Some surface aspects of the photo-oxidation of polystyrene films as revealed by ESCA: Part I—Irradiations in oxygen

Polymer Degradation and Stability 8 (1984) 213 227

Some Surface Aspects of the Photo-Oxidation of Polystyrene Films as Revealed by ESCA: Part I - Irradiations in Oxygen D. T. Clark* & H. S. M u n r o Department of Chemistry, University of Durham, South Road, Durham DH1 3LE, Great Britain (Received: 22 July, 1983)

ABSTRACT ESCA has been used to follow the changes in the surface chemistry of polystyrene films irradiated (2 > 290nm) in an oxygen atmosphere. Increases in C--O, C ~ O and O - - C = O functionalities are observed arising from extensive oxygen uptake in the surface regions. A decrease in the n ~ n* shake-up satellite intensity indicative of loss of aromaticity is also noted. Photo-oxidation has been studied as a function of light intensity, irradiation time and temperature, along with the possible rdle of hydroperoxide formation and singlet oxygen in these reactions.

INTRODUCTION Polystyrene is one of the most extensively investigated polymers with respect to photo-degradation and photo-oxidation. A considerable body of literature I - 13 has been published over the years but a consistent theory has failed to emerge and, in some cases, contradictory hypotheses have been presented. This is evidence of the complexity of the mechanisms involved. Previous studies of the photo-oxidation of polystyrene films have primarily involved the monitoring of changes in bulk chemistry. 2- 13 In order to obtain information pertaining to the oxidative processes * Present address: New Science Group, ICI plc, The Heath, Runcorn, Cheshire, Great Britain. 213

Polymer Degradation and Stability 0141-3910/84/$03.00 © Elsevier Applied Science Publishers Ltd, England, 1984. Printed in Great Britain

214

D. T. Clark, H. S. Munro

involved in natural weathering, model studies should employ uv radiation of wave-lengths > 290 nm. The investigations reported, utilising wavelengths < 290 nm, are of limited value as different mechanisms may well be involved. For example, 254 nm radiation is able to excite directly the pendant phenyl groups of polystyrene,l,9 - ~3 a situation which is unlikely to occur with sunlight at the earth's surface. It is currently believed that the initiation of photo-oxidation for ;t > 290 nm is due to the absorption of light by chromophoric impurities, e.g. hydroperoxides ~4'15 and carbonyl groups. 14-16 The possibility of oxygen-polymer chargetransfer complexes, extending the uv absorbance tail of polystyrene above 300nm, being involved in the initiation stage has also been considered. 9,17 As all solids interact with their environment via the surface, a knowledge of the changes in chemistry at the gas/solid interface is of prime importance. We have previously shown that ESCA may be successfully employed in the study of Bisphenol A polycarbonate photo_oxidation~8 - 21 and now extend these investigations to the surface photo-oxidation of polystyrene films. EXPERIMENTAL Orientated polystyrene films (Dow Chemical Co.) (25 #m) were exposed to uv radiation (2 > 290nm) in an oxygen atmosphere, as previously described. 19'2° Surface hydroperoxide formation was followed by the SO 2 labelling technique. 18'19 Singlet oxygen was produced by a microwave discharge in oxygen as previously described. 23 ESCA spectra for each sample were obtained on an AEI ES200B spectrometer employing Mgk~12radiation. The F W H M for the Au4fT:2 level at 84.0eV, used for calibi:ation purposes, was I-2eV. Line shape analyses were carried out with the aid of a DuPont 310 curve resolver. Binding energies were referenced to the C - - H component at 285.0 eV. RESULTS A N D DISCUSSION Studies as a function of light intensity and irradiation time

As a starting point for the investigation of the changes in surface chemistry during photo-oxidation (2 > 290nm) it is convenient to

Surface photo-oxidation of polystyrene." Part 1

215

300 1s

!;

,20

x10/~ ~ 536

53~,

532

-£

60 ~ ,.,.~....'...~',:','-/,.. 293

291

Z89

Z87

Z85

283

293

Z91

Za9

Za7

__

zs~, zs3

BINOING ENERGYt,V~

Fig. 1.

C ls and O ls core levels for polystyrene films irradiated in an oxygen atmosphere (2 > 290nm, 1o= l O W h m -2 h -1) for varying periods of time (min).

consider the irradiation of polystyrene films in a flowing oxygen atmosphere at a photon flux of 10 Wh m - 2 h - z. The relevant Cls and Ozs core levels are shown in Fig. 1. From a starting C is profile consisting of a main photo-emission peak at 285.0eV (arising from the C ~ H components in the backbone chain and pendant phenyl groups) and a n ~ n * shake-up satellite centred at ~292.3eV (diagnostic of the aromaticity in the system) additional peaks appear during irradiation. These components arise from C ~ O , C_~O and O - - C ~ O functionalities and are indicative of extensive photo-oxidation. The corresponding O~s signal of low intensity in the unexposed material indicates that a small degree of surface oxidation is present, although no corresponding Cls component could be detected. The intensity of this signal increases on exposure. Studies of the photo-oxidation of polystyrene utilising wavelengths > 300 nm have shown that an induction Period exists for oxygen uptake, s'~° It is clear from a consideration of the relative Ol~/Cls intensity ratios shown in Fig. 2 that there is no induction period for photo-oxidation (2 > 290 nm) in the surface regions. Oxygen incorporation into the surface is rapid and this is also evidenced by the contribution to the total C~s intensity made by the oxidised carbon species, also shown in Fig. 2. After 2 hours' irradiation there is a marked decrease in the rate of oxygen uptake, indicative of a steady-state situation representing a

216

D. T. Clark, H. S. Munro

.70" .60' l(~s

.50'

~

_

.

.

_

.

a

~

~

ICts J,0" .30'

o t~

.20 .I0 0

30

~)

120 3"~

IRRADIATION TIME(MINUTES)

Fig. 2.

01,ICes

intensityratios and per cent oxidisedcarbon atoms contributingto the total C, signal for the spectra in Fig. 1.

balance between further photo-oxidation and the evolution of low molecular weight species. From a knowledge of the appropriate instrumental sensitivity factors, the C: O stoichiometry after ~ 2 hours' irradiation is ~ 2: 1. The nature of the oxidative functionalisation is more apparent from an examination of the Cls component analysis in Fig. 3 for the spectra in Fig. 1. The C_~H component (carbon not directly bonded to oxygen) decreases in intensity with increasing irradiation time. There are concomitant increases in C_--O, C_~O and O ~ O functionalities and a decrease in the n--+ n* shake-up satellite. The extent of oxidative functionalisation and the loss of shake-up structure are indicative of the oxidation of the pendant aromatic ring. It has been suggested by Kova~evi6 et al.1° that oxidation of the benzene ring may occur during long wavelength irradiation (1 > 300 nm) of polystyrene films and the data discussed above provide confirmatory evidence that such a reaction is present in the surface regions at least. A feature of the bulk photo-oxidation of polystyrene at wavelengths > 290 nm, which has been reported in the literature, is the presence of an induction period. 4'1° The oxygen uptake in the surface regions, noted above, does not indicate an induction period, however. Examination of

217

Surface photo-oxidation o f polystyrene: Part I

C-H D

~H 0-~'= 0

'l*ClsI

~-~o c.:o

20+

10 .

30

Fig. 3.

.

.

.

.

.

60 IRRADIATION TIME(MINUTES)

120

300

Cls component analysis for the spectra in Fig. 1.

the formation of C_~O and O - - C = O functionalities, as shown in Fig. 3, reveals that the former appear after ,-~ 10 minutes' exposure and the latter after ,-, 15 minutes' exposure. The initial oxygen uptake may therefore reflect the formation of hydroperoxides and this is considered in greater detail in the next section. For comparison purposes, Fig. 4 shows the core level spectra for polystyrene films irradiated for varying times at a higher photon flux of 52.5 Wh m - 2 h - 1. A significantly greater extent of oxidation is apparent for a given irradiation time compared with the lower photon flux. The component analysis in Fig. 5 reveals the same trend with a much higher build up of C - - O and O - - C ~ O structural features at an earlier TABLE 1 Component Distributions and Oxygen Uptakes for Polystyrene Films Exposed to Similar Total Photon Fluxes at High and Low Lamp Intensities

60 min ( 1 0 W h m - 2 h -1) 10 min (52.5Whm-Zh -1)

Total Cls

C--H

C--O

C~O O--C~O

100

64

15

6

100

54

20

6

7t~ rt*

O ls/Cls

10

5

0'52

17

3

0.73

218

D. T. Clark, H. S. Munro

15

/" /

$36

s~4

s3~

~3

2~1

2~9

2~7 28s

2~3

BINDINGENERGYI eV) Fig. 4. C1, and O1, core levels for polystyrenefilms irradiated in 02 (2 > 290nm, I o = 52.5 Wh m-2 h- 1). stage than for the low lamp intensity. It is of interest to draw a comparison in terms of functionality distribution and oxygen uptake for comparable total fluxes received for samples irradiated at the higher and lower lamp intensities and this is shown in Table 1. It can be readily seen that the extent of oxidative functionalisation, in particular the levels of O - - C ~ O groups, and hence the intensity of the oxygen signal, are greater at the higher photon flux. These data suggest that the initiation mechanisms for photo-oxidation are very sensitive to the intensity of radiation and that the nature of the oxidised surface is not linearly dependent on the level of the incident flux. This becomes more apparent from a consideration of the data presented in Fig. 6 where the various C is components for polystyrene films irradiated for a fixed period of time (15 min), are shown as a function of the incident photon flux.

Surface photo-oxidation of polystyrene: Part I

219

90

C~H 50 %C1S 2O

c-o

O-C: 0 10 C=O l - l ~ i-I~×

25

Fig. 5.

5 10 IRRADIATION TIHE(HINUTES)

15

Cls components for the spectra in Fig. 4.

The decrease in C - - H and increase in C - - O functionalities occur steadily with increasing photon flux although they both appear to reach a constant level at ,-,39 W h m -z h - 1. The formations of the C ~ O and O - - C ~ O components are of particular interest. The induction period, as evidenced from the data in Fig. 3, is also apparent in the present case. The level of carbonyl functionalities present after 15 minutes' exposure is 90

[-H 50 %

0-[=0

CIS

/

16

/

....

eo

12

~o

Fig. 6.

2'o

~o

;o

s'o -

PHOTON FLUX (Whm-2h-1) C,, components for polystyrene films irradiated in oxygen as a function of photon flux.

220

D. T. Clark, H. S. Munro

601

30 I~

15~

10H

5H

0

170 168 166 B1NOIN[iENERGY(eV) Fig. 7. S2p core levels for photo-oxidised polystyrene films exposed t o

S O 2.

greater than that for the carboxylate for increasing fluxes up to , - , 3 3 W h m - 2 h -1. The intensity of the former then decreases with increasing photon flux whilst the latter continues to rise. The ~t ~ it* shake-up satellite, diagnostic of the aromaticity present in the system, is also seen to decrease. It is interesting to note that the centroid of this peak tends to move to lower binding energy with increasing flux. A similar situation also exists for irradiations as a function of time (i.e. at 10 Wh m - 2 h - 1 with increasing time). These observations, coupled with

Surface photo-oxidation of polystyrene." Part I

221

the fact that although the surface is extensively oxidised the intensity of this component tends to be relatively high, may be consistent with the formation of carbonate structural features. Surface hydroperoxide formation

The formation and subsequent decomposition of hydroperoxides are believed to be of primary importance in the bulk photo-oxidation of polystyrene. 14,15 As noted above, the oxygen uptake during the initial stages of irradiation was reflected by an increase in the intensity of the C - - O component in the Cxs spectra. This may be indicative of the formation of hydroperoxides in the surface. A technique for the labelling of these groups by the direct reaction with SO 2 has been described previously, enabling a semi-quantitative study of surface hydroperoxide formation to be made. 1a'19 This technique has been utilised to study the r61e of hydroperoxides in the surface photo-oxidation of polystyrene films (2 > 290nm, 1 0 W h m - 2 h - l ) . The ESCA spectra in Fig. 7 show the Szp core levels for photo-oxidised films exposed to sulphur dioxide. Although four component peaks are apparent, the spin-orbit splitting of the S2p level gives rise to a doublet due to contributions from the S2p.,2 and S2p3s2 levels in the ratio of 1:2. Consequently, only two chemical environments of sulphur are present. These correspond to sulphones (-,~ 168.0 eV) and sulphates ( ~ 169.2 eV), the latter arising from the reaction of SO 2 with hydroperoxides. From a knowledge of the instrumentally dependent sensitivity factors, the intensity of the sulphate component may be directly related to the hydroperoxide level in the surface (i.e. C - - O O H ) and the data in Fig. 8 show the formation of these groups as a function of time. A low level of functionalisation in the unexposed material is present, but on exposure it rapidly increases and reaches a maximum after ~ 30 min. The growth in the carbonyl component is also shown in Fig. 8 for comparison. It is readily apparent that the hydroperoxide level reaches its maximum whilst the C C_~O functionality has only attained ,-~20 ~ of the ultimate value. These data confirm that hydroperoxides are formed in the inital stages of surface photo-oxidation of polystyrene films. Surface chemistry as a function of temperature

The experimental conditions employed in the literature for photooxidation studies vary considerably, e.g. photon flux and temperature.

222

D. T. Clark, H. S. Munro

K

[=0

%CIS

1{

J

J

jjJ

,

15 Fig. 8.

| 60

310 IRRADIATION TIME [minutes)

C - - O O H levels in the surface of photo-oxidised polystyrene films (I o = 10Whm-2 h-l).

TABLE 2 Comparison of the C ls Components and O~,/Cls Intensity Ratios for Polystyrene Films Irradiated in 0 2 (I o = 52.5 W h m -2 h-1) as Functions of Temperature and Time T (°C)

Time (rain)

Total C1~

C--H

C--O

30

2.5 5 10 15 2.5 5 10 15 2.5 5 10 15

100 100 100 100 100 100 100 100 100 100 100 100

76 69 54 0.53 83 70 58 49 83 74 58 55

8 14 20 21 8 14 17 22 8 11 19 22

50

70

C~O

5 6 6 4 2.5 4 6 5 2 5 4 3

O--C~O

6 8 17 20 2.5 9 17 22 3 7 17 18

n--* n*

01s/C1,

5 3 3 2 4 3 2 2 4 3 2 2

0.2 0.46 0.73 0.80 0.2 0.44 0'65 0.77 0.16 0.38 0.71 0-75

Surface photo-oxidation of polystyrene: Part I

223

The former has been considered in relation to polystyrene films above and, in this section, the influence of elevated temperature on the nature of the surface chemistry during photo-oxidation is considered. The data in Table 2 reveal the distributions of the various Cls components and the relative Ols/Cls intensity ratios for polystyrene films irradiated with a photon flux of 52.5 Wh m - a h - 1 for various periods of time at 30, 50 and 70 °C. On increasing the temperature from 30 to 50 °C the main difference in the component distribution is the increased intensity of the carboxylate functionality at the higher temperature. At 70 °C this level is reduced to below that at 30 °C. These results indicate that temperature under the conditions of the experiment has only a slight effect. This may well arise as a result of the balance between rate processes leading to oxidative functionalisation and desorption of low molecular weight materials from the surface. Indeed, the data in Table 2 appear to show that the oxidative processes are enhanced on increasing the temperature to 50 °C from 30 °C whereas desorption rates are enhanced at 70°C.

The r61e of singlet oxygen (1AO2) in the surface oxidation of polystyrene films The results discussed above revealed that the irradiation of polystyrene films with wavelengths > 290 nm in a pure oxygen atmosphere leads to extensive oxygen uptake in the surface regions. Oxidation of the carbon atoms in the backbone chain cannot alone account for the degree of uptake. At a photon flux of 10 Wh m - 2 h - 1, 55 % of the C1s signal after --~5 hours' exposure arises from components with carbon atoms directly bonded to oxygen. Examination of the n ~ n * shake-up satellite, diagnostic of the aromaticity of the pendant phenyl group, reveals a decrease in intensity during photo-oxidation indicative of this moiety being oxidised. This process could account for the extent of oxygen uptake and has been postulated for bulk 1°'17 and surface 1 photooxidation studies utilising wavelengths <290nm. In these cases the reaction probably arises from the direct excitation of the benzene ring. However, Rabek and R~nby iv have suggested that singlet oxygen (IAO2) may be responsible for the ring opening reaction. It is possible that the observed ring oxidation in the surface at wavelengths > 290 nm could arise from such a reaction between the phenyl moiety and 1AO 2 and this is considered in detail below.

224

D. T. Clark, H. S. Munro

The C is and Ols core levels for an electron take-off angle of 70 o, shown in Fig. 9, reveal the changes in the surface chemistry of polystyrene films exposed to a stream of singlet oxygen from a microwave discharge for 60 and 120 min. (The conditions of the experiment employed were 60W, 2 torr, and the atomic oxygen and ozone also produced were removed by the continuous distillation of mercury through the discharge region.)2a Oxygen incorporation into the surface is evident from the increase in the complexity of the Cls signal, arising from C_~O, C----O and O----C--~O components, and the increase in the intensity of the Ols envelope. The nature of the surface is more apparent from the Cls component analysis and the relative O~s/C~s intensity ratios in Table 3. The distributions of components are in essential agreement with those reported by Dilks 22 and are not dissimilar to the results obtained for Bisphenol A polycarbonate. 23 The surface specificity of the oxidation is

1'

(

536

53'L+

532

293

291

289

2B?

285

283

BINDING ENERGY(eV)

Fig. 9.

C1, and O1, core levels for polystyrene films exposed to 'singlet oxygen'.

Surface photo-oxidation of polystyrene." Part I

225

TABLE 3 Cls Components and OljCls Intensity Ratios for the Spectra in Fig. 9 Time

Total

(min)

C~,

0 = 70 °

0 60 120

0

0 60 120

=

30 °

C--H

C--O

C--O

O--C--O

rc~n*

01~1C1,

100 I00 100

92 61 53

2 19 27

0 12 7

0 7 11

6 1 2

0'0 0"58 0"73

100 100 100

93 69 66

1 13 15

0 10 10

0 5 6

6 3 3

0.02 0'40 0"40

revealed from the consideration of the data obtained with an electron take-off angle of 30 o, also shown in Table 3. Both the degree of oxidative functionalisation and oxygen uptake are greater at 0 = 70 o. From the discussions presented previously 23 it is unlikely that the effluent reaching the polystyrene films is completely free of contaminant reactive oxygen species (i.e. atomic oxygen). Consequently, experiments were repeated over a 5-h exposure period with N O 2 also being added to the effluent to remove residual traces of oxygen atoms. No oxidation of polystyrene could be detected with ESCA. Control experiments where cispolyisoprene films were exposed downstream of polystyrene revealed •-v'~.,CH-CH2"-..P h - - ~ - - C H 2 " ~ - f ' ~ ' - - '

I

I

@ CHARI3E

@

\OH

y~

R.

I -°H l

@

II

TRANSFER COMPLEX _ " ~('\

f-o. I

C-CM2"-',.r~"~-A- -CH2"~I'÷ R.

® I

^ ....~,,.o2

OH

,.~

C -o RH

0 2 Er~ancemenf of

"CARBONATE ~

Spin-Forbidden So-,,l', trQnsition

Fig. 10.

Scheme for some of the possible reactions occurring in the surface of polystyrene films during photo-oxidation.

226

D. T. Clark, H. S. Munro

oxidation of the former and not the latter, indicative of singlet oxygen reaching the samples. 2a These results show that polystyrene does not react with singlet oxygen and are in agreement with the conclusions of MacCallum and Rankin. 24 Consequently, singlet oxygen is not responsible for the observed ring opening reaction in polystyrene films. Weir a has proposed that phenyl ring oxidation is due to attack by hydroxy radicals resulting from the decomposition of hydroperoxides: R O O H hv R O ' + ' O H From the study of surface hydroperoxide formation during the photooxidation of polystyrene films (2 > 290nm, I o= W h m - 2 h -1) it is apparent that the decrease in the rt ~ n* shake-up satellite does not begin until the rate of hydroperoxide formation is high and this may well support the hydroxy radical theory. The nature of the reaction in the surface during photo-oxidation is extremely complex and Fig. 10 shows a scheme for some of the possible processes.

CONCLUSION The range of oxidative functionalities and, in particular, the extensive oxygen uptake and loss of aromatic character, are consistent with photooxidative reactions involving both the main backbone chain and the pendant phenyl groups of polystyrene. It is not possible, from the data presented, to resolve the mechanisms which are responsible for the ring opening reaction, but the r61e of singlet oxygen in initiating this process may be discounted.

REFERENCES 1. J. Peeling and D. T. Clark, Poly. Deg. and Stab., 3, 97 (1980). 2. B. R~nby and J. F. Rabek, Photodegradation, photo-oxidation andphotostabilization of polymers. John Wiley and Sons, London (1975). 3. N. Grassie and N. A. Weir, J. Appl. Polym. Sci., 9, 963 (1965). 4. N. Grassie and N. A. Weir, J. Appl. Polym. Sci., 9, 987 (1965). 5. N. Grassie and N. A. Weir, J. Appl. Polym. Sci., 9, 999 (1965). 6. J. F. Rabek and B. Rhnby, J. Polym. Sci., Polym. Chem. Edn., 12, 273 (1974).

Surface photo-oxidation of polystyrene: Part I

227

7. G. Genskens, D. Bayens-Volant, G. Delannois, Q. Lu-Vaih, W. Piret and C. David, Eur. Polym. J., 14, 291 (1978). 8. N. A. Weir, Eur. Polym. J., 14, 9 (1978). 9. J. R. MacCallum and D. A. Ramsay, Eur. Polym. J., 13, 945 (1977). 10. V. Kova~evi~, M. Bravar and D. Hace, Bulletin de la Chimique Beograd., 14, 167 (1980). 11. L. A. Wall, M. R. Harvey and M. Tyron, J. Phys. Chem., 60, 1306 (1956). 12. L. A. Wall and M. Tyron, Nature, 178, 101 (1956). 13. L. A. Wall and D. W. Brown, J. Phys. Chem., 61, 129 (1956). 14. P.J. Burchill and G. A. George, J. Polym. Sci., Polym. Letts. Edn., 12, 497 (1974). 15. G. A. George and D. K. C. Hodgeman, J. Polym. Sei., Part C, 55, 195 (1976). 16. G. A. George and D. K. C. Hodgeman, Eur. Polym. J., 13, 63 (1977). 17. J. F. Rabek and B. Rhnby, J. Polym. Sei., Polym. Chem. Edn., 14, 1463 (1976). 18. D. T. Clark, Pure and Applied Chem., 54, 415 (1982). 19. D. T. Clark and H. S. Munro, Poly. Deg., and Stab., 4, 441 (1982). 20. D. T. Clark and H. S. Munro, Poly. Deg. and Stab., 5, 227 (1983). 21. D. T. Clark and H. S. Munro, Poly. Deg. and Stab., 8, 195 (1984). 22. A. Dilks, J. Polym. Sci., Polym. Chem. Edn., 19, 1319 (1981). 23. D. T. Clark and H. S. Munro, Polymer (1984) (in press). 24. J. R. MacCallum and C. T. Rankin, Makromol. Chem., 175, 2477 (1974).