Cu multilayers

Cu multilayers

Journal of the Less-Common 451 Metals, 144 & 165 (1990) 451-457 ION BEAM MIXING OF LA(OH$/CU and LA(OHI~/SRO/C~ MULTILAVERS J.P. MATHE"ET*, A. TR...

360KB Sizes 0 Downloads 30 Views

Journal of the Less-Common

451

Metals, 144 & 165 (1990) 451-457

ION BEAM MIXING OF LA(OH$/CU

and LA(OHI~/SRO/C~ MULTILAVERS

J.P. MATHE"ET*, A. TRAVERSE*, S. MEGTERT+ and J. CHAUMONT* * Centre de Spectrometrie Nucleaire et de Spectrometrie de Masse, INZP3-CNRS, F-91405 Campus Orsay, France + Laboratoire de Physique des Solides, Bat. 510, Universlte Paris XI, F-91405 Orsay Cedex, France

We have submitted bilayers of La(OHls/Cu and multilayers

of La(OH)s/SrO/Cu

to energetic heavy ion irradiation at 300 the possibility of preparing investigate

K and 700 K in order to LaaCuOI and (La1_xSrx)2Cu0 4 through the ion beam mixing technique. For comparison purposes, similar annealed at 700 K. Results are given concerning the samples were modification of the atomic compositions (La, Sr. Cu. 0 and Hl, the atomic depth distributions, the compound formation, induced either by irradiation or by thermal annealing. The specific role of the ion beam mixing process is emphasized.

1. INTRODUCTION The

goal

possibility step,

the

of

experiments

of preparing

using

the

thermodynamical

ion

described

beam

on metallic

mixing

multilayersl,

induced by the irradiation beam

technique.

If

the

target

diffusion

to

investigate

the

This

technique

is

far

from

layers of each component. From results

we know

that

through

elastic

layer interpenetration.

if the heat of mixing of the layer components

temperature

collisions

(scale time 10-13s1, atoms may jump from one

layer to the other, leading to a progressive mixing is enhanced

is

; it consists in sending an energetic (keV to MeVl

equilibrium

ion beam through a stack of alternating obtained

here

thin films of LazCuOl and (La1_xSrx12CuOq in a single

during

irradiation

is

high

Rate of

is negative.

enough

for

atomic

to take place, within lo-" s (a diffusion enhanced by defects due to

irradiation), ordered final systems may build up. The

idea

here

is

to use

11

the amount

of

energy

deposited

in atomic

collisions to induce mixing between La/Sr/Cu layers and ill the possibility of enhanced diffusion usual

ones,

i.e.

(La1_xSrx12CuOl has multilayer

to prepare the final system at a temperature about been

9OO'C'. made

by

Another

two

implanting

steps Sr

in

lower than the

attempt an

to

evaporated

prepare La/Cu

followed by annealing in oxygen at a temperature > 800 K to obtain

superconducting

samples'.

First experiments3 on La/Cu bllayers have already shown that an evaporated La layer even covered with a 40 nm Cu layer evolves

0022-5088/90/$3.50

towards La(OH13 during

0 Elsevier Sequoia, Printed in The Netherlands

452

exposure

to ambiant

Cu instead the

of

initial In

system

this

compared depth

atmosphere.

induces

paper,

to

we

those

under

partial

starting

of

components

shown3 that

results

obtained

on

on La(OH13/Cu bilayers,

La,

Cu, Sr,

Au irradiation

are

La(OHls and

irradiation

at 300 K of

H desorption.

present

obtained

distributions

formed

Hence the

LaaOJ and Cu02. We have also

La(OH13/Sr/Cu

with

0 and H and the

respect

nature

at 300 K and 700 K or under

trilayers

to

of

the atomic

the

thermal

compounds

annealing

at

700 K. A spontaneous samples,

the

boundary

is

initially

mixed

interpenetration

scale thicker

the substrate

irradiation,

thermal

annealing.

identification

the

does

layers

not

evidenced

grains,

being

presence

of

the

layer.

not

increase

there

is

certain, seem

the substrate

is

or

very

only

New compounds,

compounds

with

the

atomic

Hydrogen

partial

possibly are

to

Sr

The

at

the

The

mixed of

irradiation

desorption

some

as-deposited

depth

totally

the

whatever

desorbs

at 700

at 300 K or under

La2CuOI and

formed

display

in

unknown.

much under

or under annealing.

while

being

These

relation

in

layer

of

which,

temperature

K under

ones.

of

(La1 xSrx12CuOa,

expense

of

preferential

the

the

initial

orientation

in

structure.

2. EXPERIMENTAL Bilayers thick

of

La/Cu and trilayers

and Sr layers

by electron

gun evaporation,

0.15

As

rim/s..

of

20 nm thick)

already

induces

an evolution

despite

the 40 nm Cu coverage,

(La and Cu layers

on

bilayer3,

about

lo-’

exposure

La to La(OH13. We found although

were

at 300 K on carbon

under a vacuum of

shown

of

La/Sr/Cu

were deposited

that

a definite

Pa at a rate

to

ambiant

Sr also

40 I-MI

substrates around

atmosphere

trapped

identification

oxygen,

between

SrO

or Sr(OH12 is not possible. Irradiations’ f luence before

of

and after

He particles’ atoms, “N

were

8~10’~

ii)

irradiation

to measure using

the

on H to measure

diffraction identify

performed

with

energetic

Au ions

: i) by Rutherford

the amount and depth

resonant

(XRBl patterns compounds present

The thicknesses

nuclear

were

recorded

on a 8-28

evaporated

here

are

not

those

Although we use

by RBS to detect

6.385 in

the

powder

a

2 MeV

Cu Sr and 0 MeV incident

target.

X ray

diffractometer

to

in the stack.

in RBS experiments. possible

(FlNRl of

La,

to

in situ

(RBSl of of

distribution

; however they were chosen

substrates,

backscattering

reaction’

the H amount and depth

MeV) up

were analyzed

distributions

La2CuOl composition

on carbon

(3.65

ions. cms2, at 300 K and 700 K. Samples

not

high

required to

Tc superconducting thanks

only

components

to

their of

reach

well

films

them because, the heavy

to

provide

are

the

defined not

deposited

low mass, the

target

final peaks

it

is

such as

453

La, Sr, Cu but also the light one such as 0. A precise characterization samples

is a prerequisite

for the understanding

of the processes

of our

involved

in

trilayer before irradiation

is

ion beam mixing.

3. RESULTS A typical RBS spectrum recorded on La/Sr/Cu shown in Fig.

1 (dotted line). By comparison

between

experimental

in the target. The numbers of La, Sr and Cu correspond

to those expected from

within 5%. while the number of 0 is approximately

the evaporation

peak area

it is possible to deduce the amount of each component

and those of standards,

equal to 3

times the number of La plus the number of Sr. The RNR spectrum detected on the same trilayer, plotted in Fig. 2 (square before irradiation),

indicates that a

layer, about 30 nm thick just below the surface remains free of hydrogen. Part of the Cu layer and rest of the sample contain an H amount, H/La a 3. These two results are in favour of an initial stack formed of La(OHls/SrO/Cu.

Au

I

FIGURE 2

FIGURE 1 Number of MeV "N on (crosses) versus the

RBS spectrum of a trilayer before (dotted line) and after (full line) Au irradiation at 700 K.

Part of an experimental showing Fig.

La and Cu peaks

3.

A

good

interpenetration the bottom,

fit

RBS spectrum detected

is compared requires

the

layer. Approximately boundary.

about

30%

of

on a La(OHls/Cu bilayer

to a simulation assumption

with

of

a

and

the RUMP' code in

spontaneous

partial

on the initial layers. The stack is made of a pure La(OHls at

then a layer containing

mixed

7 due to the RNR of 6.385 H before (square) and after Au irradiation at 700 K N energy.

When the

Cu, La, 0 and H atoms,

13% of the La and Cu atoms a Sr

total

layer

number

is present, of

heavy

are

the

(La, Sr

then a pure Cu

in this

spontaneously

interpenetration and

Cul

atoms.

concerns Through

454

RBS, it is not possible to determine if it is at the atomic or at the grain scale that this interpenetrationtakes place.

300

320

310

Channel FIGURE 3 Simulation (dotted line) and experimentalLa and Cu peaks (full line) of a RBS spectrum recorded on a bilayer. An XRD spectrum,Fig. 4 a, recordedbefore irradiation,displays the (1001, (2001, (3001 lines of the La(OH)s phase. The wide peaks are the (0021, (IOO), (0041, (110) lines of the badly crystallizedcarbon substrate.The absence of other

La(OHls peaks

indicates a

preferential orientation along

the

crystallographica axis. Signals coming from SrO or Cu are too weak to be detected, due to the thin layer thicknesses and low linear absorption coefficientwith respect to the incidentwavelength (Cu Kcrradiation).

b

FIGURE 4. X Ray spectrum recordedbefore (al and after (bl Au irradiationat 300 K. The evolution of a trilayer under Au iradiationat 700 K is shown Fig. 1 (full line). The Au peak correspondsto the irradiatingAu ions stopped at a

455

depth

of

900

nm

redistribution displacement

below

the

the

of

substrate

species

surface.

(widening

and

Although

there

lowering

of

of the La peak). it is obvious that an homogeneous

reached. Such an homogeneous the experimental

distribution

is

the

a

depth

Sr

peak,

mixing is not

is simulated Fig. 5 and compared to

spectrum. Increasing the irradiation fluence does not lead to

more in depth homogeneity. Results obtained after irradiation at 300 K are

FIGURE 5 Comparison between the experimental RBS spectrum taken after Au irradiation of (dotted line) of a a trilayer at 700 K (full line) and the simulation homogeneously mixed system containing the same amount of La, Cu, Sr and 0 atoms. similar.

The amount

incident

Au beam,

remains

of Cu is slightly

since

identical within

The amount

the Cu

layer

the experimental

of 0 is lowered by about

about 50% after

decreased,

irradiation

due

to sputtering

is the overlayer uncertainties,

10% after

at 700 K. Hydrogen

by

the

; the amount of La La being

irradiation

below Cu.

at 300 K and by

totally diffuses

out of the

sample when it is irradiated at 700 K as shown on Fig. 2 (crosses), by going either outside or in the carbon substrate. Annealing at 700 K during the same time as irradiation previously substrate

noted,

temperature,

the La(OH1s lines LazCuOl or allow

leads to a partial H desorption only

a

partial

is

only

(75%). At 300 K, as

measured.

XRD spectra taken after irradiation

Whatever

phase

The

identification

lack of

supplementary

(other systems

such

peaks

as LaCue

cannot be excluded). However, this is in favour of a preferential growth.

the

no longer display

; new peaks are present whose 26 values may correspond to

(LaX_XSrXlaCuOo peaks.

a unique

desorption

does

not

or LazOs

orientation

456

4. DISCUSSION AND CONCLUSION To build up LazCuOl or (La,_xSrxlzCuOl thin films, we have chosen an unusual starting high

technique,

materials,

Tc

the ion beam

of multilayers

La(OHla, SrO and Cu. The carbon

superconducting

atomic

irradiation

films

characterization.

epitaxial

The

growth,

as-deposited

sustrate,

was

used

interpenetration

of the layers, which has not been explained

interpenetration

being

mixing

increase

does

whatever

favoured

the temperature,

collisions density,

not

very

the Sr under

although

is on the average

metallic

by much

This

a

precise

up to now, the

initial

ion beam

spontaneous irradiation, in atomic

In such a range of deposited

are usually

mixed

over

for

spontaneous

the amount of energy deposited

300 eV/i.

multilayers

layer.

energetic

unusual

not good

to allow

display

stack

to check

and

several

energy

nm depth.

The

absence of mixing in these experiments could be due to the fact that we try to mix

oxide-metal

layers

instead

of

mixing

metallic

However

layers.

leave the sample (partially or totally depending on the substrate under irradiationl. it increases

are

the H

locally

compounds.

temperature

Irradiation is more efficient than thermal annealing since loss. As a result

atomic bonds between atoms

the

: hydrogen and oxygen

composition of the spontaneously mixed layer is modified

of

irradiation

La and 0 or H and between

present,

The saturation

new

atomic

bonds

or thermal

treatment,

Sr and 0 are broken. may

form

to

build

of ion beam mixing may be understood

As Cu

up

other

in the sense

that once the spontaneously mixed layer has evolved under irradiation

towards

a new compound, the latter is stable against further ion beam treatment. It is

interesting

to note

systems formed under

that either

the La(OHls

irradiation display a preferential

on the carbon substrate

compound

or

the new

orientational

as indicated by the absence of conventional

intensity peaks seen in a powder X ray diffraction have recently been observed"

pattern.

growth highest

Similar

effects

in implanted systems.

What can we expect concerning the formation of LazCuOl and (Lal_xSrx12Cu04, after

these results

multilayers

? The

concerning starting

instead of 2 required) temperature possibly

higher

irradiation

system

of La(OHls/Cu

is already

rich

and La(OHls/SrO/Cu

in oxygen

(3 0

per

La

: no further oxygen charging may be needed. Instead of

than 800 K as in Ref. 4, a temperature

lower, no intermediate

of only 700 K (or

temperature having been checked)

is necessary

for a total H loss accompanied with formation of new compounds. This indicates the specific role of energy deposition under ion beam treatment. The structure of the substrate

seems important

in building

was observed even on the badly crystallized

up oriented

phases,

since this

carbon substrate. This encourages

us to use substrates such as MgO or SrTiOs since they are known to favour good epitaxial

growth9

of

the

superconducting

phase.

Moreover,

as

they

are

insulating materials, they should be helpful for measuring the conducting properties of

our

samples, a

supplementary and

necessary too1

for

characterizationof our final systems.

ACKNOWLEDGEMENT The SEMIRAMIS group of the Centre de Spectrometrie Nucleaire et de Spectrometriede Masse is gratefullyacknowledgedfor its constant and helpful assistanceduring the irradiationand characterizationexperiments.

REFERENCES 11 H. Wollenberger,Nucl. Instr. and Meth. B 48 (19901 493 2) see for example : P.M. Grant, S.S.P. Parkin, V.Y. Lee, E.M. Engler, M.L. Ramirez, J.F. Vazquez, G. Lim, R.D. Jacowitz and R.L. Greene, Phys. Rev. Lett. 58 (19871 2482 31 J.P. Mathevet, A. Traverse, J. Chaumont, M. Gasgnier and S. Megtert, in print in Jou. Mat. Research 41 A.E. White, K.T. Short, J.P. Garno, J.M. Valles, R.C. Dynes, L.F. Scheenemeyer,J. Waszczak, A.F.J. Levi, M. Anzlowar and K.W. Balwin, Nucl. Instr. and Meth. B37/38 (1989) 923 51 E. Cottereau,J. Camplan, J. Chaumont and R. Meunier. Materials Science and EngineeringB2 (19891 217 6) U.K. Chu, J.W. Mayer and M.A. Nicolet, Backscatterlng Spectrometry, Academic Press (19781 71 B. Maurel and G. Amsel, Nucl. Instr. and Meth. 218 (19831 159 81 L.R. Doolittle,Nucl. Instr. and Meth. B9 (19851 344 91 0. Meyer, F. Weschenfelder,J. Geerk, H.C. Li and G.C. Xiong, Phys. Rev. B 37 (19881 9757 10lB. Rauschenbachand K. Helming, Nucl. Instr. and Meth. B42 (19891 216