University of Southern California Dissertations and Theses

Addition compounds of trimethylphosphine oxide

(USC Thesis Other)

Addition compounds of trimethylphosphine oxide

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ADDITION COMPOUNDS OF
TRIMETHYLPHOSPHINE OXIDE
A Thesis
Presented to
the Faculty of the Department of Chemistry
University of Southern California
In Partial Fulfillment
of the Requirements for the Degree
Master of Science in Chemistry
by
William Edgar McKee
September 1950
UMI Number: EP41587
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI EP41587
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Dissertation Publishing
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 4 8 1 0 6 -1 3 4 6
!$) . M)**
This thesis, written by
William...Edgar .McKee.....; ..
under the guidance of h£S...Faculty Committee,
and approved by all its members, has been
presented to and accepted by the Council on
Graduate Study and Research in partial fu lfill
ment of the requirements fo r the degree of
Master of Science
D a t e . . . October JL550>^
Faculty Committee
Wat hairman
jr .....
..
ACKNOWLEDGMENT
The author is indebted to Dr. A. B. Burg
for his valuable assistance and encouragement
during this investigation.
TABLE OP CONTENTS
CHAPTER PAGE
I. THE PROBLEM................................ 1
Theoretical introduction to the problem . . 1
Historical background ................ . . 6
II. PREPARATION AND PURIFICATION OP REAGENTS ... 15
Trimethylphosphine oxide ................ 15
Purification of phosphoryl chloride. ... 19
Purification of chloroform .............. 19
Purification of ethylene dichloride ... 20
Purification of methylene chloride .... 20
Purification of benzene ....... 20
Purification of toluene . ........... 20
Purification of Skellysolve A .......... 20
Purification of ether ................. 20
Purification of ethyl alcohol ......... 20
Purification of boron trifluoride .... 21
Preparation of sulfur dioxide .......... 21
Preparation of sulfur trioxide .......... 22
III. THE REACTION OF TRIMETHYLPHOSPHINE OXIDE WITH
BORON TRIFLUORIDE ....... .......... 23
IV. THE REACTION OF TRIMETHYLPHOSPHINE OXIDE AND
SULFUR TRIOXIDE ......................... 31
Preparation Method 1 .................... 32
lv
CHAPTER PAGE
Studies toward purification of the
product............................. 35
Preparation Method 2 .................... 38
Preparation Method 3 .......... 38
Preparation Method 4 . . ................ 39
V. THE REACTION OF TRIMETHYLPHOSPHINE OXIDE
AND SULFUR DIOXIDE....................... 45
VI. DISCUSSION AND CONCLUSIONS ........... 57
The reaction of boron trifluoride with
trimethylphosphine oxide ................ 57
The reaction of sulfur trioxide with
trimethylphosphine oxide ................ 57
The reaction of sulfur dioxide with
trimethylphosphine oxide ................ 59
VII. SUGGESTIONS FOR FUTURE INVESTIGATIONS .... 63
VIII. SUMMARY................... .'........ 65
BIBLIOGRAPHY........... 68
LIST OF TABLES
TABLE
I. Addition Compounds of Trimethylphosphine
Oxide......................... .........
II. Addition Compounds of Trimethylamine Oxide
of Interest in the Present Study ..........
III. Combining Ratio of (CH3)3PO and BF3 . . . . .
IV. Combining Ratio of (CH3)3PO and SO3 ........
V. Analysis of the Product formed by the
Sulfur Trioxide-Trimethylphosphine Oxide
Addition Compound and G2H^0H ..............
VI. Analysis of the Sulfur Trioxide-
Trimethylphosphine Oxide Addition
Compound (Method 4) .....................
VII. Pressure-Composition Isotherms of System
(CH3)3P0-S02
1.192 X 10“3 moles (CH3)3PO in System . . .
VIII. Pressure-Composition Isotherms of System
(CH3)3P0-S02 at 0°C.
1.192 X 10“3 moles (CH3)3PO in System . . .
IX. Pressure-Composition Isotherms of System
(CH3) 3P0-S02 at -23°C.
1.192 X 10“3 moles (CH3)3PO in System . . .
vi
TABLE PAGE
X. Vapor Pressures of SC^-CCHgJ^PO Mixtures
in Constant Pressure Range
Moles SC^/CCH^J^PO Approximately 0.2-1.1 . . 55
XI. Equilibrium Pressures of SC^-CCH^J^PO
Systems.............................. .. 60
LIST OF FIGURES
FIGURE PAGE“
1. Sublimation Apparatus for (CH3)3PO ........ 18
2. Apparatus for Preparing SO3 and Adding it
to (CH3)3P O ............................... 18
3. Apparatus for the Reaction of (CH3)3PO with
bf3 ............................... 27
4. Titration of (CH3) 3P0-S03 with 0.01948 N
NaOH ........... 42
5. Apparatus for Treating (CH3) 3PO with SOg ... 46
6. Vapor Pressure-Composition Isotherms of
S02-(0H3)3P0 Mixtures ..................... 53
CHAPTER I
THE PROBLEM
Trimethylphosphine oxide is formally similar to
trimethylamine oxide which has been shown to have great
external bonding power as a Lewis type base.1,2*3>5 it
would be of interest to. know whether the two compounds have
similar basic properties and to attempt a comparison of
their electron donor strengths. This should be possible by
making various addition compounds of trimethylphosphine
oxide and comparing their properties to similar addition
compounds of trimethylamine oxide.
I. THEORETICAL INTRODUCTION TO THE PROBLEM
g
Lewis postulated a dative nitrogen-oxygen bond and
1 A. B. Burg, J. Am. Chem. Soc., 65/ 1692 (19^3).
2 A. B. Burg and J. H. Bickerton, J. Am Chem. Soc.,
6 7. 2261 (19^5).
3 J. H. Bickerton, Master's thesis in Chemistry, The
University of Southern California, Los Angeles, 194-3*
^ W. K. Godfrey, Master's thesis in Chemistry, The
University of Southern California, Los Angeles, 19^0.
5 H. Z. Lecher and W. B. Hardy, J. Am. Chem. Soc.,
I0> 3789 (1948).
^ G. N. Lewis, "Valence and the Structure of Atoms
and Molecules," Chemical Catalogue Co., New York, N. Y.,
1923, P. 13^.
2
assigned trimethylamine oxide the following structure;
CH,:N:' 0:
3 • • * *
ch3
This structure has been confirmed in a number of ways.
Measurement of the nitrogen-oxygen bond length by Lister
and Sutton^ shows a length of 1 .3 6 + .03 which is the
same as the calculated value for a single bond. Palmer and
o
Elliott0 present evidence that a purely covalent dative
link has about the same length as an ordinary single bond.
Linton^ and Phillips, Hunter, and Sutton10 use dipole
moment data to support the conclusion that it is a dative
bond. The formation of stable addition compounds, a list
of which has been compiled by Godfrey, also shows that the
oxygen atom has a ready availability of electrons as would
be expected from this type of structure.
7 N. W. Lister and L. B. Sutton, Trans. Par. Soc.,
22> ^95 (1939).
Q
0 K. H. Palmer and N. Elliott, J. Am. Chem. Soc.,
6 0. 1852 (1938).
9 E. P. Linton, J. Am. Chem. Soc., 62, 1945 (19^0)
10 G. M. Phillips, J. S. Hunter and L. E. Sutton,
J. Chem. Soc., 146 (19^5).
3
The evidence for the true nature of the phosphorus-
oxygen bond in trimethylphosphine oxide Is not quite so
readily available. To date, no work directly meeting this
specific problem has been published, although some parallel
studies are' of interest. Originally oxygen was considered
to be double bonded whenever it was attached only to an
atom in Periodic Groups IV, V, VI, or VII. However, this
theory was out of accord with many of the experimental
facts. Most of the obvious discrepancies were overcome by
Lewis1 theory11 which implied single bonds in which the
central atom supplied both shared electrons for a single
dative bond, instead of the double bond concept. It was
now possible to formulate many of these oxygen compounds
with only an octet of electrons around each atom.
1 O
Pauling and Brockway cast some doubt on such a
universal applicability of the dative bond by showing that
the common oxy-acids have oxygen bond lengths as short as
or shorter than double bonds. Phillips, Hunter, and
Sutton, 10 using data from their own investigations and from
the literature on bond lengths, dipole measurements, and
bond energies, conclude that the dative bond in this type
11 G. N. Lewis, J. Am. Chem. Soc., 38, 762 (1916).
12 L. Pauling and L. 0. Brockway, J. Am. Chem. Soc.,
52, 13 (1937).
4
of compound is not as common as was once widely believed.
They believe that the nitrogen-oxygen bond is dative in
trimethylamine oxide because nitrogen has no orbitals
available for double bond formation. The phosphorus-oxygen,
sulfur-oxygen, selenium-oxygen, and chlorine-oxygen links
have more the character of double bonds. In the case of
trlphenylphosphine oxide, their data on dipole moments
indicate that the phosphorus-oxygen link is similar to a
double bond. This example, however, is complicated by the
ease with which electrons could be drawn into the phosphorus
from the phenyl groups. Hampson and Stosick1^ and Brockway
11l
and Beach also contend that the phosphorus-oxygen bond is
double.
In view of these indications of the double bond
character of the phosphorus-oxygen bond in other compounds,
the phosphorus-oxygen link in trimethylphosphine oxide may
be more like a double bond than a single dative type bond.
Hence, the question arises as to what effect this double
bonding would have on the basic, or electron donor,
properties of the oxygen atom.
^ G. C. Hampson and A. J. Stosick, J. Am, Chem.
Soc., 60, 1814 (1938).
^ L. 0. Brockway and J. Y. Beach, J. Am. Chem.
Soc., 60, 1836 (1938).
5
If there is a strong double bond such as:
(1) CHo
• • ^
CHotP::0
O M ••
ch3
the compound should have much weaker electron donor powers
than trimethylamine oxide. However, hyperconjugation1^
may play a role in the electron configuration in that the
methyl groups could yield electrons to the 3< * orbitals of
the phosphorus atom. The molecule then could be repre
sented by resonance forms of the following type:
(2) +
HCH2::P:6:
• ch3
Considering also the form
(3) CH3x®.
* • • • 'w'
CH3:P: 0:
• • • •
ch3
as in resonance, with types (l) and (2), one may argue the
electron donor properties of the oxygen from the probable
^ G. W. Wheland, "The Theory of Resonance," John
Wiley and Sons, Inc., New York, N. Y., 19^, PP*> 85-8 7.
6
importance of the forms (2) and (3) relative to (l). The
basic strength of (2) and (3) should be quite similar to
that of trimethylamine oxide or even a little greater
because of the greater electronegativity of nitrogen. The
electron donor strength of the oxygen would be expected
to decrease with an increasing degree of double bonding
between the phosphorus and the oxygen. From the experi
mental side, therefore, a comparison between the addition
compounds of trimethylphosphine oxide and trimethylamine
oxide should throw light on the resonance situation.
II. HISTORICAL BACKGROUND
Trimethylphosphine oxide was first prepared by
Cahours and Hoffman in l8571^ by air oxidation of tri
methylphosphine and by the action of heat on tetramethyl-
phosphonium hydroxide. This thermal decomposition was
duplicated in 1883 by Collie1^ who gave the true formula
as (CHg^PO. He reported also the m. p., 137-138°# and
the b. p., 214-215°.
* 1 Q
Sauvage in 1904 showed that phosphoryl chloride
^ A. Cahours and A. W. Hoffman, Annalen der Chemie
und Pharmacie, 104. 1-39 (1857).
N. C. Collie, J. Chem. Soc. Trans., 53# 636
(1883).
- 1 -® A. Sauvage, Compt. rend., 139. 674 (1904).
7
reacts with organo-magnesium compounds to form R3PO and
RgPOOH. Two years later Pickhard and Kenyon1^ published
a procedure for preparing alkyl compounds of this type.
An ether solution of phosphoryl chloride was added to the
desired alkyl Grignard compound, also in ether. After
addition of HC1 to dissolve the excess magnesium, the
ether was distilled off and the remaining solution was made
strongly alkaline with sodium hydroxide. The resulting
paste was then slowly distilled in a copper retort. The
aqueous portion of the distillate was fractionated to
yield the crystalline, highly hygroscopic alkyl phosphine
oxide.
The literature reveals that trimethylphosphine has
not been the object of Intensive study in recent years.
The reactions studied have been limited to the formation
of addition compounds in concentrated aqueous or alcoholic
media. Pickhard and Kenyon studied trimethyl, triethyl,
tripropyl, triphenyl, and tribenzylphosphine oxide com
pounds with various acids and salts. The trimethylphos
phine oxide compounds are given in Table I. For purposes
of comparison, Table II lists a number of addition com
pounds of trimethylamine oxide which are of interest in
^ R. H. Pickhard and J. Kenyon, Chem. Soc. Jour.,
8 9. Part I, 262 (1906).
TABLE I
ADDITION COMPOUNDS OP TRIMETHYLPHOSPHINE OXIDE
Ref. Addition Compound
Melting
Point Characteristics
(19) 2(CH3)3PO*H4Pe(CN)6
7
Small colorless crystals
from alcohol
2(CH3) 3P0»HAuC12 j 94.5°
Recrystallized from water
2(CH3) 3P0-H2Cr207 Darkens at
200°; melts
at 2040
Bright red prisms from
water or' dil. HN03
2 (CH3) 3PO • HBiljj. Dec. with
heat
Bright red prisms from
dil. HI
4(CH3) 3PO-H2PtCl6 l£6° Deep red pyramidal cryst.
2(CH3) 3P0-CgH4(C02H) 2
Trimethylphosphine
oxide camphorate
91-93°
Clear crystals from evap
orated alcohol solution
(ch3)3po«cci3cooh 67°
Colorless crystals from
conc. aqueous solution
2(CH3)PO«ZnI2 168° Small white prisms from
alcohol
4(CH3)PO•H3C o(CN)g•3H2Q
?
Colorless needles
TABLE II
ADDITION COMPOUNDS OP TRIMETHYLAMINE OXIDE
OF INTEREST IN THE PRESENT STUDY
Ref. Addition Compound
Melting
. Point Characteristics
(7)
2(CH3)NO-H2PtCl5 228-9° dec. Well defined rhombohedrons
sol. in H20 and methanol
(7) (CH3)3NO•HAuCl^
200° Yellow octahedral cryst.
sol. in hot H20
(1)
(CH3)3N0* 2S02
* >
Vapor press. 570 mm.
at 0°
(1,5)
(ch3) 3no-so2 120° dec.
l62-l64° dec. Sublimes in vacuo
recrystallized from absol.
alcohol
(2)
(ch3) 3no*bf3
89° Dec. at 175°-purified by
sublimation
10
the present study.
Although the compounds made by Pickhard and Kenyon
are Interesting additions to the chemical literature, they
were not formulated with the purpose of elucidating the
nature of the phosphorus-oxygen bond or for testing the
basic properties of the oxygen atom. Because they have
such a complicated electronic structure, it is difficult
to draw definite conclusions about them. It is possible to
say that the proton- of strong acids in non-aqueous or
concentrated aqueous solutions will react with trimethyl
phosphine oxide to form oxonium type compounds. It is
interesting to note that Pickhard and Kenyon found the most
usual formula for compounds formed by tertiary phosphine
oxides with acids to be 2R3POHX, while for trimethylamine
oxide it is R3N0*HX.
Using the cane sugar inversion method, Pickhard and
Kenyon determined that at 30° the salts of trimethylphos
phine oxide with trichloroacetic acid and cobalticyanic
acid were both about 8 8. 5$ hydrolyzed when compared to acid
solutions of the same strength. They stated that the tri
methylphosphine oxide in water solution does not behave as
a basic hydroxide, because it does not affect the birota-
20
tion of dextrose. Nylen in 1938 reported that no basic
20 Paul Nylen, Tids. KJemi. Bergvesen, 18, 48-50,
(1938); as in C. A., 32:8888.
11
properties for trimethylphosphine oxide could be detected
. pi
from potentlometric measurements at 20°. In 1941 Nylen
stated further that phosphine oxides were much weaker bases
than amine oxides and measured the dissociation of
(CH3) 3p°H+ in anhydrous propionic acid. In this acid sol
vent, trimethylphosphine oxide was about 40$ uncombined.
The weak basic properties in water were confirmed in a pre
liminary manner in the present study. To a dilute HC1
solution of pH 2.78 was added sufficient trimethylphos
phine to make an approximately 0.1 N solution. No change
in pH was noted.
A comparison of the relative stabilities of tri
methylamine oxide and trimethylphosphine oxide indicates
the rather greater polar character of trimethylamine oxide.
Trimethylphosphine oxide melts without decomposition at
137-1380 and can be distilled at atmospheric pressure at
214-215°. Trimethylamine oxide melts with decomposition
above 200°, and most of its compounds also melt with
decomposition or decompose at even lower temperatures.
The amine oxide itself apparently decomposes, yielding,
among other things, dimethylamine and formaldehyde. The
compounds of trimethylphosphine oxide usually decompose at
21 Paul Nylen, Z. anorg. Allgem. Chem., 246. 227-
242 (1941); as in C. A., 36:1295-
12
higher temperatures, and In general they seem somewhat more
stable.
It is evident from the compounds of triraethylphos-
phlne oxide listed in the literature that the fundamental
aspects of this field of study have been neglected and that
it may be possible in this preliminary work to point the
way for future intensive investigations.
In order to explore and understand the Lewis type
base reactions of trimethylphosphine oxide it is necessary
to choose fairly simple additives whose mode of addition
we know reasonably well. It is desirable either to carry
the reactions out in a solvent which will not enter into
the reaction, or to dispense with the solvents altogether.
If the compound which bonds to trimethylphosphine oxide is
volatile, the addition product may have an appreciable
dissociation pressure in a practical temperature range.
In that case, equilibrium constants, heats of reaction,
and free energy changes could be calculated, thus allowing
quantitative comparisons between compounds.
Boron trifluoride is aui excellent example of a very
strong Lewis type acid. With a coordination number that
can not exceed four, boron will bond only by accepting a
single pair of electrons from an electron donor.
In agreement with this principle, Burg and
13
Bickerton2 in 19^5 prepared (CH^^NO-BF^ melting at 89° and
beginning to turn brown at 100°, a temperature at which the
dative bond was too stable to permit simple measurement of
dissociation into the amine oxide and BF3. In .view of such
strong bonding, the analogous addition of one mole of BF^
to one mole of trimethylphosphine oxide should be observed.
Sulfur dioxide is another compound in which the
simple Lewis electronic formula would indicate an incomplete
octet, and accordingly it has considerable acid strength.
Burg"1 , found that at -80° (CHjJ^NO^SC^ was formed. The
dissociation pressure of the compound could be represented
by the equation log p ^ = 19-918 - . f i f f i * , ? . The vapor pres
sure at 0° was 570 mm. After removal S^f SO2 at -20°, the
compound (CHgJ^NO'SOg remained. This eofcoound could not be
dissociated by heating in vacuo below 120°, the charring
temperature. Lecher and Hardy-* found it to be unstable in
cold water but stable in hot absolute alcohol. The melting
point at atmospheric pressure was l62°-l64° with decompo
sition. The quantitative data available on the trimethyl
amine oxide- sulfur dioxide reaction, therefore, present a
good prospect for possible comparison with any compound
which might form between trimethylphosphine oxide and
sulfur dioxide.
The reaction between trimethylamine oxide and sulfur
trioxide has not been reported. However, Lecher and Hardy^
14
prepared triethylamine oxide-sulfur trioxide by the reaction
of triethylamine oxide with triethylamine-sulfur trioxide.
This compound decomposed in water, but was stable in abso
lute alcohol. As sulfur trioxide is a stronger electron
acceptor than sulfur dioxide, It would be expected to bond
more strongly to trimethylamine oxide. Even though a com
parison is not possible at present, it would be of inter
est to bring sulfur trioxide into reaction with trimethyl
phosphine oxide to obtain data for future reference. It
would be somewhat simpler to attempt the reaction by addi
tion of sulfur trioxide directly to the trimethylphosphine
oxide than to use the necessarily less direct process used
by Lecher and Hardy^ for the preparation of amine oxide-
sulfur trioxide.
CHAPTER II
PREPARATION AND PURIFICATION OF REAGENTS
Trlmethylphosphlne Oxide. The preparative method
was an adaptation of the Grignard-phosphoryl chloride
19
process described by Pickhard and Kenyon. The over-all
chemical reactions were as follows:
3CH3MgCl + P0C1 — ~^(CH3)3PO + 3MgCl2
j ^ gj» ^ ^
MgClg + Na2C03 — 5^MgC0 3 + 2NaCl
The .following procedure was selected after trying
several variations of the original process: one-half mole
of methylmagnesium chloride in 250 ml. ether (a k molar
solution commercially available) was diluted to one liter
with dry ether in the usual type of apparatus used for
Grignard reactions. The Grignard reagent was cooled with
ice-salt mixture and one-sixth mole (2 5 .6 g.) of phosphoryl
chloride in 200 ml. ether was added with stirring during
the course of one hour. The stirring was continued for
one-half hour, and then the ether was distilled off using
a steam bath.
The residual dry powder was dissolved in 250 ml. of
water, and the Grignard complex was hydrolyzed by the addi
tion of slightly more than one-half mole of sodium carbonate
16
in 500 ml. of hot water. The resulting thick paste of
magnesium carbonate was readily filtered off and washed
free of. the soluble .trimethylphosphine oxide on.a medium
porosity fritted’Pyrex filter. The filtrate was evapo
rated to about 50 ml.j; after which.4 the sodium chloride
precipitate was removed by filtration. The evaporation
was continued over a steam bath for seven hours. The few
ml. of dark yellow viscous solution and crystals remaining
were further dried in a calcium chloride desiccator for
five days. Trimethylphosphine oxide was extracted from the
residual crystals with three fifteen ml. portions of boil
ing chloroform. The chloroform was then removed from the
extract by carefully heating the solution on a hot plate
until the temperature of the residue reached 130°. Although
a small amount of the oxide was lost at this temperature,
considerable heat was required to drive off the last traces
of chloroform. The resulting crystals were evacuated for
two hours at water aspirator pressure in a vacuum desiccator
protected from back diffusion of water by a calcium chlor
ide tube. The yield of a slightly yellow colored product
was 8 g. (52# based on methylmagnesium chloride). Pickhard
and Kenyon1^ did not give yield data.
Purification of the crude product was carried out
most expeditiously by sublimation on the high vacuum line.
The crude oxide was placed in one of the removable tubes
17
of the apparatus shown in Figure 1. With a trap cooled
with liquid nitrogen between the sublimation apparatus and
the pump, the line was evacuated to about lO-^mm. for 0.5-
1.0 hour. A first fraction containing water and a little
of the oxide was collected in the trap. After the stop
cock between the sublimation apparatus and the line was
closed, the tube containing the crude material was heated
to 40-50° by a water bath while the receiver was cooled
with liquid nitrogen. Crystals which deposited in the
line between the two. tubes were flamed gently into the
cooled tube. If the tube containing the crude material was
heated too strongly, some of the yellow, non-volatile
Impurities were carried over mechanically.
The melting point of the trimethylphosphine oxide
prepared in this manner was 137* 5-138.5° • ' Pickhard and
Kenyon1^ reported 137-138°•
While carrying out Pickhard and Kenyon's original
procedure using methylmagnesium iodide and hydrolyzing the
Grignard with sodium hydroxide, it was found that the
separation of the oxide by distillation from the final
paste of magnesium hydroxide required higher temperatures
than the available (Pyrex) apparatus would permit; hence,
it was necessary to develop a different procedure for this
step. Filtration and centrifugation proved ineffective
for removing the large amounts of magnesium hydroxide, but
1 8
FIG. I
Sublimation Apparatus for (CH3)3P0
Dehydrite
Anhydrone
Ng
P4010
FIG.- g I
Apparatus for Preparing S03
and Adding it to (CH3)PO
19
after dilution with large amounts of water the clear super
natant liquid containing the trlmethylphosphine oxide could
be decanted. Neither distillation nor extraction with
immiscible solvents were satisfactory methods for isolation
of trlmethylphosphine oxide from this solution. Chloro
form was the only solvent which was found to extract the
oxide from water, but its distribution coefficient was
very unfavorable. It also extracted iodine compounds from
the solution which reacted with the oxide.to give a red
oil. Chloroform also extracted iodine compounds from the
dried residue of the supernatant liquid from the Grignard
hydrolysis. To eliminate the iodine interference, methyl-
magnesium chloride, from a commercial source, was substi
tuted for methylmagnesium iodide. The time consuming
decantation procedure was replaced by hydrolyzing the
Grignard complex with sodium carbonate which at the same
time precipitated the magnesium in a form which was
readily filtered.
Purification of Phosphoryl Chloride. The commercial
product was redistilled from an all-glass apparatus with
precautions for excluding water. The fraction boiling at
106.5-1 0 7.5° was collected.
Purification of Chloroform. Reagent grade material
was dried over calcium chloride and then distilled from
20
phosphoric anhydride. Moisture was excluded. The purified
chloroform decomposed considerably after a few days because
*
of the removal of the alcohol which acts as* a stabilizer
by removing phosgene as it forms.
Purification of- Ethylene Dichlorlde. A commercial
product was dried over anhydrous calcium sulfate and then
redistilled in moisture-free, all-glass apparatus.
Purification of Methylene Chloride. Commercial
grade was dried over phosphoric anhydride and distilled in
moisture-free all-glass apparatus.
Purification of Benzene. Reagent grade was treated
as above.
Purification of Toluene. Reagent grade was treated
as above.
Purification of Skellysolve A. The commercial
material was treated as above.
Purification of Ether. Sodium wire was added to
reagent anhydrous ether.
Purification of Ethyl Alcohol. A newly opened bottle
of commercial anhydrous alcohol was refluxed over calcium
oxide and redistilled.
21
Preparation of Boron Trlfluoride. Two different
sources were used. The first preparation of boron tri
fluoride was by the reaction of sulfuric acid on ammonium
fluoborate, with boron oxide to remove hydrogen fluoride.
op
This method was much as described by Booth and Willson,
but the apparatus was similar to that used by Bickerton,^
with the addition of a dry-ice reflux head sealed on above
the water-cooled reflux condenser.
The second source was a commercial cylinder of boron
trifluoride. Flow from the cylinder was regulated by two
needle valves placed in series after the cylinder valve.
The boron trifluoride was purified by bubbling through
concentrated sulfuric acid saturated with boron oxide, then
passing through a tube containing glass wool and boron
oxide to remove any acid spray. In order to keep the
needle valve system in good condition, it was found neces
sary to disassemble it after use, rinse with distilled
water, followed by alcohol, and then dry It by passing air
through.
Preparation of Sulfur Dioxide. Gas from a cylinder
was passed through Drierite and phosphoric anhydride and
condensed in the vacuum line. The gas was further purified
22 H. S. Booth and Willson, “Inorganic Synthesis,”
McGraw-Hill Book Co., Inc., New York and London, 1939> p. 21.
22
by passing from a 0° tube through a -80° dry-ice-ether trap
into a liquid nitrogen cooled trap. This procedure was
repeated three times, the -80° fraction being retained for
the succeeding fractionation. The purified material was
stored in a large bulb at approximately atmospheric pres
sure.
Preparation of Sulfur Trioxide. Sulfur trioxlde
was prepared in the apparatus shown in Figure 2. After
the apparatus had been gently flamed while dry nitrogen
was passing through, sulfuric acid (10-15 ml., 30$ fuming)
and phosphoric anhydride (3-5 g-) were placed in flask B.
The mixture was heated gently until the required amount of
sulfur trioxide had collected in the trap C at 0°.
CHAPTER III
THE REACTION OP TRIMETHYLPHOSPHINE OXIDE
WITH BORON TRIFLUORIDE
The combination of boron trifluoride and trimethyl-
phosphine oxide was expected to be similar to the reaction
between boron trifluoride and trimethylamine oxide, and so
the apparatus described by Bickerton^ was used with only
minor changes. Excess boron trifluoride at the exit of
the apparatus was removed by a water aspirator with an
interposed T-tube with one side open to prevent evacuation
of the system. Connections between various parts of the
set-up were made with Neoprene tubing which is resistant
to boron trifluoride.
Because of the extreme reactivity of boron,tri
fluoride with moisture and the hygroscopic nature of
trimethylphosphine oxide, precautions were necessary to
exclude water from the apparatus. Within the dry box, a
0 .2 7 2 g. sample of trimethylphosphine oxide was placed in
the tared reaction tube. A five ml. portion of anhydrous
chloroform, which Bickerton had found to be a good solvent
for trimethylamine oxide in the similar reaction with
boron trifluoride, was added. The tube was then placed on
the system which had previously been filled with dry
24
nitrogen. After nitrogen was passed through again for a
moment, it was discontinued and boron trifluoride was
introduced from the generating flask. Flow through the
chloroform solution was regulated at about two bubbles per
second. A turbidity was noted in a few moments, and after
fifteen minutes the flow of boron trifluoride was discon
tinued and nitrogen was passed in until no more boron
trifluoride could be seen issuing from the exit* The
chloroform was removed from the fine white floating pre
cipitate by means of a filter stick. The reaction tube,
during the removal of chloroform and the subsequent wash
ing with two 3 ml* portions of chloroform and a 5 ml.
portion of ether, was protected with a calcium chloride
drying tube.
The product remaining, after drying with suction
and placing in a calcium chloride desiccator for twenty-
four hours, weighed 0.295 g* It was not possible to
calculate the per cent yield at this time, but after the
formula of the compound had been established as
(CH^J^PO-BF^, the yield was found to be 68$ based on tri
methylphosphine oxide. The crude material melted at 113-
125° and was evidently different from trimethylphosphine
oxide, because it was not deliquescent.
Experiments indicated that the material was stable
towards heat; hence sublimation was Investigated as a
25
means of purification. The crude product was placed in a
simple sublimation tube made from the closed end of a Pyrex
test tube which was sealed to a long piece of 8 mm. tubing.
The sublimation tube was connected to the vacuum line in a
vertical position through a Picein Joint, and a trap cooled
with liquid nitrogen was placed between the tube and the
pump. After cooling the tube with liquid nitrogen, and
-4 -5
evacuating to 10 -10 , the stopcock leading to the pump
was closed. The tube was then headed slowly in an oil bath
which covered it to a point about 1 cm. above the entrance
to the 8 mm. tube. At a temperature of 117°, a small amount
of material started to collect in the cool part of the tube.
As the temperature rose to 130-l40°C., the material all
sublimed slowly above the level of the oil bath. Nothing
visible collected in the liquid nitrogen trap, and when the
trap was warmed to room temperature, no change was observed
in.a manometer connected to the system. Only a trace of
solid was left on the bottom of the sublimation tube.
Apparently the material did not decompose and no gases
were evolved.
The melting point of the sublimed material was 148-
150°C. No appreciable difference in melting point was
noticed if it was taken in tubes which were evacuated to
about 20 mm. pressure and sealed off or in tubes which were
open to the air. It could be melted, cooled and remelted
26
to give the same melting point. It could be heated to at
least 215°C. without apparent change.
To test its stability in cold water, about 10 mg.
of the compound was placed in a small platinum boat and
dissolved in a little water. The boat was then placed in
a desiccator and the water was allowed to evaporate at room
temperature. After one week, the water had evaporated,
and the compound was apparently dry. The melting range of
the residue was l40-l45°C. The boat and its contents were
placed in the sublimation tube and sublimed, as before. No
noticeable residue was left in the boat. The sublimate
melted at 148-150°.
Analysis of the purified material for boron by the
method of Bickerton,3 and earbon-hydrogen analyses by Dr.
Adalbert Elek2^ gave the following results which indi
cated a formula (CH^J^PO'BF^: Boron 4.97 (Calc*d. 4.77$);
Carbon 22.60$ (Calc’d 22.53$); Hydrogen 5. 67$ (Calc'd.
5.67$).
As a further check on the composition and to see how
vigorous the reaction was, boron trifluoride and the oxide
were combined in the absence of solvent. The apparatus
shown in Figure 3 was set up in the hood for this purpose.
23 pr. Adalbert Elek, Elek Microanalytical
Laboratories, Los Angeles.
Drierite
FIG". &
Apparatus for the Reaction of
(CHg)gPO with BF3
28
The boron trifluoride used in this experiment was obtained
from a tank and diluted with about three or four times its
volume with dry nitrogen. The boron trifluoride passed
through the sulfuric acid-boron oxide trap at the rate of
about two small bubbles per second.
After filling the set-up with dry nitrogen, reac
tion tube I was removed, quickly stoppered and weighed by
suspending it from the balance pan hooks by the wire
shown attached between the two traps of the reaction tube.
Within the dry box, trimethylphosphine oxide was added to
side A. The tube was weighed again and replaced on the
apparatus. After passing dry nitrogen through the system
again for a moment, the mixture of boron trifluoride and
nitrogen was allowed to flow over the trimethylphosphine
oxide.
The reaction tube was cooled with water, but the
reaction was not vigorous. After 45 minutes, the boron
trifluoride was shut off and the remaining boron tri-
fluoride was swept out of the apparatus with nitrogen.
The stoppered reaction tube was weighed. It was replaced
and the boron trifluoride-nitrogen mixture passed through
for one-half hour while the reaction tube was heated to
110-120°. After replacing the boron trifluoride with
nitrogen, the tube was cooled and weighed. Small droplets,
apparently of sulfuric acid, had collected on the inside
29
of the reaction tube near the inlet. The compound had not
changed much in appearance except that the particles seemed
slightly fused together. The results are shown in Table
III.
The weight relations in Table III are further evi
dence of a 1:1 addition compound. The gain in weight after
the additional treatment with boron trifluoride at the
higher temperature was probably due to the deposit of liquid
near the inlet of the reaction tube.
The compound made in this manner had, after sub
limation, a melting point of 148-150°.
30
COMBINING
TABLE III
RATIO OF (CH3)3PO AND BF3
First Weighing
After heating 1/2
hour at 110-120°
wt. (ch3)3po
0 .2927 S-
0.2927 g-
Moles (CH3)3PO
3.18 X 10"3 3 .1 8 X 10"3
Wt. after reaction 0.5092 g.
0.5194 g.
Wt. BF3 gained 0.2165 g. 0.2267 g.
Moles BF3 gained 3.19 X 10~3 3 .3 0 X 10“3
Ratio BF3 to (CH3)3PO 1.003 1.037
CHAPTER IV
THE REACTION OF TRIMETHYLPHOSPHINE OXIDE
AND SULFUR TRIOXIDE
The reaction between trlmethylphosphine oxide and
sulfur trioxide was carried out in the hood in the appara
tus shown in Figure 2. No grease was used in the stop
cocks or joints and moisture was rigidly excluded. Neo
prene was satisfactory for any non-glass connections.
While nitrogen dried by passing over "Anhydrone” and
phosphoric anhydride was passed through the empty apparatus,
all parts except stopcocks and the few rubber connections
were heated carefully to about 100°. The nitrogen flow was
continued for two hours thereafter to remove all possible
water.
The reaction was tried under several different
conditions;
1. Without solvent, with excess sulfur trioxide at
room temperature.
t
2. Without solvent, with excess sulfur trioxide at
70°.
3. With solvent, with excess sulfur trioxide at
room temperature.
4. With solvent, with excess trimethylphosphine
oxide at room temperature.
32
Preparation Method 1_. A two-gram portion of sulfur
trioxide was distilled into the 0° trap C, as outlined in
Chapter II. Stopcock 2 was open to the atmosphere through
the Dehydrite tube, while stopcock 3 was closed. Reaction
tube I was removed, stoppered, and weighed. In the dry
box, trimethylphosphine oxide was added to part D. The
tube was again stoppered, weighed, and replaced on the
apparatus.
Tube C was brought to room temperature, flask B was
stoppered at A, and nitrogen was passed into the system
through stopcock 1 at about 25-50 ml. per minute. Stop
cocks 3 and ^ were open and 2 was closed. Sulfur trioxide
was thus carried in a stream of nitrogen over the oxide and
out through stopcock 4. The exit tube was bent downward
to prevent sulfuric acid from running back into the reac
tion tube.
The reaction between sulfur trioxide and trimethyl
phosphine oxide apparently evolved only a small amount of
heat. There was no charring, but the oxide could be seen
to change in texture and swell slightly. The mixture of
sulfur trioxide and nitrogen was passed over the oxide for
forty-five minutes at room temperature. Three times during
the run, sulfur trioxide was frozen out of the gas on to
the trimethylphosphine oxide with dry ice, after which it
was allowed to evaporate again. When all the sulfur
33
trioxide had evaporated from C, the reaction tube D was
heated to 60-70° while nitrogen passed through the system
to take out excess sulfur trloxide. After 45 minutes, no
more sulfur trioxide was issuing from the exit tube and
so reaction tube I was removed, stoppered, and weighed.
About a milligram of the reaction product was taken
from the tube for inspection. It fumed and quickly picked
up moisture from the air. To determine whether more sul
fur trioxide could be removed, the tube was then connected
to the vacuum line through a liquid nitrogen trap, heated
to 140° G. and evacuated for one-half hour. There was no
apparent effect on the material, but some sulfur trioxide
was condensed in the trap. The tube was filled with dry
nitrogen, stoppered, and weighed. The results are given
in Table IV. The weight relationships indicate a 1:1
addition product. The material did not fume but it still
rapidly picked enough moisture from the air to dissolve.
It had a very light tan color and apparently was crystal
line.
The melting point of the material was taken in
specially prepared capillaries which had been heated to
the' sodium-flame temperature and sealed while under
aspirator vacuum through a calcium chloride drying tube.
The capillaries were opened in the dry box, the material
added, closed by a plug of paraffin, and then sealed in a
34
COMBINING
TABLE IV
RATIO OP (CH3)3PO AND
CO
0
CO
Method 1 Method 3
Conditions of No solvent; Solvent;
Reaction Excess S03;
Room Temp.
Excess S03;
Room Temp.
Wt. (CH3)3PO 0.2788 g.
0.169 g.
Moles (CH3)3PO 3 .0 3 X 10“3 1.84 X 10“3
Pinal Wt.
0.5329 g. 0.335 g.
Moles S03 gained 3 .1 7 X 10“3 2 .0 7 X 10”3
Moles S03/(CH3)3P0
1.05
1.12
35
small flame below the paraffin plug. The melting range of
the crude product was 180-210°. The material In both open
and sealed tubes could be heated to 250° without apparent
change. It was interesting to note that the little column
of the material in the capillary tube seemed to melt at
the top first, as if it were dissociating slightly at that
point.
Studies Toward Purification of the Product. The
melting range of the material indicated that further puri
fication was needed. Because of the previous success in
purifying the boron trifluoride addition product by sub
limation, the same procedure was tried on the sulfur tri
oxide addition product. In this case, however, the results
were not satisfactory. The apparatus used for the purifi
cation of the boron trifluoride addition product was again
employed. The tube was heated to 190° at 10"1*'-10"’^mm.
before the material melted and slowly redeposited in the
cooler narrow neck. At 200°C., it evaporated very quickly,
almost as if it had boiled or decomposed, and was deposited
in a crystalline ring in the neck. Slightly above the
crystalline material was a viscous liquid suggestive of
sulfuric acid.
The melting point of the sublimed material was taken
in the same manner as the unsublimed. The range was 110-
36
135°C., indicating that it might have been impure trimethyl
phosphine oxide. There was no evidence of any decomposi
tion into a highly volatile gas. The melting range was
the same in open or closed capillaries.
> The extremely hygroscopic nature of the reaction
product complicated the purification by recrystallization.
Lecher and Hardy^ were able to purify triethylamineoxide-
sulfur trioxide by washing it with cold absolute alcohol
in which it was insoluble. Working in the dry box, a
similar procedure was attempted with the trimethylphosphine
oxide compound. It dissolved readily without apparent
reaction. Addition of dry ether caused the precipitation
of beautiful colorless needles. The melting point of this
very hygroscopic material, observing the usual precautions,
was 87-88°. The hypothesis that the alcohol had reacted
with the -sulfur trioxide to form ethyl sulfuric acid which
reacted with the trimethyl phosphine oxide to form the
salt, or trimethylphosphoxonium ethyl sulfate, according
to the following reaction
(CH3)3P0-S03 + c2h5oh ---^ (CH3)P0H0S02OC2H5
was confirmed by the analytical data given in Table V.
24
Traube and Zander report a similar reaction between
W. Traube and H. Zander, Ber., lo45 (1924).
37
TABLE V
ANALYSIS OP THE PRODUCT FORMED BY THE
SULFUR TRIOXIDE-TRIMETHYLPHOSPHINE
OXIDE ADDITION COMPOUND AND C2H50H*
$ Carbon $ Hydrogen
$ Ethoxy
OC2H5
By analysis
27.15$
6.46$ 2 0. 18$
Theory for
(CH3) 3PO • HSO3OC 2Hc> 27.51$
6.94$ 2 0.20$
$ Dev. from
theory
-1.3$
~6. 9$ -0:i$
* These analyses were performed by Joseph Pirie of the
University of Southern California Chemistry Department.
38
trimethylamine-sulfur trioxide and alcohol.
Preparation Method <2. In an attempt to get a purer
initial product, the reaction between SO3 and trimethyl
phosphine oxide was carried out at 70°C. without the use.
of solvent. The reaction went forward without incident.
However, the product was slightly darkened and charred in
a few small spots. It appeared as if a small spot of
sulfuric acid formed and then the charring spread outward
from it. During the evacuation at 60° to remove excess
sulfur trioxide, the apparatus apparently leaked and
allowed some moisture to enter, for the product became
slightly moist. The weight gained was about 27$ higher
than would be accounted for by a 1:1 compound. As the
product was discolored anyway, this particular modification
of preparation was not investigated further.
Preparation Method 3, . Ethylene dlchloride, which
was inert to sulfur trioxide, was found to be a satisfac
tory and non-reactive solvent for trimethylphosphine oxide.
Trimethylamine oxide removes HC1 from ethylene dlchloride.
A preliminary experiment showed that the sulfur trioxide
addition>product formed in the solution and precipitated.
After the usual preparation of the apparatus,
O.I69 6- of trimethylphosphine oxide was added to part D
of the reaction tube, and dissolved in 6 ml. of ethylene
39
dichloride. The tube was replaced on the apparatus and
about 2 g. of sulfur trioxide in nitrogen was passed over
the solution in the course of an hour. Twice sulfur tri
oxide was frozen out upon the solution by means of dry ice,
to insure contact. After the reaction, the remaining
ethylene dichloride was evaporated off by aspirator vacuum
through a dry-ice trap and a calcium chloride tube. The
reaction tube was weighed after evacuating at lO'^-lO-^ mm.
for 10 minutes on the high vacuum line, and then again
after 45 minutes of evacuation. The second high vacuum
treatment removed no more weight. The results are given
in Table V.
The product from this reaction was almost white.
Its melting range was l65-190°C.
Preparation Method 4. The previous experiments had
indicated that slightly more sulfur trioxide was being
absorbed than would account for a 1:1 compound. An attempt
to carry out the reaction.using an excess of triraethyl-
phosphine oxide seemed logical. The excess oxide could
then be removed by solvent wash and sublimation. In
addition, formation in a solvent seemed to give a better
looking product.
About 0.3 g. of trimethylphosphine oxide was dis
solved in 2 ml. of ethylene dichloride in part D of the
4o
reaction tube. Sulfur trioxide was passed through the
apparatus in a slow stream of nitrogen. Twice a small
amount of sulfur trioxide was frozen out of the gas stream.
After about one-half hour the nitrogen supply was changed
to stopcock 2, and the sulfur trioxide in tube C was cooled
by dry ice. Nitrogen was passed through the system until
no more sulfur trioxide fumes issued from the exit. The
remaining solvent was evaporated with aspirator vacuum.
The tube was removed to the dry box where the residue was
washed twice with 3 nil. of ethylene dichloride. After
evaporation, the wash solutions left large crystals of
trimethylphosphine oxide showing that there had been an
excess present. The rest of the volatile materials were
removed on the high vacuum line for three hours at 30-35°.
After this time, no more material could be frozen out by
liquid nitrogen on the line leading to the vacuum system.
The yield of compound was 0.265 g. of a very white fluffy
material. The melting range was l65-200°C.
The unsublimed (CH^^PO-SO^ compound was extremely
water-soluble giving an acid solution. A precipitate
formed instantly when barium chloride was added to the
solution to which a few drops of nitric acid had been
added. Hydrolysis was shown to be very rapid. A 5 mg.
sample of the material was added to about 5 ml. of water
having an original pH of 6.9* In less than five seconds
41
the pH was 2.1, constant during ten minutes. The titra
tion curve of 5*648 nig. in approximately 5 ml* of water
with 0.01948 N carbonate-free sodium hydroxide is shown in
Figure 4. The titrations were performed with a syringe
type microtitrator driven by a micrometer in which one
division-on the scale indicated a displacement of 2.761
mm.3 by the syringe plunger. The pH was determined with a
Beckman pH meter using microelectrodes.
Considering the neutralization point to be at pH
7. 0, the equivalent weight was calculated to be approxi
mately 91 g* per mole. The equivalent weight of
(CH^^PO'SO^* assuming complete hydrolysis to sulfuric
acid, would be 8 6. The high value could easily be due to
water absorption during handling and weighing.
Carbon and hydrogen analyses on the material from
Method 4 were performed by Dr. Elek.2^ nis results are
given in Analysis #1 in Table VI. If a 1:1 addition
compound is assumed, the value for carbon is low and the
value for hydrogen is high by about the amount that would
be caused by approximately 5*5$ water as an impurity. As
facilities for handling such a deliquescent material were
"not available, this result was to be expected. At a later
date, analyses were carried out by Joseph Pirie J with
25 Joseph Pirie, of the University of Southern
California Chemistry Department.
FIG. 4
TITRATION OF (CH3)3P0-S03 WITH 0.01948 N. NaOH
9
8
7
6
5
4
3
2
i.o 2.0 3,0
43
TABLE VI
ANALYSIS OP SULFUR TRIOXIDE-TRIMETHYLPHOSPHINE
OXIDE ADDITION COMPOUND (METHOD 4)
Carbon
Analysis*
#1 -
Hydrogen
Analysis*
#1
Carbon
Analysis**
#2
Hydrogen
Analysis**
#2
By Analysis 19-92# 5. 68# 2 0. 32# 5.40#
Theory for
(CH3)3P0-S03
20.93# 5.27# 20.93# 5.27#
# Dev. from
Theory
-5.1# +5-9# -2.9# +2.5#
* These experiments were performed by Dr. Adalbert Elek,
- of Elek Microanalytical Laboratories, Los Angeles.
** These experiments were performed by Joseph Pirie, of
the University of Southern California Chemistry
Department.
44
greater safeguards against water absorption. A small tube
of platinum foil about 4 mm. in diameter and closed at one
end was weighed, filled in the dry box, pinched closed,
weighed again, and then placed in the combustion, tube.
The; end of the platinum tube was opened after it was in
position and in a current of dry oxygen. The latter
analyses are also shown in Table VI as analysis #2. In
this case again, the carbon is a little low and the hydro
gen is a little high, showing the probable presence of
traces of water. The results are acceptably close to the
theoretical value for a 1 :1 compound, especially in view
of the figures for weight gain and equivalent weight.
CHAPTER V
THE REACTION OF TRIMETHYLPHOSPHINE OXIDE
AND SULFUR DIOXIDE
A preliminary experiment in passing sulfur dioxide
over trimethylphosphine oxide indicated a vast difference
between the reactivity of trimethylphosphine oxide with
sulfur dioxide, and the reactivity of trimethylamine oxide
with sulfur dioxide. Instead of a violent reaction as in
the latter case, with the formation of a solid addition
product, trimethylphosphine oxide merely absorbed sulfur
dioxide and liquefied with little evolution of heat.
Evaporation of the sulfur dioxide from this liquid left
the original unchanged trimethylphosphine oxide whose
melting point was 136-1380, compared with its original
melting point of 137-138.5°C.
To study the possibility of weak compound formation,
a series of composition-vapor pressure measurements at 18°,
8°, 0°, and -23° were made. The apparatus used is shown
in Figure 5. ha" is the storage reservoir for sulfur
dioxide. "B" is a calibrated gas burette and leveling
bulb for pressure measurements. "C" is the reaction tube
which held the trimethylphosphine oxide in the small bulb
in the bottom which could be immersed in slush baths of
the desired temperatures. The temperature of -23° was
46
FIG. 5
Apparatus for Treating (CHg)gPO with SOg
47
attained by a carbon tetrachloride slush, 0° by an ice and
water slush, 8° by an ethylene dibromide slush, and 18° by
a benzophenone slush. Although these temperatures are not
necessarily the accepted melting points of the.compounds,
by proper addition of a few drops of liquid nitrogen now
and then the temperatures could be maintained at +0.2° for
the required length of time.
The volume of the empty reaction tube C and the
connecting tubing was calculated by pressure-volume rela
tions using a known amount of sulfur dioxide which had
been measured in the gas burette. In two experiments, the
volume was found to be 26.12 and 26.01 ml. The average
was 26.06 ml. The volume of the trimethylphosphine oxide
was neglected.
Using precautions to prevent moisture contamination,
—3
and working in the dry box, 0.1097 g« (1.192 X 10 moles)
of trimethylphosphine oxide were placed in the bottom of .
the reaction tube by means of a long-stem funnel. After
sealing the reaction tube to the system with "Varno”
cement, the tube was cooled to -80°C. and evacuated. By
proper manipulation of the 3-way stopcocks 1, 2, and 3 and
the mercury level, small amounts of sulfur dioxide could be
removed from the reservoir and measured in the burette.
The differences between mercury levels in the burette and
the leveling bulb were measured with a cathetometer.
Turning stopcock 3 open to the reaction tube allowed the
sulfur dioxide in the burette to react with the trimethyl
phosphine oxide. The leveling bulb was then adjusted so
that a convenient volume, usually 10 -1 5 ml., was main
tained in the burette.
In general, 30-45 minutes were allowed for attain
ment of equilibrium after each addition or removal of
sulfur dioxide. In any case, however, two pressure read
ings at least five minutes apart had to agree within 0 .1
mm. before the final reading was made. The pressure
readings corresponding to small amounts of sulfur dioxide
in comparison to the trimethylphosphine oxide were taken
after a period of 2 -3 hours, and then the pressures still
seemed to be too low. The attainment of equilibrium
appeared difficult under these conditions, but was aided
greatly by agitating the reaction tube with the finger.
The composition of the trimethylphosphine oxide-
sulfur dioxide phase was calculated from the difference
between the known amount of sulfur dioxide which had been
added and the amount calculated to be present in the
gaseous phase. The vapor pressure of the trimethylphos-
r
phine oxide, being very small at these temperatures, was
neglected.
The data obtained above are given in Tables VII,
VIII, and IX. Although the values for the mole ratios of
49
TABLE VII
PRESSURE-COMPOSITION ISOTHERMS OP SYSTEM
(CH3)3P0-S02
1.192 X 10~3 moles (C^^PO IN SYSTEM
l8°c.
Mole ratio of
S02 to (GH3)3PO
in cond. phase
Equil. S02
Press,
ram. Hg
8°C.
Mole ratio of
S02 to (CH3) 3P0
in cond. phase
Equil. S02
Press,
mm. Hg
0.203 10 3 .2
0.253
7 6 .0
0.463 10 6 .2 O.525
76.9
0 .7 6 2 1 0 6 .0 0.811 7 6 .1
O .863 107.1 0.875 76.3
0.913
106.8
0.979 77.1
0.978
123.5 1.03 92.5
0.988 1 2 7 .0 1.05 93.5
0.994 124.2
1.31
145.0
1.075
154.0 1.36 16 9.O
50
TABLE VIII
PRESSURE-COMPOSITION ISOTHERMS OF SYSTEM
(CHq)3P0-S02 AT 0°C.
1.192 X 1 0 -3 moles (CH3)PO IN SYSTEM
Mole ratio of Equil. S02 Mole ratio of Equil. SO2
S02 to (CH3)3PO Press. SO2 to (C^^PO Press,
in cond. phase mm. Hg In cond. phase mm. Hg
0.105
0 .2 8 9
0.567
0.847
0.9H
1.013
1.058
1.093
1.100
1.108
1.173
50.4
53.0
53-3
55.0
5|.3
57^1
6 7 .8
6 7 .6
6 7.0
77.9
1.188 8 6 .0
1.2 5 3
93.0
1 .3 2 9 108.3
1*3XS
10 7 .8
1.388 122.2
1 .4 5 4
137.7
1.473
145.0
1 .5 1 3
154.1
1 .7 0 7 202.7
2.91 450.3
3.24
583.3
• *
51
TABLE IX
PRESSURE-COMPOSITION ISOTHERMS OP SYSTEM
(CH3)3PO-SO2 AT -23°C.
1.192 X 10“3 moles (CH^PO IN SYSTEM
Mole ratio of Equil. SO2
S02 to (CH3)3PO Press.
in cond. phase 1 1 1 1 1 1 •
0.1 5 9
18.4
0.624 1 8 .8
0.9 1 0
19.1
0.971 19.3
1 .0 1 8 19.4
1.082
19*3
1 .1 5 0 2 1 .8
1 .1 8 8 2 3 .6
1 .2 8 0
30.3
1.560 47.0
52
sulfur dioxide to trimethylphosphine oxide are arranged in
the order of their increasing magnitude, each determina-
tion was not necessarily made in sequence. Some of the
values were obtained after several additions and subtrac
tions of sulfur dioxide to the system. The data are also
shown graphically in Figure 6. The points corresponding
to the mole ratios of 2 .9 1 and 3*24 on the 0° isotherm
were not plotted as they extended the curve to an incon
venient size. These two points, however, fall on the
smooth extension of the 0° curve. No other inflections
were in evidenc e.
The first isotherm determined was at 0°. Solid,
presumably trimethylphosphine oxide, was present in the
reaction mixture during the interval of constant pressure.
With the addition of sulfur dioxide, the amount of solid
decreased and the liquid increased. At the approximate
mole ratio of 1:1, the last visible traces of solid dis
appeared and the sulfur dioxide pressure began to increase
rapidly after each addition. The rate of increase, however,
was only a very small fraction of what would be expected
if the sulfur dioxide were insoluble in the reaction
product. It was also much less than if it were dissolving
in the reaction product and obeying Raoult's Law. For
example, at 0° with the mole ratio of sulfur dioxide to
trimethylphosphine oxide equal to 1.1, Raoult's Law would
b‘ -
VAPOR mESSm-E-COMPOSITIOU
of
so2-(ch5)spo mixtures
iso'
#
flG. 6
54
give a vapor pressure of sulfur dioxide equal to 802 ram.
The experimental pressure was 67.6 mm., which was only
8.4$ of the theoretical.
These findings led to the idea that perhaps instead
of the formation of a 1:1 molecular addition compound,
there might have been formed merely a saturated solution
of trimethylphosphine oxide in sulfur dioxide (deviating
extremely from Raoult’s Law), which at 0° happened to have
a 1:1 composition. Therefore, the data for 18°, 8°, and
-23° were obtained to determine whether or not there was a
change in composition, at the rise-point of the vapor
pressure curve, with a change in temperature. As can be
seen from Figure 6, the rise-point shifts from a mole ratio
of about l.l at -23° to about 0.9 at l8°C.
The averages of the vapor pressure values on the
flat portion of each isotherm are given in Table X. Vapor
pressures corresponding to mole ratios of 0.2 and less
were not included because of the difficulty in attaining
equilibrium at the lower concentrations of sulfur dioxide.
The approximate range is from mole ratios of from 0.2 to
1.1. Some solid was present during the flat portion of
the curve at each temperature, as would be expected from a
consideration of the phase rule. At constant temperature,
the presence of three phases would require constant
pressure.
55
TABLE X
VAPOR PRESSURES OF S02-(CH3)3P0 MIXTURES
IN CONSTANT PRESSURE RANGE
MOLES S02/(CH3)3P0 APPROXIMATELY 0.2-1.1
Temp. Ave. experimental
S02 Press.
Calculated S02
Press.
-23°C. 1 9 .2 mm. 19.4 mm.
0°
54.5
54.1
8° 76.6
74.5
18°
106,5
108.0
5 6
The vapor pressure-temperature relation can be
expressed within less than 2.7$ error in the range studied
by the equation
log Pmm. « - + 6.592
The values calculated by means of this equation are shown
in Table X.
CHAPTER VI
DISCUSSION AND CONCLUSIONS
I. THE REACTION OP BORON TRIFLUORIDE
WITH TRIMETHYLPHOSPHINE OXIDE
Evidence for the formation of the white crystalline
compound (CH^^PO’BF-^, m. p. 148-150° was quite conclusive.
The formula was indicated both by the reaction balance
(1:1 combination of the two reactants), and by analytical
data for carbon, hydrogen, and boron. There was no evi
dence of other addition compounds under the conditions
studied, but they may exist at lower temperatures.
The bonding stability of the compound was appar
ently comparable to that of (CH3) 3NO*BF3. It could be
purified readily by sublimation in a high vacuum system,
and was soluble and stable in cold water. The addition
compound was not appreciably hygroscopic. Both trimethyl
phosphine oxide and trimethylamine oxide held boron tri
fluoride so firmly that it was not possible to say from
present information which was the stronger base.
II. THE REACTION OF SULFUR TRIOXIDE
WITH TRIMETHYLPHOSPHINE OXIDE
Sulfur trioxide and trimethylphosphine oxide, when
58
combined under a variety of conditions yielded a substance
which combining weights, elementary analysis, and acid
equivalent weights indicated to be (CI^^PO’SOj. The
compound was extremely hygroscopic and hydrolyzed to form
sulfuric acid and trimethylphosphine oxide.
Efforts to purify this compound by sublimation, and
preliminary attempts at recrystallization, were not suc
cessful. All batches of the material as prepared under
the four conditions studied had about the same melting
range of 165-210°. This wide melting range may be due
either to an inert impurity or to decomposition. Repeated
' melting points with the same melting point capillary gave
about the same range, indicating that heat did not cause
extensive permanent chemical change.
A direct comparison between the basic strengths of
trimethylphosphine oxide and trimethylamine oxide was not
possible in this case, as the analogous trimethylamine oxide
compound with sulfur trioxide has not been reported in the
literature. However, in respect to absolute ethyl alcohol,
the compound (C2H5)3N0*S03, as reported by Lecher and
Hardy,5 was more stable than (CH3)3P0*S0 3. This was shown
by the formation of the trimethylphosphine oxide salt of
ethyl sulfuric acid, (CH3)3P0.H0S020C2H5(m. p. 87-88°) in
alcoholic solution. The triethylamine oxide-sulfur
trioxide addition compound did not react under similar
59
conditions. If it is assumed that the basic properties of
trimethylamine oxide are similar to those of triethylamine
oxide, it might be inferred, at least qualitatively, that
trimethylphosphine oxide should be a weaker base towards
sulfur trioxide than is trimethylamine oxide.
III. THE REACTION OF SULFUR DIOXIDE
WITH TRIMETHYLPHOSPHINE OXIDE
Present evidence is insufficient to determine with
finality whether a liquid compound (CH3)3P0*S02 or a
saturated solution of trimethylphosphine oxide in sulfur
dioxide with a composition of approximately 1:1 was formed.
The fact that the rise-point of the vapor pressure-composi-
tion curve shifts from a mole ratio sulfur dioxide to
trimethylphosphine oxide of 1.1 at -23° to about 0.9 at
18° suggests the latter possibility, and that the solu
bility of trimethylphosphine oxide decreased with decreas
ing temperature.
On the other hand, consideration of the equilibrium
pressures at the rise-points indicates that if there is no
definite compound formation, there is, at least, extreme
solvation of trimethylphosphine oxide. Table XI, which
compares the equilibrium pressures calculated from Raoult's
Law with observed pressures, clearly shows this fact.
6o
TABLE XI
EQUILIBRIUM PRESSURES OF (ch3)3po-so2 SYSTEMS
Temp. Mole Fract.
S02
Calculated
press.
Observed Ratio
press. Observed/cale*d.
18° O . W 1440 mra. 106.8 mm. 0.0742
8° 0.495
1035 ram* 77*1 mm. 0.0745
0°
0.503 769 ram. 54.3 mm.
0.0708
-23°
0 .5 2 0 213 ram. 19.3 ram. 0.0906
6l
As stated previously, the large negative deviations
from Raoult’s Law were also present in the unsaturated
portions of the curves.
The isolation and analysis of the liquid reaction
product would have given a clearer picture of its nature,
but the task was not undertaken in the present study.
In comparing trimethylphosphine with trimethylamine
oxide, it can be stated with a fair degree of certainty
that trimethylamine oxide is a stronger electron donor to
sulfur dioxide than is trimethylphosphine oxide. Burg1
found that the compound (CH^J^NOSO^ was extremely resis
tant to dissociation, for its formation was not reversed
by heating in vacuo. A second mole of sulfur dioxide
added to give (CHg^NO^SOg, which at 0° had a equilibrium
pressure of 570 mm. The equilibrium pressure of the
equimolar addition product of trimethylphosphine oxide and
sulfur dioxide was approximately 55 nun. at 0°. It is evi
dent, therefore, that after trimethylamine oxide has added
one molecule of sulfur dioxide, there is still enough
residual bonding power left to hold a second molecule of
sulfur dioxide with almost as much strength as trimethyl
phosphine oxide holds its first molecule of sulfur dioxide-.
Burg calculated ^XF° = 0 for (CH^J^NOSC^'SOg at
2.4°. The present data suggest, that £F° = 0 for
62
(CH2)3po*S02 at 84°. In other words, the former compound
has a vapor pressure of 7^0 mm. at 2.4°, while the latter
reaches 7 6 0 mm. at 84°. Thus it is seen that the stability
of (CH^^PO'SC^ lies between that of the mono-S02 addition
compound of trimethylamine oxide and the di-S02 compound.
CHAPTER VII
SUGGESTIONS FOR FUTURE INVESTIGATIONS
The extraction of trimethylphosphine oxide from the
hydrolyzed Grignard reaction mixture is complicated by the
presence of large amounts of inorganic salts. It is
believed that these could be removed by an ion exchange
column of a suitable mixture of anion and cation exchange
resins which were in the hydrogen and hydroxyl form, res
pectively. Trimethylphosphine oxide, as it is such a weak
base, should pass through without being absorbed.
Although (CH^^NO’BF^ tends to decompose at higher
temperatures, it would be of interest to attempt measure
ment of its vapor pressure at various temperatures after
complete volatilization of a small amount and to compare
them with those of (CH^^POBF-j. Data thus obtained might
show relative amounts of dissociation.
For a more legitimate comparison of the reactions
of sulfur trioxide with trimethylamine oxide and tri
methylphosphine oxide, it will be necessary to prepare the
(CH^J^NO-SO^ addition compound. To study the relative
strengths with which the two bases hold sulfur trioxide,
it may be possible to add anhydrous trimethylamine oxide
to a suspension of (CH^J^PO-SO^ in an inert solvent for
trimethylamine oxide and to determine whether or not
64
trimethylamine oxide takes sulfur trioxide away from
trimethylphosphine oxide. Conditions may be such that a
measurable equilibrium may be attained.
Added work is indicated on the purification of
(CH^^PO'SOg to determine the cause of the wide melting
range.
To confirm the hypothesis of the formation of
(CH^)3POH-(C g H ^ p u r e ethyl sulfuric acid could be
prepared and combined with trimethylphosphine oxide. Its
ionic nature might be investigated in some non-aqueous
solvent such as alcohol or dioxane.
The determination of the melting point-composition
curve of sulfur dioxide-trimethylphosphine oxide mixtures,
especially in the vicinity of 1 :1 composition, may help in
proving or disproving the existence of a 1 :1 addition com
pound. Valuable information concerning this problem might
also be obtained from: a more complete set of vapor pressure
curves at higher and lower temperatures.
CHAPTER VIII
SUMMARY
The reactions of trimethylphosphine oxide with boron
trifluoride, sulfur trioxide, and sulfur dioxide have been
investigated, and some modifications have been made in the
original method of Pickhard and Kenyon19 for the prepara
tion of trimethylphosphine oxide..
Boron trifluoride reacted with trimethylphosphine
oxide to form (CH3)3PO*BF3 (m. p. 148-150°) which was not
hydrolyzed in cold water, was non-hygroscopic, and was
sublimable in high vacuum. It was not possible, from
present information, to determine whether trimethylphos
phine oxide or trimethylamine oxide was the stronger base
toward boron trifluoride.
Sulfur trioxide formed the compound (C^^PO^SC^
(m. p. l65-210o) which was extremely hygroscopic and
hydrolyzed in water to give trimethylphosphine oxide and
sulfuric acid. The compound reacted with absolute-alcohol
to form (CH3)3PO•HOSO3 C2H5, m. p. 87-88°. Satisfactory
methods for the purification of (C^^PO^SC^ were not
developed. The analogous compound (CH^^NO-SC^ has not yet
been reported, but from the stability of (C2H5)3N0S03 '
toward alcohol it may be inferred that trimethylamine oxide
probably is the stronger base toward sulfur trioxide.
66
Sulfur dioxide formed with trimethylphosphine oxide
either a saturated solution which was about fifty mole per
cent trimethylphosphine oxide in sulfur dioxide, or a
liquid compound of 1:1 composition. Evidence was not
sufficient to establish which was the case. Trimethyl
amine oxide thus is a much stronger base than trimethyl
phosphine oxide, in respect to sulfur dioxide.
BIBLIOGRAPHY
BIBLIOGRAPHY
Bickerton, J. H., Master's thesis in Chemistry, The
University of Southern California, Los Angeles, 19^3*
Booth, H. S., and Willson, "Inorganic Synthesis," McGraw-
Hill Book Co., Inc., New York and London, 1939, p. 21.
Brockway, L. 0., and J. Y. Beach, J. Am. Chem. Soc., 60,
1836 (1938).
Burg, A. B., J. Am. Chem. Soc., 6^, 1692 (1943).
, and J. H. Bickerton, J. Am. Chem. Soc., 6 7, 226l
[ 1943) .
Cahours, A., and A. W. Hoffman. Annalen der Chemie und
Pharmacie, 104, 1-39 (1857).
Collie, N. C., J. Chem. Soc. Trans., 53, 636 ( 1883).
Godfrey, W. K., Master's thesis in Chemistry, The Univer
sity of Southern California, Los Angeles, 1948.
Hampson, G. C., and A. J. Stosick, J. Am. Chem. Soc., 60,
l8l4 (19385. ~
Lecher, H. Z., and W. B. Hardy, J. Am. Chem. Soc., 70,
3789 (1948).
Lewis, G. N., "Valence and the Structure of Atoms and
Molecules," Chemical Catalogue Co., New York, N. Y.,
1923, p. 134.
_______, J. Am. Chem. Soc., 3 8, 762 (1916).
Linton, E. P.,. J. Am. Chem. Soc., 62, 1945 (1940).
Lister, N. W.. and L. B. Sutton, Trans. Par. Soc., 35,
495 (1939).
Nylen, Paul, Z. anorg. Allgem. Chem., 246, 227-242 (1941);
as in C. A., J6:1295.
_______, Tids. KJemi. Bergvesen, 1 8, 48-50 (1938); as in
C. A., 32:8888.
6 9
Palmer, K. H., and N. Elliott, J. Am. Chem. Soc., 60,
1852 (1938).
Pauling, L., and L. 0. Brockway, J. Am. Chem. Soc., 59.
13 (1937).
Phillips, G. M., J. S. Hunter and L. E. Sutton, J. Chem.
Soc., 146 (1945).
Pickhard, R. H., and J. Kenyon, Chem. Soc. Jour., 89.
part X, 262 (1906). '
Sauvage, A., Compt. rend., 139. 674 (1904).
Traube, W., and H. Zander, Ber., 57. 1045 (1924).
Wheland, G. W., "The Theory of Resonance," John Wiley and
Sons, Inc., New York, N. Y., 1944, pp. 85-8 7. .
University of Southern California Literary

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Addition compounds of trimethylamine oxide

PDF

Complex compounds of trimethylamine oxide

PDF

An experimental investigation of sulfur compounds of boron hydrides

PDF

Addition compounds of trimethylamine with boron fluoride and its methyl derivatives.

PDF

Bicyclic compounds of coordination number five

PDF

A study of the reactions of phosphorus pentachloride with trimethylamine

PDF

A study of the polargraphic behavior of decaammine-[mu]-peroxodicobalt (III,IV) perchlorate

PDF

Cryoscopic and conductometric investigations of manganese (VII) and chromium (VI) compounds in sulfuric acid

PDF

A critical study of the 8-hydro-oxyquinoline method for the determination of beryllium

PDF

An experimental investigation of amine addition compounds of trimeric and tetrameric phosphonitrilic chloride

PDF

A study of the quantitative spectrographic determination of beryllium

PDF

An experimental study of the phosphorus chloronitrides and their reactions with trimethylamine

PDF

A study of the reaction of 1-bromo-2-oxoapocamphane in base

PDF

A comparison of the electron donor properties of 1,3- and 1,4-dioxane

PDF

A method of determining molecular weights of vapors by effusion

PDF

A preliminary investigation of the formation of molecular addition compounds by phosphorus chloronitride

PDF

An experimental investigation of the reactions of alkyl-tin compounds with boron compounds.

PDF

A kinetic study of the base hydrolysis of chloroamminebis (dimethylglyoximato) cobalt (III)

PDF

Amino compounds of silicon and of boron

PDF

A study of the structures and properties of triaryl compounds of boron

Asset Metadata

Creator McKee, William Edgar (author)

Core Title Addition compounds of trimethylphosphine oxide

Degree Master of Science

Degree Program Chemistry

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemistry, organic,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Burg, Anton B. (committee chair), [illegible] (committee member), Warf, James C. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-791862

Unique identifier UC11348029

Identifier EP41587.pdf (filename),usctheses-c17-791862 (legacy record id)

Legacy Identifier EP41587.pdf

Dmrecord 791862

Document Type Thesis

Rights McKee, William Edgar

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA