Polymer Journal, Vol. 28, No. 10, pp 901-910 (1996) Spectral Analysis of Polystyrene, Polypropylene, and Poly(methyl methacrylate) Polymers in TOF SIMS and XPS by MO Calculations Using the Model Oligomers Kazunaka ENDo,t Naoya KOBAYASHI, Masayuki AIDA, and Takahiro Hosm * Tsukuha Research Laboratory, Mitsubishi Paper Mills, Ltd., 46 Wadai Tsukuba-shi, lbaraki 300-42, Japan * Analytical Laboratory, Ulvac-phi Inc., 2500 Hagisono, Chigasaki 253, Japan (Received April 5, 1996) ABSTRACT: Spectra of the polystyrene (PS), polypropylene (PP), and poly(methyl methacrylate) (PMMA) polymers in time-of-flight secondary ion mass spectrometry {TOF SIMS) and valence X-ray photoelectron spectroscopy were analyzed by the MO calculations using the model oligomers. For TOF SIMS, we tried to predict where the scission of polymers can occur on sputtering, due to bond-orders of the model 5- or 3-mers by a semiempirical MO calculations. We also determined the probable structural formulas of the secondary positive-ion fragments in the range of 0--100 amu by ab initio MO calculations using HONDO7 program. The valence X-ray photoelectron spectra (XPS) of the polymers were simulated by a semiempirical HAM/3 MO method using the trimer model molecules. The theoretical spectra showed good agreement with the observed spectra of polymers between 0-40 eV. KEY WORDS Time-of-Flight Secondary Ion Mass Spectroscopy {TOF SIMS) / Valence X-Ray Photoelectron Spectra (XPS)/ Ah Initio and Semiempirical MOs / Positive Secondary Ions/ Bond-Order /Polymer/ Polystyrene/ Polypropylene/ Poly(methyl methacrylate) / Time-of-flight secondary ion mass spectrometry (TOF of polymers involving carbon, nitrogen, oxygen, and SIMS) and X-ray photoelectron spectroscopy have fluorine, we tested the performance of semiempirical become powerful tools for studying polymer surface. hydrogenic atoms in molecule, version 3 (HAM/3) MO Informations 1- 5 obtained from TOF SIMS included method21 - 23 in that the results can be directly compared monomer molecular weight, fragmentation pathways, with experiment, because it uses the idea of transition molecular weight distribution of oligomers and spectra states24 rather than Koopmans' theorem to predict characteristic of a specific polymer family. On the other vertical ionization potentials (VIPs). hand, there has been no study about spectral analysis of The present paper also offers observed and simulated polymers in SIMS by MO calculations, since the emis­ spectra in X-ray photoelectron spectroscopy for atactic sion process of secondary ions, atoms or molecules on and isotactic PP, PS, and PMMA polymers. We found sputtering will be very complicated. In the present study, a little difference of atactic and isotactic PMMA between our aim is to predict where the scission of (polypropylene observed spectra at around 14 eV. The difference of the (PP), polystyrene (PS), and poly(methyl methacrylate) tacticity in valence XPS was unable to be explained by (PMMA)) polymers can occur on sputtering, due to the theoretical result by HAM/3 MO calculations using bond-orders of the model 5- or 3-mers from semiempir­ the model molecules. The simulation of valence XPS was ical MO calculations using AMl program.6 We also aim performed for trimer models using the standard con­ to determine the probable structural formulas of the volution techniques by a Gaussian lineshape and using secondary positive-ions by ab initio MO calculations the Gelius model 25 for molecular photoionization cross­ using HONDO7 program. 7 Experimentally, we obtained section. the different spectra of the stereoisomer (atactic and isotactic) PMMA in the range of 800-1200 amu in TOF MO CALCULATIONS SIMS, as obtained by Zimmerman and Hercules. 8 Some X-ray photoelectron spectra (XPS) studies9 - 13 Simulation of Valence XPS of model oligomers and saturated hydrocarbons dem­ The electronic structure of model trimers [H-(CH2- onstrated that information on the conformation and C(CH3)COOCH3h-H, H-(CH2-CH(CH3)h-H, and tacticity dependence can be obtained through spectral H-(CH2-CH(C6 H 5)h-HJ for isotactic, syndiotactic, simulation by MO calculations. Delhalle et al. 13 found and heterotactic types were calculated using a new ver­ evidence of folded structure at the surface of polyethyl­ sion of HAM/3 program extended by Chong. 26 For the ene lamellae in the XPS valence band. In our previ­ geometry of the molecules, we used the optimized car­ ous papers, 14 - 20 we used syndiotactic model molecules tesian coordinates from the semi-empirical AM 1 (ver­ for analysis of XPS of polymers, because we found that sion 6.0) method. the tacticity had little effect on the calculated ener­ In the HAM/3 program, we can obtain the three sets gy structures, in contradiction to the results of other of relative atomic photoionization cross-section as workers. 12 ·13 For better assignment18 - 20 of valence XPS permanent data: (a) Gelius empirical parameters for Mg-Ka (1253.6eV) radiation, (b) theoretical values from t To whom correspondence should be addressed. Nefedov et al. 27 for Mg-Ka radiation, and (c) theoretical 901 K. ENDO et al. values from Nefedov et al. for Al-Ka (1486.6 eV) radi­ 31 G basis set for H, C, and O atoms without electron­ ation. In this paper, we report the results from set (c) correlation. only, because of Al-Ka radiation was used in the pres­ ent experiment. In order to simulate the valence XPS EXPERIMENTAL of polymers theoretically, we constructed from a super­ position of peaks centered on the VIPs, Ik. As was done Materials in previous works, 18 - 20 each peak was represented by We used commercially-available polypropylene (PP) a Gaussian lineshape function. The intensity is estimated (Scientific Polymer Products, Inc.; atactic and isotactic from the relative photoionization cross-section for Al-Ka types), poly(methyl methacrylate) (PMMA) (Aldrich radiation using the Gelius intensity model. In the case Chemical Co., Inc.; atactic type Mw 93300 and isotactic of linewidth ( WH(k)), we used WH(k) = 0.10 Ik for the type M w 300000), and polystyrene (PS) (Scientific Polymer models, as adopted in previous works. 18 - 20 Products, Inc.; atactic type M w 280000 and isotactic type MW 400000). Bond-orders of Model Oligomers and Fragment Jons for Samples for XPS measurements were prepared by Mass Spectral Analysis in SIMS cast-coating the polymer solution on an aluminium plate, The electronic structure of model oligomers [H-(CH2- while toluene and chloroform were used for (PP, isotactic C(CH3)COOCH3h-H, H-(CH2-CH(CH3)) 5-H, and PMMA and isotactic PS) and (atactic PMMA and atactic H-(CH2-CH(C6 H 5)h-HJ were calculated using a PS) polymers, respectively. The film was estimated to be semiempirical AM 1 program (version 6.0). Figure 1 a few tens of micrometers thick. shows the bond-orders of the model oligomers of PP, In the TOF SIMS measurement to obtain the differ­ PMMA, and PS polymers for the optimized results, as ent spectra of the stereoisomers (isotactic and atactic obtained by the energy gradient method. polymers), solutions of 1 g L - l of the polymers were The electronic structure of the fragment positive-ions produced, from which 1 to 5 µL of the solutions was for three polymers were obtained by ab initio calculations deposited on the silver foils. The procedure resulted in using a HOND07 program. For the geometry of the the deposition of a mono layer or less on the silver surface. fragment positive ions, we used optimized cartesian coordinates from the AM I program. In the ab initio XPS Measurements calculations, we used a restricted Hartree-Fock (RHF)/4- The experimental photoelectron spectra of the poly­ mers were obtained on a PHI 5400 MC ESCA spec­ trometer, using monochromatized Al-Ka radiation. The PP(5-mer) spectrometer was operated at 600W, 15kV, and 40mA. The photon energy was 1486.6eV. A pass energy of 35.75 eV was employed for high-resolution scans in a Yo989 Yo989 valence-band analysis (50 eV of range). The angle be­ 0 0989t~ o.976 t):::cQ977y,_0976 c=~!::c~y tween the X-ray source and the analyzer was fixed at c.986 45°. The spot size in the measurement was 3 x 1 mm. c0.989 c.989 The use of dispersion compensation yielded an in­ strumental resolution of 0.5 eV with the full width at half-maximum on the Ag3d line of silver. Multiple-scan PMMA (3 -mer) averaging on a multi-channel analyzer was used for the valence-band region, although a very low photoelectron emission cross section was observed in this range. ra.977 10.980 1Q986 Gold of 20 A thick was deposited on the films of the C 0.973 { o.959 Co.959 o.954 C 0.911 <;: polymer samples using an ion sputter unit (Hitachi E u901 o.904 Iu912 1030) for scanning microscope. A low-energy electron 0.94-0.96 1 037 1.049 uQ 1029 ~~&2 flood gun was used in order to avoid any charging effect C-H on the surface of the sample. We used the Au4f core level of the gold decoration films as a calibration i0.947 0.943 ~-947 reference. The Cls line positions of CH2 groups on the polymer films could be fixed at 285.0 and 285.2 for (PP, PS) and PMMA, respectively. PS (3-mer) TOF SIMS Measurements c 09s1 091oc 0910 097oc7:09a1 We operated a PHI TOF SIMS TFS-2000 spectrometer 0.974 0.974 0.986 with a primary Ga+ ion beam (12 ke V Ga+, pulse with 7 7 13 ns, repetition rate 10 kHz, a primary ion current of 0.95-0.98 H5 H5 H5 C-H 700-800 pA measured as a continuous beam).