Mechanism of UV- and VUV-Induced Etching of Poly(methyl

Chapter 26

Mechanism of UV- and VUV-Induced Etching of Poly(methyl methacrylate) Evidence for an Energy-Dependent Reaction 1

1

1

Nobuo Ueno , Tsuneo Mitsuhata , Kazuyuki Sugita , and Kenichiro Tanaka

2

Downloaded by FUDAN UNIV on February 2, 2017 | http://pubs.acs.org Publication Date: October 31, 1989 | doi: 10.1021/bk-1989-0412.ch026

1

Department of Image Science and Technology, Faculty of Engineering, Chiba University, Yayoicho, Chiba 260, Japan Photon Factory, National Laboratory for High-Energy Physics, Tsukuba, Ibaraki 305, Japan

2

We performed a mass spectroscopic study of the vaporized species generated during ultraviolet (UV) and vacuum ultraviolet (VUV) irradiation of poly(methyl methacrylate) (PMMA), and found a remarkable difference between the spectra measured during UV and VUV irradiation. In the case of VUV irradiation, mass peaks were observed, which can be ascribed to products of the direct main-chain scission of PMMA. The results are explained by an energy-dependent and site-selective photochemical reactions in PMMA. The present finding indicates that the use of energy dependent photochemical reactions is potentially useful in the photoetching of polymer resists. The effects of temperature and oxygen gas on the photoetching of PMMA are also described in this article. Among a variety of etching techniques i n microlithography, photoinduced etching i s becoming important as a non-destructive method. High energy photons, i . e . synchrotron radiation including soft X-ray, were used for the etching of various materials, and found to be useful i n microlithography (1-3). In general, the absorption c o e f f i c i e n t s of solids consisting of l i g h t elements tend to increase with photon energy and reach a maximum i n the photon energy range of about 10 - 30 eV [vacuum u l t r a v i o l e t (VUV) region]. For photo-induced etching, large penetration depth of the incident l i g h t into materials i s not required. Hence, we can r e a l i z e more rapid etching by an e f f e c t i v e use of VUV l i g h t which i s absorbed at the surface region. Further, we expect a d i f f e r e n t photochemical reaction by VUV i r r a d i a t i o n , since VUV can excite solids to higher energy states. In fact, Ueno et a l . (4) showed the efficacy of VUV i n photoetching of S i (100) and PMMA [poly(methyl methacrylate)] using a conventional l i g h t source. In the case of PMMA, for example, VUV excites electrons l o c a l i z e d along the polymer backbone as well as those i n the side chain, while UV excites only the π electrons i n the side chain. The absorption c o e f f i c i e n t i n the VUV region i s more than several 0097-6156/89/0412-0424$06.00/0 o 1989 American Chemical Society

Reichmanis et al.; Polymers in Microlithography ACS Symposium Series; American Chemical Society: Washington, DC, 1989.

26. UENOETAL.

UV- andVUV-Induced Etching of Poly(methyl methacrylate) 425


hundreds times larger than that i n the UV region (see Figure 1) (5.) . Thus, we can expect more e f f e c t i v e reactions which produce v o l a t i l e species by VUV-induced main-chain s c i s s i o n . The use of such energyselective or s i t e - s e l e c t i v e reactions (6) would become more important not only for the photoetching applied i n the fabrication of semiconductor substrate but also for r e s i s t technology (7,8) i n microlithography. We present here the results of a mass spectroscopic study of the vaporized species produced during UV and VUV i r r a d i a t i o n of PMMA. A difference was observed between the mass spectra of the vaporized species produced by the two methods of i r r a d i a t i o n , indicating d i r e c t main-chain s c i s s i o n induced by VUV absorption of PMMA. Further, the effects of heating and introduction of oxygen on UV etching w i l l be described. Experimental We used two experimental setups as shown i n Figure 2. Mass analysis of species resulting from UV i r r a d i a t i o n was performed with method A (Figure 2a), including the studies on the effects of heating and introduction of oxygen on the etching reaction. The inner surface of an etch tube was coated with PMMA (thickness < 0.1 mm) and i r r a d i a t e d with the UV l i g h t of a commercial D lamp (Original Hanau D200F) after f i l t e r i n g by a quartz window. Thus the f i l m was exposed to UV l i g h t (hv < 6.8 eV) i n high vacuum. The vaporized species produced by t h i s exposure were introduced into the main vacuum chamber of a mass spectrometer and analyzed by a QP mass f i l t e r (ULVAC MSQ-150A) after ionization by 70 eV e~ bombardment. A l i g h t shutter was placed between the l i g h t source and the window. The f i l m could be heated and i t s temperature was measured with a thermocouple attached to the etch tube. For the etching experiment under oxygen, we introduced oxygen gas (>99.99%) into the etch tube at a pressure of 10~ Torr. Method Β (Figure 2b) was used for VUV etching of PMMA. The l i g h t source for t h i s setup was a capillary-gas-discharge lamp (9). The lamp was mounted on the main chamber with a d i f f e r e n t i a l pumping unit and a l i g h t shutter. In t h i s experiment, hydrogen gas was used f o r the discharge, since i t can emit intense VUV l i g h t i n the wavelength region of 85 - 167 nm (7.4-14.5 eV) (10). The l i g h t emitted from the discharge lamp was collimated by two glass tubings of an inner diameter of 1 mm, and i r r a d i a t e d the surface area of a specimen of about 4 mm i n diameter. As shown i n Figure 1, these photons can excite the main chain of PMMA. A PMMA f i l m (thickness< 1 μπι) spincoated on a S i wafer was placed on the rotatable sample holder. Both heater and thermocouple were mounted on the sample holder. For a d i r e c t comparison of etching products produced by UV and VUV i r r a d i a t i o n , we also used this setup for UV i r r a d i a t i o n by simply exchanging the capillary-discharge lamp for the D2 lamp. The l i g h t power density at the sample position was measured with a thermopile to be 33 mW/cm for UV and less than 0.02 mW/cm for VUV. The temperature r i s e s of the PMMA films resulting from the absorption of photons were n e g l i g i b l e . Although the intensity of VUV was much weaker than that of UV, the vaporized species by VUV i r r a d i a t i o n to PMMA were detected. This can be ascribed to the large absorption c o e f f i c i e n t of PMMA i n the VUV region (see Figure 1). For example, the penetration of VUV (Lyman α, λ=121.6 nm) into a PMMA f i l m i s compared i n Figure 3 with that of UV ( λ=213 nm) . 2

2

2

2


POLYMERS IN MICROLITHOGRAPHY

426

WAVELENGTH 200

0.2

(nm)

100

50

PMMA UV

VUV

< ι > υ LU LU Ο Ο

ζ

ο

0.1

H

/

Il

Γ

N

/

11

/MAIN CHAIH7

~~

1 / SIDE CHAIN Downloaded by FUDAN UNIV on February 2, 2017 | http://pubs.acs.org Publication Date: October 31, 1989 | doi: 10.1021/bk-1989-0412.ch026

ο

r-f/

> 12

18

30

24

ENERGY (eV) Figure 1. Absorption c o e f f i c i e n t of PMMA. The energy regions are i l l u s t r a t e d for main-chain and side-chain excitations. (Reproduced with permission from r e f . 5. Copyright 1978 American Institute of Physics.) Dt L A M P QZ-WINOOW ETCH

TUBE

PUMP

(a )

METHOD

A

Figure 2. Experimental setups for the mass analysis of vapor products a, with etch tube. Continued on next page.


UENO ETAL.

UV-and VUV-Induced Etching of Poly (methyl methacrylate)


H

(b)

GAS

2

METHOD

Β

Figure 2. Continued, b, With rotatable sample holder.

5000

2500 DEPTH (A)

Figure 3. Attenuation of UV and VUV l i g h t i n PMMA f i l m . : UV (λ =213 nm). — : VUV (λ =121.6 nm) . The intensity of l i g h t before and after absorption i s I and I, respectively. 0


428



Results and Discussion UV Etching. A t y p i c a l mass spectrum of the vaporized UV etching products i s shown i n Figure 4, together with a background spectrum obtained without UV i r r a d i a t i o n . The comparison c l e a r l y shows that UV i r r a d i a t i o n causes an increase i n intensity for various mass peaks. For example, the intensity of the peaks of m/e=15, 31, 59 and m/e=41, 69 increased d r a s t i c a l l y by UV i r r a d i a t i o n . The former three are due to side-chain scission caused by UV absorption at the C=0 unit, while the l a t t e r two are due to main-chain scission i n i t i a t e d by side-chain s c i s s i o n (11). The structure and mass numbers of t y p i c a l vaporized species are shown i n Table I. From here on, we use the spectral i n t e n s i t y after the background i s subtracted. Figure 5 shows the time dependence of the spectral intensity for m/e=31, 41, 69 and 100 during i r r a d i a t i o n . The intensity for the species of larger mass numbers increases more slowly with time than that f o r the smaller species. This means that the species of m/e=41 and 69 are not dominated by the fragments of MMA monomer (m/e=100) due to the e" bombardment i n the ionization chamber but by the etching products vaporized from the PMMA f i l m , since the fragments produced by the e~ bombardment of MMA should show similar time dependence to that of MMA. The difference i n the time dependence of the spectral intensity for the species with d i f f e r e n t mass numbers tended to be smaller when PMMA was i r r a d i a t e d at higher temperature. The tendency can be ascribed to the more rapid increase of the d i f f u s i o n constants for species of larger mass number i n PMMA with increase i n temperature. Such a consideration i s supported by experimental results that the etch rate increased with the f i l m thickness and with temperature i n the photoetching of the p o s i t i v e working r e s i s t (12-15), where the d i f f u s i o n of decomposed species i n the f i l m plays an important r o l e . In passing, we note that the intensity of MMA monomer ion was about 10^ times larger at 90°C than at 50°C, while the intensity of peaks of lower mass numbers due to side-chain s c i s s i o n did not show a comparable increase. In Figure 6, the drastic increase of the species due to main-chain scission as a result of heating i s shown; the increase of the species generated by side-chain s c i s s i o n i s also shown f o r comparison. We consider that the drastic increase i n intensity of MMA i s a result of e f f e c t i v e unzipping reaction i n vacuum occurring even at 90°C which i s s u f f i c i e n t l y lower than the c e i l i n g temperature Tc=164°C at 1 atm (16), since the vaporization of MMA reduces the concentration of MMA i n the f i l m followed by a reduction of Tc. Generally, the dominant mechanism of the UV etching of degradable polymer r e s i s t i n the presence of oxygen i s believed to be a reaction of polymer molecule with oxygen atoms produced by photodissociation of molecular oxygen (17). Therefore, i t i s interesting to compare the mass spectra of the etching products by UV-irradiation to PMMA i n vacuum and i n oxygen gas. The results are shown i n Figure 7, where the intensity variations of the mass peaks for m/e=44 (C0 ) and m/e=41 (ΟβΗβ*) are compared. A remarkable increase i n intensity of mass peak of m/e=44 (C0 ) was observed upon UV i r r a d i a t i o n i n oxygen, while other species did not show comparable increases. That i s , the integrated carbon number i n a l l of the +

+

+

2

+

2



26.

UENO ET A L

UV- and VUV-Induced Etching of Poly (methyl methacrylate) 429

m/e Figure 4. Typical mass spectra before (lower) and during (upper) UV i r r a d i a t i o n of PMMA at 50°C without 02 gas.

Table I. Molecular structure and mass numbers of selected products by photoetching of PMMA

15

16

28

29

CH

0

CO

CHO

31

0 3

55

59

CH 1

ι CH

CH

CH -C-CH 2 2

ι CH -C 2

3

1

CH -C

1

2

CH

1

C=0

3

2

100

CH

1

0

CO

3

83

69

C=0 3

44

41

J

CH

CH -C-CH 2

1

C=0

-c 2

2

1

c=o 1

0

I CH


430



100

60 120 EXPOSURE TIME (MIN)

Figure 5. Intensity variations of mass peaks of m/e=31(ο), 4 1 ( ο ) , 69(D), and 100 (Δ) as a function of UV-exposure time of PMMA at 50°C without 0 . 2

5

to

0

60 120 1Θ0 EXPOSURE TIME (MIN)

Figure 6. Temperature dependence of sum of peak i n t e n s i t y f o r m/e=41, 69, 100 (products by main-chain s c i s s i o n + unzipping ) (•;50°C, β;90°Ο and f o r m/e=15, 28, 29, 31, 44, 59 (products by side-chain scission) (o;50°C, #;90°C).


26.

UENO ET A L

UV- and VUV-Induced Etching of Poly (methyl methacrylate) 431


100

Ê

50h

0

Ê

(a)

m/e = 44

1 l 1 60 120 180 EXPOSURE TIME (MIN)

50h

60 120 EXPOSURE TIME (MIN) +

Figure 7. Intensity variations of mass peaks of m/e=44 (CC>2) (a) and m/e=41 (C3H5 ) (b) as a function of UV-exposure time f o r PMMA with and without Ο2 gas . · , A :With 0 gas . ο , Δ -.Without 0 gas. Temperature of the sample was 50°C. +

2


2


432

vaporized species did not increase appreciably by introducing oxygen gas during UV i r r a d i a t i o n to PMMA. We conclude that reactions of PMMA with oxygen during UV i r r a d i a t i o n do not play major role i n the etching reaction under the pressure range used i n our study. From the above r e s u l t s , the main UV-etching reactions of PMMA are considered as follows,

CH

CH

-CH -C- + C=0 2*

-CH -C-


2

Ο I CH

-CH -C- + Ο 2

CH + CH =C-

(1)

ι C=0 I ο I CH

CH

CH

-CH -C-

+ CH =C

(2)

,

C=0

CH

C=0

C=0

I Ο I CH

CH with Ο, Ο

-CH -C2 , C=0 I ο

CH -CH -C- + C=0 2 ·

(3)

VUV Etching. As i n the case of UV etching, we found that the mass spectral pattern changed with VUV-irradiation time. The results are shown i n Figure 8, where the contribution of the background i s subtracted. Further, the comparison of mass spectra during VUV and UV i r r a d i a t i o n without oxygen gas i s shown i n Figure 9. A clear difference was observed between the mass spectra observed for UV and VUV i r r a d i a t i o n . The main difference i s that the species of m/e=55 and 83 were c l e a r l y observed for VUV i r r a d i a t i o n . Such intense peaks of m/e=55 and 83 were never observed i n the decomposition of PMMA by UV i r r a d i a t i o n and heating. The structure of these species i s shown i n Table I. We ascribe the appearance of these species to the results of d i r e c t main-chain scission by the electronic excitation of the main chain by VUV absorption i n the following ways,


UENO ET AL.

UV- and VUV-Induced Etching ofPoly(methyl methacrylate) 433

1


xlO

.

2

VUV 60 min

\

8

69 55 \ „

..J.

xl 0

VUV 120 min

28 .

\

-83

/

J

LJ

1 4

J

20

L. VUV

xl 0

^

^

55

180 min

T

100

Jl, JL!

40

60

80

100

m/e Figure 8. Time dependence of mass spectra of vaporized products during VUV i r r a d i a t i o n to PMMA at 30°C. The time after the switch-on of the i r r a d i a t i o n i s indicated i n the figure. The intensity i s 10 times enlarged above m/e=60.



434

POLYMERS IN MICROUTHOGRAPHY

20

40

60

80

100

m/e Figure 9. Comparison of mass spectra of vaporized products by VUV (lower) and UV (upper) i r r a d i a t i o n of PMMA. The intensity i s 10 times enlarged above m/e=60.


26. UENO ET A L

UV- and VUV-Induced Etching ofPoly (methyl methacrylate) 435

νυν

VUV

CH

CH

Ç«3

-CH -C-

-CH„-C-CH -C-CH -C2 , 2 , 2. C=0 I 1 0 1 1

C=0 I 1 0 1 1

CH

3

CH

2

C=0 I 1 0 1 1

3

CH

3

+

C H

ι

+ C=0

C=0

I

"tc=o "Ί ο

I

CH CH„

CH Downloaded by FUDAN UNIV on February 2, 2017 | http://pubs.acs.org Publication Date: October 31, 1989 | doi: 10.1021/bk-1989-0412.ch026

(4)

CH =C 2

" Ι ο

I ο I

3

I

CH =C

-CH -C2, C=0

I

+

CH -C-CH2

-ic=o

Ο

I

ο I

CH

CH,

2

•c-1c=o

(5)

"["

0 1

CH,

Acknowledgments The present work was partly supported by Grants-in-Aid for S c i e n t i f i c Research from the Ministry of Education, Science and Culture (No. 61470018 and No. 62470100).

Literature Cited 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Urisu T.; Kyuragi H. J. Vac. Sci. Technol. 1987, B5 1436. Kyuragi H.; Urisu T. Appl. Phys. Lett. 1987, 50 1254. Yamada H.; Hori M.; Morita S.; Hattori S. J. Electrochem. Soc. Solid-State Science and Technology 1988, 135 967. Ueno N.; Mitsuhata T.; Sugita K.; Tanaka Κ. Jpn. J. Appl. Phys. 1988, 27 1723. Ritsko J. J . ; Brillson L. J.; Bigelow R. W.; Fabish T. J . J . Chem. Phys. 1978, 69 3931. Eberhardt W.; Sham T. K.; Carr R.; Krummacher S.; Strongin M.; Weng S. L . ; Wesner D. Phys. Rev. Lett. 1983, 50 1038. Ueno N.; Doi Y.; Sugita K.; Sasaki S.; Nagata S. J. Appl. Polym. Sci. 1987, 34 1677. Mochiji K.; Kimura T.; Obayashi H. Appl. Phys. Lett. 1985, 46 387. Ueno N.; Ikegami Α.; Hayasi Y.; Kiyono S. Jpn. J. Appl. Phys. 1977, 16 1655. Samson J. A. R. Techniques of Vacuum Ultraviolet Spectroscopy; John Wiley and Sons.: New York, London, and Sydney, 1967. Hiraoka H. IBM J. Res. Develop. 1977, 21 121. Ueno N.; Sugita K. Jpn. J. Appl. Phys. 1986, 25 1455. Lanagan M.; Lindsey S.; Viswanathan N. S. Jpn. J. Appl. Phys. 1983, 22 L67.


436


Ueno N.; Konishi S.; Tanimoto K.; Sugita K. Jpn. J. Appl. Phys. 1981, 20 L709. 15. Sugita K.; Ueno N.; Konishi S.; Suzuki Y. Photogr. Sci. Eng. 1983, 27 146. 16. Ivin K. J. Trans. Faraday Soc. 1955, 51 1267. 17. Vig J. R. In Surface Contamination, Genesis, Detection and Control ; Mittel K. L. Ed.; Plenum Press, New York and London, 1979; Vol. 1. RECEIVED June 29, 1989


14.


Mechanism of UV- and VUV-Induced Etching of Poly(methyl

Recommend Documents