Journal of Electron Spectroscopy and Related Phenomena 80 (1996) 349-352

Observation of Polymer Alloy by Spectral Soft X-Ray Microscopy with a Laser Plasma X-Ray Source

Akihiro TAKEICHI, Hirozumi AZUMA and Shoji NODA

TOYOTA Central Research and Development Laboratories Inc., Nagakute, Aichi 480-11, Japan

A polymer alloy film on a photographic plate (ILFORD, Q-plate) was exposed to the dispersed soft X-rays in the wavelength region between 8.1 nm and 33.6 nm. The image on Q-plate was magnified by an optical microscope. In the microscopic images of the polymer alloy containing silicon, circular spots of about 30 gm in diameter scattered in a matrix was observed with certain wavelengths of soft X-rays. The image contrast between the spots and the matrix was reversed at arotmd the wavelengths of 12.6 nm and 17.5 nm. This reversal of contrast suggests that these circular spots, or islands, are silicon-rich phases because the L-absorption edge of silicon is 12.3 nm.

1. INTRODUCTION

Soft X-ray microscopy can have higher resolution than optical microscopy and induces less radiation damage than electron microscopy. By using a soft X-ray having a wavelength suitable for a given object, a high-contrast image will be obtained because of selective absorption of the soft X-ray.

Polymer alloys are useful engineering materials. Determining microstructures of polymer alloys is important for their wider application. By using suitable target materials and suitable filters, microstructures of polymer alloys with known elements have been made observable in high contrast, l) Howeve r , the obse rva t ion of microstructures of polymer alloys with unknown elements is very difficult because a suitable wavelength cannot be found in advance. By determining the wavelength at which the image contrast is reversed and the absorption edges of an element, a constituent element of a specific region of the microstructure will be identified and a suitable wavelength for close observation of the microstructure can be found. 2)

2. EXPERIMENTAL

h'on rod targets were irradiated with a pulsed laser of 532 nm wavelength, 8 ns pulse duration and 0.4 J ] pulse energy at a repitation rate of up to 10 Hz deliverd from a commercial Nd:YAG laser (CONTINUUM, model NY81-10). The laser beam was focused with a spherical lens of 20 cm focal length to a spot of less than 50 Iam in diameter. The

0368-2048/96/$15.00 (c~ 1996 Elsevier Science B.V. All rights reserved PII S0368 - 2048 (96) 02989- I

irradiation intensity of approximately 2.5 u, 1012

W/cm 2 was achieved, which is sufficiently high togenerate soft X-rays in the wavelength range of 1- 40nm. l)

The block diagram of spectral soft X-ray microscopy with a laser-produced plasma as a soft X-ray source is shown in Figure 1. The spectrometer with a concave varied-spacing diffraction grating with 1200 grooves/ram (HITACHI, Model 001-043 7) was used to make spectrally dispersed soft X- rays. For microstructure imaging, the slit was removed and soft X-ray optics were slightly shifted from the exact alignment in the spectrometer, as described in the following section. The soft X-rays fi'om the laser-produced plasma were collected with a 2 deg grazing incidence Au-coated toroidal mirror of 40 mm in length and 20 mm in width. A specimen was placed on a photographic plate

Diffraction grating Specimen L . . . (120Ogrooves/mm) _/ aser plasma| soft x-ray .,::!,...~-~ _ I~r

Fe-target ~[::i:iiii • , ::::i:i;:i: :!!i:! ........ • w ~ [i[ 1

~:'::::!:~:ii iii:::.:i:.i!!~:.':::;:::?::::'""~ Dispersed • .~. " ;::i:?::i:i 11, soft X-rays

Lens'l" ~ Q-plate

............... !.~.La.ser..l~a.m. .......... To.roi.da.I mirror

YAQ Laser system E=0.4J, t=8nsec, R=10Hz

Fig. I. Block diagram of the spectral soft X-ray microscope with a laser-produced plasma as a soft X-ray source.

350

(ILFORD Q-plate), or a detector, which is sensitive vertically dispersed but horizontally narrow (about 2 mm wide) soft X-rays. An optical microscope was employed to magnify the image of the microstructure recorded on the Q-plate.

Acrylic copolymer (methyl methacrylic acid (MMA), n-butyl acrylate, and 2-hydroxy ethyl methacryl ic acid (HEMA) as component monomers), acrylic copolymer with silicone (MMA, HEMA and trimethoxy methyl silane (TMS)) and hexamethylene diisocyanate as a curing agent were blended in a solvent mixture of toluene and methyl- isobutyl-ketone, and the blended polymer solution was sprayed on a slide glass. The resulting polymer of about 10 tam in thickness coated on a slide glass was annealed for 1 hour at 150°C in air. The polyrner film was peeled fi'om the slide glass and used as a specimen.

3. RESULTS AND DISCUSSION

Figure 2 shows the spectral soft X-ray image of the specimen recorded on the Q-plate with about 30 00 shots of dispersed soft X-rays. Bright parts in the image are strongly exposed regions. In this figure, no clear line spectra from the iron target are observed and the intensity of the soft X-ray image changes gradually. This is because in the present microscope observation, the slit in the spectrometer was removed and the soft X-ray optics was slightly shifted fi'om the exact alignment in the spectrometer in order to defuse the images of the line spectra. The difference in the exposure intensity seems to result from absorption of the soft X-rays by the specimen. The exposure intensity is lower in the wavelength region longer than 25 nm, because the soft X-ray spectrum of iron target have hight intense lines between 4 nm and 25 nm. Since the microstructure of the specimen was unclear in this figure, the

soft X-ray images at different wavelengths on the Q-plate were magnified using an optical microscope.

Figure 3 shows the projected and magnified images of the specimen recorded on a Q-plate shown in Figure 2. In this figure, bright and dark parts of the image are the reverse of those in Figure 3; dark parts of the image are strongly X-ray- exposed regions. The images recorded directly on to the Q-plate at thirty different wavelengths of soft X- rays ranging between 8.1 nm and 33.6 nm were magnified and typical images at six different wavelengths of soft X-rays are shown in Figure 3. In Figs. 3(a), 3(d) and 3(e), many dark circular spots in the bright matrix are observed, while in Figs. 3(b), 3(c) and 3(f), bright circular spots in the dark matrix are observed. The spot size in Figs. 3(a), 3(d) and 3(e) is almost the same as that in Figs. 3(b), 3 (c) and 3(f). In other words, bright circt, lar spots in the dark matrix are observed in the wavelength range between 12.6 nm and 17.5 nm, while dark

(a) {d)

E

(b) (e)

(c) (f)

i I I i I t I i I ~

5 7.5 10 15 20 25 30 35 40

Wavelength (nm)

Fig.2. Spectral soft X-ray image of the specimen with about 3000 laser shots fi'om an iron target.

100 #.m

Fig.3. Magnified soft X-ray images of the specimen exposed to spectral soft X-rays in the wavelength region between 8.1 nm and 33.6 nm. The nominal wavelengths exposed are (a) 8.6 nm, (b)13.2 rim, (c) 15.2 nm, (d) ! 9.4 nm, (e) 24.1 nm and (f) 33.6 nm.

circt, lar spots in the bright matrix are observed at wavelengths outside this range. The magnified images indicate the microstructure in which circular spots of about 30 pm in diameter are scattered in the matrix. Their contrasts are weak in the wavelength range between 25 nm and 33.6 nm. It is interesting to note that the image contrast between the spots and the matrix is reversed at around the wavelengths of 12.6 nm and 17.5 nm. These images were taken at different places of a specimen. Thus, it is impossible to compare directly the images recorded from different places on the Q-plate or recorded at the different soft X-ray wavelengths. However, the microstructural inhomogeneity of the specimen is believed to be not so serious in the present study. A few dark small circular spots were also observed in magnified images in all wavelength regions. Careful observation indicates that these small dark circular spots are bubbles in the s pec i me n.

The linear absorption coefficients of the relevant chemical elements can be calculated using

It = Io exp [ -2roZt -~.12i.fi ] , (1)

where I t is the intensity of the transmited X-ray, I 0

is the intensity of the incident X-ray, r 0 is the

classical electron radius, /l is wavelength, t is thickness of the specimen, f2i is an imaginary

component of the atomic scattering factor of the i-th chemical element, and fi is the number of atoms in a unit volume for the i-th chemical element. 5) The linear absorption coefficients of the polymer alloy film were calculated from eq. (1) with the number of atoms in a unit volume for the constituent elements of the polymer alloy film (C : 56.07 %, H : 8.50 C/c, N : 1.86 c/c, O : 31.57 %, Si : 2.01%).

Figure 4 shows the calculated linear absorption coefficients for acrylic copolymers and silicon. In this calculation, the number of silicon atoms in a unit volume is assumed to be 112.8 atoms/nm3. In the wavelength region longer than 12.6 nm, the value of the linear absorption coefficient for silicon is smaller than those for the constituent elements of the polymer alloy film, because the wavelength of the L-absorption edge for silicon is 12.3 nm. Since the line spectra were diffused by slit removal and slight off-alignment in the optical system, the image contrasts did not drastically change at the wavelength of the silicon L-absorption edge. The contrast reversal in the microstructure image arot, nd the wavelength of the absorption edge of silicon

351

indicates that the circular spots are silicon-rich regions. These results have been confirmed by line analysis of the specimen using an electron probe microanalyzer (EPMA). Taking into consideration the absorption coefficients for acrylic copolymer and silicon at each wavelength explains the contrast in Figure 3, except for that in the wavelength range between 15 nm and 33.6 rim. Note that the images recorded on the Q-plate are superpositions of a given soft X-ray (first order) and higher-order soft X-rays fi'om the diffraction grating.

Assuming the intensity ratio of the higher-order X-ray to the first-order X-ray to be rather high, the first-order X-ray (e.g., ~'1) as well as the higher-

order X-rays (Z / n; n = 2, 3 ... . ) from the grating, will pass through the specimen and irradiate the Q- plate. Therefore, the estimation of the transmittance of the soft X-ray through the specimen is important for understanding the spectral contact microscope image.

The transmittance is expressed by

X-" l( Z/n ) rn( 2/n ) exp[-p( ,~Yn ) t]

T(Z) = (2) I (M

where T(2) is the transmittance, I(Z) is the intensity of the soft X-ray of the iron target, r,~(Z) is the

intensity ratio of the n-th-order X-ray to the first-

C K-edge

10 , .i, e - . . . . . . . . . . . , , . . . . . . . . . . . .

E° l°s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . li i v

10 4

103 ............................ o

.~ 10 2 lic CoPolymer } ................ ! ........... ]

V 1 ..Q <

10 0 . . . . . . . . . . . . . . . . .

0 5 10 15 20 25 30 35 40

Wavelength (nm)

Fig.4. Linear absorption coefficients of the acrylic copolymer and silicon. The characters a, b, c, d, e and f indicate 8.6 nm, 13.2 nm, 15.2 nm, 19.4 nm, 24.1 nm and 33,6 nm.

3 5 2

order X-ray, p(2/n) is the linear absorption coefficient of the acrylic copolymers or silicon, n = 1, 2, 3, 4, 5, and t is the thickness of the specimen. The transmittances for the acrylic copolymer and silicon of 10 !am thickness calculated for first-, second-, third-, fourth- and fifth-order soft X-rays are shown in Figure 5. The figure shows the transmittance determined mainly by the absorption edges at about 4 nm for the acrylic copolymer and at about 13 nm for silicon. In the wavelength ranges between 4 nm and 13 nm and between 18 nm and 26 nm, acrylic copolymer has higher transmittance than silicon. Moreover in the wavelength region longer than 25 nm, the transmittance of acrylic copolymer suddenly decreases. Since acrylic copolymer is the main constituent of the specimen even in a silicon-rich region or a circular spot, the transmittance of acrylic copolymer will determine that of the specimen. This consideration is consistent with the experimental results. Therefore, the spectral contact microscope image includes the images produced by the first-order soft X-rays as well as the those by the second- and higher-order X- rays of shorter wavelength. This causes the indistinct contrast of microstructures observed in the present spectral contact microscopy. More detailed quantitative analysis, for instance, examination of the silicon contents in the microstructure, is

0.04

. . . . . . . . . . . . . . . ,Acr,,,ccooo,ymer "ii . . . . . . . . . . . . . . . . . t 0.03 ........ l s t i - - / ........ i ........ I ................. i ........ i ........ i ........

° iii "d o .. . . . . . . i ........ i . . . . . . . . . . . . . . . .

. . . . . . ~ . . . . . . i . . . . . . . . . ! . . . . . . . . . ~ . . . . . . . . ~ . . . . . . . . 4 . . . . . . . . 4 . . . . . . . . . i. . . . . . . . .

. . . . . ' " ~ /~.,~....~:./...~ . . . . . . . . . i - 2 n d . . . . . . . i . . . . . . . . i . . . . . . . . . i. . . . . . . . .

I - 0 . 0 1 ....... , , , ~:,',',',',',',','w, ~4tlh i i i i

0 10 20 30 40 50

AiIAA A a b c d e f

Wavelength(nm) Fig.5. Calculated transmittances of the acrylic copolymer and silicon of 10 pm thickness as a function of wavelength. The characters a, b, c, d, e and f indicate 8.6 nm, 13.2 nm, 15.2 nm, 19.4 nm, 24.1 nm and 33.6 nrn.

required. In this experiment, it was difficult to obtain separately the transmittance of the circular spot in the specimen because the spot size ( < 30 pm ) was smaller than the window size ( 10 pm x 1 O0 pm ) of the micro-photometer.

4. CONCLUSIONS

Spectral soft X-ray contact microscopy with a laser plasma as a soft X-ray source has been found to be useful for observation of microstructures of a phase-separated polymer alloy. A polymer alloy specimen on a Q-plate was exposed to spectrally dispersed soft X-rays. Bright circular spots of about 50 pm in diameter in a dark matrix were observed in the wavelength region between 12.6 nm and 17.5 nm, whereas dark circular spots in a bright matrix were observed at wavelengths outside this range.

The reversal of contrast in the above wavelength region suggests that these circular spots are silicon- rich regions, because the absorption edge of silicon is 12.3 nm. These results demonstrate that spectral soft X-ray contact microscopy with a laser plasma as a soft X-ray source is an effective method of determining a suitable wavelength region for clear microstructure imaging, as well as of identifying constituent elements.

REFERENCES

1. H.Azuma, A.Takeichi and S.Noda: Jpn. J. Appl. Phys. 33 (1994) 4622.

2. H.Azuma, A.Takeichi, I.Konomi and S.Noda: Rev. Laser Eng. 21 (1993) 724.

3. B.L.Henke, S.L.Kwok, J.Y.Uejio, H.T.Yamada and G.G.Young: J. Opt. Soc. Am. B 1 (1984) 818.

4. R.L.Kelly: J. Phys. Chem. Ref. Data 16, Suppl. 1 (1987) 737.

5. B.L.Henke, P. Lee, T. J. Tanaka, R. L. Shimabukuro and B. K. Fujukawa :At. Data & Nucl. Data Tables 27 (1982) 1.