Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Institutional Repository
Characterisation of
long-chain branching in poly
(vinyl acetate) and poly
(vinyl alcohol)
This item was submitted to Loughborough University's Institutional Repository
by the/an author.
Additional Information:
Publisher:
c T.A. Coleman
VariablePrint Quality
Characterisation of Long-Chain Branching in
by
Technology.
1984
Do-D4rtment of Chemistry
by T. A. Coleman 1983
T
AB,")'LRA(.
changes on hydrolysis.
fractions
showed that, -in the molecular weight range studied,
ally distinguished.
Loughborough, and for his help and encouragement without which this
work may not have been completed. I would also like to thank
my grant and to Dr. Salmon and Dr. Ramsey at the U. K. A., E. A., Harwell,
past four years. Finally to my typist for her speed and accuracy,
many thanks.
Originality
The work presented in this thesis has been carried out by the
author, cxcept where otherwise acknowledred, and has not been pre-
Pa-e
I. T2 C,DU,,'T I-1,
I 11. 1
2. THEORY 6
2.4.1 Calibration of G. P. C. 25
EXPBRI?, T;NTAL 33
M Acetylation Poly(vinyl 41
of alcohol)
3.7 Characterization 45
3.7.1 Viscosity 45
Poly(vinyl acetate) 45
33-7-1-1
Pag,e
3.7.2.2 Poly(vinyl 47
alcohol)
3.7.3 Osmometry 48
4.3-1.1 By Reacetylation. 52
4.3-1.2 By Hydrolysis 54
ACi,ýTATE) 62
5. LONG BRANCHING IN POLY(VINYL
62
5.1- Introduction
63
5.2 Application of the Ram and 1jiltz %ethod
Initial Procedure 63
5.2.1
6. POLY(VINYL ALCOHOL) 72
6.4 Conclusions 75
7. CONCLUSIONS
AND RECOX1,
MDATIONS 77
References
l. INTRODUCTICN
1.
Much work has been carried out on both polymers due to their
branching in PVOAc and PVOHhas been studied for many years. The
(W)
A combination of viscometry and lightscattering was suggested
branching (LCB) in
by Hobbs to determine the degree of long chain
6)
order of 2x 10 in benzene, suggested that the non-Newtonian
had already been derived for several model branched chains. by the
49
theoretical treatments of Zimm and co-workers - With the advent of
.
irrespective [_QY,
polymers, of type of structure, a plot of
shed the curve, knowing the elution volume and n] for a polymer of
U. C. method, and hence its extensive use, has mainly been due to a
effort has been directed to deriving suitable methods for the inter-
gyration of the polymer molecule in a good solvent (< s2 >), and the
2
(<s >)
unperturbed mean square radius of gyration have been pro-
0
posed, but supporting evidence has so far been inconclusive. For
most work the hydrodynamic volume Ufllrl is still the easiest para-
Viscometry, both on and off line, has become the most often
weight data were derived from a U. C. curve using I'S standards. Park
94 the
and co-workers reported the analysis of FVOAc via method of
reviewed the methols of Drott and ýiendelson, and Ram and 1-tiltz with
gave essentially the same results, and showed that, for PE, was
not constant.
thp above methods dependent on molecular size, have not been reported
PVCH.
2. THEORY
6
to consider: -
1) Initiation
heat or u. v.
I 2 Re
initiat6r 2 radicals
Re + --C=C- R-CI-C'*
monomer propagating
species
2) Propaýzation I
chain,
IIKP
R-C-C- + --C=C'
propagation and so on
rate constant
3) Transfer
has been transferred. Transfer reactions can taice place with almost
Termination
(propagating
can occur either by the combination of two radicals
chains,
IIII IIII
P-C-C. P-C-C-C-C-P (2.4)
1111
combination
III
+ P-c-c-
11
disproportionation
The transfer reactions with monomer and polymer are those which
3
lead to branching. Depending on the type of transfer reaction,
1) Transfer to Poly
Kt II
r, '? (2.6)
P-C-Ce + P-t-
-OH--CHZ- I
III transfer
propagating OAc UAC
rate
species constant
polymer chain
(repeating unit) + Lonomer
I'! NI
UAC
II Ktr, II
P, P-C-C-H + --CH-CHF
P-C-C. + -CH-CHF-
transfer rate u
I
C=O constant C=O
I I
Uli3 CH2
+
-ýH-CHF > -CH-CHi- (2-7)
0
1y
U=U C=O
II
Ulll2c FT
2) Transfer to monomer
[".1
IIK tr, m?
P-C-C- + CH212121
ýCH - CH =C0+ > CH =0- (2.8)
UAC URC UAc
vinyl acetate
monomer )
(VAcf,!
C)
C=0
(I; ti 3 ul'2 CHE-
CH + RO R-CH +I (2.10)
2ýC- Z-6-
UAC UAC
+
CH =TH + RO R-CH. -CH > R-CH H
2 I 2-T
OAc- UAC- OAc-
polymers are formed, regardless of the route, for the same type of
branching.
have internal saturation and will therefore polyr, ýerise very slowly
29
if at all.
H:
ý Pa,, H
HC0
1
C=O
I
H-C-H
I
H
derived,
K
V= p K, (2.12)
-14 =
2(-r Kd Kt) [I]
V is
where the kinetic chain length defined as the number of monomer
initiate chains
where ý' and Xn are the number average molecular weight and degree of
n
polymerisation, and M is the molecular weight of the repeating
rp
unit. For termination by combination 7, 2V and for disproportion-
n=
ation Xn=V. In practice, combination and disproportionation may
d^Y K [N E) N0 (2.14)
dt tr, p
de
(2.15)
and dt =X
p [1-:(1
-] - e)
Dividing 2.14 by 2.15 gives equation 2.16,
dy
d ()
eNo CN EV (i - E)
Kp (I - E) p0
Integrating Y/11
2.16 and setting
Oe =P gives
1-
Cpe (1+ ln (i -e
12
typical high polymer reaction but is also important for the manu-
7
catalyst used. The general reaction is)
change occurs. Long side chains, which are attached to the main
chain through the oxygen of the acetoxy group (figure 2.1) are split
off, and although the polymer may be reacetylated, the long chains
4
remain as separated molecules. Since it is generally true that the
in length and hence molecular weight, more than shorter ones with a
lesser number or even no branches. This leads to a naxrowing of
than 99%.
2.3 POLYKERSOLUTIONTHEORY
Polymer Dimensions
2.19.
2 2 ýl - cose
<r >= na (1 +
0 cos 8
CLASS OF PVOH TIYDROLYSIS -VISCOSITY GRADEScP. (a)
DEGREE
98 4..lOt4Op56
Partially hydrolysed 88 4., 8, l8p4O
-6tal 0
to
C-)
S
4)
I-
a
CL*
a
CN
trw
%Z
c
I
C4
.
LL LA
0
Ld
91
2,
Eý
LZ
AIISOOSIA OISNDJINI 6ol
0
(71
OD
0
N
0
4D
0
3
0
0
be
co
CD
M
14
LA
0
1.4
distance between the ends of the polymer chain. The rms distance
1 (2.20)
2
Long branches in the polymer chain decrease the value of<s 6>1
11 2 I<r 2 (2.21)
<r 5,0
242i
<r :> is the actual rms end-to-end distance of the polymer and <r
0>
I OC
is called the un-Y.
)erturbed dimension, i. e. the value of<"r'ýý2 when
is unity.
OL becomes unity. 31
with solvent I is so small that
15
Viscosity Equatiom;
'ý 'fl
where and axe the viscosities of solution and solvent respec-
s
tively.
involves the double plot of the Hug ins and Kraemer equations, 2.23
1ý9
and 2.24 respectively.
-qsp (2.23)
in In'q (2.24)
C
In -q nj + DE-q ul
2
E
ý
E-ql, E-q], C2
E-qJ
+ K,
a +K
Clearly, when IKH3 =1 (Y'J' is equation 2.24 accurately
only
into the determination ofEfl]. Again due to the nature of the Taylor
lead
can to large in
errors [-Q.
There have been numerous further equations proposed to overcome
34,36'37
these difficulties. Table Z. 2 summarises some of these-, yet
32,33
more are considered by Berger. In this work, the Huggins
equation (equation 2.23) has given good linear fits to the data
2 (. 3)
[m]= 2>is
is constant. I. ý the term <& substituted
where a un-liversal
2.31 is given,
r, 2 -11/2 (2. JI)
1<r, >l
L-7-J
2-,./ is if CL>>l it
m a first approximation<r, L, constant and,
.Lo
is found that Z-31 can then be re-written
-Equation
['fl] o. 8 (2-3z)
= K!,.';
[-q]is p,Iven by
(<r 2 /T,
T) 3/2.
where KT
[-fl] ja (2-34)
ý 11
The iii equation is based on, and is a good approximation to, the
and 2 -'13.
(by E-q]
weight an absolute method) and of a series of polymers of
T-
F, ii (2-35)
number average =- 14
n i
2
Ili "i
weight average (2-36)
>N Ili
(1+a)-
yi,,
li
viscosity average (2-37)
v-ZNi Ili
i-
but over an extended range of molecular weight the log I-q] vs. log Ii
39
plot is both theoretically and experimentally curved. The constants
Ko ,M0.5 (2-.38)
40
is obeyed. The constant Ko is found to be very to K 0, the
close
characteristic value in a (3 solvent at the sane temperature in the
19
though conditions
absence of large, specific, solvent effects, even
wide range of molecular weights may exist, but all having the same
reacetylation.
Ni, ý 'Ti,
b
(Tti. (2-39)
a 3+1)
where subscripts 'b' and 'at refer to before and after hydrolysis
NA (2.40)
0z
1:
EL
-I-
I-
21
weight, viz
F-"*ilb lli,
b
Im
n (b) -T Ni, b
F- "i, 11i,
a a (2.42)
n(a) Tli,
a
of branch b3'
and defining the number average number points
(2.43)
n, 3_ (b)
FN F-(Ni,
n (b) i (a) b
(2.44)
"i (b) YN
Fin(a) (b)
-i
'n(b) (2.45),
"n(a)
is tiat of the ratio of the radii of ", ration of branched and linear
47
molecules, of the same molecular weight.
zb
(2.46)
2,
<s
2 3/2
I'M (X <r
(intrinsic
>
viscosity)
21 (rns)
where<r >, is the root mean square end-to-end distance of the
polymer chain.
22 ['ffl 3/2
For linear molecules<r > cx<, I.-ý>, therefore (X<-s >.
L;T
1
9TE
for monodisperse systems, and
(2 + 13, + -fL3,
6 w)ý w] (2.49)
8*1 ln
3, P ;%3,_ z 13,
w + ýTb,
w) wJ
-(Z
for polydisperse systems.
P'SCB ý
El/(s+,
)]xEl + s(1-2r+2f2-2f3) + S2(_:r+4f2-f3)] (2-50)
to the polymer rain chain, whereas short branches axe usually pendant
favour b,
of workers a function of the type !ý' = r, where the vabic of
24
b is found (er-,.,,,
)irically) to lie between 0.5 and 1.0. *,any authorý, ý
data to theory.
'o far, there has been no method developed for directly measur-
'Ytw
or xý (2
-52-)
branching index
presu. ned to is
remain constant, with 14,3 varying
46
linearly with molecular weight. This assumption, however, is not
X
universally valil. For polystyrene has been shown to be more or
5(), 1 8,79 94
less constant but for poly(ethylene)7 . and PVUAc
,
-rouni to vary considerably, with molecular weight, though much con-
2.4 CHA:q-ACT_'-RTATI.,YTi
diffusion measurements have also been made. All these methods have
22
been the subject of a recent review.
taking care and expertise by the operator. For this reason most work
light scattering detectors for GP(I are becoming more widely available
(and reliable) the current situation will no doubt change. For the
this work has used the viscometry/GP(, method since both these
-oresent
confined to these.
45,52
2.4.1.1 Calibration by Narrow Stand, 3xds
obtained for each of the fractions and the elution volumes correspon-
(2-54)
peak nxý, Iw
For anything but the most common polymers narrow standards are
not usually available. Even when they are, a few are reasonably
more difficult and less accurate. For this reason, several methods
calibration are generally only available for the more common polymers.
14,53
However, since the introduction by Benoit of the concept of the
"=2
[rn]2 (2-55)
However useful this theory is, there has been imuch discussion
The view exDressed by Casass a56 and co-workers was that only
Ance GPC alone cannot furnish unambiguous data, use has been made of
the intrinsic viscosity of the polymers, and data aw- usually mani-
outlined here.
46,51
2.4.2.1 The ].ýethod of Drott and Mendelson
_fLwIl-2
= constant (2.56)
I'll
= X)
f(. 11
6,1w
IT
b
Using the definition of g' (2-50))
ITO'
(where subscri- ptzi 1 and b refer to lin ear and branched species),
29
-ql
b
(2-59)
- (<3>w) b
Im 7- (2.62)
whole = E-fo
1 wi
where w1 is the weight fraction of speciez i. Accepting that the
E _f]
E-flIF
whole
Mb F) ]b
.a ý 1,
-' _ hi: !ýEf
i Uj X
X,
A value for b is chosen and, assuming constant an iterative
64
2.4.21. ý 'The Ilethod of *Jild, Ranganath and liyle
61
2.4.2.4 The Ram-and "iltz 'ýethod
ILhe method employed by Rain and ", iltz vias intended to circumvent
23
ln ln K+a ln +b ln Ii +c 1-n 65)o
for polymers whose molecular weight was sufficiently high enough for
0
branching to be present. Below this minimum molecula. ý weight (,.,,
0)
the ordinary A-Ark-Houwink equation was used.
Since f or 11 = 1,:
0
2
b ln ln3 yýo = C, (2.66)
1*0+c
('-. 67)
Cb In 1
0
2
(a + blnil + cln
n1calc Efl]ihi ZEhjKMj (2.68)
ýF
=
for b is iterated so that I'Mcalc matches measured.
a value '
This value of b is then used in the extended i-IH equation, again with
required).
required.
fact that, when dealing with broad chromatograms which may extend
merisation was chosen as the technique for the purposes of this work.
9 0 0
N .0 Cd .0
0 r-I 0
ýX4 Il: w +3 Cd 4-ý
0 C) 4-3 0.0 43 C: C:
9* r-I (1) 9ý rl (1)
0
H 4- 3 r +3
ný 0 $-4 0 0
C:
0
El a) 0 Q ýq El 0
E- Cd 4.3 0 0 Cd 0
V
(d 43 0 4
rt (D
Q) (1) CL (D a) (1) El
tLo ttr) 0 bo to tLo 0
0 ly, 0
10 ph
o
fn 00 4>
r4 ý4 m
M to 1-4
>1 ul ý--b m
rq :>11
0 0 ý- r-f
0 r-q
e r
: 31- Qý 0
ýr. V
I
0
. 0 ý4
0
;4
H
>1
0) 00 a) 00
X. -ri r, 4) 9 xi -f-I 9
4 3 43 CO 4> + 4 3 4-3
0)
91N
E
4 cg
0
0
-13 -ri
10 00
4
4)
-
ý! fn 1 CD . al 5
*Fý
m
s
0 -r4
: li P
Q
" A0
1ý -r-I "I I
ý-
.m
P- r i0
10 -H
W
E-
0 0 0 0
ý-4 r4 0 r4
"-I $3
43 ýl 43 -P 0
-ri 4j r4 -0-4 -M
10 10 Id 4-3
-H
Itj ro 10 qj
C:ý C) m a) CF) V2
E-4 P4 > 0 H
m
0 PLI IL4 r0-I cd 0
P.
C-) 0
C+4
0
43 ol
0.4 r-I 0
C)
04
10
V) 0-
-r4
43
9 0) ý =1 -r4 43 m
0) -H :>-. mm0
V) m 4-1 m0 (1)
u
Cd CS 4-4 -ri Cd e4--4 (1) tto W
rf
fri r4 0m w0" Cd 4-3
C-1)
C')
0 0 0
e
1
ý4
ý 4
I- PL.
0
ccl
-4
E-
34
contain less than water and it has been reported3 that up to 5ýj
PreDaration of Initiator
transler 5
reactions.
6
According to Burnett et al. polymerisation of VAcn below QOOCr
initiated by the photo-decomposition of A20N, resulted in predominantly
VAc- (170 ipl) was introduced into the reaction vessel 3nd cooled
to GOC. A solution of initiator in ice cold VAL (5 ml) was added
S. D
L
V)
a
L.
. L
a '.
CLI
V
0
C
W
L-
:3
cn
LL
-0/
W
0
8m A
u
LP E.E 0
9
52 1:,'
)
F 4
>
C ýý \o
CL
E
Ei
J
r, 5
f
0 (D Co rý, (0 to le
.a4
0 41 ýi 41
ýo
ý4)4)41je
iZ je xw
vuc Ai Aj 41 J 163
- 31
w (3, je 4) 00 ul
-I ZZ -0 v8w0-, 13
: ,0v gl. m
v0 >
'o m -0 ýK - -0 in 14 in
vV22mc -4
cu
c0
LL u8u0
r. E
44,
) w0 km Ch m0080
9uHx ,4CZ
4)
E. U) () () ca to to 94 0u
c;
1ý rz co 0;.4 .ý4A4; - .4 -1 ýZý'; ý4 -4 .4 dj4ý;
94 94 f4 94 jýj
f4 e4 N
35
was assembled inside a fume cupboard, whose glass front had been
The precipitated polymer was t'hen dried in vacuo an,.:L weighed. Final
was added to form a thin "soup" and the whole added dropwise, with
hexane alone).
in a vacuum desiccator.
36
vacuum line by the freeze-thaw method, then the ampoule sealed. The
(section 3.1.4.1).
a) Alcoholysis:
b) Hydrolysis:
c) Direct hydrolysis:
H
2 (3.4)
PV-OAc + nNaOll PV-OH + nNaOAc
d) Aminolysis:
IH20
PV-f2Ac + n.T; R1R2 >PV-CH + nAcAlli 1R2
e) Ammonolysis:
NH Cl or NH4>OAc
'4 (3.6)
PV-OAc + nNH3 PV-OH + nAcNH2
20> (3-7)
ROAC + NaCH RIM + NaOAc
14 or -01i (3-9)
V-OAc + R'OH =>CH CHO + ROAc
3
if HQAc (3.1; ))
V-OAc +H20 :4oGHCHO +
3
(V CHI.CH,,-)
.
The characteristics of each method are summarized in table 3-2-
a) Alkaline hydrolysis
OH cl--ý UH
rv-O-C=o FV-C-O PV-o-C-o
I V- I ýK- I
Xei. e .;e
43
4a)05
r-4
:: ) "
9-1
0
0
0
0
ý-& 10 In.
13
10 04
ly.
" r -A r -4 0
En CQ ca 0 P4
:4
ýo 0
$ý4
0 10
P4 C.)
E-4 "-I
C-4 0 0
+)-
4-4 0 0
C-) -ri 4> 4-3
lic 0 43
a) -rq cd k ;4 4-4
0
0
4-3 '-4
0
V0 r-I C) 0 0
e
C-ý 88
(D cd IZ W
4-)
Cd 0
ca 431 43
V
0 cd 10,
E-4 r-4 43 43
r-4 r. "A
Cd 10
"_4 0
ril
co
9 *5 8
0
0, 9 C-4
- 0
C34. 0
0
0
$4
0
-4
43
z
cv X
.14
r-4 0 4D
ý-4 (n
Cd 4-4
0
m
V
Cd -r-4
4Z
(D in
rl.
0 (D
0 4ý'
v 0
0 cc cd
1ý
00 4-ý
0
I
w i
:4
=Cmc, I
C-4
E- 0
I
ý4
E-4
1'119
e
OH Ä OH
I
FV-0-C-11 11 + 111-0-ii pv 'j
I!le
3alt
b) Acid hydrolysis
+
PV-0-C=C + HA PV-')-C=CH +A
H ý*+ R
0H OR
Ii' Oil
-
+ HA
1.'ie
for converting FVOAc to PVOH the following points have been borne in
mind:
Alcoholysis
completion.
Acid hXdrolysis
C) Alkaline hydr6lySiS
00 H 14ý0
R-C , R-C! -+RI Oli (3-11)
OR' .0
salt alcohol
Two methods were readily available for the conversion, one using
a high and one a low amount of alkali. Table 3-3 summarizes the
is kept 50 0C
about 30 minutes. Thereafter the mixture at with
viaorous stirring for a further hour. During the reaction PVOH pre-
Temperature (OC) 50 40
Required NaOH (g) 0.03 0.20
(a)Vigorous Stirring
(b)Standing
B. D. H. 500POOO -
PVOH
OURCE
SS
B D. H. 14.,000
solution is added, the mixture stirred once and then allowed to stand
for 30 minutes. During this time PVOH precipitates and the polymer
1 that it requires less ti. -,ie, a lower ter, -,perature and no sI-Decial
and the ratio, of alkali to polymer is about six times lower than in
over NaOH pellets), glacial acetic acid (Fisons, 99.5ýL) and acetic
24 hours. After this time excess reagent was removed using a rotary
evaporator and the product dissolved in methanol. The PVOAc was pre-
i comparison is made between the two polymers, this must be taken into
account. The way chosen here was to hydrolyse PVOAc to PVOlf, then
69
nolymers containing large numbers of hydroxyl functions. Infra-re
e hydroxyl
whoýý, content is reliably Imown.
67,70
3.4.1 Method 1: Reacetylation
acetylating'reagent.
dried over NaOH pellets) and acetic anhydride (0.7 ml, Fisons 97ý)
boiling tube containing PVOH (o. 1 g). An air condenser was fitted
and the mixture heated over a steam bath for 8 hours. Toluene (1.5 ml
S. L. R., distilled) was added and the boiling tube stoppered and shaken
and the whole titrated, with shaking, with 0.2 N aqueous sodium
by titration.
43
0.1 11NaOH was used. ) The flask was stoppered, shaken well, and
and the contents of the flask slightly acidic. Excess acid ensures
and the contents of the flask titrated with 0.1 N NaOH (A. R. )
similarly.
table 3.4.
carry out the process for PVOH this is a complicated and often
unsatisfactory procedure, and is seldom performed, though some work
72
has been reported.
44
methanol (S. L. R., 1500 ml) was added to the vessel and distilled
The temperature to 30 0C
give a clear solution. was raised and the
solution slowly stirred for approximately one hour. The vessel was
tion remained.
modif i
corical
held 30 0C the
was raised once more and at with stirring until
mer and the resulting solution removed through the central glass
procedure.
3.7.1 Viscosity
3.7.1.1 Poly(vinyl acetate)
(BDH A. R., stabilized with 0.1%* quinol) at 250C was measured with a
distilled 0
filtered, water and oven dried at 30 C. Flow times for
toluene and THF were -188 and - 157's respectively. Maximum and
35 0C in bath to
a water prior use. Each solution was carefully
LowerTiming Mark
Capillary
CAassSinter
the intrinsic n]
to obtain viscosity I cLtý'.
The viscometer was not modified since glass sinters proved too much
water and the whole boiled to effect solution. The solution was
(0-5 first
then filtered and a small sample - 1.0 ml) evaporated at
concentration of the solution. The flow time for water was - 130 s-
Chromatography (GPC)
3.7.2 Gel Permeation
Some GPC work was carried out by the author using a modified
Waters 502 ALC/GPC with TEY (BDH, A. R., stabilized with 0.1"/o'Ouinol)
effects were found unnecessary except for work carried out by the
TSK 3000 Pw type column in series, from the Toyo Soda Company of
polymers. The exclusion limit for PEO was - 600,000. A sample con-
1o injection
centration of ff/ with a 50 ul volume was used, each solu-
in solution).
3.7.3 Osmometry
Osmometry was performed on samples of unfractionated poly(vinyl
4o).
4. POLY(VINYL ACETATE)
RESULTS
49
K' CIA
[13 ý+ Cm (4.2)
R=V
n
Since the initial monomer concentration has been constant, and con-
l/x- should be linear. This plot is shown in figure 4.1. Values for
n
Xn were calculated from I-In given by GPC. The GPC was calibrated
(14
V-; 6-
Ln
2i
cn c0 0
0X
_5')
tem-Derature.
(PVOAc's A, B and C) are shown in figures 4.2 and 4.3. Table 4.1
gives molecular weight data for the polymers. The data are not con-
an increase in D.
Expressions for the Fý Aw
values of and of a mixture of
M-
polymers have been derived (equations 4.3 and 4.4).
D
C71
EL
u>;, --
ý4
Lr)
I)
I
a'
0
(D
D
Cl
LT-
10
a.-
t'.
ý_i CD 1-1 w bi ci C) C)
0 a's cý Co r- tD 13 en C,4
z =:
SAMPIE Rn(PS) Fiw(ps) D (a)
Cl
PVQALc 102,300 516,100 5-0/+
mix "L". The Rw values are in good agreement but a higher "'I value
n
The highest molecular weights for any one polymer reported here
5.
axe
F1 105 and -Mw- 5, x 10 Using more complicated cooling systems,
n_
6
Burnett claimed to be able to produce linear polymers of in up to
n
500,000, though in some cases the value of D as derived from osmo-
21
metry and light scattering was <1. Atkinson and Dietz prepared
00
high molecular weight linear PVOAc at -2 under essentially similar
produced.
temperature (designated B3 and B4) are shown in figures 4.4 and 4.5.
27
As predicted by theory the molecular weight distributions of these
polymers are much broader than for vinyl polymers of low to medium
when analysed by GPC. The high molecular weight tail appears to have
7 (figure
than 10 4-5). This makes inter-
molecular weights greater
4-3.1-1 By Reacetylation
-[CH 2-CH(OH)I-
Ali
1 mole of PVOH (100% hydrolysed PVOAc) moles hydroxyl. (OH) groups
-11
44
where 1.1 is the molecular weight of the polymer and the molecular
r
weight of the repeat unit is 44
"'r 1
1g polymer = 7Tý x 7-- ý 22-73 x 10 -3 moles of OH groups.
I-Ir
other being complexed with pyridine. For every mole of -OH groups
The blank solution will require 2 moles of NaOH per mole of AA.
(Blank titre 3arlDle titre )x L' 1073 per gram. of polymer (4.5)
-
9 x
polymer is given by
V" (4.6)
A /" Olf =wý 17 xVxLx 10-3 x loo = 1.7
POH 99
of groups).
17 X 1ý10H
100 1700 (4-7)
w (44XI,,! )+ (T -N ) x8F; -1
POH 44 + 86( T-
OH OH 1)
OH
MOM
No. of moles of OH 100 oil (4.8)
The mole OH-"'POH x
Total moles of all groups 1T
T 100
(4.9)
1.1 m
OH POH
l7kýOH
1700
and W (4.10)
POH 44 + loo 86 - 0.421-ý
- 1)86 OH
POH
54
86",4
POH
Mol % OH = Degree of Hydrolysis = 17 + 0.42W
POH
PVOHare detected.
M
OAC
MPOAC ý -T x 100
(mole 1%OAc)
T= 100
"ýOAc
OAc
591ýOAC
wPOAC (4.14)
IV+ + 0.42i4,
pOAC
44d
POAc (4.15)
000 ýIPOAC
59 - Oo42W
POAc
59oo - 86w
POAc (4.17)
III - 59 - o. 42W
POH POAc
55
the original NaOH put in, and ensure acid is in excess. The excess
neutralize the acetic acid, and hence the original acetate content.
VxN (4.19)
weight of OAC 1000 x 59
59VN (4.20)
weight of OAc per gram of polymer = 1000
g
ý. 9VN (4.21)
weight percent OAc (W
POAc 9
results. This also serves to show that the chosen method of conver-
before and after reaction) and secondly to demonstrate that the method
V)
>-1
rI
Cq
I
el
t
57
little effect is seen. Figures 4.4 and 4.5 demont3trate the chan-es
4.4 FRACTIONATION
Further fractions were not collected, since branching was not con-
plotted in fiGure 4.7. Also in the figure is shown the same data
for the fractions. The first of the collected fractions were dis-
tion was made into the cause of the discolouration; however, it may
carded fractions have been taken into account in figure 4.7. The
4.5 CHARACTERISATI ON
4.5.1 Osmometry
cally in figure 4.12 and in table 4.10. Figure 4.13 shows a compari-
curnkildive
weigli %
60 50 40
Mwt PS)
LOG
FRACTIUI Pn (PS) Pw (PS) PJLYDISPEIISITY
RATIO D
tD
9
Lr)
RgLre411,A2 rent Curnjajiye-ý6(=ht Disto ulm-d
cumulative Branched PVOAc B4
LogM(PS)
Figure 1.12
05 10 ConcentrationP/o) 15 20
SAMPLE iýI A2X103 iin (PS, GPC)
SAMPIE R, (OS) *1
before (a) after (a) n
LINEAR KI BRANCHED KI
FRACTIMS EV FRACTIONS
55-
Figure 4.13:
'-,L I Com an son of Vn
Osmometry-golC
2Eý
[whole branched po(yme
LOGMn (Gf
(PS)
49 50 51 52 53 54 5,5 56
3.
In.
59
equation 2.45. Table 4.11 gives values Of-ý, n(number average degree
4.5.2 liscometry
4.5-2.1 Fractions
the Huggins equation (equation 2-23). Tables 4.12 and 4.13 give the
though the points for the branched fractions appear a little more
the range of molecular weight studied, long chain branching of the noo-
The relative slopes of the plots are 0.77 for THF and 0.61 for
toluene. For the same molecular weight, PVOAc has a lower viscosity
is 27,43
expected since toluene is almost a theta-solvent for PVOAc
21
under these conditions, whereas THF is a good solvent. The values
SAMPM n] KI sp.ýTm n] K,
)LUENE
of P
Veight
6o
PVOAc, and 0.34 to 0.42 for branched. In this work, the average
0.36 and 0.43 respectively, while in toluene this was 0.47 and o. 43,
and linear species with respect to molecular weight, and this small
difference in K values is therefore considered insignificant.
Viscosity data for whole polymers are given in table 4.14. The
average value for K is 0.37, re-inforcing the view that the varia-
branching.
volumes are given in tables 4.15-4.17. The figures V(F4n), V(T.. and
D)
V(Q elution volume (viz time) corresponding to the
are values of
need not necessarily lie between Pn and riw). Having obtained values
elution volume
5.1 IlaRODUCTIO'N
with conventional GPC) is not possible. Some work has been done
is required and many methods have been used. itostly GPC has been
All these methods are considered valid, since the physical quantities
obtained.
19,82,83
The combination of GPG with either on or off-line
branching parameters.
5.2 APPLICATION OF THE RA,,,, AND MILTZ i,,ETHOD FOR LONG BRANCHING
Until now the Ram and Miltz (R/1-I) method has not been applied to
follows: -
whole polymers. From the plots, the elution volume parameter most
sistent with observed results. Hence values for the hark-Houwink (!,Ui)
64
The R/14 mEithod was applied to all fractions and whole polymers,
C)
was made.
values for various elution volume parameters were evolved according r>
EM 4.13-
to figure 4.16. These are plotted against from tables
Clearly the data are best correlated using V(FIw) as the elution
figure 2-5). The scattering of the points in the other two plots is
ii i
Q- 14*
D
21 0
.LL
c
C)
0
co
"0
8E
ý5
C) 0
ul
3:
0 0
00
C)-a
Ln
Ln
ILO
t
Ap
LL
C&5
I
FKNre 5.3- Intrinsic Viscosity of PVOAr. versus-f-luhQn. VoluMg
0.5- -
cxfni
0.4-
-0.
.-o.
-0.
-0,
-0.5
65
whole polymers however fall on the same curve as the fractions. This
than at the main chain, and this frequency increases with tempera-
3,8,11,24
ture. The different polydispersities of the polymer., do
values for the Mark Houwink (NH) constants E' and a are required.
16,21,84,85
Though these may be obtained from the literature
, agree-
ment between various authors is poor, and this warrants the re-deter-
=
n1pslim
1.1
PVOAc E'IUPVOAc . at constant Ve*
shown in figure 5.3 that V(Flw) gives the best interpretation of the
66
data, this was chosen. The values of vjVOAc were thus derived
value
n]. for K and a were by
and plotted against Values obtained
R K-1/a+l (5.2)
W.
(-7ýa-+j)
j!
I
[TO
Method 1: From and GPC
polymers of the same type (in the same solventz) are obtained, then
R, 2.63 0.65 25 84
Rn 0.682 0.766 23 85
Ri 0.51 0.791 25 19
Rp 1.56 0.708 30 21
j, a/a+l )l
w
(5-5)
['T02 /a+ 1
q.
j
should occur.
97-P9,
Other workprr, using these methods, have found similar
renultn; the nethodn, work with reasonable accuracy for narrow 1':AD
(low D) polymers but tend to give poor results when polymers of
go
broad MWDare used. It has been suggested that the classical
Q
method is better than the GIý method. The accuracy of the latter
The values chosen for K and a for PVOAc have been established
(equation 5-1). Values for D01YStyrene (PS) were taken from the
21 4
literature as a=3.724 and K=1.16 x 10-
branched from linear. Since, for two molecules of the same mole-
lens than the true value. Also as expected, 1-1wincreases (over the
B3M B4M
R, mw Mn -mw
TABIE 5-4: Results from the Ram and Miltz Method on PVQAc.
E-Qlb ftw
SAME
An
E -T 1 b=0.5 b=0.8 b=1.0 b=1.5
5.2.4 Influence of
that the calibrations for the branched species do not deviate from
figures 5.8 to 5.11. These show more clearly not just how branching
does not match the measured one. The possible cause of this is too
given sample above this value, decreases. Thus the values for
.2
8
.8
C. '
I0
--T---, - -- --..
32
t'. i 81
a)
E
D
C
0
a)
2= Z:
1 .5
CDP
r, 4D
FKpre 58- ColibrOtIon Curves qneýated b Rom cnd Miltz
logm
linE
logm
109ml
70
linear
50ý-- 1111111111
520 540 560 580 600,620 640 660 680 700
Figure 5Lll: Calibration Curves generated by Rom and Miltz
Prncpdurp forMo=200000
logm
linec
520 5LO 560 580 600 620 640 660 660 7bO
elution volume
71)
C'M 1-fl]
in orler to fit to If i,. is too hi-h,
V>
calculated measured' 0
pro-, ram.
Conclusions on R/,',l
_5.2-5
The values 7,.,, Y(Ai ) obtained from the ! (/I-!&
of and corresponding
,,
method have been plotted in figures 5.12 and 5.13. Figure 5.12 shows
the log I-, vs. jorý[-q]plot with the derived kark-llouwink equation,
W
and figure 5.13 the shift in values, caused by the Elki procedure.
Though this method has been used favourably both in this work
61,62
and others ,a criticism which has not been previously highligh-
of the polymers.
5.3 BRANCHINGPARK-MEM
calculated via equations 2.47 and 2.49. Table 5.5 gives values of
M
;: bntirp ';.17-.Pint of vpr-, u-, Mw ior Whole Polvmprcý
56 58 rilo 62 6,4
logpw
Figure 5.13- QQmpononof Data for WhQle lymers fro
logn)
OPC and Rom and Miltz
0.51
0,3
/0//13314
0.2 epjý
3c*4stDr
1 0.1
r
BDNLQ,
0"0
750 700 // 650 600
9ý4 vliýwl
ut
-01
-0-3
-0-4-
//
-&5-
71
here that branching through the acetate group (for polymers prepared
0
at 80 C) resulting in hydrolysable branching is far more important
that, though this may be the case, detection of such low degrees of
Assuming that more than 705' of these are hydrolysable leaves about
branch per molecule. It Is hardly surprising that this remains
undetected.
6. '-,
l'OLY(VINYL ALCOHOL)
0.1 ItE,,;UIT'-.
MI'l -)
V("-w V(R, )) in table 6.2. The data from tables 6.1
and
n) - p
6.2 in figures 6.4 to 6.6. As in the case of PVOAC
and are plotted
PVO'H shows best correlation of the data. All the points fall on a
tion (U. C. ) equation 2.55 was used. This required knowledge of the
since the value ofn] for the samples was known (table 6-3). The
LINEAR FRACTIONS BRANCHEDFRACTIONS WHOLEPOMERS
SAMPLE SAMPLE [-qj SAMPLE
Lf 4 1.0`76 Bf 3 1.208 BAKER 0.526
50 eldion vdume 60 70
Figure 63G PChromIcLoLgramsot Whote PVOII
naýmolised
height
. -. % r-4 UN r-4 -4
Ix
N 0 Cý
r-
C\l 100
(7% to
UN CW\ %D to to %I,
%
r-I to U-\ to to
E-4 t; r4
@
-, % CV% cn %lD C\l CY% -4
0 ý
01: 'A 0; U -\ %D \0
m t^ o t- to clý
m %.
4-4 4-4 4-4 4-4 f*4
to
U3
m cn in m cn $4
4)
to m m lZ -7
clov
WN WN
C)
o
ý4:
to xr\ C\d
Cý C; %A
%.
-I LIN .10 .10 \0 ID %0
0
r-4 rl H
Eli
to
00
al
(D
iý
c"D
0.
.1 CD
cl Cl)
1.0
001
co
V)
%A
0c
0 L4
r
>
Q- u
0
4,
E00
D
p" =. O. ý
to0 0
cloo
(10 1
Ll 04
Cý
f-W c"j"
logrl 1 Figure 66 Elution- lume ver5us Intrin5ic V!%os, yýd Pvo
-
02
01.
0 ZIB
-Q2
-04
o oraxnea ircuions
0 whole poyn-*rs
-OLS
mw(PE 0) R4(FVOII)
E-ql
SAMPLE (GPC) (cale. )
from a knowledge of the molecular ý:ei, -,ht anrl structure of the YNAc
calculated data is given in table 6.4 for ione fractionn. 'I"he values
cular weights derived from the P!ýO/P! ýG calibration are too high.
Oince it has been shoicn that the fractions and whole reacetylated
regardless of the true values for hw and P! the ratio of these will
n'
be approximately constant. A re-appralsal of the method of determin-
sidered necessary.
The method also requires knowledge of the constanto for PEO/PýZ, so,
Lf 14 1.37 Bf 8 1.38
Lf 15 1.32 Bf 9 1.29
POLYYM C1T TS
-Houtil
Kx10 a Kx10: 1 a
generally a poor solvent for FVCIIi and most renorted values for #a'
2
lie in the region 0.55-0.65 depending on temperature.
M 1.78 (6-3)
Vi = 12.5 x 10-5 1-1
FVOH, the value of [-q] had been measured, [^QR was obtained from
I'T a+l/a
DIm--
r. (6.4)
1/a
K
The elution volume parameter used was V(2w). From figure 6.7 it can
be seen that the U. C. curves are not coincidental, which offer3 an
does not hold for either one or both polyiners on the column uned.
not been fully revealed) found that PVOH and POO fell below the U. C.
Dolymers with the column gel, was proposed. DubinlO5, usinf: TSKPW
gel also noticed a displaced U. C. Plot for PEO compared with dextrans,
Qn Pj-Qt-.to(-P-E-Q-Qn-d
Figure 67: Unryers(I Cglibr_Qti.
-,
a IMHconsionts)
0-78
0 56 prn-,limp,
.
050ý001'f !ýftkle
064
75
6.4 Coll"CLUgolis
This work has shown that under the conditions employed, branching
weights of less than 200, OCO. Higher molecular weight polymers may
reveal most about the structure of the polymer; a true theta solvent
other physical properties (table 2.1). This has most often been
FVOH still retains a certain amount of acetate groups (i. e. not fully
that, though long chain branches may be absent in FVOH, short chain
branches formed by intramoleculax chain transfer may be present, and
these would not necessarily be detected by viscometry (section 2-6.2).
76
ranges, when it is known that the equation is not linear under such
of K and a are not suitable for use with polymers of broad molecular
weight distribution.
The main effort of this work has been to study long chain
branching In PVOAc and FV011. Though kinetic studies have shoi-rn the
branch type. have also been studied, making it possible to predict the
unknown.
meter which holds for all samples of the same polymer type. As out-
this work the methods of characterisation chosen have been used many
61
times before for, for example, polystyrene (PS)'and polyethylene
commercial polymers, PVOAc included, are most often and easily manu-
factured by the free radical techniques3, leading almost exclusively
Perhaps one of the reasons why little work has been carried out
on PVOAc is the fact that, compared with for example PE, the amount
axe not impossible, though they may be just that to detect. It has
viscometry techniques.
noted that the universal calibration (U. C. ) plots for these polymers
are not coincident. Sufficient data has not been produced here to
for this reason that 1OLY%hydrolysed FVý)Ii was used in this work.
viscometric methods.
molecular weýght polymer was not used in this work for these
reasons.
FF
urthermore, the higher the molecular weight, the higher is the
theN
A + 1)
= nxi-I n(l-x)l. i
pol r OAc r OH
hydrolped
where Mr and N, are the molar masses of the acetate and alcohol
OAc r OH
repeat units.
A-
-[CH2-CH(OCOCH -[CH 2 -CH(OH)]-
a )+ ) (A. 2)
*01.11 pol = n(x(Ill -Zr i"',
r OAc OH olf
44) (A. 3)
= n(42x +
'IPVOAc
But since RPVOAc 86n n= 81-6-
PVOAc (A-4)
then KP., (42x + 44)
86
it (A. 5)
86 Pol 42x + 44
=
'ýVOAC
T.
"Dol 44 (A. 6)
P6 A--
72-
J%PVOAc
(A. 7)
36
rý-2-
"TVAC
M, (A. 8)
8600, (l
degree of hydrolysis = 100(1_x) = 42 _ A,
I"ýVOAc
44 -
79 "'PVOAC
REFZRZNC',
-IS
19. 'J. S. Park and 'J. W. Graessley, J. Pol. Jci. (Pol. Phys. ), 15,71
(1977).
(1967).
24. O.L. Wheeler et al., J- Pol. Sci-, 2,157 (195ý
York (1953).
26. Vinyl and Allied Polymers, Ed. G. Matthews, Gliffe, London (1972).
U. S. A. (1971).
30. Organic Chemistry, '.-iorrison and Boyd, 3rd Ed., p. 677ff, Allyn
39. !-,1. Kurata and W.H. Stockmayer, Adv. Pol. Sci., 2,196 (1963).
40. Molar Mass iieasurements in polymer Science, N. C. Billinghan,
(1967).
Wiley, U. S. A. (1961).
69. Encyclopaedia of Polymer Science and Technology, ,ýd. i. i%. Bikales,
M Science, 65 (1963).
1e&Ln1queF; of Polymer
79. G.R. Williamson and A. Cervenka, Europ. Pol. J., 10,295 (1974).
92.3.114. Agarwal et al., J. Appl. Fol. Sci., 27, ', o. 1,113 (1982).
(10/59).
94. Polymer J., 8, No. 2,196 (1976).
Y.
.. oriý-, hina ot
London (1966).