1058 Macromolecules 1989,22, 1058-1064

Ceric Ion Initiation of Vinylidene Chloride from Poly(viny1

alcohol)
Robson F. Storey* and Leslie J. Gofft
Department of Polymer Science, University of Southern Mississippi,
Southern Station Box 10076, Hattiesburg, Mississippi 39406-0076. Received May 19, 1988;
Revised Manuscript Received August 2, 1988

ABSTRACT The mechanism of ceric ion induced free radical formation on the poly(viny1alcohol) (PVOH)
backbone was studied with and without the presence of vinylidene chloride (VDC) monomer. The grafting
of VDC from PVOH was carried out at various concentrationsof PVOH and constant CeIVconcentration.
Differences in the copolymer products from these reactions are discussed in terms of the possible initiation
sites along the PVOH backbone. CeN oxidizes PVOH most rapidly at 1,2-diolsinherent in PVOH. This reaction
causes cleavage of the PVOH backbone and produces a block copolymer segment rather than a true graft.
Exhaustive oxidation of the 1,2-diolsin the subject PVOH sample, using either Ce" or periodic acid in the
absence of VDC, showed the PVOH contains 1.9 mol % 1,2-diols. The resulting diol-free PVOH, when reacted
with CeIVunder conditions of [CelV]/[OH]<< 1,yielded a pseudo-first-order rate constant of 0.040 min-' at
40 "C. The redox reaction between Ce'" and 1,2-diolswas examined by monitoring the CeIV concentration
and the PVOH molecular weight under conditions of low Ce" concentration. The Ce"/l,Z-diol stoichiometry
was found to vary with temperature, i.e., from 2.0 at 0 "C to 1.5 at 30 "C. Within this temperature range
a pseudo-first-orderrate constant for reaction of CeN with 1,2-diolswas measured with a temperature dependence
defined by k = 6.1 X 10l2exp(-18.3 kcal/RT) min-'.

Introduction characterize the mole fraction 1,2-diol content, A, through

Incorporation of vinylidene chloride (VDC) into coatings the equation
polymers may potentially improve such important prop- A = Mo(1/Mn - 1/Mno) (1)
erties as oxygen and water transmission, fire retardancy,
gloss, and oil and solvent resistance. Aqueous ceric ion where Mn0is the number average molecular weight before
grafting is a particularly attractive route of incorporation degradation, and for PVOH, Mo = 44 g/mol. Equation 1
of VDC due to the ever-increasing interest in water-borne is generally useful for any oxidation process which cleaves
paint and coatings. If an initially water soluble backbone the 1,2-diols of PVOH, including CeIVoxidations. For a
polymer is chosen, it appears possible to produce a self- nonexhaustive degradation, A becomes simply the mole
stabilized, aqueous suspension of graft copolymer product fraction repeat units which were cleaved to lower the
which forms a water-impervious film upon drying. In a molecular weight from M,, to M,.
previous paper1 we described the grafting of VDC from It has been known for many years that Ce" leads to C-C
poly(viny1 alcohol) (PVOH). We found that the incidence bond cleavage a t 1,2-diol structures. Mino et al." dem-
of homopolymer formation was very low, and we estab- onstrated that pinacol was cleaved at the 1,Zdiol by Ce",
lished conditions under which percent weight gain due to and the resulting free radicals were capable of initiating
grafting could be optimized. This work clearly showed that vinyl polymerization. Hintz and Johnson12 studied the
homopolymer formation during ceric ion grafting is very oxidation of cyclic alcohols and cyclic 1,Zdiols by Ce" and
low for systems in which the substrate, in this case PVOH, found that C-C bond cleavage occurred only with the
is by far the strongest reducing agent in the polymerization 1,2-diols. For example, cyclohexanol yielded cyclo-
mixture. Strongly contrasting behavior has been observed hexanone as a near quantitative oxidation product, while
in graft polymerizations of acrylamide,2 where oxidation both cis- and trans-1,2-cyclohexanediolyielded >90%
of the monomer itself brings about homopolymer forma- adipaldehyde.
tion. Ogiwara and Uchiyama5 examined the kinetics of re-
Oxidation of PVOH by Ce'", with3+ and without7 the duction of CeIVby PVOH. They suggested the progress
presence of polymerizable monomers, has been actively of Ce'" reduction can be dividied into three stages, each
studied over the years. It is generally accepted that Ce" stage displaying a different rate of reduction. The initial
forms radical on the PVOH substrate most readily by stage is a very rapid reduction of Ce" which was attributed
oxidation of 1,2-diol units, Le., vicinal hydroxyl groups, to oxidation of 1,2-diol units. This is followed by a second
which occur as a result of occasional head-to-head mono- stage of slower rate, and a third stage of very slow rate.
mer incorporation during polymerization of vinyl acetate.8~~ The authors observed chain cleavage of the PVOH during
Head-to-head linkages in PVOH are reported to occur to stage one, and they showed that the majority of grafting
the extent of 1-2 mol % along the backbone.1° Mino et occurred during stage one as a result of oxidation of 1,2-
al.,7 by reacting CeIVwith PVOH which had been freed of diols. From these results the authors concluded that the
1,2-diols,conclusively proved that isolated hydroxyl groups, copolymer product was principally a block copolymer.
derived from normal head-to-tail monomer addition, were Iwakura and Imai4 and Mino et al.7 also suggested the
also oxidized by Ce'", but a t a much slower rate. formation of block copolymers during ceric ion induced
Early work by Flory and Leutner* demonstrated that grafting from PVOH. However, Mino et aL7 argued that
exhaustive periodic acid oxidation of the 1,2-diolsin PVOH only under conditions of a very large excess of PVOH, Le.,
caused cleavage of the C-C bonds which connect the vi- [1,2-di0l]/[Ce'~]> 10, would radicals be produced only a t
cinal hydroxyl groups. The number-average molecular the ends of the fragmented PVOH, leading to block co-
weight of the PVOH after degradation, M,, was used to polymers. Under normal grafting conditions, a mixture
of block and graft copolymers would be expected.
Current address: Plastics Technology Division, General Electric Early investigations by Duke and c o - w o r k e r ~ con-
~~J~
Company, Highway 69 South, Mt. Vernon, IN 47620. cerning the mechanism of oxidation of 1,2-diols by ceric
0024-9297/89/2222-l058$01.50/0 0 1989 American Chemical Society
Macromolecules, Vol. 22, No. 3, 1989 Ceric Ion Intiation of Vinylidene Chloride 1059

ion led to the general acceptance of the intermediacy of into reagent grade acetone. VDC (Dow Chemical) was washed
a coordination complex between CeN and oxidizable diols. with 5% NaOH, washed with water, and dried over magnesium
Monoalcohols such as methanol, ethanol, and 2-propanol sulfate. It was then vacuum distilled such that the temperature
are also believed to form complexes with CerVand to be of the monomer never exceeded 26 "C. Ceric ammonium nitrate
(CAN) (Fisher Scientific) was used as received. A standard
oxidized in a similar manner.' In fact, the use of such solution of 0.156 M CAN in 0.1 N HN03was prepared for grafting
complexes for the qualitative colorimetric determination reactions. All other chemicals were reagent grade and used as
of alcohols was advanced very early by Duke.15 The com- received.
plexes have been shown to be rapidly established in an Preparation of Diol-Free PVOH. Approximately 7.5 g of
acidic aqueous medium and are readily observed by using PVOH ( M , = 74 000) was dissolved in 100 mL of RO water and
UV/visible spectroscopy.16 Although CerVhas a coordi- charged into a three-neck flask. A 10% solution of HJ06 (periodic
nation number of six and could thus conceivably coordi- acid) (Fisher Scientific) was prepared by using RO water, and
nate with six monoalcohols or three 1,2-diol~,'~J~ it has 10 mL of this solution was charged to the PVOH solution while
been demonstrated that the complexes involve one mole- stirring. Stirring was continued for 60 min under COz. The
cule of diol (or monoalcohol) and one Ce1V'2J6and that polymer was precipitated in excess acetone and collected by
filtration. The polymer was then dissolved in 100 mL of RO water
the ceric ion complex with a diol is a chelated complex. and allowed to stir with 20 mL of a 2% solution of sodium bo-
According t o the well-accepted m e c h a n i ~ m , ' J ~the
-~~ rohydride (Fisher Scientific) prepared from 1 N NaOH. After
rate-determining step of the oxidation is the unimolecular 15 min the pH of the mixture was adjusted to 1.0 with 0.1 N HC1.
disproportionation of the complex C to yield cerous ion, The polymer was precipitated in excess acetone and dried in a
a proton, and a free radical on the alcohol substrate: vacuum oven at room temperature. Finally, the polymer was
extracted in a Soxhlet extractor using a water-2-propanol azeo-
OH OH
trope for 100 h to remove borohydrides.
-c-H
I + CeIV 'x, c -c*
I + CeIII + H+ (2) Graft Polymerization. In a typical graft polymerization,20
mL of an aqueous PVOH solution of appropriate concentration
I I was charged into a chilled (0 "C) glass ampule. Then, to the
If the alcohol is a 1,2-diol, the carbon-carbon bond between ampule were added, prechilled,0.1 N nitric acid solution (2.0 mL),
the hydroxyls cleaves to yield an aldehyde or ketone and 2.4 g of VDC, and 0.156 M CAN solution (2.0 mL). This produced
a two-phase reacting system with a 23-mL aqueous phase sus-
a free radical: pended over -2 mL of VDC. The ampule was subjected to three
OH OH OH 0 freeze-pumpthaw cycles and sealed under vacuum. The reaction
I I
-c-C- + Ce" ec
K
- k 11 + Ce"'
-c' I + C- + H+
was carried out at 40 "C for 3 h, during which each ampule was
agitated every 20 min. The products were precipitated in
I 1 I I (3) methanol, collected by filtration, and dried to constant weight.
Oxidation of PVOH (Method I-Data for Figure 3 Only).
The fate of the free radical in both cases is supposedly a A number of ampules were charged with 10 mL of a 20 g/L
fast oxidation by another ceric ion: aqueous solution of PVOH. Nitrogen was bubbled through each
ampule for 30 min, and then 0.5 ml of 0.156 M CAN solution was
OH 0 charged to each ampule. The ampules were sealed and placed
-c*
I + ceIV 2
k
-c-
11 + Ce"': + H + (4) into a 40.0 "C water bath and reacted for varying times. The
reactions were stopped by pouring the contents of each ampule
I into excess methanol. The polymers were precipitated again in
The ceric ion induced grafting from PVOH thus emerges acetone and dried in a vacuum oven overnight at room temper-
as a complex process in which vinyl polymerization of ature.
grafted chains occurs simultaneously with degradation of Oxidation of PVOH (Method 11-Including Determina-
Concentration). Into a three-neck flask was charged
tion of CeTV
the PVOH backbone. Oxidation of 1,2-diols and isolated 100 mL of a 22 g/L aqueous solution of PVOH. While stirring,
hydroxyls both occur, but the former is much the faster 10 mL of 1 N sodium nitrate (Aldrich Chemical)was added. The
reaction. Oxidation of 1,Zdiols cleaves the PVOH chain system was purged with nitrogen and equilibrated to the desired
and yields a radical at the fragment end. Thus the graft reaction temperature. The 0.156 M CAN solution was equilibrated
copolymer product is best viewed as a mixture of block and to the reaction temperature and injected (5 mL) into the reaction
graft copolymers. vessel to yield [Ce'"] = 6.8 mmol/L. The reaction was allowed
In an earlier paper' we described ceric ion initiated to proceed under a nitrogen blanket. Samples were withdrawn
grafting reactions of VDC from PVOH. The reaction is at recorded times and quenched in an excess of 0.05 N ferrous
heterogeneous, consisting of a n acidic aqueous phase ammonium sulfate (Fisher Scientific) solution in 0.5 N sulfuric
containing CerVand PVOH and an organic phase con- acid. The excess ferrous salt was titrated with 0.05 N ceric sulfate
(Aldrich Chemical),using o-phenanthrolineindicator. The ferrous
sisting of VDC monomer. The product is a suspension of salt solution was standardized before use with 0.05 N potassium
PVDC/PVOH block and graft copolymer in a n aqueous permanganate (Aldrich Chemical),and the ceric sulfate solution
medium. was standardized with the ferrous salt solution.
This paper describes a detailed kinetic investigation of For pseudo-first-orderreactions the ceric ion concentration was
our polymerization system, undertaken to better define the 0.34 mmol/L. The same procedure was employed with appro-
nature of the initiation reaction and to understand the role priate adjustments in titrant concentrations.
of the various alcohol structures during graft initiation. Molecular Weight Determinations. Molecular weight de-
Studies were carried out on both virgin PVOH and PVOH terminations were performed by using a Knauer membrane (MO)
which had been freed of l,2-diol units (diol-free PVOH). and vapor pressure osmometer (VPO).
The rate of consumption of CerVwas monitored by titra- Results and Discussion
tion, and the M,, of the PVOH was monitored during ox-
idation to characterize the extent of 1,Zdiol cleavage, using Characteristics of PVDC Grafting from PVOH. In
eq 1,which had occured at any time during the reaction. our initial paper describing VDC grafting from PVOH,'
we indicated that the percent weight gain due to grafting
Experimental Section (70wt gain = (wt of graft product - wt of PVOH)/wt of
Materials. PVOH (Air Products) of M , = 74000 was 96% PVOH) increases with an increase in PVOH concentration
hydrolyzed according to the manufacturer. It was purified by at a constant concentration of ceric ion. We have since
dissolving in reverse osmosis (RO) water at 90 "C and precipitating performed more extensive investigations covering a broader
1060 Storey and Goff

Table I
Ceric Ion Grafting of VDC from PVOH
[PVOH]," [PVOH]," I
I
run g/L wtgain, % run g/L wtgain, %
1 17.4 65 7 9.6 118
2 14.8 85 8 7.8 89
3 13.0 60 9 6.1 136
4 13.0 70 10 3.5 263
5 12.2 114 11 1.7 250
6 10.4 117
temp = 40.0 "C VDC = 2.4 g (2.0 mL)
reactn time = 3 h [CeWla= 6.8 mmol/L
Concentrations are based on 23-mL aqueous volume. Total
(I

reaction volume including monomer = 25 mL.

250
i
-1
*
*

** *
* *
** *

Figure 1. Percent weight gain during grafting of VDC from

PVOH, as a function of [PVOH]at constant initial [Cerv] = 6.8
mmol/L (concentrations based on 23 ml aqueous volume).
Temperature = 40.0 "C. Reaction time = 3 h. VDC charge =
2.4 g.

range of PVOH concentrations, and we have found that,

in fact, percent weight gain is relatively constant within
the narrow range of conditions reported earlier but in-
creases dramatically a t very low [PVOH]. Table I and
Figure 1 outline the results.
There was a noticeable difference in the appearance of
the graft product formed at low concentrations of PVOH.
At [PVOH] < 4.0 g/L and [CerV]= 6.8 mmol/L, a hard
precipitate settled from the aqueous layer, which, when
filtered and dried, formed a powder. Higher concentra-
tions of PVOH (5.0 g/L and higher) yielded stable sus-
pensions of copolymer product, which upon removal of
water formed films.
It was of interest t o determine the extent of homo-
polymer formation, if any, under these varius reaction
conditions. The powdery product obtained from 4.0 g/L
PVOH and 6.8 mmol/L CeIV (which is approximately 4
times the ceric ion concentration for normal grafting re-
actions) was extracted for 24 h at 90 "C with RO water and
24 h at 70 "C with N-methylpyrrolidone in a Soxhlet ex-
tractor. In neither case was PVOH or PVDC homopolymer
extracted from the graft copolymer product.
As discussed earlier, the PVOH/CeIV system is com-
plicated by the existence of two types of oxidizable hy-
droxyls. One type is the isolated hydroxyl which results
from normal head-to-tail addition of vinyl acetate and
represents over 98% of the hydroxylic sites along the
Macromolecules, Vol. 22, No. 3, 1989 Ceric Ion Intiation of Vinylidene Chloride 1061

(shown as a narrow, shaded band in Figure 2), the

[Cew"]I[1,2-diol] ratio is approximately 1:l. Since reaction
of Ce is faster with 1,2-diols, this system is thought to
produce predominantly block copolymers with a lesser
percentage of grafted chains. As the [CeN]/[1,2-diol] ratio *Or------
60
is decreased further (moving to the left in Figure 2), the
small ceric ion population reacts almost exclusively with m
1,2-diols, producing solely block copolymers. Since 1,2- I
0 -
diols are in large excess, one also obtains pseudo-first-order 4 -
kinetics with respect to CeIV. 40.-
The difference in physical properties of these two sys- X I
tems may be rationalized in terms of the number of grafted I*
or block segments and by their relative length. The high 2 :
[CelV]/[PVOH] system (system 1) produces more but
20 -
shorter chains of PVDC. The lower [CeN]/[PVOH] sys-
tem (system 2) produces fewer but longer PVDC chains, *
which in fact are predominantly block, resulting in con- -1
siderably less branching. With less cleavage of the PVOH * *
chain in system 2, the PVOH segments theoretically would o * 0 o *
0 , , # , I > , ,
, / , , , / / , , , / , / , , , , , / , I , , , , / , , , / , ,

be larger. 0 10 20 30 1

The block and graft copolymer produced in system 2 Time (min)

formed slightly yellow, brittle films. Film formation may Figure 3. Exhaustive oxidative degradation of PVOH by Cew
be attributed to fewer attached segments of PVDC, re- as a function of time: (*) PVOH (0)diol-free PVOH. [PVOH]
ducing the tendency for extensive PVDC crystallization. = 19 g/L. [Cew] = 7.4 mmol/L. Temperature = 40.0 "C.
Copolymers produced in system 1 displayed properties
very similar to PVDC homopolymer, indicating high sumption of each diol requires at least two ceric ions. As
crystallinity. In fact, as Figure 1 shows, copolymers pro- shown earlier in eq 3 and 4, one ceric ion is consumed
duced a t low [PVOH] are, by weight, predominantly during disproportionation of the complex and generation
PVDC. It was also observed that precipitation or phase of a radical (eq 31, and another is consumed during oxi-
separation (formation of suspension) occurred earlier in dation of the radical in a subsequent fast reaction (eq 4).
system 1than in system 2. This would also suggest higher Major support for the operation of eq 4 or some similar
PVDC content and higher crystallinity in system 1. unimacromolecular termination mechanism comes from
Oxidations of PVOH in the Absence of Monomer. the absence of cross-linking and gelation in ceric ion in-
The M , of the unoxidized PVOH was determined to be duced grafting reaction^.^ However, the data in Figure 3
74 OOO by MO. Periodic acid was used to cleave the PVOH shows no indication of Ce" depletion, as would be expected
chain at each 1,2-diol group to produce a diol-free PVOH. if each 1,2-diol consumed two ceric ions. The molecular
After reaction, the M, of the diol-free PVOH was found weight (Figure 3, asterisks) was observed to fall to the
by VPO to be 2200. When a similar oxidation was carried limiting value of 2200, which by the periodic oxidation
out by using CeN, the M , of the PVOH was also found to reaction was shown to represent cleavage of 100% of the
be 2200. 1,Zdiols. Thus, it may be concluded in this case that each
This result indicates that CeIV oxidation of PVOH, 1,2-diol consumed no more than 7.4/8.4 = 0.88 ceric ions,
which leads to chain cleavage, occurs exclusively and but perhaps considerably less. The stoichiometry cannot
quantitatively at the 1,2-diol structures. Although oxi- be known precisely from this data alone because an un-
dation of isolated hydroxyls also occurs, it does not lead known concentration of CeN remained after depletion of
to chain cleavage. On the basis of these results, the mole all the 1,2-diols. Thus, in view of these results, it seems
percent 1,2-diol groups, A, in the subject PVOH was cal- very likely that some other unimacromolecular termination
culated to be 1.9% using eq l given by Flory and Leutner.* mechanism is operative, in addition to or instead of eq 4.
This agrees well with Mino et al.," who found A = 1.4% A bimacromolecular radical disproportionation mecha-
for PVOH prepared in their laboratory. nism is conceivable, in view of the structure of the radicals
involved:
The kinetics of 1,2-diol oxidation by CeN was monitored
by observing the resulting decrease in M,,. The first set
of conditions examined were typical of those used for ef-
ficient grafting, i.e., 40 O C , with 1,2-diols (8.4mmol/L) only
in slight excess to Ce" (7.4mmol/L), but isolated hydroxyl
groups in large excess. This ratio of CeIV to 1,2-diols, L H _I H (6)

hereafter referred to as "grafting conditions" was found However, the likelihood of this is not very great for two
to produce optimum weight gain in our earlier studies.* reasons. First, Hintz and Johnson12 found that at 0 OC
The data in Figure 3 (asterisks) demonstrate that during both hydroxyls of a 1,2-diol pair were converted to al-
the oxidation reaction, a precipitous decrease in M , occurs dehyde; although it is quite relevant that they anticipated,
within the first few minutes. Thereafter the M, undergoes and in fact found, that 2 mol of Ce" were reduced per mole
a slow, asymptotic approach to the final M , of about 2200. of 1,2-diol oxidized. This is contrary to the results pre-
Similar behavior in the presence of periodic acid indicates sented in Figure 3, but it should be noted that their re-
that chain cleavage occurs a t the 1,2-diols. Isolated hy- actions were run in aqueous perchloric acid solution, and
droxyl functionalities are oxidized at a slower rate but do more importantly, a t lower temperature. Second, this
not cause cleavage of the PVOH backbone. mechanism does not adequately explain the results; i.e.,
These results bring into question the stoichiometry of it cannot explain the consumption of less than 1 mol of
the reaction between CeIVand 1,2-diols. The traditional CeIv/mol of diol oxidized.
view of ceric ion oxidation of PVOH7J2-14holds that con- A different radical consumptive reaction that reconciles
1062 Storey and Goff Macromolecules, Vol. 22, No. 3, 1989

Table I1
Temperature Dependence of CeIVOxidation of 1,2-Diols
CeN/l,2-diol reactn final M,,
temp, "C stoichiometry time, min g/mol q, mmol/L 0
0
0.0 2.0 161 44 500 0.17
15.0 1.9 28 43 200 0.18 Ll *
23.5
30.0
1.7
1.5
12

[PVOH] = 19 g/L
5
42 000
39 500
0.20
0.22 "i
i=
L l
0

[Ce"] = 0.34 mmol/L

total reaction vol = 0.115 L

IC. I
75 4
I
1
*
I l / ' l l ' l 1~ 1 ~1 1 1 1 1 1 ~ f I ~ " 1 ~ I ~ I ~
0 '0 20 30 40
TIME (min)
Figure 5. [CeN]as a function of time under grafting conditions:
(*) PVOH; (0)diol-free PVOH. [PVOH] = 19 g/L. [Ce"] = 6.8
mmol/L. Temperature = 40.0 "C.

35
n
cl
\
0.1 Po *
0
0 *
0 20 40 60 13

Time (min)
Figure 4. Partial oxidative degradation of PVOH by Ce", under
;
E
0

pseudo-first order conditions, as a function of time: (0)

(0)
30.0 "C;
23.5 "C; ( a )15.0 O C ; (A)0.0 "C (inset). [PVOH] = 19 g/L.
[CetV] = 0.34 mmol/L.
v
i * O

the latter result is a reduction of the free radical inter-

mediate by Ce"' as follows:
OH

*CH*
I + Ce"' - H+
*CH~-OH + ~e'" ( 7)
0 UI a0
Time (mm)
120 10

0001 , ! ! I / l!!85l!ll
I 1, 1 1 I I I I I I I I I I I I I I l l I I ( I ! , l I l , 1 '

I t is conceivable that this reaction is operative a t high

temperatures, e.g., 40 "C, and is responsible for the low
CeN/ 1,Pdiol Stoichiometry observed. As the temperature
is lowered, this pathway is possibly suppressed, and Hintz
and Johnson,12 working a t 0 "C, would indeed find a
CeIV/1,Zdiol stoichiometry of 2.
Further experiments were conducted to explore the
temperature dependence of the Ce'"/ 1,2-diol stoichio-
metric ratio. A series of oxidation reactions were run for
which the temperature was varied from 0 to 30 "C, and
the CeN concentration (0.34 mmol/L) was approximately
1/25 the concentration of 1,2-diols (8.4 mmol/L). Under
these conditions the concentration of 1,2-diols remains
virtually constant, pseudo-first-order kinetics should be
observed, and the CeIVshould react exclusively with the
1,2-diols. In Figure 4 and Table I1 it can be seen that
under these conditions there is also a significant loss in
PVOH molecular weight due to backbone cleavage. Sig-
nificantly, the molecular weight asymptotes at long times
are greater than 2200, indicating incomplete oxidation of
the 1,2-diols, and are dependent upon reaction tempera-
ture.
The cause for incomplete oxidation of 1,Zdiols observed
under these conditions is depletion of available CeN. The
Flory equation (eq 1) allows calculation of the millimole
Macromolecules, Vol. 22, No. 3, 1989 Ceric Ion Intiation of Vinylidene Chloride 1063

Table I11 -0 *
Pseudo-First-OrderRate Constants for Reaction of Ce'"
with Excess l,2-Diols 4
*
temp, "C k, min-' temp, "C k, min-'
0.0 0.013 30.0 0.48 - 1 1-1
15.0 0.085 40.0 0.89 4 *
23.5 0.20
Ce" and yield an overall requirement of less than one ceric
ion per 1,2-diol. This hypothesis is presently under in- *
vestigation.
The kinetics of oxidation of PVOH by CerVwas also
followed by monitoring the disappearance of ceric ion by 4
t i t r a t i ~ n . ' The
~ results obtained under grafting conditions
and under pseudo-first-order conditions are shown in
Figures 5 and 6, respectively. Under grafting conditions,
[Ce"] zz [1,2-diol],but isolated hydroxyl groups are in large
excess. This should reduce the reaction with isolated
hydroxyls to a pseudo-first-order reaction with respect to
ceric ion.
-4i 4
J
*

The curve for PVOH in Figure 5 (asterisks) can be di-

vided into two processes. At first, there is a'rapid decrease Figure 7. Arrhenius plot for reaction of Ce" with PVOH 1,2-diols
in CeIVconcentration due to reaction with 1,2-diols7fol- under pseudo-first-order conditions.
lowed by a slower, first-order reaction with isolated hy-
droxyls. The break in the curve occurs at [Ce'"] = 1.4 subsequently treated with sodium borohydride.
mmol/L. If it is assumed that this point represents de- As shown in Figure 3 (circles) the molecular weight of
pletion of 1,2-diols, the Ce1"/1,2-diol stoichiometry was diol-free PVOH was unaffected during oxidation by CeN.
-0.65. The pseudo-first-order rate constant for reaction This reaffirms that oxidation of isolated hydroxyls does
of CeN with isolated hydroxyls was measured to be 0.040 not lead to chain cleavage and therefore would initiate a
min-'. The initial slope of the rapid decrease portion of true graft copolymer chain.
the curve yielded a value of 0.21 min-'; however, this re- The pseudo-first-order rate constant for oxidation of
action is not first-order since it exhibits an accelerating diol-free PVOH with Ce'" was determined to be 0.040
downward curvature. min-* from the slope of the diol-free PVOH curve in Figure
When the oxidation reaction was run under pseudo- 5 (circles). This value is in agreement with the rate con-
first-order conditions, as shown in Figure 6, the downward stant for the oxidation of the commercial PVOH after all
curvature was not observed a t 40 OC, but it was evident the 1,Zdiols had been oxidized (Figure 5, PVOH curve for
at lower temperatures toward the end of the reaction. The time > 10 min). Also, this confirms that the second, slower
precise cause of this rate acceleration is unknown at this rate is due to the isolated hydroxyls on the PVOH back-
time. However, the plots are quite linear during the early bone, and it demonstrates that the majority of 1,2-diols
stages of the reaction, yielding pseudo-first-order rate are oxidized within the first 10-15 min of the reaction.
constants which are listed in Table 111. These were ob- This is in agreement with previous work.5
served to follow an Arrhenius temperature dependency,
as depicted in Figure 7. The activation energy and Conclusion
preexponential factor were found to be 18.3 kcal/mol and Dramatic differences in the PVOH/PVDC copolymer
6.1 X 10l2min-l, respectively. The value of 12 = 0.20 min-' products were observed as the [Cerv]/[1,2-diol] concen-
obtained unknown 23.5 "C is very reasonable since Mino tration ratio was varied over a large range. These differ-
et ala7obtained a value of 0.11 min-' at 20 "C. As a further ences may be explained in terms of the competition for
note, if one compares the initial slopes of the 40 OC reac- Ce'" between 1,2-diols and isolated hydroxyls. The final
tions under pseudo-first-order conditions (0.89 min-') and M,, of the PVOH segments is dependent on the mole
grafting conditions (0.21 min-l), it is seen that the rate of percent 1,2-diol content and the initial concentration of
depletion of CeN is much greater under pseudo-first-order CeN. Under conditions of high [CeN], the product is en-
conditions, i.e., when [Cerv]/[1,2-diols] << 1.0. This sug- visioned as a small PVOH segment with numerous block
gests that under grafting conditions, complexation of CeW and grafted segments. This causes the product properties
occurs with both 1,Zdiols and isolated hydroxyls, and this to be dominated by the PVDC segments. In systems where
effectively limits the CerVavailable for reaction with 1,2- [CeN] is lower, the PVOH segment is believed to remain
diols. As pointed out earlier, reaction with both 1,2-diols larger, and the product is predominantly a block co-
and isolated hydroxyls should lead to a mixture of block polymer. Because this product retains a high weight
and graft copolymers. However, as the [CeIV]/[ 1,2-diol] fraction of PVOH, it displays unique properties that are
ratio is lowered, oxidation tends to occur more and more different from either homopolymer.
exclusively with 1,2-diols7which produces block rather than The oxidation of PVOH occurs most readily at the
graft copolymers. Thus, if one desires an exclusively block 1,2-diol linkage of the PVOH backbone. The reaction of
copolymer product, the [CeW]should be kept very low, for CerVwith 1,2-diols was studied by observing the decrease
example, by a starve-feed mode of addition of CerVto the in M , of PVOH. Exhaustive oxidation by either Ce'" or
reaction mixture. periodic acid led to PVOH fragments of M,, = 2200. This
Oxidation of Diol-Free PVOH. For comparison, result allowed calculation of the mole percent 1,Z-diol
diol-free PVOH ( M , = 2200) was subjected to the same content, which in this case was 1.9 mol %. It also showed
reactions as was virgin PVOH. Diol-free PVOH was pro- that a t relatively high temperatures, Le., 40 "C, the
duced by periodic oxidation of PVOH,7 and to ensure the CeIv/1,2-diol stoichiometric ratio was actually less than
absence of terminal aldehyde groups, the polymer was 1.0.
1064 Storey and Goff Macromolecules, Vol. 22, No. 3, 1989

This result lead to further reaction studies in which CeIv A = MO(l/Mn - l/Mno)
was chosen to be the limiting reagent. Under these con-
ditions of low ceric ion concentration the stoichiometry of where
CeIVto 1,2-diols was observed to increase from 1.5 at 30 M, = 44 g/mol
"C to 2.0 at 0 "C. The value of 2.0 at 0 "C is in agreement
with prior work of others.12 These results suggest that the Mn0 = 74000 g/mol
traditional view of Ce" oxidation of 1,2-diols,i.e., the fast and a t 23.5 OC
oxidation of previously formed free radicals by additional
ceric ion (eq 4),must be modified to include some other M, = 42000 g/mol
radical-consumptive mechanism operative at higher tem- multiplied by the molar concentration of PVOH repeat
peratures. It is suggested that this might involve a fast units in the oxidation reaction:
reduction of free radicals by Ce'" (eq 71, which is formed
in the initial step of the reaction. 4 = A([PVOHl /Mo)
From CeIVtitration data under "grafting conditions", a = 0.20 mmol/L
majority of the CeIVwas shown to be depleted early in the
Registry No. PVOH, 9002-89-5; CAN, 10139-51-2; (V0H)-
reaction due to oxidation of 1,2-diols. A second rate of (VDC) (block copolymer), 117556-92-0; (VOH)(VDC) (graft co-
oxidation due to isolated hydroxyls was seen thereafter. polymer), 108917-82-4; VDC, 75-35-4.
The pseudo-first-order rate constants for this second rate
and that of diol-free PVOH are the same, Le., 0.040 m i d References and Notes
a t 40 "C. Under conditions of low [CeIV],titration data (1) Storey, R. F.; Sudhakar, D.; Goff, L. J. J. Makromol. Sci.,
yielded a pseudo-first-order rate constant with respect to Chem. 1987, A24(9), 1051-1064.
ceric ion with a temperature dependence defined by k = (2) McDowall, D. J.; Gupta, B. S.; Stannett, V. T. Prog. Polym.
Sei. 1984, 10, 1-50.
6.1 X 10l2exp(-18.3 kcal/RV m i d . The value of k = 0.20 (3) Mino, G.; Kaizerman, S. J. Polym. Sci. 1958, 31, 242.
min-l at 23.5 "C is in reasonable agreement with Mino et (4) Iwakura, Y.; Imai, Y. Makromol. Chem. 1966, 98, 1-6.
al.,' who found a value of 0.11 min-l a t 20 "C. An accel- (5) Ogiwara, Y.; Uchiyama, M. J. Polym. Sci., Polym. Chem. Ed.
erating downward curvature was observed in most of the 1969, 7, 1479-1488.
first-order plots toward the end of the reaction; this phe- (6) Narita, H.; Okamoto, S.; Machida, S. Makromol. Chem. 1969,
125, 15-23.
nomenon is poorly understood at this time. (7) Mino, G.; Kaizerman, S.; Rasmussen, E. J . Polym. Sci. 1959,
These results indicate that the redox reaction between 39, 523-529.
CeIV and PVOH is indeed a complex one. The reaction (8) Flory, P. J.; Leutner, F. S. J.Polym Sci. 1948,3(6), 880-890.
stoichiometry, and therefore the mechanism, has been (9) Molyneaus, P. Water-Soluble Synthetic Polymers: Properties
and Behavior; CRC: Boca Raton, FL, 1984; Vol. I.
shown to change dramatically over a relatively narrow (10) Finch, C. A. Polyvinyl Alcohol; Wiley: New York, 1973.
temperature range. However, the reaction conditions can (11) Mino, G.; Kaizerman, S.; Rasmussen, E. J. Am. Chem SOC.
be controlled to optimize the physical properties of the 1959,81, 1494.
resulting PVOH/PVDC copolymer, and if proper condi- (12) Hintz, H. L.; Johnson, D. C. J. Org. Chem. 1967,32,556-564.
(13) Duke, F. R.; Forist, A. A. J. Am. Chem. SOC. 1949, 71,
tions are employed, it is conceivable that predominantly 27962792.
block copolymers can be produced. (14)
, , Duke. F. R.: Bremer. R. F. J. Am. Chem. SOC. 1951. 73. I .

5179-5181. '
Acknowledgment is made to the donors of the Petroleum (15) Duke, F. R.; Smith, G. R. Ind. Eng. Chem., Anal. Ed. 1940,12,
Research Fund, administered by the American Chemical 201.
Society, for the support of this research. (16) Young, L. B.; Trahanovsky, W. S. J. Am. Chem. SOC.1969,
91(18), 5060-5067.
Appendix A (17) Mino, G.; Kaizerman, S.; Rasmussen, E. J.Polym. Sci. 1959,
38, 393.
The millimole chain cleavages per liter, q, is given by (18) Saha, S. K.; Choudhury, A. K. J. Polym. Sci., Polym. Chem.
the fraction of chain repeat units cleaved, A Ed. 1971, 9, 1505.