Radiation effects of ion and electron beams on poly(methylphenylsilane)

~)

Radiat. Phys. Chem. Vol.48, No. 5, pp. 539-544, 1996

Pergamon

Copyright © 1996Publishedby ElsevierScienceLtd Printed in Great Britain.All rights reserved 0969-806X/96 $15.00+ 0.00

P I h SO969-806X(96)O0078-3

RADIATION EFFECTS OF ION A N D ELECTRON BEAMS ON POLY(METHYLPHENYLSILANE) SHU SEKI, 1 HIROMI SHIBATA,2 HIROSHI BAN, 3 K E N K I C H I ISHIGURE 4 and SEIICHI TAGAWA *t JThe Institute of Scientific and Industrial Research, Osaka University 8-1 Mihogaoka, Ibaraki, Osaka, 567, Japan, 2Research Center for Nuclear Science and Technology, University of Tokyo, 2-22 Sirakata-Shirane, Tokai-mura, Ibaraki, 319-1I, Japan, SNippon Telegraph and Telephone Corporation (NTT), 3-1 Morinosato, Wakamiya, Atsugi, Kanagawa, 243-01, Japan and 4Department of Quantum Engineering and System Science, Faculty of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113, Japan Abstract--Radiationeffects of ion beams on poly(methylphenylsilane),PMPS are described in the present paper. PMPS solid films irradiated by high energy H +, He +, N + ion beams and electron beams show changes of solubility with a large LET effects. Ion (2 MeV) and electron (20 and 30 keV) beams induce mainly crosslinking of PMPS, while it was reported that UV light and y-rays caused predominantly main chain scission on PMPS. The G-values of crosslinking increase with the values of LET of incident beams. Copyright © 1996 Published by Elsevier Science Ltd

INTRODUCTION Polysilane derivatives have attracted great interest as a new category of polymer materials (Miller and Michl, 1989). We have already reported on the transient species and the electronic properties of polysilanes, employing electron beam pulse radiolysis spectroscopy (Ban et aL, 1987, 1988). Silicon containing polymers and their reactions induced by electron beam irradiation play a significant role in two layer resist processes which is one of the most important future technologies of electron beam microlithography (Miller and Michl, 1989; Bowden, 1984). In addition, polysilanes may play an important role in ion beam lithography, a technology which is receiving renewed interest as a candidate for the manufacture of semiconductors in the future (Taylor et al., 1981). It was reported by Trefonas et aL (1983) that films of high molecular weight poly(n-hexylmethysilane) showed a UV absorption spectral shift and molecular weight reduction upon exposure to UV light (the wavelength was 313 nm) (Trefonas et al., 1983). Zeigler et al. (1985), Hofer et al. (1984) and Miller and Michl (1989) reported that photovolatilization was caused by excimer laser irradiation (248-306 nm) for alkyl substituted polysilanes. Based on their results, polysilanes have been investigated as potential positive photoresist materials because of these results of UV photolysis. In addition to the UV studies, the radiation effects of y-rays on a variety of tTo whom all correspondence should be addressed.

polysilanes, and the G-values for main chain scission and crosslinking (number of chain scissions or crosslinks/100 eV of absorbed dose) were determined as well. According to the data, main chain scission is predominant and G ( s ) / G ( x ) ratios are more than l0 in each sample (Miller, 1990). Further, several groups have investigated the potential of polysilane derivatives for electron-beam lithography. Miller et al. (1988) reported that poly (di-n-pentylsilane) could be imaged by electronbeam, and the sensitivity of the material was very high (the fluence of 10-20 pC/cm-') as a positive resist material(Miller et al., 1988). Taylor et al. (1988) studied different materials in similar experiments on the electron-beam imaging characteristics of three polysilane copolymers. These materials showed low sensitivity as positive resists (the fluence of 50 pC/cm 2 or more) and low contrast for the electron-beam, leading the authors to speculate that the high vaccum conditions made the efficiency of crosslinking increase, and as a result, the sensitivity and contrast got worse. To date, we are unaware of LET (Linear Energy Transfer eV/A: energy deposition of a incident particle per unit length) effect on polysilane derivatives as ion beam resist materials, though ion beams have been applied to the modification of polysilanes (Venkatesan et al., 1983). In this work, radiation sensitivity of polysilanes for several kinds of ion and electron beams, especially G-values of crosslinking, have been studied from the view point of ion and electron beam lithography and future modification.

539

Shu Seki et

540

PMPS sample was measured by gel permeation chromatography (GPC) with tetrahydrofuran (THF) as eluent. The PMPS used in this experiment has bimodal molecular weight distribution, and the high molecular weight peak was cut off by filtration. The sample had a molecular weight of 1.1 x 104 determined by polystyrene calibration standards. The PMPS samples were dissolved in xylene and

EXPERIMENT

Poly(methylphenylsilane); PMPS was prepared by the reaction of methylphenyldichlorosilane with sodium in refluxing toluene. The reaction was carried out under an atmosphere of predried argon. The chlorosilane was purchased from Shinetsu Chemical Inc., and distilled prior to use. The molecular weight of

1.0

al.

(a) --o- 2OkoVC

= I / , 7 ~ y~i / / .4" ? /" / / ./ /

- . - ~o,ovo 0.8 --

-~

/"

/"IZ

--m-- 2 M e V H + 2MeVN +

o z

O /~/,/

A / /

•/ ~ o 77"

0.6 --

/.

"z

-~

/

0.4

~/2"

d

I -

/

/

I

/ o.2-

I

/

, I

/ 0 0.1

,*

I

° o~ !

#i

i

10

I00

Absorbed dose (MGy) 1.0

(b) ----O-- 20 keV e--O-- 30 keV e-C~-

0.8 --

7

2 MeV H+

--I-- 2 M e V H + 2 MeV N+

/ /'it

: ~Z

0.6 --

//A

ill

/

/

Z.

o 0.01

I

/

= :

=L=/ 1

I

i

ii

•

i ,I

i I

•

!

/

/

"••

2xN / 'it

/ /

/

/

A?

0.4--

'll

~ " /

/

/

=:°

lllll ~ • i i

~ . ,

Ell

i '

I 100

./.

io/ l

t/ •

" 104

Fluence (l~C/cm 2)

Fig. 1. (a) The relation between normalized thickness (gel fraction) and fluence. (b) The relation between normalized thickness and absorbed dose.

Radiation effects on PMPS

541

Table 1. Variation of gelatlon fluences and doses. Absorbed doses are calculated by the procedure described in the text Radiations LET(eV/A) Gelation fluences (/tC/cm2) Gelation doses (MGy) G(x) 2 MeV N ÷ 160 0.089 0.88 0.21 2 MeV He + 24 2.8 3.3 0.17 2 MeV H ÷ 2.5 4.5 x 10~ 6.8 0.12 30 keV e0.21 6.8 x 102 8.5 0.082 20 keV e0.16 1.0 × 103 9.4 0.078 Co6° 0.02-0.03 0.014 y-rays Data from Ref. 3, molecular weight is 738,200.

spin-coated o n Si wafers. The thickness o f the films was 0.5 mm. These films were irradiated by 20 a n d 30 keV electron beams, 2 M e V H +, H e ÷ a n d N ÷ ion b e a m s f r o m a V a n de G r a a f f accelerator (Kouchi et al., 1989) in a v a c u u m c h a m b e r ( < 10 -6 T o r t ) at r o o m temperature. I o n b e a m irradiation was done at the Research C e n t e r for Nuclear Science a n d Technology, University o f Tokyo. Electron b e a m i r r a d i a t i o n was carried o u t at 20 a n d 30 keV by the LSI l a b o r a t o r y , N T T corporation. After irradiation, the solubility o f irradiated P M P S films was evaluated in iso-amylalcohol, a developer used to o b t a i n positive p a t t e r n s in the U V photolysis o f polysilanes, a n d the irradiated P M P S was confirmed n o t to dissolve in the solvent. All samples were developed in xylene for 2 m i n a n d rinsed in methyl alcohol. The irradiated p a r t o f film, where gel was generated, was insoluble in xylene. T h e thicknesses o f the remaining films after d e v e l o p m e n t were measured. The n o r m a l ized thickness was defined as the ratio o f the thickness after i r r a d i a t i o n to t h a t before. T h e loss o f kinetic energy o f ions in traversing the P M P S thin films was estimated from the s t o p p i n g powers o f each c o n s t i t u e n t element o f P M P S o n the

basis o f Bragg's additivity rules (Northcliffe a n d Schilling, 1970). A l t h o u g h the stopping powers o f each element depend o n the energy of incident ions, P M P S films were so thin t h a t the energy losses were very low c o m p a r e d with initial energy o f the ions, a n d could be neglected except for N ÷ ions. In the case o f N ÷ ion beam, the energy deposition in P M P S film a m o u n t s to 30% o f initial energy. T h r o u g h this calculation process, the a b s o r b e d doses o f each sample were obtained. T h e effects o f backscattering were estimated only for electron beams, because they were negligible for ion beams. I R spectra were m e a s u r e d for samples irradiated by ion b e a m s with F T - I R spectrometer. Investigation of molecular weight change with irradiation was carried out by G P C for 2 M e V He + irradiated samples. RESULTS AND DISCUSSION Figure 1(a) shows sensitivity curves o f P M P S films for i r r a d i a t i o n o f 20 a n d 30 keV electron beams, 2 M e V H ÷, He + a n d N ÷ ion beams. Gel was generated for any kinds o f beams, even in the case of electron beams, a n d the gel fraction corresponds to

0.25

0.2

0.15

............................. 2 M e V N + ......... + ~Mc.Y....I-I.~..........................................

................................

...................................i.................................

i

i

i

r~

O2MeVH +

0.1 _g ~-

o 0.01

20 k e V e"

30 k e Y e"

0.05

0

1

[

Co 60 T-rays

'i

[

I 1

100

LET (eV//~) Fig. 2. The relation between LET of radiations and crosslinking G-values.

542

Shu Seki et al.

Si--Si Si--C Phenyl Ph--Si Me--Si

\

(5) C--H bonds (in methyl) / C--H bonds (in phenyl) /

.=

(2)

(3)~

Si[ C

Siloxane structure

(1)

[ 4000

I 3500

I 3000

~

Si--H

L I 2500 2000 Wave number (cm-1)

I 1500

I 1000

450

Fig. 3. Changes in IR spectra as increasing dose of He + ion beam:(a) before irradiation, (b) after irradiation (572 MGy). the normalized thickness. Figure l(b) shows the sensitivity curves which were calibrated by the absorbed dose. The gelation fluences and doses are summarized in Table 1. Even in the values of gelation dose, the gelation dose becomes small with increasing values of LET, as shown in Fig. 2. According to the statistical theory of crosslinking and scission of polymers induced by radiation, the behaviour of gelation is expressed by the following equation (Charlesby-Pinner relationship) (Charlesby, 1954; Charlesby and Pinner, 1959),

where G(x) is the G-value of crosslinking and G(s) is the G-value of main chain scission. With these equations, crosslinking G-values are calculated and summarized in Table 1 together with gelation fluences and doses. It is clear that the efficiency ofcrosslinking becomes larger with the increase of the LET of incident beams, and it is so-called LET effects. The following measurements can give some more information about the reactions caused by these irradiation. I R spectra o f P M P S

s + s ~/2 = Po/qo + m/qo(Mn)oD,

s=l--g, where P0 is the probability of scission, q0 the probability of crosslinking, s the sol fraction, g the gel fraction, m the molecular weight of a unit monomer, (M,)0 the number average molecular weight before irradiation, and D is absorbed dose. And then, each G-value is related to the values of p0 and q0 as follows, G(x) = 4.8 x 103 × qo G(s) = 9.6 x 103 × Po

IR-spectra of PMPS film were changed by the irradition of 2 MeV He + ion beam as shown in Fig. 3. Peak assignment which is shown in the figure is based on several previous studies (Venkatesan et al., 1983; Ban and Sukegawa, 1987; Zhang and West, 1984a, 1984b; Wesson and Williams, 1980; Holler, 1982; Schilling et al., 1990). Sharp peaks become broader and larger with the irradiation. Broadening of peaks (4) in Fig. 3 is maybe due to oxidation, and it is likely to produce siloxane structure. Peaks (5) become broader, too, and it shows the increase of S i C bonds. According to the data of SiC IR-spectra, solid SiC has a strong and

Radiation effects on PMPS

543

/ / - - "~\ \ / --~t /

/ /_ 10

100

tn)

// 1000

\\ 104

105

106

Molecular weight (Mw) Fig. 4. Changes of molecular weight induced by He ÷ ion beam irradiation: (a) initial distribution, (b) 1 #Clcm2, (c) 3 #C/cm 2. broad absorption feature at 800 cm-~. Comparing peaks (2) and (3) before irradiation with those after, the height of the peaks after irradiation becomes considerably low. This is ascribed to the loss of methyl and phenyl substituents from polymer backbone, and similarly the dissociation can be observed in peaks (1) which correspond to C - H strechings of methyl and phenyl substituents. Another slightly increasing change is observed around 2200 c m - 1. It may be due to increase of Si-H bonds as a result of main chain scission and hydrogen terminated chain ends.

Molecular weight changes around gelation dose Film samples for molecular weight measurement were irradiated at the fluence of 1 and 3 pC/cm 2 by He ÷ ion beam, and the data shown in Fig. 4 with initial distribution. The shoulder in the high molecular weight region (105-106) rises with the irradiation, which is proof of the crosslinking described before. However, one peak appears in the low molecular weight side. It is also due to the dissociation of substituents from polymer backbone shown in IR-spectra. The results of these measurements indicate considerable changes in molecular structure of solid state PMPS, especially breakdown of Si-Si bond structure in the polymer backbone. In spite of this fact, the macroscopic behavior of solid PMPS is a typical behavior of a crosslinking polymer. It is considered that the increasing Si-C bonds grow into silicon carbide network, therefore large molecules may be produced. On the other hand, the Si-C bonds cause rearrangement of polymer backbone, and as a result, the succession of s-conjugation is broken

down. It is supported by the UV absorption band decrease with the ion beam irradiation (unpublished results).

CONCLUSION

Ion (2 MeV) and electron (20 and 30 keV) beams induce mainly crosslinking in PMPS even though UV-light and v-rays cause predominantly main chain scission. The G-values of crosslinking become larger as the values of LET of radiation increase. These phenomena are due to so-called LET effects. That is, the polysilane resist behave as a negative resist for high LET radiation, and also behave as a positive resist for low LET radiation. Further, in the case of low energy electron beam (20-30 keV), which is one of the middle LET range beams, the predominant reaction can be changed by polymer molecular structure. Therefore, it is quite important to elucidate the difference in the electron beam induced reactions of polysilanes. Ion beam irradiation induces molecular structure changes of PMPS, such as dissociation of substituents, or generation of siloxane structure and S i C bonds. Especially generation of S i C bonds is important because increasing Si-C bonds maybe grow into silicon carbide network, and they also mean the rearrangement of polymer backbone. It is a very interesting characteristic of PMPS that the main chain scission and rearrangement are simultaneously induced by the ion beam irradiation.

Acknowledgements--We thank Dr S. Hayashida (NTT corporation) and Dr Y. Yoshida (Osaka University) for their experimental supports.

544

Shu Seki et al. REFERENCES

Ban H., Sukegawa K. and Tagawa S. (1987). Macromolecules 20, 1775. Ban H. and Sukegawa K. (1987). J. Appl. Polym. Sci. 33, 2787. Ban H., Sukegawa K. and Tagawa S. (1988). Macromolecules 21, 45. Bowden M. J. (1984) ACS Symposium Series 266. Am. Chem. Soc., Washington DC, p. 39. Charlesby A. (1954) Proc. R. Soc. Lond. Ser. A 222, 60; 222, 542; 224, 120. Charlesby A. and Pinner S. H. (1959). Proc. R. Soc. Lond. Ser. A469, 367. Hofer D. C., Miller R. D. and Willson C. G. (1984). Proc. SPIE 249, 16. Holler I. (1982). J. Electrochem. Soc. 129, 180. Kouchi N., Tagawa S., Kobayashi H. and Tabata Y. (1989). Radiat. Phys. Chem. 34, 453. Miller R. D., Rabolt J. F., Sooriyakumaren R., Fickes G. N., Fleming W., Farmer B. L. and Kuzmany H, (1988) ACS Symposium Series 360, Am. Chem. Soc., Washington DC, p. 43.

Miller R. D. and Michl J. (1989). Chem. Rev. 89, 1359. Miller R. D. (1990) Advanced in Chemistry Series 224. Am. Chem. Soc., Washington DC, p. 413. Northcliffe L. C. and Schilling R. F. (1970). Nucl. Data. Tables AT, 233. Schilling F. C., Bavoy F. A., Lovinger A. J. and Zeigler J. M. (1990) Advanced in Chemistry Series 224. Am. Chem. Sot., Washington, DC, p. 341. Taylor G. N., Wolf W. W. and Moran J. M. (1981). J. Vac. Sci. Technol. 19, 872. Taylor G. N., Hellman M. Y., Wolf T. and Zeigler J. M. (1988). Proc. SPIE 920, 274. Trefonas P., West R., Miller R. D. and Hofer D. (1983). J. Polym. Lett. Ed. 21, 823. Venkatesan T., Wolf T., Allara D., Wilkens B. J. and Taylor G. N. (1983). Appl. Phys. Lett. 43, 934. Wesson J. P. and Williams T. C. (1980). J. Polym. Sci. 18, 959. Zeigler J. M., Harrah L. A. and Johnson A. W. (1985). Proc. SPIE 539, 166. Zhang X. H. and West R. (1984). J. Polym. Sci. 22, 159. Zhang X. H. and West R. (1984). J. Polym. Sci. 22, 225.