Laser-induced synthesis, deposition and etching of materials

Bull. Mater. Sci., Vol. 11, Nos 2 & 3, November 1988, pp. 137 157. ~¢' Printed in India.

Laser-induced synthesis, deposition and etching of materials

S B O G A L E

Department of Physics, University of Poona, Poona 411 007, India

Abstract. The field of laser-induced synthesis, deposition and etching of materials is reviewed with an emphasis on the emerging trends and novel adaptations of the basic laser processing concepts. A number of examples are cited to illustrate the issues involved. These include rapid synthesis of titanium nitride by pulsed laser induced reactive quenching at Ti:liquid NH3 interface, laser deposition of good quality thin films of such materials as hot oxide superconductors, zinc fcrritc, iron oxide, stainless steel, etc. and laser etching of superconductor films.

Keywords. Laser processing: reactive quenching: etching: deposition.

I. Introduction

Beam solid interactions has become a major area of research in recent times in view of its scientific richness and relevance to high technology (Picraux et al 1987).

A number of techniques based on the use of energetic beams of ions, electrons and photons are already in use in the microelectronics industry and have been responsible for the realization of very large scale integration of device features. Yet the field appears to be far from saturated both in terms of the emergence of new processing concepts and in their implementation. This is evidenced by the growth rate of innovative scientific publications in this area. It is true, of course, that with the newer developments in the field as well as the changing requirements of modern technology, the emphasis of the field has been changing from time to time.

Amongst the beam processing concepts the most important ones are ion implantation (Carter and Colligon 1968; Poate et al 1983; Ullirich 1985; Picraux et al 19871, laser processing (Wood et al 1984; Ullirich 1985: Donnelly et al 1987), and electron beam treatment (Picraux et al 1987) of materials. These techniques, either separately or jointly, have contributed significantly to the rapid growth of the science of advanced material processing. In this article we shall only focus on some important aspects of laser material processing. Lasers attracted the attention of materials scientists for the first time when it was demonstrated that nanosecond pulsed lasers can anneal out damage in ion-implanted materials, especially silicon, leading to a good-quality, device-worthy epitaxial layer on the surface with a very high degree of electrical activation of the dopants (Celler et al 1978~ Foti et at 1978;

Hoonhout 1981) This led many laboratories all over the world to plunge into this area, with an emphasis on laser annealing of semiconductors, Significant work has already been generated on this aspect and it will be reviewed in another contribution to this volume. Recently, however, the emphasis of laser processing activity has drastically shifted to other areas that are equally vital to technology, viz. laser-induced synthesis, deposition and etching, because laser applications in these areas have been found to be truly interesting, innovative and novel in their scientific content as well as technological utilization. In this article I shall focus on these newer applications of lasers in materials sciencc and shall try to bring out the fact that there is great scope for imaginative research in this field.

137

138 S B Ogale

The processes of laser-induced deposition, synthesis and etching offer a number of advantages over the conventional means of achieving comparable results, and under certain specific situations lead to characteristically novel results which can have far-reaching consequences (Emery et al 1984; Allen et al 1984; Black et al 1987). The directionality and monochromaticity of laser beams and the possibility of compressing the time-scales of their operation by pulsing are important elements of the processes mentioned above. These features render the field of laser material processing technologically important and academically interesting. Also, unlike electron mad ion beams, laser beams can be propagated through a wide variety of chemically reactive environments. The directionality of the laser source allows the energy to be precisely aimed at an area as small as the wavelength of the beam itself, and this can lead to localized deposition. Monochromaticity of the beam and tunability of laser wavelength can be used very effectively to deliver energy to the electronic or vibrational excitations of specific reacting species and such selective energy flow into the system allows the deposition or synthesis to occur at tem- peratures much below those required by equilibrium considerations. When pulsed lasers with pulse durations of a nanosecond or smaller are used, metastable phases can be synthesized by overcoming the barriers of equilibrium solid solubilities.

Metastable phases and solid solutions can also be synthesized using CW lasers provided such laser beams are rapidly scanned over a given sample area.

The physical character of laser-material interaction and subsequent material synthesis and modification can be conveniently classified into two categories, viz.

thermal and photochemical. In figure la is shown a typical arrangement used in a laser processing experiment aimed at achieving localized synthesis, deposition or etching; in figure lb is shown an alternative arrangement for large-area deposition operations. When thermally driven chemical reactions are to be used for deposition or synthesis, the laser is made incident on an absorbing substrate so as to generate a local hot spot which drives the chemistry between the participating atoms, molecules and molecular complexes in the neighbouring ambient. Whether the substrate itself participates in the reactions or not depends on its chemical reactivity and thermodynamic properties. Clearly for such an operation to be successful the ambient itself is required to be reasonably transparent to the laser wavelength. The opposite conditions apply when one desires deposition via photochemical routes.

Thus the reactants absorb laser energy directly and chemically react in their excited states so as to finally settle on an adjacent substrate at reasonably low temperature via diffusion-limited transport mechanisms. It is possible to control and use both the thermal and photochemical routes by using multiple laser beams of different wavelengths. Thus the basic advantageous features of laser-induced material deposition and synthesis are: (i) high spatial resolution (region selectivity) and control; (ii) fast processing leading to higher deposition rates; (iii) localization of heating in thermochemical processing and etching, and low-temperature processing in photochemical deposition; (iv) increased purity of deposited and synthesized products; and (v) ability to interface easily with other laser-related techniques such as laser annealing, laser alloying, etc.

The process of laser chemical vapour deposition (LCVD) differs from the conventional chemical vapour deposition (CVD) process in its dependence on some surface reaction kinetics and substrate characteristics. Since the hot spot is strongly localized in the case of LCVD, much higher reaction temperatures can be employed

Laser-induced synthesis, deposition and etchin9 of materials 139 than those possible in the case of CVD. The localization of heating also eliminates preheating of large substrate volume and thereby the degree of unwanted contaminants. It further helps to achieve very high reactant concentrations in the

"point sink" geometry employed in laser deposition experiments, leading to increased diffusion of reactants towards the surface. This enhancement of diffusion rate in turn allows higher temperatures to be accessed and thereby deposition rates higher by many orders of magnitude than the conventional CVD rates can be achieved in LCVD processes. The rates of deposition or synthesis in the LCVD and related processes are limited by diffusion of reactants to the active reaction zone and convection effects in participating fluid ambients. The convection effects also have a bearing upon the homogeneity of deposits and care must be exercised to achieve good film deposits by LCVD.

Finally it is important to point out the intrinsic nonlinear self-limiting characteristic of the laser-related deposition and synthesis processes. When the LCVD process is used to deposit a highly reflecting metallic film on an absorbing substrate, an initial film deposition reduces the subsequent coupling of laser density into the medium and the metal film deposition rate drops down. On the other hand an increase in the coupling of laser density can occur when an absorbing film is

I LASER

+-Y

I ~ SUBSTRATE

(B)

m-

~TO VACUUM

~_m

PHOTON FLUX CHAMBER WITH ARRANGEMENT FOR REACTIVE GAS INCORPORATION

~ L A T E R A L LASER BEAM

=.-SUBSTRATE

~-'-SUBSTRATE HEATIN6

Figure 1. Laser chemical vapour deposition--system configurations for (A) localized synthesis, deposition or etching; (B) large-area deposition.

140 S B Ogale

deposited on the substrate. Thus the physics of the LCVD process can be learnt primarily by time-resolved measurements and not by simply characterizing the final state of the material obtained. Since understanding of the physics of laser-material interactions has a direct impact on the possibilities of designing new experiments and reaction routes, it is essential that any programme on laser-solid interactions is adequately supported by a programme aimed at performing time-resolved measurements (Auston et al 1978; Ferris et al 1979; Lo and Compaan 1980; Aydinli

et al 1981; Stritzker et al 1981; Galvin et al 1982) of such properties as reflectivity, transmittivity, resistivity, etc.

2. Laser-induced synthesis

The LCVD and related techniques have been applied to a wide variety of materials and systems. These include metals such as W, Fe, Mo, Cr (Dentsch et al 1979; Allen and Tringubo 1983; Grossman and Karnezos 1987); semiconductors such as Si, Ge, GaAs; (Ehrlich et al 1981; Allen et al 1982; Andreatta et al 1982) and compounds such as Si3N 4, SiO z, aluminium nitride and oxynitride, TiN, T i O 2 (Allen 1981:

Demiryont et a! 1986; Sugimura et al 1987). In most of these experiments a solid- vapour interface is processed with a pulsed or CW laser, and for details of such work I would refer the reader to the cited references. I would prefer to discuss a new process recently developed in my laboratory (Ogale et al 1986, 1987a; Patil et al

1987) and subsequently studied elsewhere (Dijkkamp et al 1987a; Ogale et al

1987b). In this process a high-power pulsed laser was used to process the interface between a given solid and a suitably chosen reactive liquid so as to achieve rapid compound synthesis via reactive quenching at the interface. The details of this process and its implications have been discussed in another article in this volume.

Here only two examples are discussed to illustrate the method for the case of technologically important materials, viz. titanium nitride and tungsten carbide.

2.1 Nitridation: T i / N H 3 case

This experiment was performed in collaboration with the FOM Institute, The Netherlands (Ogale et al 1987b). A pulsed XeCI excimer laser (2=308 rim, pulsewidth=25 ns) capable of giving maximum energy per pulse of 500 mJ was used. Energy densities of up to 3 J/cm z could be obtained within an acceptable degree of homogeneity by suitable beam guiding and focusing to a spot size of a few mm 2. While processing the samples in liquid ambients, the depth of liquid covering the sample surface was maintained at 1"5 mm. The details of the preparation procedure of pure liquid ammonia on the sample has been described elsewhere.

Figure 2 shows a Rutherford backscattering spectrum for polycrystalline titanium foils pulsed laser treated in liquid ammonia at a temperature of -60°C. A reference spectrum of untreated Ti is also shown. The spectrum shows a significant depletion of Ti signal near the surface and a small nitrogen peak. This peak, obtained after appropriate substraction of the background Ti signal, is also shown enlarged in the figure, together with the native oxygen peak of the untreated sample. Simulation of the spectra revealed a reasonable agreement between the depletion of Ti signal and the area of the N peak. It was found that about 2000/~ of Ti had reacted. The N surface concentration after a single laser pulse was found to be as high as 40 at. %.

Laser-induced synthesis, deposition and etchin.q ol materials 141

1¢,0

12C

10C

q

2C

Figure 2.

ammonia.

. •

N 0 . 3 1 / c m 2

x 10

0'.s

ENERGY (MeV)

RBS spectra of unirradiated titanium and titanium ~rradiated in liquid

After N incorporation was shown by Rutherford backscattering spectroscopy (RBS), X-ray diffraction (XRD) was employed to find out whether compounds had formed in the surface layers. The corresponding result, shown in figure 3, reveals the formation of TiN compound in the surface layers.

2.2 Carbide formation: W / C 6 H 6 case

The use of the reactive quenching process was demonstrated for carbide phase formation also (Ogale et al 1987a) We chose tungsten carbide as the test system because of its technological importance. In this experiment appropriately cleaned tungsten foils were pulsed ruby laser treated under benzene at energy densities of 4.7 J/cm 2 (sample A) and 6-7 J/cm 2 (sample B). The samples were characterized by using the technique of small angle XRD. The XRD pattern corresponding to a virgin tungsten sample is shown in figure 4a. This pattern consists of four characteristic lines at 20 values of 40.20 °, 58"20 '~, 73"20 °, 87'00 °, representative of W.

The line intensity ratios do not match with the data reported for powder samples owing to the presence of texture effects induced during the synthesis of the foil itself.

The XRD pattern for sample A (figure 4b) shows significant changes with reference to the untreated sample. The analysis of different diffraction peaks indicates the formation of --- W~C (hexagonal) and - WC (WC~ -x) phases.

In sample 13. Jreared al highcl encrg3 density, formation of equilibrium

W3C

phase is observed (figure 4c). Interestingly enough, W3C phase with A-15 type structure has been reported only recently by Bhat and Holzl (1982) and Srivastava et al (1984).

In the experiments described here it was shown that the reactive quenching process can be used to achieve interesting surface synthesis results. The quenching aspect of processing under a liquid ambient can also be used in another interesting way in the context of laser annealing of silicon. We discuss this briefly in the following.

142 S B Ogale

:l

o

O.Z

0.2 34

TIN 111 T i ~ .

... ] i...+"++. 4 ... ..I""

i I I I

36 38

Ti

002

\ } ...

,'o

011 Ti TiN 200

• ".: ...+,....+... ... : -,j-...

42 44

2 0

Figure 3. XRD spectrum of titanium irradiated in liquid ammonia ( - - - ) . The reference spectrum for untreated titanium is also shown ( . . . ).

t

> - I-- Z LU I,- Z i-,4

• ] :

+++ ++i

e,.+A + ~.+-•,.+

20

t.O 60 80

- - 20 (DEGREE)

Figure 4. Glancing angle XRD patterns of (a) an untreated tungsten foil; (b) and (e) tungsten foils laser treated in liquid C6H 6 (benzene) at energy density values of 4.7 J/cm 2 and 6.7 J/cm 2.

90

2.3 Silicon pulsed laser annealing under water

In this w o r k ( P o l m a n et al 1987) laser i r r a d i a t i o n in w a t e r a m b i e n t was c a r r i e d o u t o n low e n e r g y A s - i m p l a n t e d Si samples. S i ( 1 0 0 ) was i m p l a n t e d at r o o m t e m p e r a t u r e by 1 0 k e V As + i o n s to a d o s e o f 3 x 1015cm -2. U s i n g this low i m p l a n t a t i o n e n e r g y s h a l l o w i m p u r i t y profiles a r e o b t a i n e d . Arsenic is k n o w n to

Laser-induced synthesis, deposition and etching of materials 143 diffuse rapidly in liquid silicon on a nanosecond time scale and does not exhibit segregation effects at the liquid-solid interface under the present nonequilibrium con- ditions. Pulsed laser irradiation was performed using a Q-switched ruby laser (2 = 694 nm, pulsewidth = 32 ns). Laser energy was varied using volume absorbing neutral density filters. A quartz guide diffusor was used to obtain 5% uniformity over a 6 mm diameter spot. A mean energy density at the diffuser tube end was calibrated to within 3% using calorimetry. The details of the experimental set-up have been described elsewhere.

Arsenic concentration profiles before and after irradiation were determined by RBS using a 2 MeV He + beam in a random direction close to the surface normal.

Reflected He particles were detected and energy analysed by a surface barrier detector (energy resolution 14 keV) at a scattering angle of 105 °, which gave in a depth resolution of 150/~. Channelling was employed to study the sample's structural state after irradiation. Figure 5 shows the full width at half maximum of As profiles as a function of laser energy density in the end of the quartz tube.

Results for samples treated in water and samples treated in air are compared. The thresholds for epitaxy (epi) are shown in the figure for two cases. For all samples broadening of As profiles is observed relative to the profile for an As-implanted sample.

First we will discuss the As peak width after irradiation in air (filled circles). For low energy densities (<0.5

J/cm 2)

40

30 :E T LL

/

" 0 / i I l I i I I

/ Laser irr.

/ in AIR

i I O / I

/ / i epl ///

20

p,

o t l l •

r ,

k . j _ r -

I ~%y~te~ I ~pi

I resotution I =

( ] I I [ I I I

0 0.4 0.8

- - LASER ENERGY DENSITY ( ] / c m 2) /

¢

/ L a s e r irr in WATER

I i

1.2 1.6

Figure 5. F W H M of As profiles obtained from RBS spectra without correction for the system resolution as a function of energy density in the diffusor end. Results for irradiation in water and in air are compared.

144 S B Ogale

detector resolution of 7 channels, and therefore changes in the profile cannot be accurately detected. Between 0.5 and 0-7J/cm 2 the peak width increases significantly owing to increased melt depth in the amorphous material. Above the threshold for epitaxy (0.7 J/cm z) the melt depth is expected to increase further with energy density and as a consequence we see a considerable increase in As peak width.

The As peak width behaviour after irradiation under water (open circles) is distinctly different from the behaviour described above. Up to the threshold for epitaxy only slight changes in the As profiles are detected. The As peak width at the threshold for epitaxy is smaller than that at this threshold in the case of irradiation in air. For higher energy densities the difference in As spreading between the two cases becomes even more distinct. Above the threshold for epitaxy the As peak width increases only slightly with energy whereas for irradiation in air a rapid increase is found. Thus the use of liquid overlayer during laser annealing of silicon compresses the melt duration due to an associated heat transfer effect. It is also important to mention that in the case of treatment under water no significant quantity of oxide is formed on the surface. This is probably due to the diffusion- limiting action of the initial surface oxide formed.

3. Laser-induced deposition

Having discussed some interesting applications of pulsed laser induced interface processing I shall now present some results from another example of the use of lasers in materials processing, viz. thin film deposition by pulsed laser induced vaporization/ablation from bulk.

3.1 Deposition of thin films by pulsed laser evaporation technique

The motivation for the use of pulsed laser for film deposition is derived from the highly nonequilibrium nature of the attendant vaporization process. Such an evaporation process can be expected to lead to near-stoichiometric emission from a complex target and correspondingly it can be hoped that a film with stoichiometry close to that of the bulk could be obtained by this route. That this is indeed true can be seen from the examples discussed below. It is important to point out that pulsed lasers in the UV, visible or IR range can be used for this purpose (Dijkkamp et al 1987b; Wu et al 1987; Hansan and Robitaille 1988; Joshi et al 1988) but the detailed nature of the vaporization can be different in these cases. The UV lasers are particularly interesting because these lasers have been shown to be capable of ablating delicate materials such as polymers without breaking the chains. (Dijkkamp e; al 1987; Hansan and Robitaillc 1988) This cannot be realized by a,,dng ~i,,ibk' light or IR lasers.

3.2 Thin films of Y-Ba-Cu-O high temperature superconductor

The idea of using pulsed lasers for synthesis of hot superconductor films as well as a successful demonstration came from Venkatesan's group at Bell Communications Research and Rutgers University. This group showed that tile pulsed vaporization character of the process leads to transfer of the stoichiometry from the bulk to the film without any significant modification, and, as such, superconducting films can

Laser-induced synthesis, deposition and etchin,q of materials 145 be obtained without difficulty. In the following we briefly discuss the experiments of the Bellcore/Rutgers team because a discussion on pulsed laser deposition cannot be complete without this exciting development.

In the experiment of Venkatesan and others (Dijkkamp et al 1987c; Wu et al 1987; Allen et al 1984; Emery et al 1984), pellets with normal compo- sition YBazCu30~_x were prepared in the usual way. Electrical measurements of this bulk material using a four-point probe with indium-soldered contacts showed an onset temperature of 95 K and a transition width of 0.3 K. A pellet (1 cm in diameter, 2 mm thick) was mounted in a small vacuum system with a base pressure of 5 × 10 7 Torr, and irradiated through a quartz window with a KrF excimer laser (Lambda Physik EMG200E, 30 ns full width haft maximum, 1 J/shot) at 45 ° angle of incidence. A quartz lens was used to obtain an energy density of approximately 2 J/cm 2 on the target. SrTiO3 (100) (which is lattice-matched with the Y-Ba--Cu-O perovskite along specific orientation) and vitreous carbon (for RBS studies) substrates were mounted at a distance of typically 3 cm from the pellet surface, close to the normal from the centre of the laser spot. The laser was fired at a repetition rate of 3~-6 Hz, for a total of a few thousand shots. With each shot, a plume of intense white light emission could be observed normal to the sample surface. During deposition, the pressure in the system rose to about 1--2× 10 -6 Torr. To obtain a more constant deposition rate and avoid texturing, the pellet was slowly rotated in subsequent experiments.

The films were deposited with the substrates heated to 45ff~C. This resulted in shiny, dark-brown films with strong adhesion, but still electrically insulating. The films were annealed in an oxygen atmosphere for typically 1 h at 9 0 0 C followed by slow cooling to room temperature. These yielded the desired low resistivity (in the 10 -3 ohm-cm range) and superconductive transition.

Film composition was examined by RBS. The structure in the film was determined by XRD measurements. The RBS spectrum of a film deposited at 450~'C on vitreous carbon showed the film composition to be Y~Ba2Cu30 6, which is closely matched with the bulk pellet composition (see figure 6). This result clearly showed that the bulk stoichiometry is adequately transferred to the thin film without a major modification. The results of XRD study performed on the films showed preferential orientation, with the c-axis of the 3-fold stacked perovskite YBazCu30 7_x normal to the plane of the film (see figure 7).

The resistivity versus temperature (measured with a four-point probe) for the film on a (100) SrTiO 3 substrate showed an onset of transition at 95 K and a zero resistance state at 87 K (see figure 8). Venkatesan's group has also succeeded in achieving channelling in these films demonstrating their epitaxial orientation. Very recently this group has also been able to reduce the highest process temperature to less than 650 ('. whi,.!~ i:, ~-, i~p,,,t"~'~l .,.chi:,.L~v,,:~ iVc~k,tc.~:~, p,i,~.:c communication). Thus, pulsed laser evaporation is a very attractive dry process for deposition of thin films of comI.zlex oxides. We illustrate another example to emphasize the point. This refers to the deposition of zinc ferrite films (Nawathey et al, in press).

3.3 Zi~w.lerrite (ZnxFe3_ x O j films

We used the zinc ferrite system as a vehicle for experimentation because this ferrite shows a number of interesting physical properties which are also useful in

146 S B Ooale

O J

Q LU N

O Z

ENERGY (MeV)

1.5 2 - 0 215 3 ' 0

! m m !

Deposited on carbon

. . . . Simulation of 190 nm Y1 Ba2Cu306

.J

3OO

3 . 5 .

!

Ba

O

xlO

!

Cu

, L | ,

coo 5oo 6Go 7oo 8oo 96o

CHANNEL

4.0

Figure 6. RBS spectrum (4 MeV He 2+, 175 ° backscattering angle) of an as-deposited laser-evaporated superconductor film on a carbon substrate (solid line), overlayered with a simulation (dashed line).

1 0 0

90

A 8 0

• - ~ 70

- ~ L _ 60

<

" - " 50

>-

I-- ,..., 40

O3 Z uJ 30

I--- Z O n

(001)

SrTiO 3

(005)

I

SrTiO 3

10 15 20 25 30 35 40 45 50 55 60 2 e (Degrees)

Figure 7. XRD spectrum of an annealed 350-nm-thick superconductor film on (100) SrTiO 3. The strong presence of (001) YBa2CU3OT_ ~ reflections indicates preferential orientalion along the c-axis.

a p p l i c a t i o n s ( D o b s o n et al 1970; S r i v a s t a v a et al 1976a, b). Specifically, t h e z i n c f e r r o u s ferrites ( Z n ~ F e 3 _ x O 4 ) s h o w a p p r e c i a b l y h i g h e r s a t u r a t i o n m a g n e t i z a t i o n

Laser-induced synthesis, deposition and etchin9 of materials 147

1500

v

I(]OO

750 500 ' " 250

0! 100 150 2(]0 25o

TEMPERATURE (K)

1

50 300

Figure 8. Resistivity vs temperature curve for 350 nm thick and annealed superconductor film deposited on (1001 SrTiO3. The transition width (90% 10%) is about 2 K.

compared to Zn-substituted Ni and Mn ferrites and thus can be used with advantage in several applications requiring high values of 4riMs. When studied for their magnetic properties by susceptibility and M6ssbauer measurements the zinc ferrous ferrites, as well as other zinc substituted ferrites, show significant relaxation effects at T/Tc values much less than l, in contrast to the case with other, non-zinc ferrites. Such effects in the zinc-based systems have been attributed to domain wall oscillation and relaxation phenomena related to magnetic microstructure in the system and are found to be sensitive functions of the zinc concentration. This makes the zinc-based ferrite a suitable candidate for the present study since one of the objectives of this study is to examine the relation between the film and the bulk stoichiometry.

The zinc ferrite pellets used as targets for laser vaporization were prepared by conventional chemical processing methods. The structure of the pellet was deter- mined by X-ray powder diffraction method and the pellets were found to be essen- tially ZnxFe3_xO4 single phase materials with x close to 1. The average particle size in the pellet was determined to be ~ 1 pm. In a given deposition experiment one such pellet was mounted in a chamber pumped by a Varian Diffstack system capable of rendering an ultimate vacuum of 5 x 10-7 torr. The ruby laser beam was derived from J K laser (System 2000) capable of yielding a maximum energy of 10 J/pulse at a pulse repetition rate of 2-3 pulses/min. The laser pulse duration was 30 nsec. The depositions were carried out on single crystal AlzO 3 substrates at substrate temperatures of 200°C and 450°C and at oxygen partial pressures of 1 x l 0 - 6 Torr and 5 x 10 -4 Torr. In a given deposition 300 laser shots were used.

Typical deposition rate was found to be in the range of 10~15 ,~/pulse depending on the deposition conditions.

The deposited films were characterized by small angle XRD measurements (performed using a Rigaku Rotaflex RU 200B using Cu K, radiation). In the small angle XRD measurements Seeman-Bohlin geometry was employed (Tu and Berry 1972) and the grazing angle of incidence was kept at 1 °. The films were also studied by conversion electron M6ssbauer spectroscopy (CEMS) employing a constant acceleration M6ssbauer spectrometer with Co57:Rh as the source. The hyperfine interaction parameters were obtained using the standard MOSFIT (Kreber 1976) code.

When laser deposition is performed at an oxygen partial presssure of 10 - 6 Torr onto single crystal A1203 substrate held at a temperature of 200°C the resulting film has an X-ray structure corresponding to the XRD spectrum shown in figure 9a.

From the lattice plane signatures indicated on the XRD pattern it is clear that the

148 S B Ogale

®

o" ~ o" o ~'

¢ . . rl

N N N L 1

- o ~ o ~ ~ ~- o

4O-00 6O-OO 801]0 go~o

2Q (Degrees)

Figure 9. Small angle XRD patterns for ZnxFe3_xO ~ films deposited on single crystal AI203 substrates at substrate temperature of 200C and oxygen partial pressure of (a) 10 -6 Torr and (b) 5x 10-4 Torr.

deposited film is composed of Fel_xO phase as well as ZnxFe3_xO4, the contribution of Fe~ _~O phase being significant. The presence of Fe~-xO phase in the film indicates that under the given conditions of deposition there is a degree of phase decomposition in the material transfer from the pellet to the film. When laser deposition is performed at the same substrate temperature of 200°C but at a higher oxygen partial pressure of 5 × 10-4Torr there are significant changes in f i l m properties. This very fact should reflect the nature of the pulsed vaporization and subsequent deposition process. The small angle'XRD pattern for this case is shown in figure 9b. The major X-ray signatures together represent a typical pattern for ZnxFe3_~O 4 phase, though the line intensity ratios do not correspond to the bulk case, presumably owing to preferential orientation effects in the film structure. Also, the lines at 20 values of 36'88 °, 40'52 °, 62'0 ° and 65.92 ° indicate presence of ZnO 2 phase in the sample.

We shall now present results for the case of deposition at a higher substrate temperature of 450°C. The XRD pattern of the sample deposited at T~ub of 450~'C and oxygen partial pressure of 10 -6 Torr is shown in figure 10a. It is interesting to compare this pattern with that rn figure 9a which is of a film deposited at a T~ub of 200~C. The entire spectrum in figure 10a is composed of Contributions from ZnxFe3_~O , phase with modified line intensity ratios due to strain-induced texture effects and from ZnO phase. It is thus clear that changing substrate temperature can lead to a change in the nature of the phase formation process. This could be partly due to the nature of stability of intermediate phase(s) under the condition of temperature-dependent absorption~lesorption kinetics at the substrate surface.

When deposition is carried out at a substrate temperature of 450°C and oxygen partial pressure of 5 x 10 -4 T o m the sample gives the small angle XRD pattern shown in figure 10b. The XRD pattern in this case clearly corresponds to that of ZnxFe 3_~O~ phase closely resembling the case of the starting pellet itself except for the fact that in the film a high degree of texturing is present leading to the

Laser-induced synthesis, deposition and etchiny

~?f

~2

I -

,q

U') Z 14J I -

, 6 " ~ ff ¢

~v~- ~ ~ v ' ~

. ~ g . . " ." go.

: e

40.00 60.00

®

20.00 80.00 90.00

2 O (Degrees)

Figure 10. Small angle X R D patterns for Zn,F% .,04 films deposited on single crystal AI20 3 substrates at substrate temperature of 450 C and oxygen partial pressure of (a} 10 ~"

Torr and [bt 5 × 10 ~ Torr.

appearance of only the two most dominant planes with d values of 2.94,~ and 2.51 ,~. In order to examine the local atomic order in the film deposited at substrate temperature of 45OC and oxygen partial pressure of 5 × l0 -4 Torr, it was subjected to M6ssbauer spectroscopic examination using the CEMS technique. The CEMS spectrum is given in figure 11. This spectrum can be fitted by deriving guidelines from the work of Dobson et al (1970) and Srivastava et al (1976a,b). Thus our spectrum can be fitted with the following parameters:

(i) sextet: IS=0.38 mm/s, QS=0.002 ram/s, IMF=360.0 kOe, (ii) sextet: IS=0.37 ram/s, QS=0.06 mm/s, IMF=423.0 kOe, (iii) doublet: IS = 0.45 ram/s, QS = [.46 ram/s,

where IS, QS and IMF represent isomer shift, quadrupole splitting and internal magnetic field respectively. These parameters correspond to ZnxFe3_xO4 phase with x close to 0-6. The quality of X-ray and M6ssbauer data for the film deposited at 4 5 0 C and at oxygen partial pressure of 5× 10-4Torr clearly indicate an essentially single phase stoichiometric character of the film. It is important to mention that the conversion of Fel xO phase into Fe304 phase is facilitated at the higher temperature of 45OC and hence the growth of ferritic single phase is thermodynamically favoured. All the data reported so far thus show that by appropriate control of ambient conditions and substrate temperature it is possible to obtain good quality films of complex material systems such as zinc ferrite by using the pulsed laser induced vaporization process.

Having presented some results for complex oxide based systems, I shall now discuss the case of a complex metallic alloy, viz. stainless steel, which can be obtained in a thin film tbrm by laser deposition.

3.4 Stainless ,steel thin Jilm

In this experiment a pulsed ruby laser (2 = 694 nm, pulse width 20 ns) was focused

150 S B Ogale

I i

I I I I I I

r = i i i i i

JD

> - I - -

z

L d I - -

z

• ...- y

, t I I I [ I I I I !

-10 - 8 - 6 - 4 -2 0 2 4 6 8 10

VELOCITY ( m m / s e c )

Figure 11. C E M spectrum for a Zn:,Fe3_xO 4 film deposited on single crystal A1203 substrate at substrate temperature of 450°C and oxygen partial pressure of 5 x 10 - 4 Torr.

onto a solid 304 stainless steel target in a stainless steel vacuum chamber pumped to ultimate pressure of 2 × 10 -v Torr. Substrates of glass, iron and vitreous carbon were mounted on an electrical heater assembly which was placed at a distance of about 4 cm in front of the target. More details of the experimental arrangement are described elsewhere (Koinkar et al, in press). During laser evaporation, the substrates were kept at constant temperature of 300°C to enhance the adhesion of the deposited thin film to the substrate surface. The laser irradiation was carried out at an energy density of 10-12 J/cm 2 on the target. The films thus produced were characterized by employing CEMS and RBS. The film deposited on carbon substrate was used for RBS measurements to avoid the complications of mass and depth analysis which was carried out by using RUMP programme. The CEM spectrum of as-received 304 stainless steel is shown in figure 12a. It is a single line spectrum with IS value of -0.14 mm/sec (with respect to -Fe). Figure 12b gives the CEM spectrum of a thin-film sample on glass substrate, obtained by pulsed laser evaporation of bulk 304 stainless steel. There is a significant difference between the two spectra. The M6ssbauer spectrum of deposited thin film could be fitted with a broad doublet with IS value of 0.14 mm/sec and QS value of 0.562 mm/sec. The broadness of the resonance lines essentially reflects variation in the local environments in the thin film. It is interesting to note that there is considerable overlap between the RBS spectrum of deposited thin film on carbon substrate and that of source material (figure 13), indicating thereby the preservation of overall stoichiometry in the thin film. Thus, even though the overall stoichiometry is preserved in the pulsed vaporization process, the microstructure of deposited thin film is different from that of original material. The FCC structure of 7-iron phase contains a large number of octahedral interstitial voids which are occupied by

Laser-induced synthesis, deposition and etchin9 o/ materials 151

ILl Z

r l

. ..j.,

a I , I ~ I

- I 0 - 8 - 6 -/. - 2

I ^{i • I I I}

0 2 ~ 6 8 10

VELOCITY ( m m / s e c )

Figure 12. Room-temperature CEM spectra of (a) as-received 304 stainless steel and (b) film deposited from 304 stainless steel on glass substrate by pulsed ruby laser evaporation.

interstitial elements such as carbon or nitrogen. Every interstitial is surrounded by six atoms and, in certain austenitic steels such as 316SS, one or two nearest- neighbour (nn) positions are occupied by substitutional elements such as Cr, Ni, etc.

Various combinations of such atomic configurations in austenite phase perturb the local symmetry of the iron atom and give rise to quadrupole splitting distribution.

The present M6ssbauer data for the deposited thin film do show such a distribution, which elucidates the effects of substitutional as well as interstitial atoms upon the microstructure of 7-iron phase existing in stainless steel.

152 S B 0 9 a l e

P ,

20

> - Q

ENERGY (MeV)

30 1.0 1.2 , , 1.4 1.6 1.8

[ I I i ]

0! ~ ~ J ' % ' - - •

150 200 250 300 350

CHANNEL

Z

Hgure 13. RBS spectra of as-received 304 stainless steel ( ) and film deposited from 304 stainless steel on vitreous carbon substrate by pulsed ruby laser evaporation ( -).

Resistance to thermal oxidation is one of the important properties of austenitic stainless steel. It has been observed that the formation of a thin protective oxide layer on stainless steel inhibits further oxidation of the substrate material. It is interesting to explore the oxidation processes associated with laser-evaporated thin films of stainless steel. Since iron foil is known to be prone to oxidation, it was chosen as substrate material for the deposition of 304 stainless steel film, which should act as protective coating. This sample and an iron foil were annealed in air at 400 C for 5 min to observe the oxidation process. The C E M spectrum of oxidized Fe foil shows (see figure 14b) formation of higher iron oxides such as F e 3 0 4 and ~-Fe20 3. However, the CEM spectrum of Fe foil with a deposition of stainless steel film shows a single resonance line along with the underlying ~-Fe but no iron oxide (see figure 14a). The thermal process seems to be responsible for redistribution of interstitial and substitutional atoms in stainless steel film, giving rise to a slightly broad single line instead of a doublet. The oxidation of iron foil is clearly prevented by the deposited stainless steel film. In conclusion, the pulsed laser evaporation technique can be used to deposit thin films of complex metal alloys as well.

4. Laser etching

In the previous sections I presented some new results on laser-induced synthesis and deposition. Etching is another area of microelectronics wherein lasers are finding increasing applications (Osgood et al 1982; Reksten et al 1986; Datta et al 1987; Rothschild et al 1987a,b). In this context pulsed as well as CW lasers have been employed, either in the presence of reactive ambients in the vapour or liquid phases or in the absence of such environments, purely for ablation purposes. In some cases lasers have been used to impart specific modifications to selected areas

Laser-induced synthesis, deposition and etchin,ct c?['materials i53

U z t

~ ' ° ° ~" .

Z

W ~ I I I

I

I

-10 - 8 - 6

I I I / I

- 4 - 2 0 2 4

VELOCITY ( mm I s e c )

. . . . ° - . °

s t ~

, , i F ~ o ~

' ~ ' F e

6 8 10

Figure 14. Room-temperature CEM spectra of (a) film deposited from 304 stainless steel on iron substralc by pulsed ruby laser evaporation and (b) iron foil, both air-annealed at 400°C for 5 rain.

and such areas have been subsequently etched out in reactive media. The interested reader is referred to the papers mentioned above. In the following ! present some results on an interesting etching application recently reported by Inam et al (1987) once again in the context of the hot oxide superconductor films. I have chosen this example to emphasize the dry-processing aspect of pulsed laser etching.

4.1 Pulsed laser etching of high T~ superconducting films

Laser-induced etching for dry processing of materials has recently been a rapidly growing field of investigation. Inam et al (1987) have recently applied this method to Y B a - C u O thin films. In this experiment a KrF excimer laser (Lambda Physik EMG200E) producing 0"8 J pulses of 30 ns duration at a wavelength of 248 nm was used. The uniform part of the laser beam was focused through a 0-5 mm × 1 mm aperture in a thin metal sheet kept about 0.5 mm away from the film surface and was incident perpendicular to the sample surface. A wide range of incident energy fluences was explored. Incident beam energy density was varied from 0"1 to

154 S B Ogale

3"8 J/cm 2 with the use of quartz attenuators. The samples used were 1.O-1-6/~m thick annealed films of Y - B a - C u - O on sapphire substrates deposited by laser eva- poration. The experiment was performed in air and the substrate was held at room temperature. Measurements of etch profiles were performed using a mechanical stylus (Alfa Step). The accuracy of this measurement was better than 30 nm. Auger electron spectroscopy (AES) was also used for characterizing the film composition in the etched and unetched regions of the film.

First, etch depth dependence on the number of laser pulses was studied as a function of incident fluence. A profile of a typical etch is shown in figure 15. The original high T~ film surface roughness is about 20-40 nm. The bottom of the etch is not flat but has a slightly convex shape. This is probably due to a combination of two effects: (i) As the material is ablated, the rising material couples with the incident laser beam and attenuates the central portion of the beam. Therefore, the beam intensity is lower in the central region of the etch, effectively removing less material there. This effect is commonly observed in the laser drilling of materials. (ii) Some of the etched material redeposits into the ablated area, thereby contributing to the formation of the observed hump at the centre of the etch. Auger data revealed that the film composition is essentially the same at the etched and unetched sites on the film. This encouraging result indicates that the etching process is a layer-by-layer removal of the superconducting material that does not affect the stoichiometry of the underlying area (within Auger resolution). For each value of the incident fluence, ten sites were etched with the number of laser pulses varying from 1 to 10 at each site. For all incident energy densities, the etch depth was found to be linear with the number of pulses, as can be seen from figure 16.

The slopes of the plots in figure 16 give the etch rates at various incident energy densities. Figure 17 shows the results. The threshold value of incident energy density (Eth) below which no etching occurs can be ascertained from the plot by extrapolation and is found to be 0.11 J / c m 2. Furthermore, we can identify at least two distinct regions in the plot. For incident energy densities above the threshold and below 2 J/cm z, the etch rate rises rapidly as a function of incident energy density. If a well-stabilized laser is utilized for etching, this is the desirable region where etching can be performed efficiently. If, however, the laser output is not easily definable or stable, then etching in the region where the incident energy is higher than 2 J / c m 2 is clearly more suitable. Thus pulsed excimer lasers can be used very effectively to dry-etch oxide superconductors.

-r"

I-- o. 0 bJ C ) T -/I t,J h - I.u - 8

-12 ^{I I !}

5013 I000 151313

LATERAL DISTANCE (jurn)

Figure 15. Surface profile of site on oxide superconductor film sample etched with nine laser pulses at laser energy density of 0.68 J/cm 2.

Laser-induced synthesis, deposition and etchin 9 of materials 155 16

A <I

v

- r I-- Q.

LIJ C}

-I"

o 4 l-- ILl

0

Energy density (]'/cm 2} , / ~ 4 -'--'-3.8

-o-o- 0.7

0.3 j ~ r

-o-o-0.2 _ . , 7

2 4 6 8 10

NUMBER OF LASER PULSES

Figure 16. Dependence of etch depth on the number of laser pulses at four different laser energy densfiies.

1-5

1-C

Ixl

~

0. ~.

0~ 0

...o_.---~

1.0 2-0 3!0

LASER ENERGY DENSITY ( ] / c m 2) /..0

Figure 17. Variation of etch rate as a function of incident laser ene:gy density. The etching threshold energy for oxide superconductor film has a value of 0.11 J/cm 2 as inferred from the graph.

5. Conclusions

Lasers are extremely powerful tools for material synthesis, modification, deposition and etching applications. The laser beams can be transported in a variety of reactive or non-reactive media with considerable ease and therefore lend themselves to imaginative use to achieve specific objectives such as localized synthesis or removal of compounds or deposition of complex atomic systems such as oxide superconductors or ferrites.

Acknowledgements

The author is grateful to the members of the Submicron Materials Group of the Department of Physics, University of Poona, for co-operation in the work reported here as the Poona University contribution. Thanks are also due to Prof. Franz Saris of the FOM Institute, The Netherlands, and Dr T Venkatesan of Bellcore, USA, for fruitful discussions during the course of the author's stay in their laboratories. The author is grateful to Mr P P Patil and Rashmi Nawathey for help in preparing the manuscript.

References

Allen S D 1981 J. Appl. Phys. 52 6501 and references therein

Allen S D, Jan R Y, Edwards R H, Mazuk S M and Vernon S D 1984 Laser Assisted Deposition, Etching and Doping, SPIE 459 42

156 S B O y a l e

Allen S D and Tringubo A B 1983 J. Appl. Phys. 54 1641

Allen S D, Tringubo A B and Teisinger M L 1982 J. Vac. Sci. Technol. 20 469

Andreatta R W, Abele C C, Osmundsen J F, Eden J G, Lubben D and Greene J E 1982 Appl. Phys. Lett.

40 183

Auston D H, Golovchenko J A, Simons A L, Slusher R E, Smith P R, Surko C M and Venkatesan T N C 1979 in Laser-solid interaction and laser processing--1978, (eds) S D Ferris, H J Leamy and J M Poate, AIP Conf. Proc. No. 50, page 11

Aydinli A, Lo H W, Lee M C and Compaan A 1981 Phys. Rev. Lett. 46 1640 Bhat D G and Holzl R A 1982 Thin Solid Films 95 105

Black G, Ehrlich J, Rothschild M, Doran S P and Sedlacek J H C 1987 J. Vac. Sci. Technol. B5 419 Carter G and Colligon J S 1968 Ion bombardment of solids (UK: Grasham Press)

Celler G K, Poate J M and Kimerling L C 1978 Appl. Phys. Lett. 32 464

Datta M, Romankiw L T, Vigliotti D R and von Gutfeld R J 1987 Appl. Phys. Lett. 51 2040 Demiryont H, Thompson L R and Collins G J 1986 J. Appl. Phys. 59 3235

Deutsch T F, Ehrlich D J and Osgood R M Jr 1979 Appl. Phys. Lett. 35 175 Dijkkamp D, Wu X D, Chan S W and Venkatesan T 1987a J. Appl. Phys. 62 293

Dijkkamp D, Venkatesan T, Wu X D, Shaheen S A, Jisrawi N, Min-Lee Y H, McLean W L and Craft M 1987b Appl. Phys. Lett. 51 619

Dijkkamp D, Gozdz A S, Venkatesan T and Wu X D 1987c Phys. Rev. Lett. 58 2142 Dobson D C, Linnett J W and Rehman M M 1970 J. Phys. Chem. Solids 31 2727

Donnelly V M, Herman I P and Hirose M 1987 (eds) Photon, beam and plasma stimulated chemical processes at surfaces, See also other proceedings in the same series.

Ehrlich D J, Osgood R M, Jr and Deutsch T F 1981 Appl. Phys. Lett. 39 957

Emery K, Boyer P K, Thompson R, Solanki R, Zarnani H and Collins G J 1984 Laser Assisted Deposition Etching and Doping, SPIE 459 9

Ferris S D, Leamy H J and Poate J M 1979 (eds) Laser~olid interaction and laser processing--t978, AIP Conf. Proc. No. 50

Foti G, Campisano S U, Rimini E and Vitali G 1978 J. Appl. Phys. 49 2569

Galvin G L, Thompson M O, Mayer J W, Hammond R B and Percy P S 1982 Phys. Rev. Lett. 48 33 Grossman V M and Karnezos M 1987 J. Vac. Sci. Technol. B5 843

Hansan S G and Robitaille T E 1988 Appl. Phys. Lett. 52 81

Hoonhout D 1981 Pulsed laser annealing of ion implanted silicon Ph.D. Thesis, FOM Institute, The Netherlands

Inam A, Wu X D, Venkatesan T, Ogale S B, Chang C C and Dijkkamp D 1987 Appl. Phys. Lett. 51 1112 Joshi S, Nawathey R, Koinkar V N, Godbole V P, Chaudhari S M and Ogale S B 1988 J. Appl. Phys.

(In press)

Koinkar V N, Chaudhari S M, Kanetkar S M and Ogale S B Thin Solid Films (in press)

Kreber E 1976 Universite des Saarlandes, Saarbruken; Mosfit code for fitting of Mossbauer spectra. This was adopted for an ICL 1904S computer by S K Date from the National Chemical Laboratory, Pune, India

Lo H W and Compaan A 1980 Phys. Rev. Lett. 44 1605

Nawathey R, Vispute R D, Chaudhari S M, Kanetkar S M, Mitra A, Date S K and Ogale S B J. Appl.

Phys. (in press)

Ogale S B, Patil P P, Phase D M, Bhandarkar Y V, Kulkarni S K, Kulkarni S, Ghaisas S V and Kanetkar S M 1987a Phys. Rev. B36 8237

Ogale S B, Patil P P, Phase D M, Kulkarni S A, Ghaisas S V and Bhide V G 1986 Materials Research Society symposium on beam-solid interactions and transient processes, Boston (unpublished)

Ogale S B, Polman A, Quentin F O P, Roorda S and Saris F W 1987b Appl. Phys. Lett. 50 138 Osgood R M Jr, Sanchez-Rubio A, Ehrlich D J and Daneu V 1982 Appl. Phys. Lett. 40 391

Patil P P, Phase D M, Kulkarni S A, Ghaisas S V, Kulkarni S K, Kanetkar S M and Ogale S B 1987 Phys. Rev. Lett. 58 238

Picraux S T, Thomsen M O and Williams J S 1987 (eds) Beam~olid interactions and transient processes, MRS Symp. Proc. See also other proceedings in the same series

Poate J M, Foti G and Jacobson D C 1983 (eds) Surface modification and alloying by laser, ion and electron beams (New York, London: Plenum Press)

Polman A, Roorda S, Ogale S B and Saris F W 1987 in Beam-solid interactions and transient processes (eds) S T Picraux, M O Thomsen and J S Williams (to appear)

Reksten M, Holber W and Osgood R M Jr I986 AppI. Phys. Lett. 48 551

L a s e r - i n d u c e d synthesis, deposition and etchin9 o f materials 157

Rothschild M, Sedlacek J H C, Black J G and Ehrlich D J 1987a J. Vac. Sci. Technol. B5 414 Rothschild M, Sedlacek J H C and Ehrlich D J 1987b J. Vac. Sci. Technol. B5 1400

Srivastava C M, Shringi S N and Srivastava R G 1976a Phys. Rev. B14 2041

Srivastava C M, Shringi S N, Srivastava R G and Nanadikar N G 1976b Phys. Rev. BI4 2032 Srivastava P K, Rag T V, Vankar V D and Chopra K L 1984 J. Vac. Sci. Technol. A2 1261 Stritzker B, Pospieszczyk A and Tagle J A 1981 Phys. Rev. Lett. 47 356

Sugimura A, Fukuda Y and Hanabusa M 1987 J. Appl. Phys. 62 3222 Tu K N and Berry B S 1972 J. Appl. Phys. 43 3283

Ullirich B M 1985 (ed.) Ion beam modification of materials; Nucl. Instr. Meth. See also other proceedings in the same series

Venkatesan T (private communication)

Wood R F, White C W and Young R T 1984 (eds) Pulsed laser processing of semiconductors (New York:

Academic Press)

Wu X D, Dijkkamp D, Ogale S B, Inam A, Chase E W, Miceli P F, Chang C C, Tarascon J M and Venkatesan T 1987 Appl. Phys. Lett. 51 861