plik

For Review Only

STUDY OF THERMAL DEGRADATION OF PVC PLASTICIZED FORMULATIONS

COSTABILIZED WITH D-SORBITOL AND TRIPHENYL PHOSPHITE

Pilar Haro-Gutiérrez, Oscar Rodríguez-López, Jesus Arellano, Eduardo Mendizábal, Carlos F. Jasso,

Martín Arellano and Luis Javier González-Ortiz

Centro Universitario de Ciencias Exactas e Ingenierías, Universidad de Guadalajara, MEXICO

Abstract

PVC is generally degraded during processing,

producing  several  undesirable  effects.  In  this  work,
mixtures  of  calcium  and/or  zinc  stearates  and  epoxidized
soybean oil are used as stabilizers. In addition, to improve
overall  stability,  D-sorbitol  or  triphenyl  phosphite  were
added  as  costabilizers.  The  formulation  composition  was
systematically

varied

considering

the

following

parameters: a) presence of epoxidized soybean oil, b)
CaSt

/ZnSt

ratio and, c) presence and type of costabilizer.

Thermal  stability  was  followed  during  isothermal  heating
by  determining:  a)  the  accumulation  rate  of  some
conjugated  polyenes  and,  b)  the  changes  in  the  tensile
properties.

Introduction

Unless it is stabilized, PVC is degraded during

processing.  Such  thermal  degradation  is  the  result  of  a
process  called  ‘‘zipper  dehydrochlorination’’,  which
generates  polyene  sequences  in  polymer  chains  that  may
produce  an  undesirable  color  in  the  material  [1].  In
addition,  since  polyenes  are  highly  reactive,  they  may
react  to  form  crosslinked  polymer  chains  [1],  as  well  as
low  molecular  weight  molecules  [1-2].  As  a  consequence
of such secondary reactions, the mechanical behavior and
the  color  of  the  formulations  [1]  may  be  importantly
modified. It is generally accepted that stabilizers may react
with  labile  chlorine  atoms  in  PVC  chains  (preventing
further  dehydrochlorination)  and/or  react  with  the  HCl
generated  by  the  degradation  process  (which  accelerates
the degradation process) [1,3]. K, Ca and, Ba carboxylates
are  mostly  HCl  scavengers  [1].  However,  Zn  and  Cd
carboxylates  are  able  to  scavenge  HCl  and,  also,  to  react
with labile chlorine atoms [1]. An undesirable effect of the
stabilizing action of zinc stearate (ZnSt

) is the production

ZnCl

which

can

promote

the

sudden

dehydrochlorination of PVC chains. [1, 4-5] However, it
has been claimed that such sudden process occurs only
after the ZnCl

concentration reaches a certain level [5]; in

such process, ZnCl

is consumed [4]. Besides, it has been

experimentally  demonstrated  that  formulations  prepared
with  mixtures  of  Zn  and  Ca  carboxylates  show  a
synergistic  stabilizing  action  [6-9],  mainly  due  to  the  Ca
carboxylates, which act as ester-exchangers with ZnCl

[1,

6-7]. Epoxidized compounds are recognized as HCl
scavengers [1, 8], but can also participate in other

stabilization reactions, where they react simultaneously
with HCl and ZnSt

[8]. In the catalytic presence of ZnCl

epoxidized  compounds  react  with  such  chloride  to
produce  a  chemical  compound  that  is  able  to  remove
chlorine  atoms  in  PVC  chains  through  an  etherification
reaction [1].

Costabilizers are added to PVC formulations to

improve the efficiency of the stabilizers; some of them are
D-sorbitol [1] and triphenyl phosphite [1]. The stabilizing
mechanism  of  such  costabilizers  has  been  partially
reported  [1],  being  clear  that  their  action  is  dependant  of
the  chemical  characteristics  of  system.  Although  the
degradation-stabilization  mechanism  of  formulations
containing  PVC  has  been  extensively  studied,  there  are
few  studies  that  consider  the  combined  effect  of  three  or
more  components  in  industrial  grade  formulations.
Therefore,  in  this  work,  the  formulation  composition  was
systematically

varied

considering

the

following

parameters: a) presence of epoxidized soybean oil, b)
CaSt

/ZnSt

ratio and, c) presence and type of costabilizer.

The accumulation rate of different conjugated polyenes (6
≤

n ≤ 20) and the changes in tensile properties produced

by isothermal heating after processing were determined.

Experimental

Materials

PVC resin from Policyd S.A. de C.V (Vinycel G-30;

K-Fikenstcher  value  of  70)  was  used  to  prepare  the
formulations.  Di  (2-ethyl  hexyl  phthalate)  (DEHP;  purity
of 99.5%) was purchased from Síntesis Orgánicas S.A. de
C.V. ESO was acquired from Resinas y Materiales S.A. de
C.V  (Pantopox;  oxirane  number  6.95%  oxygen).  CaSt

(CaO essay 9.0-10.5%) and ZnSt

(ZnO essay 12.5-14%)

were  purchased  from  Alfa  Aesar.  D-sorbitol  from  Alfa
Aesar (98 wt %) and triphenyl phosphite from Aldrich (97
wt  %)  were  used  as  costabilizers.  Two  different
DEHP/ESO  ratios  were  used:  45/0  or  45/6  (phr/phr).  For
each  one  of  the  DEHP/ESO  ratios,  the  following
CaSt

/ZnSt

ratios (phr/phr) were used: 0.0/1.0, 0.2/0.8,

or, 0.4/0.6; in all formulations the total content of stearates
was  1.0  phr.  For  each  ESO  concentration,  the  three
following  series  were  considered:  a)  without  costabilizer,
b)  D-sorbitol  (0.25  phr)  and, c) triphenyl phosphite (0.25
phr).

Page 2 of 7

ANTEC

For Review Only

Sample preparation

The samples to be post-processed were prepared as

follows:  (a)  dry  blending  the  components,  (b)  pelletizing
the  dry-blend  (twin-screw  extruder  Leistritz  276L/32D)
and,  (c)  extruding  the  pellets  (to  obtain  ribbons).  The
samples  were  heated  in  a  forced  convection  oven
(FRELAB) at 150±5°C (0.5, 1, 1.5, 2 or, 3 h).

Characterization

The changes on the stress-strain behavior of

degraded  samples  (ASTM  D638  type  II  specimens)  were
determined  using  a  Universal  Testing  Machine  from
United  (SFM10),  characterizing  the  samples  at  room
temperature  and  with  a  cross-head  speed  of  50  mm/min.
The  concentration  of  polyenes  was  determined  by  UV
spectrophotometry  (UV-visible  Spectrophotomer  Perkin
Elmer Lambda 25).

Results and discussion

The stress-strain curves showed a monotonic stress

increase  with  deformation;  therefore,  the  Young  modulus
(E), ultimate stress (σ) and, deformation at break (ε) were
considered as representative parameters of the mechanical
behavior.  For  each  set  formulation-degradation  time,  4
equivalent  samples  were  mechanically  characterized,  and
the  correspondent  average  values and standard deviations
(s

, s

or s

) were reported. In order to evaluate the

modifications on the tensile properties that were produced
by the isothermal heating, the following normalized
properties were defined: a) normalized modulus
(E*(t)=E(t)/E(0)),

normalized

ultimate

stress

(σ*(t)=σ(t)/σ(0)) and, c) normalized deformation at break
(ε*(t)=ε(t)/ε(0)). In last relationships, the degradation time
is indicated in parentheses. The use of such normalized
variables allowed separating the plasticizing effect of the
ESO from its stabilizing effect.

Regarding the mechanical behavior of non-

postprocessed  samples, Table 1 shows that such behavior
is practically independent on the formulation composition,
which  is  a  consequence  of  the  low  degradation  level  that
samples  underwent  during  the  processing  and  the  low
plasticizing effect of the ESO.

The changes in the modulus values produced by the

degradation process are evidenced in Figure 1. In such
Figure, a catastrophic increment on the modulus can be
observed for the sample stabilized only with ZnSt

, which

also  showed  the  highest  diminution  on  the  elongation  at
break (57% after 3 h of heating). However, its σ value was
maintained  almost  constant  during  the  degradation
process. Such behavior can be explained as a consequence
of  the  domain  of  branching  reactions  over  the  chain
scission ones, during the whole degradation process.

Analyzing the effect of the CaSt

presence on the E

values, it can be affirmed, for the formulations without
ESO, that its presence is beneficial for the thermal

stability of formulations without costabilizer. However, in
the  formulations  stabilized  with  D-sorbitol,  its  presence
did  not  produce  important  changes  on  the  respective
stabilities  and,  in  the  formulations  containing  triphenyl
phosphite,  its  presence  produced  a  slightly  detrimental
effect.  Otherwise,  in  formulations  containing  ESO,  the
presence  of  such  stearate  produced  a  slight  improvement
in  the  thermal  stability  of  formulations  containing
triphenyl  phosphite,  showing  a  negligible  effect  on  the
other  formulations.  On  the  other  hand,  Figure  1  shows  a
beneficial  effect  on  the  thermal  stability  caused  by  the
addition of ESO. With respect to the costabilizing effect, it
is  clear  that,  in  general  terms,  the  addition  of  D-sorbitol
improved  the  thermal  stability  of  the  modulus.  However,
the  addition  of  the  other  costabilizer produced a negative
effect on such stability.

With regard to the ultimate properties, Table 2 shows

that  the  post-processing  treatment  does  not  produce
important  changes  on  the  σ  values;  in  fact,  it  can  be
considered  that  such  changes  fall  within  the  expectable
experimental  error  (~15%).  Nevertheless,  the  heating
produced  important  decrements  on  the  ε  values  (between
18 and 57%). The highest ε decrements were observed in
the  formulations  without  costabilizer  (especially  in  the
formulation lacking CaSt

). In contrast, the best stabilities

were obtained with formulations stabilized with D-
sorbitol. In addition, in general terms, formulation stability
is not dependent on the CaSt

/ZnSt

ratio; the only

exception is shown for the formulation stabilized only
with ZnSt

, which, as it was explained, showed the highest

degradation  level.  Finally, the information given in Table
2  clearly  indicates that, in the most unstable formulations
(without  costabilizer  or,  containing  triphenyl  phosphate
series) the ESO presence improved the thermal stability of
the  respective  formulations.  However,  such  improvement
cannot be observed in formulations containing D-sorbitol.

Figures 2-3 present the initial polyene concentrations

°) and the accumulation rates of polyenes (dP

/dt) for

both sets of formulations. It can be observed that the
presence of triphenyl phosphite improved the processing
stability for all formulations (lower values of P

°). On the

other hand, in general terms, the worst processing stability
was observed in the D-sorbitol series (especially in the
presence of ESO). Analyzing the effect of the CaSt

/ZnSt

ratio on the processing stability, it can be noticed that the
differences produced by the changes on such composition
parameter were small, falling within the expected
experimental error (~15%), preventing the adequate
establishment of any trend.

Regarding the post-processing stability, the most

unstable formulation was the one stabilized only with
ZnSt

, whose values of UV absorbance for degradation

time ≥ 0.5 h were above the validity limit of the Lambert-
Beer law and, therefore, the polyene concentrations cannot
be calculated. For this formulation, the costabilizer

Page 3 of 7

ANTEC

For Review Only

addition  notably  improved  the  post-processing  stability.
For  formulations  45/0,  the  D-sorbitol  was  a  better
costabilizer  (lower  values  of  dP

/dt) than triphenyl

phosphite. The increment of CaSt

concentration

improved the post-processing stability of formulations
without costabilizer and of those costabilized with
triphenyl phosphite.

In formulations without costabilizer and formulations

containing triphenyl phosphate, the post-processing
stability is improved by the presence of ESO. However,
for formulations costabilized with D-sorbitol, the addition
of ESO did not produce significant changes on the dP

/dt

values. Finally, the addition of CaSt

to costabilized

formulations produced a negative effect on the post-
processing stability. Otherwise, in formulations without
costabilizer,

since

the

(dP

/dt)

Formulation

with

CaSt2

(dP

/dt)

Formulation without CaSt2

ratios are larger than 1.0, only

for some n values, the global effect of the CaSt

presence

on post-processing stability cannot be adequately
established.

Conclusions

The initial mechanical behavior did not strongly

depend on the formulation composition. The addition of
triphenyl phosphite improved the processing thermal
stability of all tested formulations. Finally, the addition of

D-sorbitol showed to be beneficial for the post-processing
stability of 45/0 formulations.

References

1. R. Bacaloglu, M.H. Fisch, J. Kaufhold, H.J. Sander,

in: “Plastics Additives Handbook”. 5th ed., Zweifel H,
Ed. Hanser, Munich 2001, chap. 3.

2. R.P. Lattimer, W.J. Kroenke, J. Appl. Polym. Sci., 25,

101 (1980).

3. T. Hjertberg, E.M. Sörvik, J. Appl. Polym. Sci., 22,

2415 (1978).

4. E.D. Owen, K.J. Msayib, J. Appl. Polym. Sci. Part A

Polym. Chem.

, 27, 399 (1989).

5. H. Baltacioglu, D. Balköse, J. Appl. Polym. Sci., 74,

2488 (1999).

6. K.B. Abbas, E. Sorvik, J. Vinyl Addit. Technol., 2, 87

(1980).

7. N.L. Thomas, Plast. Rubber Compos. Process Appl.,

19, 263 (1993).

8. R.F. Grossman, J. Vinyl. Addit. Technol., 15, 25

(1993).

9. L.J. González-Ortiz, M. Arellano, C.F. Jasso, E.

Mendizábal, M.J. Sánchez-Peña, Polym. Degrad.
Stab.

, 90, 154 (2005).

Key Words: Poly (vinyl chloride), Stabilizers, Mechanical
Properties.

Table 1. Tensile properties of samples after processing.

CaSt

/ZnSt

ratio

Without costabilizer

D-Sorbitol

Triphenyl phosphite

E ± s

(MPa)

σ ± s

(MPa)

ε ± s

εε

(%)

E ± s

(MPa)

σ ± s

(MPa)

ε ± s

εε

(%)

E ± s

(MPa)

σ ± s

(MPa)

ε ± s

εε

(%)

DEHP: 45 phr. ESO: 0 phr

0.0/1.0

25±4

19±2

294± 25

40±4

21±1

243±24

35±5

21±1

247±12

0.2/0.8

32±5

20±2

244± 17

41±6

21±1

247±14

38±4

20±1

281±23

0.4/0.6

38±4

20±1

248± 14

36±2

20±1

240±10

31±2

20±2

309±37

DEHP: 45 phr. ESO: 6 phr

0.0/1.0

25±3

19±1

259±12

28±6

20±1

261±22

22±3

18±1

268±16

0.2/0.8

24±2

19±1

257±6

26±5

20±1

274±19

25±2

19±0

253±10

0.4/0.6

26±4

20±1

249±11

28±4

20±1

290±12

26±4

19±1

252±13

Table 2. Rate of change of σ* and ε*of the formulations.

Without costabilizer

D-Sorbitol

Triphenyl phosphite

CaSt

/ZnSt

ratio

dσ

∗

/dt, h

-1

dε

εε

∗

/dt, h

-1

dσ

∗

/dt, h

-1

dε

εε

∗

/dt, h

-1

dσ

∗

/dt, h

-1

dε

εε

∗

/dt, h

-1

DEHP 45/ ESO 0

0.0/1.0

0.01

-0.19

0.01

-0.08

-0.02

-0.13

0.2/0.8

0.02

-0.09

0.03

-0.06

0.03

-0.10

0.4/0.6

-0.01

-0.10

0.01

-0.08

-0.00

-0.11

DEHP 45/ ESO 6

0.0/1.0

-0.00

-0.06

0.00

-0.08

0.03

-0.07

0.2/0.8

-0.02

-0.06

0.00

-0.06

0.03

-0.05

0.4/0.6

-0.02

-0.06

0.01

-0.08

0.03

-0.06

Page 4 of 7

ANTEC

For Review Only

0 phr ESO

6 phr ESO

Figure 1. Evolution of the normalized Young modulus with degradation time, corresponding tor formulations prepared with
the indicated CaSt

/ZnSt

ratio and ESO content. Series: (•••)

without costabilizer,

(───)

D-sorbitol

(───) triphenyl

phosphite

0.0

0.5

1.0

1.5

2.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

t, h

0.0

0.5

1.0

1.5

2.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

t, h

0.0

1.0

2.0

3.0

4.0

5.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

t, h

0.0

0.5

1.0

1.5

2.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

t, h

0.0

0.5

1.0

1.5

2.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

t, h

0.0

0.5

1.0

1.5

2.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

t, h

0.0/1.0

0.2/0.8

0.4/0.6

Page 5 of 7

ANTEC

For Review Only

Figure 2. Initial polyene concentration (P

°) and accumulation rate of polyenes (dP

/dt) for formulations DEHP 45/ESO 0,

prepared with the indicated CaSt

/ZnSt

ratio. Series: (•••)

without costabilizer,

(───)

D-sorbitol

(───) triphenyl

phosphite

µµµµ

0.0/1.0

0.2/0.8

0.4/0.6

Page 6 of 7

ANTEC

For Review Only

Figure 3. Initial polyene concentration (P

°) and accumulation rate of polyenes (dP

/dt) for formulations DEHP 45/ESO 6,

prepared with the indicated CaSt

/ZnSt

ratio. Series: (•••)

without costabilizer,

(───)

D-sorbitol

(───) triphenyl

phosphite

µµµµ

0.0/1.0

0.0/0.1

0.2/0.8

0.4/0.6

Page 7 of 7

ANTEC