o s: u c I c o Q -p CO c •H « ^^ 0) -p CO m z: o o 6 3 p o a CO Tl u CO V. c M I I OS o CO o o o CM 80UBq.q.TtasuBaq. %uao jsj

PAGE 74

62 o

PAGE 75

65 Tlie infrared spectrum of (CH,);,NBH, was unaffected by heating for twenty-four hours at 100°. The data showed that exchange between dimethyl or trimethyl amine borane and deuterium did not occur in one hour at 100" but that exchange did occur after twenty-four hours at 100". The data are summarized in Table 14. III. Heating of (CH,)2DNBH^ to determine if exchange occurred between ND and BH within the molecule . — N-deuterodimethylamine borane was heated at 100° for eleven hours in a sealed glass tube. The material was handled in the vacuum system or in the Dri-Lab controlled atmosphere box. The solid material was dissolved in spectral grade carbon tetrachloride and an infrared spectrum (Figure 5) obtained. This spectrum was compared to a spectrum of an unheated sample of (CH^)2DNBH^ (Figure 2). Beer's law was used to calculate the concentration of compound containing NH, ND, BH, and BD using the CH deformation peak as the internal reference (see Table 5). The spectra (Figures 2 and 3) showed an enrichment in the percentage of deuterium contained in the unreacted material, and a decrease in the percentage of hydrogen-containing material. The concentration of NH containing compounds decreased from 22 per cent to 15 per cent, ND increased from

PAGE 76

6^ CM &4 g P3 OJ W o o o M EH O U -P en o m o rt ft 2 W •P o ft CQ H O g o o O erf •H f^ P o O ft erf S EH O •H -P O erf o H o d o o o o o o H PI o O o

PAGE 77

65 78 per cent to 85 per cent, BH decreased from 97 per cent to 92 per cent and BD increased from 3 per cent to 8 per cent when the compoiind was heated. During heating, dime thylaminoh or ane was formed, which had an absorption peaJc in the same region (1750 cm" ) as the BD absorption. Therefore, the absorption peak at 1750 cm"* in the spectrum of (CH,)2DNBH2, after heating, could not be attributed solely to BD containing compounds, and the calcxilation of 8 per cent BD was an over-estimate. Any calculations or quantitative considerations of this 8 per cent BD would be questionable. IV. Heating of (CH;,)2DNBH, and (CE^)2^imD^ to determine if amine exchange occurred . — Dimethylamine borane-d, and Ndeuterodimethyl amine borane were sublimed into a tube equipped with a capillary break-off tip, and the glass was sealed. Duplicate tubes were heated at 100" for twenty-four hours. The non-condensible product was removed and analyzed in the mass spectrometer (see Table 15). The solid products also were analyzed in the mass spectrometer (Figure 21), The experiment was dpne in duplicate. In neither case did the mass spectra show a peak at the mass to charge ratio of 6$. The peak at this m/e would occur only if the completely deuterated compound were present in the solid materisLl. If amine exchange occurred, then this peak would

PAGE 78

IT\ _

PAGE 79

67 "be present; otherwise, the highest mass to charge ratio would he 62, the parent ion peak for (CH:z)2HKBD;z. Therefore, amine exchange does not occur under the reaction conditions. In the msLSs spectrum for (CH;,)2H1TBD^, the ratio of the intensities of peak 51:62 was 7.9. The mass spectra, taken on the solid material after heating, contained a ratio of the peak intensities of 61:62 of 5.^ and 5.8 for the two experiments. In the recorded spectra, the peak at m/e of 62 was a shoulder on the peak at m/e of 61, The same base line was used to get both peak heights. This may not be the actual height of peak 62, but it may actually be less intense than measured, which would tend to bring the ratio of 61:62 more into line with that for the spectrum of

PAGE 80

DISCUSSION OP RESULTS Possibility of 3iydroj:^en-deuteriiim exchange Every isotope study must be thoroughly checked to make sure that an exchange process does not vitiate the conclusions. Exchange between hydrogen and deuterium, either in the gas phase or when bonded to other atoms within a molecule, could invalidate the experiment. Therefore, it becomes of primary importance to ascertain if any exchange reactions could occur under the experimental conditions. Even a kinetic isotope effect in the hydrogen elimination reaction X'^ould cause enrichment in the unreacted compounds of the less reactive isotope, and thus affect the measured ratio of HpiHDrDo from a long term reaction. Therefore, the critical experiments to determine the isotope effect were run for only 1 to 2 per cent hydrogen elimination to avoid this possibility. There are five processes by which isotopic interchange could occur: (a) H2+ D2 ^ 2 HD [6] (b) (CH^)2HNBH^ + D2 !^ (GH^)2HNBH2D + HD . [?] (c) (CH,)2DNBHj ti (CH^)2H1TBH2D [8] (d) (CH^)2HI:TBH^ + (CH^)2DKBD, == (CK,)2DNBH, + (iCE^)^mBII^ C9] 68

PAGE 81

69 (e) Exchange between reactants prior to amineborane formation. Each, of these possibilities was examined experimentally. The mass spectriom of a mixture of D2 and H^ with mercury vapor, heated for twenty-four hours at 100°, did not contain a peak at m/e of 5. Therefore, exchange reaction [6], even over mercury metal, did not occur. Dimethyl amine borane and trimethyl amine borane were heated with deuterium gas to determine if exchange occurred. After one hour at 100°, neither iCE^^^^IBE^ or (GH,)^imH^ had exchanged with the D^ gas. A mass spectrum of the gas from the reaction tube showed no HD, and the infrared spectrum of the solid materials showed no BD absorptions in the 1750-1800 cm" region. However, after twenty-four hours at 100°, both (GH^)2H1TBH;, and ( CH;, ) ;,NBH;, had exchanged to some extent with the Do gas, since their infrared spectra showed BD absorptions in both instances. The reaction products when dime thy 1 ami ne borane was heated were hydrogen gas and dimethylaminoborane, according to the equation: (CH^)2HNBH^ M^t_^g^ ^ (CH^)2NBH2 ClO] At 100°, the dimethylaminoborane will disproportionate (7) according to the equation: $(CHz)2NBH2 ^ [(CHj)2N]2BH + (GHj)2NB2H^ [11]

PAGE 82

70 Noeth (51) lias reported that (CH2)^N2*2BHj will give (CH,)2NBH2 and (CH2)2HNBH, when heated, and at 100*' the principal product v/as (CH,)NBH2 and H^ with some [(CH,)2N]2BH and (CH^)2NB2Hc being produced. Therefore, under the experimental conditions, the reaction products would not be just hydrogen aad the aminoborane, but a more complicated mixture of compounds. Since aminodiborane is known (6) to exchange with deuterium gas, the argument co;ild be made that it was this species or possibly the aminoborane which was exchanging with the deuterium gas and not the amine borane. But, the BH bonds in trimethyl amine borane exchanged with deuterium gas iinder the same conditions. Therefore, it is not unlikely that exchange also occurred between dimethylamine borane and Dp gas. Deuterium gas has been reported to exchange also with diborane (6,12,37) and with the BH bonds in borazine (10). From the experimental data, it can be concluded that exchange between deuterium gas and dimethyl or trimethylamine borane did not occur to a measurable extent in one hour at 100", but that it did occur in twenty-four hours at 100°. This result must be considered in the interpretation of the experimental results. The intramolecular exchange (equation [8]) did not occur when (CH,)2DNBH;, was heated. Since the starting material contained some NH and BD bonds, the change in

PAGE 83

71 relative concentrations in the infrared spectra, on heating, had to be calculated using Beer's law. The change in IM contciining compound gave the more accurate results^ and indicated an enrichment of the deuterium containing compound in the unreacted material. The enrichment in ND bonds, in the material remaining after heating, implied that the NH bonds were lost much more readily. A relative change in the moles of KH to the moles of ND can be calculated from the spectral data. If one considered the total moles of dime thylamineborane, a, to be the sum of the moles of NH and ND, then before heating there v/ere 0.22a moles of NH and 0.78a moles of ND, Assuming that the reaction was 30 per cent complete according to equation [10], then 0,50a total moles of both NH and ND reacted. After heating, the moles of NH were (0,15) (l,00a-0,30a) which was 0.105a and the moles of ND were (0,85) (1.00a-0,30a) which was 0,60a, This corresponded to a decrease in moles of NH of 0,115a or 52.3 per cent and in moles of ND of 0,18a or 25,1 per cent. These calculations indicate the following: (1) A hydrogen atom was eliminated 2,3 times more readily than a deuterium atom from the nitrogen, (2) There was an increase in concentration of the deuterium-containing compound in the unreacted material. ^See page 16. This figure should be reasonably accurate considering the data in Table 12,

PAGE 84

72 (3) Exchange "between the WD and BH was not occurring to any significant extent during the reaction time. Otherwise, a greater decrease in WD would be expected and less of a decrease or even an increase in WE would be expected since there were 3 BH bonds per WD bond available for exchange in the original molecule. The analogous calculation using the BD or BH absorptions woiild contain too large an error to be meaningful. Since ND and BH absorb in the same region, the BH absorption could not be used for the calculation. The BD absorption could be used, but to do so the following assumptions must be made: (1) The absorptivities ratio of 3H:CH is the same as that for BD:CH. (2) The absorptivities of BH:CH in (CH^)2HNBH^ is the same as that for BH:CH in (CH^)2NBH2. The second assumption was checked by heating (CH^)2HNBH^ for thirty-four hours at 100** and then subliming out the most volatile portion of the reaction mixture. The infrared spectrum gave a ratio of absorptivity of BH:CH of 1.30 compared to that of 1.90 for (GH^)2HNBH,. The lower value for the absorptivity ratio would cause an underestimation of the amount of BH compound actually present if the value of 1.90 were used in the calculation. But, the

PAGE 85

73 most significant point in this spectrum was an absorption at 1750 cm" . Th.is absorption meant that a reaction product also absorbed in the same region as the BD vibration. Thus, an over-estimation would be made of the amount of BD containing material in an infrared spectrum after the compound was heated; the 8 per cent BD calculated from the spectrum ai^ter heating was probably much greater than the actual amount of BD. Therefore, the experiment indicated that any intramoleculcir exchange between hydrogen and deuterium on heating was insignificant. To determine whether amine exchange occurred between the amine borane molecules, N-deuterodimethylamine borane and dimethylamine borane-d^ were heated at 100" for twentyfour hours, A mass spectrum of the solid materials did not contain a peak at m/e 63 which would be present if amine exchange had occurred to form N-deuterodimethylamine boraned,. The mass spectriom did show an intense peak at m/e 51 which had a shoulder at m/e 62. If (CH,)pDNBD;, had been present a low intensity peak at m/e 63 and a more intense peak at m/e 62 would have been expected. The intensity of the peak at m/e 62 did not increase, therefore no amine exchange took place. Moreover, its intensity could be fully accounted for by the presence of unreacted starting material. Therefore it was concluded on this basis that amine exchange did not occur.

PAGE 86

7^ However, tlie possibility of exchange between the gaseous elimination products and the dimethyl amine boranes could account for the absence of the completely deuterated amine borane. Even if amine exchange occurred to give (CH;,)2DNBI>2j asubsequent reaction with H^ could conceivably have reduced the parent mass peai: to an undetectable level. Thus, the above conclusion is open to some doubt. The failure to prepare (CH;,)pDKBH^ and (CH^)2HNBD, in which there was not also an impurity of BD or ND bonds implied that an exchange reaction was occurring before adduct formation. The only other method to produce the impurity would be an intramolecular exchange in the adduct, but experimental evidence discounted this possibility even when the adduct was heated. The compounds were prepared by condensing diborane and excess dimethylamine together and warming to -78° to form the adduct. An exchange reaction must have occurred at a rate comparable to the rate of adduct formations at this temperature, Dahl and Schaeffer (10) have reported that N-deuterodiethylamine exchanged with the BH bonds in borazine at -30° within three minutes. Since the ND bond in diethylamine exchanges with the BH bond in borazine, it would be reasonable to expect that a NH bond in a secondary amine could exchange with a BH bond in diborane. The infrared analysis of 'the dimethylamine boranes, prepared

PAGE 87

75 from the amine and diborane, seemed to indicate that this exchange did occur to some extent. The preparation of (CH;,)pDNBH;, from (CH,)plTDpCl and excess LiBH^ contained even more BD bonds according to the infrared spectrum than v;ere present in the compound prepared from amine and diborane. The infrared spectrum of (CH5)2DNBH^ implied that 72.8 per cent KD bonds and 13.2 per cent BD bonds were present in the compoiind. The calculation was made from Beer's law. using the CH deformation at 1^75 cm" as the internal standard. The comparison of 13.2 per cent BD bonds to 27.2 per cent NH bonds indicated that an exchange reaction occurred prior to the reaction to produce the amine borane. Otherwise, the per cent BD bonds should be equal to the per cent HE bonds if the exchange occurred after the amine borane was formed. If the exchange reaction did occur before amine borane formation, as the data implied, then the un re acted LiBH^ should have contained some BD bonds. N-deuterodimethylamine borane, prepared from the amine and diborane, contained 3 per cent BD as compared to 13.2 per cent in the compound prepared from (GE^^r) 2^^2'^'^ ^^^ LiBH^. These data indicated that the exchange reaction between (.CH,)2ND'^2 ^^^ ^^"h. occurred to a greater extent before adduct formation than was observed for the amine diborane reaction.

PAGE 88

76 Heatinp: mixtures of dimethyl amine "boranes containing; various distributions of hydrogen isotopes for one hour In order to determine if the kinetic isotope effect in the elimination reaction could be attributed to the nitrogen or boron-bonded hydrogen isotope, mixtures of dimethylamine boranes were heated at 100° for one hoiir. A reaction time of one hour was used to minimize the reaction extent so that the observed product compositions could be related unequivocally to a known isotopic distribution in the reactants. In each mixture, either the boron or nitrogen bond was one hydrogen isotope, and the other hydrogen bonds in that molecule and in the other molecules were the other hydrogen isotope; for example, one mixture was (CH-,)H1TBD, and (GH-,)I)NBD^. The isotopic distribution was t> t> varied to determine if a NH or a KD bond was eliminated more readily or v;hether a BH bond reacted more readily than a BD bond, The results in Table 13 show the following: (1) If one molecule contained a NH bond, and the rest of the hydrogen in the system was the deuterium isotope, then the principal product v;as HD, the ratio of H2:HD:D2 being 0,15:1.00:0.39. (2) If one molecule contained ITD bonds, and the rest being the hydrogen isotope, then the principal product was Yl^-, "the ratio of H2:HD:D2 being ^,75:1,00:0,0, (3) If one molecule contained BH bonds, and the rest being

PAGE 89

77 the deuterium isotope, then the principal product was HD, the ratio of H2:HD:D2 heing 0.52:1.00:0.6?. W If one molecule contained BD bonds, and the rest "being the hydrogen isotope, then the principal product was Hp, the ratio of H2:£D:D2 being 1.32:1.00:0.05. The data implied that a hydrogen atom was eliminated more readily than a deuterium atom since in each instemce the primary product was either Hp or HD and not Dpi even in the cases where there was only one M or BH bond in the system. The data were consistent with a BD bond being eliminated more readily than a IH) bond. In the experiments which compare these two bonds, the H2:HD ratio was 4-, 75:1. 00 for the ND case and 1.32:1,00 for the BD case. This implied that HD was eliminated between a BD bond and a NH bond approximately 3.6 times faster than betv;een a WD bond and a BH bond. The HD:D2 ratio of 2.56, v/hen (CH^)2DN3D^ and (CH;7)pH]IBD;r were heated, compared to the HD:D2 ratio of 1.39, when (CHv)2DNBD:, and (CH;,)2DKBH^ were heated, indicated that HD was eliminated betX'/een a BH bond and a ND bond 1.8^ times faster than between a NH bond and a BD bond. The HD:D2 ratio of 2.56, v/hen (CH,)2DKBD, and (CH,)2H1\[3D;, were heated, and the H2:HD ratio of ^.75 » when (OE^) 2^^^'^7, ^^^ iCE:,) 2^^^^'^x were heated, indicated a large kinetic isotope effect to be occurring when the hydrogen isotope bonded to the nitrogen atom v;as varied. Thus, the

PAGE 90

78 ratio of the rate constants for the nitrogen atom eliminating ( ^mi\ a hydrogen or deuterium atom \ ^-^ ) experimentally determined, varied between 2.7 and ^.8. Edwards (15) has predicted a ratio of the rate constants for the bond breaking process involving NH and ND bonds! ;f-^ ) of 8.5. The average experim l^)-^-' mental isotope effect for the I ;r^ 1 of 3.8 was less than the predicted ratio of 8.5. This implied that there was considerable, but not complete, loss of the Mi stretching vibration in the activated complex. The HD;D2 ratio of 1.39, when (CH^)2DNBD, and (CH^)pDNBH;z were heated, and the E^tBI) ratio of 1.32, when (CH,)2HNBD, and (CH^)2HiTBK;, were heated, indicated a small kinetic isotope effect to be present when the hydrogen isotope bonded to boron was varied. Thus, the ratio of the rate constants for the boron atom eliminating a hydrogen or deuterium atom r— ) , experimentally determined, was V ^BD / 1.3 to 1.4. Hawthorne and Lewis (17) calculated the ratio I ^BH ^ of I -r — to be 4,2 for the expected effect of isotopic substitution on boron from the loss of the BH stretching vibration at 2300 cm" . The observed isotope effect was much less than that predicted, implying that there was only a small loss of the BH stretching vibration in the activated complex.

PAGE 91

79 The maximiim isotope effect would be obtained when the bond to hydrogen or to deuterium was essentially completely cleaved in the activated complex, and the isotope effect would decrease with increasing bonding in the activated complex (^4). Therefore, the isotope effects observed should allow some predictions about a possible configuration of the activated complex in the hydrogen elimination reaction. The larger effect when ND was substituted for NH than when BD was substituted for BH .predicts that the NH(ND) bond was more affected in the activated complex than was the BH(BD) bond, Hawthorne (I7) has reported a similar situation in the hydrolysis of pyridine diphenylborane with water or deuterium oxide in acetonitrile solution. The reaction was found to be first order in both pyridine diphenylborane and water, and a primary kinetic isotope effect was determined. The ratio of the rate constants \ ij— of 6,90, when deuterium oxide was substituted for water, was nearly as large as that predicted for a complete loss of the OH stretching vibration in the activated complex of 9.9. The ratio of the rate constants (If)-52 observed on isotopic substitution on boron was much less than the predicted! ^^ of 4.2 for complete loss of the BH stretching vibration in the activated complex.

PAGE 92

80 Hawthorne (17) proposed tiie transition state (I) for the BH bond hydrolysis and suggested that this type of non-linear transition state may be general for hydride transfer. ^H H-^^ --B PyThe similarit;^ between the isotope effect observed by Hawthorne J and that observed here in the hydrogen elimination reaction of dimethylajnine borane, suggests that the reactions might be occurring through a similar activated complex. Namely, that the activated complex was not a linear configuration involving WE and BH bonds but that the NH bond was stretched more along the bond axis than was the BH bond, in a manner analogous to that for the OH and BH bonds in I. These data must be interpreted in the light of possible exchange reactions occurring faster than the elimination reaction. Since hydrogen and deuterium gas did not exchange when heated for 24 ho-ors at 100° , this possibility could be discounted. If amine exchange between boranes occurred, there would be no net change in the systems, so this would not cause any difficulties. Deuterium

PAGE 93

81 gas when lieated at 100** for one hour with, dimethyl amine borane did not affect the infrared spectrum of the solid material. Therefore, it may he concluded that no isotope exchange reactions occurred during one hour at 100** , and that the data were not subject to any uncertainties due to exchange. The starting materials were not as pure in their isotopic distribution as would be necessary to determine a precise kinetic isotope effect in the hydrogen elimination reaction. The infrared spectra showed the following: (1) (CH;,)2l>NBH^ contained $ per cent BD bonds relative to the per cent CH bonds, (2) (CH,)2HNBD;, contained 7 per cent ND bonds relative to the per cent CH bonds. (5) (CH;,)2DNBH;, contained 1$ per cent BD bonds relative to the per cent CH bonds, ^ (Prepared from (CH,)2ND2"^ and BH^"), Compounds (1) and (2) were prepared by condensing dimethylamine and diborane together and forming the adduct at -VS**, Compound (3) was prepared from (CH,)2ND2"'' and BH^*" In each instance some exchange must have occurred before the reaction producing the amine borane adduct. ^This compound was not used in any of the hydrogen elimination experiments.

PAGE 94

82 The presence of isotopic impurities introduced during synthesis would account for the Do produced when (CH^)2HNBD^ and (CH,)pHNBH^ were heated and for the Hp produced on heating (CH,)2HNBD:, with (CH,)pDNBD;, and (CHj)2BNBH, with (CE^)^^'^'^^, The data from the one-hour heating experiments did not unequivocally distinguish between a unimolecular and a bimolecular reaction mechanism. But, the analogy to Hawthorne's work with the hydrolysis of pyridine diphenyl"borane (1?) an-d. the large percentages of HD obtained in each case does favor a bimolecular reaction. However, the molecularity of the reaction was resolved by the experiment in which (CH,)2DNBD, and (CH^)2HKBH^ were heated. Heating mixtures of dimethyl amine boranes containing various distributions of hydrogen isotopes for twenty-four hours Dimethylamine borane and H-deuterodimethylamine borane-d,, in 1:1 molar ratio, were heated at 100° for twenty-four hours and the non-condensible products were analyzed in a mass spectrometer. If the hydrogen elimination reaction were unimolecular, then the gaseous product should contain H2, Dp and perhaps some HD due to incomplete deuteration. If the reaction were bimolecular, the gaseous products should be H2, HD and D2 in the ratio of 1:2:1, respectively, neglecting isotope effects.

PAGE 95

83 The experimentally determined ratio of HpiHDcDp was 5,8:^,3sl»0, The data implied that the reaction was bimolecular and that there was an isotope effect favoring H2 eliminations in the reaction. Dimethyl amine borane-d;, and N-deuterodimethyl amine horane, in a 1:1 molar ratio, were heated for twenty-four hours at 100° and the non-condensihle reaction products were analyzed in the mass spectrometer. If the reaction were unimolecular the gaseous product should be HD with the Dp and Ho due to incomplete deuteration. If the reaction were bimolecular, the gaseous products should be Hp, HD and Dp in ratio of 1:2:1 neglecting isotope effects. The gaseous products had a Hp:HD:D2 ratio of 3.1:3.6: 1.0. Since the gaseous products of the reaction contained more Hp and Dp than could be accounted for by incomplete deuteration, the reaction must have been bimolecular. The data were in close agreement with the previous results and indicated that it made no difference whether the deuterium atoms were all in one molecule or partly on the nitrogen in one molecule and partly on the boron in the other molecule. The results in both cases gave the same percentages of Hp, HD and Dp in the gaseous product. The total deuterium percentage in each of the reaction systems, (CH,)2HNBH, (CH^)2DNBD^ and (CH,)2DNBH, ^Por mass spectral analysis of the solid reaction products see Figure 21.

PAGE 96

8^ (CH,)2HNBD,, was almost the same. From the infrared spectra, the system (CH;,)2DNBH, (CH^)pHlNrBD;, contained ^7.^ per cent deuterium bonds and from the pyrolysis products of the dihorane-dg, the system (CH;,)2HNBH, (CH,)2DNBD, contained a minimum of 4-5.^ per cent deuterium bonds. Therefore, it was not surprising that under the same conditions, if the reaction were bimolecular or if an equilibrium reaction were established, for these two systems to give the same ratio of HocHDrDo in the gaseous elimination products. However, the interpretation of these results was made questionable by the possibility that exchange occurred during the reaction between deuterium and the reactants or other reaction products. Some exchange did occur between dimethylamine borane and deuterium gas on heating for twentyfour hours at 100** , the same conditions as in these experiments. In order for the two experiments to have had the same ratio of H2:HD:D2 in the gas phase with exchange occurring between the gases and other compounds in the system, the reaction mixtures in each experiment must have reached the same equilibrium. Equilibrium would have been established according to the following equations for the boron containing reaction products (7). C(CH,)2NBH2]2 ^ 2(CH^)2NBH2 • [12]

PAGE 97

85 5(CH^)2NBH2 ^ C(CH^)2N]2 BH + (CH^)2NB2H5 [153 Burg (6) reported that the aminodiborane in the presence of excess D^ after seventy-four hours at 100® was found to be 65 per cent deuterated. In experiments using Dp gas to deuterate a compound, a large excess of D2 gas and long reaction times were used to be sure an equilibrium was established. In the hydrogen elimination reaction of (CH,)2DNBD, and (CH,)2HNBH,, the per cent reaction was 51.3 per cent. Therefore, of the original 1.59 mmoles of amine borane, there remained 1,09 mmoles unreacted amine borane, with 0,50 mmoles of gas and 0,50 mmoles of aminoborane produced. The 0,50 mmoles of gas contained 0,05 mmoles ©2, 0,22 mmoles HD and 0,20 mmoles H2. If an exchange reaction were occurring between the gaseous hydrogen isotopes and the boron-nitrogen compounds, then the ratio of H2:D2 in the gaseous product should be the same as the ratio of H2:D2 over any other catalytic system. Essentially the boron-nitrogen compounds would be serving as a catalyst for the hydrogen-deuterium exchange reaction, E^ + "Op ^ 2HD. Considering the total moles of gas to be A, then the amount of hydrogen gas available in the (CH,)2HN3H^ (CH,)2DNBD, system would be 0,26A moles and of deuterium would be 0,2^A moles. Using the equilibrium constant^ 5.^8 at ^00°K, calculated by Urey (^2) for the deuterium exchange, the moles of HD were calculated to be 0.20A; and thus, the

PAGE 98

86 concentration at equilibrium of Hp would be 0.08A and of Do was 0,06A. The ratio of Hp:Dp at equilibrium would therefore be 1.33. The observed ratio of HotDp was 5.33, implying that the gaseous products were not at equilibrium for the exchange reaction. If an exchange reaction were occurring between the gaseous hydrogen isotopes and the boron-nitrogen compounds, then a consideration of the difference in zero-point energy between a BH and a BD bond compared to the difference in H^ and Do bonds should give an indication of the thermodynamically favored reaction. For two isotopes in an otherwise identical bond, the difference between the two zero-points of energy is given by E = ^h(-i)-"U ), where V and V refer to bonds containing the lighter and heavier isotopes X and X', respectively. The difference iii'^gg ^^^"^BD ^^^ilculated from only the stretching vibrations at 2350 cm"-*(BH) and 1875 cm""^ (BD) was 1,36 kcal/mole, and the difference in't'trTT andl^j.^. calculated from the frequencies for the fundamental vibration transitions (1) of 4159 cm""'' (HH) and 2990 cm""^ (DD) was 3.3^ kcal/mole. Thus, a comparison of the differences in zero-point energies indicated that the reaction to produce Dp would be thermodynamically the most favored and the gas phase should be enriched in Dp. This was contrary to the observed kinetic isotope effect and the observed ratio

PAGE 99

87 of 41.5 per cent Hp:^?.? per cent HDzll.O per cent ©2 in the gaseous product. Therefore, it does not appear likely that the exchange reaction between H2, HD or D^ and the amine boranes or aminoboranes had reached equilibrium in twenty-four hours at 100°, The exchange reaction, thus, has only a secondary effect on the eliminated gases and the reaction must be bimolecular as was inferred by the experiments on heating the dimethylamine boranes for one hour.

PAGE 100

CONCLUSION The reaction of dimethyl amine horane to eliminate hydrogen was bimolecular and a kinetic isotope effect occurred during the elimination of hydrogen. The data showed that a hydrogen atom was eliminated faster than a deuterium atom and that a BD bond reacted more readily than a ND bond. The ease and speed of deuterium and hydrogen atoms exchange limits many of the possible experiments which might be used to elucidate the behavior of the dimethyl amine borane system on heating, and in studying the kinetic isotope effect in the reactions of the amine boranes. More work needs to be done to establish the conditions and possibly the rates of the deuterium-hydrogen exchange reaction in the BN containing compounds. A case in point being R. B. Davis' report (11) that the rate of deuterium exchange in acidic DpO with trimethyl amine borane was much more rapid than the hydrolysis reaction. In this work, evidence was presented that for N-deuterodimethyl amine borane-d,, the rate of exchange with acidic HoO was approximately equal or only slightly faster than the rate of hydrolysis. This inferred a possible order of magnitude for the rate of the exchange reaction. 88

PAGE 101

89 The unsuccessful attempts to prepare (CH,)2DNBH, or (CH,)2HNBD2 with only the isotopic distribution indicated, suggested that before any experiments can be successfully designed for these compounds, more information about hydrogen-deuterium exchsmge reactions of boron-nitrogen compounds would be necessary. These exchange reactions, or merely the possibility of such, means that any experiments with isotopicsLlly labelled amine boranes must be examined critically to make sure that it is not an exchange reaction which gave the measured results. In this work, the hydrogen elimination reaction of dimethylamine borane has been considered in view of the possibility of an exchange reaction. The evidence strongly inferred a bimolecular reaction.

PAGE 102

SUMMARY To determine the mechanism of the dehydrogenation of dimethylamine borane, (CH^)2HNBH^, the following compounds were prepared, N-deuterodimethyl amine borane-d^, (CH:,)^DNBD,, Ndeuterodimethylamine borane, (CH,)2DNBH, and dimethylamine borane-d^, (CH,)2HKBD^ were prepared by condensing together the appropriate isotopically labelled diborane and dimethylamine. Deuterated diborane was prepared either by displacement from (CH^)^KBD^ with BF^(g) or by reaction of NaBD^ with BF^(g). The (CH^)^NBD, was prepared by an exchange reaction between acidic DoO and (CH-,),KBH,. N-deuterodimethylamine was prepared by hydrolysis of LiN(CH,)p with D2O, Lithium dimethylamide was prepared from n-butyl lithium and dimethylamine. The following mixtures of compounds were heated for twenty-four hours at 100°: (CH,)2DNBD^ and (CH,)2HNBH,; (CH^)2DNBH^ and (CH^)2HimD,. The follov/ing mixtures were heated for one hour at 100°: (CH^)2l>KBH^ and (CH^)2H1TBH^; (GH^)2HNBD, and (CH5)2l»NBD^; (GH^)2HWBD^ and (CHj)2HNBH^; (CHj)2DNBH^ and (CH^)2DNBD,. 90

PAGE 103

91 The results indicated that the dehydrogenation reaction was bimolecular and that a kinetic isotope effect occurred during the reaction, favoring the elimination of hydrogen rather than deuterium. The possibility of an isotopic exchange reaction occurring which could cast doubt on the results was considered and determined not to be significant in this work. The larger isotope effect when deuterium was substituted for hydrogen on the nitrogen atom than on the boron atom indicated the presence of a non-linear activated complex in which the NH bond was more stretched along the bond axis than was the BH bond.

PAGE 104

PART II. SYNTHESIS 0? TRIMETHYLAMINE CHLOROBORANES IITTRODUCTIOIT Alkyl (or aryl) substitution on the boron atom has been done primarily on a boron atom in the three-coordinate state, Grignard reagents, organolithium reagents, and other organic-metallic reagents have been utilized in addition to disproportionation reactions between diborane and alkylboranes, A general review of these methods may be found in either V, Gerrard's Organic Chemistry of Boron (15) or H, C. Brown's Hydrob oration (3), Halogen substitution on the boron atom in the amine boranes has been accomplished using hydrogen halide gases (^0,32), the halogens (52), or the trihaloboranes (32,56). The different halogen reagents reacted to different extents with the amine boranes. Amine exchange can occur between the amine borane molecules and a free amine. The transamination reaction under certain circumstances has been also found (26) to give boronium ions containing two amines attached to the boron and the overall group having a positive charge, Since amine boranes are known to be reducing agents toward aldehydes and ketones (22), as well as toward silver 92

PAGE 105

93 (I) ions, it would be of interest to investigate what happened to the boron compo\nid during these reactions. By proper choice of reactants and conditions, the redox reaction might be used as a means to substitute a group, or atom, into the boron compound. In ar attempt to obtain an x-ray diffraction pattern of trimethylamineborane, mercuric chloride was selected as the internal reference. The two solids were placed in a mortar and ground together. As the solids were groiind, the color of the material turned gray. The color change implied that a reaction had occurred and the physical appearance indicated that mercury metal was being produced. Further study of this reaction showed that the chloroborane adducts were produced. The method proved to be a good preparative route to (CH,),NBHCl2. These observations suggested a simple sequence of events. The two electrons gained by Hg(II) would most probably come from the hydridic BH bond. If the boron lost a hydride ion, the site formerly occupied by this atom could be occupied by a chloride ion from mercuric chloride.* This would be possible if the BN bond remained intact and no other reactions occurred to degrade the molecule. Thus, this reaction sequence would give a simple route to boronsubstituted amine boranes.

PAGE 106

9^ Using, as a working hypothesis, the idea that a metal ion in a high oxidation state can accept electrons from a BH bond to give H"*" (or H), the metal in a lower oxidation state, and a boron atom with an acceptor site which may be occupied by the strongest base in the reaction system, a variety of reactions may be considered. This method would be widely applicable and it could conceivably give many compounds which were previously difficult to prepare. The major practical limitations to this method would be unfavorable solubilities of the salts and difficulties in recovery of the products from the reaction mixture before further reactions could occur. Other oxidizing agents considered were (CH^);,NHC1, SO2CI2, SOCI2, SbCl^, and SbCl^, The order of reactivity of these compounds with trimethyl amine borane was established. Evidence was found that the reactivity of the BH bond as a reducing agent decreased with increasing chloro substitution on the boron atom.

PAGE 107

EXPERIMENTAL Nomenclature The compounds f oimed by the reaction of diborane and an amine have "been named as amine adducts of borane or of substituted boranes. The following is a list of compounds considered in Part (II): trimethyl amine borsuie, (CH,),NBH, ; trimethyl amine monochloroborane, (CH,)2NBH2C1 ; trimethyl amine dichloroborane, (CH,),NBHCl2 ; trime thy 1 amine trichloroborane, (CH,)^NBCl2 • Reagents and purification Trime thyl amine hydrochloride was obtained from Eastman Organic Chemicals and dried in a vacuum desiccator before use, Trime thyl amine trichloroborane was prepared from BCl,(g) and (CH,),N(g) and recrystallized from ethanol before use. Antimony pentachloride was obtained from Allied Chemical Company and used without further purification. Antimony trichloride was the Eisher certified reagent grade. It was sublimed under vacuum at 90° before use* 95

PAGE 108

96 Mercuric chloride was the Fisher certified reagent grade. It was ground to a fine powder in a mortar and was used without further purification, Trimethylamineborane was obtained from Gallery Chemical Company, It was purified by vacuum sublimation from room temperature onto a glass cold finger at -78**, Diethyl ether was the Fisher technical grade. It was dried over CaH2 or "dri-Na" unless otherwise specified. Anhydrous ether used was Fisher reagent chemical. Carbon tetrachloride used in recrystallizations and extractions was the technical grade. It was used without further purification. Ethyl alcohol used in extractions was 95 per cent pure and used without further purification. Anhydrous hydrogen chloride was taJcen directly from a Matheson lecture cylinder. Instruments A 100 ml stainless steel autoclave was used. In certain experiments a removable glass liner was inserted into the autoclave to facilitate the removal of the reaction mixture, A V/heelco Model 293 temperature controller with a Chromel-Alumel thermocouple was used to regulate the temperature of the upright autoclave furnace.

PAGE 109

97 An all-glass vacuum system was used according to the standard technique (21); non-condensible gases were measured hy transferring them into a calibrated volume with a Toepler pump. Melting points were taken in capillary tubes in a stirred oil bath. Solvents were removed either in a Rinco rotating evaporator or by using a water aspirator vacuum to evaporate the solvent. All infrared spectra were taken as KBr pellets on a Beckman IR-10 Spectrophotometer. Elemental analyses were done by Galbraith Laboratories, Inc., I^oxville, Tennessee. Extraction of BN compounds from the reaction mixture ' The first extraction of BN compo;inds was by filtering the reaction mixture. In this work, this solvent was diethyl ether. Then approximately an equal volume of ether was added to the solid residue, the mixture was stirred for five to ten minutes and then filtered again. The ether solutions usually still contained some Hg2Cl2 which came through the filters or which was produced in the filtrate due to a continuing reaction. Even fine porosity sintered glass filters did not prevent some HgpClp contamination in the filtrate.

PAGE 110

98 Mercury metal was then added to the ether solution and the mixture shaken vigorously, precipitating a dense black material, most likely Hg2Cl2 and finely divided mer Clary. The mixture was allowed to sit for two to three hours or longer before refiltering. The clear ether solution, thus obtained, was evaporated using a water aspirator vacuum. A white solid was left behind. The white solid was recrystallized from hot CCl^. The material was dissolved in boiling CCl^, but it was not completely soluble. The hot solution was filtered and the filtrate cooled in a freezer compartment of a refrigerator. The cold CCl^ was filtered to give the BN compounds. Next, the solvent was removed from the filtrate with a rotating evaporator. The material recovered from the filtrate wsls either recrystallized from a smaller portion of CCl^ or was washed with cold CCl^. The latter usually proved to be sufficient to obtain a pure product. The next extraction of the reaction residue was made using excess boiling CCl^.^ The CCl^ extraction solution did not contain as much Hg(II) ion as the ether extraction solution judging from the amount of black precipitate '^At this point, a very strong odor of HCl was noted. The rest of the procedure was the same as for the ether extraction.

PAGE 111

99 produced when mercury was added to the respective solutions. The filtrations had to be done on boiling solutions. It was possible to distinguish the chloro adducts by their infrared spectra. Thus, infrared analysis of the extracted solids could be used to determine if the solids were pure compoiands or to ascertain the identities of components in a mixture. The solubilities of (CH^),NBE^, (CH,),1TBH2C1, (CH:,),NBHCl2, and (CH,),KBC1^ varied only slightly as more chlorine atoms were substituted for hydrogen atoms and decreased with progressive chlorine substitution. The solubility of (CH,),NBH, and (CH^)jNBClj differed widely enough that they could be easily separated. But if (CH,)2NBHpCl and (CH;,),lTBHCl2 were added to the mixture, the solvents used (diethyl ether, CCl^, ethanol or benzene) would not separate the compounds quantitatively. However, a final extraction with hot ethanol would have removed any (CH2),1IBC1, from the reaction residue if it were present. Infrared spectral analysis The infrared spectra of these compounds were sufficiently different that each adduct could be distinguished in a spectrum teiken on a mixture of the adducts. Table 15 lists the infrared absorptions of the compounds. The most characteristic region of the spectrum was the 300-600 cm" •

PAGE 112

100 ?? EH m § o O EH O Ah 03 O W o OJ r-i O o O OJ o o H

PAGE 113

101 It was this region which allowed rapid positive identification of the chloro-substituted adducts. The spectrum of (CH,)^imH2Cl (Figure 22) was characterized mainly by the shift to higher wave numbers of the doublet BH stretching frequency, and the 500-600 cm" region. The spectrum of (CH^);,lTBHGl2 (Figure 23) was characterized mainly by the singlet BH stretching frequency at 2^80 cm"^, two peaks at 1065 and 10^0 cm and the 300-600 cm" region. In identifying a mixture of (CE^)j.WBE-^ and (CH,),NBHGl2, the doublet BH stretch of the (CH^),NBH^ and the two peaks at 1065-10^0 for (CH^),NBHCl2 were the most prominent distinguishing features. A comparison of the 300-600 cm~ region in the infrared spectra of (CH,)^KBH,, (CH,)^NBH2C1, (CH,),NBHCl2 and (CH,),NBCl5 is shown in Figure 24-. Tentative assignments were made based on Taylor's assignments for (CH;,)^NBH^ and (CK,);,NBC1:, (^1), noting that the peaks shift to higher wave numbers as chlorines are substituted onto the boron. Nuclear magnetic resonance spectra The B nuclear magnetic resonance spectra of trimethylamine borane, trimethyl amine monochloroborane, trime thy 1 amine dichloroborane, trimethyl amine trichloroborane, trimethyl amine tribromoboreine, and trimethyl amine

PAGE 114

102 o o o o o

PAGE 115

103 o o o o o o o r-t O I O E -^ o rH U 0) XI E C > o 00 o o o cv o o CM O o o rH cn O o o 0) D, (0 T) 0) u u c I— { I I • cv H O 00 o o o ioue^'^xujsuBJ'; •;uao jaj

PAGE 116

CJ! 104

PAGE 117

105 trifluoroborane were obtained at 19.3 megacycles^ using a Varian DP 60 spectrometer. The results are given in Table 16. The B NMS spectra were obtained to determine if on chloro substitution the electronic density about the boron atom decreased. This was predicted by the inductive effect which was used to explain the change in reactivity of the BH bond. The results confirmed the arguments. The measured chemical shift of (CH,)^NBH of +26.1 referred to BCOCH^), agreed with that reported by Phillips, Miller, and Muetterties (33) of +2^.9. The spectra of the monochloro and dichloro adducts showed splitting due to coupling between the boron and hydrogen atoms.' The monochloro adduct showed the boron peak split into three peaks (21 + 1) and the dichloro adduct showed the boron peak split into two peaks. These spectra may be seen in Figtire 25 and 26. The trend in the chemical shift of iOE^)^WSBr^ > (CH^),NBP, > (CH,),NBC1^ agreed with the trend found for the tri ethyl amine, pyridine, and (CcHc)2C0 adducts of the tribal oboranes by Gates, McLaughlan and Mooney (1^).' The chemical shift for (CH:,);,NBP, was greater than that for (CH,)2NBC1^ which implied a greater electronic density on the boron atom for the trifluoro adduct than the

PAGE 118

106 TABLE 16 ai B -^ NUCLEAR MAGNETIC RESONANCE SPECTRAL RESULTS Compound

PAGE 119

107 c^

PAGE 120

108 r^

PAGE 121

109 trichloro adduct. This was contrary to what wo\ild be expected from the consideration of electronegativities. However, the trend in the chemical shifts may be e3q)lained by a back donation of the fluorines' electron density onto the boron atom by a p^-p^ type system. This effect has been found to occur in other boron-fluorine compounds, and used to explain the behavior of the compounds. Reaction of (CH,)^imH, and HSCI2 General procedure for reaction in ether . — Trimethylamine borane (13.7 mmoles) was dissolved in 200 ml of ether and stirred magnetically. Mercuric chloride (64-, 8 mmoles) was added to the d;her solution and the stirring continued for the desired time. The amounts of reactants were those found convenient to use. A reaction temperature of 0** to J>^°C did not change the products which were produced. A reaction time of thirty minutes to several hours did not appear to control which products were recovered. This reaction procedure did not give complete reaction considering the reaction stoichiometry. In all cases a mixtiore of amine chloroboranes was recovered, and the HgClo did not react completely to give Hg2Cl2 which was identified by an x-ray diffraction pattern, Even large excesses of HgCl2 would not drive the reaction to completion.

PAGE 122

110 Slow addition of tlie HgClo via a Soxhlet extractor was found to give a complete reaction to (CH,)^NBHCl2 whenever four moles of HgClo per one mole (CH;,),NBH2 were used. When two moles HgClp per one mole (CH;z);rNBH, were used, in attempting to prepare (CH;z);,I)rBH2Cl, the recovered product contained a small amount of (CH^),NBHCl2 in addition to the desired product. The method for the extraction of boron-containing compounds from this reaction system was the same as that described in the reaction of (CH^),NBH^ and HgClp at ether reflux. Benzene could be used as an extracting solvent, but an ether extraction followed by a hot COl^ extraction was found to be sufficient, provided excess solvent was used in each case. The results of these reactions were listed in Table 17. Reaction of ( CH^, ) ;,KBH;, and HgClo on a large scale Trimethyl amine borane (0.137 mole) was placed in a two liter, three-necked round bottom flask containing one liter of diethyl ether. One neck of the flask contained a Soxhlet extractor with a filter thimble containing mercuric chloride (0.555 moles), the center neck of the flask contained a Teflon wing-stirrer, and the last neck contained a glass plug.

PAGE 123

Ill OJ w ^ cd m •P o rJ •d o Ph H •H •H EH O fH Ctf t EH C\J 05 rf\ r-l KN O W O rH rCN OJ O WWW KN rOi rfN H OJ O rH cvj o @ g O OJ w ro> OJ rH O W pq O OJ W io> OJ H O w o rov OJ w w o OJ O w w C\J tO\ ro> PCS KN rO\ t<\ rA hf\ rA rOi rA W W W W W hh o o o o o o rA rr\ w w o o fA rov rCN W WW O O O W W o o OJ o O o

PAGE 124

112 •p o o IN EH w +3 o o iH Ph •H •H EH Fh :3 -p a o & 0) EH (D /--N o w o o CO m O LA rH • U 0) -P Jh ^ o -P o 0) ^ O w pq rOt CM OJ H OJ rA O rH rH f\JO rH o w m o •-N /^^ /'-^ ^^ /'-> r>CN rCN ros fO( rCN • ^1 r r^ "T*^ ^T* ^^ HM HH HM KM (-M O O O O O v.^ v^ v^ ^^ v--' LPs rov CO CD — P CO rt ^ a o O LA CO

PAGE 125

113 As the ether refluxed through the Soxhlet, it dissolved the HgClo. This gave a relatively slow addition of HgCl2 to the (CH^)^imH, solution. The reaction was followed by infrared analysis of the reaction mixture. Aliquots of the ether solution were removed, the solvent .removed from the aliquot portion, and an infrared spectrum taken of the solid residue. The infrared spectrum etfter the reaction had refluxed for eighteen hours and held at room temperature for twelve hoiirs was that of pure (CH,):2NBHCl2. The reaction was continued for ninety-five hours, but no change was observed in the infrared spectra. The BN compounds were extracted from the reaction mixture as described previously. The procedure gave a white product, recrystallized from CCl^, in 80.5 per cent yield. A commercial analysis gave: Per Cent C H N CI B ' Found 2^.68 6.89 9.59 ^7.77 7.95 Calculated for (CHj)jNBHCl2 25.^1 7.11 9.88 50.00 7.62 This material softened at 1^5° and was a liquid at 15^°. In a sealed tube it melted at 144 to 149<'C.

PAGE 126

114 The infrared spectrum agreed with the spectra obr tained for (CH;,),NBHCl2 when prepared by other methods. The spectra showed a singlet BH stretching frequency at 2480 cm" , two intense peaks at 1065 and 1040 cm , and less intense peaks at 500, 440, 345, and 315 cm as the distinguishing peaks. Reaction of (CH^)^KBH2C1 and HgCl2 Trimethyl amine monochloroborane (3.6 mmoles) was placed in a two-neck round bottom flask. Mercuric chloride (11,4 mmoles) was placed in a tipping tube connected to the reaction flask in one neck; in the other neck of the flask, there was a stopcock adapter for connection to the vacuum system. The reaction flask x>ras cooled to 0° and evacuated, „ Ether (75 nil) was distilled from CaHo into the reaction flask at liquid nitrogen temperature; the reaction flask was warmed to 0** and stirred magnetically. The HgClo was tipped into the ether solution. After two hours at 0**, the reaction was cooled to liquid nitrogen temperature and less than 0,5 nun pressure was measured in the vacuum system. The reaction system was then warmed to room temperature for eighteen hours, and again the reaction was cooled but no non-condensible gas pressure was measured. Commercial "anhydrous" ether was distilled into the reaction system to determine if trace amounts of water might

PAGE 127

115 cause a reaction. The reaction system was then warmed to 0° for two hours, and then to room temperatiire for fourteen hours with no visible changes occurring. The solvent and other volatile reaction products were transferred from the reaction flask, and the reaction flask contfidning the solid reaction residue was then set aside at room temperature with the stopcock closed. After thirteen months and twenty-one days, the solid reaction residue was heated with CCl^. The hot CCl^ solution was filtered. The filtrate was reacted with mercury metal, heated and again filtered hot. A product was precipitated on cooling the CCl^ solution. This product gave 26,7 per cent pure yield of (CH^),i^HCl2. This product melted at 150-15^°, Its infrared spectrum agreed with that of (CH;,);,lNrBHCl2. There was no spectral evidence for (CH^),NBC1^ in this product. Attempted reaction of (CH;,);,NBHCl2 with HgCl2 Trimethylamine dichloroborane (2.75 mmoles) was dissolved in 100 ml of diethyl ether, and cooled in an ice hath to 0**. Mercuric chloride (5«59 mmoles) was added to the solution and the reaction mixture stirred magnetically using a Teflon coated stirring bar. The reaction flask was closed to the atmosphere by a glass stopper. After one hour at 0°, the HgCl2 had dissolved and the reaction system was

PAGE 128

116 homogeneous. The reaction, mixtirre was then warmed to room temperature. After the mixture stood for twenty-four hours, mercury metal was added to convert the Hg(II) to Hg(I), The mixture was filtered and the residue washed with excess ether. The ether was evaporated from the filtrate and a solid product remained. The recovered product melted at 1^6-1^8°, and had an infrared spectrum identical to that of (CH,),NBHCl2. Ko spectral evidence for (CH^);,NBC1, was observed. The starting amine borane was recovered in 93 per cent yield. The same two reactants were mixed in a ratio of one mole of (CH^)^I:TBHCl2 to two moles of HgClp, then slurried in ether. The ether was evaporated and the solid material remained at room temperature for twenty-six days. The (CH,)3,NBHCl2 was recovered by the previous procedxire in 85,^ per cent yield. The infrared spectrum agreed with that of the starting material,' ITo spectral evidence of (CH;j),KBCl^ was observed in the recovered material. Relative rates of reaction of (CH;,),1TBH, with HgClo and HCl in ether at 0° Trimethylamine borane (1.59 mmoles) was sublimed into a tube which could later be turned so the solution would flow into the reaction flask, and could be closed

PAGE 129

117 from the reaction flask by a stopcock. The trimethyl amine o horane was dissolved in 15 ml of anhydrous ether. The tipping tube was connected to a three-neck 300 ml round bottom flask, whose center neck had a stopcock adapter for connection to the vacuum system and the other neck was plugged. Mercuric chloride (5.20 mmoles) was placed in the reaction flask and 50 ml of anhydrous ether added. The entire reaction set-up was condensed and evacuated, and the solutions degassed by warming to room temperature, condensing and evacuating the set-up. Hydrogen chloride gas was condensed into the vacuum system and a measured amount (1.60 mmoles) was condensed into the reaction flask. The reaction flask was warmed to room temperature and stirred magnetically. The HgClo would not dissolve completely. There was a slight residue of finely powdered HgClp suspensed in the ether. The reaction flask was then cooled to 0** in an ice bath and the stirring continued. The tube containing the (CH,),NBH^ was washed into the reaction flask by condensing the ether vapor in the tube with dry ice and letting this liquid flow back into the reaction flask. By using this procedure all the iCE^)^l<[BE^ was transferred into the reaction flask within a few minutes. The reaction flask was stirred at C for one hoxir. The products were then condensed and the non-condensible

PAGE 130

118 gas v;as transferred, using a Toepler piimp, into a cali"brated bulb. The amount of non-condensible gas was 0.56 mmoles, i This implied that only 0.'35 mmoles of the 1.59 (20 per cent) of (CH,);,KBH;, reacted with the HCl gas. In other experiments on this scale and under these conditions HgClo was found to react with (CH,)^NBH, to give (CH,),NBH2C1 and some (CH,);2lN[BHCl2. It was concluded that in the reaction time the (GH;,)^I:^H^ would be completely reacted and that the reaction with the HgClp was faster than the reaction with HCl by at least a factor of four. Reaction of (CH^)^^^;, and HgClo in the autoclave General procedure . — Trimethyl amine borane and mercuric chloride were placed in the stainless steel autoclave which was immediately sealed. After a set period of time at the desired temperature, the gaseous products were released through a liquid 1^2 tirap, a mercury bubbler and collected in a water-filled inverted graduated cylinder. In this manner, the amoiint of non-condensible gas was determined. The solid materials were removed from the autoclave and extracted with hot carbon tetrachloride. The compounds extracted in the hot carbon tetrachloride were analyzed by obtaining infrared spectra. The results of the experiments are given in Table 17.

PAGE 131

119 In each reaction, there was unreacted HgClp and mercury metal present in the reaction mixture. Qualitative tests for other reaction products showed that the noncondensible product contained no infrared active material; a strong odor of HCl(s) was noticed in the workup of the compounds. Thus, the following equation may be written: 2(CH,)^NBH, + $HgCl2 "* 2(CH^)^NBC1^ + 5Hg + 6HC1[+H2]^ Cl^] It may now be concluded that HgClo v;ill react with (CH,),NBH, at 100° to produce the completely chlorinated trime thy 1 amine borane adduct, and that the Hg(II) was reduced to Hg(o). Reaction of (CH;,)^NBH^ and HgClp in presence of acetic acid in benzene Trime thyl amine borane (1^.0 mmoles) was dissolved in 100 ml benzene (dried over "dri-Ka") and 0.8 ml of glacial acetic acid (1^,0 mmoles) pipetted into the solution. A slurry of HgCl2 (28.2 mmoles) in benzene was added to the amine borane acetic acid solution. Judging from the appearance of the white precipitate, the reaction was slower than in ether or water solvents. After two hours and ten minutes at room temperature . the reaction mixture was filtered through excess anhydrous K2CO, and the benzene removed from the filtrate using a rotating evaporator. The (CH5)5NBH3. ^The hydrogen was produced by the reaction of HCl and

PAGE 132

120 infrared spectrum of the material recovered from tlie benzene stiowed that the compoxind was primarily (,0E^) ^^WSEpGl with a small amount of (CH;,)^NBHCl2. Considering the product to be pure (CH;,);,KBH2C1, the yield was 82 per cent. The reaction was repeated using a larger molar ratio of acetic acid to amine borane. Trimethyl amine borane (1^.1 mmoles) and 8 ml of glacial acetic acid (1^0,0 mmoles) v;ere dissolved in benzene and solid HgClp (28.6 mmoles) was added. After two hours and ten minutes at room temperature a product was recovered from the benzene using the procedure in the previous experiment. The infrared spectrum of the material showed it to be mainly (CH;,),NBH2C1 with a small amount of (CH^)^NBHCl2. The material was recovered in 73*0 per cent yield of pure (CH^)^NBH2C1. Thus, the acetic acid did not appear to affect the reaction under these conditions. Reaction of (CH^)2NBH^ and HgGl2 in water and in potassium chloride solutions General procedure . — Trimethyl amine borane dissolved in water (or KCl solution) was added to a water (or KCl) solution of mercuric chloride and the solution stirred magnetically for the desired period of time. A product was then extracted by adding an equal volxime of benzene to the reaction solution and the mixture shaken vigorously. The water-benzene phases were separated using a separatory

PAGE 133

121 fxinnel, and the benzene layer dried over anhydrous potassixm carbonate. After removal of the benzene in a rotating evaporator, an infrared spectrum was taken of the residue. The results of the experiments are given in Table 18. Attempts to isolate other BN containing products were not successful, No other 3N compounds were evidently produced during the reaction in water. The yield of monochloro adduct would imply that it was produced in a side reaction. An x-ray diffraction pattern of the water— insoluble residue from the reaction was similar to that of mercurous chloride, the only difference being a broad peak at 26" which was not present in the pattern of Hg2Cl2, The change in pH with time for these reactions v/as examined. Reaction of (CH^)^ITBH;, and EgClo in water — the change in pH with time The pH was followed by a Model SR Sargent Recorder equipped with a resistance-matching adapter (S-72172 Sargent pH adapter) for measurement with Beckman glass and calomel electrodes. The apparatus was calibrated using standard buffer solutions to have a pH range of to 12,5 over a one millivolt range. The electrodes were placed in a mercuric chloride solution, the chart speed set at one inch per minute, and then a trime thy 1 amine borane solution was added to the cell, A white precipitate formed and the pH dropped immediately.

PAGE 134

122 TABLE 18 EEACTIOIT OF (CH^) NBH^ AM) EgGl^^ IN VATEH AED KCl SOLUTIONS. (CH^)^NBH2 HgClo Solvent Reaction (CH,)-,NBHoCl mmoles mmoles Per Cent Yield 15.8

PAGE 135

123 In the reaction of (CH-,)^NBK^ (5.02 mmoles) and HgClp (30.1 mmoles), the pH changed from 3.15 in the HgClo solution to 1.10 within six seconds after the addition of the amine borane solution. After twenty-five seconds the pH was 1.05 and did not change further in thirty minutes. Considering the two possible reactions which could produce a strong acid in the solution and calculating the pH of the solution for complete reaction, the resulting pH, in a volume of 275 nil, of the experiment (1,05) agreed more closely with the reaction to produce boric acid (1,0^), The possible reactions are: (CH,)jNBH^ + 6HgCl2 -* (CH^)^K3C1^ + 3Hg2Cl2 + 3HC1 pH = 1,26 (CHj)jlTBH^ + 6HgCl2 (CHj)^mCl + B(OK), + 3Hg2Cl2 + 5HC1 pH = 1.04 Another reaction was carried out using the same procedure but a different stoichiometry. Trimethyl amine borane (O.5I8 mmoles) and mercuric chloride (3.00 mmoles) were reacted and the resulting solution was titrated with O.IM NaOH solution. In this reaction, the pH changed from 4-,90 in the HgCl2 solution to 2,55 within six seconds after the addition of the amineborane. After two minutes the pH was 1.75 and remained constant for five minutes. (See Figure 27.) The solution was then titrated to a pH of 9.0 with O.IM UaOH solution and the pH remained constant for 13 minutes,'

PAGE 136

124 c^ VO vn ^ rN 0) -p c B e H en C\J u t) +> c •H CM rH O C CO CQ c^ o «H o c o •H -+J o -p bO C •H v. 3 •a o 6 X -(-> Q> hfl C cd X o I I • CM •rl O O NO O o o o o o o o CM o o HCJ

PAGE 137

125 The end point of the titration at pH of 5.0^ showed that 2.5^ mmoles of H,0"^ had been produced by the reaction. If the amine borane-merciiric chloride reaction had given (CH,),N3Cl;j then 1.55 mmoles of H^^O"*" would have been present in the solution; if the reaction had gone to boric acid then 2,59 mmoles of H^O"^ v/ould have been expected. Again, the data implied that the end result of the amine boranemercuric chloride reaction in water was boric acid and that any trimethyl amine monochloroborane recovered from the reaction must have been produced in a side reaction. This point was considered further in the discussion of the experimental results. Reaction of (CH^)^KBH2C1 and (CH^)^imHCl2 with HgCl2 in water — the change in pH xvith time The apparatus to measure the pH was the same as that used in the trimethyl amine borane experiments, Trimethylamine monochloroborane (1.57 mmoles) was dissolved in 100 ml water using ethanol (2 ml) to wet the amine borane, and the electrodes were placed in the solution. The pH of the solution decreased slightly before the addition of mercuric chloride-water solution. On addition of the mercuric chloride, a dark precipitate formed but it turned white within a few minutes. The rate of change in pH was much At this end point only the strong acid was titrated.

PAGE 138

126 slower tlian in the trimethyl amine borane reaction as may be seen in Figure 28. The same procedure was followed in the reaction between trimethyl amine dichloroborane (2,5^ mmoles) and mercuric chloride (5.17 mmoles). Trimethyl amine dichloroborane would not dissolve in 100 ml water or in a 20:1 water-ethanol mixture. The electrodes were placed in the trimethyl amine dichloroborane-water-ethanol slurry and the pH decreased steadily. The addition of the mercuric chloride solution did not affect the rate of change of pH in the solution, but the pH continued to decrease steadily (see Figure 28). No visual evidence of a Hg2Cl2 precipitate was noted in this reaction as had been in the previous reactions. The data implied that the amine dichloroborane was reacting with the water and that mercToric chloride did not significantly affect this reaction. Reaction of (GE-;r)yl^E^ and excess HCl(g) Trimethyl amine borane (13.8 mmoles) was placed in a glass lined autoclave. The autoclave was closed, attached to a vacuum system, immersed in liquid nitrogen and evacuated, Hydrogen chloride was condensed into the vacuum system and a measured amount (37.1 mmoles) transferred into the autoclave. The autoclave was warmed to room temperature and then heated at 100+5° for twenty-four hours.

PAGE 139

127 5.00 ^.00 3.00 2.00 10 20 30 Time, minutes Fig. 28. — Change in pH with time during the reaction of (CH3)^NBH2C1 and (CH3)3NBHCl2 with HgCljin water. ( • ) (CH3)3NBH2C1 and ( o ) (CH^)3NBHCl2 .

PAGE 140

128 The autoclave, at room temperature, was opened and tlie volatile portion was allov/ed to pass througli a liquid No trap, a mercury bulibler, and into a waterfilled inverted graduated cylinder. A non-condensible fraction (25.^ mmoles) was collected, A solid product was removed from the glass liner,' This material was sublimed from room temperature to -78°, It was then recrystallized from hot CCl^. The pure yield was on the order of 45 to 50 per cent (CH,),]TOHCl2. Analysis of the product showed two moles of chlorine per mole of boron. Per Cent B Cl~ Found 7.92 48.4, 47,6 Calcd, 7.62 50,00 The melting point of this material was 147-148°. The infrared spectrum, identical to that of (CH,),NBHCl2» showed a singlet BH stretching frequency at 2480 cm" , two intense peaks at 1065 and 1040 cm" , and less intense peaks at 500, 440, 545 » and 315 cm" as the dist ingui shing pe aks , Reaction of (CH^)^ITBH;, and ECl(g) in benzene Trimethyl amine borane (149,0 mmoles) was dissolved in $00 ml of benzene, and a mixture of trimethyl amine borane and the monochloro adduct (9.5236 g) was added to the solu-

PAGE 141

129 tion. Hydrogen cliloride was bubbled slowly through, a mercury bubbler and into the amine borane-benzene solution at room temperature.. After six hours, the benzene solution was yellowish in color and bubbling in the solution had ceased. The solution was then filtered and the solvent partially removed from the filtrate by distillation at atmospheric pressure until approximately 50 ml of benzene remained. This was removed using the rotating evaporator. The remaining solid material was slightly yellowish in color, had a mp of 84—85'*,^ and the infrared spectrum showed only (GE-r) ^1^E2C1 to be present. The exact yield of the reaction could not be determined since the amount of amine borane was not known. The weight gained (5.9^^ g) by the solid material during the reaction indicated that the reaction must have been almost a quantitative one. This is a recommended method for the preparation of trimethyl amine monochloroborane. Reaction of (CH2)^NBH-t and concentrated HGl(aQ) in water 2-2 2 _A trimethyl amine borane-saturated water solution (10 ml) was added to concentrated hydrochloric acid (2 ml) and shaken vigorously for five minutes. The solution was then neutralized by adding 5 ml of 6M NaOH solution. A literature value (52) is 85*". The water layer turned pink on the addition of phenolphthal e in ,

PAGE 142

130 product was extracted into benzene by twice adding approximately 25 ml of benzene, shaking and then separating the benzene-water phases. The benzene phase was dried over anhydrous KoCO;,, filtered and the solvent removed,' The infrsired spectrum of the recovered material showed a mixture of (CH^)^EBH, and iCEy^) ^mE^Gl , The reaction was repeated using 15 lal of saturated amine borane solution and 3 nil of concentrated HCl(aq), After neutralization the solution was cooled to room temperatiire before extraction v/ith benzene. The infrared spectrum of the recovered product shox^red (CH;z);,iroH2Cl but it also contained peaks v;hich v/ere similar to those in a spectrum of boric acid. Reaction of (CH^)-,KBH^ and concentrated HCl(aq) in benzene 2__2 2 I Trimethyl amine borane (1^.7 mmoles) was dissolved in 50 ml of benzene in a separatory funnel and 2 ml of concentrated hydrochloric acid were added to the solution. After shaking vigorously for ten minutes, the water layer was removed; and the benzene solution was set aside for one hour. An infrared spectrum on the material remaining after removal of the benzene showed the product to be a mixture of (CHj)^NBH^ and (CH^)^NBH2C1, The experiment was repeated on a larger scale but the results were the same as in the previous experiment.

PAGE 143

131 Tlie reaction product was recovered in a yield of 40 per cent if it had been pure (CH^)3NBH2C1. Reaction of (CH:.)-,RBH^ and concentrated HCl(aq) in carbon ^ ? ? ? te tracM oriae Trimetliyl amine borane (1^.2 mmoles) was dissolved in 50 ml of carbon tetrachloride, 5 nil of concentrated hydrochloric acid were added, and after ten minutes, the water layer was removed. Since there had been no visual evidence for a reaction occurring in the CGl^ layer, e.g., bubbling, the solution was set aside open to the atmosphere. After three weeks, it v;as noticed that the CCl^ had evaporated and a crystalline material was in the reaction flask. The infrared spectrum of this material shov;ed it to contain only ( CH;r ) ^zNBHoCl . Therefore, a reaction must have occurred in the CGl^ phase. Reaction of (CH^)^KBH^, (CH^)^KBH2C1 and (CH^)^NBHCl2 with ^t^^)^^A(^L ^-^ ^-^ General procedure . — The amine borane and the amine hydrochloride were placed in a stainless steel autoclave and heated for a set period of time at the desired temperature. After the autoclave cooled to room temperature, the gaseous products were released. In some experiments, the water-insoluble gaseous products were measured but since in all but one reaction the product was a mixture of compoxmds

PAGE 144

152 tiie volume of gas had little meaning. The trimethyl amine produced in the reaction could not be removed completely because a small quantity of it tended to remain adsorbed in the reaction mixtixre. A product was extracted from the solid reaction residue with boiling carbon tetrachloride, the solution filtered hot, and the carbon tetrachloride removed from the filtrate^ using a rotating evaporator. Infrared spectra were obtained on the solid residues. The results of the experiments aregiven in Table 19. Reaction number I in Table 19 produced piire (CH,)^NBHGl2. The compound had a melting point of 145-145°. A commercial analysis of this compound gave the following: Per Cent C H N CI B Found 25.56 7.06 9.85 50.09 7.55 Calcd, for CCHj)5N3HCl2 25.41 7.11 9.88 50.'00 7.62 The yield of product from this reaction was 40 per cent. These reactions gave poor yields and in most cases the product contained a hygroscopic material, probably some of the unreacted amine hydrochloride. Occasionally, the solid reaction mixture removed from the autoclave was blue-green in color. The colored material was water soluble and on addition of dimethylglyoxime a red

PAGE 145

133 o o o pq O W o o 0^ H <3j EH H EH n g /"^ o o O H EH O < CO CO -p.H O W fd rH o d rH W •H f>s O O k! O Pi WEh CO Pi u o W •P »• o o «5 a 0) -H WEH H O o CO r-\ O o o o o O LA LA lA [N CO r-< r^ i-{ LA M OJ (\J rH rH H O O O OJ H W W t<^ rA rOi /--N ^-^ ^N rA rA rA W K W o o o H AJ rH OJ O t-H O H OJ O OJ O HI w tn w rA rA rA rA /-N ^-^ /-^ /— S fOv N"N rA rA W M ffi H! o o o o •p LA OJ rA « % • fH t> iH rH CO -p O o ft CO d o o ooooooooo OlAlAOlAiAOlAlA lAC^COlAO-COlACJDOO r-\rHr^r-\t-ii-{r-\r-ir-\ LA l>LA • « • lA LA rA d.^h LA OJ O O LA LA CM • CM ON o OJ 00 o lA CO rH OJ H o w ^-\ tA w o <1> h 3 o lA CO CO P iH O o Pi ft CO IS o CO OJ a ta o to 0) H o tA

PAGE 146

13^ o d -^ •H !>5 o •H P o O ft CO o W •H O O O o o r-l o Eh r^ o CM tx! o H O tx! O rCi o o « o u •H a (D •H O U u o OH o nd nd o o O lA LTN OH rH o O •LTN Ui H -P O •ri o Pi KN LA « • O OJ OJ H CO o CO CC © CQ 05 OJ rH rA w o rA tA rf\ rA O O Hpq I 05 O-H CS5 >i-p ^ ct5 05 a-P'd •^ f1 GOO fH-H O ft ft OrcJ O-H PI o P! o H CO -d O o o o O CA lA Lr\ ocxD r-\ CM rA rA 00 OJ OJ o « 1>H •P CD CO CD fH ft g o t o o -P o -p o H d a CQ d rH
PAGE 147

155 precipitate formed. This implied that Ni(II) was present in the reaction mixtxire. The only way this coiild he accounted for was by some side reaction between a reactant or product and the walls of the stainless steel autoclave. Reaction of (CH,),N3H2C1 with SbClc Antimony pentachloride (1.56 mmoles) v;as pipetted into a 50 ml round bottom flask containing trimethyl amine monochloroborane (1.03 mmoles) in a Dri-Lab controlled atmosphere box. An immediate violent evolution of gas occurred. The flask was fitted with a vacuum system adapter, attached to the system and evacuated. All volatile materials were transferred from the flask. A Nujol mull was prepared on the non-volatile reaction residue and the infrared spectrum obtained using NaCl plates. The spectra did not show any 3H stretching vibrations in the 2300-2500 cm" region, but all the peaks in the spectra agreed with either the spectrum of Nujol or the spectra of (CH,)^NBC1,» The reaction was rerun under better controlled conditions. Antimony pentachloride (5.91 mmoles) was distilled in the vacuum system into a flask containing trimethyl amine monochloroborane (2.10 mmoles) and the reaction flask was warmed to -78°. Since there was no change of the press\ire in the system SLfter several minutes, the flask was warmed

PAGE 148

156 to 0", An immediate increase in pressure was noticed wlien the flask was first warmed, but at 0° tlie pressure continued to increase slowly. On warming to room temperature, a faster increase in pressure occurred and after forty-five minutes the flask was immersed in a liquid N^ bath; a small amount of non-condensible gas was noted. The condensible gaseous product was distilled from the reaction flask into a flask containing ^0 ml of O.IM NaOH solution and warmed to room temperature, A back titration of the excess base with 0,1M standard HCl solution indicated that the gas contained 2,85 meQ. of acid, but the reaction product also contained a material which reacted with the phenolphthalein indicator which made the end point of this titration somewhat in doubt. The solid reaction product was washed with 60 ml of diethyl ether to remove any antimony chlorides. Hot ethanol was used to extract a product from the ether insoluble material. , The ethanol extracted product contained (0H2):zNBCl;, according to its infrared spectrum. The ether from a portion of the wash solution was distilled in the vacuum system and a yellowish residue remained. The residue txirned a dark grey-black slowly over a period of two days. The other portion of the ether wash solution remained visually unchanged in the same period of time. The dark residue was washed with ether and hot ethanol.

PAGE 149

157 A subsequent qualitative test for antimony (III) (18) on the remaining grey pov;der sifter reaction with nitric acid proved to be positive. The ether in the remaining wash solution was removed and the same color change was noted. After one week, two large crystals were noticed in the top of the flask. The sealed tube melting pointB of these crystals and of sublimed antimony trichloride were both 73-7^°. The reported (9) melting point is 73. ^4-°. From the reaction of (CH^);,EBH2C1 and SbClc it may be concluded that the reaction produced an acidic gas, (CH,)^II3C1;,, and a grey powdery material which did contain antimony (or antimony III) and that the reaction with SbCl^ did not occur in ether solution but did occur slowly at room temperature when the ether was removed. Reaction of (CH^)^NBH^, (CH^)^KBH2C1, (CH^)^imHCl2 and (Cli^)^N3Cl, with SbCl^ 9 2 2 Exploratory reactions betv/een the amine boranes and antimony trichloride were done in the Dri-Lab controlled atmosphere box. The follovn.ng observations were noted. SbCl, and (CH,)^1JBH,: The solids turned black upon contact and after two days the flask was coated with a metallic mirror. SbCl^ and (CH^)^iroH2Cl: The solids reacted immediately to give a black material and liquid phase from which

PAGE 150

158 gas evolved. SbCl^ and (CH^)^NBHCl2: On first contact, tlie solids changed to a liquid phase which after several minutes began to have d£Lrk specks of material in it which increased with time. The infrared spectrum of the liquid phase in Nujol on NaCl plates showed only the starting material, (CH^);,NBHCl2 and perhaps some (CH^)^NBCl,. The relative intensity of the BH stretching peak at 2^80 cm" to the rest of the spectrum was less than it would have been if the material had been pure (CH;2)^N3HCl2. SbCl;, and (CH:z)-,lTBCl,: No evidence was noted for a reaction. When heated in an evacuated flask the ShCl^r sublimed to cooler portion of the flask aind no reaction occurred. Reaction of (CH^)^NBH;, with SOCI2 Thionyl chloride^ (6,96 mmoles) was distilled into a 50 ml round bottom flask containing trime thyl amine borane (0,51 mmoles). The reaction was warmed to -78° and after twenty minutes only a 1.5 mm increase in pressure occurred. On warming to -6$", after thirty minutes the pressure was 5.5 mci and after thirty minutes at -23** the pressure was 70,5 IQ21. The reaction v:as then warmed to 0° and after three hours the pressure v;as 125.0 mm. The reaction was condensed after an additional one ho\ir and twenty minutes at room temperature (pressure 176.0 mm) and no non-condens ible gas v;as present. ^01 impurity removed by distilling from -78** into -196° trap.

PAGE 151

139 All tlie volatile material was transferred from the reaction flask^ leaving a yellow residue, Tlie infrared spectrum of tlie volatile reaction products implied mainly HCl (19). The material was fractionated "by distilling from a -78° trap through a -119" trap into a -195° trap and infrared spectra obtained on each fraction. In the -195° trap only HCl (1.21 mmoles) was present, in the -119" trap mainly SOp (35) (0,331 mmoles) with a small amount of SOClo was present, and in the -78° trap v;as SOClo (25) with a slight imptirity of SO^, The flask increased in weight by 0.0985 g due to the reaction. A product was extracted from the yellow residue with boiling CCl^ and the infrared spectrxim of the material showed only (CH^)^I)[BHCl2 to be present. An infrared spectrum on the CCl^ insoluble yellow residue indicated that (CH;,)^NHC1 was present as the principal infrared active material. There were peaks at 1200 cm" , 925 cm" and 550 cm" which could not be attributed to the amine hydrochloride. These could be due to SO or SCI bonds in another non-volatile and CCl ^-insoluble reaction product. Reaction of (CH^);,ieH, and SO2CI2 Trimethyl amine borane (0.53 mmoles) was sublimed into a flask containing sulfuryl chloride (5.20 mmoles). The flask was warmed from -195° to -53" for thirty minutes and a

PAGE 152

140 slow increase in the pressure occurred; on warming to -25°, the pressure initisLLly increased rapidly and then increased steadily. After one hour the flask was warmed to 0° sind no further increase in pressure occurred. On recondensing the flask in liquid Np, a small amount of non-condensible gas was present in the system. All the volatile material was distilled from the flask, leaving a white residue. A gas phase infrared spectrum of the volatile material showed the presence of sulfuryl chloride (25), sulfur dioxide (35) » and hydrogen chloride (19). An attempt to separate the HCl and SO2 from the SO2CI2 by distilling a fraction from -119° into a -196*' gave quantitative separation of HCl from SO2CI2 ^md S02» and the infrared spectra indicated only HCl in the most volatile fraction. The gas phase spectra on the material remaining in the -119° trap was primarily that of SO2 since SOo has a higher vapor pressure than SO2CI2. The increase in weight of the reaction flask, if the product were pure (CHv):zNBCl^, implied a 97 per cent yield.' The infrared spectrum of this material indeed was primarily that of (CE:,)-rIiBCl-, but also indicated a trace of (CH:z),KBHGl2 and some evidence for the (CH^);zNH'^ ion. Reaction of (CH:2):rNBH, and ZnCl2 in glacial acetic acid Trimethyl amine borane (14.0 mmoles) was added to 50 a ml of glacial acetic acid containing solid zinc chloride. ^On standing the zinc chloride dissolved in acetic acid.

PAGE 153

1^1 The solution evolved gas vigorously. After ten minutes, approximately 75 ml of benzene was added to the solution and the mixture shaken before separating the two phases. The benzene phase was inadvertently lost. The acetic acid phase was neutralized with anhydrous KpGO, and again extracted with benzene. The mixture was filtered and the benzene removed from the filtrate in a rotating evaporator leaving a clear film on the walls of the flask. This material was slurried with CCl^ and stored in the freezer compartment of the refrigerator overnight. The next morning needle-shaped crystals were filtered from the cold OGln solution. The infrared spectrum (Figure 29) of this material contained a BH2 stretching vibration peak at 2350-2^10 cm and an intense absorption at 1690 cm , possibly a carbonyl absorption. Qualitative test showed the compound contained boron, BH bonds, trimethyl amine and no chloride. The material melted sharply at 58-60® and was soluble in carbon tetrachloride, benzene and water, No change in the melting point occurred on recovering the compound from a water solution. Not enough material was recovered for a quantitative elemental analysis, but the data suggested that a possible II formula of the material might be (GH2)2N3H2(0-C-CH,). However, the experiment was not reproducible and the same

PAGE 154

142 o 00 o o o 80UBq.q.xiasuBa^ r^uBO jsj

PAGE 155

1^3 material could not be prepared again. All attempts to do so failed. Reactions of (CH,)^NBH, and (CH,),NBC1, in an autoclave and the stability of the monoand dicliloroborane adducts under these conditions Various ratios of trimethyl amine borane and trimethylamine trichloroborane were heated in a stainless steel autoclave at 150° for nine to ten hours. A solid product was removed from the autoclave and recrystallized from hot CCl^. The results of the experiments are given in Table 20. The infrared spectrum of the product produced by the reaction of one mole of (CH^);zNBH, with two moles of (CH,),NBC1;7 was the only spectrum indicating a pure compound, ((CH,),NBHCl2). The melting point of this compoxmd was 158160**, and it was recovered in 90.6 per cent yield. A commercial analysis gave: Per Cent H N 01 B Pound 25.17 6.93 9.86 50.2? 7.87 Calcd. for (CH^),NBHCl2 25.-4-1 7.11 9.88 50.00 7.62 The tendency for (0E:,):.WSRGl2 to produce (CH,),NBC1, when heated was greatly reduced when (CH;z),NBHCl2 was heated in the presence of (GH^),N as may be seen in Figure 50, This implied that the (CH:,),N was a possible intermediate in the reaction to produce (CH^),NBG1,.

PAGE 156

1^^ o PQ g H en <^ o CO o pq o w o a o H ro, W ^f^ o P4 o o M EH O O bO O -P O td ^ I o (h o pq O C6 •H ?^ •P o O Ph Q) ri a fH 0:3 «EHP i1 o •H P O O Fh rt s ps O'H O W W O r-i O -ci o i* o o OJ OJ O O OJ w CVJ o rH CVJ O rH OJ O ro» rA W ffi w w w O O O O O o

PAGE 157

1^5 o o NO § o souB'^i^-f.uisuBj':}. %uQO aej o o o o o o o o e c 0) > -p c (U to 0) a. a: o p CQ -P c a o -p o x: -p c -P •H rH O a: CQ o o o o o o o o o o 00 o o o 30u-e%'^zvisuTej.% %ubo aaj e o c CI to (D U Q. en e d C > -p o p •H C •H -P nJ 0) c o -p o 3 n o a o c o to •H t. rt Oh e o o I I • o en to •H

PAGE 158

DISCUSSION OF RESULTS AlTD CONCLUSIONS Reactions of (CH^),NBH^ and HSGI2 Trimethyl amine borane reacted with mercuric chloride in diethyl ether to give the following products: (CH^)^NBH2C1, (CH^)^NBHCl2, Hg2Cl2, HCl(g) and H2(g), according to thg following equations i ' (CH^)^NBH^ + 2HgCl2 -* (CH^)^N3H2C1 + Hg2Cl2 + HCl [153 (CHj)^NBH2Cl + 2HgCl2-* (CH,),NBHCl2 + Hg2Cl2 + HCl [16] (CH^)^NBH, + ECl (CHj),NBH2Cl + S.^ [1?] The products isolated from the reaction mixtiire depended upon the reaction time and method of addition of the HgCl2. In ether solution, the temperature did not affect which compounds were produced in the reaction. An infrared spectrum of the white solid material recovered from the ether solvent agreed with the superposition of the infrared spectra of pure (CH,),NBH2C1 and (CH^),NBHCl2. The ether solution was acidic to pH paper and had a strong odor of HCl, The white insoluble residue resulting from the reaction gave an x-ray diffraction pattern identical to that of a pure sample of reagent grade mercurous chloride. The non-condensible product (H2) could be explained as a product 146

PAGE 159

1^7 of the reaction between trimethyl amine borane and the HCl produced by the reaction with HgClo* It has been reported (^0) that HCl will react with trimethyl amine borane to give the monochloroborane adduct and Hp gas. It has been confirmed in this laboratory that trimethyl amine borane and HCl, in diethyl ether at 0**, formed the monochloro substituted adduct and hydrogen gas. This side reaction would account for the non-condensible product formed in the reaction. It would also account for the fact that slightly less than two moles of HgClo were required per mole of BCl produced. The equations were written in a stepwise sequence which was implied from the experimental observation that (CH,),NBH2C1 reacted with HgCl2 to give (CH,)^NBHCl2. The next logical step did not occur in the ether solvent: (CH2)2J^KCl2 did not react with HgCl2 under these conditions, The preparation of monoor dichloro substituted amineboranes were not as clean-cut as the equations would tend to imply. If the reaction time was less than one hour or if the HgCl2 was added as a solid to the trimethyl amine borane-ether solution, the reaction would not give the products expected from the stoichiometric quantities. If the molar ratio was two moles HgCl2 per mole of (CH,),NBH,, the products were primarily (CH,),NBHCl2 and a lesser amo\mt

PAGE 160

1^ of (CH,),NBHCl2. If the ratio was four moles HgCl2 per mole of (CH^),NBH^, or greater, tlie products were primarily (CH,),NBHCl2 and a lesser amount of (CH;,)^NBHpCl, Even thougli excess HgCl2(s) was added to the trimethyl amine borane-ether solution and allowed to sit for several days, the reaction would not go completely to the dichloro adduct hut some monochloro adduct would remain in the reaction mixture. It appeared as if the HgCl2 became unreactive toward the trimethyl amine borane. An e3q)lanation for HgCl2 becoming passive could be that the mercury metal in a very finely divided state, as it was produced in the redox reaction, reacted with the HgCl2 on the surfaces of the crystalline HgCl2 to form Hg2Cl2. In this manner, the HgCl2 crystals surfaces could " become covered with unreactive, insoluble HgoClpj and thus prevent the crystal from dissolving or reacting further with the amineborane. The reaction of (CH,);,NBH^ with HgCl2 could be made to go completely to the dichloro adduct. If the HgCl2 was dissolved from a Soxhlet extractor filter thimble by refluxing ether and added, in this manner, to a trimethyl amine borane-ether solution, the reaction would go completely to the dichloroborane trimethyl amine adduct. In this case, the HgCl2 was added as a homogeneous solution to a solution of trimethyl amine borane. The time required for the addition

PAGE 161

1^9 of HgCl2 to tlie (GH,)^NBH^ was limited by the solubility of the HgClo in dietliyl ether. The reaction was complete at the dichloro product and further refluxing of the reaction mixture did not cause the reaction to proceed further even though Hg(II) was present. Therefore, this was found to be a good method to prepare useful quantities of (CH,)^NBHCl2. Trimethyl amine dichloroborane was characterized for the first time in this work. Schlesinger, Flodin and B\irg (40) had claimed this compound as the result of longterm heating at 100** of HGl(g) and (CH^)2NBH,, but they gave no experimental details or properties of the compound. In this work, the compound was found to be a white solid having a melting point in the range of 1^5-150*. The infrared spectrum is given in Figure 23 and the NMR spectrum in Figure 26. The reaction of trimethylamineborane with two moles of HgClo to prepare (CH;,),NBHpCl did not give a pure product, Invaxiably, the isolated product was a mixture of (CH2),NBH2C1 and (CH,),l^rBHCl2. The material was primarily the mono substituted product, but it could not be made the sole product of the reaction. The side reaction of (CH,)^NBH, and HCl increased the yield of (CH,),NBH2C1 and at the same time reduced the amount of HgClp necessary to produce only the (CHx)5NBH2Cl, The difficulty (or im-

PAGE 162

150 practicality) of separating the (CH,);,NBH2C1 and (CH;,);,lTBHCl2 made this method unattractive as a means to prepare (CH^),NBH2C1. Experimentally it was observed that in ether solution the HgClp did not discriminate to any detectable extent in reacting with (CH^)^NBHj and (CH,)2NBH2C1, but it did discriminate in reacting with (CH,),NBH2C1 and (CH,),NBHCl2. Mercuric chloride reacted with (CH^)^NBH2C1 but did not react with (CH,),NBHCl2 in refluxing ether. Thus, mercuric chloride reacted to substitute two chlorine atoms for hydrogen atoms on the boron atom in trimethyl amine borane, but would not substitute the third chlorine atom for the last hydrogen atom on the boron atom in diethyl ether. However, under appropriate conditions, all three BH bonds can be reacted with excess HgCl2 to give (CH,);,NBC1^. When (CH,),NBH;, and excess HgCl2 were heated in an autoclave for nine hours at 100°, the reaction did not go to completion but it did produce a substantial amount of (CH,);zNBCl, in addition to the other two chloro substituted adducts. In this reaction, the Hg(II) reacted to form mercury metal instead of the usually recovered Hg2Cl2 product. Thus, the BH bond in (CH,);,NBHGl2 required more energetic conditions to react, Trimethyl amine borane reacted with mercuric chloride in a water solution to give approximately 2 per cent yield

PAGE 163

151 of the monochloro adduct, but the primsiry product of the reaction was boric acid and hydrochloric acid,' A high concentration of chloride ion in the water solution did not affect the recovered products of the reaction to any significant extent, but the reaction was slower in the chloride solution than in water solution. The solubilities of the reactants appeared to be different in the chloride solution, the amine borane less soluble and the mercuric chloride more soluble. The enhanced solubility of mercuric chloride could be explained by the formation of tetrachloromercurate(II) ion in the chloride solution and the decreased rate of reaction, by the tetrachloromercurate(II) ion being less reactive toward the BH bond than toward the uncomplexed mercury (II) ion. The change in pH during the reaction of (CH;z),NBH, and HgClo in water indicated that the reaction was essentially complete within six seconds. The change in pH and the amount of strong acid produced implied that the reaction had proceeded according to the following equation: (CH,),NBH, + 6HgCl2 + 3H2O -* (CH,),UHC1 + B(OH), + 5Hg2Cl2 + 5HC1 [18] Thus, the reaction to produce (CH;j)^NBH2Cl must have been a side reaction, or the (CH2),NBHCl2» if produced from (CH2);,NBH2C1 must have hydrolyzed immediately.

PAGE 164

152 The rate of change in pH during the reaction of (CH,),NBH2C1 and HgCl2 in water implied that the reaction was slower than the reaction with (CH,),NBH7. The pH of (CH,),NBHCl2 slurried with a water-ethanol mixture (20:1) decreased slowly at a steady rate, and the addition of HgClp did not affect the rate of decrease in the pH. Therefore, Hg(II) ion did not react with the BH bond in (CH,)5NBHCl2 in water solution to any measurable extent* This suggested that the initial reaction between a BH bond and Hg(II) ion resulted in the formation of boric acid and not in the formation of a BCl bond, which then underwent hydrolysis. Since the reaction rate of (CH,),NBH2C1 with HgCl2 was slow, if (CH,)2NBH2C1 had been a product or intermediate in the reaction of (CH,),NBH^ and HgCl2 in water the pH change curve should have had a point where the rate of change in pH decreased since the reaction rate of (CH^)^NBH2C1 with HgCl2 was slow. This break point would have corresponded to a build up of the (CH,),1TBH2C1. But, since no change in the rate of decrease of the pH occurred when (CH^),NBH, reacted with HgCl2 in water, it may be concluded that no significant amounts of (CH,),NBH2C1 were produced as an intermediate in the reaction, but that the reaction produced boric acid and HCl as the primaxy products . The reaction between the BH bond and the Hg(II) ion may be considered to be an oxidation-reduction reaction

PAGE 165

155 witli the BH bond as the reducing agent and the Hg(II) ion the oxidizing agent. Considering the experimental results, a plausible mechanism would be one in which a two-electron transfer step occurred. The Hg(II) ion was attracted to the electronic density of the BH bond and accepted two electrons from the bond. This would give mercury metal, H"*", and any intermediate C(CH,),NBH2'*']. The positive site on the boron attracts a base (X") to give (CH^),EBH2X. This process most likely occurred in one step. The activated complex would most probably be one in which Hg(II) accepted the electrons from the BH bond as the Cl" approached and the H left. rather than a stepwise process involving complete ionization. The mechanism may be graphically depicted by the following sequence of configurations: H g H , (CHj)jN— B"^ * (CH^)^^— B-H H Cl-Hg-Cl HCljHg-^1 [19] H /^ (CH,),N— BC^ + Hg + HCl (OH;,);,!!— B+ . 5 3 i^ci ^ 5 3 ^ CI' H^ • ^ Hg CI" Subsequent reactions : Hg + HgCl2 •* Hg2Cl2 C20] (CH5)5NBH, + HCl (CH,),NBH2C1 + Hg C21]

PAGE 166

15^ This is essentially a similar mechanism to the one suggested by Havrthorne (1?) for the hydrolysis of a pyridine diphenyl horane and to the one proposed as a mechanism for the reaction of dimethylamine borane to eliminate hydrogen in Part I of the dissertation. The intermediate (CH,),NBH2'*" would be expected to be very reactive. Three-coordinate boron species are usually reactive due to the empty orbital and a three coordinate boron species having a positive charge would be expected to be even more reactive. The expected reactivity would argue against a stepwise process. The mercury metal reacted with the Hg(II) in the system to give Hg(I) which could precipitate as the insoluble salt. Mercury metal was a reaction product in the autoclave reactions where no solvent was used and the Hg(o) and Hg(II) did not have as good an opportunity to come in contact. The H"^ reacted with the hydridic BH hydrogen to give H2 and another boron intermediate which immediately reacted with abase to give (CH,),NBH2X. Thus, the mechanism accounts for all the observed products. If an electron transfer step was proposed where the Hg(II) accepted only one electron from the hydridic BH bond, the hydrogen retained the other electron, and leaving the boron with an acceptor site, this would accoiint for the HgpCl2, but it would also give hydrogen atoms which would require that Ho be a major reaction product, which it was

PAGE 167

155 not. It would not account for the HCl or mercury metal observed in the reactions and would be contrary to the observed data. The proposed activated complex for the mechanism implied that the compound bonded to the boron atom after reaction must be the same component as originally present in the primary coordination sphere of the Hg(II) ion. The experimental results were consistent with such a configuration for the activated complex. In ether solution, the HgClp would be present primarily as the molecular species and the chlorine atoms would be close to the mercury atoms. In the solid phase reactions the chlorine atoms would be closest to the mercury atoms, and in both cases the observed reaction products were the chloroborane adducts. In water solution, the mercury(II) ion would be surrounded by solvent molecules and in the activated complex it would be a water molecule instead of a chloride ion which would be forming a bond to the boron atom. The resulting species would be expected to hydrolyze immediately. The small amount of monochloro adduct recovered from the water solution could result from Hg(II) ions which still had a chloride ion remaining in the primary coordination sphere. The failure to increase the amount of monochloro adduct produced by increasing the concentration of chloride ion in the solvent and the apparent decreased reaction rate implied a difference in the

PAGE 168

156 reactivity toward the BH bond of a simple Hg(II) ion and the tetrachlormercurate(II) ion. ; Thus, the proposed mechanism, where Hg(II) ion accepted two electrons, explained and was consistent with all the observed experimental results for the reaction of trime thy 1 amine borane and mercuric chloride. The side reaction between the unreacted trimethylamine borane and the hydrogen chloride produced by the reaction with Hg(II) raised the question as to which reagent reacted faster with trimethyl amine borane. Mercuric chloride reacted faster than HCl with (CH^),NBH, at 0** in diethylether. An ether solution of (CH,),NBH, was poured into an ether solution of HCl and HgCl2. The molar ratio of reactants was 1(CH,^H^: lHCl:2HgCl2. The E^ produced was found to be approximately twenty mole per cent of the hydrogen which could have been produced by complete reaction of the HCl with the amineborane. The reaction time was sufficiently long so that the HgClp would have reacted completely with the trimethyl amine borane, if no HCl had been present. This means that 80 per cent of the trimethyl amine borane reacted with the HgClg. The HgClp was not completely dissolved in the ether, but a small portion of the solid was suspended in the solvent. The HCl was in solution and each time a HgCl2 molecule reacted, it produced one of HCl, Therefore, the HOI should have had the advantage of being in more

PAGE 169

157 homogeneous contact with the trimethyl amine borane and of being replenished by the competing reaction. The HgClo reaction being faster than the HCl reaction agreed with the qualitative observations. An ether-trimethyl amine borane solution and a HgClo-ether solution, when mixed, gave an immediate white precipitate. This precipitation appeared to be complete within a few minutes and no further visible changes in the system were noticeable. Nevertheless, the reaction of HCl(g) with trimethylamine borane, in ether at 0° or benzene at 25**, was a good method to prepare usef\il quantities of (CH,);,NBH2C1. The reaction was complete at the monosubstituted adduct tinder these conditions. To prepare the dichloro adduct from these reagents it was necessary to heat the two compounds in an autoclave at 100° for twenty-four hours. This had been reported (^0) previously and was confirmed in this laboratory. A report in the patent literature by Borer and Dewing (2) of the preparation of the monochloro adduct from trimethylamine borane and concentrated hydrochloric acid was investigated and found to be correct. In the preparation procedure reported (2), it would have been possible for the monochloro adduct to have been formed during the benzene extraction by a reaction of amine borane and HCl, which would both be extracted into the benzene phase. Borer and

PAGE 170

158 Dewing 's procedure was repeated except that the water solution was neutralized with NaOH solution before extraction with benzene; a small amount of the monochloro adduct was still recovered from the benzene extraction phase. Thus, the monochloro adduct was formed in the concentrated hydrochloric acid. The monochloro adduct was prepared, using concentrated HCl(aq) as a source of HCl, by adding the concentrated acid to a benzene solution of trimethylamine borane. After shaking, the water-benzene phases were separated and the HCl which extracted into the benzene reacted with the amine borane. The monochloro adduct could be recovered in approximately ^5 per cent yield but the infrared spectra implied that the reaction wsls not quite completed in forty-five minutes. Other organic solvents, e,g,, CCl^, could be used for this reaction procedure. The reaction mechanism for the reaction of a BH bond with HCl would be one in which H"*" acted as the oxidizing agent, accepting electrons from the BH bond to form hydrogen gas. This is the same mechanism as has been previously suggested for the Hg(II) reaction. The mechanism is consistent with the observed experimental results. Reaction of (CH;,),NBH^ and (CH,),NHC1 Trimethylamine borane reacted with trimethyl amine hydrochloride to give the chloro-substituted adducts. On

PAGE 171

159 heating the solid reagents to 150-185° for fifty hoxirs, monoand dichloro adducts were isolated. In order to get a more complete reaction, it was necessary to remove the trimethylamine product about halfway through the reaction time and then reheat the mixture. The reaction products implied that the following reactions occurred, (CH,)zNBH, + (CH^)^NHCl (CK^)jNBH2Cl + iGE^)^Ii + H2 [22] iCE^^^mU^Gl + (CH,)^NHC1 (CH^)^NBHCl2 + (CH^)3N + H2 C23] (CH,),NBHCl2 + (CH^),NHC1 (CH^)^NBCl^ + (CH^)3N + Hg [2^3 If the starting materials for the reaction were (CH;,);,NBH2C1 and (CH2)2NHC1, equation [25], then the reaction product contained the trichloro adduct, The trichloro adduct was not obtained in any of the experiments where the starting reagent was trimethyl amine borane. But, the products recovered were strongly affected by the trimethylamine produced in the reaction. The compounds were not very stable at these high temperatures (150-185°) for extended periods of time (2^-50 ho\ars) and the products were recovered only in small yields ( '^ ^0 per cent) ; usually contaminated with the amine hydrochloride. However, the initial preparation of pure trimethyl amine dichloroborane was accomplished by this .method. Considering the

PAGE 172

160 reaction products and their dependence on the presence of trimethyl amine the following mechanism was proposed. The protonic hydrogen in the trime thy 1 ammonium ion accepted an electron pair from the hydridic hydrogen in the BH "bond to produce hydrogen, trimethyl amine, and an acceptor site on the boron atom. A competition for the site on the "boron occurred between trimethyl amine and the chloride ion, and the experimental conditions determined which was the stronger base toward the boron atom. Miller, Chamberland, and Muetterties (26) have reported that the following reactions occur: (CH^)^NBH, + (CH^),NHG1 i^^ [(CH,)^N]2BH2C1 + H2 [25] [(CHj)^N]2BH2Cl 18^-^QQ°^ (GH^)^NBH2C1 + (CH^)^N [26] Therefore, the evidence suggested that at lower temperatures the trimethyl amine was the stronger base, forming a boronixim ion, and at higher temperatures the chloride ion was the stronger base. The mechanism was the same except the trimethyl ammonium ion attacked the BH bond. There were two possible mechanisms which could acco\Hit for the second (or third) step of the reaction sequence. The first, and most probable one, would be a repetition of the first step with the chloride ion being the stronger base toward boron than the trimethyl amine, under the experimental conditions. The

PAGE 173

161 second would be a dissociation of the amine monochloroborane into the amine and monocbloroborane, followed by a disproportionation of the monochloroborane into borane and dichloroborane which then recombined with the amine to form trimethyl amine dichloroborane according to the following equations: (CH,)NBH2C1 !5 (CH,)xN + BHgCl [2?] aBHgCl ^ BH, + BHCI2 [28] (CH,),N + BHCI2 -* (CH,),NBHCl2 [29] (CH,),N + BH, (CH^)^NBH^, repeat cycle [30] Diborane and trichloroborane have been reported (5) to disproportionate and the various chloroboranes have been trapped as the diethyletherates. In our e3cperiments a sample of (CH,),NBH2C1 after heating at 185° for twentyfour hours was foxmd to contain some (CH,),NBHCl2» Both mechanisms account for the experimental results and neither can be completely discounted. Reaction of (CH,)^NBH2 and SbClc Antimony pentachloride reacted violently with trimethylamine monochloroborane to form an acidic gas (HCl), (CH,)^NBC1,, SbCl,, and a grey powdery material (Sb). The following equations were suggested by the data:

PAGE 174

162 (CH,)^NBH,C1 + 2SbCl5 (CH^)^NBCl, + 2SbCl, + 2HC1 [31] 3(CH,),NBH2C1 + ^SbCl^ 3(CH^)^NBC1^ + ^Sb + 6HC1 C32] The antimony trichloride produced in the reaction did not react with the BH bond in ether solution but as soon as the ether was removed a reaction occurred slowly to produce a grey material. Thus, the etherate of antimony trichloride stabilized the antimony (III) toward further reaction with the BH bond. The dichloroborane adduct also reacted vigorously with SbClc to produce the grey powder and a gas. The grey reaction product occurred in each of the reactions, said visually appeared to be a metallic powder. The powder was oxidized in nitric acid, the solution buffered with ammonium acetate, and the orange-red antimony oxysulfide complex formed on the addition of Na2S20, crystals proving the powder to contain antimony, but not what oxidation state it was originally, A similar mechanism could be proposed for the reactions of SbClc and SbCl^ as was given for the reaction of Hg(II) ion in which the metal atom accepted electrons from the BH bond. Reaction of (CH,),NBH, and SbCl^° Qualitative reactions between antimony trichloride and trime thy 1 amine borane and the chloroboranes confirmed

PAGE 175

165 the fact that they react to give a grej-black residue which, in the case of the trimethyl amine borane reaction, resulted in a metallic mirror covering the flask. The metallic mirror indicated that the grey-black material was indeed metallic antimony, The experiments showed that the order of reactivity of the amine boranes with SbCl^ was -BH^, > -BH2CI > -BHCI2. Antimony trichloride and (CH;,);,NBHCl2 formed a clear liquid when the solids came into contact which slowly tiirned dark. The infrared spectrum in the range of ^000-500 cm" indicated no change in the spectrum of the starting material other than perhaps some formation of (CH,),NBC1,, No e3q)lanation other than the formation of a low melting eutectic mixture of the solid materials seemed reasonable unlesssome sort of liquid polymeric species was formed in which the boron bonded chlorines were acting as a Lewis acid and the SbCl, acting as a Lewis base. But no reaction occurred between (CH,)xNBCl, and SbCl;, which could be reasonably expected if the SbCl^, acted as a Lewis acid to accept an electron pair from a chlorine bonded to boron to form a liquid polymeric species containing chlorine bridges. Thus far, the oxidizing agents used here to react with the BH bond have been either a high oxidation state metal, that is, Hg(II), SbCV), Sb(III) or a protonic species. Compounds of non-metsillic elements in high oxidation states,

PAGE 176

16^ tliat is, P(V), S(VI), S(IV), should undergo analogous reactions; and therefore, the reactions of sulfiiryl and thionyl chloride with the amine borane were considered. Reaction of (CH^),NBH, and SO2CI2 Sulfuryl chloride reacted with trimethyl amine borane in a straight-forward manner according to the equation: (0Hj)5NBH^ + 3SO2CI2 * (OH^)^NBOl^ + 3H01 + 5SO2 [533 The reaction occurred on warming slowly from -78° to -25°, The weight increase during reaction implied 97 per cent yield of (CH,),NBC1,, but the infrared spectrum implied a small impurity of (CH,)^NBHCl2. Each of the products was identified by its infrared spectrum; HCl (1.50 mmoles) was separated from the SO2 and S0pCl2, but the other two components could not be separated. The same electron transfer mechanism as previously outlined was also proposed for this reaction. The sulfur atom accepted two electrons forming the stable SOp molecule which was given off as a gaseous product just as the Hg(II), Sb(V) and Sb(III) have been proposed to accept electrons to form stable species of a lower oxidation state. The evidence for a small amount (CH^),NBHCl2 in the solid reaction products' infrared spectriim of (CH,);,NBC1, suggested that the formation of the trichloro adduct also occurred in steps.

PAGE 177

165 Reaction of (CH^)^NBH, and SOCI2 Thionyl ch.loride reacted with trimethyl amine borane to form the dichloroborane adduct, HCl, SO2, and a sulfur containing residue. Considering the reaction to be analogous to the reaction with sulfiiryl chloride, the following equations may be written. (CH5),KBH, + 2SOCI2 (CH^),NBHCl2 + 2HC1 + 2S0 [3^] 2S0 S + SO2 C55] The intermediate compound SO which was proposed by these equations has not been reported as a stable species, but was suggested by analogy. The reaction occurred very slowly at -65°, slightly faster at -23° and at a convenient rate at 0**. The gaseous products were separated and infrared spectra confirmed their identity. The solid reaction residue was separated into (CH;,),NBHCl2 and a yellow residue and the infrared spectra obtained confirmed the dichloro adduct. The spectrum of the yellow residue contained (CH;,);,NHC1 as the major infrared active component but it also showed an intense peak at 1200 cm" and less intense peaks at 925 cm~ and 550 cm , The stoichiometry of equation [3^] was not confirmed by the meas\ired amounts of HCl and SO2 gas. More HCl(g) (1,21 mmoles compared to 1,01 mmoles) was recovered than would be implied by equation C3^] and less SO2 (0,53 mmoles compared to 1,01 mmoles) was recovered. The weight

PAGE 178

166 of non-volatile products of tlie reaction was also larger (0.98g) than would be predicted by equation [5^] (0.0880g). Thus, equation [3^] as written does not describe completely the stoichiometry of the actual reaction. The proposed mechanism would be the same electron transfer process as previously suggested with the sulfur atom accepting the electrons followed by a decomposition of the reduced sulfur containing species. This was the same mechanism as in the previous cases except that the reduced sulfur compound in this instance was not a stable compound. Conclusions from the reactions Bach of the reactions discussed so far may be considered, and have been considered, as an oxidation-reduction reaction with the BH bond as the reducing agent and a variety of different oxidizing agents. Each reaction involving the BH bonds was found to be a stepwise reaction and the niimber of BH bonds reacted per molecule varied with the oxidizing agent or with the experimental conditions. Hydrogen chloride reacted with trimethyl amine borane to give the monochloro adduct at room temperature, mercuric chloride gave the dichloro adduct, and antimony pentachloride gave the trichloro adduct under the same conditions thus implying an ordering in the relative strength of the oxidizing sigents toward the BH bond in the amine borane and indicating a

PAGE 179

167 change in ttie BH bonds' reactivity during the reaction. For the reagents which gave a product which still confc ained a BH bond, the reaction occurred to a further extent if the conditions were made more energetic. This showed that the reducing ability of a BH bond decreases as chlorine atoms t were substituted for hydrogen atoms. The work of Noeth and Beyer (52) was also consistent with a change in reactivity of the BH bond as the substituents bonded to boron changed. They reported that trimethylamine borane reacted with HBr, HI and Ig to give the monosubstituted adduct and that chlorine, bromine, and HP reacted with the amine borane to give the tri-substituted adducts . The decrease in reactivity of the BH bond as a reducing agent on chloro substitution may be explained in the following manner. Once the more electronegative chlorine atom is bonded to the boron atom, it would tend to shift the electron density in the bond more toward the chlorine. This in turn woxild cause a shift toward the boron of the electron density in the remaining BH bonds, the net result being that the attached hydrogens were less hydridic and would therefore be weaker reducing sites. A second chlorine bonded to the boron would make the remaining BH bond even less hydridic and, by this argument, it would not be expected to be a very

PAGE 180

168 strong reducing agent. This should also cause a strengthening of the BN bond, but this would be countered by increased steric strain due to the size of the chlorine atoms. Essentially, this same inductive argument has been used by Muetterties (28) to explain the reduced reactivity of BH bonds in the boronium ions. Q?he inductive effect argument was confirmed by th.e B chemical shifts observed in the nuclear magnetic resonance spectra of the amine borane and chloroborane compounds. The chemical shift, which may be considered as a measure of the electronic density about the nucleus, suggested that the order of electronic density about the boron atom decreases in the following manner, (CH,),NBH;, > (CH,)^NBH2C1 > (CH,),NBHCl2 > (CHj),NBCl,. Thus the hydridic character of the BH bond should decrease in the same order as the electronic density on the boron atom. The largest change in the chemical shifts occurred between the monochloro and dichloro-substituted adducts and it was between these two compounds that the greatest chamge in the reactivity of the BH bond was noted. Prom the experimental conditions and from the extent of reaction it was possible to predict an ordering of the strength of the oxidizing agents toward the BH bond. The , experimentally observed order of reactivity was SbClc > SOgOlg > SOClg > HgCl^ > HCl > (CHj),HHCl. Antimony

PAGE 181

169 tricliloricLe was not included because a quantitative esqperiment using it as the oxidizing agent was not attempted, but the qualitative indications implied that it would probably be intermediate between sulfuiyl and thionyl chloride. The ordering was based primarily on the following considerations: 1. SbClc reacted to give only (CH;,)xNBCl,. r 2, SO2CI2 reacted to give (CH,),NB01x but also some (CHi)2iraH01p remained in the system, 3.' SOCI2 reacted to give (CH,),NBHCl2 below 0°. ^', HgCl2 reacted at 0° to give (CH^)^NBHGl2. 5. HCl reacted at room temperature to give (CH^):5NBH2C1« 6. (CH,),NHC1 reacted only at elevated temperatures. Thus, from these considerations it would be possible to pick the oxidizing agent to give the best preparation of a desired product. Thermal stability of the adducts The thermal stability toward disproportionation of trimethyl amine borane and the chloroboranes could determine which products were recovered from an experiment involving high temperatures for long periods of time. Therefore, a preliminary study was made of the stability of these adducts toward interconversion of one adduct into another. Trimethyl amine borane and trimethyl amine trichloroborane, in a 1:1 molar ratio, were heated for nine to ten

PAGE 182

170 hours at 150° in an autoclave; the recovered product was found to contain the mono and dichloro adducts with no evidence for the presence of the starting materials in the infrared spectrxim. If the stoichiometry were varied to two moles of trime thy 1 amine trichloroborane to one mole of trimethylamine borane, the recovered product was trime thy 1 amine dichloroborane. But if the stoichiometry were the reverse of that, namely, two moles amine borane to one mole amine tricMoroborane, the recovered product was the monochloro adduct; a small amount of the dichloro adduct was evident in the infrared spectrvim. The reactions may be summarized in the following equations : (OH,),NBH, + (CH^)^NBCl^ (CH,)^NBHCl2 + (CH,),NBH2C1 [36] (CHj),NBH, + 2(CH,),NBC1^ 5(CH,)^NBHCl2 C37] 2(GH^),NBH, + (CH^)^NBCl^ ^ 3 ( CH, ) , NBH2CI C+ (CH,)^NBHCl2 small amount] [38] The results suggested that the dichloro adduct was the most stable one under these conditions. This idea was further supported by the heating of the monochloro adduct alone and finding evidence for the dichloro adduct in the recovered material . The monochloro and trichloro adducts when heated under the same conditions gave primarily the dichloro adduct with

PAGE 183

171 a small amoxint of tlie trichloro adduct in the recovered material. The results implied that there was an extra stability associated with the dichlorohorane adduct which the other adducts did not have. Ratajaczak (56) reported that triethylamine dichloroborane appeared to have some special stability and that when one mole of (C2Hc)2NBH, reacted with more than one mole of BClz(s) only (C2Hc):5NBHCl2 was isolated. The dichloro adduct when heated for twenty-four hours at 185° gave as recovered products the starting material and the trichloro adduct. But, when heated in the presence of trime thy 1 amine the recovered product was greater in yield and was principally the dichloro adduct with only a small amount of the trichloro adduct present. The trimethylamine thus suppressed the reaction. In view of the experimental results, the following mechanism was proposed for the rearrangement reactions of the borane adducts. The first step v/ould be a dissociation of the amine borane into the amine and the borane, the second step would be a disproportionation reaction of the boranes, and the third step would be a recombination of the borane and amine. The effect of trime thyl amine on the tendency of the dichloroborane adduct to undergo disproportionation indicated that the mechanism involved trimethylamine. Thus, a dissociation of the amine borane occurred

PAGE 184

172 , (•> during the reaction. This is the same mechanism presented as a possibility for the second step in the reaction of (CH,),NBH, and (CH,)^imCl. Therefore, it may be concluded that the recovered products from an experiment using high temperature for extended periods of time may or may not be the primary reaction product but may have been produced by subsequent disproportionation reactions.

PAGE 185

SUMMARY Trimethyl amine chloroboranes were prepared by reacting trimethyl amine borane, (CH,)^!^!!,, with mercuric chloride, hydrogen chloride, trimethyl amine hydrochloride, antimony pentachloride, antimony trichloride, sulfxiryl chloride, thiouyl chloride, and trimethyl amine trichloroborane (CH,);,NBC1,, The extent of chloro substitution on the borane was a function of which reagent was used and the reaction conditions, Trimethyl amine dichloroborane, (CH,),NBHCl2, was prepared and it was characterized for the first time in this work. The reactions were all considered to be oxidationreduction reactions in which the BH bond was the reducing agent. An order of reactivity of the oxidizing agents was determined to be: SbClc > SO2CI2 > SOCI2 > HgClg > HCl > (CHj),NHCl. The reactivity of the BH bond decreased on chloro substitution of the boron atom in the order: -BH, > rBH2Cl > -BHGI2. This decrease in reducing ability of the BH bond was explained by an inductive effect in which the chlorine atoms . 173

PAGE 186

17^ withdrew electronic density from the "boron atom, and thus 11 from the BH bond. The model was consistent with the B nuclear magnetic resonance spectra of the compounds. A mechanism was proposed in which the oxidizing agent accepted an electron pair from the hydridic BH bond. The mechanism proposed a non-linear activated complex and was similar to the mechanism proposed by Hawthorne (1?) for hydride transfer reactions. The mechanism was consistent with all the experimental results. The basic idea of using the reactivity of the BH bond as a means to prepare B-substituted amine boranes appeared to be one worthy of further investigation.

PAGE 187

BIBLIOGRAPHY 1. Gordon M. Barrow, Introduction to Molecular Spectroscopy , New York: McGraw-Hill Book Company, Inc., 1^62, p. ^^. 2. K. Borer and J. Dewing, Brit. 881,576, Nov. 1, 1961. 3. H. Cr Brown, Hydro"boration , New York: W. A. Benjamin, 1962. 4. H, C. Brown and P. A. Tierney, J. Am. Ghem. Soc, 80, 1552(1958). — 5. H. C. Brown and P. A. Tierney, J. Inorg. Nucl. Cliem. , 2, 51(1959). 6. A. B. Burg, J. Am. Chem. Soc, 2i» 13^0(1952). 7. A. B. Burg and Carl L. Randolph, Jr., J. Am. Chem. Soc, 21, 953(1951). 8. Chemical Rubber Publishing Co. , Handbook of Chemistry and Physics , Wth ed. , 1958-1959, pp. 2356, 2572. 9. Ibid., p. 536. 10. Gerd H. Dahl and Riley Schaeffer, J. Am. Chem. Soc, 8^, 303^(1961). 11. R. E. Davis et al ., J. Am. Chem. Soc, 85, ^87(1963). 12. J. P. Ditter et_al. , J. Phys. Chem., 6ff, 1682(1960). 13* J. 0. Edwards, Inorganic Reaction Mechanisms , New York: W. A. Benjamin, 196'4-, p. ^^8. 1^. P. N. Gates, E. J. McLaughlan and E. P. Mooney, Spectrochim. Acta, 21(19), 1445(1965). 15. W. Gerrard, Organic Chemistry of Boron , New York: Academic Press, 1961. 16. W. A. G. Graham and P. G. A. Stone, J, Inorg. Nucl. Chem., ^, 164(1956). 175

PAGE 188

176 17* M. p. Hawthorne and E, S. Lewis, J. Am, CtLem, Soc, 80, ^296(1958). 18. T. R. Hogness and V, C. Johnson, Qualitative Analysis and Chemical Equilibria , ^th ed, New York: Holt, Rinehart and Winston, Inc, 195^. p. ^27.' 19. Infrared Spectral Data, American Petroleum Institute Research Project ^^, Serial Number 315. 20. Ibid,, Serial Number '4-38, 21. W, J, Jolly, Synthetic Inorganic Chemistry , New Jersey: Prentice-Hall, Inc., 1960. p. 9^. 22. H, C, Kelly, Mario B, Giusto and Frank R, Marchelli, J, Am, Chem, Soc, 86, 3882(1964-), 23. H, C, Kelly, Frank R. Marchelli and Mario B, Giusto, Inorg, Chem,, ^, 4-31(1964-), 24-. Newton Levy, Jr., Solvolysis Kinetics of the Methylamineb cranes . Ph. P. Dissertation, University of Florida, August, 1964-. 25. Dowell E, Martz and Robert T, Lagemann, J, Chem, Phys,, 22, 1193(195^). 26. N. B. Miller, B. L, Chamberland and E, L. Muetterties, Inorg. Chem., ^, 1064-(1964-). 27. J. M. Miller and M, Onyszchiik, Can. J. Chem,, 4-1, 2898(1963). 28. E. L. Muetterties, Pure Appl. Chem., 10(1), 53(1965). 29. Kazuo Nakamoto, Infrared Spectra of Inorganic and Coordination Compounds . New York; John Viley and Sons , Inc., 1963. p. 111. 30. Ibid., p. 104-. 31. H. Noeth, Z.Naturforsch, ]^, 327(1960). 32. H. Noeth and H. Beyer, Chem. Ber. , 21» 2251(1960). 33. W. D, Phillips, H, C. Miller and E. L, Muetterties, J, Am. Chem, Soc, 81, 4-4-96(1959).

PAGE 189

177 5^, Raymond H, Pierson, Aaron N. Fletcher and E, St. Clair Gantz, Analytical Chem. , 28, 1218(1956). 55» Santiago R, Polo and M, Kent Wilson, J. Chem, Phys,, 22, 900(195^). 36. S, Ratajaczak, Bull. Soc, Chim, Prance, I960 , No. 5t 487. 37. J. S. Ridgen and V, S. Koski, J. Am. Chem. Soc, 83 « 3037(1961). 38. E. R. Roberts, H, J. Emeleus and H. V. A. Briscoe, J. Chem. Soc., 1939 , 41. 39* George W. Schaeffer and Elaine R. Anderson, J. Am, Chem. Soc, 21» 2143(19^9). 40. H. I. Schlesinger, Nestor V. Flodin and A, B. Burg, J. Am. Chem. Soc, 61, 1078(1959). 41. Robert C. Taylor, Adv. Chem. Series 42, 59(1964). 42. Harold C. Urey, J. Chem. Soc, 1947 , 562. 43. Allen N. Webb, John T. Neu and Kenneth S. Pitzer, J. Chem. Phys., 12, 1007(1949). 44. Kenneth B. Wieberg, Chem. Rev., ^, 713(1955).

PAGE 190

BIOGRAPHICAL SKETCH James William Wiggins was born MarctL 5* 19^0, at Paris, Arkansas, In May, 1958, he was graduated from Paris High School. In J\ine, 1962, he received the degree of Bachelor of Science from the University of Arkansas, In September, 1962, he enrolled in the Graduate School of the University of Florida,' He worked as a teaching assistant and a graduate assistant while pursuing his work toward the degree of Doctor of Philosophy. From September, 1964 until June, 1965, he was the DuPont Post Graduate Teaching Fellow. James William Wiggins is a member of the Methodist Church, American Chemical Society and Alpha Chi Sigma. 178

PAGE 191

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 was submitted 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. December 17, 1966 Dean, Col Arts and Sciences Dean, Graduate School Supervisory Committee:

PAGE 192

•^ ^ A/A//7'A


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EOT8KJQDW_7898KB INGEST_TIME 2017-07-19T20:31:33Z PACKAGE UF00097887_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES