Citation
Synthetic and polymerization mechanism studies of unsaturated quaternary ammonium salts

Material Information

Title:
Synthetic and polymerization mechanism studies of unsaturated quaternary ammonium salts
Creator:
Angelo, Rudolph John, 1930- ( Dissertant )
Butler, George B. ( Thesis advisor )
Reid, C. E. ( Reviewer )
Brey, W. S. ( Reviewer )
Gropp, A. H. ( Reviewer )
Fyner, Mack ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
1955
Language:
English
Physical Description:
111 leaves. ; 28 cm.

Subjects

Subjects / Keywords:
Amines ( jstor )
Boiling ( jstor )
Bromides ( jstor )
Diamines ( jstor )
Liquids ( jstor )
Monomers ( jstor )
Polymerization ( jstor )
Polymers ( jstor )
Quaternary ammonium compounds ( jstor )
Resins ( jstor )
Ammonium salts ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Polymers and polymerization ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
Previous work in this laboratory has shown that insoluble resins of the poly-quaternary ammonium type possess anion exchange properties, Fuoss and Cathers) have made similar observations with the quaternized derivatives of polyvinylpyridine. With this property in mind, it was decided to prepare a series of unsaturated quaternary derivatives of diamines and study the anion exchange characteristics of the resultant resins. The main objectives in this investigation were as follows: (1) To provide a successful synthetic route to the bis-quaternary derivatives of unsaturated 1,n tertiary diamines. (2) a. To study a different approach to the preparation of the 1,4 and 1,5 tertiary diamines and their corresponding quaternary derivatives, since some preliminary work involving their preparation had proved unsuccessful, b. To propose a mechanism to explain the failure, and the various side reactions, of the previous attempts to prepare the 1,4 and 1,5 tertiary diamines. (3) To correlate the effect, if any, of the degree of cross-linking and the swelling coefficient on the capacity and possible Ion-screening effect of the resins, (4) To study the effect of variable distance between quaternary ammonium exchange centers on the properties of the resin, especially In regard to selectivity of Ions of various size and charge. (5) To study the mechanism of the polymerization of this type of unsaturated quaternary ammonium monomers, since they appear rather unique in their property of yielding a solvable and, consequently, noncross-linked resin from a monomer possessing two unsaturated groups.
Thesis:
Dissertation (Ph.D.) -- University of Florida, 1955.
Bibliography:
Bibliography: leaves 107-110.
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:
021767944 ( AlephBibNum )
13301719 ( OCLC )
ACX5629 ( NOTIS )

Downloads

This item has the following downloads:


Full Text






Synthetic and Polymerization Mechanism Studies

of Unsaturated Quaternary Ammonium Salts













By
RUDOLPH J. ANGELO











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











UNIVERSITY OF FLORIDA
June, 1955
















ACKNOWLEDGI-L1TS


The author wishes to express his sincerest

appreciation to Dr. George B. Butler whose tcochinC and

direction have greatly contributed to the completion of

this work. The author is indebted to Dr. A. H. Gropp

for his oncouragmonent, enthusiasm and criticism tlhough-

out this course of study. The advice and suggsctions

offered by the author's Supervisory Cormittce and by his

student associates is Gratefully acknowledged. The author

especially wishes to thank Mr. G. D. Price for carefully

proof-reading the manuscript, Mr. B. A. Parkin for the

benefit of stimulating discussions, and irs. Mlary Joy

Breton for typing the manuscript.

The author is indebted to the Atomic Energy

Commission for the support of this work.













TI'IL OF CO;TEJ.;TS


ACkICjIOL..LDG1i;TS

LIST 0. TA I ,S

LIST 0.' ?I'I ....S

Chapter

I. IIITiODUCTION

II. PRlPAi- TIO:O OF l,n TERTIARY DI/S!I:IS;

A. Tetraallylethylene diamine

B. Tetraallyltrjizathyle ne dianine

C. Totraally!hexamrcthyl-no diamine

D. Tetraallylheptanothylene diaxmino
E. Tetr-.allyloctamnthylcne dianine

F, Tetraallylnonamethylene diamino

G. Tetraallyldecamethylane diamirDe
III. P ,;PJiJ I.'IOC OF 1,4 AN D 1,5 Tlr, IA.TY DIA.'I.. S

A. Ui-s.ccecsful Attempts to Prepare 1,4
and 1,5 Tertiary Diamiiies

B. 'rcpar:~-tion of 1,4 and 1,5 Tertiary
Diari-ioes

C. Discussion of Results and Proposed
ilo clcni sm


iii


Pago
ii

v

vi


1

3

4

4

5

5
10

12

14
16


17

24

27









Chapter
IV. PREPAiTATION OF BIS-QuATE.. :A:Y Ai.ONIUM SALTS
A. Preparation of the IHezaallyl
Derivatives
B. Preparation of the Dimethyltetraallyl
Derivatives
V. SUGSP: ISIOI PO YiljZATICO; 0. BI3S-Q' AT: I A .T
SALTS
A. General Discussion of Method
B. Experimental Results
VI. EVALUATION OF CAPACITY STUDIES
A. ilethod of Analysis
B. Surimary of Experimental Results of
Capacity Doterminctions
C. Disc:asion of Capacity Determinations

VII, POLYiESiI. ATI0:T ILiC -:iIr.i S. UDt1S
A. Monomror Syntheese
B. Polyme:-ization Products and Results
C, Discussion of Results
VIII. SUIIAiY
BIBLIOGRAPHYP
BIOGRAPHICAL ICTISC


Page

34

35

o0

47
47
52

54
54

58
62

67
68

74
96
104

107
111










LIST OF TA3LES


I. Ilfrared Spectral Analysis 24

II. Smunnry of Physical Properties of l,n Tertiary
Diaminos 32

III. Proporties of Hexaallyl Diammnonium Dibromide
Derivatives 45

IV. Properties of Dimethyltetraallyl Dianmmonium
oibromido Jerivativas 46

V. Polymne ization Results 53

VI. SLummary of Ca.pacity Determinations 60

VII. Azo Initiated Polymerization of Hoxaallyylethlene
Diaamoniuw Dibromide 88

VIII. Azo Initiatod Poly-erization of Diallyldiethyl
Ajmmoniu n Bromide 89

IX* Analysis of Poly-diallyldiothyl Anmonium Bromide 91

X. Hydror-enation of Poly-diallyldiethyl Ammonium
Bromi de 93

XI. Comparison of Polymerizations Open to Atmosphere
and under Nitrogen 94

';II. Approximation of the Derroo of Polymerization 97













LIST OF FIGURES


Figure Page
1. Equilibrium Exchange Capacity 57

2. Hexaallyl Derivatives 64

3. Dimethyltetraallyl Derivatives 64







I INTRODUCTION


Previous work1-6) in this laboratory has shown

that insolul:le resins of the poly-quaternary ammonium

type possess anion exci.ance properties. Fuoss and Cathers(7)

have made similar observations with the quaternized deriva-

tives of polyvinylpyridine. With this property in mind, it

was decided to prepare a series of unsaturated quaternary

derivatives of diamines and study the anion exchange

characteristics of the resultant resins.

The main objectives in this investigation were as

follows:

(1) To provide a successful synthetic route to

the bis-quaternary derivatives of unsaturated l,n tertiary

diamines.

(2) a. To study a different approach to the

preparation of the 1,lj and 1,5 tertiary diamines and their

corresponding quate!rnry derivatives, since some p-elimi-

nary i:ork(5) involvin:- their preparation had proved un-

successful.

b. To propose a mechanism to explain the

failure, and the various side reactions, of the previous

attempts to prepare tlie 1,4 anld 1,$ tertiary diamines.

(3) To correlate the effect, if any, of the

1







2

degree of cross-linking and the swelling coefficient on the

capacity and possible ion-screeninC effect of the resins.

(4) To study the effect of variable distance
between quaternary ammonium exciharcie coiters on the proper-

ties of the resin, especially in regard to selectivity of

ions of various size and charge.

(5) To study the mechanism of the polymerization

of this type of unsaturated quaternary ammonium monoamers,

since they appear rather unique in their property of

yielding a soluble and, consequently, noncross-linked resin

from a monomer possessing two unsaturated groups. ()






II P;: P J Tli; OF 1,n TERTIJ/Y DIliIInS


In attempting to peopcre l,n tertlury diami:.es,

the reaction of the appropriate dihaloalk-ne or di(aryl-

sulfonoxy)calk.ine with diallyl arinoi proved very successful

except in the cases of the 1,4 mend 1,5 derivatives. The

successful method of obtcininT the 1,4 a-d 1,5 derivatives

follow in Section III.

The yields of the l,n tel-tiary diamines were 70-

954 in all cases except for the hexamothylene derivative
wlich was obtaL-.e in 21.75 yield. All temperatures are

listed as de-rees centigrade and are uncorrected.

The carbon, hydrogen and nitro-on analyses were

carried oft by Peninsular Chemroscarch, Inc* and Clark

Microanalytical Laboratory. The bromide analyses were

determined by the Volhard method as riven by Pierce and

laenish.

The following section deals with the preparation

and purification of the l,n tertiary diaml.es except for

the 1,4 and 1,5 derivatives*







A. Tetraallylethylene diamine

(CHe=CHICHs) al (CHO) ai (CHaCH=CH )

This product was prepared in a manner similar to

that reported by Angelo. ( Three moles (291 g.) of

diallyl amine was added to a paste of 168 g. (2.0 M) of
NaiICO0 and 100 ml. of water. The mixture was heated to a

gentle reflux and 187.8 g. (1.0 M) of ethylene dibromide
was added slowly. Reflux conditions were maintci:ied for
seven and one-half hours after the addition. The mixture
was cooled, filtered and the solution treated with an
excess of 40% iJaOH solution until a distinct separation
occurred. The oil was separated, dried over solid IaOCI
and distilled at reduced pressure. One hundred and fifty-
seven crams (71.14 yield) of clear liquid boili.ig at 72-3e/
ae
0.3 mm, and having4 nD 1.4705 was obtained. The compound
was previously reported as boiling at 790/0.5 mm. and having
28
nD 1.4702.

B. Tctltaallyltrimethylone diamirne

(C2,,=.Lli i)2:. (Cu2)a (L-A2C-;*=C^-1)8

The product used in this work was that prepared by

Angelo;(9) the procedure was that of Laakso and Reynolds(10)

involving the reaction of diallyl amine with 1,3-di(p-toluene-
sulfonoxy) propane. The yield was 80.~'; boiling point
88-90/0.5 rm.; nD 1.4711. Analysis calculated for

C,5HeN8O: N, 11.95%. Found: i., 11.57%.






C. Tetraall ylhexarnethylene diamine

(C i}e=CiIC 1;a )aN; (CHis), N(CHe CH=CaII ) B

The product used in this work was thct prepared by
(n)
Angelo; the procedure involved the reaction of allyl
bromide with hexamethylene diamine mixed with a paste of
ila2CO3 aad water. The yield was 21*75; boiling point

124*/0.8 mm.; n6- 1.4701. Analysis calculated for

CaIsiaiQ: 2, 10.13. Found: h, 10.21J.

D. Tetraallylheptamethylene diamine

(CH2=C;J C. ,) (C:Ia) 7. (C:12--=CIHa)

1. 1,5-dibro:nop :.itaE, e
Four hundrod and seventy-two grams of concentrated

H1SO0 was adJi. with cooli.-; to 1610 .. (9.57 M) of 43L
HBr solution. To this mixture was added 137 g. (1.595 .,)
of tetrahyd2ropDr3an. Tho mixtur-o was allowed to reflux
gently for threo- and a half hours, waz cooled and separated.

The bottom layer was filteird, :as 'ed with water and driod
over XgCO3. 'Thl dark liquid lua distilled -id 237 g. (72.3%

yield) of 1,5-dlbronoo.reitie oboilin,:: at 88-900/6-7 mm. was
obtai.i..-:d,

2. Pimelic nitrile
Two hundred and twenty grama (i.5 NI) of uaCI was
added to 300 ml, of water and the mixture heated gently
until the 'iaCU dissolved. Pour hundred and six grams







(1,76 M) of 1,5-dibromopentane dissolved in 1 liter of
95% ethyl alcohol was added carefully to this hot solution
until a vigorous reaction occurred. The remainder of the
bromide solution was added as fast as the reaction would

permit. The mixture was allowed to reflux for forty-two

hours after the addition. The ethyl alcohol was distilled

at reduced pressure and the residue cooled a'd filtered.
The filtrate was extracted twice with ethyl acetate

(300 ml. portion and 100 ml. portion). The ethyl acetate
was removed from the extract by distillation and the residue

was distilled at reduced pressure. One hundred and fifty-

seven grams (73.0% yield) of product boilinG at 153-6*/
(12)
3-3.5 am. was obtained. Pimelic nitrile is reported
as boiling at 175-60/14 mm.

3. Esters of piLmlic acid
a) ';iethyl pimelate was prepPred in a nenner

sililrf. to the method -iven by Pdms aand arvel.(13)

Five hundred 'Frr's of concentrated H2SO4 was added, with

cooling and stirrin-, to 500 r. of 95 ethyl alcohol. One
hundred and fifty-seven grars (1.29 M) of pimelic nitrile

was added -;d the mixture was allowed to reflux for ten
hours. The mixture was then cooled, poured into an equal

volume of ice water, separated and the top layer dried

over CaCla. The liquid was distilled at reduced pressure
through a Claisen apparatus and 202 g. (72.7% yield)






of diethyl pimolate boiling at 1),9-SL0/19-20 amm. was

obtal.ned. Diethyl pimelate(1) is reported as boiling

at )16-52*/22 mm.

b) Dibutyl pinelate was obtained in a 9C0J yield

by the following gonersl procedure: Two moles of pi.Aclic

acid was adCod to ol ht moles of n-butanol in an equal

volume of toluene and 2 g. of co-'3cettrated :.i':04c was added

as a catalyst. The solution was refl;-xed with stir:-lig

while the water of reaction- was roroved as the azeotrope

and collected in a water separator. The acid was neutra-

lized wCith an excess of "3aCOs while the solution was still

warm. After Ciltrati.n, the tolucne and butanol were

removeJ by distillation. The dibutyl pi~r.clte was fi.-all;

obtaine,0 by distillotion at reduced prs.Lure.

4. hieptanethylene glycol

a) The general procedure(15) as -iven in "Organic

Reactions," Vol. VI for the reduction of esters using

LiAlH4 was followed. Nineteen and five-tenths grains

(0.513 H; a 10% excess) of LiAlH4 was added to 500 ml.
of dry ether in a five liter flask. The slurry was cooled

by an ice-salt bath; stirring~ was maintai.ied with an

induction-typo motor; all possible precautions to exclude

noisturo were taken. A solution of 100 g. (0.368 A) of

di-n-butyl pimelato and 250 ml. of dry ether was added








very slowly to the well-stirred slurry of hydride. After
adding for one hour, the stirring became difficult and
500 ml. of dry ether was added to the reaction mixture.
The temperature was maintained at -3o to 0 during the
addition period of two hours. Two hundred milliliters of
ether-ethyl alcohol solution (50% by volume) was added
slowly to destroy the excess hydride. After all the excess
hydride had been destroyed, 1100 ml. of 10 HSO4 solution
was added until two clear layers were obtained. The layers
were separated; the ether layer saved and the water layer
extracted four times with 100 ml. portions of ethyl acetate.
The combined ether layer and extracts were dried over

Drierite. The solvent was removed and finally 40 g.

(32.5% yield) of heptamethylene glycol boiling at 140-
145/8-10 mm. was obtained. heptamethylene glycol(16)

is reported as boiling at 146-49/l0 mm.
b) The previous experiment was repeated using
20.0 C. (0.525 M) of LiAlI4 and 100 g. (0.368 M) of di-
n-butyl pimelate. Forty-five and one-tenth grams (92.7%
yield) of hoptamethylene glycol boiling at 145-488/8 mm.
was obtained.

5. IIeptamethylene dibromide
The general procedure(17) given in "Organic
Syntheses," Coll. Vol. II, for the preparation of bromides
from alcohols was employed. Sixty-six grams (0.5 M) of





9
heptamethylene glycol was placed into a flask equipped with

a gas inlet tube, magnetic stirrer and condenser leading to
a tared trap of -IaOH solution. Dry HBr gas from a cylinder

was introduced into the flask while the temperature was
maintai-ed at 100-200. The addition of HBr took three and

one-half hours, at which time the trap contai.-i:g NaOH

solution boran to increase in weight rapidly. The layers

were separated and tho bromide layer was treated with one-

third its volume of concentrated Hl20&. Water was added to

this emulsion until a distinct separation occurred. The
mixture was separated and washed with water-methanol solution

(50 by volume) until the product was neutral to litmus.
The pro.hiuct was dried over CaCI2, distilled at reduced
pressure and 65.0 r. (50.3w yield) of clear liquid boiling

at 125-260/8 mm. was obtained. Mullr and Vane(18) report

the boiling point of heptancthylene dibromide as 1230/11 nmm

6. Tetraallylhoptamethylene diamine

To a slurry of 242 g. (2.5 M) of diallyl amine,

126 g. (1.50 M) of .aiiCO3 and 75 ml. of water, heated to a
gentle reflux and stirred vigorously, 65.0 g. (0.252 M)

of heptamethylene dibrcnide was added slowly for one and

one-half ho:rs. The mixture was cooled, treated with an

excess of LC4 NaOH solution, separated and the amine layer
dried over NaOH pellets. The excess diallyl amine was






10
distilled and the residue was distilled at reduced pressure
through a Claisen apparatus. Sixty-four groans (87. yield)
of light yellow liquid boiling at 124-26*/0.1 rmn. and having
sB
an n 1.4719 was obtained. Analysis calculatod for

Ce114 s: C, 78.6 H, .8; ,, 11 ; i 9.6o5. Found: C, 78.51 ,

H, 11.50o; ;i, 9.51.*

E. Tetraallyloctanethylene diamine

(C i,=CHCi ); (C Ha)I ; (CHGCJ =C Ia) ,

1. Suberic nitrile
This preparation was carried out exactly as
described in the previous preparation of piiolic nitrile.
One hundred and ninety-seven grams (0.82 M) of hoxanethylene
dibromide dissolved in 500 ml. of 950 ethyl alcohol was
added to 100 g. (2.05 M) of NaCI; dissolved in 150 ml. of
water. After the hexawiethylene dibromide was treated in
the same rmnnvor as the previous nitrile preparation, the
mixture yielded 96 g. (87.1% yield) of clear liquid boiling

at l75-78*/10 a. Deutsch and v. Braunn(19) report the
boiling point of suberic nitrile as 176-8/11 mm.

2. Diethyl suberate
This preparation was carried out exactly like the
previous preparation of diethyl pimelate. Ninety-six
Crams (0.705 M) of suberic nitrile was allowed to reflux
fifteen hours with a mixture of 400 C. of concentrated





11
IiSO20 and 40C ml. of 95. ethyl alcohol. After the stberic

nitrile was treated in the sa:no irmnner as tVie previous
nitrile hrydolysis, the mixture yielded 119 g. (73.5% yield)
of clear liquid boiling at lO-L,'O/8 mrm. and having an
n8 1,4324. Karvonen(20) reports the followilt constants

for diethyl suberate: boiling point, 136-80/8 mn. and
as
n, 1.43278.

3. Octarethylene rlycol
The reduction of diethyl suberate was carried out
in the scje manner as the previous reduction of di-n-butyl
pimelate. One hundred and nineteen rrams (0.517 1i) of
diothyl suberate dissolved in 250 ml. of dry ether was
carefully added to a cooled slurry of 25.0 g. (0.66 M) of
LiAIi4 in 500 il. of dry etheo. The reaction mixture
yielded 73.5 r. (97.4* yiold) of a clear liquid boiling at
158-60*/9-10 nam. The boili.g; point of octamethylone glycol
is reported(21) as 160-620/9*5 am.

4. Octar~ithylene dibror-ide
Octamethylene dibromide was prepared in the sane
nanner as the previous prepLration of hoptamethylene di-
bromide. ieventy-three aid five-tenths grams (0.503 M)

of octamethylene -lycol was treated with Gaseous iir at
100-20* for two hours. After being treated in the same
manner as the previous bromination, 86.5 g. (63.2% yield)







of clear liquid boiling at 170-1*/33-34 mm. was obtained.
Dionneau(22) reports the boiling point of octamethylene

dibromido as 173*/35 mm.

5. 2etraallylocta-methylone diamine
To a slurry of 303 g. (3.18 II) of diallyl aniLe,
168 g. (2.0 M) of .aHCICO and 120 ml. of water, heated to a
eontle reflux and stirred vigorously, 8,.5 g. (0.318 M)
of octamothylone dibromide was added slowly for forty
minutes. Afte, the addition, the mixture was allowed to
reflux for an additional eight and one-half hours. The
mixture was then cooled, treated with an excess of LO' .aOli
solution, separated and the aminc later dried over iaoH
pellets. The excess diallyl amine was distilled an-d the
residue was distilled at reduced pressu-re thiro,'h a Claisen
appa-atus. The prodAuct frothed very badly; class wool and
paraffin were added to relievo the foaming. Eishty-three
and five tent'.s grams of light yellow liquid boiling at

133-40*/0.l mm. was obtained. This liquid was fractionated
sad 73.0 r- (75.5` yield) of light yellow liquid boiling at

134-80/0,09-0.1 mrn. ind having: an nS 1.4651 was obtained.
Analysis calculated for CloHal3a: C, 78.9,; 'I, 11.-32;

i., 9.2%. Found: C, 78.81/; H, 11.755; N, 9.15C.

F. TetraallTlnonamethylono dianine
(CHIi=CHCsii )2N(C1H, ), (CH, l=CLal )






1. v'rom tile disulfonate
The preparation of 1,9-di(p-toluenesulfonoxy)nonane
was accomplish. od in accordance with the procedure of Laakso
and lHeynolds.(lo) Eirhty-six and two-tenths grra:s (0.54 M)

Oj. noinam.ethylone Cl:rcol dissolved in 10 ml. of dry pyridine
was treated with 206 g. (1.08 A:) of p-toluene sulfonyl
chloride dissolved in 280 ml. of dry pyridine. Folloiuiin
the method cited, 217 r. (85.5 yLold) of pro.i:.ct was
obtc.i-ed. Lieltini point after third recrystallization from
absolute alcohol was 75-70. Analysis calculated for

CaaliaSs0e: C, 5).95 ; H, 6.j:>. Found: C, 56.87,;
:, 6.o9;.

-.noty-o:e grams (-.195 M) of 1,9-di(p-toluone-
sulfonoxy) no:a:le was refluxed and stirred for fort,-eiLiht
hours '.ith 33 g. (4.0 M) of diallyl amine in accorda.ice
with the procedure of Laakso and Remnolds.(10) Afte working:
up the reaction r:.iture, 42.5 g. (68.3; yield) of liuiht

yellow liquid boiling at 11i8-52/0.06-0..1 m. mrnd having
an n52- 1.4710 was obtaoined.

2. From the dibromide
To a slurry of 170 g. (1.75 M) of diallyl Laine,

84 g. (1.0 M) of ':aHCO3 and 50 ml. of water, heated to a
-cotle reflux and stirred vigoro.:sly, 50 g. (0.175 M) of
nonanothylene dibromide was added over a two and on---.alf







hour period. The mixture was allowed to reflux an addi-
tiaoal sevon aod one-half hou-s. The mixtju.o was thon
cooled, treated with an excess of 40' MaO0i solution,
separated and the amine laTyer dried over s0'Hi pellets.
The excess diallyl amine was distilled and the residue was
distilled at red-iced presoFre. ?ifty-two gra-s (9m4.5

yield) of li-ht yellow liquid boiling at 146-8/0.06-0.07 mm.
and hrvin.- an nDi 1.4723 was obta.i ed. Analysis calculated
for Cs,Ha38,: C, 79.3/; ri, 11.93. Fo'-nd: C, 79.8-5;

i, 12. 01.

G. 'otraallyldecaWr3tnilen1e dimline

(GHC=C:ICHs2 )a (CH2) eol( CH;G=C=GH2 ) 2

1. From t-e disulfonate
he preparration of 1,10-di(bo .zoenesulfonoxy) docame
wrs accomplisrhed in accor-dance with t~ prococd.-re of Laakso
and Rernolds. (10) One hundred and one gram-s (0.58 M) of
docrenethyl-no glycol (n.p. 70-1) dissolved in 200 ml. of

dry pr:-idln- :.s treated with 205 g. (1.16 M) of ben:cne
sulfor-yl chloride. Follow!:- the method cited, 226 g.

(35.65 yield) of product was obtained. -lelting point after
third recrystallization from absolute alcohol was 46-308
/Aialysis calculated fcr CmaHa oSgO: C, 50.14'; i, 6.66u.
F'ouLd: C, 57.6T/; H, 6.79?.

One hundred and thirty-six Crmis (0.3 M) of
1,10-di(benzenesulfonoxy) decane was refluxed for forty-





15
oi.rht hours 'with 3uU ;. (5.15 ') of diallyl amine in accor-

dance with the proced-ire of Lackso and Reynolds.(0 The

reaction ni::ture yielded 82 g. (u2.2J yield) of a dark

liquid, boili..- at 155-6C0/05 rm. This was fractionated
and 72 g. of a li ht yellow liquid boilij:! at 154-5/
30
0.1-0.2 --i. and havlin an nD 1.4701 was obti: .ed. Analy-

sis calculated for CG22I1402: -43:. found 2, 8.27,.

2. -.ro i tV dibro:-ido
To a sl-:r'r of 162 -. (1.66 :) of diallyl ami.-o,

84 (. (1,0 M) of !;a:iCO3 arn .50 nl. of water, heated to a
?-tle reflux nid stirred viGorousl-, 50 -. (0.166 M) of
decamethylone dilb omido was added over a period of one and

threo-fourths hours. The rmi-.ti, u was efl''.xed for an
additional fivq hours, cooled, treated with an e::c.ss of
IC" U'r:0Hi s< l't.io., separated and driod over 'iaOi pellets.

The excess dicll-jl v&ine was to]:en off and the residue
distilled at reduced pressure. Thirty-nine graC1s (70.5-

yield) of a lirht, v--llow liquid boili: at 152-6*/0.07 mm.
v id havi:- an ini- 1.725 was obtrl.-ed.








III PRiPARATIOu OF 1,4 AnD 1,5 TEi'TIARY D)IA11iL-S


Attempts to prepare tetraallyltetramethylcne

diamine and tetraallylpentamnethylene diamine by the

previously successful reactions of the 1,4 and the 1,5

dihalo- or di-(arylsulfonoxy)alkanes vith diallyl a.ine

were unsuccessful. (23) In the reaction with the dihalo-

alkanes, the products were diallyl pyrrolidinium bromide

and diallyl piperidinium bromide. In the reaction with the

di-(arylsulfonoxy)alkancs in the presence of a larr'e excess

of diallyl ainne, the products were allyl p.'rrolidine and

triallyl amine from tihe 1,4 derivatives, and allyl piperi-

dine and triallyl amine from the 1,5 derivatives. The

products of the reaction can be explai.ied on the basis of

an intramolecular cyclization to the five and six ::embered

cyclic quaternary ammonium salt, followed by an allylation

of the excess diallyl amino by the quaternary ammoniun salt

to produce triallyl aminc and the appropriate allyl substi-

tuted hetrocyclic amine.

This section deals with the isolation and identi-

fication of th'e products of the above reactions, an ex-

planation of the course of the reaction and, finally, the

successful method of preparation of the desired 1,4 and

1,5 tertiary diaminos.






A, Unsuccessful Attempts to Prepare
1.4 and 1.5 Tertiary Diamin-s


1. Reactio:i of 1,4-dibromobutane with diallylamine

a) A mixture of 44 5 g. (0.206 M) of 1,4 dibrono-

butane, 60 .. (0.618 Mi) of diallyl amino, 34.6 g. of IaliC03,

and 20 ml. of water was reflu :ed for seven hours. After

filtration and addition of 2iaOH to the filtrate, the organic

layer was distilled; however, no material was obtained

boiling in the expected ranre for tetraallyltetramethylene

diain.e.

Several other experiie-nts were perfor.;ed varying

the ratio of dihalide to amine, cha.n-gin solvents and

chagrin the base, but in all cases none of the desired

product was isolated.

b) To four milliters of 1,.-dibromobutane was

addod twelve milliliters of diallyl amine at a temperature

of 35. While the liquid was beirn- stirred, the temperature
rose to 1000. After standing overnight, the crystalline

product was collected, washed i ilth aceto.ne and recrystal-

lized from hot acetone and a small amount of absolute

alcohol. The product started to decompose at 205* and

melted completely at 315-220 with decomposition. A

bromide analysis of the product showed 34.77Z bromine.





18

The calculated Br percentage for the dihydrobromide,

H H
(CHI2=CCIIC a)aI(CII )4N(C!aC=Clsi)2, is 38.951. The calculated
+ +
Dr" Br"

Dr percentago for diallyl pyrrolldiniun bromide,
Br-
I \+ iGCE=Caii

CHaCH=CHI,, is 34.49%.

Since diallyl pyrrolidinium bromide has not been

reported in the literature, a method for its preparation

follows:

N-allyl pyrrolidine was prepared by treatinT

pyrrolidine with allyl bromide in the presence of NaHiCO3,

The boiling point was 129-30* and n8.- 1.II086. Analysis

calculated for Cyia U: C, 75.64; H, 11.71;; N, 12.61".

Found: C, 75.35,; H, 11.87%; N, 12.,1%.

To a cold mixture of 21,0 g. (0.19 M) of ;-allyl

pyrrolidine and 75 ml. of acetone was carefully added

25.0 c. (0.207 M) of allyl bromide with cooling. The

hygroscopic salt which precipitated was washed with

acetone and dried in a vacuum desiccator. Thirty-nine

grans (90.5% yield) of product was obtained. After re-

crystallization from hot acetone with a small amount of

absolute alcohol, the product started to decompose at

2020 and melted completely with decomposition at 315-20,.

A mixed melting point with the reaction product from above




19

showed decomposition starting at 198o; the mixture melted

completely with decomposition at 314-220.


2. ieactioi, of 1,5-dibromopentane Uith diallyl amine

A mixture of 4 ml. of 1,5-dibromopontane and

12 ml. oi dicllyl amine was allowed to stand overnight.

The hygrioscopic crystals were collected, washed with

acetone and dried in a vacuum desiccator. After recrystal-

lization from hot acetone i.ith a small amount of absolute

alcohol, the product gave a flash melting point of 194-60.

Upon slow continuous heating the product started to de-

compose at 209 and melted completely at 297-302 t:ith

docomposition. A bromide analysis of the product showed

32.318 bromine. The calculated Dr perce-.tare for the
dihydrobromide,

H H
(CiS=C3,LHa) (i'I)6:i(ciihi.*u=ic)as is 37.66%.
+ +
Br" Dr"

Tio calculated Br content for diallyl piporidinium bromide,

r-


\C;CU=G;ig, is 32.o'. Diallyl piporidinium

bromide is roported(l) to Ilavo a flaai inoltiLg, point of 190.


3. ioaction of l,4-di(p-toluoGiesulfonoxy)butane with diallyl-
0anine

Aftor a number of unsuccessful attempts to obtain

totraallyltotra-iethyleno diaminine by the method of Laakso







and Reynolds,(10) the following modification was adopted:
Two hundred and twenty-eight Crams (0.577 1) of 1,4-di(p-
toluenesulfonoxy)butane and 632 g. (6.5 M) of diallyl amine
were rofluxed for seventy-two hours. The mixture was
allowed to cool and the layers separated. The lower layer
was treated with an excess of 40 0 iaOH solution, allowed
to stand overnight and filtered. Fran the filtrate, 3 g.
of a yellow liquid boiling at 100-1101/0.3-0.5 mm. and
84
having an nD 1,4721 was obtained. The original upper

layer was distilled and 180 g. of diallyl amine was re-
covered. The residue was distilled at reduced pressure

and 11.5 c. of a light yellow liquid boiling at 100-110o at

0.3-0.5 mm. and having an n$4- 1.4720 was obtained. A
nitrogen analysis showed 9.33,. The calculated nitrogen
percentage for tetraallyltetranethylene diamine (C eHa8Ia)

is 11.27.
These fractions appear to be impure tetraallyl-
tetramethylene diamine because this compound was prepared
by another method and found to have a boiling point of
96-8o/0.1 mm. and an ns- 1.4716. All attempts to purify
the fractions failed; the attempts to prepare derivatives
for identification purposes also proved unsuccessful.
Considering the total of 14.5 g. of crude product as
tetraallyltetramethylene diamine, the yield would be 1015.






howeverr, during the distillation of the above
reaidue at reduced pressure, a large amount of material
was collected in the cold trap. This material was
fractionated r~d a small amount of diallyl amine was
recovered. This was followed by 20 g. boiling in the
ratire 111-147, and 21 r. boilingJ at 147-?149 and having
an ne- 1.44 89. With allyl bronide, this latter fraction

(147-49*) 7ave tetraallyl3 ari.onium bromide molting at
183-40. A mii;xe meltinG point with an authentic sample
of tetraallyl aunonium bromide was 182-83*. The reported
constants for triallyl arrine are: Boiling point 143-90;

2n'- 1.4502.
The intornediate fraction (111-47) should contain

allyl piyrolidine (bp, 129-30) if the previously cited
disproportionation rnochanism is correct. However, allyl
p rrolidilo was not isolated front the fraction at that time.

4. Reaction of diallyl pyrrolidinium bromido with
diallylamino
In order to support the above proposed course for
the reaction, 36.0 r. (0.155 ') o' diallyl pyrrolldinium
bromide and 120.0 g. (1.24 M) of diallyl amine were re-
flu::od for thirty-six hours. Turin;~ this time, the liquid

ti-'pe.-ature remained at 1090, the boiling point of dic.llyl
amine. After thiE time, the salt became viscous, the color
changed to a licht -ed and teo temperature rose to 120o.







Refluxinc was continued for an additional twelve hours.
The mi:.ture was cooled, filtered and the dark red solution

fractionated to yield: Diallyl arino, 68 g.; boilijC

point, 110-13,; n1f- 1.4397; all-l pyrrolidino, 7 C. ;
boiling point, 130-1i0*; n25- 1.,1471; acd ttiallyl ani.e,

4.5 r.; boiling point, 146-90; n~ 1.lU99.
Treatiicut of t.e allyl p.rrolcidlne fraction with

allyl bromide gave the quaternar- ammoniiun salt melting

at 313-20 with decomposition. mixedd molting point with
an authentic sample of diallyl pyrrolidiniui bromide was

314-210 with decomposition.
Treatment of the triallyl anine fraction with

allyl bromide gave the quatornar-;' amoniIu salt melting

at l84-5". Mixed melti:ig point with an anthontic sr-e mple
of tetreallyl annonilm bromide was 184-5*.


5. i;oaction of 1,5-di(benzonesulfo..oxy)pont e9- with
diallyl amine

As in t'ie case of the 1,4 derivative, a nnumbe" of
unsuccoszsful attempts to pr-eoare tetrea llylpentamethyleno

diamine by the method of Laakso and Reynolds(10) were made

and the following modificaticn was adopted. 'Tw1o h'. .drod
and seventy-nine gra-,s (0.727 M) of 1,5-di(benzenesulfE..oxy)-

pontane and 776 r. (8.0 M) of diallyl amine were refluxed
with stirring for forty-eight hours. The product was

separated as described above, and 501 g. of excess diallyl








amine removed b: distillation. Following! the diallyl
amine, l5 g. of a colorless liquid boiling at 1I7M-9*
was obt.aled, a !d 40 o, o.f a dark yellow liq~3i boiling
at 60-100 at 0.05 is. iAftor refracticnration of the
fraction boiling at 147-1:90, a prod-ict boilin.- at 1l8.-503
a.d having an nD I.l4539 was obtained. On the basis of

the above proposed mechLanmis for the reaction this
product sho Id be a mi.:tu-c of triallyl amino (b.p. 1.3-

50, nD 1.1502) asd allyl piporidine (b.p. 14'- ,o

nD 1.4577).
Since it was ir'possible to separate t;ese com-
pounds by distillation, an iifrsar:i analysis was node
"sl."i a Porkin ULei double-boam infrared spectrop`.oto-

meter with a cell thic lnss of 0,025 mr:. Various mixtures
of pure nacmol.s of triallyl aonine and allyl piporidine
were a~de in an attempt to match the index of refraction
of the reaction rIlixtu-o. A min-ture contain. C 51e by
we-it of triallyl amine hsd a -;DL" 1.4533 compared to

the nl'- 1.4539 of the reaction Tni:tu e. The infrared

spectral anal-yis .'-rve tile folloi.in-r results:







TABLE I

IinFAil^J AL;CTiiAL Am.LV.IS


Sample nas Absorption band (mici-ons)
Sample n D


Triallyl
amine

Allyl pipori-
dine


1.4502


1.4577


Authentic
Mixture
(51, by wt. 1.4533
triallyl
amine)

Reaction
Product 1.4539
Mixture


3.30 sh




3.38 sh


9.32


---- W


7,23 7.70 8.33


9.35




9.35


7.22 7.70 0.33




7.22 7.70 8.33


'these results show conclusively the presence of

both triallyl amine and allyl piperidine in the reaction

product mixture, and the refractive indices indicate

approxinately equal molar quantities of tihe two compounds.


B. Preparation of 1,L aoid 1,. Tertiary Diariinos


1. Tetraallyltetramethylene diamine

(Cea=ChCH )NI;(CH)4N (CiiaCli=0iis)a
a) Preparation of L,.i,1'.1,'--tetraallyl suc-

cinamide.--To a well-stirred solution of 211 g. (2,168 i0) of

diallyl amine in 200 ml. of dry benzene was added a solution

of 84 g. (0,542 M) of succinyl chloride (bp. 87-9/18 mm.)


---


---


"44


---- I





25
in 100 ml. of dry benzono while t:h tempornture was nmin-

tailod ,-t 10-20', titrrfL: i was co:tinl'.od at this tempera-

ture for one hor afto:-- addition w conrileted. The Solu-

tion was filter?' ar.d the Ciallyl amidI hydrochloride

crystals were washed with Jr-y benzc:-s. The wcshinirs were

added to the ori "'inal filtrate and the bensene removed.

Tieo rveadno wfs distilled at re'd- ed nress-'ro to obtain

100 g. (80o,6; yield) of liquid boililn: at 160-2/0. 1 "n.,

having an n, 1.5925 and d4 1.0060. The infrared

spectral- showed absorntlon bacds for aliide carbonyl a-:d

carbon-c'r','o dc:ble bo:d. Anclyslis calculated for

C.e:40s: C, 69.:,; H, 8,38;; 1T, 10.13., Fo~nd:

C, 69. C'; 9.11 ; -;, 9.7 "T: calc-'letd.: 81.86.
IL-D fo'd: 81.15.

b) :,dJ ctio-.. of :1,: '--tetrnallyl s c-

ci:l;,ide.--To 16 g. (0.h2 H) of LiA1H4 as a sl'-rry in

300 ml. of td,- ethLor, was added during one anid throo-
fourths hx-.., 62 p.* (0,22 IT) of tetrc-llyl s-.ccina-ide

dissolved in 2CC ml, of dry ethe'r, A tihre-neck five

liter flask equi::e" Milth ice water condenser, thermometer,

droppi:ri f-Tunol, sealed stirrer, and cooled by an ice-solt

bath wrs omplo~ed. The tempe-"ature of the reaction ml1:ture

was nrii-itrui-ed below 15* durir- the addition:. After reflu;zing

fifty hers, a mixture of ethanol and water was added to

destroy tho excess hydride. The mixture was then treated







with 800 ml. of 10% NaOH solution and solid "8aOH pellets
until a distinct separation occurred. The ethor layer was
separated and the water layer was extracted twice with
200 ml. portions of ether. The combined ether layers were
dried over solid NaOH, and the ether removed. The residue
was distilled at reduced pressure to give 27.5 g. (49.3%
yield) of clear liquid boiling at 96-8*/0.1 mm. and having

an nD 1.4716. The infrared spectrum of the compound
showed the absence of the carbonyl bond and the presence
of the carbon-carbon double bond absorption. Analysis
calculated for C,eIlleeNI C, 77.4l%; H, 11.3,; hI, 11.2,.
Found: C, 77.50; H, 11.3N; 1, 10.7". The dihydrobromide

was prepared by treatment of a dry benzene solution of the
above amine with dry HBr gas. The compound melted at 107-8.
Analysis calculated for C,1ois3aoBra: Br, 38.95r. Found:

Br, 38.621.

2. Tetraallylpentamethylene diamine
(CHa=CiiCH2 ) 3;(CHB),N(CHaCH=CHja)a

a) Preparation of -I;J.11'.Ii'--tetraallyl rlutar-
aaido.--UsinG 64 g. (1.64 M) of diallyl amine and 72 g.

(0.425 M) of glutoryl chloride (b.p. 1000/10 mm.), and
following the procedure above for the preparation of
tetraallyl succinamide, 113 g. (91.7 yield) of totraallyl
glutaramide was obtained boiling at 165-7*/0.03 mm, and
having an nia- 1.5020. Analysis calculated for C,7HSOeiHs:





27
c, 70*3/; H, 8.86"; i, 9.t65. Found: C, 69.9c; H, 8.366;
i, 9.26%. The infrared spectrum showed absorption for
amide carbonyl and carbon-carbon double bond.
b) Induction of li.il.U'..I'--tot.,ually7l lutar-
saido.--Usinr 21 r. (0.553 I) of LiAlH4 and 87 r. (0.30 1)

of tetraallyl rlutaramide, and following the procedure
described above for the reduction of tetraallyl succinamide,
e::copt that this solution was allowed to reflu-< for four

hours instead of fifty hours, 53 g* (63.[L yield) of a
clear liquid was obtained. This product boiled at 96-70/
0.02 n2i. and had an nr- 1.4743. The infrared spectrum

showo'e the absence of the carbonyl bond and the presence
of a stronC terminal carbon-carbon double bond. Analysis
calculated for C71Ho1- : C, 77.8'; H, 11.4.6 ; I, 10.68;.
Found: C, 77.3oj; H, 11.2LI; N, 10.33T. The dihydrobromide
was pre)ared by treatment of a dry benzene solution of the
above amin:e with dry IIHr ras. The compound molted at 115-

7. Analysis circulated for C,7i1haiBrs2: Br, 37.66'.
Found: Br, 37.58%.

C. Discussion of Rosults andl Proposed IMechanism

In the reaction of l,,L and 1,5 dihaloalkenos with
diallyl zrline, it appears likely that the first attack
would result in the formation of the 4 or 5-bromoalkyl
diallyl amnino hydiobromide. In the presence of an excess
of diallyl amine, the free bromoalkyl diallyl amino should






28
be liberated, since it should be the weaker base. The second
step probably involves the formation of a cyclic quaternary
(24)
ammonium salt. It has been pointed out that aliphatic
bases containinC amine and halogen in the 1,4 position
exist only as salts and on liberation of the bases condense
to pyrrole derivatives. It has also been shoun(25) that
+
azin.-s of the type x(CHs)nNRa yield (CHe)nNRax" when n

has a value of 4, 5 or 6. It appears, therefore, that the
mechanism of reaction of 1,4 and 1,5 dihaloalktans with
diallyl amine is as follo~.s:

HBr
Br(CH,)4, Br HIfl(CHICIH=C:I)- Br(C 4 I)a'(e2CIl=C=g),

Hi (CHaC T=CII2 )a

( C)C =T(C:2C I"I 0:a )12 Br (CHs) e ( H( C2CIi=CH2 ) I


Although disproportio.'ation was not observed in the
reaction of diallyl amino, eith-r with 1,4-dibromobutane or
1,5-dibronopcntano (probably since a large excess of diallyl
amino was not .seod), heon diallyl pyrrolidinium bromide was
heated with a large excess of diallyl amine, both allyl
pyrrolidine and triallyl amine were isolated and identified.
It appen.as that the reaction is as follows:
Br"
\+ l3HaCIICHs
+< C=C + 2HN1(CH1ChCH) reflux in
1\CW=CHU s larGe excess of
+Br- HN(CHaCH= CHCa)
Iahl(CH2CiH=CIia)s + N(CIIHCI=CI[a)3 + D .CHSaCH-=CHS




29

This would appear very likely in view of the work of Zliel

&ad Pecklam(26) involving the migration of the furfuryl

group in the disproportionation of fur'uryl quaternary

anronium derivatives. These authors havo shown the

followii g:

I -(Ch) ) + reflux r (Cis)3l +


(0.05 M) (0.6 M) s +

+4 llg
I"
(W3' yield)

Thic tTpe of disproportionation also occurred in

attempting to use the nothod of preparation involving the

di(aryl-3ulfonoxy) alkanes in the presence of a large excess

of diallyl amine. It appears that in this case the reaction

involves first an intramolecular cyclization, as shown

before, followed by a disproportionation involving migration

of an allyl -roup, as shown above. This would be indicated

as follows:

(lihere A is an arylsulfonate group)

reflu.x in A CH=
A(CHg)4 GA excess (

i- i (CO IaC =C IT= )a CC.ia- CH=Ca

2 (Cg(...=CJ i)a <+ ILa(C.lCi=C~Ha) + (3) i CIIeClI=CHe







Although Laakso and ieynolds(0) reported a yield

of 77.5, in the preparation of 1,l-bu s(dibutylamino)butane,
and 66.5 in the preparation of 2,5-bis(dimorpholino)hexane,

the yield was only 10% in the preparation of 1,5-bis(di-

benzylamino)pentane. It has been shown by von i3aaun, Kahi,
and Goll(27) that the relative firiness of attachment of

hydrocarbon residues to hetrocyclic nitrogen in quaternary
aAmnonium salts increases in the follo':in- order: allyl,
benzyl, methyl, ethyl, propel, etc. These authors also

point out that rethyl groups were lost, in the presence of
primary or secondary amines, from quaternary ammonium salts

of hetrocyclic nitrogen compounds (even piperazine, the
least stable hetrocyclic compound studied) without rupture

of the hetrocyclic ring.

Allyl or benzyl groups are lost in preference to
alkyl groups because of the resona-.ce stabilized ions forced.

The weakness of the carbon-nitrogen bond caused by the

formation of such ions accounts for the allyl migration
shown in this study and for the poor yield of 1,5-bio(di-

benzylamino)pontane. () This would also tend to support
the theory of a carbonium ion(26) as an iitoermdiate in

the disproportionation reaction.

The fact that furfuryl trimethyl ammonium iodide
reacts with piperidine(26) to give a 48% yield of N-fur-

furyl piperidine is evidence that the relative firmness






of attach-eont, as discusseJ by von draun, Ku=n, and

Gloll,(2 of furf-r:-- is less than that of methyl. This

would uynp'-ar quite reasonable since the .'urfuryl group

contains Lhe allylic structure. The fact t"it this

reaction o2oce:ided to :S completion in three hours

duri-i: rrflux at appro-inmitoly 1300, while diallyl pyr-

rolidinirLm b.-oinide, after thirty-six ho rz at 1090, had

shown no sirns of reaction, would indicate that the furfuryl-

nitrocon bond in frrfuryl trimethyl aiiionizuci iodide has a

lesser doaree of attachraent and a greater degree of ionic

character than the allyl-nitrogcn bond in diallyl pyrroli-

dinium bromide. This view is also supported by the fact

that the j.li'uryl carbonium ion has a greater number of

stoailir-in; resonance structures than the allyl carbonium

ion.








































































H Wsr

r-t 3 0e r- SS 9
fr r 0 *




H00 HI4
+)4. r4 43 f
I C-4 CA


r-
r-4


H
i-l










o'
C.-
(N








N<





H


H
r-t
N-







Co



i
('a



0












HI
C.-

























H 0 40
r,-
r8
0





043




E-4
QD~


0


H(
at S


h


a
0










0
4-4
,-p



*f'
q.


V
0
e ,o 00








E-4 E-
r-

as
3
0


H


*
0A-

0 0
O Q
o o
C 0

0 0


0 0
0 0


0 0
r-H H
o 0



00

0 0
-O O-












> o
i- r



0
0 a
(4<










A, (X
(8 r







1A C i.
*rl * * *
CO l^ s0 0m '0 "-


r0 r1 1 F4


0 V\ r~\4
f-i '"\ co-

a *
0 0





%-4 CC ) CO
oo
4A 3 -
*** I






,-4



-4 H H -4


H 60 6,-I o i
0 * I
4 0 c0 Co











0- C'- C'- i

r a H
0 Hr- H HD q






o %4 (s F, C
fr I r-I -I ,-I <


S- 1 4 Ccd
$ *0 C 4 o
u 43 -1 03 Q


NQ CUN N

) leP 0r 0 0r
E-4 H H-4 H


B o E e *
c a '
*







K000
',o C
0 -0












-4 r- i b s
O H5,4 els cr-4 r-40






? 4 (.30 D ? s
-4 4#$ 0 +O 4
*. 0 ra






C'T0 Cd O





H E-0 0 e
0 -s.0 I4~C u.L;, su 0
U 4-CU~r 4.)'r4 ~ d -G~t*
02 0I 2
~E4 -( r









IV PRLPARATIO;1 OF BIS-LuAT.r.AiY AM.nI:IO.:IU SALTS


The bis-quaternary ammonium salts prepared in this

work were made by adding the appropriate halide to the

unsaturated tertiary diamine dissolved in acetone or aceto-

phonone. In the preparation of the hexallyl derivatives,

two moles of allyl bromide were added per mole of diamine.

In the preparation of the tetraallyldimethyl derivatives,

dry gaseous methyl bromide was added in an excess which was

collected in a cold trap.

All of the hexaallyl derivatives were formed quite

readily by gentle heating on a steam bath. The tetra-

allyldimethyl derivatives were prepared at the temperature

of an ice bath, using a magnetic stirrer and allowing the

excess methyl bromide to escape into a cold trap while

the mixture remained at room temperature ovornirht.

In most cases the salts obtained were extremely

hygroscopic and extreme care had to be taken in the puri-

fication of these compounds. In all cases the product

from the reaction mixture was quickly filtered and then

washed with cold dry acetone and dry eth-r several times.

The salt was then quickly placed in a vacuum desiccator

and allowed to remain at room temperature at less than

05 mm. from six hours to one week. The salts were then

recrystallizod from hot acetone with addition of absolute

34





35
ethanol until solution occurred. Only in some cases could

absolute ethanol or higher alcohols be used for recrystal-

lization, since this usually resulted in a solution from

which a crystalline product could not again be precipitated.

Totraallyldimethylmeth ethylene diammonium dibromide, which

was extremely hygroscopic, was the only salt which could

not be recrystallized.

The salts usually melted over a one or two degree

rcice and in sone cases decomposed slightly.

A. Preparation of tho Hexaallyl -Derivatives


1. Hexaallyletyllonediammoniumr dibromide
+Br- +Br-
(CiHi=C HCIIa)3 aE(H) a(CHaCI=CHa))a

Sixty and five-tcnths grarns (0.5 M) of allyl bromide
was added to a stirred solution of 50 g. (0.227 M) of tetra-

allylethylene diamine dissolved in 100 ml. of acetone, This

mixture was allowed to heat rently on a steam bath for five

and a half hours. The product was then quickly filtered,

washed with cold acetone and dry ether and placed in a
vacuum desiccator. Seventy-six grams (72.83 yield) of

white crystalline product was obtained i.hich melted at

1119-500 after recrystallization. The product was previously

reported(28) as melting at 147-8*.








2a iexaallyltrimethylenediamnonium dibromide
Br- Br"
(Ca=C HCHa ) l' (CHsa)a :(C HCIi=C )

The same ge:seral procedure was followed; eirht
and eirht-tenths grains (0.0724 M) of allyl bromide was
added to 8.5 g. (0.0362 M) of tetraallyltrimethylene
diamine dissolved in 20 ml. of acetophenone. The raixtu-e
was allowed to heat gently on a steam bath for one-half
hour. Fifteen end eighty-five hundrodtis gra-s (91.7%
yield) of product melting at 185-6* (with slight decom-
position) was obtained. The product was previously
reported(28) as melting at 185-7* (with decomposition).

3. iexaallyltetramethylenediaramonium dibromide
+Br- +,Br-
(CH32=CHICIa) ai (CHeB),4 (C(Ai=CiiC ls)a8

Usini the sane general proce-ure, 14.5 g. (0.12 M)
of allyl bromide was added to 15 g. (0.06 M) of tetra-
allyltetramethyleno diamine dissolved in 30 ml. of aceto-
phonone. The mixture was allowed to heat for one hour.
Twenty-seven grins (91*7, yield) of product melting at

130-41* was obtained. This was recrystallized from acetone,
ethanol, and ethyl acetate mixture. The product meltoL at

168-70*. Analysis calculated for C2He38iiHaira Br, 32.59%.
Found: Br, 31.70, 31.7h%.







4. ilexaallylpentamethylonediammonium dibromide
+B" .+Br

Usili the same general procedure, 13.5 g. (O.154 li)
of allyl bromido was added to 20.0 ,. (0.076 '-) of tetra-
allylpon:tanothylene diamino dissolved in 25 ml. of acetone.
IIeatin,' was continued for fifteen minutes and after the
resulting viscous liquid crystallized, 35.5 g. (92.2,: yield)
of product melting at 153-5@ was obtained. Analysis
calculated for C23i140s3rs: Br, 31.695. Found: Br,

31.70, 31.75,.

5. ioxaallylhexamethylenedia: morniLun dibroinide
3r" Br-
+ +
(CIHa=CIICH) aN (CI)CC.) iOCI!=CHa) a

The product used in this work was some of that
preporod by Angelo(28) in previous work. The compound was

made by addi;~g 28.2 g. (0,23 M) of allyl bronide to a
solution of 27.6 g. (0.1 2) of tetranllylhoxariothylene
diarLiAc and 25 ml. of acetophenone. Forty-seven and three-
tontlhs (rc.os (91.5 yield) of product melting at 179-800
(with sli-:ht decomposition) vcs obtaliod. Analysis
calculated for Cai, 2at2r:. Br, 30.83". Found: Br,

30.770.









6. Hexaallylheptamethylenediarmmonium dibromide
+Br" +r"
(Ciia=CCi:e)aW (Ca)7 H(Cial.=Ciia2 )

Using the same general procedure, 25.4 g. (0.21 M)
of allyl bromide was added to 30.0 g. (0.10l4 4) of tetra-
allylhoptamothylene diamine dissolved in 60 ml. of acetone.
The mixture was heated for ten to fifteen minutes. rorty-
threo gri:s (77.81 yield) of product meltiia at 192-30
(with slight decomposition) was obtained. Analysis cal-
culated for C2a,141B3ra: ITr, 30.02%. Found: Br, 29.94,

29.965.

7. He=nallyloctamnethylenediammonium dibromide
+3r" +Br"
(CH,=CHCHa)a i. (CliS) N(Cli LCH=CH)am

Usinr the same Coneral procedure, 16,5 g. (0.136 M)
of allyl bromide was added to 20.0 g. (0.068 4) of tetra-
allyloctamethylene diamine dissolved in 25 ml. of acetone.
The mixture was heated for ten to fifteen minutes on a
steam bath, whereupon a viscous liquid separated. This
was placed in a refrigerator overnight and 15 g. (40.5'
yield) of product melting at 195-7* (with decomposition)
was obtained. Analysis calculated for C2eH4,NsBri:
Br, 29.214. Found: Br, 28.57, 28.518.







8. ie.xa lylnon:amthylnediar- rionium dibronide
+. +ar-
(CI!=C [iC. )a.iJ( i [i )9:. (CHaC if=ClI )

Usin- the saoe general proce.-ure, 18.2 g. (0.15 N)
of allyl bromide was added to 16 g, (0,05 M) of tetra-
allylnonamethylene diisine dissolved in 40 ml, of aceto-
nhenone. The mixture was heated for fifteen minutes,
Twenty-three and six-tenths grams (8L, yield) of product
meltinr at 201-2o was obtained. Analysis calculated
for C27ii4Sii2Brq: Br, 28.26%. Found: Br, 28.41, 28.52/.

9. ie -aallyldocamenthylcned;omia noniium dibro.iide
+B"r +Br"
(CH=Cic.;)3 ;;a (C),o(CHaCiI=CHa)

Usiri: the same -eneral prc-cedure, 24 '. (0. M)

of allyl bzrondide was rLddod to 30 g. (0. 09 I) of tetra-
allyldecanIothylone diamino dissolved in 50 ml. of aceto-
phenone. The mixture was heated for thirty minutes. Forty-
three rre-is (83' yield) of crudo product was obtaiiLed. This
was recr'ystallized fror. n-anmy alcohol. With the bath
previously- heated to 180o, the product melts at 185-6
(with decomposition). Analysis calculated for Cgaoliiao:Bra:

Br, 27.'d2. Found: Br, 27.76, 27.71.









B. Preparation of tho DLmetlhyltetraallyl Derivativos


1. Dimeiethyltetraallylethylonediammoniuun dibromide
+TBr +Br-


CbT CIe T2

The product used in this work was some of that

prepared by Anigelo'(2 in previous work. To illustrate

the ge.ieral procedure, the preparation will be given.

Dry methyl bromide was allowed to bubble through a gas

inlet tube of capillary size into a mixture of 28 *-.

(0.127 M) of tetraallylethylcnc diamine and 50 ml. of

acetone. The mixture was stirred with a magnetic stirr-er;

cooled by an ice bath; excess C.:aBr vapors were collected

in a cold trap inr-.rsed in a mixture of dry ice and acetone.

Precautions to exclude moisture were tkl:en. The addition

of methyl bromide was discontinued after six hou:s and the

excess CIG3r was allowed to evaporate into the cold trap

as the mixture romalaed at room temperature overnight.

The mixture was filtered, washed with cold acetone and dry

ether, and placed in a vacuum desiccator. Fifty grans

(96.2/ yield) of product iielting at 191-2 (with slight

decomposition) was obtained. Analysis calculated for

CGeH3oNaBra: Br, 38.97%. Found: Br, 38.86".







2. DimethyltetraallyltrLnimthyljnediamEronium dibromide
+i Br'
I(, \ ,,



710 pa cuct used in tills worL u~a prepared(28)

previously Iusi. the above pi'oce'.ure. Thirty-ei;ht gtrt -s

(89.*6; yiold) of product nmoltin- at 1.,9-11 (closed
capillary) un; obtalioed. !LilyciZs calculLt-jd for

C, 7nI3a...-rs: -r, 37.67,. .ouid: 3r, 37.63, 37.924.

3. DiLiethyltetraallyltetra)iethiyl-onelae -.onium dibromide
+J3r- +-r"
(CHesC:Cne) :(CHe)s"(CHC.-c=Ce:=),


*.sii.: t::c sa:'o general procedJre, rnothyl bromide
was L-ced to 16 g. (0.064 I) of totr-aallyltetral-ethyl-ne

diamino disXsolved in 1C00 ml. o- acetone. Twenty-four

grc- .. (38,' yicld) of p: :duct m3tti-. at 163-5 (with

deco: iorition) was c-btcil.d. Analysis calculated for

Ce i'4-i'-. 2: -2, 36.47- I:''-d 1Lx, 36.' 5, 36.36L .

4. Dirie thy:ltetraallylpoe:- ta.nethylond lammoniumi dibrormide
+Dr- +.r



Usin the c-ar. general procedure, methyl bromide
:as added to 20 g. (0.0763 ;I) of totraallylpentamethylone







diamine dissolved in 100 ml. of acetone. Thirty grar-s

(87.3 yield) of product melting at 133-5* was obtain-d.
Analysis: calculated for C9,sineil~Bra: rr, 35.2 Found:

Br, 35.36, 35.26,.

5. Dimnethyltetraalllylhixajmerthyl lned irr.ron-iu dibromide
+Br" +Br




The product used in this work was prepared(26)

previously usii:V thle samer procedure. Porty-ai.-Iht iram'-.
(quantitative yield) of prod-ict meltinr at 202-3 (with
decomposition) was obtained. Analysi- calculated for

CeoliaNa3Bre: Br, 3h.28-. Pound: Br, 3U.284


6. Dimothyltetraallylhepta--.ethll;-nedi i Lr.o:luru dib. oride

+(H= +jr"
(CH=C):(CH (HgC;=C


'-sin~ the so- r.-" .rcl ?;ace auzro, r"ot.1rl brol..o.ide
was dc'3i1 to 20 C. (0.069 ) o tetaallylhpt.othy
diamni- dsloled in 100 nl. of aceto so. -'t,:;ty-el ht
a.;:i five-tonths gcrp's (86.2- yi-1d) of pr-od-uct .elting

at 193-"., (with docomposition) was obtained. Analysis
calculated for Csj1olaS13r2a 3r, 33.27%. Found: Br,

33.36, 33.32 .







7, Dimethyltetraallyloctamethylenediammonium dibromide
+Br- +Br"
(CHC; -:=C;i ) ( CHa ) (CH, C i=Ci )


Usiin the somc general procedure, methyl bromide
was added to 20 g. (0.068 M) of tetraallyloctamethylone
diamine dissolved in 100 ml. of acetone. Nineteen and
sevon-tonths grcas (58.7% yield) of product melting at

178-80* was obtained. Analysis calculated for Ci-4aisaBrs2
Br, 32.33,. Found: Br, 32.37, 32.4L0..

8. Dime thylt etraallylnonamethyloned iemmonium dibromide
+Br" +Br"
(CH=iCtlCla) (CHa)9 (CHCCHCH)
C IIa
-sing the same general procedure, methyl bromide
was added to 22 g. (0.069 M) of tetraallylnonamethylene
diamine dissolved in 100 ml. of acetone. Twenty-oight
grams (80% yield) of product melting at 187-90 (with
decomposition) was obtai;ind. Analysis c-lculvtod for

C3aHa441:Brs: Br, 31,45. FomLd: 3r, 31.23, 31.25!.

9. Dimothyltetraallyldecanothylenedianmmonium dibromide
+Br" +3r"
(CII,=C;lC, [,)21(CHia) o'(CGHC:i=CHa1)a

s13 C1a3
*sinr the same general procedure, methyl bromide
was added to 16 g. (0.06[15 M) of tetraallyldecamethylene






44
diamine dissolved in 100 ml. of acetone. Twenty-four grams
(8.% yield) of product melting at 163-50 (with decomposition)
was obtained. Analysis calculated for C,2li4il4Brs:a Br,
30*59. Fo-unid: Br, 30.61, 30. .










r CO ,- Cr- N IUA CO lA1 0 O
0 * * * 0 *
N r-I H r-N H 0 -:r (n
( E- 01c- c 0 o t'- -,t o co



1 t 0 1^1A C- r0 co Nc i 0 _:t
aa C- I a% *o\ v tr\ _t i\ a- F--
o ar< 8 8 I * a* * * * *
S0 I rI 1 r- 0 0. Co cO C:O --
.n n (n ( n f r\m\ c j ) (v m cm CU C\j
H H~
4 S

0% f N 4 1 0 N
S I \D CoO N Ns co
a r P * * *
S I I rN r c 0 C0

S0 I I- tM N N- N
* -
P3 0 N


0 0 0 0
O O O C0
H U'1 0 C,- nIl o r C 4 F.


S r- r- r-I H t r rH r-' C


Q 41 0 at 0 ca
a M
O a 0 In 0 00 C


0 0 gto V t 0 0
So df o e o t- e

0 0 0 0 0 0 0 0


o r-0 0 r-0 0 0 0 k
ca 3 0s r-I H 0T CT w o
E-4 r-I ,- P,
4o l 0 0 0- c 0
4) 04

S40rt 4z c 0 0 r 0



c -P 0 0 0 Q 00
B S r-i H i A Hs .


0 H H 0 H 0 0 H
o4 .> o 0 SI 0 g









0 00 0 Cd 0 0
r-1 H H H H H4 H "

O- It r- r -l (- I il












N ( o 0 on C N E- 0 0
o0 0% 0 cO 0 V\o
w 0 0 CO CO cO c 0 Co co



to-- :1 Bo r'oN ms sUO Co CC 0 r-o, No m -0
01> %0 0 0 (n C-\ f (MN N e i-; j NC\j 10 -t
H G2 ** * * *9* ** 9 * *
E-40 P- CO c -t' -0- ,.O,0 1A \ -I tn C N H -I-! 00

f- *


CC

- 0 0 0 0
odo i r4 r c3 0 0
(r n cF"-nco C- m CH



C% ao 3 C0 A o t v (D 0





#-4 0 r ri Hr t
* * *I I
S0 0 P 0 0 0

0 C.




0 o a C c 1 0 H
o 0 0 0 0
H E 0 0 e 0 O o i



H HaH ct H H


0 40
Q m f -4l 4q 4-I r 4 =;
;Hi a I H a 0 eI



z-4



P-H rl r0-i |A z H-i 0- A r-i H0 r-l I


W' 0 H H H Hd (D rH C cH H H0H
ro oo fr fr or o oo






2cs 0 0 H 0 0

4) 0 (0D 04c 0 0 00 0 0 00


a

00 00 00 00 00






V S.,S? ..,IO.; PGLYLE.I:!ATIOi: 02 JIS-.UAT.LiAi.Y AJHIC.LIU SALTS


A, General Jiscussion of Method

In an attempt to prepare the rosins in a rather uni-

form size ond shape, and to allow for dissipation of the

heat of reaction which could causo decomposition of the

quaternary ammonium derivatives, the suspension polymeri-

zation technique was a-ployed.

Using this method a number of difficulties are

encountoered. Constant high speed stirring must be main-

tained; the proper shape stirrer must be employed a:d the

proper carrier solvent must be used. An attempt to solve

these difficulties was tradee by carrying out various sus-

pension polymerizations usi-in tetraallyl ammonium bromide

as the mononer.

The stirring problem was finally resolved in the

folloriri, manner: A Palo Myers' motor, controlled by a

variable voltage rer-ulator, was employed at very high

stirri-IG rates. The voltage setti-ng for the high speed

varied froia 45 to 60 volts. This was due to slight

differences in the binding of the stirring sleeve. After

experimonting with various stirring sleeves, a heavy-

walled piece of glass tubing inserted in a rubber stopper

was used. This compensated for slight vibrations at very

47





h8

high speeds. The stirrer was made of -lass and was shaped

like a cork screw. The stirrer and stirrer sle:ve were

fitted as tightly as possible. Any initial binding was

compensated for by allowing the stirring device to proceed

at high speeds, usi-g- graphite and niA- ral oil as lubricants.

This was done previous to each experirLent and resulted in

more uniform stirring. The stirring sleeve and the stirrer

wore excessively and had to be replaced quite often.

Further, this wear caused the deposition of a fine lass

powder in the reaction flask.

The most difficult phase encountered in trying to

obtain uniform bead particles was the sticky period, which

caused merging of individual dispersed particles into

agclomerations.

Hohenstein and Mark(29) indicate that polymerization

in a heterogeneous suspension in which the monomer is

mechanically dispersed in a liquid not a solvent for it

or for any species of polymer molecules, and in which the

initiator is soluble in the monomer or monomer phase,

results in a polymer bead or pearl. In such cases the

polyr;erization takes place in each monomer- lobule and

converts it gradually into a polymer bead or pearl; the

liquid plays only the role of a carrier, which favors

heat transfer and aritation, but does not interfere with

the reaction.





19
However, after a period of tire, the globules, which

now represent a rather concentrated solution of polymer in

monomer, become 1usmmy and upon mutual contact in the s:et.am

stick together and form agClomerations which cannot easily

be broker up into individual lobules of the original size

or shnpe. It frequently happens t:-i t upon continued stir-

ring the ar-lonerations do not disintegrate, but coalesce

into laroe:r units. 'Te authors(29) describe several ways

of brin.ii.gL the system over the sticky period without

mergi ,- of the individual particles:

(1) Suspension stabilizers can be used that can

cover the s'.rfaces of the -lobule with thin layo s of

inorrt nic or organic substances which do not interfere

with the rEction, but prevent or diminish tho toredency

of thv globulos to stick together during the gunnmy period.

(2) The intorfacial tension between the carrier

solvent aCnd t'-o mo..onorr phase can be increased by dis-

solvi-r electrolytes in the carrier solvent if it is water

(3) The density of the carrier liquid can be

adjusted to the doesity of the globules during the cgmy

state. In doi-.- so, the tendency of the stick-y Clobules

to accumuleto either on the surface or at the bottom of

the suspeosio:i is removed.

(!I) The viscosity of tho carrier liquid can be

increased and thereby moko it more difficult for the dis-

persed globules to collide vigorously enot h to mere.








In this respect, many experiments wlore car'rice

out with tetrEallyl arimonrlimn bromide as the noiromcr;

the p:'cedi.43 su-.rcstions were ecnplo.-od until the bost

conditions wer found. Durin.- these e:oerimrints stabili-

zers 3such as bentonite and talc were used; carrier solvents

s ch as ethyl benzeae, chlorobenzene and toluone wore used;

increased viscosity of tihe carrier liquid was investigated

- various additions of mineral oil; since the cnrrilor

liquid in this particular poly-jrrization was not water,

adjustment of the interfocial tension by the addition of

electrolytes was not employed.

The best conditions for carryi;: out this particu-

lar polymerization were as follcw;i: A stirring rate ad-

justed to l0-60 volts as indicated previously, a tempera-

tuie level of 600, a carrier solvent co.nsistinr, of 150 ml.

of mineral oil anrd 50 ml. of ethyl boa.zone, a modiomer

phase consisti-irn of a half a milliliter of water per Lrar

of quaternary salt, and an initiator co-:centration of two

drops (approximatel- 0.012 g./drop) of 6G~ t-butylh: dro-

peroxide solution per gram of quaternary salt. In all

cnses, using totraallyl aimioniuTr bromide as the rLonoi;mer,

90-1lC.- yield of i-solub.-li resin was obtained. However,

in no case were any wAll-shaped boads forced. The product

approached a spherical sliape, but had much random shape dis-

tribution. Iuch of the size distribution of the particles

obtained was in the desired 15-60 mesh region.




51
The stirri-g at 600 was continued for forty-eight

hours after which the product was filtered and washed

succeesively w-ith heptrne, acotor.e, L 1cohol and other,

The p-.oduct was then so:':old for apprlo:-iately one hour

in hot hontan~ and thon filtered and washed as before.

The product was thon dried for two;ity-four hours in an

oven at 60o. This proccdu:s i.isured the removal of the

film of nincr.al oil* At this time the swolli,,-- cooeficient

of the rczin was dot;e i:ned. The ratio of the wet bromide

volume to the dry bromide volume of the rosin is the

swellinii: coefficient. It was zaeasur-ed by allowing the

dry broriid foio of the ro ain to settle in a r,-aduated

cylinder and detormini--,t the volume. The wet volume was

deterlaiied by allowing the stne rosin sriplo to soak in

distilled water until no further chaae in volume vas

appornt.

This ,-eneral p:.ocedlure was applied to the polymeri-

zation of the quaate:.nary derivatives of the diamiAiles des-

critod previo, sly. In al. cas2s, except for te: he::aallyl-

d .ec ethylenene end the diI.athyltotruallylhe:mr.ethylene

di-crnmonliuri dibro.rde derivatives, tLe. rocin product vWs

not perfectly bhad sI-ool. In :ost cases t'ip p.-oduct was

of rmido:n slhpo, but usually in the desired 1"-6' nosh

reG' ioe. o. risi-i particles of t:.iso two dc:-ivatives

ijore perfectly bead shaped; only the docamethylone





$2
derivative retai.ied its board shapo th.roivhoat all operations,

such as washi:lg, exchanCg or recycll.i

The poly..er-ization of th.; quatcrnary derivatives

of txo pot Iotlylm-e an'. oct.:intLhylone dcrivativos dif-

fered fro:. the -enoral prccadure describcd abcvo. T>ese

pol7~oriPizatio.js wore carried out in the bulT prise ;sriSg

t'.e sac.ie mo:ionecr, initiator c.d uater ratio, 'ile resulting

resi.is were ground to the desired 15-60 raesh size alid

treated s:-nilarly.

B. Ti!perirm.ntal liesults

The following table gives a sur.m.iary of the results

of the polymerization of the quaternary derivatives of the

preceding diamiines. It is believed that any yield over

10e0 is caused by glass particles deposited from the

stirring mechanism. In this section the resins will be

given the following notation: CaH = polymer of hexa-

allylothylenediammoniur dibromide, CUDT = polymer of

dimethyltetraallylethylene disamonium dibroride, etc.






2 L. -.:i V


Total wt. Wt. obt:inod Swell.
asin OLtaLi. ed ( ..) in 15-60 .;e1sh (c.) Yield coeffr
.. -_ ? - -__ l~ l- ] l l l l l ll l I. . ,, .... !,, l - -


C.H
COH
C4H

CaH
COH
CTH








C3DT
C4DT


CGDT

Ca-DT
C 9DT

Ce DT
cOT
CoDT


31.3
23.8

15.65

17.7
19.5

19,3
8.2
17.6

20.5

32.4
12.25
21.8

19.1

15.1
18.95

i4.95
18,9

21.0


30.5
14.2
14.0
16.9
15.1

7.4-
7.5

14.2
19.8
29.0

11.5

5.3
18.1
14.8
16.65


2.65

17.8


100+

100+

78.3
88.5

96.5
96.5
91.1
88,0
100+
100+

76.1
100+

95.5
94.0
94.8
99.6

94.5
100+


1.45
1.52
2.97
1.88

1.69
1.64

2.33
1.76

1,43
1.48

1.67

1.44
2.11
2.21

1.83
1.69
1.80

1.50


- II "---- I~








VI EVALUATIOUi OF CAPACITY SiTU3IES


A, Method of Analysis

The method of determiininC the capacities of the

resins in this investigation was essentially a titration

procedure similar to a neutralization titration of a

strong base and a strong acid, A known quantity of the

poly-quaternary ammonium bromide form of a resin was

quantitatively converted to the poly-quaternary ammonium

hydroxide form of the resin. An excess (compared to the

theoretical number of milliequivalents of poly-quaternary

ammonium bromide) of a salt of a strong mineral acid was

added to the poly-quaternary ammonium hydroxide. An ex-

change took place and hydroxyl ions were released into

solution. Suitable titration of the hydroxyl ion content

by a strong mineral acid then determined the exchange

capacity of the resin in terms of milliequivalents of

hydroxyl ion exchanged per gram of dry bromide form of

resin.

This method is essentially that described by

Butler and Bunch, (1) and Butler and Ingloy,(2) in which

the actual titration was carried out using a Beckman pH

meter, adding 1.0 ml. increments of standard acid at

3.0 minute intervals and measuring the pH of the solution

54







at the end of each three minute interval. This routine

was continued until the titration was complete. The pH

was plotted against milliliters of acid added and the

capacity was obtained from the amount of acid required to

cause a sharp break in the curve. However, the capacity

of the resin measured by this method never approached the

theoretical capacity of the resin, and the break in the

titration curve usually showed a shoulder which was

interpreted at the time as amine capacity. (3)
Ilusa(6) developed an "equilibrium" titration

method for detenrining the capacity of this type of anion

exchan-o resin. His method consisted of preparing equal

samples of resin in the bromide form, quantitatively con-

vorting each sample to the hydroxide form, and adding

various known quantities of standard KBr solution and HBr

solution. The pH of each sample was measured initially

and at various intervals, extending to 420 hours in some

cases. A plot of pH versus milliliters of standard HBr

originally present gave a titration curve which appeared

very similar to a standard acid-base titration and from

which a relatively accurate determination of capacity could

be obtained. The advantage of this type of titration is

that the measured capacity is the capacity attainable after

the system has reached equilibrium. The disadvantage of

this method is the prolonged period of time involved and

the number of samples measured.







The equilibrium capacities, measured by Husa, (6)
often approached the theoretical values and were much
higher than the values determined by the rapid titration

method. Differences in equilibrium time, pH of solution,

and ionic strength were suggested as factors causing

variation of the results given by the rapid method and
(30)
the equilibrium method. The equilibrium titration

curves showed no amine capacity, as was suggested pre-
(3)
viously, but gave a smooth curve with a sharp break.

As an illustration, the equilibrium titration curve for

the resin of hexaallyldecamethylene diansnoiium dibromide

follows, This curve appears by the courtesy of W. J.
Husa(31) from his Ph.D. Dissertation, University of

Florida, 1953, P. 82.

The following procedure was used to determine the

capacity of the bis-quaternary ammonium resins obtained

in this work. A weighed quantity of resin (approximately

0.2000 g.) was placed in a sintered glass filteriig funnel

(medium). The funnel was stoppered and the resin soaked
in ten milliliters of 14 UaOl solution; the resin sample

was filtered, washed and soaked every day for 10-12 days.

After this time, the turbidity test for bromide ion,

precipitated as silver bromide, was less than ten parts

per million. The resin was then washed free of any

hydroxyl ion by using distilled water and checking the

filtrate with phenolphthaleini



























CO
Ok












0





C
r r-< I *










d











S0 6y
''^i f-: l -






















C
3 I








0 -C
r- _^ r1





.'o "' 0-
F: t









0 '-C















CLA
in 0- =

< ,. C





0 C

s- C-; i-
;\ --f **
i. r t -
rt1-S

c, I'.
W U(
^ ^^


fal ri
^ *f-l
W 4i





5p3

The resin was placed in a 250 m2l. beaker with 75 ml.

of distilled water. To this mixture was added 50 ml. of

0.20 i KBr solution (ten milliequivalents of bromide ion).

After standing for five minutes, the mixture was titrated

with 0.0195 1N HDr solution by using a Beckman Model K

automatic titrator. The endpoint was set at pH 7 by

adjustment with a standard buffer solution, and the

anticipation rate was set at 7. 'he first 15-20 ml of

standard acid was titrated rapidly, but thereafter the

titration proceeded rather slowly until the end of the

allotted one hour period. There was still so.o exchange

takii-,q place at the end of one hoi.r, as was observed when

somie samples were allowed to titrate further. This type

of asymptotic approach to complete neutralization was to be

expected from tho data of (usa.3

The theoretical number of rilliequivalants was

calculated for each resin based on monomer units. The

actual number of milliequivalents exchsaned per hour was

calculated and was expressed as the fraction of theoretical

exchange per hour.

B. Summary of Experimental Results
of Capacity Determinations

The capacity determinations of the bis-quaternary

ammonium resins were carried out in the manner described.

A summary of the data obtained follows in Table VI. In





59
this table thie zoAosll.-s are -ivan th-o same notation as before,

whero 0C21= res1n of hoxaa.llyletbAyar3 nndia monium dibromide,

CaDT = resin of

dibronide, atc.














0




4
41,





0













Sr








0
I o
*









H're
a













0
s,
*








r


.0 CO CO C ooz CO
0 O0 N00 0o
oo L- E- sO 0
'.o c- co a'. r o IA '.0
* * * * *
0 0 0 0 0 0 0 0




'\ U\ r- O' 0 lC- 'O t

* * * 9 *





o, H,- H rn 0 1UN 0 0
0 0 0 0 0 0 0 0
c rt O \0


0 C O NO c CO (D 0-


o Noo oD -
('n ^ r< r, r`- CND(l

0 00 0 0 0 0
* * * *



o N c H m -- 0
o H (O H C'- !\- 0
0 0 0 0 0 0 0 0

o c ; 0* 0 < 0




L tS- co o01 - C
-r A -O t '0 C"j t-
* 0 0 0 9 9 0 *


S 0 4 U 0 a E- E-


I0 o (n
O H or
* *





I\CO (co

* *





000






e
0 0 0







c Cm C\J




cr- r-
* *




0 4\ .O
m 0 0
* 9






* *





61






04

"qa 0 0
,, N O NO ,0 ,O t.
S 0 0 0 0 0 0 0




COm r- O3% 0 H
rit rH r- -H 4 H %3
O rI A' t\ IAJ U\ -.;
0 0C 0 000 0





r: H N o ( N c
SO 0O

o*




: H O NC CO O E r- C-
S E * * *
CT C 0 0 c 0 0 0
MH ! al N W cM cN cuN











4
H 0 -4- \I.C 0 r-f r 0% 0 X
6 p o o 00 o o o
H OH ZX N (M W) N N
-C0 0 0 0 0 0 0



*I r* * *

09 0 9 0 6
-a .ti s < -^ r.. l? cy 0, 0 0

I N H H- r-I H

0


0 0 0 0 0 0
0 0 0 0 0 0 0







C, Discussion of Capacity Determinations

The type of rapid titration determination which was

carried out in this work proved to be quite adequate. The

time and number of samples were kept at a minimum. The

Beckman Iodel K automatic titrator proved to be satisfac-

tory in measuring the fraction of theoretical exchange per

hour. It is believed that suitable experimental techniques

could be developed, involving the automatic titrator, to

differentiate between the initial stage of exchange,

apparently due predominately to a mass action effect, and

a latter equilibrium stage of exchange, apparently due to a

diffusion rate factor. It is conceivable that relative

diffusion rates of various ions could be determined and

differentiated from total exchange rates, which were

probably measured in this work.

It is believed that an equal milliequivalent basis

would be more reliable than an equal weight basis of com-

parison of the resin exchange rates. In the previous

determinations it should be noted that the theoretical

number of milliequivalents of the resins ranged from

0.707 to 0.977. However, the adequate excess of initial
bromide ion should compensate for differences in mass action

effect due to the unequal unit milliequivalent basis; the

initial quantity of bromide ion consisted of 10.0 milliequi-

valents.







The most interesting effect, as seen from the

exchange capacity data, is the definite indication of a

maximum exchange per unit time of the pentarnethylene

derivatives. This effect is shown in Figures 2 and 3.

It could be merely a coincidental factor dependent upon

some non-uniform polymerization technique. Although every

polymerization was carried out as uniformly as possible,

there were, in many cases, noticeable differences in the

induction period, the globule size during the polymeri-

zation, and the final particle size and shape. However,

the effect of maximum exchange of the pentamethylene

derivatives is a factor that could be considered from the

basis of differences in the unit structure of the polymer.

A suggestion as to the uniqueness of the struc-

tural configuration of the pentamethylene derivatives is

the possibility that uniform linear chain growth through

the allyl double bond would result in a polymer with five

carbon atoms between each exchange center. This type of

linear growth would cause only the pontamethylene deriva-

tives to have equal numbers of carbon atoms between suc-

cessive nitrogen atoms. This is illustrated in the

following scheme:







Fir'-urc 2


1.0



~ 9



0. 8



0.7



0.6



0,5


6 7 8 9 10


Figure 3


2 3 4 5 6 7 8 9 10
Number of Carbon Atoms in Methylene Chain


2 3 4 5


Hexaallyl Derivatives


0O 0
0\


0.9



0,8


0.7



0o6


= i f i





Ra Ra
\+ +/ "n
CH,=CHCH2I- (CH) 6-1a-C 1,CH"=C H C+Z*
Initiation with free radical Z-

\+ + H H
CH9=C;C: .i(C;i a)5l-CIIC--C : Z
H
PropaCation with monomer molecule

HCH


Cja=CIC1\1(CHa)e"-C3C-sC-C-CJiaLl-(CH) BGi-CHsCBH=C1a


Ca C4, > C t etc.
Since there has been no definite proof of structure

of these derivatives, this suggestion is to be considered

as a possible course only,

The capacity data of Table VI seems to show no

definite correlation bctwc-n the swelling coefficient

relatedd to the apparent degree of cross-linking) and the

e :chanco capacity of the resins. It was expected that

replacement of one allyl Croup of a quaternary anmonium

derivative by a methyl group should give a resin having

less cross-linking and higher swelling, there being fewer

allyl double bonds available for cross-linking. Conse-
quently, the methyl quaternary resins should have higher

excha.re values if we consider a low degree of cross-linking

to allow easier ion movement in a less tangled network. How-

ever, the data in Table VI does not bear out this contention





66

and, contrary to expectations, the methyl derivatives have

less exchalce capacity per tuuit time tl.in the correspo-iding

allyl derivatives in practically every case.

There was no expo~imental evidence obtai.Lud to show

a connection between the ex:cr-;go capacity of the recias,

particulac-rly ac rogards their selectivity towards ions of

different size and charge and their respective swelling

coefficients, apparent decrees of cross-linking or

variable distances between quatom ar7 ammonium centers.
(32)
However, some work32) is presently being carried out in

this laboratory by L. G. Kulkarni, relative to the possible

selective exchange properties of these resins.











VII 2CLY ..IATIO CiAiI2:: z:,r: .

Previous work(1-5) carried out in this laboratory

pertainiir to the free radical initiated polymerization

of allyl or substituted allyl quaternary ammonium deriva-

tives gave infusible and water-insoluble polymers only

when derived from monomers containing three or more allyl

groups. The product resulting from the attempted polymeri-

zation of a monomer containing one or two allyl groups

either was not polymerized or was assumed to be non-

crossed linked, since it was completely soluble in water.

These results are contrary to the accepted views

that mon omers containing one double bond result in a

linear chain type product possessing some degree of solu-

bility; monomers contahing two or more double bonds result

in a t-Lre:-dimonsio:al, cross-linked product possessing

little or no solubility. Since the products of the polymer-

ization of allyl quaternary ammonium salts have appeared

somei~rint abnormal in their properties, it was decided

to investigate the mechanism of this type of free radical

initiated polymerization.

A piperidine ring structure for.ied by an alter-

natinr intramolecular-intermolocular chain reaction was

67







oonsider3d as a possibility which coo-ld preve'; cross
linking in the polymerization of q~.,atornary ~umnoni,

derivatives containing only two allyl do-ble boiuds.

Simpson, ::olt and ZCite(33 have indicatC frc.,i the

corrolatio.l of data on ti:e polz:iorlzation of diallyl

phthalate that at the Cel point m.uch fc' the cazrboo"-ca&-bo:;

double bond, suppocosly free to crose link after; tie

Efel point, is tied up as a cylic structure due to intra-
moleculr polynorization.

It umas decided to approach this subject fro,!, the

aspects of suitable stnthetic adaption of mononer struc-

ture, der.radation studies, infrared anal-ses and molecular

weight determinations.

A. Monomer S.itheses


1. Preparation of diethyldiallylarrrionium bromide
+ Br-
(Ca1Gcd) aN(CHci1=CH= ),

Sixty and five-tenths grams (Q.5 M) of allyl

bromide was added to 50.0 g. (0.44 M) of diethylallyl amine

dissolved in 100 ml. of acetone; the diethylallyl amine
had a boiling point of 110 and n 1.4198; Libermann

and Paal 1 report the boiling point as 110-13o. Upon

addition of the bromide, the mixture became cloudy and

crystals began to form. The product was washed and






decanted several tL.ies with cold acetone, filtered and
dried LiL a vacuLiu doeiccUtor. i.A. -hy-seven grans (804.1
yield) of white, hygroscopic crystals, melting at 1550
(closod capillary) we-e obtained. .aecrytallizatlon from
hot acetone and a miLall iam1-nt of absolute etimanol gave
the saXe1 -..eltlI; point.

2. Preparation of triethylallylarmonium bromide
+ Br
( 3C T~ ) z r (CHnC' IC=Cl )
.LiIrty-sis and t:-i.L.-tenths grams (0.3 :) of
allyl ~ibroc i.i was added to 30.0 r. (0.293 I) o' triothyl
aiinc. '2h1 sane -roc-d'-e was followed and 60.5 g.
(92.1"J yield) of uhite hygroscopic crystals was obtai..od.
The product :.ialted at 229-31* (with deccrposition; closed
capillary).

3. Preparation of 1,1l-bis(diethlylallarrnonium)butene-
2 dibronilde
+Br- +r-
+ +


CIIa=C_'C Ci =C;Iz
The intermediate dianine, 1,4--bis(diethylamino)butene-2,
was prepared in a 31.83 yield by the procedure of Amundsen. (35
One hundred and sovonty-eicht grams of light yellow liquid
boiling at 118*/20 mmr and having an n2'- 1.4573 was ob-

tai.ied. Amu-ndsen reports a boiling point of 115-16/20 mm.
and an nD'. 1.4532. The product rapidly decolorizes when
allowed to remain in a sealed container at room temperature.







Twenty gr'sa (0.10 ") of l,13-bi((diethlamino)
butene-2 dissolved in 50 ml. of acetone ;oas allowed to
react wIth 26.6 c. (0.22 ;7) o-' allyl jromide in the same
manner as bL.JiAro. The cr-ue prn:d:ct waz rocr-ystalliz d
froi hiot acotone aind absolut h o thanol; 26.5 rg (59 ryielrl)
of white, powLeri product .: ltiA, at 185.-6 (uitlh de-
composition) was obta-ioed. Co-t- (36) repc-rt the meltiIg

polnt of this pr-oDct t ;it'out rocyst3allization as 172-3o.

4. Preparation of diallyltetraothyldocanothylenodia~r.oniuni
dibr-Oiide
+Br- +Br-
C:i3Che) a-(Clio) ,oe-'(CICIieH)8

CGls=CH zIGCriH=C
heo intermediate diamine, tetraothyldecanethylone
diamine, was prepared in the sane manner as tetraallyl-
decariethylene diir.ine in Section II. One hundred and
twenty-six gra':s (1.73 M) of diethyl anino in a slurry
of 81: r. (0.5 A) of ..aHICO aeid 50 ml. of water, was
allowed to reflu.: with 52 g. (0.173 ::) of decamethylene
dibromide for ten hours. Thirty-nine gra-!s (79.5$ yield)
of light yollIo liquid boili at 120-Co/0.02-0.0.5 ;.
and having an nD'- 1.4493 was obtained. Analysis cal-
culated for COa84o:0: C, 76.2,; :~ 14l..1; N, 9 .86.
Found: C, 75.97*; H, 14.Gb8; ;, 9.,2-.
Fifteen Grams (0.053 M) of tetraethyldecamiethylane
diamine, dissolved in 50 ml. of acetone, was allowed to





71
react t:thi 13.3 g. (0.11 ) of nlll ..ro-idc. After being
treato... i tLho U ai- i.rL.1 as Lioore, 17.0 g. (61.2

i-eld) of ::2le, vor, y-'ocopic salt minl.tlC. at 2$-70
(,it, dico.:.position; cluodu capilla.y) was obtai.icd.

5. Attempted preparation of quaternary salts of 4-diethyl-
zLihi.oj c..-t,.i s-1
+
(COTCTV, (C ,IICC H= ) IsCiH)


The intermediate amine, 4-diethylaminobutene-1,

was prepared by roflu:din for ei-hteon hours a mixture of

190 r. (2.6 M) of diethyl amino, 84 g. (0.9 M) of NJa;;CO,

50 ml. of watei and 30.0 g. (0.33 H) of 14-chlorobuteno-1.
(The l-chlorobutene-l was rcdistilled and the portion

boiling at 75.0-75.5* anid having ann nu 1.4238 was used;
(37)
Juvala reports the boiling point as 75.0* and the
so
nD as 1.*233). Aftar separation and purification, 15.L g.
(36.6.J yiold) of clear liquid boilinL at 134-5* and having
as
an nD 1.4250 was obtained. An infrared spectrum indi-

cated the characteristic terminal double bond absorption
ihilch was deslled in this preparation.

Attempts to prepare quaternary ammonium derivatives

of 4-diethylaminobutenel- did not prove very successful.
Only in the case of the allyl quaternary was a crystalline
product obtaoi.ed. The reaction of equlmolar portions of

allyl bromide and 4-dlothylaminobutcue-1 dissolved in





72
acetone, gave a white, hygroscopic product which melted at
208-100 (with decomposition).
The reaction of equimolar portions of 5-bromo-
pentene-I and 4-diethylaminobutone-1 dissolved in acetone

gave an oil, which, after many recrystallization attempts,
gave a very hygroscopic semi-solid. This semi-solid was
obtained by careful washing, decanting with cold acetone
and allowing the residue to remain in a vacuum desiccator
at leas than one millimeter for several days.
The mixtures of 4-chlorobutene-l or 6-bromohexene-l
with equimolar portions of 4-diethylaminobutene-1 dissolved
in acetone showed no signs of reaction after gentle heating
and after remaining at room temperature for six months.

6. Attempted preparation of quaternary salts of 1-
di(pentene-4)amino butane
+ X"
(CHp=CIICHC,;mGCilgC) al' C IC aCHaClla


The intermediate amine, l-di(pentene-l)amnno butane,
was prepared by refluxing for three days a mixture of 21.9 g.

(0.30 M) of n-butyl amine, 87.0 g. (0,7 M) of iHaCO3*HO,
100 ml. of water ard 100 g. (0.67 1) of l-bromopentene-4
(b.p.-53-4L*/51-2 mm. and nD 1.4636; Juvala37) reports

the b.p. as 56*/75 mm.). After separation and purification,

41.5 go (66.2% yield) of light yellow liquid boiling at
228-37 was obtained. This was redlstilled and a light
yellow liquid boiling at 114-160/9-4 rm. and having







73
0so
an nD 1.4525 was obtained. An infrared spectrum gave the
characteristic terminal double bond absorption desired in
this preparation. Analysis calculated for C,4HaTNI
C, 80.3,; H, 13.9j 1N, 6.7,. Found: C, 80.41 ; H,
12.60,; N, 6.73/.
All attempts to prepare crystalline derivatives
of l-di(pentone-h)amino butane were unsuccessful. The
amino was reacted with methyl bromide, methyl iodide,
allyl bromide, benzyl chloride, hydrogen chloride and
hydroron bromide in various solvents, but no crystalline
product was obtained.

7. Attempted preparation of quaternary salts of
1-di(hexene-5)amino butane
+X"
(CH,-=C..C IsCiCI4 ) 81 (CiC8i iCiIl )
/
The 1-bromohexene-5, used to prepare the inter-
mediate amine, was made by treating at 0-3* a dry pyridine
solution of 5-hexene-l-ol with phosphorous tribromide.
Seventy-two and one-tonth grams (Qi4.2- yield) of 1-bromo-
hexeno-5 boiling at 78-800/61 rm. and hEvvi~ an na~- 1.4676
was obtained. Price(38) reports the boiling point as

75-80/61-2 mri. and the na as 1.4620.
The intermediate amine, l-di(hexene-5)amino
butane, was prepared in the same manner as before.
Fourteen and five-tenths grams (0.20 M) of n-butyl amine,









82.5 g. (O.0o M) of K4GOa1 1/2(H0o), 2.o g. of copper
powder, 100 ml. of water and 70.1 g. (0.43 M) of 1-bromo-
hexene-5 were refluxed for three days. After separation
and purification, 29 g. (61.1 ; yield) of a light yellow

liquid boiling at 175-840/64 mm. and having an nD 1.h51:6
was obtained. An infrared spectrum gave the characteristic
terminal double bond absorption desired in this prepara+i? f .
Analysis calculated for C,8Hsi3,,: C, 80.93"; H, 13.17;'

N, 5.90. Found: C, 80.41W; iH, 12.8r. ; N, 5.98:.
All attempts to prepare crystalline derivatives

of l-di(hexene-5)amino butane were unsuccessful. The amine
was reacted with methyl bromide, methyl iodide, allyl
bromide, benzyl chloride, hydrogen chloride, and hydror-en
bromide in various solvents, but no crystalline product
was obtained.

B. Polymerization Products and Results

In an attempt to explain the water soluble property
(apparent lack of c:oss-linking) of resins obtained from
the polymerization of quaternary ammonium derivatives con-
taining two allyl groups, an alternating intramolecular-
intermolecular chain reaction forming a piperidine ring

structure, which could cause a non-cross linked type of
chain growth, was proposed. The following scheme il-
lustrates the possible mechanism. (Z'= initiator, R

saturated group).







+ sCIIa=Cilh
(1) Z-+ fH1
-7,C -=C]1 =






H H
(2) 1II

H A2 CIC-C-ZI1





+ Cli C-C-Z
(3) R2l
,H\
HCis-CHa




.CHeC-C-Z
R/ \p"
Rp Ili


H H
S CIIC-C-Z


CHaCIi=CHm



+ CHaC-0-Z

H 2
XaC40-c-z


Initiation





Intramolecu-
lar Growth


+i IiaCH=Ci

CaC Bgl


Internolecu-
lar Growth


C112=C *-. I



HH


(tI) 1n2i(


Intramolecular
Growth


etc. 4







As a possible approach to the structure of those

poly-allylquaternary ammonium derivatives, a study of the

reactions of suitable monomors and the degradation of the

polymers was attempted.

1I Degradation of poly-tetraallylammonium bromide

A decomposition of the poly-quaternary ammonium

hydroxide, leading to simpler products which could be

identified and possibly associated with a ring type struc-

ture, was carried out. Since a quantity of poly-tetraallyl

amn-onium bromide was on hand from the previous suspension

polymerization trials, this product was converted to the

hydroxide form and degraded,

Seventy-eight grams of poly-totraallyl ammonium

bromide was soaked in 250 ml. of 1' 1aOH solution, filtered,

and washed with distilled water. This process was carried

out once a day for 28 da-s until the test for bromide ion

was less than 10 parts per million. The product was

washed free of hydroxyl ions and filtered as dry as

possible before it was paced in a vacuum desiccator. The

product remained in the vacuum desiccator at a pressure

less than 1 mm. until it reached a constant weight. This

took one day and the product weighed 62.5 g.; the calcu-

lated weight of hydroxide form should be 60.2 g. from the

initial 78 go of bromide formn







The converted product was placed in a distilling

pot connected to a receiver immersed in a Dewar flask

containin:1 Dry Ice and acetone, A second trap cooled in

a similar manner was connected in series. The product was

heated at 300-500 for four hours while e the pressure was

maintaieod at 10 nm. Tho contents of the flask b-came a

tarry residue. A total of 14*0 ml. of a two layer product

was obtained. IFurther heating for an hour at 3500 and

less then 1 mmn. ave no further product,

A total of 12,8 :. (14.0 ml.) of product was

obteaied; 5.0 ml. of a blach upper layer and 9.0 ml. of

a light yellow layer. The black upper layer was re-

distilled; 1.1 ml. of liquid boiling at 60-80O/760 mm.

and having an n3l- 1.4620 was obtained; 1.5 ml. of very

dark liquid boiliAir at 60-110/0.05 ru;. anu having an
as
nD 1.51 was obtained (the liquid was too opaque to give

an accurate nI"); some tarry residue was left in the distil-

ling pot. Both fractions turned darker upon standing and

the higher boiling fraction became quite tarry. Both

fractions gave a positive test for nitr'ogeon. An infrared

spectrum was obtained for each fraction, but no clues to

structural features were evident. Attempts to prepare

derivatives of these fractions failed.

A quantity of tle residue left in the original

distillil. pot was extracted with ethanol in a Soxhlet

extractor for one day and the alcohol solution evaporated.






78
A residue was obtained which upon f-rth.-tr r3crystallizatlon

yiolod 0.2 g. of white solid, Thiis product decoiiposod

over a long ran- and ras not coi.pl't4ly decomposed at

360,. Fulrt'ler attempts to ide.itify theoe picducts were

abandoned.

2. Degrada.tion of poly-dial lyldiethyl amnionium bromide

Twenty drops of 60% t-butylhydLiopero:-ide (ap.roxi-

rntely 0,012 g./drop) was added to a solution of 8.0 g.

of diallyldiethyl airmionium bromide and 4.0 ml. of iwter.

The mixture was allowed to remain open to the atmosphere

in an oven at 60 for 48 hours, Tho resulting i:hite

hygroscopic product was ground to a fine powder and dried

for several da:.- in a vacuum desiccator. Eight grams of

product, meltinC with consider.jle decomposition at

346-514, was obtained. The product was quite soluble in
water' and ethanol and gave an i-Lediate halogen test i en

treated with lAgI;g solution.

The product was converted to the hydroxide form in

an attcipt to degrade it. It was bclievod that tlhe loss

of ethylene and water from the hydroxide form would result

in a poly-tertiary ,mine or simple.- products that could be

studied. Sevcn and one-tenth grc-is cf poly-diallyldiethyl

ammonium bromide, dissolved in 200 ml. of water, was allowed

to pass through an ion-exch-nge column of Nalcite SAR (con-







verted to the hydroxide fonu). The solution was recycled

three ties and tested for bromide ion; after the third

cycle, the test was less than ten parts per million.

The poly-diallyldiethyl ammonium hydroxide solution

showed a faint pink color when tested with phonolphthalei;.i

tho pl value obtained from a Boclca- pHl meter flluctuated

considerably .d was ti:okin as 11.75 to 12.05.

lhe solution was ovaporatoe to on:e-fourth of the

orlriial volurie on a hot plata and to di-yness on a stewa

bath. The tarry residue vioighed appro::i.atel, five crens.

The p-ro.uct v:a vato insoluble and for:..d a gel when

heatrd with water. The p.-oduct was very slightly soluble

in ethyl c1co.icl, carbon disulfid,, dinlethyIl ormamide,

carbon tetrac'_loride and chlorcJor..; insoluble in benzone

and diethyl ether, lhero ta, co.-,slderable doubt as to

wfet.heor the product was soluble to any extent in these

solvents, or Ut .:t :..r a small q.c.itlty o- p.-o.~-ct formed

a I,.l.

The tr:i,- residue was diCested in carbon tetra-

chloride ;nd yielded a product which could be ground into

a finc poooi dor atez vacuum dryinz. The product decomposed

over a \:id. id:-n.-e znid i:as not co:.pletoly decomposed at 360*.
Analysis found. C, l:9.1, li9.5i; .., 8.39, 8.23X; P, 7.6.

InfraroJ sp.ctre of the pi:oduct we:-e obtai-ed; the tociinique

of depociti:A' a fili frori et:anol, carbon totrLchiloride







or carbon disulfide solutions was employed; t:ih tec:'iique

of incorporation a small qnrltity sample in KBr powder

and pressing a salt plate was also seod. .. 31cs(39)

kindly carried out this pr'ccedu_-e.

When bromine (35 in CC14) was added to a rLixture

of product in hot CC14, an orr--'o precipitate iwas foi-.cd.

This addition product was filtered, dried a.-d ground into

a fine powder. The product dCco.iposed over a ijidc rr.ce

and was not conplstely docompos3d at 360. Analysis found:

d, 5:' .s-

The addition of bromirno a:-d the absorption bsnd

at 6.10,p indicate so e degroo of unsaturation of the

poly-tertiary ctriLle resid-Lo. Colthup((0) lists the

stratchi-n fir. -L-ncy rinee of an ulnconjugated carbon-carbon

double bond fro.,i 6.06 to 6.25~,u

It was hoped that a molecular weight of the original

pol7-"uater?'nry ax-ioni:m bromide cold be ap--ro..irated

from a molecula-r ireiCht determi.ntion of the poly-ari.e.

Since oxtromo difficulties arc encountered in molecular

wei-ht detoirm-inatio-:s of polyelectrolytes, a molecular

weight of the poly-amine could be used to deduce a molecu-

lar .:?i 'ht of the: original poly-quaternary derivative by

assu.i:n- an overall loss of Cj2s r from the poly- 'uatornary

amonirmn bro:i-de. A determination of this type would prove

very useful, since a value of the constant in the Staudiaer







equatio-, relating viscosity end r olccul.r weil-At, could

be c.clculatd and furth-.1 molecular weight determinetions

of poly- uatern ry .;ro,'.uctc could be ap .roximated. liowevcr,

all attempts to detcrmi:ln the molecular weight of th:r poly-

rin. ~roved unsuccessful. ZThe boiling point el-:vnaion

and 2r-..ozin. poi:t depression methodIs wee retrIicted by

the insolubility of the product iand the t mpersature

differe-nti cl, 'eb folloin-: solvc-nt : ver. tri, ond

proved 'unsat is:fectory: water, eth;nol, t-bultcnol, eyclo-

heoanol, triothanol amine, diei!anol eITnin.., nonem th-lene

clyCcol, cr.:tpho:', npthal<-. b, benzene, ce-clohexene, -lacial
acetic acidc, nittrbonzene, 2,4,6-t.irib:olqo'oO.lili.O, bromo-

bcnzene, ci-d oth:l:l..-c dibroraide.


3. Polyme.-ization of diallyl amine hydrochloride
By polymorizin- diallyl amine hydrochloride and

treatl ~r the polymor i;ith LaOH solution, it was thought

tLat a poly-mmin:, similar to the poly-amine in the

previous section, could be obtained and studied.

Ten craas of diallyl amino hydrochloriJo (ro-

cr~7tallized from- acel-one-ethanol three tirns; o..p. 164-

50) :c.s dissolved in 5 nl, of water and treated -with

25 drops of 6C," t-I'vtyl hy-dropcro--ide solution (0.012 g,/

drop). (ir mi::tur rc.ral..cd in an open b,'l:oe- at 600 fcr

three days. Ten grams of product was obtained. The

product was insoluble in water and formed a gel. The







product was treated with ethanol but was insoluble. The

wash ethanol was treated with acetone and only a faint

cloudiness was obtained which indicated there was very

little unroacted diallyl amina hydrochloride present.

After soaking in an acetono-ethanol mixture or picking up

moisture from the atmosphere, the product exhibited elastic

properties. The product did not react with iNaOH solution

to give a poly-amine, as was hopod. However, the properties
of this polymer definitely indicate some degree of cross-

linking. This was the only monomer in this study with

two allyl groups to show characteristics of cross-linking

upon polymerization.


4. Polymerization of 1,4-bis(diethylallylammonium)
butone-2 dibromide

Since Butler and Goette41) have indicated the

butene-2 double bond in 1,4-bis(diethylallylamnionium)

butcno-2 dibromide did not enter into the polymerization,

oxidation of those butene-2 double bond was considered

as a point of attack for degradation studies. Although

the polymer was made and some of its characteristics

studied, further degradation studies were not carried out.

The polymer was prepared from lS.0 g. of 1,4-bis(diethyl-

allylarmaonium)butcno-2 dibromide and 45 drops of 60O

t-butyl hydropero.lde dissolved in 9 ml. of water. After

5 days at 600 in an open beaker, 18 g. of polymer was ob-








tainted. It was soluble in water and othanol. It was re-

crystallized from 100 ml, of ethanol by addi:; 200 ml. of

50,' acetone-dioxane solution. The rosidue was very vis-

cous; after washing, decanting and drying in a vacuum

desiccator, a hygroscopic product melting at 348-$54

(with considorable decomposition) was obtained.


5. Attempted polymeoizations of monomers with the double
bond farther removed than the allyl position

As was shown in the proposed mechanism involving

intramolecular-intermolecular polymorization, the allyl

groups offer an ideal situation for the formation of

strain-free six-membered rigrs. It was hoped that prepara-

tion of monomers containing the double bond farther re-

moved from the nitrogen center would eliminate the possi-

bility of forming a six-mo:libered ring structu-e. If the

intramolecular-intexmolecular type of growth was necessEry,

monomers of this type would require formation of r..ngs

larCor than six-ncmbered and, consequently, the probability

of their formation would bo greatly decreased. In most

cases, the attempts to prepare monomers of this type were

unsuccessful anid polym rization studies could not be per-

for-.d. In the following attempted reactions, it was diffi-

cult to determine whether polymerization had occurred.








a) Attempted polymnrization of diethyllallyl-

butene--yl ammonium bromide.--A nmiture of 1.15 G. of

diethylallylbutene-3-yl semmonium bromide, 3 drops of 60

t-butylhydroperoxide solution and 1.0 ml, of watsr w:as

reacted at 600 for three days. A brown solid was obtained

which was soluble in water,

b) Attempted polymerization of diethylbute:ie-3-yl

Dpnten-h-vl ammonium bromide.--A mixture of 0.5 g. of the

very hygroscopic sorn-solid diethylbut oi-3-yl penten-l.-yl

ammonium bromide, 5 drops of 60/ t-butylhydrop oroxide

solution, and 3.0 ml. of water was reacted at 600 for five

days. A dark viscous liquid was obtained; it was soluble

in water and ethanol.

c) Attn~mpted polvymerization of diallyltetraethyl-

decamethvlene diemmoniurm dibro.Tide.--A mixture of 3.0 g.

of diallyltotratethy1decamethylene diawimonium dibromide,

6 drops of 60% t-butyllTydroperoxide, and 1.5 ml, of water

was reacted at 60S for three days. A light yellow solid

was obtained which was soluble in water.

If polymerization occurred in the above cases, it

is assumed that the products were not cross-linked since

the products were water-soluble,

6. Attempted infrared studies

Since it was postulated that an alternating intro-

molecular-intermolocular chain reaction could causo a






piperidino ring structure throughout the poly.mr, an infra-

red study of related struct'-es was carried out, In this

work hctrocylic ring structures, aminos, amine salts,

quaternairy derivatives and many of the polymoeic deriva-

tives discussed in this section wore studied. IRany of

the con.pounds were obtained conmlcrrcially and the others

were propa2ed. Spoctia o2 the following compounds were

obtaLied; pipeiil.ie, 2-~1.;thylpiperidino, 2,3-dixrethyl-

pipe:iidic, 1,2-dpiiperidinoethlan, 1,2-dimorpholino-

et'haiio, diallylpiporazino, pyrrolidino, Ni-allylpyrrolidine,

diallyla.ni:e, tr-iallylaE1i.e, dimethylallylamine diethyl-

allylamL-. and aminc sclts and quaternary derivatives of

some of thIoze compounds.

ilow.-ver, all the attempts proved unsuccessful,

since a ri--r structure could neither be identified from

the s.ectra of the above compounds nor correlated to t'-e

spectra of the polrie.ric derivatives,


7. Solvo:t effect on polymerization

The possibility of solvent interaction during the

polymerization must be considered as a factor affecting

the nature of the polymer--. If a growing chain-fre. radical

could react with a solvent radical in preference to a mono-

mer molecule, the degree of polymerization would be reduced.

Thus, it mirht be possible to have a monomer molecule con-

taininr several unsaturated groups and yet obtain a polymer





L6
of a low degree of polymer-i"ation and sli-ht cross-linki:.c.

The following espe.1 ients were carried out vw-:ioii tho sol-

vent Pnd solvent conce traction.

a) A mixture of 3.0 ". of trinllylbutyl c1-monium
bromide, 3.0 ml. of water, and 6 drops of 6C-, t-bc.tjylydro-

pero::ide solution was reacted at 600 for 76 hours. A hard,

water-soluble solid uas obtain ed. The product was dissolved

in 10 ml. of water; 10 drops of initiator was added a~d the

mixture was reacted fcr an additional ieQk at 60. The

product w-s water-soluble*

b) Two grc.ms of triallylbutyl azr-.onium broilide,

4 drops of initiator, and 2.0 ml. of vater were reacted
at 1000 for 38 hours. The product was wator-solublo,
c) Two grcas of triallrlbutylammoniumi blo-nido,

2 drops of initiator, nnd 0.08 ml, of wate-:; wei.' reacted

at 1000 for 38 hours. The resulting Z]le, glossy p-oJ-uct
was socked in hot water, filtoeed, aLd dried. One and

seven-tenths gre:-s (085 yield) of water-insoluble product

was obtained. Tie procedure was that of Bunch, (42) who

reported a 52.5% yield.

d) Three grams of diallyldiethyl anmnonium bromide,

10 drops of initiator, and 0.3 ml. of water were reacted at

1000 for 38 ho; rs. The product was water-soluble.

e) Three grams of diallyldiethyl amnioniuiu bromide,

10 drops of initiator, and 50 drops of dimethyl formamide

were reacted at 1000 for 38 hours. The product was water-








soluble.


8. Initiator ef-fct on polymerization

The effect of other initiators, with rosnact to

t-butylh7droporo-dide, was determi.id in the following ex-

porinents. Va:iousa initiators were tried with monomers

containir:l one, two, and three double bonds to se? if

t-b-;.,tylh-dropero-:ide is unique in its property of yiIldii

non-cLoslinurd poly:-rs with monorn-rs containing, two allyl

rarouips*

One gram each of triethylallyl ammonium bronide,

diallyldiet'.hyl arz-oniun bromide and 1,4-bis (diethylall'rl-

aruonium)butono-2 dibromide were dissolved in 1.0 ml. of

water and reacted for one waek at 600 with 0.05 g. each

of t-butylh-;dro'pl:-oxide, di-t-b;:tylpero-:'de, and benzoyl-
n-=ro:ldo.

Tle lack of cross-linki :i was indicated in every

case by the wator-solubility of the products. Poly:-.o~i'a-

tion took place only in the rs -octive reactions of diallyl-

diethyl arnonlni bromnde and 1,4.-bis(diethyltallyl7anronium)

butene-2 dibro:-ido with t-b:;tylh:rdropero Aido. In the othir

cases, the strrtin nonomors were rocovere'd and identified.

In an attempt to ovrluate the initiating property

of 2,2'-azoisobutryonitrile, the azo initiator was reacted

with hoxaallylethylene diamr'onium dibroride, since it was

thou'-it that an easily isolated, insoluble product would






88
be obtaiiod if polyeri-zatioi occur-i. T.i mono.', dcis-

solvod in 10 ml. o2 dLC:thyl foi-._-;idJ was allo-.;oJ to react

for two W32.-:3 at 60 in an opjn vusjoi.


TAISL VII
AZ0 IIITIATED POLYM_ IZATIO: OF


ITT: AAL,'U' YL'TJ-iYLW::1; JIAnt"0iiI~4


Wei-b.t of Vei-t of Perccnt :it. of H1O
ionomer Initiator Initiator Insoluble nemaliks
Prod ct


5.00 g. 0,013 g.

5.00 g. o.025 -.


5,0o g. 0,050 c.

5*00 -. 0.10 g.


5.00 g.


5.o0 g.


0.1


0.5


1.0


2.0


0.25 g.


0.50 g.


0. 0


Soluble LigL-t viscous liq.
after twio weeks

Soluble Ligrht viscous liq.
after two wo ks

Soluble Viscous liq. after
two we -a s

0.75 r'. sav7 viscoua liq.
after 2 weeks

3.26 g. Gelatinous after
12-15 hr. Flex-
ible solid 2 wks.

4.37 g. Gelatinous after
6 hr. Grittle
solid at 2 wks.


The initiating effect of 2,2'-azoisobutr7onitrile

upon the polymerization of diallyldiothyl ammonium bronide

was doternmiied in a similar manner* Two rrars of diallyl-

diethyl tcmmoonium bromide (mI.p. 155*) dissolved in 10 ml. of

dimethylformamide was allowed to react with varyinr amounts





69

of azo iaitiator at 750 frr firtooCn dc;s. hle JirduLct was

water soluble lin voir cas.;; tohe Letcn.aiL' ~io: of the aUount

of pol-,i.. ization was rather difficult Unjreaacted monomer

U;0s Jo3.ratjad 2o." tile roLction prod.u.ct by extraction dth

hot aceto..o. Attumpta r1ccrystallizatioln iom acetooe and

Othanol rOi sultUdL in l ll lu-~iation.

TA LE VIII

AZO I:.ITIr.Tr-D POLYiLOr I:~ TIc OF DIALLYLDIETHYL A1~O:'IUJM



Wt. of Ut. of .-
rionoi:er Initiator ,ciursB


2.iX .*:. 0.02 -. 1.0 'atcr soluble; portion
insoluble in hot acetone
m2neted at 295-300*

2, .10 5.0 ..:rter soluble; nclted
at 305-7*

2.00 g. 0.20 g. 10.0 Water soluble; melted
ct 306-7o


9. "',Vect of o-::e'- in the pclT,' riza.tion

It was considred that c::cen could exiit so:-o

inhfi.Itirn factor in the polTicrizatio'. and coittri' -te to

the f.ct that sol'.ble ,:--d non-crozsli:nro pol -r.o s were

obtani. d i:ith rinolo.-;:-s cc:.tair.n-i tiro -llyl Crcups. The

follcwing c::.,crimernt were perfori:'cod to determine the

ef-ect of o:-:-'cn.






90
a) A mixture of 8.0 g. of diallyldiethyl anunonium

bromide (moip 155*) 4,0 ml. of water, and 33 drops of

60cv t-butylhydroperoxide solution (appro.dimately 0.012

g,/drop) was allowed to react for 76 hours at 60-659, The
reaction was carried out in a nitrogen atmosphere, nitrogen

was passed through a combustion tube filled with copper

turning, two wash solutions of alkaline pyrocallol, a

CaSOc dryiie tower, and into the reaction vocsel. The

system was flusled. with nitrogen before the reaction was

started. Eight grams of hydroscopic product was obtained

after dryiln for several days in a vacuum desiccator at

less than 1 mm. pressure, The product decomposed over a

wide rnnrm and nelt d with decomposition at 355-600 The

product was soluble in water but was relatively insoluble

in ethanol, forming some gel. The degree of solubility

of this product in ethanol differed from the solubility

of polydiallyldiethyl samonium bromide, Which was prepared

open to the atmosphere.

A determination to show any possible difference

in the analysis calculated to include complete catalyst

incorporation in the polymer, and the actual analysis of

the polymer, was made. Since the conditions prevented the

reaction of atmospheric oxygen, the only oxygen present

in the polymer would be contributed by the catalyst or

from solvent interaction by the OH radical.







Calculations


iMoles of catalyst = 33 drops x 0,6 x 0.012 g./drop x
Smole = 0.00264 mole
90 g.
1 mole = 034 mole
Molos of monomer = 8.0 g. x 23 0034 mole
234#2 g.


0.034 mole nonomer
C.00261: :-lole cat.'ilvst


= 12 mole moimomor
mole catalyst


(assuming com-
plete catalyst
incorporation
in polymer)


TA :L: I::

I..AL'-'.:. OF .POL.-DIJ LL'L.7JJ -_iriL A:I,...,T:JII 3- I1" 1


Monomer Polymer Actual Analysis
(theory) (Assumiii: com-
plete catalyst
incorporl. ationl) I II Avlg

c 51.31 51.30 51.2 51.5 51.35
H 8.56 8.63 8.38 8.26 8,32
i 5.98 5.82 5.46 5.53 5.50
;'Br 34.15 33.21 33.8 34.0 33.90
%0 o 000 1.04 Total 99.07

Total 100.00 100.00 Oxygen by difference 0.931

An infrared absorbtion band at 6.10M4 indicated
some degrco of unsaturation of the polymer; a measurement
of the doCroo of unsatbration was attempted. This method
consisted of catalytic hydrogenation at atmospheric pres-









sure and was carried out according to the modification of
Parkin. (3)

A weighed quantity (approximately 0*06 c.) of

platinum oxide and 20 ml, of distilled water was placed
into a hydrogenation flask equipped with a magnetic stirrer

and a rubber-tipped side arm. After the system was evacuated

and flushed with hydrogen three times, the magnetic stirrer

was started. After complete absorption of hydro-en by the

platinum oxide, the stirring was stopped and a weighed

sample, dissolved in 3.00 ml. of water, was introduced

through the rubber tip from a calibrated hypodermic syringe.

At this point, the pressure was equalized to compensate

for the addition of the 3.00 ml. sample. The stirring was

started and the reaction continued until further absorp-

tion of hydrogen had ceased. Since the stirring mechanism

caused some heat to be generated, the system was allowed to

core to equilibrium before measurements wo:-e made. The

following table gives a summary of the data,

Since all the attempts to recrystallize the pol-mer

sample proved unsuccessful, the crude sample was hydro:ena-

ted after careful drying. However, a later method proved

successful in extracting any monomer left in the crude

polymer sample and a value of pure polymer fraction was
determined, With this correction, an average value of 0.193

moles of hydrogen absorbed per mole of monomer unit charged









00 0 7 Io A 0
* r 0 0 0 0

00 0 00o0 0 0
0 0 0 0 0 0 0 0
000 000


* A 0 0 0 o o
0 Q O0 C\l C'- H XfE '>Q
N(O -^' O N NO H 00
S'001 AO l 00 00 8
C * *00 0 0 0 0
0 0 0 * * *
000 0 0
o o o' c o' o


H




H



S H




H


rt

'3



H













a

o
0


0 c
oN 0
o o o

0 0 0
0 0 0


S0 0




0 0

G A 4

a ,a ,o
o 0 0





Pi


0 0 0
o 0 0







o a 0


0 i A r co
M O O oN
0- 0 O; 0 0 0
0 00 * *
0 0 0


4)
*r4


4





SV

0 2
a I



H


t P
P -j
0





H - Tr H
to H 0
43 40, 3 g



030 0 4
eC a





0r 0
04O H E1
o ) *rc




E as rj o


0
o







E-


01.

0






0





0
0

C-

























*rs
C-l
0





























a
o)

'-




Full Text

PAGE 1

Synthetic and Polymerization Mechanism Studies of Unsaturated Quaternary Ammonium Salts By RUDOLPH J. ANGELO A DISSERTATION PRESENTED TO THE GRADUATE COUNQL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA June, 1955

PAGE 2

UNIVERSITY OF FLORIDA 3 1262 08552 3321

PAGE 3

>^The author xvishes to e:q>ress his sincerest appreciation to Dr. George B, Butler whose teaching and direction have greatly contributed to the completion of this work. The author is indebted to Dr« A. H, Gropp for his encouragement, onthusissia and criticisn throughout this cov.i^se of study. The advice and suggestions offered by thcj author's Supervisory CoimiLittee and by his student associates is gratefully acioiowl edged. The author especially wishes to thank Mr, G, D, Price for carefully proof-reading the manuscript, r4r, B, A, Parkin for the benefit of stimulating discussions, and ilrs, l-lary Joy Breton for typing the manuscript. The author is indebted to the Atomic Energy Commission for the support of this work. ii

PAGE 4

tablji op contents Pago ACKiJ0V/LEDGI4EIITS ii LIST OF TABES V LIST OP PIGUPuES vl Ctiaptor I. IITTRODUCTION 1 II. Pltt^PAR/iTION OP l,n TERTIAHY DIAMINES 3 A» Tetraallylethylene diamin© l\. B. TetraallyltririiGthylene diamine ij. C. xetraallylhexamethylone diamine 5 D, Totraallylheptamotiiyleno diamine 5 E, Tetraallyloctamothylcne diamine 10 P. Tetraallylnonaraathylene diamine 12 G, TetraallyldocarriethylGne diamine 111 III. PPJUPARATION OP l,lj. AIID 1,5 TSRTIAUY DIAI'-II..-SS 16 A. Unsiccecaful Attompts to Prepare l,Ij. and 1,5 Tertiary Diamines 17 B. Preparation of l,h, and 1,5 Tertiary Diamines 2!{. C. Discussion of Results and Proposed Mechanism 27 iii

PAGE 5

Chapter Pac® IV. PRSPAfJATIOlJ OP BIS-QUATEW;AKY AMI40HIUM SALTS 3I4. A. PrGparatlon of the llexaallyl Derivatives 35 B. Preparation of the Dimethyltetraallyl Derivatives lj.0 V. SUSPENSION PO-YMERIZATIOiT OP BIS-QUATEl^JARy SALTS kl A« General Discussion of Method I(.7 B« Experimental Results $2 VI, EVALiJlTION OP CAPACITY STUDIES 5I| A. Method of Analysis ^k B. Suraraary of Experiraental Results of Capacity Deterainc.tions $Q C. Discussion or Capacity Deterninations 62 VII. POLYI-ERIZATIOn lECHANISI-i STUDIES 6? A, Monomer Syntheses 68 B, Polyraerization Products and Results Jl^ C, Discussion of Results 96 VIII. SUI^S'lARy 10l|. BIBLIOGRAPIiy 107 BIOGRAPHICAL ITH-IS 111 iv

PAGE 6

LIST OF TABLES Table Pace I. Infrai^d Spectral Analysis 2i|. II. SiDTimary of Physical Properties of l,n Tertiary Diarainos 32 III. Properties of Hexaallyl Dlainmonitm ulbroralde Dorlvatlvos k$ rv. Properties of Dlmethyltetraallyl Dlaumonltan JibroiTildo Jerivativos U6 V» Polyraei'izatlon Results 53 VI, S-uimnary of Capacity Determinations 60 VII, Azo Initiated Polymerization of Hexaallylethylene Dlammoniura Dibromide 88 VIII. Azo Initiated Polyraerlzation of Diallyldiethyl Araraonium Bromide 89 DU Ano.lyEls of Poly-diallyldiethyl Ammonium a?oraide 91 X, Iiydror;enation of Poly-dlallyldiothyl Ammoniiita Bromido 93 XI. Comparison of Polymerizations Open to Atmosphere and under Nitror.en 9if XII, Approximation of the Decree of Polymerization 97

PAGE 7

LIST OP FIGURES Flgm-*e Peg© 1, rilquilibrltim Exchange Capacity 57 2, Ilexaallyl Derivatives 61^. 3« Dimethyltetraallyl Derivatives 6I|. vi

PAGE 8

I INTRODUCTION Previous work^ " In this laboratory has shown tloat insolublo resins of the poly-qiia ternary ananoniura type possess anion exchange properties, Fuoss and Catiiers^^) have made similar observations with the quaternized derivatives oi* polyvinylpyridine. With this property in mind, it was decided to prepare a series of xxnsaturated quaternary derivatives of diamines and sttidy the anion exchange characteristics of the resultant resins. The main objectives in this investigation were as follows: (1) To provide a successful synthetic route to the bis-quaternai-'y derivatives of unsaturated l,n tertiary diamines. (2) a. To study a different approach to the preparation of the 1,1; and 1,5 tertiary diamines and their corresponding quaternary derivatives, since some pi-eliminary worlc*-^' involving their preparation had proved unsuccessful, b. To propose a mechanism to explain the failure, and the various side reactions, of the previous attempts to prepare the 1,U and 1,5 tertiary diamines. (3) To correlate the effect, if any, of the X

PAGE 9

2 degree of croso-linklng and the swelling coefficient on the capacity and possible Ion-screening effect of the resins, (If.) To study the effect of variable distance betii?een quaternary ainmonltxra exchange centers on the properties of the resin, especially In regard to selectivity of Ions of various size and cliarge* (5) To study the mechanism of the polymerization of this type of unsaturated quaternary aiamonlum monomers, since they appear ratiier unique in their property of yielding a solvable and, consequently, noncrosslinked resin from a monomer possessing tvjo unsaturated groups. ^ '

PAGE 10

II PREPAHATIOK OF l,n TSHTIARY DIAMINES In attempting to prepcre l,n tertiary diamines, the reaction of the appropriate dihaloallcene or diCarylsulfonoxy)all«uie v;ith diallyl amino proved very successful erccept in the cases of the l,l\. and 1,5 derivatives. The successful method of obtaininc the l,l\. a:id 1,5 derivatives follow in Section III. The yields of the l,n tertiary diamines vere 7095^ in all cases except for the hexamethylene derivative wliich was obtained in 21.7^ yield* All temperatures are listed as degrees centigrade and are uncorrected. The carbon, hydrogen and nitrogen analyses were carried out by Peninsular Chemresearch, Inc. and Clark Microanalytlcal laboratory. Tlie broriide analyses vjore determined by the Volhard metliod as given by Pierce and Haenish. ^^^ The following section deals i;ith the preparation end purification of the l,n tertiary diamines except for the lfl\. and 1,5 derivatives*

PAGE 11

A, Tetraallvleth?/"lene diamine (CHe=CHCHa ) aN(CIis ) sN(CiIsCH=CHa ) g This product was prepared in a manner simllfir to (9) that reported by Angelo. " Three moles (291 g, ) of dlallyl amine was added to a paste of 168 g« (2.C 11) of NallCOa and 100 ml« of water. The mixture was heated to a gentle reflux and I87.8 c. (1,0 li) of ethylene dlbromide was added slov/ly, Refl\ix conditions were maintained for seven and one-hslf hours after the addition. The mixture was cooled, filtered and the sol^ition treated with an excess of ij-O^ i\!aOH soliition until a distinct separation occur-red. The oil was separated, dried over solid KaOII and distilled at reduced pressure. One hundred and fiftyseven grains ijl^l^o yield) of clear liquid boili;ac at 72-3 V & 8 0.3 mm, and iiavin?^ n^^ l,l!-705 vjas obtained. The compound was previously reported as boilirif^ at 79Vo«5 ^i* and liaving nj^l.li.702. B, Tetraallyltrimethylene diemine i Ciia~0.iivJi.ig / a^' v^-iia / giJ ( i-.JiiaCii=^Ci.ig ) a The product used in this work was that pi-epared by Angelo; trie procedur-e v;as tliat of Laakso and iieynolds^ ' involvinr the reaction of diallyl amino with l,3-di(p-tol^lene sulfonoxy) propane. The yield was 80,5?o> boiling point 88-9Vo«5 JTin. J n^^l,i|711. Analysis calculated for CjeHae^iaJ H, 11,95^. Pound: K, 11,57^,

PAGE 12

C, Tetraallyliioxeuiethylene diamine (CH8«CHCKa ) aincila ) oN(CHBCH=CIIa ) a The product used In this v/ork was that prepared by Angelo; the procedure involved the reaction of allyl bromide vdth hexamethylene diamine mixed with a paste of NaaCOg and water. The yield was 21.7a^j boiling point 12k°/0,Q mm.; nj 1,1^701. Analysis calculated for CjeHaaNa: U, 10.13;^. Found: U, 10.21/^. D» Tetraallylheptaraethylene diamine (CHa-CIICIia ) sHCCHs ) 7K(CHsCH=CHa )« 1. 1,5-dibromoperitaj.ie Four hiindrod and seventytxw grams of concentrated H2SO4 was added with cooling to I6IO ~, (9.57 l'^) of i^.3^ IIBi' solution. To this mixture was added 13? g. (l,i^91^ i^O of tetraiiydrop^Tant The mijcture was allowed to refliox gently for tliroo and a half hours, was cooled aid separated. The bottom layer was filtered, ii?ashed with water and dried over KgCOg. Tlio dark liquid was distilled and 237 g. (72.8?^ yield) of l,5~dlbroraop9ntane boiling at 83-90 V^-7 i'ara. was obtained, 2, Pimelic nitrile Two hundred and twenty grams (ii.,5 M) of NaCN was added to 3OC1 ml, of water and the mixture heated gently tmtil the NaCW dissolved, Fovir hundred and six grams

PAGE 13

6 (1,76 M) of 1,5-dibromopentane dissolved in 1 liter of 95^ ethyl alcohol was added carefully to this hot solution until a vigorous reaction occurred* The remainder of the bromide solution was added as fast as the reaction vjould permit. The mi^xture was allowed to reflux for Torty-two hotirs after the additioxi. The ethyl alcohol was distilled at reduced pressure and the residue cooled and filtered. The filtrate was extracted twice v/ith ethyl acetate {300 ml, portion end 100 ml. portion), Tlie ethyl acetate was removed from the extract by distillation and the residue was distilled at reduced pressure. One hundred and fiftyseven grams (73»^/^ yield) of product boiling at 153-6*/ (12) 3-3«5 ^^'C'l* was obtained, Plmelic nitrile is reported as boiling at 175-6°/!^ ™i» 3* Esters of pimelic acid a) Diethyl plmelate Wfs prepared in a manner (13) similar to the method given by Adnms and Marvel. Five himdred f^reras of concentrated H3SO4 was added, vrith coolin,r and stirring, to ^0^ F» of 95% ethyl alcohol. One hundred and fifty-seven grarrs (1.29 M) of pimelic nitrile was added and the mixture was allowed to reflux for ten hours. The mixture was then cooled, poured Into an equal volume of ice water, separated and the top layer dried over CaClg. The liquid was distilled at reduced pressure through a Claisen apparatus and 202 g. (72. 7;^ yield)

PAGE 14

of diethyl plraolate bolllnfat lU9-5irVl9-20 mm. vms obtained. Diethyl plmelato ^ is reported as bolliiig at li|6-52V22 ram. b) Dibutyl pimelate was obtained in a 90^ yield by the foliov/ing general procedure J Two moles of piraelic acid v;a£ added to eight moles of n-butanol in on equal volume of toluene and 2 g, of concentrated liaSO^ was added as a cetal^rst. The solution was refl^lxed with stirring while the water of reaction was removed as the azeotrope and collected in a iirater separator. Tiie acid was neutralized with an ejccess of BaCOg while the solution was still wanm. After filtration, the toluene and butanol were removed by distillation. The dibutyl pimelate was riiially obtairiei by distillation at reduced pressure. i|., Heptamethylene glycol a) The general procedure -^' as given in "Organic Reactions," Vol. VI for the reduction of estei's using LIAIH4 was followed, Illneteen and five-tenths grama (0,513 M; a 1|Q^ excess) of LIAIH* was added to 500 ml. of dry ether in a five liter flask. The slurry was cooled by an icesalt bath; stirring was maintained with an induct iontypo motor; all possible precautions to exclude moisture were taken, A solution of 100 g. (0,368 I-I) of di-n-butyl plmelato and 2^0 ml. of dry ether was added

PAGE 15

6 very slowly to the well-stirred sliirry of hydride. After adding for one hour, the stirring became difficult and 500 nil# of dry ether was added to the reaction mixture. The temperature was maintained at -3° to 0** during the addition period of two hours. Two hundred milliliters of ether-ethyl alcohol solution (50^ by volume) was added slowly to destroy the excess hydride. After all the excess hydride had been destroyed, 1100 ml. of 10^ H2SO4 solution was added until two clear layers were obtained. The layers were separated; the ether layer scvod and the water layer extracted fourtines with 100 ml. portions of ethyl acetate. The combined ether layer and extracts wore dried over Drierite. Tiie solvent was rsaovod and finally i|0 g, (82.5^ yield) of heptamethylaie glycol boiling at li|01U5V8-10 can. vms obtained. Heptamethylene glycol '^' is reported as boilinr at Ik^-kS^/lO ifca* b) The previous experiment was repeated using 20.0 g, (0.525 M) of LiAlH4 and 100 g. {0,368 M) of din-butyl pimelate. Forty-five and one-tenth grams (92,7^ yield) of hoptamethylene glycol boiling at lk$''l\Q°/Q mm, was obtained, 5» lieptamethylone dibromide The general procedure^ '' given in "Organic Syntheses," Coll. Vol. II, for the preparation of bromides from alcohols was enployed. Sixty-six grams (0,5 M) of

PAGE 16

9 heptamethylene glycol was placed into a flask equipped with a gas inlet tube, magnetic stirrer and condenser leading to a tared trap of NaOH solution. Dry HBr gas from a cylinder was Introduced into the flask while the temperature vxas maintair-ed at 100-20°. The addition of HBr took three and one-half hotirc, at which time the trap containing NaOH solution began to Increase in weight rep idly. The layers were separated and the bromide laj&r was treated with onethird its volume of concentrated 112304, Water was added to this emulsion imtil a distinct separation occiirred. The mixture was separated and washed with water-methejiol solution ($0^ by volurae) until the product vm.s neutral to litmus. The product was dried over CaClg, distilled at reduced pressure and 65*0 g, (50.3^ yield) of clear liquid boiling at 125-26°/8 mm. was obtained, MuJLlor and Vane report the boiliiog point of heptamethylene dibromide as 123 Vn ^m, 6, Tetraallylheptamethylene diamine To a slurry of 21.l2 g. (2.5 M) of diallyl amine, 126 g. (1,50 M) of ImHCOg and 75 ml. of water, heated to a gentle reflujx and stirred vigorously, 65,0 g, (0,252 M) of heptamethylene dibromide was added slowly for one and one-Iialf hours. The mixture was cooled, treated with an excess of kC^ KaOH solution, separated and the amine layer dried over KaOH pellets. The excess diallyl amine was

PAGE 17

10 distilled and the residue v;ac distilled at reduced pressiire throufjh a Clalson apparatus. Sixty-four gi^oEis {Qj,'j% yield) of light yelloij liquid boiling at 12U-26Vo«l ^^» an^ having an n-. 1,[;719 was obtained. Analysis cslculatod for C.eHa^Ns: C, 7Q,6%; H, 11. 8^ j IJ, 9.65^. Pound: C, 76.5^; H, 11.5Q^J N, 9.51/^. E» Tetraallyloctai-gethylene dioriine (CHa=CHCH8 ) sK ( CHa ) qN (CH8CII=CHa ) s 1. Suberic nitrile This preparation was carried out exactly as described in the previous preparation of piiaolic nitrile. One hundred and ninety-seven grains (0,82 M) of hoxanethylene dibroraide dissolved in 500 ml, of 9^% ethyl alcohol was added to 100 g. (2.05 K) of ilaCII dissolved in I50 ml, of water. After the hexaraethylene dibromido was treated in the same nannor as the previous nitrile preparation, the mixture yielded 96 g, (87,1^ yield) of clear liquid boiling at 175-72°/lO mm, Doutsch and v. Braunn^"'-^' report the boiling point of suberic nitrile as I76-8VII ra^i. 2, Diethyl suberate This preparation was carried out exactly like the previous preparation of diethyl pimelate, Hlnety-slx grams (0,705 M) of suberic nitrile was allowed to reflux fifteen hours with a mixture of UOO g, of concentrated

PAGE 18

11 HaSO^ and [|.00 ml. of 9$% ethyl alcohol. ATter the suberic nltrlle was treated in the same manner as the previous nitrilo hydi-'olysls, the mlxtijre yiolded 119 g. (73«5^ yield) of clear liquid boiling at l[;0-i{.Ii.*'/8 rnn, and havinf^ an S2 I (PO) njQ l,i.!32i^.. Karvonen**^ ' reports the following constants for diethyl suberate: boilliig point, 13G-Q°/d mm. and s< D n^°1,1:3278 . 3. Octanethylene glycol The reduction of diethyl suberate was carried out in the saj";ie mannei' as the previous reduction of di-n-butyl pimelate. One hundred and nineteen grains (c.,517 H) of diethyl suberate dissolved in 2^0 ml. of dry ether was carefully added to a cooled sliirry of 25*0 g. (0,66 H) of LlAlli^ in 500 ml. of dry ethei", Trie reaction mixture yielded 73. > g. (97 'k^ yield) of a clear liquid boiling at I58-6OV9-IO mm, Ihe boiliiig point of octamethylene glycol is reported ^^'' as l60-62°/9«5 mm. l\.0 Octamethylene dibromid© Octamethylene dibromide was prepared in the same manner as the previous preparation of hoptamethylene dibromide. Seventy-three cuid five-tenths grams (0,503 M) of octamethylene glycol was treated v;ith gaseous Iii3r at 100-20° for two hours. After being treated in the same manner as the previous bromination, 86.5 g» ic>3»2% yield)

PAGE 19

12 of clear liquid boiling at 170-lV33-3U ran, was obtained, Dlonneau'^^' reports the boilinr; point of octamethyleno dibromido as 173°/35 rm* 5, Totraallyloctamethylone diamine To a slurry of 3 08 g, (3,18 M) of dlallyl amine, 168 g, (2,0 M) of I-IaHCOs and 120 ml, of water, heated to a gentle reflux and stirred vigoroxisly, 86,5 g» (0,313 H) of octamo thy lone dibromide v/as added slowly for forty minutes. After the addition, tlie mixture was allowed to rofliLJc for an additional eight and one-half hours. The mixture was then coolad, treated with an excess of h.CJ(^ NaOH solution, separated and the amine la7^er dried over NaOH pellets. The excess diallyl amino was distilled and the res5.due was distilled at reduced pressure tiirough a Claisen appai'atus. The product frothed very badly; glass wool and paraffin were added to roliovo tho foaming. Eightytiiree and five tenths gre:is of light yellow liquid boiling at 133-UO°/0»l iiroi, was obtained, Tb.is liqiiid was fractionated aiid 73.0 g, (75.5/^ yield) of light yellow liquid boili:\g at 13U-8Vo» 09-0,1 ram, and having an n^^l,iL65l was obtained. Analysis calculated for CeoHgeWsS C, 78,9^; H, 11,82,^; H, 9.2^, Pound: C, 78.8l?^j li, 11,75/^5 I-, 9.l5f^. P. TetraallTlnonanothyleno diamine ( CKa-CHCxis. ) 8 K { Ciig ) g II( CKg ^H=CiIa ) «

PAGE 20

13 1, i^'roBi tne disulfonate The preparation of l,9-di(p-toluenesulf onox7)nonano was accompli sned in accordance with the procedure of Laakso and iJeynolds. ^^^^ Eighty-six and two-tenths grams (0,5U H) Ox nonamethylsne glycol aissolved in 10. ml, of dry pyridine was treated vjith 206 g. (l,08 M) of p-toluene su.lfonyl chloride dissolved in 280 ml, of dry pyridine. Pollo;d.ng the method cited, 21? n« (65.5^ yield) of pioduct was obtained. Melting poiat after third rocrystallization from aosolute alcohol v;as 75-7°. Analysis calculated for CasiioaSsOo: C, 5ci*95^; H, b,b6%. Found: C, 5<3.
PAGE 21

hoTir period. The mixture was allovred to reflux an additional seven and one-half houi-s. The raixtioro was then cooled, treated v;ith an excess of 1|P^ NaOH solution, separated and the amine layer dried over NaOH pellets. The e;:coss diallyl amine was distilled and the residue vjas distilled at redi^ced pressure. Fifty-two grams {9ii-»5/^ yield) of li.r^ht yellow liquid boiling at li|6-8Vo. 06-0,0? ram. and having an n^®l,ij.723 was obtained. Analysis calculated for CaiHsQlJs: C, 79.3/^5 H, 11,93/^. Pound: C, 79.88^j H, 12,01^, G. TotraallYldecaj.ioth7lene disnine (CH8=CiICHs)ax^(CHs),oN(GliaGiI=Ciig)2 1, Prom tae disulfonate The preparation of 1,10-di (benzenesulf onoxy) decane was accomplished in accordance with trie proced'ore of Laakso and Reynolds. ' One himdred and one grams (0,58 M) of docamethylene glycol (n*-p. 70-1°) dissolved in 200 ml, of dry pyridine was treated with 205 g. (1,16 II) of benzene siilfonyl chloride, Pollovri.ng the method cited, 226 g, (85»6^ yield) of pi-oduct was obtained. Melting point after third recrystallization from absolute alcohol was ij.6-8°, Analysis calculated for CaaHgoSsOe: C, 58»li4^» H, 6,66^, Found: o, 57,67^; H, 6,79/^. One himdred and thirty-six grai.is (0,3 M) of l,10-di(ben2enesulfonoxy) decane was refluxed for forty-

PAGE 22

15 eight houi-s with 50O g, i^^lS W) oi" diallyl amine in accor(10) dance v;ith the procediire of Laakao and Reynolds, The reaction mixture yielded 62 g. (62»2/j' yield) of a dark liquid bollinr at li^5-6oVo»5 J^« This \ias fractionated and 72 g. of a lipht yellow liquid boiling at 15^^-5°/ 0,1-0,2 ran. and having an n^ 1,14.701 was obtained. Analysis calculated for CsglUoWs: H, Q*h3%* Pound JT, 8.27^2, Prom the dibromido To a slurry of 162 c* (1»66 II) of diallyl amine, dk g, (1,0 H) of NaliCOa and $0 nl. of x;atar, heated to a gentle rcfl^^x and stirred vicorously, 50 S« (0,166 II) of decaraethylene dibromide was added over a period of one and throe-fourths hours. The mixture was refluxed for an additional five hours, cooled, treated with an excess of kO^ IJaOli solution, separated and dried over NaOII pellets. The excess diallyl ariine was talren off and the residue distilled at reduced pressiu^e. Thirty-nine grams (70»5/^ yield) of a lirht, yellow liquid boilinr at 1^2-6 Vo«07 inra, and having an n®*l,li725 was obtained.

PAGE 23

Ill PREPARATION OF l,lj. AND 1,5 TERTURY DIAI-lIiJES Attempts to prepare tetraallyltetramethylcne diamine and tetraallylpontamethylene diamine by the previously successful reactions of the l,h and the 1,5 dihaloor di-(aryl3ulfonoxy)allcanes vjith diallyl ariine (23) were iinsuccessfiil. In the reaction xd.th the dihaloalkanes, the products were diallyl pyrrolidinium broraide and diallyl piperidiniura bromide. In the reaction vjith the di-(arylsulfonoxy)alkanes in the presence of a large excess of diallyl amine, the products were allyl pyrrolidine and triallyl amino fro:n the l,kderivatives, and allyl piperidlne and triallyl amine from the 1,5 derivatives. Tne products of the reaction can be explained on tiJO basis of an intraraoleciilar cyclization to the five and six membered cyclic quaternary ammonium salt, followed by an allylatlon of the excess diallyl amine by the quaternary amraoniiim salt to prodViCe triallyl amine and the appropriate allyl substituted hetrocyclic amine. This section deals Kith the isolation and identification of the products of the above reactions, an explanation of the course of the reaction and, finally, the successful method of preparation of the desired 1,[|. and 1,5 tertiary diamines, 16

PAGE 24

17 A. Uhsuccessfvi], Attempts to Prepare l.lt end l.lj Tertiary Dlaiaii^ioai' 1, Reaction of l,Ii.-dibroraobutane xvlth diallylamine a) A mixture of iUi-.5 g» (0,206 M) of 1,1; dibronobutane, 60 g, (0.618 M) of dlallyl amino, 3lj.,6 g» of KaHCOg, and 20 ml. of iirater was rofluxed for seven hours. After filtration and addition of IJaOH to the filtrate, the orcanic layerwas distilled; however, no material v;as obtained boiling in the expected ranre for tetraallyltetramethylene diamine. Several other exporiiiients vxere perfonned varying the ratio of dihalide to amine, che.ncing solvents and changing the base, but in all cases none of the desired product was isolated. b) To four milliliters of l,Ii.-dibromo butane was added twelve railliliter-s of diallyl amine at a temperature of 35** • 'Ahile the lirjuid was beinr; stirred, the temporature rose to 100°, /ifter standing overnifrht, the crystalline product was collected, washed with acetone and recrystallized from hot acetone end a smell amount of absolute alcohol. The product started to decompose at 205** and melted completely at 315-22** with decomposition. A bromide analysis of the product showed 3k»n% bromine.

PAGE 25

18 The calculated Br percentage for the dlhydrobroraide, H H (CHa=CIICIIa)aN(CH8) 41^(0118011=0118)8, Is 3Q.95/^^ The calculated + + 3rBr" Br percentage for dlallyl pyrrolldiniioia bromide, ^s^^iIaOH=:CHa ^CiisCH=CHs, is Jh*h^%» Since diallyl pyrrolidinium bromide has not been reported in the literature, a method Tor its preparation follows: N-allyl pyrrolidine was prepared by treating pyrrolidine v;ith allyl bromide in the presence of IJailCOa, Tiie boiling point vras 129-30° and n^°l.Iii;.86. Analysis calculated for C^H.aN; C, 75.6Ji; ii, 11. 71^^; ^, 12,61/^. Pound: C, 75.3^; H, 11,87^; N, 12, lU/^. To a cold mixtiire of 21,0 g. (0,19 M) of K-allyl pyrrolidine and 75 ml. of acetone vjas carefully added 25.0 c. (0.207 M) of allyl bromide with cooling, Tlie hygroscopic salt which precipitated was waslied Xirith acetone and dried in a vacui;:n desiccator. Thirty-nine graias (90.5/^ yield) of product vjas obtained. After recrystallization from hot acetone v;ith a small amomit of absolute alcohol, the product started to decompose at 202** and melted coirrpletely with decomposition at 315-20**. A mixed melting point with the reaction product from above

PAGE 26

19 shox/ed dQconposition startlnf^ at I98®; the mlxtvre melted completely with decomposition at 31i|-22°. 2. Reaction of 1,5-dlbromopentane vdth dlallyl amine A mlxtiire of l\. ml, of 1,5-dlbroraopontane and 12 ml. of diallyl amine was allov/ed to stand overnight, Tho hygroscopic crystals v^ore collected, washed with acetone and dried in a vacuum desiccator. After rocrystallization from hot acetone with a small amount of absolute alcoiiol, the product gave a flash melting point of 19l.|-6°, Upon slow continuous heating the prodvict started to decompose at 209*' and melt ed completely at 297-302* with decomposition. A bromide analysis of the product showed 32.8I1/J bromine, 'fhe calculated Br percentage for the dihydrobr oroide , H H (CHa=CHCHs)8N(CH2)6N(CHaCn=CilE)a, is 37.66^. + + BrBr* The calculated Hp content for diallyl piperidlnium bromide. ar" y \;il2CII=Ciig, is 32.1]^. Diallyl piper idinium bromide is reported ^-^^ to have a flasli melting point of 190**, 3, Reaction of l,[i.-di(p-toluenesulfonoxy) butane with diallylai.iine After a nixciber of unsuccessful attempts to obtain totraallyltotrariethylene diamine by the method of Laakso

PAGE 27

20 and Reynolds, the following modification vjas adopted: Two hundred and twenty-eight grams (0.577 M) of l,I|-di(ptoluenesulfonoxy) butane and 632 g. (6,5 M) of diallyl amine were rofluxed for seventy-tv.'o hoiirs. The mixture v/as allowed to cool and the layers separated. The lower layer was treated with an excess of i+Q^ NaOH solution, allowed to stand overnight and filtered. Pron the filtrate, 3 g, of a yellow liquid boiling at 100-lloVo»3-0«5 inn. and having an n^ l,[j.721 was obtained. T le original upper layer was distilled and 18 g, of diallyl amine was recoverod. The residue was distilled at redi«;ed pressure and 11,5 g. of a light yellow liquid boiling at 100-110® at 0, 3-0,5 irm, and having an nj*1,I|.720 vras obtained. A nitrc^en analysis showed 9«38/^« The calculated nitrogen percentage for tetraallyltotratnethyleno diamiiie (CjeHssNg) is 11.27. These fractions appear to be impure tetraallyltetramethylene diamine because this compound was prepared by another method and found to have a boiling point of 96-8°/0«l Sim. and an n^^l,l!.716. All attempts to pi.n>ify the fractions failed; the attec^ts to prepare derivatives for identification pui-poses also proved unsuccessful. Considering the total of ll;..5 g, of crude product as tetraallyltetramethylene diamine, the yield would be IQJw,

PAGE 28

21 However, during "the distillation of the above residue at reduced preasuj^e, a large amount of material v/as collected in the cold trap# Tills material was fractionated and a small amount of diallyl amino was recovered. This was followed by 20 g. boiling in the ra2TgQ 111-1Il7°j and 21 g. boiling at l/.t.7-l|.9*' and having an nj"l,l^)|89, V/ith allyl bromide, this latter fraction (11|7-!j.9**) gave tetraallyl ammonium bromide molting at l83-U**« ^ mixed melting point Xirith an authentic sample of tetraallyl ammonium bromide v;es 182-83°, The reported constants for trlallyl amine are: i3oiling point lli.o-9°» nj*1,1^502, ITi© intermediate fraction (111-I;7**) should contain allyl pyrrolidine (b.p, 129-30°) if the previously cited disproportionation meclianism is corrects However, allyl p^rrrolidino was not isolated from the fraction at that time. k. Reaction of diallyl pyrrolidinium bromide with diallylamine In order to support the above proposed course for the reaction, 36,0 g, (0.155 M) of diallyl pyrrolidinium bromide and 120,0 g, (l,2l!H) of diallyl amine were reflu2:ed for thirty-six hours. During this time, the liquid temperature remained at 109°, the boiling point of diallyl amine. After thir. time, the salt became viscous, the color changed to a light red and the temperature rose to 120°.

PAGE 29

22 Refluxinc was contlnuod for an additional twelve hours. The mi:ctxir'e was cooled, filtered and the dark red solution fractionated to yield: Diallyl amine, 68 g« ; boiling point, 110-13°i n^**1.^1397; allyl pj-rrolidine, 7 L'. J boiling point, 130-iiO*'; ng^l^kkHi and triallyl amine, U.5 g.i boiling point, ll|.6-9''j ng^1,101.99. Treatment of the allyl pyrrolidine fraction with allyl bronide gave the quaternar: axnmoniimi salt ra el ting at 313-20** with decomposition. Mixed molting point with an authentic sample of diallyl pyrrolidinium bromide was 3114.-21* with decomposition. Treatment of the triallyl amine fraction with allyl bromide gave the quaternary amraoniiim salt melting at iQU-"^^* Mixed melting point with an authentic scmple of tetraallyl ammoninm bromide v:as 181^.-5**. 5. Reaction of l,5-di(benzenesulfonoxy)pGntane with diall^rl amine As in the case of the lfl\. derivative, a member of imsuccessful attempts to prepare tetraallylpentaincthyleno diamine by the method of Laakso and Reynolds v/oro made and the follov;ing modification was adopted. Two hundred and sevonty-nino gratis (0,727 M) of l,5-di(benzenesulf onoxy)pontane and 77^ g. (8»0 M) of diallyl amine were rofluxed v;ith stirring for forty-eight hours. The product was s^arated as described above, and $01 g. of excess diallyl

PAGE 30

23 amine reraovod by distillation. Following the diallyl amine, U5 S« of a colorless liquid boiling at l[i7-9*' wac obtained, and l}.»0 g» of a dark yellov: liquid boiling at 60-100* at 0.05 EETi. After refractionation of the fraction boiling at ll!.7-h9°, a prod-act boilin<: at li>.8-50*» and having an nj°l,li539 was obtained. On the basis of the above proposed mechanism for the reaction, this product should be a mixture of triallyl amino (b,p, ll^S50°, n^^1,]|502) aiid allyl piporidino (b,p, llt.8-50**, ng^1.H577). Since it v;as impossible to separate these compounds by distillation, an infrared analysis was made using a Poi^kin Elmer double-beam infrared spectrop-iotometer with a cell thickness of 0»025 mm. Various mixtures of pure sample E of triallyl amine and allyl piper idlne were made in an attempt to match the index of refraction of the reaction mixture, A mixtirre containiiig ^1% by weight of triallyl amine had a ^°1«U533 compared to the nj°l.ii539 of the reaction mixture. The infrared spectral analysis gave the following results:

PAGE 31

2k TABLE I IHPRAHED SPEGTiUL AiiALYSIS Sample n^*' Absorption band (iTilci-ons) ''tSne^'^ 1^^502 3*38 9.32 Allyl piperldine Uk$n 7#23 7.70 8.33 Authentic Mixture (51?^ by wt, 1.U533 3.30 sii 9.35 7.22 7.7O 3.33 triallyl aiTilne ) Reaction Product 1.1-539 3*3^ sh 9-35 7.22 7.70 C.33 Mixture These results show conclvisively the presence of both triallyl amine and allyl piperidine in the reaction product mixture, and the refractive indices indicate appro:-dLinat el y equal molar quantities of the two compoimds. B. Preparation of l.k and l.g Tertiar-r Pianinos 1» Totraallyltetramethylene diamine ( CHa=CHCKa ) a^ ( CHg ) 4N ( CiiaCIi=Ciia ) a a) Preparation of N.lj.N'»Iv'« — totraall?! sue cinamide .— To a well-stirred solution of 211 r> (2.168 M) of diallyl amine in 200 ml. of dry benzene v/as added a solution of 8I1. e. (0,51]2 M) of succinyl chloride (b.p. S7-9V18 i^. )

PAGE 32

25 in 100 ral, of dry benzene v^llo the tamporati-ire was maintained at 10-20*', Stirring was continued at this teiTiporat\ire for one hoiir after addition v/as completed. The cclution was filtered and the diallyl amino hydrochloride crystals were washed with dry benzene. The washings were added to the orir;-inal filtrate and the benzene removed. The residue was distilled at reduced pressni'^e to obtain 100 e. (60,6^ yield) of liquid boiling at 160-2V0-1 ^nm, , having an n^, « 1»5925 end d4 1,0060, The infrared spectriiPi showed absorption bands for aiTiide carbonyl and carbon-carbon double bond. Analysis calculated for C.eHs^OgNs: c, 69.5^i H, 8.88,^; IJ, 10,3X Po^Jind: C, 69.0^i H, 9.11|^; N, 9.7?^. H% calculetod; 81. 86, 14% found: 81,15, b) Keduction of H.Ha^' J'H— tetraallyl six elngmide.— To 16 g, (O.kS M) of LiAlH^ as a sltirry in 300 ml. of dry ether, was added during one and threefourths houj's, 62 g. (0,225 H) of tetraallyl sr.ccinapiide dissolved in 200 ml. of dry ether, A threo-neck five liter flask equipped with ice water condenser, thomomoter, droppinn funnol, sealed stirrer, and cooled by an ice-salt bath was employed. The temperature of the reaction mi.^itxye was maintained below 15° during the addition. After reflvucing fifty hoirrs, a mixture of ethanol and water was added to destroy the e-xcess hydride. The mixture was then treated

PAGE 33

26 with 800 ml, of 10^ NaOH solution and solid NaOII pellets until a distinct separation occxirred. The ethor layer was separated and the water layer was extracted twice with 200 ml. portions of ether. The combined ether layers vjore dried over solid NaOIi, and the ether removed, Tlae residue was distilled at reduced pressure to give 27»5 S« (k-9»3% yield) of clear liquid boiling at 96-Q^/0,l mm, and having go an Hp 1,Ll716. The infrared spectrum of the compound showed the absence of the carbonyl bond and the presence of the carbon-carbon double bond absorption. Analysis calculated for C.elleelJel C, 77*kf^; H, 11.3/^5 K, 11.2,^', Pound: C, 77.5/^; H, 11,3/"^; N, 10,7/^. The dihydrobroxaide was prepared by treatment of a dry benzene solution of the above amine with dry HBr gas. The compound melted at 107-8°, Analysis calculated for C,6HaoNgBrs: Br, 38,95^, Pound: Br, 38,62^, 2. Tetraallylpentamethyleiie diamine (CHa=CIICH3)aK(CnB)6N{CH8CH=CHg)a a) Preparation of li.II.II'.:!' — tetraallyl nlutar araido ,-'Usin/t 6l^ g, (l,6i|. M) of diallyl amine and 72 g, (O.I|25 M) of glutaryl chloride (b,p, 10G°/lO mm,), and follov;ing the procedure above for the preparation of tetraallyl succinamide, II3 g, (91»7^ yield) of tetraallyl glutaramide was obtained boiling at 165-7 V^* 03 mm, and having an ng^I.502O, Analysis calculated for CjyHssOsNa:

PAGE 34

27 C, 70«3^; H, 8.86;^; N, 9.65J^. Found: C, 69.9/^; H, 8,86;^} N, 9»26^. Tlie infrared spectrum showed absorption for onide carbonyl and carbon-carbon double bond. b) Reduction of N«N.N* .rJ«— 'tetraallyl p^lutar e^iide.— Usinc 21 g. (0,^53 14) of LiAlII* and 0? g. (O.3O M) of tetraallyl glutai-amide, and following the procedure described above for the reduction of tetraallyl succinamide, except tliat this solution was allov;ed to reflu:>c for four hours instead of fifty hours, 53 g» (63,I{^ yield) of a clear liquid was obtained. This product boiled at 96-7°/ 0,02 isra, and had an n^'^l,Ii.7l.!.3. The Infrared spectrum showed the absence of the carbonyl bond and the presence of a strong terminal carbon-carbon double bond. Analysis calculated for C^^Hqo^^h'' C, 77.8/^* H, ll,Ij.6/i;j li, 10.68?5, Pound: C, 77.3/5J H, ll,2l\fo; If, 10,33)i. Tlio dihydrobromide was prepared by treatment of a dry beraene solution of the above amine with dry IIBr gas. The compoiind melted at ll^7°. /inalysis calculated for CiyllssligBrs: Br, 37,66/^, Found: Br, 37,53>1, C» Discussion of Results and Proposed Mechanism In the reaction of 1,U and 1,5 dihaloalkanos tjith diallyl amine, it appears likely ths.t the first attack wo-ald result in the formation of the l\. or 5"t>romoalkyl diallyl amino hydi'obromide. In the presence of an excess of diallyl amine, the free bromoalkyl diallyl amino should

PAGE 35

28 be liberated, since it should bo tho weaker base. The second step probably involves the Tormation of a cyclic quaternary (2b) ammonium salt. It has been pointed out that aliphatic bases containing amine and halogen in the l,l\. position exist only as salts and on liberation of the bases condense (25) to pyrrole derivatives. It lias also been shovm that anines of the type x(CHa)^IIRe yield (CIia)^im3x" vih&n n lias a value of Ij., 5 02? 6. It appeal's, therefore, that the meclianisiu of reaction of l,ij. and 1,5 dihaloalkanes x-jith diallyl amine is as folla-js: Br(CHs)4 8 Br HlT , (CH aCIIf:CIIg.)a, mz) :i(CHsCi3i=GIl2)s
PAGE 36

29 This would appear very likoly in view of the work of Zliel axid Pecldiam^^^' involvinf^ the migration of the fui^furyl group in the disproportionation of fiirfiiryl quaternary amnoniun derivatives. These authors have shovoi the follov;ins! ^I/ \ reflux _^ (0113)311 + (0.05 M) (0.6 K) ^^ll^l IiWo yioid) This type of disproportionation also occiirred in attempting to use the method of preparation involving; the di(aryl-3ulfonoxy) alkanes in the presence of a large excess of diallyl amine. It appears that in this case the reaction involves first an intramolecular cyclization, as shown before, followed by a disproportionation involving migration of an allyl r^roup, ac shown above. This would be indicated as follows: (Where A is en arylsulfonate group) reflux in A* A(CHe)4^gA excess ^ ^ (wiaT^Tj^^^ mi (Cil8CH=CIIs ) a Gii3GIl=C A~ "7 HaN(CnaCII=CIIa)8 + inKCIia 011=0113)3 + (qna)4,5_^N OllaOI^OIIa

PAGE 37

30 Although Laakso and iJe^nolds reported a ylold of 'n*Sfo In the preparation of l,l.j-bis(dibutylamlno) butane, and 66.5^ In the preparation Ox 2,5-'ols(dlEiorphollno)hexane, the yield was only 10^ In the preparation of l,5-bls(dlben2ylanlno)pentane. It has been shown by von Baaiin, Kuhn, (27) and Goll* '' that the relative firmness of attachment of hydrocarbon residues to hetrocyclic nitrogen in quaternary ammonium salts Increases in the follo-;;ing order: allyl, beiizyl, methyl, ethyl, propyl, etc. 'Hiese autl-iors also point out that methyl groups ^^^ere lost, in the presence of primary or secondary amines, from quaternary anmonlum salts of hetrocyclic nitrogen compoimds (even piperazine, the least stable hetrocyclic compound studied) v;lthout ruptirre of the hetrocyclic ring. Allyl or benzyl groups are lost in preference to alkyl groups because of the resonance stabilized ions formed. The weakness of the carbon-nitrogen bond caused by the formation of such ions accoujits for the allyl migration shovm in this study and for the poor yield of l,5-bis(dibenzylamino)pentano. Tliis would also tend to support the theoi»y of a carbonium ion^ ^ as an intermediate in the dispropor-tionation reaction. The fact that furfuryl trimethyl ammoniuja iodide reacts vjith piperidine^^^' to give a l^% yield of N-furfuryl piperidine is evidence that the relative firmness

PAGE 38

31 of attachmont, as discussed by von Braiin, Kiilin, and (27) Goll, " or furfuryl is less than that of methyl. This would appear quite reasonable since the furfuryl group contains the allylic structure. The fact that this reaction proceeded to l^o% corapletion in throe hours during reflux at approxiraately 100°, while dlallyl pyrrolidinium broiTiide, after thirty-six hours at 109°, had shown no signs of reaction, would indicate thst the furfiirylnitrogen bond in furfxiryl trinethyi armnoniura iodide has a lesser degree of attochrient and a greater degree of ionic character than the allyl-nitrogen bond in dlallyl pyrrolidiniuci bromide. This view is also supported by the fact that the furfuj?yl carboniu-tti ion has a greater numbei' of stabilizing resonaiice etructiires than the allyl carbonium ion»

PAGE 39

32

PAGE 40

B •H o o H C3 cri H •H •a a o 01 •H H H o in; H CiJ O o 33 CM o '^. O W o o «

PAGE 41

IV PIffiPARATIOK OP BIS-QUATERMRY Al-ttlOillUI-I SALTS The bis-quaternary aramoiiium salts preparocl in this work wore made by adding the appropriate halide to the imsaturated tertiary diamine dissolved in acetone or acetophonone. In the preparation, of the hexallyl derivatives, two moles of allyl bromide were added per mole o£ diamine. In the preparation of the tetraallyldimethyl derivatives, dry gaseous methyl bromide was added in an excess which was collected in a cold trap. All of the hexaallyl derivatives were foxroed quite readily by gentle heating on a steam bath. The tetraallyldimethyl derivatives were prepared at the temperatiu'e of an ice bath, using a magnetic stirrer and allovxing the excess methyl bromide to escape into a cold trap while the mixture remained at room temperature overnight. In most cases the salts obtained were extremely hygroscopic and extreme care had to be taken in the purification of these compounds. In all cases the product from the reaction mixture was quickly filtered and then waslied with cold dry acetone and dry eth-^r several times. The salt was then quickly placed in a vacu'um desiccator and allowed to remain at room temperature at less tiian 0,5 mm. from six houi-s to one week. Tlie salts were then recrystallizod from hot acetone vjlth addition of absolute 31^

PAGE 42

35 ethanol until solution occurred. Only in some cases could absolute ethanol or higher alcohols be tised for recrystallization, since this usually resulted In a solution from which a cr7;-stalline product could not again be precipitated, Tetraallyldimethyltriniethylene diaramoniura dibromido, which was extremely hygroscopic, v;as the only salt which could not bo recrystallized. The salts usxac.lly melted over a one or two degree range and in some cases decomposed slic'htly, A, Preparation of the Hexaallyl Derivatives 1, Hexaallylethylenediaramonium dibromide +Br+Br(Ciia=GHClIa ) alUCHa )aiHGHaCIf=CHs ) a Sixty and five-tenths grams (0,5 M) of allyl bromide was added to a stirred solution of 50 g. (0,22? H) of tetraallylethylene diamine dissolved in 100 ml. of acetone. This mixture v/as allov;ed to heat gently on a steam bath for five and a holf hovirs. Tlie product was then quickly filtered, waslied vd-th cold acetone and dry ether and placed in a vacuum desiccator. Seventy-six grams (72,8^ yield) of vjliite crystalline product was obtained v.hich melted at l[|.9-50** after recrystallization. The product was previously reported as melting at l)|.7-8®.

PAGE 43

36 2» Hexaallyltrlmethylenedlainrioniimx dibromlde BrBr(CHa=CHCHg)3lJ(CH8)3N(CHaCH«CHa)a The same general procedure was follov;ed; eight and eight-tenths grams (0, 072U M) of allyl bromide was added to 8,5 g. (0,0362 M) of tetraallyltrimethylene diamine dissolved in 20 ml. of acetophenone. The mixture was allovjed to heat gently on a steam bath for ono-half hour. Fifteen end eighty-five hTindredths grams (91. 7J^ yield) of product melting at 185-6** (with slight decomposition) was obtained. The product was previously reported as melting at 185-7° (with decomposition). 3, Hexaallyltetramethylenediammonium dibromide +Br+Br(CHs=CHCH3 ) 3N (dig ) ^U ( CIIa=CiiGHa ) a Using the same general procedure, lii.,5 g« (0,12 II) of allyl bromide was added to 15 g. (0.06 M) of tetraallyltetramethylene diamine dissolved in 30 ml, of acetophenone. The mixtixre was alloivred to heat for one hour. Twenty-seven grems {91»7?^ yield) of product melting at 130-ii.l° v;as obtained. This was recrystallized from acetone, ethanol, and ethyl acetate mixture. The product melted at 168-70°. Analysis calciaated for CaaHaallaBrs: Br, 32,59^, Found: Br, 31*70, 31.7l4^.

PAGE 44

37 l^ Hexaallylpentamethylonedianmonlxira dibroraide (CIl8=CiiCiIa ) gWCCHa ) olUCiieCB=CHa ) a Using the aeme general procedure, 18»5 G« (O.l^lj. M) of allyl brortiido x^as added to 20,0 g, (0. 076 ll) of tetraallylpontanethylene diamine dissolved in 25 ml. of acetone. Heatinc was continued for fifteen minutes and after the resulting viscous liquid crystallized, 35,5 g. (92.2/d yield) of product melting at 153-5° was obtained. Analysis calculated for CaaIi4oI>!e3raS Br, 31,69,'^. Pound: Br, 31.70, 31.75/^. 5# Hexaallylliexamethylenediaramonlum dibroraide (CHb^CIICEs ) glUCHg ) oN(CHsCE=CHa ) g The produ.ct used in this vjork xvas some of that ^ X, . n (28) prepared by Angelo in previous work. The compovdid was made by adding 28,2 g, (0,23 M) of allyl bromide to a solution of 27*6 g, (0,1 M) of tetraallylhexamothylena diamine and 25 ml. of acetophenone. Forty-seven and tiireotonths grai-ns (91»5% yield) of product melting at 179-80° (with slight decomposition) vw.b obtained. Analysis calculated for Ga^II^gileBTs: Br, 30,83?^, Found: Br, 30.77^.

PAGE 45

36 6« HexaallylheptamethylGnedlararaonivm dibromide (CHa=CHCIls)3li{CHs ) 7H(CHaCK=CHa )a Using the same general procodure, 2$,k. g* (0,21 M) of allyl bromide was added to 30. g, (O.lOi}. 14) of tetraallylhoptamothyleno diamine dissolved in 60 ml, of acetone. The mixture was heated for ten to fifteen minutes, x-ortythree grains (77,8% yield) of product melting at 192-3** (with slight decomposition) was obtained. Analysis calculated for CgeliA^iisBra: Br, 30,02/^. Poimd: Br, 29,9k» 29.96/f. 7t Hexaallyloctainethylenediammonium dibromide (CHB==CHCHa ) a^i ( Clis ) a N ( CilsCI^CHg ) g Using tiie same general procedure, 16,5 S« (0. I36 M) of allyl bromide vjes added to 20.0 g. (0,068 M) of tetraallyloctejnethylene diamine dissolved in 25 ml. of acetone. The mixture was heated for ten to fifteen minutes on a steam bath, whereupon a viscous liquid separated. This was placed in a refrigerator ove2?night and 15 g. O+O.p/'^ yield) of product melt in?: at 195-7** (vjith decomposition) was obtained. Analysis crlculetod for CaeH4QNaBr8J Br, 29. 2i^^. Pound: Br, 23,57, 2o,56^.

PAGE 46

39 8, Hexaallylnonaraathyl(3nediar!iraoniuDi dlbronide (ClIa=CPICiIa)aN(CHg)9lUCHaCI^CHa)3 Using the same general procedure, 18,2 g, (0,15 ^0 of allyl bromide v;as added to 16 g» (0,05 M) of tetraallylnonamethylene diamine dissolved in UO ml, of acetophenone. The mixture was heated for fifteen minutes. Twentythree and six-tenths grams (81}^ yield) of product melting at 201-2° was obtained, Anal^/sis calculated for Ca7.H4aNeBrs: Br, 28,26^. Pound: Br, 28,itl, 28,52?6. 9» Hexaallyldocamethylenediammoiiiian dibromide +Br" +3r" (CIIs^CHCHa ) all ( CJig ) , oK ( CIIsCiI=GKs ) a Using the same general procedure, 2k g, (0*l\. M) of allyl bromide t-ias added to 30 g. (0, 09 M) of tetraallyldecaniethyleno diamine dissolved in 50 ^« of acetophenone. The mixttu^e vjas heated for thirty minutes. Fortythree graiiis (83^ yield) of crude product vms obtained. This was recrystallized from n-amyl alcohol. With the bath previously/heated to I80*, the product melts at 185-6° (v/ith decomposition). Analysis calculated for CaoHooNaBraJ Br, 27. £32/^^, Pound: Br, 27.76, 27. 7^!^.

PAGE 47

14.0 B, Preparation of tho Dlnethyltetraallyl Dei-ivatlvag 1» Diraethyltetraallylethylenediarraioniioin dibromlde (CHe^CHCIIs ) ^:i ( die ) glv ( C:[3Cn=CIl2 ) B sJo.uuiie;3.au ci., cii a ^^^3 Tiie product used in this work was sone or that prepared by Angelo*^'-^' in previous work. To illustrate the general procedure, the preparation will be given. Dry methyl broraide was allovied to bubble tlirough a gas inlet tube of capillary size into a mixture of 28 g« (0,127 M) of tetraellylothylene diamine and $Q ml, of acetone. Ihie mixture v;as stirred with a luagnetic stirrer; cooled by an ice bath; excess CHaBp vapors wore collected in a cold trap iraroersed in a mixture of dry ice and acetone. Precautions to exclude moisture were taken, Tlie addition of methyl bromide was discontinued after six houi-s and the excess Cria3r vras allowed to evaporate into the cold trap as the mixture remained at room temper a tu-re overni^t. The mixture was filtered, washed with cold acetone and dry other, and placed in a vacuum desiccator. Fifty grams (96,2^ yield) of product melting at 191-2° (with slir:ht decomposition) was obtained. Analysis calculated for CieHaoNsBTa: 3r, 38.97^» Found: 3r, 38.86;!^.

PAGE 48

la 2, Dlraethyltetraall/ltrlraethylanedlammoniiim dlbronlde (CHa=CHCHa)s^J(0IIa)3lJ(CIlsCfI=CIIe)s cms ^Ha (28) 'jHho prodiict used in this work was prepared previously usiiig the above procedure, Hairty-eiglit grama (89«6;J yijld) of product melting at 109-11*' (closed capillary) was obtained. Analysis calculated for CiyHagiisBrg; Br, 37.6?;^. Pound: Br, 37.63, 37.92^'. 3, DiraethyltetraallyltetrairiethylGnediaranoniuin dibroraido +i3r~ +3r(Cli^CllCllz ) yUCKs ) ^.i^CCHsCIfeCIIs ) g G^ CH3 using the same general procedure, methyl bromide was added to 16 g. (0,06l}5 H) of tetraallyltetraiaethylone diamine dissolved in 100 ml, of acetone, Txirenty-rour gra-ms (85/^ yield) of product melting at 163-5° (X'.'ith decomposition) was obtained. Analysis calculated for C.eHs^NaBra: Br, 36.^75^. Found: Br, 36.14-5, 36.36/^. l\.m Dinethyltetraallylpentamethylenediammoniuai dibromide +Br^.Br( CHs=CHCHs ) 2 1^ ( CHa ) gll ( CIIsCI?=CKg ) « 3C Ci CH3 CH3 Using the came general procedure, methyl bromide was added to 20 g, (0,0763 M) of tetraallylpontamethylene

PAGE 49

diamine dissolved in 100 ml* of acetone. Tlilrty grans {87.3^ yield) of product meltliie at 133-5** was obtained. Analyslc calculated for C jgiiaglfaBra: Br, ^$»2.h^o. Found: Br, 35.36, 35.28^. 5« Diiaethyltetraallylhexarnethylonedianjraonium dibromide ^.Br" +Br" (CIia=CiICn2 ) aiUCila ) eiiCGHaCI^Cila )« The pi'oduct used in this work was prepared^ ' previously using the same procedure. Forty-eight grams (quantitative yield) of product melting at 202-3** (vith decompoaltion) v;as obtained. Analysis calculated for CaoIIssNaBre: 3r, 3[|..28,<. Pound: Br, 3h»^^% 6» Dimethyltetraallyiheptaiaethylenediaiamonlum dibiomidn +3r+3r~ (CHa=CHCHc )e> (Oils ) 7K(CIIaCH=sCH8)a m^ CH, tJsinn the same general procedure, methyl bromide was added to 20 g, (0,069 M) of tetraallylhoptaBiethylene diamine dissolved in 100 ral» of acetone. Twenty-eicht and five-tenths greras (86, 2J^ yield) of product melting at 193-^° (v/ith decomposition) was obtained. Analysis calculated for Ca.n^oi'eBTat 3r, 33«27/^. Foxind: Br, 33.36, 33.3i|^.

PAGE 50

1^3 7» Dinethyltetraallyloctaraothylenediamraonii-ira dibromlde (CHBCH~CHa)a;^(Cn8)oIl(CH3CII=CHa)3 Cfla CHa Using the same general procedure, methyl bromide \ms added to 20 g. (0,068 M) of tetraallyloctamethylone diamine dissolved in 100 ml. of acetone. Nineteen and seven-tenths grams (58»7J* yield) of product melting at 178-80** was obtained. Analysis calculated for Cga^Ua^'aBTsl Br, 32,33/^. Pound: Br, 32.37, 32,kOfo, 8. Dimethyltetraallylnonamethylenediararaoniura dibromide +Br"" +Br*" ( GHa«CHCHa ) oN ( 011^ ) 9 K ( CIl8CH=CiIs ) « Clla CII3 Using the same general procedure, methyl bromide tms added to 22 g, (0. O69 M) of tetraallylnonamethylene diamine dissolved in 100 ml. of acetone, Tv/enty-eight grains (80^ yield) of product melting at 187-9° (v-ith decomposition) was obtained. Analysis calculated for Cs3H44l.'aBrs: Br, 31. US/^. Pound: Br, 31.23, 31.25/^« 9. Dimethyltetraallyldecamothylenediammonium dibromide ^Br" +3r"* (CIIa=CIICIl8 ) a'HCHa ) , olUCIIaCIlsCHa ) a C^a CHa Using the same general procedure, methyl bromide v;as added to 16 g, (C, 06ii5 ^'0 of tetraallyldecamethylene

PAGE 51

diamine dissolved in 100 ml. of acetone. Twenty-foxir grama (85J^ yiold) of product molting at 163-5** (with decoriposltion) wes obtained. Analysis calciilated for C24li43NaBra: Br, 30.595^. Pound: Br, 30.61, 30.1|0^.

PAGE 52

kS 03 > M < Q O K M M M Q < CO H Eh o H •H >H 03 O H o o o H 3 o CO o I 02 a o H H lA H CO ir\ o o A ^^ CO CO O^O C»-CO fHCVJ -.D-^ 0_=iOlA I •••• • •••••••• I (HHHf-l O OC>cOcOCOCDt>~r«I r<^(^rarO oo C\J(\|C\JOJOJCVlC\Jf\J vO I o t lA CO CO O (0 n 0) «0 Oi a o H -P O H t^ H H 0) !<; O c3 o H -p o H H >I] o -p o § is -p o +J H H H cd O CO • • iH O CO fn I LA H CI o o H Q 0) P. rH H H W S o CO I l>H o ot 4 41 O o • © o •d o rA I O « C\J 01 ^ "lb 01 o 5it a ^© 2 ^ O tH H 1>5 -p © o H C5 A] AJ o o H C} f^ p^ (§ J§ o © © H •P © a •p o o H ;=^. H H Cd CO © AJ CO • • AJ CM I H O CO h (9 o IH o © o H .d -P o o H cd od © I \A 00 H 01 at o o © o p © § © H h> H H cd © cd

PAGE 53

U6 H O to TJ o o CD o • • o CO CO OCO Of CO OJ r^ CO CO O o CO r^^CM C\J • • • CM f\l H iH O • UN CO HO sO_;t • • o o r«-

PAGE 54

V SUSPEIISION POLYl'IERIZATIOW OP BIS-qUATE?JIABY AMMONIUM SALTS A* General Jlacusslon of Method In an attempt to prepare the roolns in a rather tinlforra size and shape, end to allow for dissipation of the heat of reaction which could cause decomposition of the quaternary ainmoniura derivatives, tlie suspension polymerization teclinique was employed. Using this method a number of difficulties are encountered. Constant high speed stirring must be maintained; the proper shape stirrer must be enqsloyed aiid the proper carrier solvent must be used. An attempt to solve these difficulties was made by carrying out various suspension polymsrizations rising tetrsallyl ammonium bromide as the monomer, Tlie stiri'ing problem was finally resolved in the follouing manner: A Palo Myers' motor, controlled by a variable voltage regulator, was aaployed at very high stirriiig rates. The voltage setting for the high speed varied from li$ to 6C volts. This v;as due to slight differences in the binding of the stirring sleeve. After experlJiienting with various stirring sleeves, a hoavywalled piece of glass tubing inserted in a rubber stopper was used. Tills compensated for slight vibrations at very

PAGE 55

high speeds. The stirrer ijvls made of glass and was shaped like a cork scret-/. The stirrer and stirrer sleove v/ere fitted as tinhtly as possible. Any initial binding v;as compensated for by allowing the stirring device to proceed at high speeds, using graphite and mineral oil as lubricants. This was done previous to each experiment and resulted in more uniform stirring. The stiiririg sleove and the stirrer wore oxoessively and had to be replaced quite often. Further, this wearcaused the deposition of a fine glass powder in the reaction flask. The most difficult phase encountered in trying to obtain uniform bead particles was the sticky pei-iod, v;hich caused merging of lndivid\ial dispersed particles into agglomerations • (pq) Hohenstein and Mark* ^' indicate that polymerization in a heterogeneous suspension in which the monomer is mechanically dispersed in a liquid not a solvent for it or for any species of pol^^mer moloculos, and in which the initiator is soluble in the monomer or monomer phase, results in a polymer bead or pearl. In such casos the polymerization takes place in each monomer globule and converts it gradually into a polymer bead or pearlj the liquid plays only the role of c. carrier, which favors heat transfer and aritation, but docs not interfere v;ith the reaction.

PAGE 56

I|.9 However, after a period of time, the clo^^lQS, which nov; represent a rather concentrated solution of polymer In monomer, become r'osum.j and upon mutual contact In the system stick together and form agglomerations which cannot easily be broken up into individual globules of the original size or shape. It frequently happens that upon continued stirring the agrloiierations do not disintegrate, but coalesce into larger units. Tae authors^ °' describe several ways of bringing the system over the stlclcy period without merging of the individual particles: (1) Suspension stabilizers can be used tliat can covei' the sujpfaces of the globule with thin layers of inorganic or organic substances which do not interfere with the reaction, but prevent or diininish the tendency of the globules to stick together during the gummy period, (2) The interfacial tension between the carrier solvent and the monomer phase can be increased by dissolvi:ig electrolytes in the carrier solvent if it is water* (3) The density of the carrier liquid can be adjusted to the density of the globxilos during the gummy state. In doing so, the tendency of the sticky globules to accumulate either on the surface or at the bottom of the suspension is removed, (h) The viscosity of the carrier liquid can be increased and thereb:/ make it more difficult for the dispersed globules to collide vigorously enough to merge.

PAGE 57

$0 In this respect, many experiments v;crc carried out v.'lth tetrf;allyl ammonitan bromide as the monomer; the precedixif': su^^gostions were employed until the best conditions were fotmd. During these experiments stabilizers such as bentonite and talc were used; carrier solvents such as ethyl benzene, chlorobenzene and toluene were used; increased viscosity of tJie carrier liquid was investigated b7 various additions o£ mineral oil; since the carrier liquid in this particular polymerization ;^^as not water, adjustment of the intarfacial tension by the addition of electrolytes was not eriploycd. The best conditions for carrying out this particular polymerization were as folloivs: A stirring rat© adjusted to 1^0-60 volts as indicated previously, a temperature level of 60°, a carrier solvent consisting of 1^0 ml. of mineral oil and $0 ml. of ethyl benzene, a r:onomer phase consisting of a half a milliliter of v/ater per gram of quatemai^y salt, and an initiator coxicentration of two drops (approximately 0,012 g./drop) of 60J^ t-butylhvdroperoxide solution per gram of quaternary salt. In all cases, using tctraallyl amaoniuni bromide as the laonomer, 90-100^ yield cf ixxsoluby^ resin vias obtained, Uowever, in no case were any wsll-shaped beads formed. The product approached a spherical shape, but had much rendom ^ape distribution, aich of the size distribution of tixe particles obtained was in the desired 15-60 mesh region.

PAGE 58

51 The stirring at 60° w&s continued for forty-eight hours after which the product was filtered and washed successively with heptane, acetone, rlcohol and ether. The product was then soaked for appro:?dLiTiately one hour in hot heptane and then filtered and washed as before. The product was then dried for twonty-four hours in an oven at 60**, This procedure insured the removal of the film of mineral oil. At this time the swelling coefficient of the resin v/es deter>mined, Tho ratio of the wet bromide volume to the dry bromide volu:iie of the rosin is the swelling coefficient. It was measured by allowing the dry bromide form of the resin to settle in a graduated cylinder and determining the volume. The vjot volTome was detewiined by allowing the stme resin saznple to soak in distilled water until no further change in volume iijas apparent. This general procedure was applied to the polyi?ierization of the quaternary derivatives of the diamines described previously. In all cases, except for the hexaallyldecamethylene end the dimothyltotr£allylho:fcamethylene dlaramonium dibromide derivatives, the resin pi'oduct was not perfectly bead shaped. In most cases the product vreis of random shape, but usually in tiio desired 15-60 mesh region. The resin particles of these two derivatives were perfectly bead shaped; only the docaraethylone

PAGE 59

52 derlvativG retained its bead shape thpovisliout all operations, such as washing, exchange or recycling. The polymerization of thtquaternary derivatives of the pontaiaothylene and octa^iethylene derivatives differed from the general prccediire described above. Hiese polyraoriaatio-xs wore carried otit in the bulk phase using the same raonomer, initiator and xrater ratio, Tlie resulting resins were ground to the desired I5-6C raesh size and treated similarly* B. Bxperiirantal liesults The following tabic gives a surai?iary of the results of the polymerization of tlie quaternary derivatives of the preceding diamines. It is believed tiiat any yiold over 100^ is caused by glass particles deposited from the stirring mechanism. In this section the resins will be given the follov;ing notation: C3H polyraer of hexaallylethylenediarmnoniun dibromide, G3DT = polymer of diraethyltotraallylethylono dismmonlum dibromide, etc.

PAGE 60

53 TABLE V P0LYM2iiIZATI0N fffiSULTS Hesln

PAGE 61

VI EVALUATION OF CAPACITY STUDIES A» Method of Analysis The method of detemiining the capacities of the resins in this investigation was essentially a titration procedure similar to a neutralization titration of a stronT base an.d a strong acid, A Icnown quantity of tiie polyquaternary ammonium bromide form of a resin was quantitatively converted to the polyqua ternai^y ammonium hydroxide form of the resln« An excess (compared to the theoretical nuaiber of mlllieqtilvalents of polyquaternary ammonium bromide) of a salt of a strong mineral acid was added to the poly-q^xaternai'y ammonium hydi-^oxido. An ex* chan£':e took place and hydroxyl Ions were released into solution. Suitable titration of the hydroxyl ion content by a strong mineral acid then determined the exciiange capacity of the res3n In terms of mllllequlvalents of hydroxyl Ion exchanged per gram of dry brcanlde form of resin. This method is essentially tiiat described by Butler and Bunch, and Butler and Ingley, ' In which the actual titration was carried out using a Beckman pH meter, adding 1,0 ml, increments of standard acid at 3,0 minute intervals and measuring the pH of the solution

PAGE 62

55 at the end of oach three minute Interval. This routine was continued until the titration v;as complete. The pH was plotted against milliliters of acid added and the capacity was obtained from the sjnount of acid required to cause a ^arp break in tiie curve. However, the capacity of the resin measured by this method never approached the theoretical capacity of the resin, end th© break in the titration curve usually shov;ed a shoulder which i^ras interpreted at tlie time as amine capacity, ^^* Iiusa^°' developed an "equilibritmi" titration method for determining the capacity of this type of anion excliango resin. His method consisted of preparing equal samples of resin in the bromide fomn, quantitatively converting oach sample to the hydroxide form, and adding various known quantities of standard KBr solution and HBr solution. The pH of each sample vjbs measured initially and at various intervals, extending to l\2Q hours in some cases, A plot of pH versus milliliters of standard II3r originally present gave a titration curve vdiich appeared very similar to a standard acid-baso titration and fiom which a relatively acc-urate determination of capacity could be obtained. The advantage of tliis type of titration is that the measured capacity is the capacity attainable after the system has reached equilibriujm. The disadvantage of this method is tlae prolonged period of time involved end the number of samples measured.

PAGE 63

56 The equilibriiim capacities, measur-ed by Husa, often appi-^ached the theoretical values and were rtrach higher than the values determined by the rapid titration method. Differences in equilibrium time, pH of solution, and ionic strength were suggested as factors causing variation of the results given by the rapid method and (30) the ecjiilibrium method. The equilibrium titration curves shwjed no amine capacity, as v;as suggested pre(3) viously, but gave a smooth curve with a shaip break. As an illustration, the equllibriuir. titration curve for the resin of hexaallyldGcaiiiGthylene di ammonium dibromlde follows. This curve appeal's by the coxirtesy of Iv, J. Husa^^ ' from his Ph.D. Dissertation, University of Florida, 1953. P. 82. The following procedure was used to determine the capacity of the bis -quaternary ammonium resins obtained in this work, A weigiied cpiantity of resin (approximately 0.2000 g. ) was placed in a sintered glass filterii^g funnel (medium). The funnel was stoppered and the resin soaksd in ten milliliters of l\^ iJaOH solutionj the resin sample was filtered, waslied and soaked ©very day for 10-12 days. After this time, the tiirbidity test for bromide ion, precipitated as silver bromide, vxas less tloan ten paints per million, Tiie resin was then wasiied free of any hydroxyl ion by using distilled water and checking the filtrate with phenolphtlialelnr

PAGE 64

^7 a> •r-l u •H o H <; O (1, 9 C r-l "we © K C -P "i-. O O o •H i! « K !-. rH H JJ ^^ o ^C o (0 o O -P U 3 3 2 _ c j5 "^ w 'XI CO X3 ^ t^ O O o w O rH &< H 3

PAGE 65

58 The rosin was placed In a 2^0 ml. bealcer v;ith 7^ ml, of distlllod water. To this lulxtup© i/as added $0 ml. of 0.20 i'l KBr solution (ten milllequlvalents of broiulde ion). After standing for five minutes, the mixture xtas titrated with 0»C195 N HBr solution by using a Beckman Model K autoniatic titrator. The endpoint was set at pH 7 t>y adjufltment vi^lth a standard bufi'er solution, and the anticipation rate was set at 7» '2hQ first 15-20 ml. of standard acid was titrated rapidly, but thereafter the titration proceeded rather slowly iintll the end of tlTD allotted one hour period. There was still sa.e exchange takirig place at the end of one hour, as was observed when soiiie samples Xijere allowed to titrate further. This type of asymptotic approach to complete neutralisation was to be (31) e:cpect©d from the data of ifc.sa. The theoretical number of milllequlvalsnts was calculated for each resin based on monomer units. The actual number of milliequivalents exclianged per hour was calculated and was eicpressed as the fraction of theoretical exchange pei^ hour, B, Summary of ExD , erimental Results of Capacity Determinations The capacity determinations of the bis-quatei^nary ammonivuQ resins were carried out in the mianner described. A suraiiiEry of the data obtained follows in Table VI, In

PAGE 66

59 this tabic the roslias are given the seme notation as before, whero Cgll = resin of hexaallylethylsnediaitcrionium dibronide, C»DT = resin of dimethyltetraallyletliylenediaramoniim dibromide, etc»

PAGE 67

60 H > fH CO o M H O < ft) O CO I CF> H O • O 0} rJ O o c; o i^ o > o a ci o •H P O Pi -P •H EH O P< •H O -P
PAGE 68

61

PAGE 69

62 G, Discussion of Capacity Determinations The type of rapid titration determination which was carried out in this work proved to bo quite adequate. The time and number of samples v/ere kept at a minimum. The Beckman Model K autoKiatic titrator proved to be satisfactory in measuring the fraction of theoretical exchange per hour. It is believed that suitable experimental techniques co-old be developed, involving the automatic titrator, to differentiate between the initial stage of exchange, apparently due predominately to a mass action effect, and a latter equilibrium stage of exchange, apparently due to a diffusion rate factor. It is conceivable that relative diffusion rates of various ioxis could be determined and differentiated from total exchange rates, vihich were probably measi-ired in this vrork. It is believed that an equal malliequlvalent basis would be more reliable tlian an eqiial weic:ht basis of comparison of the resin exchange rates. In the previous determinations it should be noted that the tlieoretical number of mllllequlvalents of the resins ranged from 0,707 to 0,977, However, the adequate excess of initial bromide ion should compensate for differences in mass action effect due to the unequal \init milliequivalent basis; the initial quantity of bromide ion consisted of 10,0 mllliequivalents.

PAGE 70

63 The most Interesting effect, as seen from the exchsnge capacity data. Is the definite Indication of a raaxlnuin exchange per tonlt time of the pentaraethylene derivatives. This effect is shown in Pigxires 2 and 3, It could be merely a coincidental factor dependent upon so.ne non-uniform polymerization technique. Although every polymerization was carried out as uniformly as possible, there v:ere, in many cases, noticeable differences in the induction period, the globule size during the polymerization, end the final particle size and shape. Hoi'jever, the effect of raaxiraxrai exchange of the pentamethylene derivatives is a factor that could be considered from the basis of diff eJ:'ences in the unit s tincture of the polymer. A suggestion as to the uniqueness of the structural conf ig\u*ation of the pentamethylene derivatives is the possibility that uniform linear chain growth thro'ogh the allyl double bond would result in a polymer with fivo carbon atoms betvjeon each exchange center. This type of linear growth would cause only the pentamethylene derivatives to have equal numbers of carbon atoms between successive nitrogen atoms. This is illustrated in the following scheme:

PAGE 71

Pi)T:iAre 2 64 •H P u P4 o El •w O ti o •H 4J O OS f< 23lf56 789 10 Figure 3 0.9 0,8 0.7 0,6 0.5 Dimethyltstraallyl Derivatives 23^56789 10 Number of Carton Atoms in Methylene Chain

PAGE 72

65 Cn8=CHCHaN(CHa ) 6-N-CHaCH=CIIa+Z» V Initiation vjlth free radical Z" +/ H H CHa=C:iCIi8N(CIl3)eN-CIIaC— C : Z • H Propagation xd.th monomer molecule Z HCH R. 'Re / % /« \+ +/ / H H \+ +/ CHs=CIICIl2lJ (CHs ) eH-CHaC : C— C-CKsil(Cria ) ei^-CHaCS=CIIa H H • Since there has been no definite proof of structure of these derivatives, this suggestion is to be considez^ed as a possible course only. The capacity data of Table VI seems to show no definite correlation between the swelling coefficient (related to the apparent degree of cross-linkiiig) and tiie ©."Change capacity of the resins. It was expected that replr cement of one allyl group of a quaternary ammonium derivative by a methyl group should give a rosin having less cross-linking and higher swelling, there being fewer allyl double bonds available for cross-liiiking. Consequ.ently, the methyl quaternary resl-is should Imve higher ezcharigo values if we consider a lov; degree of cioss -linking to allow easier ion movciaent in a less tangled network. However, the data in Table VI does not bear out this contention

PAGE 73

66 and, contrary to o:'.pectations, the methyl derivatives have less exchange capacity per unit time tlmn the corresponding allyl derivatives in practically every case. There was no e^qjoriinental evidence obtained to show a comiection between the e::cl^nge capacity of the resins, particularly as rocardc their selectivity towards ions of different size and chaa^go, and their respective swelling coefficients, apparent degrees of cross -linking or variable distances between qLiatomary aiamonium centers. (32) However, sorae uork is proscn-cly being carried out in this laborato2?y by L, G, Kulkarni, relative to the possible selective exchange properties of these resins.

PAGE 74

VII POLmERlZATIOll I>!ECHANISM STUDIES Previous work^ ~-^' carried out in this laboratory pertaining to the freo radical initiated polymerization of allyl or substituted allyl quaternary ammonium derivatives gave infusible and water-insoluble polymers only when derived fixDm monomers containing three or more allyl groups, Tho product resulting from the attempted polymerization of a monomer containing one or two allyl groups eitiier was not polTmaized or v;as assumed to be noncrossed linked, since it was completely soluble in vrater. These results are contrary to the accepted views that monomers containing one double bond result in a linear chain type product possessing some degree of solubility; monomers contaiiing two or more double bonds result in a tiiree-dimonsional, cross-linked product possessing little or no solubility. Since the products of the polymerization of allyl quaternary ammoniirai salts liave appeared somevdiat abnormal in their properties, it was decided to investigate tho mechanism of this type of free radical Initiated polymerization, A piperidine ring structure formed by an alternating intramolecular-intermolecular chain reaction was 67

PAGE 75

60 considered as a possibility vrhich could preverr": cross llnliine In the poljmerlz&tlon of quatonmry cinnonlum dei'ivatives containing only two allyl double bonds, Simpson, Holt and Zcite^^^' liave indicated rrc:;i the correlation of data on tli© polpaer-ization of diallyl phthalato that at the eel point much of the carboncai-bou double bond, supposedly free to crost: link after tlia gel point, is tied up as a cylic strixt-uro due to intramoleculcr polynorization. It xms decided to approach this subject from the aspects of suitable synthetic adaptions of nononer structure, degradation sti:idies, infrared analyses and molecular weight deter ruinations. A, HonoEier Syntheses 1, Preparation of diethyldiallylaxnraonium bromide + Br(CH»CHg ) gl-J ( CH2CH=CHg ) ^ Sixty and five-tenths gratis (0,5 ^0 of allyl bromide was added to 50»0 g. (0,1(1;^ i-I) of diethylallyl amine dissolved In 100 ml, of acetone; the diethylallyl amine had a boiling point of 110" and nj® 1.1^90; Liberraann and Paal '^'' report the boiling point as 110-13°, Upon addition of the bromide, the mixture beca^Tie cloudy and crystals began to form. The product was wasiied and

PAGE 76

69 decanted sevoral tiiiies witli cold acetone, filtered and dried in a vacuum desiccator. Eighty-seven grams (8i|»l^ yield) of white, hygroscopic c-'ystals, melting at 155* (closGu capillary) were obtained. A^ecr'/atallization f3X>m hot acetone and a aiaall amovuit of absolute otnanol gave tile sauo liieltiiig point. 2. Preparation of triethylaUylammoniua bromide + Br" (CHaCHe ) alUCHeCIJsCHe ) Tlilrty-si." and tliree-tonths grams (0,3 M) of allyl bromide was added to 30»0 g, (0,296 fl) of triothyl araine. The Sicrao proc3dure was followed and 60,5 S« (92,1/j yield) of white hygroscopic crystals was obtained. The product :^i8lted at 229-31** (idth deccnposition; closed capillary). 3, Preparation of l,l!-bis(diethylallylanmonluin)butene2 dibromlde + + (CKgCtia ) a;iCH3ClI=CHC'.l8lUCHj30IIa ) a CHaS=Cncms CIIaCifaCHs The intermediate diamine, l,[|-bis(diethylamino)butene-2, was prepared in a 31,8^ yield by the procedure of Amundsen, ^-^^' One hundred and seventy-eight greras of light yellow liquid boiling at 118 V^O mm, and Jiaving an ng'1,).|.573 was obtained, Amundsen reports a boilirig point of 115-16/20 ram. and an n§ 1,I}.532. The product rapidly decolorizes imen allowed to remain in a sealed container at room temperattire.

PAGE 77

70 Twenty grans (0,10 H) of Ijlr-bisCdiethyleialno) butene-2 dissolved in $0 rJ., or acetone v;as allowod to react t,d.tli 26«6 g, (0.22 H) o" allyl bronide in the sonie manner as beforo. The crude product was rccrystalllzed froTd hot acetone and absolute othonolj 26,5 C« (59^ yield) Ox white, powdery product neslting at l35-^** (with decomposition) was obtained, Goette -^ reports the melting point of tliis product vrlthout rocr^rotallization as 172-3°, [{., Preparation of diallyltetraethyldecaniethylenediaaimoni'uia dibroiiiide +Br+BrCHaCH8)a;^(CIia) ,oH(CIi8Cii3), '8 CHa^CHCTHe CilsCH=CHa The Intei'rnediate diamine, tetraothyldecanethylone diamine, was prepared in the seiie nanner as tetraallyldccairiethylene diaiaine in Section II, On© hundred and twenty-six grains (1.73 K) of diethyl anino in a slurry of Qk g, (0.5 M) of NaliCOa and ^0 ml, of i\«ter, was allot; ed to reflu;: with 52 g. (0,173 ^0 of deca^niethylene dibroraide for ten hom^s. Thirty-nine grams (79«5/o yield) of light yollov/ liquid boili.c at 120-6V0.02-0. 05 mm. and having an n^^I.I4J+9S was obtained. Analysis calculated for CjeHioNs: C, 76.2^5 H, lli-.l^j II, 9.86^, Found: C, 75.97/^i H, lk.OQ%; IT, 9.B2,^. Fifteen grams (C.053 M) of tetraethyldecamethylene diamine, dissolved in 50 ml, of acetone, was allovjed to

PAGE 78

71 react v;ltii I3.3 g. (0,11 II) of allyl bromide. After being treated In the same manner as before, 17»0 6» (6l,2/o j-lold) or uhite, very iiygroscopic salt noLtiiig at SO^^-y** (v.lth decoiiiposition; closed capillary) was obtai.acd. 5. Attempted preparation of quaternary salts of li-diethylaxiinobutcne-l (GH3CiIa)2n(CIlBCHaCIJ=GHa) R The intermediate amine, li-dletiiylaialnobuteno-l, was prepared by roflixxinf; for eipliteon hoxors a mixture of 190 r^, (2.6 M) of diethyl amine, 8Ij. g, (0,9 M) of IJailCOa, 50 ml. of water arid 30,0 g. (0,33 M) of l|-chlorobuteno-l, (The h-clilorobutene-1 was redistilled and the portion boiling at 75»0-75»f3** and having an n^** 1,14238 was used; (37) Juvala reports the boiling point as 75«0** and the njj as 1,[!233). After separation and purification, 15«i| g« (36»6^o yield) of clear liquid boiling at 13i>'-5° and having an n^ l,i|250 t-;as obtained. An infrared spectrtun indicated the cxiaract eristic teiminal double bond absorption iliich I'jas desired in this preparation. Attempts to prepare quaternary aramoniuBi derivatives of l4--diethylarainobutene-l did not prove very successful. Only in the case of the allyl quaternary was a crystalline product obtained. The reaction of equiraolar portions of allyl bromide end ij.-diothylaEiinobutQne-l dissolved in

PAGE 79

72 acetone, g&vg a white, hygroscopic product wliich melted at 208-10* (with decomposition). The reaction of equimolar portions of 5-t)romopentene-1 and ^-diethylaminobutone-l dissolved in acetone gave an oil, idiich, after many recr^T-stallisation attempts, gave a very hygroscopic soml-solid. This soral-solid \^^as obtained by careful waslilng, decanting with cold acetone and allowing the residuo to remain in a vacui^m desiccator at leas tlian one millimeter for several days. The mixtures of U-chlorobutene-1 or 6-bromohexene-l with equimolar portions of Ij--diethylaminobutene-l dissolved in acetone showed no signs of reaction after gentle heating and after remaining at room temperature for six months, 6# Attempted preparation of quaternary salts of 1diCpentene-li) amino butane (CHg^CIICKaCHaCHa ) aNGHaCHaCHaCHa R The intermediate amine, 1-di (pen tsne-k) amino biiteiie, was prepared by refluxing for three days a mixture of 21.9 g» (0.30 M) of n-butyl amine, 87.O g, (0,7 M) of HaaCOa'HaO, 100 ml. of water and 100 g. (0,6? M) of l-bromopaitene-Ii. (b.p.-53-ij.V5l-2 ram, and n^"^l,i|636j Juvala^' reports the b.p, as 5o°/75 ran,). After seperation and purification, Ul»5 g» (66,2/^ yield) of light yellow liquid boiling at 228-37° was obtained. Tills was redistilled and a light yellow liquid boiliiig at lli|-l6V9»ll mm, and having

PAGE 80

73 so an njj l,Ii.525 was obtained. An Infrared spectrum gave the characteristic terminal double bond absorption desired in this preparation. Analysis calculated for 0,4112711,: C, 80.3/^^; H, 13.0J^j II, 6,7^. Found: C, 80, Ul^; H, 12.60^; N, 6.73^. All attempts to prepare crystalline derivatives of l-di(pentone-i.i.) amino butane were unsuccessful. The amine was reacted with methyl bromide, rosthyl iodide, allyl bromide, benzyl chloride, hydrogen chloride and hydroften bromide in varioi:.s solvsits, but no crystalline product was obtained, 7. Attempted prenaration of quaternary salts of l-di(hexene-5)a2ii»o butane (GHB=GHCHsCHaCIIaCIi2 ) gH (CllsCIIaCIIeCIIa ) H The 1-bromohexe ne-5» used to prepare the intermediate amine, iv'as made by treatiiig at 0-3° a dry pyridine solution of 5-hexene-l-ol with phosphorous tribromide, Seventy-tvio a:id ono-tonth grams (i}i^,2/;J yield) of 1-bromohexeno-5 boiling at 78-80V6I mn. and havirig an n^*^l,it676 \m.s obtained. Price -^ reports the boiliiig point as 75-8V61-2 ram. and the nj® as l.I|620. The intermediate amine, l-di(hexene-5) amino butane, ;-;as prepared in the same manner as before. Fourteen and fivo-tenths grans (0,20 H) of n-butyl amine.

PAGE 81

7U 82.5 g. (0.50 M) of KaCOa'l l/2(lia0), 2,i^ g. of copper powder, 100 ml. of water and 70,1 c» ('^•U3 ^D oJ^ 1-bromohexQno-5 were refliixed for thi^ee days. After separation and purification, 29 g. (61,1 % yield) of a light yellow liquid boiling at 175-81;. V^U mm. and having an nj l.[j.51>6 Xiras obtained. An infrared spectrum gave the characteristic terminal double bond absorption desired in this prepera+-* '>'". Analysis calculated for C,Qii3,Ii,: C, 80,93^^^ H, 13,17f^'; H, 5»90^Pound: C, SO.lUi^; H, IZ.'dTM i^, 5.98;^. All attempts to prepare crystalline derivatives of l-di(hexene-5) amino butane were unsuccessful. The amine v;as reacted vri. th methyl bromide, methyl iodide, allyl bromide, benzyl ciilcride, hydrogen chloride, and hydrogen bromide in various solvents, but no crystalline product was obtained, B, PolT/merlzation Products and Results In an attempt to explain the water soluble property (apparent lack of cross-linking) of resins obtained from the pol^erization of quaternary ammonium dex^ivatives containing tvro allyl groiips, an alternating intramolecularintermolecular chain reaction forming a pij>erldine ring structure, viiich could cause a non-cix)ss linked type of chain growth, was proposed. The following scheme illustrates the possible mociianlsm, (Z»= initiator, R = saturated group).

PAGE 82

^CIIaCH=CiIa (1) Z»+ Rz^ H H 75 ^ CHbC-C-Z -> ^"^N^ * Initiat ion ^GH8c:i=cris ;HoCIi=CiIfl (2) B« H H HsC-C-Z • H RaM H H CiiaC-C-Z HsC-CHb LntraiiiolecU " lar Grovxth (3) RalJ H H 'CH3C-C-Z CHgC-Cha H H CHeC-C-Z + lie HaCH=CIio HaCHjfCHs CIIa=CHCH H • C-C-Cii H H IntermolecU " lar Groxrth H ^^ CHaC-CIiaZ CiIa~CHCiia +/ \ "^"^^ + Intramole cular \ V la HHs Growth .CIia-CHCila .CIIflCH-CiiaZ etc.
PAGE 83

76 As a possible approach to the atnicture of those poly-allylqxiaternery ajtimonlum derivatives, a study of the reactions of suitable monomers and the degradation of the polyTiors was attempted. 1, Degradation of poly-tetraallylaramoniiira bromide A deconposition of the poly-quatez'nary eEimoniura hydroxide, leadiix^ to sinpler products vdiich could be identified and possibly associated with a ring type structure, was carried out. Since a quantity of poly-tetraallyl armionium broxnide was on hand frora the previous suspension polymerization trials, this product vias converted to the hydroxide form and degraded. Seventy-eight graras of poly-tetraallyl ammonium bromide was soaked in 2^0 ml, of i0o Ha OH solution, filtered, and v;aaiied with distilled w-ater. This process was carried out once a day for 28 day-s until the tost for bromide ion was less than 10 parts per million. The product was ;;ashod free of hydroxyl ions and filtered as dry as possible before it vxas pla ced in a vacuum desiccator, Tho product remained in the vacuujn desiccator at a pressure less tlian 1 mm. until it x^eached a constant weight. This took one day and the product v/eighed 62.5 g« J the calculated weight of hydrtjxide foKa should be 60.2 g. frora the initial 78 g. of bromide form.

PAGE 84

77 The converted product was pieced in a distilling pot connected to a receiver iiTnaersed in a Dev/ar flask containing Dry Ice and acetone, A second trap cooled in a slrailar manner was connected in series, The product was heated at 3OO-50® for four hours v;hile the pressure was maintained at 10 mm, Tho contents of the flask became a tarry residue. A total of iLuO ml, of a two layer product was obtained. Further heating for an hour at 350** and less then 1 mm., gave no further product, A total of 12,8 g, (li^.O ml.) of product was obtained; 5*0 ml* of a black upper layer and 9,0 ml, of a light yellow layer. The black upper layer vms redistllledi 1,1 rdl, of liquid boiling at 60-8oV760 ram. and liaviiig an n®^1,1|620 was obtained; 1,5 ml. of very dark liquid boiliiig at 6O-II0/0, 05 mia, and having an n^) 1,51 was obtained (the liquid was too opaque to give an accurate n£^)j some tarry residue v;as left in the distilling pot. Both fractions turned darker upon standing and the higher boiling fraction became quite tarry. Both fractions gave a positive test for nitrogen. An infrared spectrum was obtained for each fraction, but no clues to stanxctural features Xifere evident, A.ttempts to prepare derivatives of those fractions failed, A quantity of the residue left in tho original distilling pot was extracted v;ith etiianol in a Soxlalet extractor for one day and the alcohol solution evaporated.

PAGE 85

70 A residue was obtained t^icli upon ftirther rocrystallizatlon ylolded 0»2 c» o^ white solid. This product decomposed over a Ions rang ,3 and was not conpletoly decomposed at 360'*, Further attempts to Identify these pi-cducts ^^rere abandoned, 2» Degradation of poly-diallyldiothyl awnonium bromide Tt^enty drops of 60^ t-butylhydi^ope iKJicide (ap2:!roxi~ mately 0,012 g,/drop) was added to a solution of 8»0 g, of diallyldiothyl anmonium bromide and Ll#0 ml, of x^iiter. The mixtupo was allowed to remain open to the atmosphere in an oven at 60** for UQ hours, The resi:.lting white hygroscopic product was graind to a fine powder and dried for several days in a vacmwi desiccator. Eight grams of product, melting with considerable decomposition at 3li.6-5^r°> was obtained. The product was quite soluble in water and etiianol and gave an immediate halogen test viien treated with AgllOg solution. The product was converted to tiis hydroxide foim in an attempt to degrade it. It was bolieved tliat the loss of ethylene and water from the hydroxide form would result in a polytertiary amine or simpler products that could be studied. Seven and one-tenth grams of poly-dlallyldlethyl ammoni'um bromide, dissolved in 200 ml, of water, was allowed to pass through an ion-exchsnge column of Nalclte SAR (con-

PAGE 86

79 vorted to the hydroxide foiro), Tlie solution was recycled threo tines and tested for bronlde ion; aTtsr the third cycle, the test was loss thaii ten parts per million* The poly-dlallyldl ethyl annonixin hydroxide solution sho^^red a faint pink color when tested with phono Iphthale in j the pH value obtained from a Beclcrian r>lL ineter fluctuated considerably and vsxs taken as 11, 75 to 12.05, The solution was evaporated to one-fourth of the original volur.ie on a hot plate and to dryness on a stecan bath. The tar'r*y residue weighed approxiiuatelj five grtjis. Tlie product v/ac water insoluble a:iid fomied a gol when heated with water. The pr-oduct was very slightly soluble irx ethyl alcohol, carbon disulfide, dimethyl formaraide, carbon tetracliloride and Ciiloroforiri; insoluble in benzene and dieth7/l ether. There Xias considerable doubt as to whether the product Xiras soluble to any extent in these solvents, or whvothor a siaall quantity of product formed a Gol. The tarry residue was digested in carbon tetrachloride end yielded a product irtiiich could bo ground into a fine pox'jdor after vacuuin drying. The product decomposed ovor a wide range find xi?as not completely decomposed at 360**. Analysis found. C, [.9.I, Ii.9.5/^> ii, Q.G9, 3.23^; ii, 7.65/i^. Infrared spectra of the product were obtained; the tecimlque of depositing a film from ethanol, carbon tetrachloride

PAGE 87

30 or carbon disiilflde solutions was esaployed; the teclmique of incorporatinga snail quantity sample in KBr powder (-JO) and pressing a salt plate was also usod. R. 3ilas -^^ ld.ndl7 carried otxt this procedure. V/hen bromine (3/? in CCI4,) was added to a niixture of product in hot CCI4, en oran^ic precipitate was foriaed. Thie addition product was filtered, dried a.id ground into a fine powder. The product deeoiaposed over a wide ranee end was not completely decomposed at 360**. Analysis found: Br, $3*6^* The addition of bromine and the absorption band at G,10u, indicate so-vie degree of unsatui^ation of the polytertiary amine residue, Colthup^ ^ ' lists the strstchiiig frequency range of an unconjxigated cai^bon-carbon double bond from 6,06 to G»2$u, It was hoped that a molecular weight of the original polyquaternary ammoni-um bromide could be approximated from a molecular v/eight determination of the poly-amine. Since extreme difficulties arc encountered in molecule r weight deterrainations of polyolectrolytes, a molecular weight of the poly-amine could be used to deduce a nolecular weight of the original polyquaternary derivative by assuninr an overall loss of GaHgBr from the polyquaternary ammonium bromide. A determination of this type would pi^ove very useful, since a value of the constant in the Staudinger

PAGE 88

81 ©qtiatlon, relating viscosity end moleculrr weight, could be calculated and further molactilor weiglit detorminations of poly-qiieternary products could be aporoxlrnated. However, all atteaipts to determine the molecular weight of the polyamine proved unsuccessful, Tho boilinr: point si ovation and freezing point depression methods were restricted by the insolubility of the product and the tsmper-atur© differential* I'he following solvents txere tried and proved unsatisfactory: vnat'^r, ethrnol, t-butenol, cyclohoxanol, trlethanol amine, diethanol aiainc, nonamothylene glycol, camphor, napthaleno, benzene, cyclohexene, glacial acetic acid, nitrobenzene, 2,lj.,6-tribromoaniline, bromobenzene, and ethylene dibroraide* 3, Polymerization of diallyl amine hydrochloride By polymer izin,:^ diallyl amine hydrochloride and treating the polymer v;ith KaOH solution, it was thought tlj^t a poly-amino, similar to the poly-amdne In the previous section, could be obtained and st\idied. Ten grams of diallyl amino hydrochloride (rocrystallized from acetone-ethanol three times j m.p. 161|5*> was dissolved in 5 '^» of water and treated with 25 drops of Gcf. t-fcutyl hydroperoxide solution (0,012 g./ drop). The mixture remained in an open beaker at 60® for three days* Ten grams of product was obtained, Ihe product was insoluble in v;ator and formed a gel. The

PAGE 89

82 product was treated x^ith ethanol but was Insolublo, The wash etlianol was treated vdth acetone and only a faint cloudiness was obtained which indicated tiiere was very little unreacted diallyl amine hydrochloride present. After soaking in an acetone-ethanol mixture or picking up moistuPG from the atmosphere, the product exhibited elastic properties. Tiie product did not react with waOH solution to give a poly-arains, as v;as hoped. However, the properties of this polymer definitely indicate sorB degree of crosslinking. This v;as the only monomer in this study ;-;ith two allyl groups to show characteristics of cix)ss -linking upon polymerization. ll-. Polymerization of IjIi-bisCdiethylallylarnmoniian) butone-2 dibromide (hi) Since Butler and Goette ' iiave indicated the butene-2 double bond in l,l.j.-bis(diothylallylarar.ionii:!m) butenG-2 dibromide did not enter into the polymerization, oxidation of those butene-2 double bond was considered as a point of attack for degradation studies. Although the polyjier was made and some of its characteristics studied, furtlier degradation studios were not carried out. The polymer was prepared from 16,0 g, of l,Ii.-bis{diethylallylamnonium)butcne-2 dibromide and i|5 drops of 60% t-butyl hydroperoxide dissolved in 9 ml, of water. After 5 days at 60** in an open beaker, 18 g, of polyiner was ob-

PAGE 90

83 talned. It was soluble In v/ator and otlianol. It vaa rocrystalllzod from 100 ml. of otlianol by addl:ig 200 nl. of $0^0 acetone-dloxane solution. The residue vras vory viscous; after washing, decanting and drying in a vacuun desiccator, a hygroscopic product melting at ^k^-^k" (with considorabl© decomposition) was obtained, 5. Attempted polymor'ir,ations of monomers with the double bond farther removed than the allyl position As was shown in the proposed mechanism involving intramolecular-intei'molecular polymerization, the allyl groups offer an ideal sltuBtion for the formation of strain-free six-membered rings* It was hoped that preparation of monomers containing the double bond fartiier removed from the nitrogen center would elimincite the possibility of forming a six-membered ring structure. If the intramolecular-intermolecular type of growth was necesscry, monomers of this type would require formation of rings larger than six-membered and, consequently, the probability of their formation would bo greatly decreased. In most cases, the attempts to prepare monomers of this type were unsuccessful and polymerization studies could not be perforr'ied. In the following attempted reactions, it ivas difficult to determine whether polymerization had occurred.

PAGE 91

8lt a ) Attempted pol-pioi-lzatlon of d.1.ethylalX?fl but; ; ene-3--y 3, ^irmnon3,iAn bj:^mlde ,— A mixt^ire of 1.15 C» of diethylallylbutene-3-yl RMmonlum bromide, 3 drops of 60^ t-butylhydroperoxide soltition and 1,0 ial» of water was reacted at 60° for throe da^'^s. A brovm solid was obtained which was soluble in water. b) Attempted polymerization of diethylbut^^ne~3*yl penten-lj.--vl antnonium bromide . —A mixture of 0,5 g» oi" the vory hygroscopic soml-solid diethylbuton->3-'yl penten-[;-yl ammoniiim bromide, 5 dixjps of GO^ t-bi'.tylhydroperoxide solution, and 3»0 ^1» of water was reacted at 60° for five days. A dark viscons liquid was obtained; it was soluble in water and etlianol* c) Attempted Dolymerization of ^iallyltetraethyl decamethYlfine dismmoniurn dibroriide .--A mixture of 3,0 g. of diallyltotraethyldecomethyleno diai';xraorJ.um dibromide, 6 drops of 60^ t-butylliydro per oxide, and 1.5 iti1» of water was reacted at 60° for three days. A light yellow solid was obtained which v/as soluble in water. If polymerisation occiscred in the above cases, it is assumed thet the products were not cross-linl^ed since the products were water-solvible, 6, Attempted infrared studies Since it was postuJLated that an alternating intramolecularintermol ocular chain reaction could cause a

PAGE 92

85 plperidino riag structure tloroughout the polyraor, an infrared study of related structures was carried out» ^n this work hctrocylic ring structures, amines, amino salts, quaternaiy derivatives and many of the polymoiic derivatives discussed in tiais section woro studied. I-lany of the con5)oiinds i^rero obtained conraercially and the otxiers were prepared* Spectra of the follo;;ing compounds were obtal^ied: pipexldine^ 2-raothylpiperidinG, 2,3-diniethylpiperidine, l,2-dipiperidinoetha:i£, 1,2-diiaorpholinoethane, diallylpiporazine, pyrrolidine, N^allylpyrrolidine, diallylaniiiie, triallylerairie, dimethylallylamlne, dlcthylallylamine and amine salts and quaternary derivatives of sone of thece compounds* However, all the attfflipts proved unsuccessful, since a ring struct-ui'c could neither be identified from the spectra of the above compourds nor correlated to tiie spectra of the polymeric derivatives, 7, Solvent effect on polynor-ization The possibility of solvent interaction during the pol^erization must be considered as a factor affecting the nature of the polymer. If a growing chain-freo rcdical could react with a solvent radical in preference to a monomer molecule, the degree of polymerization would be reduced, Tlius, it might be possible to have a monomer molecule containing several unsaturated groups and yet obtain a polymer

PAGE 93

86 of a low degree of polymerisation and slljTht cross -linking. The following experiments v;ere carried out V£3jrylng the solvent and solvent concer tretlon. a) A mixttiro of 3»^'^ 3« of triallylbutyl amraoniuia bromide, 3»0 ral# of water, and 6 drops of 60^ t-biitylhydroporoxido solution was reacted at 60* for 76 hours, A liard, water-soluble solid ;mB obtained. The product was dissolved in 10 ml. of water; 10 drops of initiator was added and the iiixtixce VTP.s reacted for an additional x-ieek at 60*'# The product vrco water-soluble* b) Two grroTis of triallylbutyl aznnoniuni bromide, L drops of initiator, and 2.0 ml* of v/ater were reacted at 100° for 38 hours. The product was water-soluble, c) Tv;o grans of triall^flbutylarrmxoniuui bixsnidc, 2 drops of initiator, and 0, 08 ml, of Xirate?. were reacted at 100° for 38 hours. The reciting hard, glossy product was soaked in hot water, filtered, and dried. One and sevon-tentlis grans iQ^% yield) of waterinsoluble product v;as obtained. The procodure was that of Bunch, '^"^^ who reported a ^Z,$% yiold, d) Three grams of diallyldi ethyl aiTrnionium bromide, 10 drops of initiator, and 0,3 ml, of vrator were reacted at 100° for 38 hours. The product was vreiter-soluble, e) Three rTQm& of dlallyldiethyl ammonium bromide, 10 drops of initiator, and $0 drops of dimethyl foiwamide were reacted at 100° for 38 hours. The pi'oduct vjas water-

PAGE 94

87 soluble. 8. Inltintop effect on polymerization The effect of other initiators, t/ith respect to t-butylhydroperoxide , was determined in tlie following QXt>erinents. Various initiators were tried with monomers containi:ir: one, two, and tiireo d cable bonds to seo if t-butylhydr ©peroxide is unique in its property of yielding non-cross linl^ed polyiTcrs v;ith monoxncrs containing two allyl groups* On© gram oacJi of trlethylallyl ammonium bromide, diallyldiothyl ammoniimi bromide and l,l{.-bi8(diethylallyla2TJnoniujn)b\iten0-2 dibromide were dissolved in 1,0 ml. of T>rat0r and reacted for one week at 60*» with 0,0^ c» each of t-butylhydroperoxide, di-t-butylperoxide, and benzoylperoxide. The lack of cross-linki n was indicated in every case by the water-solubility of the prodticts. Polymerization took place only in the respective reactions of diallyldiothyl arrmonium bromide and l,[!.-bis (dlethylallylsmmonium) butene-2 dibromide with t-butylhydropero cido. In the other cases, the starting monomers were recovered and identified. In an attempt to evaluate the initiating property of 2,2'-a2oisobutryonitrilo, the aso initiator was reacted with hoxaallylothylene diamrtioni-um dibromido, since it v;as thought tiiat an easily isolated, insoluble product would

PAGE 95

&6 be obtaiiiad IT pol^erizatioii occupi^ou. Eio laonoiaor, dissolved in 10 iiil, of dimethyl f orisaiiido was allowed tc react foi' two waeka at 60° in azi open vessel, TAHLF. VII AZO IKITIATED POL^fMERIZATION OP HEXAALLYLETHYLEUE DIA^Ii402JII]M DIB^.OHJDE Weirht of Uol-ht of Percent •t. of HaO Monomer Initiator Initiator Insol-uble ilemaiks Prod ct 5,00 g. 0,013 g. 0,1 Soluble Light viscous llq. after ttro weeks 5,00 g, 0,025 g» 0,5 Soluble Light viscous liq. after two v;o :ks 5,00 g, 0.050 g. 1,0 Soluble Viscous liq. after tiJO i-jenks 5,00 ,r, 0,10 g, 2,0 0«75 g. Heavy vlsco\ia liq. after 2 weeks 5.00 g. 0,25 g. 5,0 3.28 g. Gelatinous after 12-15 lu?. Flexible solid 2 wks, 5,00 g. 0,50 g. 10.0 i^37 g, Gelfetlnous after 6 hr. Brittle solid at 2 wks./ The initletin,'^ effect of 2,2'-azoisobutryonitrile upon the polymerization of diallyldlethyl ainraoniuiii bix>nide was doterraiiied in a similD.r raamier* Tvio grains of diallyldlethyl cmraonlum bromide (n,p, 155**) dissolved in 10 ml, of dlraethylfoiY/iainide vms allowed to react with varying amounts

PAGE 96

89 of azo initiator at 7^** for flTtoon dcys. The pioduct was water sorable in ©very case; the determinatloai of the aaount of polymerization was ratixor difficult* Uiireacted rionomer was separated from tiie i-*eactioa product by extractloa with hot acetono. Attempt od rccryatallization fi'om acetone and ethanol resulted in gol formation* TABLE VIII AZO liJITI/iTED POLYI-iaHISATION OF DIAI.LYLDIETHYL AmONIUM ][iiiOIIIDE Wt, of Wt. of Monoiner Initiator % Initiator Remarks 2,00 6, 2,00 gi 0.02 g« 0.10 g. 2*00 g. 0.20 g* 1.0 5.0 10.0 Water soluble j portion insoluble in hot acetone melted at 295-300° Water soluble; melted at 305-7° Water soluble; melted at 306-7*' 9. Effect of oxygen in the polymerization It was considered that oxygen could exhibit so:r.o inhibitir^(^ factor in the polymerization and contrib\:.te to the fact that soluble and non-croc slinlcod polymers were obtained with monomers containing two allyl groups. The following eicperiments were porformod to determine tl:ie effect of oxygen.

PAGE 97

90 a) A mixture o£ 6,0 c» o£ diallyldl ethyl ainnionluiri bjTOiixide (la.p, 155" )» k*^ »il» o^ water, and y^ arops of 6Q^ t-butylhydi'operoxlde solution (appro-dinately 0.012 g./drop) was allowed to react for 76 hoiirs at 60-65°» Kie reaction was carrloa out in a nitrogen atmosphere, Hitrogen was passed through a combustion tube filled v;ith copper turnings, ttjo wash solutions of alkaline pyrogallol, a CaSC4 drying tovver, and into the reaction vessel* 'Bio system vjas flusi^d with nitrogen before the reaction was started. Eight grains of hydroscopic px'oduct was obtained after dryirg for several days in a vacuum desiccator at less than 1 mm, pressure. The product decomposed over a wide range and melted with decor.rposition at 355-60°, The product was soluble in water but was relatively insoluble in ethanol, forming some gol. The degree of solubility of this pi'oduct in ethanol differed from the solubility of polydiallyldiethyl ammonium bromide, viaich vas prepared open to the atmosphere, A determination to show any possible difference in the analysis calculated to include complete catalyst incorporation in the polymer, and the actual anal^/sis of the polymer, was made. Since the conditions prevented the reaction of atmospheric oxygen, the 02ily oxygen present in liie polymer would be contributed by the catalyst or from solvcait intex'action by the OH radical.

PAGE 98

91 Calcu3,atlons Moles of catalyst = 33 drops x 0.6 x 0,012 c»/«irop x 1 mole = 0. 0026I|. mole = 0,03l| mole rr,r.-,r. r««„«^,^.v, (assijmlnc com0.03Umole monoraer ^^ 12.9 t "^o^^o^°^ pleto catalyst Molos of monomer = 8.0 g, x 90 a* 1 mole 231^.2 6. 0.0026k nolo catalyst molo catalyst incorporation in polymer) TABLE IX ANALYSIS OP POLY-DrLLYLDIETIiyL AILIOIIIUI^I BROMIDE Itonomer (theory)

PAGE 99

92 sure atid was carried out according to the modification of Parkin/ ^^^ A woislied quantity (approximately 0,06 g») ot platinum oxido and 20 ml, of distilled water was placed into a hydrogenation flask equipped with a magnetic stirrer and a rubber-tipped side arm. After the system was evacuated and flushed v;ith hydrof^en throe times, the magnetic stirrer vxas started. After complete absorption of hydrogen by the platinum oxide, the stirring was stopped and a weighed sainple, dissolved in 3,00 ml, of water, was introduced through the rubber tip frcra a calibrated hypodermic syringe. At this point, the pressure was equalised to compensate for the addition of the 3,00 ml, sariiple, ffiic stirring was started and the reaction continued until further absorption of hydrogen had ceased. Since the stirring meclianism caused some heat to be generated, the S3'stem was a.llcwod to come to equilibrium before measurements were made. The following table gives a summary of the data. Since all the attempts to rocrystalllze the polymer sample proved unsuccessful, the crude sample was hydrogenated after careful drying, Hov/over, a later method proved successful in extracting any monomer left in the crude polymer sample and a value of pure polymer fraction was determined. With this correction, an average value of 0,193 moles of hydrogen absorbed per mole of monomer unit charged

PAGE 100

93 H H 9 B cq H !^ O 9 ft o H Ch 8 •d H O o o • • O r^ • • OJ vO CM vO O P P d O O Eh o o o o o t5 •H o +3 H O H O O vO lA O • • o o H O • • o o • • o o lA O CM O H I o ^ -7 & H O ft CQ O H w -p H o o •H P O CM O O o o o lA C^ CM O O o o • o CM CO o « o o o C3 -P •H © g i C3 o H i H o § •H •P O CJ U Is Q H O Oh lA o o o o • o VA o o o o • o CO o o o o • o H o 1A H CM • ^A H vO O o o • o o H o o o O o o o tA H O o o • o o 1A H O o o o vO H o o o • o o •H •P O a Js is o i o X> O C3 C3 O H O O o o • o c> o o o • o o o -P o cd ;^
PAGE 101

9k gives a ratio of one free dot^blo bond for every five monolaer tmiits (diallyldiethyl ammoniiirn bromide) incoi'poratod In the polymer chain, b) In order to justify the previous aasuiaptlon that solvont interaction involving an Oil radical did not occiir, several expericients were niade using accurately weighed quantities of monomer and catalyst. T'o.eso determinations were candied out like tiie previous e^qperiments; in one case the polymerization was carried out open to the atnosphei'e and in the other case under an atmosphere of purified nitrogen. The product was carefully dried at 60** for sevei'-al hoiirc, dried in a vacuuiu desiccator and weighed to a construit weight, Tlie follov;ing table gives a suimiary of the results. TABLE XI COriPAr-ISClI op POLYinSRIZATIOIIS OPEIJ TO ATllOSPIIEiffi MD miDER Na Open to Atm, Under II a Vit. of diallyldiethyl ammonium „ r.,-./, o r'n'ir, bromide (g.) ^'^"^^^ ^'^279 v;t. of 33 drops of t-butylliydror, ,-ot ^ ^ .'wi peroxide (60?^ soln. ) (g, ) 0'^8l^9 0.3i|6l}. V/t. of product (g. ) 8.0606 8.6317 Wt. increase (g. ) 0,l$l^5 0,1038 Theoretical vjt, increase assur-iing com^ ocfnn n noRn plote catalyst incorporation (g. ) ^Oi?!" 0,3^00

PAGE 102

95 The actL\al Incroaso of weight on polymerization, shov.n In Table XI, Indicates that Oil radical Interaction from tlio solvent can be neglected. It was also confiidorod tliat the formation of pe-o:?:lde or hydroperoxide linkacos could acco-unt for the utilization of some of the monomei' double baads during the polymerization. However, Infrai'od spectra of the above products showed no absorption bands at 870 and 8i|.0 cm"-'-, i,foich are cliaracteristlc of peroxide ^'-^-^ and hydroperoxide,^'^' respectively, Polarographic determinations ^^'^^ of poly-dlallyldiethyl ammonium broEildo (prepared open to the atmosphorG and under nitron on) and poly-l,L|.-bis(diethylallylan3monlum)but6no-2 dibromide gave no indication of a peroxide or hydroperojcldo wave, A rataer crude approximation of the degree of polymerization can bo made from a consideration of Tables XI and XII and t:io followiiTg assumptions: (1) Only the tbutoxido radical is to be considered as an initiating frafjDient, This assumption may be valid, considering some of the work with tlie decomposition of t-butylhydi'operoxide in the gaseous and liquid phase, ^ "^ In all cases citod, fission of the 0-0 linkar^n to produce t-butoxlde and hydt'oxyl radicals was propos d to account for the formation of tho reaction products, (2) The increase in weight during polymeiization is duo only to incori:)oration

PAGE 103

96 of initiator fragment 3, (3) The means of chain termination is either by combination or by disproportionation. Hie polymerization reaction products were freed from any tinreacted monomer by extraction xdth hot acetone, Sangjles of the reaction products vere extracted in a fin© filter funnel until a constant weight was obtained upon vacuum drying. The fraction converted to polymer was found to be 9U«9^ and 96»9% respectively for the samples prepared open to the atmosphere and londer nltror;en. The followinr; table gives a siirinary of the results, C, D iscu ssion of Results Tlie investigation of the mechanism of the poljnaerization of allyl quaternary derivatives iias not resulted in any definite evidence to prove oidisprove the intramolecularintemolecular chain reaction. The degradation of poly-tctraallyl ammonium bromide and poly-diallyldiethyl aiiimoniura bromide did not yield any products v/iiich could be identified orassociated wi"h a piperidino ring atrnctur-e. An infrared study of pipei idine compounds and related struictiufes did not reveal any similarities vihich could bo correlated with the spectra of ti:ie polymers. Although some unsaturated tertiary amines were prepared with the unsaturation located farther from the nitroren than the allyl position, crystalline derivatives of these tertiary amine c could not be obtained. It was

PAGE 104

97 •TABLE XII APPROXIIIATION CF THE DEGHEE OP POLYI-ERIZATIOII Open to Atn, Under Ilg Wt» increaao (assuming only iniatop frafment incoppoi-atlon Moles of radical (assuming only t-butoxido radical as initiator) Start iix^ wt, of monoradr Fraction co:ivorted to polymer Polymer fraction e:xpressed as laolcs of mononcT Iloles of monomor per laole of radical initiator Degree of polynerization (assuiiLing chain tepiuiimtion by disproporticnation) Dogroe of pol^erisatlon (ass\a:ninr; cluiiii tci"^iliiatlon by combino.tioii) hoped tliat qimternary dorlvativos of thoso tertiary amines coiild bo formed and polymeria edj the position of the double bond, beir:g farther from the nitrogen tirnn tho allyl position, woTJld prevent intramolecular plperidine ring croij-th during polymorization. The only crystalline derivative obtained was diothylallylbuton-3*'7l ammonium bromide; the polytnerizatlon product of this derivative was water soluble. 0.1^11.5 g#

PAGE 105

9Q It is likely tixat tiio solvent does afreet tiae polymerisation, siiic
PAGE 106

99 It was found that t-butylhydropero:dLdo or 2,2'azoisobntryonitrilo v;ould initiate the polymorizatlon of allyl quatarnai'y aimnoniun dei'ivativoc. Table VII indicates that the amount of effective azo initiator is betwoon 5 and 10 percent. Tliis is hirher than the effective initiator ratio of t-butylhydroper oxide, The polyineri:!;ation appr^ai'Gd to bo soraox/hat sensitive to o::ygon, since the deci'oe of pojymerisation of tho product polyinerized iinder nitrogen vjas higher than the corresponding product prepared open to the atmosphere; this greater decz'Oo of polymerisation v/as also noticed by the different solubilitios in ethanol. Chemical, infrared and polarographic ancJ-jsec did not show the presenco of poroxido linlcagos and it was, ther-efore, concluded that oxygen Interaction ims not by the forroation of any poioxide product* It is likely that oxygen affects the rate of initiation rather than the formation of different polymer products. The polytierprovi'icts shoifed some unsaturation by their reaction xviih a bromino solution and by their infrared absorbtion ab o.lOn* luantitatlvo h;;'dro,2:enation oxperliaents, as summarized in Table X, restilted in a value of 0,193 moles of hydrogen absorbed per mole of monomer •unit cimrgod, indicating that one free double bond v;as left for every five monomer imits in the polyiuQ-' chain, Since

PAGE 107

100 the monomer, dlallyldl ethyl ammoniumi bromide, would contain a total of ten double bonds for five units and since one double bond is loft in the polymer, nine doublo bonds must linve beon utilized in the polyraerization reaction. Linear Intermolecular chain growth would liavo used fivo double bonds per five monomer units, leaving four doublo bonds unaccounted for in the polymerisation, Simpson, Holt and Zeite^^-^' havo shovm that at the gel point lii the polymerisation of diallyl phthalate much of the caj:'bon-carbon doublo bond supposedly free to cross linlc was tied up as a cylic structtire due to intraraolocular polymerization. Therefore, it is conceivable that much of double bond in the polymerization of diallyl quaternary ammonium derivatives could be tied up as an intra:nolccular cylic structure and thus account for the apparent lack of cross-linldng. Since the soluble polymer products were polyelectrolytes, a molecular weight determination of the products, as such, was very difficult. The usual molecular weight determinations, which are dependent on the colligative property of polymermolecules In solution, were not usef^ll since the degree of icnic dissociation of the polyelectrolytes vms unknovjn. An attgnpt was made to convert poly-diallyldiothyl ammonium bromide to tli© poly-hydroxide form, degrade this product by the loss of ethylene and water, and measure the molecular weight

PAGE 108

101 of the rosultlnc poly-tertlary amine* Althoiogh the polyeloctrolyte v/as degraded, a molecular v:eight determination of the product obtained was unsuccossrixl» Since this method failed, a crude approximation of the molecular weight v;ac obtaliiod by polymerizing accurately velched amounts of monomer and catalyst, determining weight increments, oxtractiii^ unreacted raonomor', and by various assi'jnptions arriving at a degree of polymerization* By as3ural:cTg that the chain tcrrainrtion process is by dis« proportionation or cor.ibination, the data in Table XIj indicates the degree of polymerization is from If^ to 30 (product polymerized open to atmosphere) and 2$ to $0 (product polymerized under nitrogen). Shis lovj degreo of polymerizations agrees with the (51) results of Bartlott and Altschul viho found the degree of polymerizatiai of allyl acetate to bo from I3 to llj. and attributed this lou degree of polymerisation to hydrogen atom transfer* The authd-s'^^ considered thct during polymerization of allyl acetate, the kinetic chain was terxTiinfitod by the transfer of an o( -hydrogen atom from a monomer to n growing free radical. This o( -hydrogen transfer from a monomer raoleculG would have resulted in the foiiaation of a radical sufficiently stabilized by resonance, by virtu© of its allylic structui-'o, to resist reaction vri.th a monomer long enough to caabino with another radical either of its

PAGE 109

102 ovm or of the clieln-propacatirif; kind. This reaction, (52) tonnod "degradative chain transfer" by Bartlott, v;onld havo caused the low degree of polymerization. Confirmation of these conclt^sions, based on kinetic work, i.'as atteiroted by i3artlett and Tato. -^-^ Allyl acetate, having the 0( -hydrogen replaced by deuterium, was polymerized. It vjas expected tiiat deuteration of tlie CX -hydr og en of the allyl group would affect only the chain terminating step, if OC -hydrogen transfer was involved in that stop. Since previous v.-orkors '' ' had shovm that the rate of proton transfer was four to ten tines as great as deuterium transfer, it v.-as thought that deuteration of the oC -position would increase the kinetic chain length by slower deuterium transfer. The average degree of polymerisation v/as found to be 2,38 times as great for the douteratcd as for the undeutoratod polymer, thus conf iiroing the conclusion, previously based on kinetic grounds, that the chain terminating stop in the polyTioriEation of allyl acetate involves the transfer of an 0< -hydrogen from a monomer to a growing free radical. Consideration of the cylic intram.olecular polymeri(33) zation of diallyl phtlmlate by Simpson, holt and Zeite and the degradative chain transfer in the polymorization of (^1-53) allyl acetate by Bartlett and covrorkers "^ loads to the belief that the polymerization or allyl quatornary derivatives is affected by both factors, Cylic intramolecular

PAGE 110

103 polyMGrization, greatly Influcncod by chr.in tcmirmtlon causod byOC-hydropen tranafer, would acco-juit for tho solublo, non-ci-csslinlced, low raoleciaai' wclclit and uncat'orated polymers obtained in this study.

PAGE 111

VIII SUMiAHY A series of unsaturatGd toptlary diamines of the type (CHs=CIICHa)elI{CIia)^N(CHeCI^CHs)a was prepared and characterized. For those conipotinds tidiore n = 2, 3 or 6-10, acceptable yields were obtaliiod by the reaction of the corresponding l,n-dlhalo or di (arylsuironoxy)alkane with diallyl amine* Tliis method was imsuccessful for those compotinds where n = h and 5> i^ the presence of an excess of diallyl amiiie, tine products of reaction were allyl pyrrolidine end triellyl amine from the l,I|--deriva» tives, and allyl plporidino and triallyl amine frora the l,5-derivntives« The products of the reaction wore erplair^d on tlie basis of an intramolecular cyclization to the five or six membored cyclic quaternary ammonium salt, followed by allylatlon of the excess diallyl amine by the quateinery ammonium salt to produce triallyl amine and the appropriate allyl substituted heterocylic amine. The reaction products were identified by chemical means vnd by infrared ^octral analysis, The desired l,li and 1,5 tertiary diamines were obtained by the reduction of the corresponding amides vjith lithium alumlnutn h3rdride« 10k

PAGE 112

105 The raethyl and tlio allyl quatoi^nary urtimoni-uia dorivativoc of til© tci'tlox-j diamines x^gpg prepared in good yields. A modified suspension polijiaorization was uood to polyniorizo tho quaternary ainuoniuii salts, Usinc t-butylhydx'opero.:ido as an initisitor and a mixture of wateiand mineral oil as a reaction medium, watei-insoluble resins possessing anion excban^je properties x^ere obtained in good yields, A rapid titration method of doteimining tho exchange capacity of the rosins x.^as dovslopod ns.king use of a Beclfflian Model K automatic titrator. This method kept tho time and tiie number of samples at a minimutu. Trie ©xclmneo capacity xvas expressed as the fraction of theoretical exchange per hourj a maximum of ©xchanG© per unit tirao was observed at the pentajiiethyleno derivatives of both the allyl mid the methyl qxiatcrnary polymer series. There was no definite correlation between the swelling coefficient and tho exchange capacity of the resins, A study of the polymeriza-^ion mechanism of allyl quaternary ammonium derivatives did not result in any definite evidence as to structxiral configuration of the polymoi'S, Synthetic variations of monomer structiu^e, degradation studies, chemical and infrared analyses, and attempted molocul^ : weight detoiminations were used to investigate the poljmc:ization. It is believed that a

PAGE 113

106 conslderablo araoxjjit of cylic intrainolecxiltir poljTierization, groatly ax fee ted by cliain termination cav^jod by oC -iiydi'Of^on transfep, xv-ould accotmt for the soluble, lot: molecular woight, unsaturatod polyjiers obtciinod in this study.

PAGE 114

BIBLIOGHAPin 1. Butlor, C-, B,, and Bunch, H, L, , J. Am. Ch ein^ Goc. 71. 3120 (I9i;9). 2. Btitlor, G, B., and In^loy, P. L,, J. /m. Ch em. Soc. 73 « 89^ (1951). 3. Biitlcr, G, B, , Bunch, R« L., and InglGy, P# L, , J. Am . Chor.1Soc. 7L. . 25li3 (1952). l|.. Butler, G. B., aiid Goette, R. L., J. Am. Ch en. Soc. 7L. 1939 (1952); 16, 21:18 (195^1 ). 5, Ancelo, R. J., 1-i.S. Theals , University of Florida, 1953* pp. 32-7« 6, Huaa, v;, J., Fh.D. Maaertation , University of Florida, 1953, PPk^-"^* 7, Pu-oss, R. M., and Gathers, G. I., J. Polmer ^ci. h , 106 (I9li9). 8, PiercG, Iv, G,, and Haonish, ri, I., Qimntitative Analysis , Joiin Viiley end Sons, Inc., How York, H.Y. , 191.^6, ?. 302. 9, Ancelo, R. J., H^S. Thosis . University of Floi-'ida, 1953, pp. 8-9. 10, Laakso, T, M,, and Reynolds, D. D,, J. Am. Ch on. Soc. 73 » 3518 (1951). 11, AiiGolo, R» J., N.S. Tnonis , University of Florida, 1953, pp. 16-17. 12, Lanre's Handbook of CheiTiistry , Seventh iiditlon. Handbook" I^iblishers, Inc., Sanduslry, Ohio, 19U9, P. 62li.. 13, Adaraa, R., and I-Iarvol, C.S., J. I'Xi, Cl ^era. Soc, 1:2. 3I3 (1920). Uj., Blatt, A, H, , Orfyanic S-vnthesos Coll. Vol. II . JoiTn Wiley and Sons, Inc., Hew York, li.Y. , 195^, P» 537. 107

PAGE 115

108 15» Adams, R#, Organic Heactlons Vol. VI « John V/iley and Sons, Inc., Ibw York, II. Y. , 1951, Che-P. 10. 16. Blatt, A, H,, Organic SYnthoses Coll. Vo l . II , Jolm V/iloy and Sons, Inc,, Hew York, 11. Y., 195^» P» l^^* 17» Blatt, A, H«, Crnanic oyntaosos Coll. Vol, II , John V/iley and Sons, Inc., ^'ew York, N.Y. , 1950* P* 2i}.6« 18. IJuller, A., a:.cl Vane, U., I3er. 77S , 669-75 (19i!^-). 19. Joutsch, II., and v, Brcim, J., Ber. 1;6 , 230 (I913). 20. livrvoncn. A., AlUli Ac&u. ^^QJ^ Torjl. (A) IQ, llo. 5, s. 17. 21. Ilellbron, I., Dictionar" of Orr^anic Gorgpoxiads Vol. IV . O^rCord University Press, London, 1953 > P» Q« 22. .Jio:xneau, R., Ann. ChJjTi. -^ ^ 2l;2-55 (1915). 23. Ancolo, II. J., M. S. ^Iliesis, university of Florida, 1953, -oP' U-6j 10-13, 2k» Hollins, C, The Synthesis of Mitrop:en Rinn Conipounds , Ernest 3enn, Ltd., London, 1921+, p. 63. 25« Lelician, M» R. , llhonpson, C. D, , and Marvel, C» S., J. Am. Chen, , Soc^,._^, 1977 (1933). 26. Eliel, E* L. , and Peckham, P. £'. , J. Am. Chexn^ Soc. 72 ^ 1209 (1950). 27. von Braun, J., Kuhn, M,, and Goll, C., Ber ^9 . 233O (1926). 28. Angelo, R. J., M.S. Thesis. University of Florida, 1953, pp. 20-26. 29» llohenstein, W. P., and ilark, H, , J. Polvraer Sci. I j 130-iLC (I9I1-6). 30. Husa, W. J., Ph.D. Dissertation . University of Florida, 1953, :^. 109. 31. Ilusa, Iv. J., Ph.D. Dissertation . University of Florida, 1953, p. 32, ~

PAGE 116

109 32. Kulkarnl, L, G, , Private Cornrnujiicatlon, 33. Simpson, W,, Holt, T, , and Zelte, R. , J. Polytnor Scl. 10 , [189-97 (1953). 3I4. Liberxaann, C, and Paal, C, Be r, 16 . 526 (1883), 35. Atnimdsen, L, H, , J. Ain. Cliem. Soc. 7^ . 2118 (1951), 36. Goetto, R, L», Ph.D. Jlssertatloa , University of Florida, 1953, P. 19. 37. Juvala, A,, Ber. 63 . 1993-i^ (1930). 38. PricG, G» D», Private Coraniiinicatio:!, 39. Silas, R,, Private Ooimaiinlcatloi'i, bo, Golthup, H. B, , J. GDtical Soc> /^. hO , 397 (I95JO). bl, Butler, G, B, and Goette, It. L, , J. /vm, Ch.au. Sec. 76 ^ 2l!l8 (195U). k2. Bunch, R, L, , Ph, D, Dlssertstlon , University of Florida, 191^9, P. 37. t|3. Parkin, 3, A,, Private Coinmunicatlon, [jli, Philpotts, A, H, , and Thain, W, , Anal. Ghem, 2l\ , 638 (1952). I^5. Shreve, 0, D, , lieether, M. R, , Knlf^ht, H, B, , and Swern, U, , Anal. Chan. 23 . 282 (1951). 14-6. Long, G, , Private CommTinication. [}.7. Kharsch, h« S«, Fono, A,, and IJudenborg, V;,, J, Qrp. , Chem. 1^^. 763 (1950). i|.8, IQiarsch, M.S., Pono, A,, Nudenberg, W, , and Bischof, B., J. Orf^. Chaa. 17 . 207 (1952). I19. Kliarsch, M, S. , Arlinto, P, S,, and Nudenberf);, V/, , J. Orr. Chom, 16 . 1556 (1951). 50, Eell, JE, H,, Raley, J, H,, Rust, F, F,, Seubold, .'\ H,, and Vauglin, W, E, , Faraday Soc. Discussion 10 . 2L6 (1951).

PAGE 117

no 51. Bartlett, P. !), and Altschul, R. , J. Am. Chein. Spc. 67. 812-16 (19U5); ^, 016-22 (19U5)« 52. Burnett, G, M, , Mecxianism of Polyner rieactiona . Interscience Ptiblishei^s, Inc., ilew York, 2{, Y, , 195U, P» 68, 53. Bartlett, P. D, and Tato, P« A,, J. Air. Choral. ::oc. 7^ . 91-^ (1953). Sk» WynneJones, W. R, K, , J, Cheai* Phys. 2 , 38I {I93I1.). S$, Wentholmor, F, H# and Nicolaides, N, J. , J» Am» Chem » 90C f 71 > 25 {19i|9).

PAGE 118

BIOGRAPHICAL ITEMS Rudolph J. Angelo was bom Jirne 16, 1930 in Ellv;ood City, Pennsylvania, After attondinc the public schools of Ellv/ood City, he entered Geneva Collefte, Beaver Palls, Pennsylvania, in Septombor, 19U7> said graduated in June, 195l> vd-th the degree of Bachelorof Science. While as an imdergraduate he served as a st-udent assistant in the Department of Chemistry. He entered the graduate school of the University of Florida in September, 195l> and received the I-Iaster of Science degree in February, 1953» He served as a graduate assistant in the Department of Chemistry from September, 1951, to June, 19^2. From Juno, 1952, to the present time, he lias served as a rosccrch asslstait on a project supported by the Atomic Energy Commission, He is a member of Gamma Sigjna Epsilon, Phi Kappa Phi and the Society of Sigma Xi, 111

PAGE 119

This dissertation was prepared under the direction Ox the chainnan of the candidate's supervisory coimnitteo and has been approved by all members of the committee. It was submitted to the D«an 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, Jane 6, 1955 Dean, College of Arts and Sciences Dean, Graduate School SUPERVISORY COMMITTEE: y Chairman c. £:. UUP) ??. ^U^k rj^^^j. 112