Analysis of Polymer Films by Diffuse Reflectance FTIR Spectroscopy: Characterization of Terminal Carboxyl Functionalities

P. de DONATO,’ J. M. CASES,’ B. HUMBERT,’ P. LUTCEN,3 and C. FEYDER3

‘Laboratoire “Environnement et Minbralurgie” L.E.M.-U.A. 235 du CNRS-Equipe de Recherche et d’Application sur la Physicochimie des surfaces et des Minbraux Industriels-I.N.P.L. Nancy, France; ’Laboratoire de Spectromkrie de Vibrations, Universitb de Nancy I, BP 239, Vandoeuvre Les Nancy, France; 3Du Pont de Nernours (Luxembourg) S.A. L-2984, Luxembourg

SYNOPSIS

Fourier transform infrared (FTIR) spectroscopy in the diffuse reflectance mode was used to study polyethylene terephthalate (PET) films. The polymer film (12 pm, molecular weight MI = 18,000) was placed on a finely powdered KBr matrix, used as a reference. Infrared spectra exhibited a new band at 1684 cm-’, not usually reported in the literature. This band is assigned to the C = 0 stretching vibration of a terminal acidic function in the presence of internal hydrogen bonds. In the carbonyl region, artifacts created by specular reflection are also discussed. The assignment of the band a t 1684 cm-’ is confirmed by transmission measurements on the overtone of carbonyl group (3335 cm-’) , using polymer films with thicknesses greater than 200 pm and by comparison with polymer of different molecular weight. These acidic functions can be used to monitor the rate of polymerization. I t is therefore possible to obtain information on the polymerization rate of PET films, using diffuse reflectance and transmission analysis, directly on the solid. 0 1992 John Wiley & Sons, Inc. Keywords: infrared spectra of P E T film from diffuse reflectance poly (ethylene tere- phthalate) , polymerization rate study by diffuse reflectance FTIR of carboxyl terminal functionalities in polymerizing P E T films from FTIR study

I NTRO DUCT10 N

Diffuse reflectance combined with Fourier-trans- form infrared spectroscopy (FTIR) is proving more and more to be an excellent method for analysis of solids.’*2 Several authors already have used diffuse reflectance spectroscopy (DRIFT) to study polymer surfaces using either pelletized or powdered KBr or other reference^.^-^

Sergides et al.3 obtained polyethylene tere- phthalate (PET) spectra by placing the film on the reference mirror of a Harrick accessory; however, with this experimental set up their spectra did not exhibit an adequate signal to noise ratio.3 This work is aimed at showing the potential of diffuse-reflec-

Journal of Polymer Science: Part B: Polymer Physics, Vol. 30,1305-1310 (1992) 0 1992 John Wiley & Sons, Inc. CCC 0887-6266/92/01201305-06

tance infrared analysis for polymer film analysis, especially in the carbonyl region. For thick materials, interference fringes on the spectra obtained by transmission measurements do not allow a precise quantitative determination. The importance of sample preparation and of comparison with trans- mission spectra were also worked out. For simplicity, the spectra recorded by the Harrick accessory will be called diffuse reflectance spectra. In fact, these measured reflectance spectra are not true diffuse re- flectance spectra: they are reflection-absorption spectra3 because part of the beam is transmitted through the film without scattering.

EXPERIMENTAL

Different thicknesses of PET films (M, = 18,000) were used 12, 100, 200 ( 2 X loo), 300 ( 3 X loo),

1305

1306 DE DONATO ET AL.

350, and 500 pm (5 X 100). Polymer films were studied by transmission, microscopy, and diffuse- reflectance analysis. In all cases the spectra were recorded with a Bruker IFS 88 Fourier transform infrared (FTIR) spectrometer.

For transmission (thicknesses 100,200,300,350, and 500 pm) and diffuse reflectance (12 pm) mea- surements, a room temperature detector (DTGS) and a mercury cadmium telluride (MCT) detector, cooled at 77 K, were used respectively. For micro- FTIR measurements ( 12 pm) , the IFS 88 spectrom- eter was provided with a Bruker A-590 microscope. The microscope has its own narrow-band mercury cadmium telluride (MCT) detector of 100 pm di- ameter, cooled at 77 K. The infrared beam width was reduced to 20 pm' by diaphragms, in order to analyze different domains of the polymer film. The diffuse reflectance spectra were obtained using a Harrick attachment. The PET film was deposited on a matrix of finely powdered KBr, used as the reference. Signal optimization was realized by ad- justing the level of the sample. The sample was then placed at the center of the two ellipsoidal mirrors of a Harrick diffuse reflectance attachment (DRA) . All the spectra are given in absorbance units ( AU) . Because it was impossible to use a decomposition program without making considerable error, the in- tegrated area of a peak was not used in this study, even though this procedure usually avoids errors due to peak shift.

RESULTS AND DISCUSSION

The spectra obtained by micro-FTIR analysis, on the different domains showed no significant chem- ical differences. They can all be represented by the spectrum presented in Figure 1 (resolution 4 cm-', 200 scans). It must be pointed out that all the main bands between 1800 and 1000 cm-' are off scale and truncated. The presence of fringes enables the cal- culation of the film thickness according to the for- mula:

N 2nA

c = -

where c is the thickness in cm, N is the number of fringes between 2685 and 2187 cm-' , n is the sample refractive index, and A is the difference in wave- numbers between the first and the last fringe.

In our case, [ N = 2, n = 1 , 6 and A = 498 cm-'1 , this leads to a thickness of 12 pm which confirms the values given by Du Pont. In the future it should

L

1000 0.0

4000 3000 2000 VRVENUMBERS CM-1

Figure 1. Infrared transmission spectrum of polyeth- ylene terephthalate (PET) (Bruker IFS 88; Bruker A-590 microscope, diameter of beam 20 pm; film thickness 12 pm).

be possible to reduce the infrared beam to 10 pm and thus to compare the results with those obtained with a 20 pm diameter infrared beam. This spectrum exhibits all the absorption bands characteristic of PET whose assignments were given by Grine and Wardg and Krimm." The comparison with the spectrum obtained by the Harrick DRA, where the signal-to-noise ratio is excellent (> 500) (Fig. 2 ) , exhibits no major difference except in the region of strong absorption, and especially in the carbonyl range ( 1800-1600 cm-') . Taking a detailed look at this region (Fig. 3 ) , the transmission spectra (Tr) , show a main peak at 1727 cm-' corresponding to the carbonyl stretching vibration in an ester f u n c t i ~ n ~ " ~ and a second peak at 1684 cm-', which has never been assigned. The diffuse reflectance spectra exhibit two bands at 1743 and 1703 cm-' and a peak at 1684 cm-' . None of the bands at 1743 (too high frequency) or 1'103 cm-' (too low fre- quency) can be assigned to the C=O stretching vibration of an ester function. It is well known that, in the region of an absorption band, the refractive index of the material changes due to anomalous dis- persion."~'2 As a result, the distortion in the ab- sorption band shape can be attributed to the con- tribution of specular reflection to the reflectance spectra.13-15 Assuming that the variations of the complex refractive index of PET in the carbonyl stretching region are globally the same as those given by Porter et al.14 for poly (methyl methacrylate), it is possible to calculate the evolution of the intensity of the specular reflection (Rs) according to the fol- lowing formula (determined under normal inci- dence ) :

FTIR CHARACTERIZATION OF CARBOXYL FUNCTIONALITIES

n

w u z

cr: 2 8 m <

E l 3

m

E 3 w u z

cr: 2 8 m <

4000 3000 2000 10.. W A V E N U M B E R S LM-1

Figure 2. Infrared diffuse reflection-absorption spec- trum of PET film. (Bruker IFS 88; Harrick DRA; film thickness 12 pm) .

RS = [(n - 1 ) ' + k 2 ] / [ ( n + 1)' + k 2 ] ( 2 )

where N = n + ik, N is the complex refractive index, n is the real refractive index (real part), and k is the absorption index (imaginary part).

1800 1700 I YRVENUMBERS CM-1

0

Figure 3. Comparison between infrared diffuse reflec- tion-absorption (Re) and transmission (Tr) spectra of PET film: Spectral region 1800-1600 cm-'. (film thickness 12 pm) .

. . .r(

+ +

0.0 LW.0

1850 > 1800

1750 1700 1650

Wavenumbers (cm- I )

Figure 4. Specular reflection (Rs) and absorption index ( k ) as a function of the wavenumbers, for PET film, in the carbonyl region. (film thickness: 12 pm) .

The evolution of both the specular reflection and the absorbtion index are given in Figure 4, as func- tions of the wavenumbers. The resulting calculated distorted band shape agrees with the observed spec- tra (Fig. 3 Re). The first peak at 1743 cm-' results from the superposition of the high-frequency range of both absorbance and specular reflection. Although the Harrick DRA is adjusted according to the man- ufacturer's instructions, the specular reflection is not completely eliminated: a part of it is collected by the DRA. The intensity of the infrared beam at the MCT detector results from the combination of two opposing contributions: a carbonyl absorption com- ponent, which uses up energy and a specular com- ponent, which increases energy, respectively. The maximum frequency of this reflection, corresponding to an energy magnification at the MCT detector, is equal to 1725 cm-' (theoretical value, Fig. 4 ) . This value is in good agreement with the minimum at 1720 cm-' observed in the spectrum (Re) (Fig. 3 ) .

The second peak at 1703 cm-' corresponds to a decrease of both absorbance and specular reflection. However, comparison with the transmission spec- trum (Fig. 3 T r ) , shows that the band at 1684 cm-' in the reflection spectrum (Fig. 3 Re) is not affected by specular reflection. This is due to the fact that, below 1700 cm-' , the intensity of specular reflection becomes negligible (Fig. 4 ) . As a first hypothesis, the band at 1684 cm-', can be assigned to the car- bony1 stretching vibration of benzoic acids, char- acteristic of the terminal acid groups of the polymer (Fig. 5 ) . Indeed, in the presence of internal hydrogen bonds, the carbonyl frequency is lowered by conju-

1308 DE DONATO ET AL.

rover tone (wF-1

1727 cm-' (S)

- flvertone (vw1----q

1727 cm-' (S)

flvertone (vw1----q

Figure 5. corresponding overtones. (S, strong; w, weak; vw, very weak).

Structure of polyethylene terephthalate (PET): two carbonyl frequencies and

gation to 1690-1679 cm-' .I6-'' As a result, the spec- tra must exhibit the corresponding overtone of the C = 0 group in terminal acidic functions, at around doubled frequency (Fig. 5 ) , i.e.:

2 X vc=o (1684 cm-') = 3368 cm-'

This band remains undetectable, because its ab- sorption coefficient is very weak, as are all overtone bands. The intensity of the band must therefore be enhanced for satisfactory observation. The trans- mission spectra obtained for PET films with a greater thickness are presented in Figure 6. Knowing that the Beer Lambert law is obeyed by all bands in a transmission spectrum and that the nature of the different polymer samples is the same (in particular, the polymerization rate and the molecular weight), the observed linear increase of band intensity at 3335 cm-' with film thickness, confirms the assignment hypothesized by the authors (Fig. 7). The band at 3431 cm-' corresponds to the overtone of the C = 0 group in the ester functions. If we assume that the C = 0 vibrators are not interacting with their neighbors, by applying the classical formula for the anharmonicity vibrations of a diatomic molecule, the anharmonicity constants of both C = 0 groups can be calculated as 11 cm-' for the C = O groups in

ester functions and 16 cm-' for those in terminal carboxyl groups. It is difficult to use the intensity ratio between the bands at 3432 cm-' and 3335 cm-' for quantitative monomer unit determination, be- cause the anharmonicity coefficients (on which the relative intensity of an overtone band is dependent) of the C = 0 groups and their integrated intensities are different. Finally, comparison with another polymer ( 12 pm) of different molecular weight ( M2 = 20,700) has been made. The intensity ratio Z, = 1684 cm-' [ COOH ( M1 ) ] / I , = 1684 cm-' COOH ( M2) ] , calculated from diffuse and transmission IR measurements is equal to 1.3 and 1.8, respectively. These values are in good agreement with the mo- lecular weight ratio (ca. 1.2) of the two polymers, which can be correlated with the ratio of terminal acidic groups. The value obtained by transmission measurements ( 1.8) is overestimated because the intensity of the band at 1684 cm-' is disturbed by the saturated intensity of C = O groups in ester functions (1727 cm-I). The band at 1684 cm-' is not affected by the band at 1727 cm-' in diffuse IR spectra, because of the presence of a minimum of absorbance due to specular reflection. Thus, the in- tensity ratio obtained by diffuse reflectance ( 1.3 ) is very close to the molecular weight ratio (ca. 1.2) of the polymers (i.e., the ratio of terminal acidic groups).

FTIR CHARACTERIZATION OF CARBOXYL FUNCTIONALITIES 1309

0.00,

Diffuse reflectance analysis is a quick and non- destructive method, which allows the differentiation between the vC=o stretching vibrations in ester functions (main peak 1727 cm-') and in terminal acidic functions (secondary peak 1684 cm-' ) . Quantitative determination of the content in car- boxylic groups of different PET films can be achieved by comparing transmission with diffuse reflectance analysis. In the first case (film thickness 2 200 pm) (see Fig. 5 ) , the intensities are calculated from measurements of the overtone of the C = 0 in terminal groups. In the second case, they can be obtained from measurements of the C = 0 stretch- ing band on films whose thickness can be less than 12 pm. This is of great importance for the knowledge of the polymerization rate of PET films and can be applied for industrial control.

l.15--& 200 100,um I

CONCLUSIONS

Infrared spectra of PET films (even thicker than 10 pm) can be obtained, with an excellent signal-to-

n

x

A

Li.urity limit or tbe DTCS detector 1.8

1,; . I .

"'i 0.0

0 100 200 300 400 500 600

THICKNESS (microns)

Figure 7. Intensities of overtones of C = O groups ver- sus polymer film thickness: linear correlation: (m) 3431 cm-'; (13) 3335 cm-'.

Figure 6 . Infrared transmission spectra of PET films of different thicknesses. Overtones region (100, 200, 300, 500 pm) .

noise ratio, by diffuse reflectance measurements on a KBr matrix using infrared Fourier transform spectroscopy. All the spectra exhibit a new band at 1684 cm-' characteristic of the uC=o stretching vi- bration of a terminal acidic function. This band has never been assigned in the literature. Spectral re- gions corresponding to both overtones and funda- mentals of carbonyl groups can be used to compare and to assess polymer films with different contents in terminal carboxylic groups. New applications and developments are now open for diffuse reflectance analysis because it gives information on other pa- rameters, particularly the polymerization rate of polymers.

For better micro-FTIR spectroscopy measure- ments and in order to control the good homogeneity of the PET film, the infrared beam size has to be reduced to 10 wm.

The authors thank Du Pont de Nemours (Luxembourg) S.A. for supplying PET films.

REFERENCES AND N O T E S

1. P. R. Griffiths and M. P. Fuller, Anal. Chem. 50,1906

2. M. P. Fuller, I. M. Hamadeh, P. R. Griffiths, and ( 1978).

D. E. Lowenhaup, Fuel, 61,526 (1982).

1310 DE DONATO ET AL.

3.

4.

5.

6.

7.

8.

9.

10.

C. A. Sergides, A. R. Chughtai, and D. M. Smith, Appl. Spectrosc. 41, 157 (1987). S. R. Culler, M. T. Mc Kensie, L. J. Fina, H. Hishida, and J. L. Koenig, Appl. Spectrosc. 38, 791 ( 1984). V. T. Teng and W. H. Shumaker, Appl. Spectrosc. 40, 421 (1986). G. Xue, S. Jiang, C. Dai, W. Zhu, and M. Seng, Poly- mer Bull., 15,363 (1986). H. Ishida, Rubber Chemistry and Technology, 60,515 (1987). 0. Barres, A. Burneau, J. Dubessy, and M. Pagel, Appl. Spectrosc. 4 1, 1000 ( 1987 ) . D. Grine and M. Ward, Trans. Faraday. SOC., 54,959 (1958). S. Krimm, Forstsh. Hochpo1ymer.-Forsh., 2,51 (1960).

11. D. A. Jarvis, I. J. Hutchinson, D. I. Bower, and”1. M.

12. G. K. Ribbegard and R. N. Jones, Appl. Spectrosc. 34,

13. M. D. Porter, Anal. Chem., 60,1143A (1988). 14. M. D. Porter, D. Karweik, T. Kuwana, W. Theis,

G. B. Norris, and T. 0. Tiernan, Appl. Spectrosc. 38, 11 (1984).

15. B. Humbert, DEA, Universit6 Nancy I (1988) 53 pages.

16. M. St. C. Flett, J. Chem. Soc., Part 11, 962 (1951). 17. C. J. W. Brookes, G. Eglington, and J. F. Norman, J.

18. M. St. C. Flett, Spectrochim. Acta, 18, 1537 (1962). 19. D. Peltier, A. Pichevin, P. Dizabo, and M. L. Josien,

Ward, Polymer, 2 1 , 4 1 ( 1980).

638 ( 1980).

Chem. SOC., Part I, 106 ( 1961 ) .

Compt. Rend. Acad. Sci., 248,1148 (1959).