poly(arylene sulfides) with pendant cyano groups as high-temperature laminating resins
TRANSCRIPT
JOURNAL OF POLYMER SCIENCE: Polymer Chemistry Edition VOL. 11. 2793-2811 (1973)
Poly(ary1ene Sulfides) with Pendant Cyan0 Groups as High-Temperature Laminating Resins
IBRAHIM HADDAD, SHAUN HURLEY, and C . S. MARVEL, Department of Chemistry, The University of
Arizona, Tucson, Arizona 85721
synopsis
Poly(pheny1ene sulfidea) were synthesized from m-benzenedithiol and aromatic dibro- midm in a basic medium of potaasium carbonate in dimethylformamide or dimethylacet- amide. The products obtained were slightly off-white with relatively low melting ranges and had inherent viscosities in the 0.2-0.4 dl/g range in hexamethylphosphoric triamide. Similar poly(pheny1ene sulfides) containing pendant cyano groups along the polymer chains were obtained by the use of 5 mole-% of either 2,Pdichlorobenmnitrile or 3,5di- chlorobenzonitrile. The products were similar to the pure polyphenyiene sulfides, efcept that they showed lower melting ranges and gave insoluble products when heated alone or in the presence of zinc chloride.
INTRODUCTION
Although many thermally stable polymers can be fabricated into fibers and films, the use of these materials as resins in laminating, molding, or coating applications often poses considerable proceasing and handling problems. Besides the required stability towards oxidative and hydrolytic degradation, polymers which are to be used as laminating resins must be low-melting or soluble in inert solvents in order to wet the surface of the coated or reinforced material and should show good adhesive properties. If the final polymeric resin is a crosslinked material, the crosslinking process should not be accompanied by the evolution of any gaseous or other volatile products, as these would cause voids that result in weak points in the final laminate or coating.
The work reported in this paper describes the preparation of polyphenyl- ene sulfides and related polymers for use as thermally stable laminating resins. Poly(pheny1ene sulfides) have been previously reported in the likrature.'-S These polymers, however, did not possess either the 801u- bility or the low melting ranges required of a laminating resin, and thus a search for a poly(pheny1ene sulfide) with the proper properties was ini- tiated. The laminating resins were obtained in the precuring stage as poly-m,p-thiophenylenes substituted with cyano groups along the polymer chains and having softening ranges around 100°C. These laminating
@ 1973 by John Wiley & Sons, Inc.
2793
N
-4
W c m 6 F
TA
BL
E I
Pr
epar
atio
n an
d Ph
ysic
al P
rope
rties
of
Mod
el C
ompo
unds
8 “0
Thio
phen
ol,
Mod
el c
ompo
und
g (m
ole)
Bro
mo
com
poun
d,
g (m
ole)
b D
MF,
K
,CO
, Ti
me,
Yie
ld,
ml
g (m
ole)
hr
%
” 10
.83
(0.0
98)
(0.1
01)
11.1
2
10.4
4 (0
.095
)
10.7
9 (0
.098
)
5.51
(0
.05)
BB
15.3
9 (0
.098
)
11.9
2 (0
.05)
11.2
1
pDB
B
mD
BB
200
15.5
5 48
35
(0
.113
)
24
86
100
16.0
3 (0
.116
)
100
15.0
8 26
70
(0
.109
)
m Ph
ysic
al
C
-m
“*
prop
ertie
sd
? bp
147
-152
/8-9
mm
Hg
mp
79-8
1 (E
t0H
)f
* T P bp
164
-170
/0.1
-0.2
m
m H
g‘
(0.0
48)
15.2
88
(0.1
13)
DB
BP
100
15.5
5 24
90
m
p 11
5-11
6 (i
Pr0H
)f
(0.0
5)
DB
N
50
7.95
24
90
m
p 10
1-10
3 (i
PrO
H)
7.16
(0
.06)
(0
.025
) w
)
24
41
mp
8-68
8 (i
Pr0H
)f
5.55
D
BE
50
8
.0
~s
*o
-@
s~
(0
.05)
8.
28
(0.0
58)
(0.0
25)
3.10
4 B
BN
60
4.
5 24
82
bp
132
-135
/0.2
mm
Hg
(0.0
28)
5.09
6 (0
.033
) (0
.028
) B
BN
40
6.
35
24
70
mp
85-8
7 (E
tOH
)h
(0.0
2)8
7.29
8 (0
.046
) (0
.04)
cd
0
F
cs a
All
reac
tions
wer
e ru
n at
150
-155
°C.
b B
rom
o co
mpo
unds
: B
B =
bro
mob
enze
ne;
pDB
B =
p-d
ibro
mob
enze
ne; m
DB
B =
m-d
ibro
mob
enze
ne; D
BN
= 1
,4-d
ibro
mon
apht
hale
ne; D
BE
=
c Pe
r ce
nt y
ield
aft
er o
ne d
istil
latio
n or
one
recr
ysta
lliza
tion.
e D
ata
of H
andb
ook
of C
hem
istry
and
Phys
ics.
6
g m
-Ben
zene
dith
iol w
as u
sed
rath
er th
an th
ioph
enol
. h
Dat
a of
Mar
vel
and
Wilb
ur.’
5 5 bi
s(p-
brom
ophe
nyl)
ethe
r; B
BN
= 4
-bro
mob
enzo
nitri
le;
DB
BP
= 4
,4’-
dibr
omob
iphe
nyl.
d M
eltin
g po
ints
are
unc
orre
cted
; Ref
eren
ce g
iven
to
liter
atur
e va
lues
alr
eady
rep
orte
d.
m
f D
ata
of C
ampb
elL6
B
z $ M
2796 HADDAD, HURLEY, AND MARVEL
resins were cured by heating at temperatures of 350-400°C or by heating to 290°C under nitrogen in the presence of zinc chloride, and it is thought that the crosslinking might proceed by the trimerization of the nitrile groups without the formation of any volatile side products. The products were determined to be crosslinked on the basis of their insolubility in hexamethylphosphoric triamide in which the noncrosslinked polymers were quite soluble and from infrared spectra that were similar to those obtained for the noncured polymers without the nitrile absorption. Conclusive evidence for the formation of the triazine molecules could not be obtained, as neither infrared nor proton NMR spectra could be effectively used in this determination.
RESULTS AND DISCUSSION
Model Reactions
It has been reported that potassium carbonate in N, N-dimethylform- amide (DMF) provided a suitable medium for the reaction between an aromatic dithiol and an aromatic dihalide to give linear polymers with sul- fide linkage^.^ However, the polymerizations which were of interest to this work were not reported. Model reactions were, therefore, carried out between thiophenol and various aromatic dibromidos in order to evaluate the feasibility of using potassium carbonate-DMF as the reaction medium for the corresponding polymerizations. It was also thought that potassium carbonate-DMF would provide a suitable medium for the reaction be- tween polyphenylene sulfides containing thiol endgroups and 4-bromo- benzonitrile to give nitrile end-capped polymers which could be used in the crosslinking experiments. Model reactions were, therefore, carried out between 4-bromobenzonitrile and thiophenol and m-benzcncdithiol.
The products of the model reactions along with the reaction conditions, yields and the respective physical properties are shown in Table I. Identi- fication of the products was achieved by infrared spectroscopy and by elemental analysis, the corresponding values of which are given in Table 11.
During the initial heating period of these reactions, a bright yellow color due to the thiophenoxy anion was formed. The intensity of this color usually decreased during the later stages of these reactions. This observa- tion was also true of the polymerization reactions described below.
As shown in Table I, the yields of all but two of the model compounds were reasonably high. This suggested that potassium carbonate-DMF should provide a suitable medium for the polymerization and endcapping reactions. Furthermore, i t appears that these reactions proceed by the expected nucleophilic aromatic substitution mechanism as, in each case, products with rearranged structures were not observed. It was, therefore, thought that the corresponding polymerizations would also proceed without positional isomerizations and thus give products of the desired structures.
POLYARYLENE SULFIDES 2797
TABLE I1 Elemental Analysis of Model Compounds
Model com-
pound8 Formula c, % H, % s, % N, % 2 Ci8HiaSz Calcd
Found 3 CiBHi4Sz Calcd
Found 4 CZ~HIBSZ Calcd
Found 5 CzzH1Sz Calcd
Found 6 CzaHi~Sz0 Calcd
Found 7 CiaH9NS Calcd
Found 8 C Z ~ H I ~ N ~ S Z Calcd
Found
73.48 73.32 73.48 73.35 77.84 77.95 76.76 76.91 74.61 73.42 73.93 74.18 69.77 69.99
4.76 4.78 4.76 4.90 4.87 5.14 4.65 4.62 4.66 4.69 4.27 4.50 3.49 3.49
21.78 21.74 21.78 21.78 17.30 17.13 18.61 18.77 16.58 16.39 15.17 6.64 15.25 6.79 18.71 8.14 18.27 8.21
8 Compound number corresponds to model compound number in Table I
Polymerization Reactions Poly(pheny1ene sulfides) with various struc-
tures were prepared in potassium carbonateDRIF, and some of their prop- erties were determined. These investigations were carried out in order to determine the most suitable backbone for the poly(pheny1ene sulfide) which was eventually to be substituted with nitrile groups and crosslinked to give the final cured laminate.
The compoistion of the various poly(pheny1ene sulfides) obtained along with the reaction conditions and results of polymerizations of m-benzene- dithiol (MBDT) with various dibromo compounds are given in Table 111. Elemental analysis data of the various polymers are given in Table IV.
As with the model compounds, a bright yellow color was observed during the initial stages of the reaction, which slowly changed to a light brownish color as the reaction proceeded. At the end of the reaction, the reaction mixture was cooled, and the polymers were obtained by precipitation with acidic water-methanol solution. With the exception of polymers PVII, PVIII, and PIX, all the polymers obtained werc stirred in a large volume of ether to dissolve the low molecular weight fractions and possibly any unreacted starting materials. Polymer PVII was precipitated from ben- zene, and polymers PVIII and PIX were extracted with boiling ethanol during the final purification stages.
As indicated in Table 111, the polymers were generally obtained in fairly high yields. As seen from the model reaction involving bis(p-bromo- phenyl) ether, i t is quite obvious that the displacement of the bromine atom by the thiophenoxy anion is quite difficult, and a polymer with this structure could be more readily prepared from m-dibromobenzene and 4,4'- bis(pheny1 ether) dithiol.
Poly(pheny1ene Sulfides).
TA
BL
E I11
Po
lym
erss
9 ?
5.60
77
mD
BB
80
12
.70
24
8.7
110-
120
0.1
? "2 U
U
Poly
mer
g
(mol
e)
g (m
ole)
b m
l g
(mol
e)
hr
g "0
dl
/&
3
Dib
rom
o co
m-
MB
DT
, po
und,
D
MF,
K
&O
a,
Tim
e, Y
ield
, M
p,
?inh
,
(0.0
4)
9.43
41
(0.0
92)
(0.0
4)
DS
G4
PI
5.60
60
pDB
B
80
12.7
0 24
7.
7 11
0-14
0 0.
15
5 (0
.04)
9.
4290
(0
.092
) U
(0
.04)
9
PI1
..
E 50
3.
97
24
2.5
240-
250
-
4.72
59
(0.0
25)
(0.0
29)
c F 4.
114
DB
BP
70
9.84
24
8.
81
160-
190
-
PIIT
PIV
(0
.031
) 9.
6721
(0
.071
) (0
.031
)
2.36
33
(0.0
1)
2.84
10
mD
BB
40
6.
35
24
4.25
85
-105
0.
11
l&+4
%.3;
PV
(0.0
2)
DB
BP
(0.0
46)
3.12
00
(0.0
1)
PIX
2.83
92
(0.0
2)
2.84
42
(0.0
2)
2.34
23'
(0.0
1)
2.34
00'
(0.0
1)
DB
N
5.71
99
(0.0
2)
6.56
98
(0.0
2)
2.36
18
DB
E
mD
BB
(0.0
1)
pDB
B
2.36
00
(0.01)
40
6.35
24
4.
67
155-
195
-
(0.0
46)
40
6.35
96
4.
41
55-7
0 0.
07
(0.0
46)
20
3.2
48
3.20
70
-90
0.14
(0
.023
)
20
3.2
48
2.96
18
5-23
5 0.
22
(0.0
23)
a A
ll po
lym
eriz
atio
ns w
ere
run
at 1
50-1
55°C
. b
mD
BB
= m
-dib
rom
oben
zene
; pD
BB
= p
-dib
rom
oben
zene
; DB
BP
= 4
,4'-d
ibro
mob
iphe
nyl;
DB
N =
1,4
-dib
rom
onap
htha
lene
; D
BE
= b
is(p
- br
omop
heny
1)et
her.
The
mel
ting
rang
es r
epor
ted
wer
e ru
n in
ope
n ca
pilla
ries
and
wer
e un
corr
ecte
d.
Vis
cosi
ties w
ere
obta
ined
at 3
0°C
in a
0.5
% s
olut
ion
of t
he p
olym
er in
HM
PA.
4,4'
-Bip
heny
l et
her
dith
iol.
0 g
Bro
mot
hiop
heno
1.
2800 HADDAD, HURLEY, AND MARVEL
TABLE IV Polymer Analyses
Polymer Repeating
Unit
PI
PI1
PI11
PIV
PV
PVI
PVII
PVIII
PIX
Analysisa
c, % H, % s, % Br, % Calcd 66.67 3.70 29.63 - Found 63.42 3 .54 27.28 5 .51 Calcd 66.67 3.70 29.63 - Found 66.21 3 .73 28.45 1.75 Calcd 66.67 3.70' 29.63 - Found 63.88 3.56 26.91 5.24 Calcd 73.97 4.11 21.92 - Found 70.85 4 .12 19.98 5.00 Calcd 70.78 3.94 25.20 - Found 69.13 3 .92 24.16 2.28 Calcd 72.18 3.76 24.06 - Found 71.95 3 .80 23.85 < 0 . 3 Calcd 70.13 3.90 20.78 - Found 64.66 3.85 18.65 8.10 Calcd 70.13 3 .90 20.78 - Found 67.56 3.84 20.18 <0 .3 Calcd 70.13 3 .90 20.78 - Found 67.06 4.34 18.84 -
* The bromine endgroups affect the analysis for carbon, hydrogen, and sulfur.
The structures of the polymers obtained were confirmed by elemental analysis (Table IV) and by infrared spectroscopy. As in the case of the model reactions, no positional isomerism on the aromatic ring is expected and the polymerizations are belicved to proceed via the nucleophilic dis- placement mechanism.
As would be expected, the melting ranges of the poly(pheny1ene sul- fides) and related polymers (Table 111) are dependent upon their re- spective structures. Polymers PI11 and PIX have the most regular struc- ture and were found to possess the highest melting ranges. This phenom- enon was most visible from results obtained from polymer PIV and PV which were obtained from a mixture of two different dibromo compounds. These results reveal that the incorporation of further m-thiophenylene units along the polymer chains has a rather large effect on the physical characteristics of the polymer and that a control over the melting range could be achieved by varying the amounts of the various dibromo com- pounds used.
The softening behavior of polymers PI, PI1 and PI11 have been studied.a These polymers exhibited distinct melting points and displayed a two-stage penetration on the softening apparatus. Generally, these penetrations occurred close to or at somewhat lower temperatures than the melting ranges reported in Table 111. In the case of polymer PI, DTA showed two distinct endotherms at 85-90°C and at 120°C which are apparently two crystalline melting points. It was, t,herefore, suggesteda that two different crystalline forms are present and that the same behavior was observed with
POLYARYLENE SULFIDES 2801
poly-p-thiophenylene samples obtained from Dow and Phillips Petroleum (Ryton) chemical companies.8
Isothermal aging of some of the polymers were run in a small furnace in an atmosphere of noncirculating air, and the samples were aged a t tem- peratures of 200, 300, 350, and 400°C. The results of these experiments are shown in Table V. All the polymers except PI11 melted and darkened a t 200°C. At the com- pletion of the isothermal aging the polymers turned into black solids that could not be separated from the walls of the glass containers, which indi- cates the good adhesion properties of these polymers even after heating at 400°C.
This behavior was exhibited by PI11 a t 300°C.
TABLE V Isothermal Aging of Poly(pheny1ene Sulfides) and Related Polymers
Weight loss, % Polymer 6 days a t
(Table 111) 200°C 8 days at
300°C 7 days at 9 days at
350°C 400°C
PI 0.34 PI1 0.41 PI11 6.65 PIV 1.48 PV 1.53 PVI 1.56
4.5 1.95
13.3 4.96 3.08 2.32
21.8 67.8 7.07 77.0
16.0 84.4 8.8 49.3 9.6 66.0
17.28 -
a Aged for 5 days.
The solubility of the polymers vaned considerably with structure. In general, the polymers did not dissolve in cold organic solvents but some of them did dissolve when heated but precipitated out when the solution was cooled. All polymers were, however, very slowly solubilized in cold hexamethylphosphoric triamide, and solutions used in viscosity measure- ments were prepared by initially dissolving the polymer in the hot solvent and allowing the solution to cool to room temperature.
Effect of Polymerization Conditions on the Molecular Weight of Poly- (mpthiophenylene). From the results of the polymerizations carried out earlier it was decided that the backbone of poly-m,p-thiophenylene would be the most suitable one to use in experiments involving the preparation of poly(pheny1ene sulfides) containing nitrile groups. This choice was based on the availability of starting materials, the better solubility and low melt- ing range of this polymer. The results of experiments run to determine the reaction conditions under wnich the poly-m,pthiophenylene of highest molecular weight could be obtained are given in Table VI. Elemental analysis data obtained for these polymers are shown in Table VII.
During polymerization 1, 2, 3, and 4, the polymers were observed to precipitat'e from the reaction media and to coagulate. Since the premature precipitation was believed to result in lower molecular weight polymers, reactions 5 and 6 were carried out using more dilute solutions. Less precipi-
TA
BL
E V
I E
ffec
t of
Poly
mer
izat
ion
Con
ditio
ns o
n th
e M
olec
ular
Wei
ght
of Po
lv-m
.p-th
ioD
henv
lene
from
MB
DT
and
DD
BB
~
Poly
- m
eriz
atio
n M
BD
T, g
PD
BB
, g
Solv
ent,
mlb
K
2C03
, g
Tem
p, "
C
Tim
e, h
r M
p, "
CC
ll
inht
dll
d
1
2 3 4 5 6 7 8 9 10
11
5.60
60
2.84
34
2.84
16
2.84
35
2.84
25
2.84
52
2.84
01
2.85
40
2.85
29
2.84
11
2.84
71
9.42
90
4.71
97
4.72
30
4.72
71
4,72
49
4.72
94
4.72
01
4.73
02
4.72
78
4.72
24
4.73
18
DM
F
80
DM
F
40
DM
F
40
DM
F
40
DM
F
80
DM
F
80
DM
Ac
40
DM
Ac
40
HM
P
40
NM
eP
40
Sna
40
12.7
0
6.35
6.35
6.35
6.35
6.35
6.35
6.35
6.35
6.35
6.35
150-
155
150-
155
150-
155
150-
155
150-
155
150-
155
165-
170
165-
170
190-
195
170-
175
170-
175
24
48
72
96
48
72
48
72
72
48
48
110-
140
110-
130
105-
125
100-
125
110-
125
100-
120
100-
120
105-
140
110-
130
120-
135
85-9
0
0.15
0.26
0.40
0.24
0.13
0.15
0.41
0.33
0.28
0.26
0.12
MB
DT
= m
benz
ened
ithio
l; PD
BB
-dib
rom
oben
zene
. b
DM
F =
dim
ethy
lform
amid
e;
DM
Ac
= d
imet
hyla
ceta
mid
e; H
MP
= h
exam
ethy
lpho
spho
ric t
riam
ide;
NM
eP =
N-m
ethy
lpyr
rolid
one;
Sn
=
0 T
he m
eltin
g ra
nges
repo
rted
wer
e ru
n in
cap
illar
ies
and
are
unco
rrec
ted.
d
Vis
cosi
ties w
ere
obta
ined
at 3
0°C
in 0
.5%
sol
utio
n of
the
poly
mer
in h
exam
ethy
lpho
spho
ric
tria
mid
e.
sulfo
lane
.
POLYARYLENE SULFIDES 2803
TABLE VII Elemental Analysis of Poly-mpthiophenylene Samplesa
Analytical data Found
Polymerization c, % H, % s, 70 Br, 70 1 2 3 4 5 6 7 8 9
10 11
66.21 65.89 77.32 66.10 66.02 66.09 65.92 65.64 65.67 66.11 62.45
3.73 3.63 3.75 3.75 3.72 3.66 3.74 3.63 3.72 3.62 3.92
28.45 29.12 29.35 29.85 29.70 29.58 29.87 29.93 28.00 29.38 28.64
1.75
0.93 <0 .3
<0 .3 < 0 . 3 <0.3 -
a Anal. calculated for (C&S)n: C, 66.67%; H, 3.70%; S, 29.63%.
tation and no coagulation occurred during these runs, however, the products obtained exhibited lower viscosities and thus lower molecular wiights.
As indicated in Table VI, an increase in molecular weight was obtained when the reaction time was raised from 24 to 72 hr. When the reaction time was raised to 96 hr, a drop in the molecular weight was observed. No plausible explanation could be given for this unexpected behavior.
Dimethylacetamide, hexamethylphosphoric triamide, pyridine, N- methylpyrrolidone, sulfolane, quinoline, and diphenyl ether were used in the polymerization reactions replacing dimethylformamide as the solvent. No polymerization occurred when diphenyl ether and pyridine were used, and only dimethylacetamide was found to be more effective than dimethyl- formamide. For that reason dimethylacetamide was used in the reactions involving the preparation of polyphenylene sulfides containing pendant cyan0 groups.
Synthesis of Poly(pheny1ene Sulfides) Containing Nitrile Groups. After the selection of poly-m,pthiophenylene as the polymer whose backbone was to be used in the preparation of the nitrile-containing poly(pheny1ene sul- fides), experiments were run to determine the best method by which such nitrile groups can be introduced into the polymer system. Three different methods to introduce the nitrile groups were attempted. The first in- volved end-capping the polymer chains with nitrile groups, while the two other methods involved substituting the nitrile groups along the polymer chains.
A poly-m,p-thiophenylene, presumably containing only nitrile endgroups was prepared by the reaction of the initial poly-m,p-thiophenylene with excess mbenzendithiol in KzCOrDMF followed by the reaction of the newly formed polymer with 4bromobenzonitrile. That a polymer con- taining nitrile groups was obtained was shown by elemental analysis and infrared spectroscopy which indicated the presence of nitrile groups after
2804 HADDAD, HURLEY, AND MARVEL
the polymer was vigorously stirred in a large volume of ether to remove 4-bromobenzonitrile. However, elemental analysis also indicated the presence of bromine which showed that the end-capping was incomplete.
TWO techniques were used to introduce nitrile groups along the polymer backbone. The first involved the bromination of the initial poly-m,p- thiophenylene with bromine, followed by treatment of the brominated polymer (mp 205-225°C) with cuprous cyanide. The final product showed a distinct nitrile absorption in the infrared spectrum, however, elemental analysis showed that a substantial amount of bromine was still left. This polymer was bright yellow, did not dissolve in all solvents tested, and turned dark brown when heated around 200°C but did not melt when heated up to 500°C. It could not be determined whether the insolubility of the polymer and its inability to melt is an inherent property of the polymer or whether it is due to crosslinking that might have occurred during its prepa- ration.
.The second technique to introduce nitrile groups along the polymer chains involved the use of a small molar per cent of a nitrile containing comonomer. In this method, a poly(pheny1ene sulfide) containing pendant cyano groups was prepared from m-benzenedithiol, 95% dibromobenzene, and 5% 2,4dichlorobenzonitrile or 3,5-dichlorobenzonitrile. The poly- mers were prepared in dimethylacetamide and potassium carbonate, since in this medium the highest molecular weight poly-m,p-thiophenylene was obtained (Table VI). The polymers were stirred in ether in the final puri- fication stages and were found to lose between 15-2074, of their weight be- lieved to be low molecular weight fractions. The polymers melted at ranges about 20" C lower than the corresponding pure poly-m,p-thiophenyl- ene of comparable inherent viscosity. The structures of the polymers were determined by infrared spectroscopy which showed a nitrile absorp- tion at 2245 cm-l and from elemental analysis which gave results that were within the allowed limits of the theoretical values.
Br 100% 95% 5%
100% 95% 5%
Thermogravimetric analysis of polymers PX and PXI are shown in Figs. Isothermal aging of the polymers I V ~ B carried in a small furnace
The results of 1 and 2. in air and the results of this test are shown in Table VIII.
POLYARYLENE SULFIDES 2805
2o
Linear polymer - - - - Cronalinkcd polymer -
T e p . (OC)
1 I I I I I I I
isothermal aging and TGA seem to indicate that these polymers possess good thermal stability.
It has been reported in the litera- ture that compounds containing nitrile groups can be trimerized in the pres- ence of an acid catalyst or simply under pressure and heat to give s-triazine molecule^.^ Because triazine molecules are known to have a fairly high thermal stability, they would be appropriate constituents in the final lamin- ate structure. For this reason poly(pheny1ene sulfides) containing cyano groups were synthesized.
Synthesis of Crosslinked Polymers.
TABLE VIII Isothermal Agmg of the Nitrilecontaining and Crosslinked Polymers
Cumulative Weight Loss, yo Polymep 10 days at 300°C 8 days at 350°C
PX PXI PXII PXIII PXIIa PXIIIa
3 .7 3.9 2 .4 2.9 2 .3 1.5
16.2 18.6 19.8 19.6 13.2 15.8
a Polymers PXIIa, PXIIIa were crosslinked polymers washed to remove zinc chloride.
2806 HADDAD, HURLEY, AND MARVEL
Teap. (OC)
I I I I I I I 100 200 300 400 500 600 700 800 900
Fig. 2. TGA (- -) of linear poly(pheny1ene sulfide) (PXI) and (-) of the correspond- ing polymer PXIII crosslinked with zinc chloride. Heating rate AT = 3"C/min; atmosphere, N2: flow rate, 98 cc/min.
Crosslinking of the polymer with nitrile endgroups was carried out, but very little crosslinking occurred. This was believed to be due to the in- complete end-capping of the polymer chains as well as to the low concen- tration of nitrile groups making the trimerization statistically unfavorable.
The synthesis of polymers PX and PXI was carried out after the failure of the polymer with nitrile endgroups to give a crosslinked product. Cross- linking of these polymers (PX and PXI) was carried out under nitrogen at 350-40O"C without catalyst or a t 290°C in the presence of zinc chloride. Initial attempts to cure PX in a carver press a t 250-300°C and 15,000 psi resulted in a soluble polymer, however further attempts to cure the same polymer a t temperatures in excess of 350" and under nitrogen resulted in a completely insoluble product. A fiber cloth was then impregnated with about 34% of its weight of a nitrile-containing poly-m,p-thiophenylene and the soft laminate obtained, when heated in a carver press a t 350-400°C with occasional application of about 10,000 psi, produced a black, hard but slightly flexible laminate without any appreciable weight loss.
Crosslinking of polymers PX and PXI with zinc chloride was carried out, since the curing process in this case can be achieved at lower tem- peratures (290-300" C). Because zinc chloride is so hygroscopic, the amount of this catalyst used could not be determined exactly; however, about 7% by weight was used. Initial heating of PX at 210°C and under
POLYARYLENE SULFIDES 2807
nitrogen failed to produce any crosslinking of the polymer chains and yielded a product that was soluble in hexamethylphosphoric triamide in which the original linear polymer dissolved quite readily. Further heating of the polymer a t 290°C for 24 hr produced an 89% insoluble polymer, PXII. Trimerization of polymers PX and PXI was then carried out by heating in the presence of zinc chloride and under nitrogen a t 210°C for 12 hr followed by heating a t 290°C for 24 hr. The product obtained by heating PX was 870/, insoluble and that obtained from PXI (PXIII) was 76'% insoluble. These results seem to indicate that PXI is probably more difficult to crosslink than PX, and it is believed that this could be due to the presence of the nitrile groups in the more crowded ortho positions.
A T: 2.5°C/min Abospbere: Static A i r
Linear polymer - - - - - Crosslinked polymer -
Tenp. (''2)
I I I I I I I I 300 350 400 200 2" 50 100 150
Fig. 3. Softening under load (--) of linear poly(pheny1ene sulfide) PX and (--) of polymer PXII crosslinked with zinc chloride. Heating rate AT = 2.5"C min; atmosphere, static air.
Identification of the products as the result of trimerization and not charring was indicated by the infrared spebtrum of the final products which were identical to those obtained for the starting materials without the nitrile absorption and by elemental analysis which gave values very close to the expected ones. Further evidence for the crosslinking was supplied by the Ehlers-Fisch softening behavior technique, in which a comparison of the amount of penetration under load and a t different temperatures was made between the linear and crosslinked polymers. Figures 3 and 4 show that the crosslinked polymers were less penetrable than the linear polymer a t higher temperatures, which suggests that some curing of the linear polymer was achieved.
2808 HADDAD, HURLEY, AND MARVEL
The crosslinked products were dark brown, almost black, and showed excellent glass adhesion which made it difficult to obtain the crosslinked polymers free of glass particles for use in the various tests. A convenient method for obtaining the polymer free of glass particles involved the coating of the inner walls of the glass tube with aluminum foil followed by the re- moval of the foil from the crosslinked product by reaction with dilute hy- drochloric acid.
Thermogravimetric analysis of samples of the crosslinked polymers still containing zinc chloride are shown in Figures 1 and 2. The results of isothermal aging of the untreated crosslinked polymers (PXII and PXIII) as well as those washed with hexamethylphosphoric triamide and
A T: 2.5'C/min Atmosphere: Static Air
Tenp. (OC)
I I I I I I I 50 100 I50 200 250 300 350 100
Fig. 4. Softening under load of polymer PXIII crosslinked with zinc chloride. Heating rate A T = 2.5"C; atmosphere, static air.
water to remove zinc chloride are shown in Table VIII. These results show that the crosslinked polymers possess a rather high thermal stability. A comparison of these results with those obtained for the linear polymers show that zinc chloride-free crosslinked polymers possessed the highest thermal stability while the untreated crosslinked polymers possessed the lowest thermal stability. These results seem to indicate that zinc chloride could be the reason behind the excessive weight loss in that it might be enhancing the oxidation and hence degradation of the polymer.
A laminate was prepared from PX on glass fiber in the presence of zinc chloride. The curing was run a t 250°C under about 15,000 psi and pro- duced a hard but slightly flexible laminate.
POLYARYLENE SULFIDES 2809
EXPERIMENTAL
Commercially obtained samples of thiophenol, bromobenzene, and m- dibromobenzene were dired and then fractionally distilled before use. Similarly samples of 4-bromothiophenol, p-dibromobenzene, 1,4-dibro- monophthalene, bis-p-bromophenyl ether and 4-bromobenzonitrile were recrystallized.
DMF dried over magnesium sulfates and DMAc and HMPA were dried over calcium hydride before distillation.
The polymerization solvents were dried and distilled before use.
rn-Benzenedithiol
m-Benzenedithiol was obtained by the reduction of m-benzenedisulfonyl chloride with zinc amalgam in sulfuric acid. The methods employed for the amalgam preparation and the reduction have been reported by Caesar.'O The only modifications made to the reported procedure were concerned with the workup. The reaction mixture was filtered under gravity and the solid material was washed with distilled water then ether. The dithiol was extracted from the aqueous filtrate with ethyl ether. The ether ex- tracts were combined and dried over magnesium sulfate. Ether was then removed by rotary evaporation, and the crude dithiol was distilled under reduced pressure on a Vigreux column to give the pure dithiol, bp 98"C/2.2 mm, in yields ranging from 60 to 80%.
H, 4.28%; S, 44.88%. ANAL. Calcd for CeHcS2: C, 50.70%; H, 4.23%; S, 45.07%. Found: C, 50.61%;
4,4'-Diphenyl Ether Dithiol
This dithiol was prepared by the use of a prcedure reported in the litera- ture." After initial workup, the crude product was sublimed three times (96-98"C/0.05 mm) to give a white crystalline product; mp, 101-103°C (uncorr.); lit., 103-104°C.
H, 4.47; S, 27.17%. ANAL. Calcd for C12HlOOS2: C, 61.65%; H, 4.27; S, 27.35%. Found: C, 61.59%;
Model Reactions
All model reactions were carried out in three-necked flasks fitted with a reflux condenser and a drying tube. A 15% molar excess (based on thiol concentration) of K&03 was generally employed. All reaction mixtures were initially worked up in a similar manner. The DMF was removed under reduced pressure, and dilute HC1 was then added to the remaining slurry. The aqueous mixture was extracted with benzene and the com- bined extracts after washing with water, drying over magnesium sulfate and filtration was rotor evaporated to remove benzene and leave the crude model compounds behind. The pure model compounds were then obtained by distillation or recrystallization as indicated in Table I.
2810 HADDAD, HURLEY, AND MARVEL
After the initial workup, the model compound obtained in reaction 8 was distilled under reduced pressure and distillate was dissolved in ether. The resulting solution was then extracted with 20% aqueous NaOH, 20% HC1 and 10% aqueous NaHC03, and the light yellow solution was dried over anhydrous sodium sulfate. After removal of the ether, the yellow oil which solidified was rccrystallized from ethanol.
Polymerizations
All polymerizations were run under nitrogen in a three-necked flask fitted with a condenser and a mechanical stirrer. Since all polymeriza- tions were carried out similarly, only a sample polymerization is described here.
Into a 200-ml, three-necked flask equipped with a mechanical stirrer, a reflux condenser, and a nitrogen bubbler were placed 2.8401 g (0.02 mole) of m-benzenedithiol, 4.7201 g (0.02 mole) of p-dibromobenzene, 6.35 g (0.046 mole) of potassium carbonate, and 40 ml of dimethylacetamide. The reaction mixture, which turned light yellow, was stirred, and the temperature of the oil bath was raised to 160°C and maintained between 160 and 170°C for 48 hr. After cooling to room temperature, the reaction mixture was poured with stirring into a cold solution of 800 ml of methanol, 400 ml of water, and 100 ml of concentrated hydrochloric acid. The solid that formed was filtered and washed well with water and methanol. The dry solid was then stirred well in 1.5 liter of ether, filtered, and dried in vucuo. The off-white solid obtained weighed 3.7 g (89yo yield) and softened a t 100-120°C. A 5% solution of the polymer in hexamethylphosphoric triamide gave an inherent viscosity of 0.41 dl/g.
The preparations of the polymers containing pendant cyan0 groups were carried out similarly. Polymer PX was prepared from 2.8457 g (0.02 mole) of m-benzenedithiol, 4.4827 g (0.0190 mole) of p-dibromobenzene, 0.1722 g (0.0010 mole) of 2,4dichlorobenzonitrile, and 6.35 g (0.046 mole) of potassium carbonate in 40 ml of dimethylacetamide and was obtained in 84y0 yield after stirring in ether.
ANAL. Calcd for (CIZHSZ)O.X, (CI~H~NSZ)~.~~, , : C, 66.52%; H, 3.68%; S, 29.49% N, 0.27%. Found: C, 66.30%; H, 3.60%; S, 29.71%; N, 0.30%.
Polymer PXI was prepared from 2.8447 g (0.020 mole) of m-benzenedi- thiol, 4.4831 g (0.0190 mole) of p-dibromobenzene, 0.1725 g (0.0010 mole) of 3,5-dichlorobenzonitrile, and 6.35 g (0.046 mole) of potassium carbonate in 40 ml of dimethylacetamide and was obtained in 82% yield after stirring in ether.
ANAL. N,0.27%.
Calcd for (C~zH&)o.gs, [CI~H~NSZ)~.OS,,: C, G6.52%; H, 3.68%; S, B.49% Found: C,66.33%; H,3.80%; S,29.43%; N,0.25%.
Crosslinking Reactions
Crosslinking of PX. A 0.5 g sample of PX was mixed wyell with 0.035 g The resulting mixture of zinc chloride by the use of a mortar and pestle.
POLYARYLENE SULFIDES 2811
was then added to a test tube with a side-outlet and equipped with a nitro- gen inlet. Under nitrogen, the contents of the tube were heated a t 210°C for 12 hr followed by raising the temperature to 290°C and maintaining it a t this temperature for 24 hr. The black solid obtained (PXII) was found to be 89% insoluble when stirred in hexamethylphosphoric triamide for 24 hr and gave an infra;ed spectrum that corresponded to that obtained for PX without the cyano absorption.
C, 65.72%; H, 353%; N, 0.34y0; S, 29.17%. ANAL. Calcd for PXII: C, 66.52%; H, 3.68%; N, 0.27%; S, 29.49%. Found:
Crosslinking of PXI. Crosslinking of PXI was carried similarly as that The crosslinked product PXVI was only 76% insoluble in hexa- of PX.
methylphosphoric triamide.
ANAL. Calcd for PXIII: C, 66.52%; H, 3.68%; N, 0.27%; S, 29.49’%. Found:
Preparation of Laminates. Laminates were prepared by impregnating three pieces of E glass 181 glass fiber cloth with a solution of the polymer with or without zinc chloride in hexamethylphosphoric triamide, followed by evaporation of the solvent under partial vacuum. The soft laminates thus obtained were placed in a carver press and heated under pressure for 24 hr to give dark colored hard laminates.
C, 66.34%; H, 3.6370; N, 0.22%; S, 29.47%.
We gratefully acknowledge the Air Force Materials Laboratory, Air Force Systems Command, Wright-Patterson Air Force Base, Ohio, for the financial support of this work.
We are indebted to Dr. G. F. L. Ehlers, Air Force Materials Laboratory, WrighbPatter- son Air Force Base, for the thermogravimetric analysis and to Dr. R. K. Fisch, of the same laboratory, for the softening curves.
We are also indebted to Dr. V. D. Todd of Koppers for a generous supply of m-ben- zenedisulfonyl chloride.
References 1. R. Lenz, C. E. Handlovits, and H. Smith, J . Polym. Sci., 58,351 (1962). 2. C. E. Handlovitz, Macrmlecular Synthesis, 3,131 (1969). 3. J. Edmonds and H. W. Hill, U.S. Pat. 3,354,129 (1962). 4. A. L. Blank, Makromol. Chem., 140,83 (1970). 5. Handbook of Chemistry and Physics, Chemical Rubber Publishing Co., Cleveland,
6. J. R. Campbell, J . Org. Chem., 29, 1830 (1964). 7. C. S. Marvel and J. Wilbur, unpublished results. 8. G. A. Loughran, G. F. L. Ehlers, and K. R. Fisch (Wright-Patterson Air Force Ma-
9. Z. Rappoport, The Chemistry of the Cyan0 Group, Interscience, New York, 1970,
10. P. Caesar, Organic Synthesis, Call. Vol. ZV, Wiley, New York, 1963, pp. 693-697. 11. P. C. Dutta and D. Mandel, J . Indian Chem. Sac., 33,54 (1956).
Ohio, 52nd ed., 1971-1972.
terials Laboratory), private communications.
p. 410.
Received July 6, 1973 Revised August 9, 1973