Low temperature anomaly of the electrical resistivity of SnMn

HAL Id: jpa-00209665

https://hal.archives-ouvertes.fr/jpa-00209665

Submitted on 1 Jan 1983

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Low temperature anomaly of the electrical resistivity of SnMn

D. Korn, K. Maurer

To cite this version:

D. Korn, K. Maurer. Low temperature anomaly of the electrical resistivity of SnMn. Journal de Physique, 1983, 44 (7), pp.827-832. �10.1051/jphys:01983004407082700�. �jpa-00209665�

827

Low temperature anomaly ^{of the} electrical resistivity ^{of SnMn}

D. Korn and K. Maurer

Fakultät für Physik ^der Universität, ^{D 7750} Konstanz, ^Postfach 5560, ^F.R.G.

(Reçu ^{le 9 dgcembre} 1982, accepté ^{le 21 mars} 1983)

Résumé. ²⁰¹⁴ Mn étant insoluble dans Sn, les deux métaux sont condensés simultanément à partir ^{de la} phase vapeur

sur un substrat à 5 K. La résistivité électrique ^{de ces films a un} coefficient de température négatif (anomalie ^de résistivité) ^comme l’effet Kondo de SnMn mentionné _par Buckel et al.

L’anomalie de résistivité croit avec la concentration de Mn indiquant ^{un moment} magnétique ^{localisé. Un}

maximum de résistivité apparaît ^au voisinage ^{de 1} % ^Mn.

La température du maximum monte avec la concentration de Mn. Les résultats expérimentaux indiquent que le désordre structurel est favorable pour l’anomalie de la résistivité.

Le coefficient de température ^{de la} résistivité de SnMn amorphe ^est grand ^et négatif.

Abstract. ²⁰¹⁴ Because of the insolubility ^{of Mn in Sn the} samples ^are produced by simultaneous vapour deposition

of both components ^{onto a} liquid ^helium cooled substrate. The electrical resistivity is measured in the temperature region ^{2 K to 290 K} and results are presented for Sn films with 0.07 at % ^{Mn to 5 at} % ^{Mn. The} samples ^{show a} negative temperature coefficient of the electrical resistivity (resistance anomaly) ^{like the Kondo effect of SnMn,} reported by ^{Buckel et al.}

The resistance anomaly increases with Mn concentration corresponding ^{to a localized moment} system.

At about 1 % ^{Mn a} resistance maximum appears shifting ^to higher temperature ^with increasing ^{Mn concen-}

tration. The experimental results indicate that besides the Mn atoms, disorder of the Sn lattice is responsible ^for

the resistance anomaly.

A large ^{value of the} negative temperature coefficient exists for amorphous ^SnMn.

J. Physique ⁴⁴ (1983) ^{827-832 JUILLET} 1983,

Classification

Physics ^Abstracts

72.15C - 72.15E - 73.60D - 75.20H

1. Introduction.

Magnetic scattering

^{centres in}

nonmagnetic

^metals

can cause a

negative

temperature coefficient of the electrical

resistivity

^{at low} temperature (resistance

anomaly).

^{If the}

magnetic scattering

^{centre consists}

of a

single

^{atom with a localized}

magnetic

^moment

then the resistance

anomaly

^is

proportional

^{to the}

impurity

concentration

(Kondo effect).

^{As an}

example

may be mentioned AuFe. In other systems ^like AuCo

[1, 2]

^{and CuCo}

[3] single

^atoms of the second component

(here Co)

^{do not show a localized}

magnetic

moment. In this case the resistance

anomaly

^varies

quadratically

^{with Co} concentration

[4, 5].

Regarding

^{SnMn the}

question

arises whether Mn has a localized

magnetic

^{moment in Sn or not. The}

answer is not easy because of

experimental

^features

of this system. ^{The solid}

solubility

of Mn in Sn appears to be

negligibly

^small [6]. ^Therefore

investigations

are done

by

^{means of}

quench

condensed films.

Hauser

[7] reported

^{that Mn in Sn has no}

magnetic

moment unless Mn is situated in

grain

boundaries.

In the

liquid phase

^a

magnetic

^{moment is found}

[8, 9].

Measurements of the electrical resistance

by

^Buckel

et al.

[10, 11]

^{show a} resistance minimum of Mn in Sn. This observation is not yet clear evidence for

a localized moment of

single

^{Mn atoms. For a decision}

the concentration

dependence

^{has to} be measured.

The

following

^sections

give

further evidence that solid SnMn is a localized moment system

although

the influence of disorder of the Sn lattice upon the resistance

anomaly

is remarkable.

2.

Experimental procedure.

A master

alloy

^is

prepared by melting

^{the two} pure

components ^{in a} tantalum crucible under an argon

atmosphere.

^The

ingot

is then filed to fine

particles by

bronze tools. This

powder

^is introduced in the cryostat ^{where the in situ} measurement takes

place.

One

particle

after another is

evaporated

^{from a hot}

tantalum foil. Each

particle

contributes less than one

atomic

layer

^{in order to avoid} fractionation. The

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:01983004407082700

828

Fig. ^{1. -} Temperature dependence ^{of the} electrical resistivity

of a SnMn film deposited ^{at 5 K.}

mean

deposition

^{rate is 1} nm/s. ^{The vacuum}

during evaporation

amounts to 10-’ mbar.

The film thickness varies between 40 nm and 140 nm and is

mainly

^{70 nm. The} thickness of the film is measured after the

experiment by

^{an interfero-}

metric method. The determination of the electrical

resistivity

of the film is based on this «

optical »

thickness. The absolute ^error of the electrical resisti-

vity

amounts to

10 %.

After

evaporation

of the film upon ^a

crystalline

quartz

plate,

^{held at 5} K, the film is

stepwise

^annealed

to temperatures

T.

(as

given

^{later on in the}

table).

Simultaneously

the electrical resistance is measured.

A

typical

^curve is shown in

figure

^{1. Below}

Ta

^the

resistance of the film is reversible.

Only

^reversible

curves are discussed in the

following (expect Fig. 5).

3. Results.

Figure

2 shows the electrical

resistivity

of metastable solid solutions of SnMn after

annealing

^to

Ta

^{= 25 K.}

Regarding

^{small Mn} concentrations there exist

simple

minima of the

resistivity

which result from the resis- tance

anomaly

^{at low} temperature ^{and the}

phonon

contribution to the

resistivity

^at

high

temperature.

The resistance

anomaly

^increases with the Mn concentration. At a concentration of 1.5

%

^{Mn in Sn}

a maximum of the

resistivity

appears ^at low tempe-

rature shifts to

higher

temperature ^with

rising

^concen-

tration. This result is similar to the resistance beha- viour of

AgMn [12, 13]

^{which is a localized moment} system. ^{As a}

quantitative

^measure of the resistance

anomaly

^{we take the}

negative slope

^{of the}

resistivity

at 5 K. These values

(-dp/dT)

^are

compiled

ⁱⁿ

Fig. ^{2. -} Electrical resistivity of SnMn solid solutions after annealing ^{to 25 K as a} function of temperature. ^Ordi-

nate zero suppressed.

the table I. The resistance

anomaly

^increases

nearly linearly

^{with Mn} concentration.

By

^{no means is}

there a

quadratic dependence

^on concentration

(Table

I )

^{as in the case} of AuCo and CuCo mentioned in section 1. The increase of ( -

dp/dT)

suggests the existence of a localized moment.

The resistance

anomaly depends

^on

annealing ( Ta

ⁱⁿ

Fig. 3).

^The

negative slope

decreases with

Ta

.and vanishes at

Ta

^{= 220 K. About} this tempe-

rature Mn

precipitates

^{and the}

superconductivity

of Sn appears for Sn with 0.2

%

^Mn

(Fig. 3).

^The

decrease of the resistance

anomaly

^with

Ta

^{is valid}

for all concentrations

(Table I).

^The

negative slope

decreases even at low temperature ^between

Ta

^{= 7 K,}

12 K, ²⁵ K, ^etc. (not ^shown

here).

^{We measured}

this for Sn + 0.7

%

^Mn.

Figure

^{4 shows as one}

example

the resistance maximum as a function of the temperature ^of

annealing.

The maximum shifts

only slightly

^with

annealing

^{but the}

height

^decreases

very

markedly.

In

figure

^{5 the}

resistivity

^{of a} Sn film with 5

%

^Mn

is

plotted.

^As

compared

^to

figure

^{1 a} steep decrease of the

resistivity

^{is added. Such a decrease} of the resisti-

vity

^is

already

observed for SnCu and is

typical

for the transition

amorphous-crystalline [14].

^{In the}

amorphous

^state the film has a

negative slope (Fig.

^6,

Ta

^{= 25}

K)

and in the

crystalline

^state

(T.

^{= 40}

K)

the maximum ^can be

recognized.

The maximum is shifted to

higher

temperature ^as

compared

^with

figure

^2.

829

Table I. ^-

Snmn jilms.

Annealing temperature

Ta,

^residual resistivity po, negative

slope - dp/dT

^{taken at 5 K.}

Temperature

T min

of ^{the minimum and}

T max. of the

^maximum of the electrical resistivity.

830

Fig. ^{3. - The} resistance anomaly ^{of Sn + 0.2 at} % ^{Mn at} different annealing stages. (Enlarged ^scale compared ^to Fig. 1.)

Fig. ^{4. - Maximum} of the electrical resistivity ^of

Sn + 2 at % Mn and its annealing dependence.

Fig. ^{5. -} Electrical resistivity ^{of Sn + 5 at} % ^{Mn. This}

film grows amorphous ^when deposited ^{at 5 K.}

Fig. ^{6. -} Temperature dependence ^{of the} electrical resisti-

vity ^{of Sn + 5 at} % ^{Mn in the} amorphous ^and crystal-

line state.

831

4. Discussion.

Buckel et al.

implanted

[10, 11] Sn films with 1060 ppm Mn and found a Kondo minimum with

a

depth

^{of 5 x 10- 3 nS2}

cm/ppm.

The measurements of Buckel et al. and our results agree in the existence of a resistance minimum within the system SnMn. The concentration

dependence

^of

-

dp/dT

(Table I) ^{and the} temperature

dependence (Fig.

^2, minima and maxima similar to

AgMn)

^indicate

a localized moment system. This observation is confirmed

by magnetic

measurements

[15] :

e.g.

Sn + 3

%

^{Mn has a}

magnetic

^moment

independent

of

annealing

up ^to

Ta

^{= 80 K.}

Magnetic clustering begins

^about

Ta

^{= 150 K.}

The resistance

anomaly

( -

dp/dT)

increases with Mn concentration (Table

I)

but, ^as shown in the

following,

it increases

additionally

^{with the}

degree

of disorder. As a measure of the

degree

of disorder

we take the residual

resistivity. Figure

^{7 shows the}

resistance

anomaly

^{as a} function of the residual

resistivity.

^With

regard

^to

annealing

temperatures of

Ta

^{= 25 K, 40} K and 80 K no

clustering

^and

precipitation

of Mn takes

place.

Diffusion is too low at these temperatures ^and

magnetic

measurements

[ 15]

reveal that the

magnetic

^{moment is}

independent

^of

annealing

^for

Ta ,

^{80 K. However,} the resistance

anomaly changes

^{in that} temperature

region

^{and it is}

large,

when the film has a

high degree

of disorder

(Ta

^{= 25}

^K)

and the resistance

anomaly

is less when the film has a lower

degree

of disorder

(Ta

^{= 80} K).

That means disorder increases the resistance

anomaly.

Fig. ^{7. -} Resistance anomaly - dp/dT ^{taken at 5 K as}

a function of residual resistivity po. Points for Ta ^{= 25} K, 40 K and 80,K.

For

Ta

^{= 150 K the} temperature ^is

high enough

^that

at first

clustering

^{and then}

precipitation

^takes

place

as may be seen

by magnetic

measurements

[15].

^In

this case the decrease of the resistance

anomaly (Fig.

7)

is due to

clustering

^and

precipitation. Conclusively

we can say that both, ^Mn concentration and

degree

of disorder contribute to the resistance

anomaly.

The strong

dependence

of the resistance

anomaly

^on

annealing

^{at low} temperature

( Ta

⁸⁰

^K)

^{is very}

remarkable ^{since the}

magnetic

measurements show no

clustering

^{of the Mn atoms. We}

explain

this observa- tion in the

following

^{manner :} the resistance

anomaly

in SnMn is not caused

by

^{the Mn atom alone but}

additionally

disorder of the Sn lattice is essential.

Disorder does not

necessarily

^mean well-defined lattice defects such as interstitials, ^{etc. It is sufficient}

that the

crystal

^{structure is not}

complete

^with respect

to the

periodicity

of the atomic

positions

^{and the}

angles

between the chemical bonds of white Sn. E.g.

the

special hybridization

existent in the

crystal

^{is not}

any ^more

completely

existent in the

liquid

^{or amor-}

phous phase.

^This

problem

^has

something

^{to do with}

the

disintegration

of bonds in

grain

boundaries, ^{or in}

the

liquid

^or

amorphous

^state. Some of this bond

disintegration

^{exists as} disorder in Sn as we get ^it

by quench

condensation even without any addition of ^a second component. ^{A second} component increases the disorder and ^{can even}

bring

Sn into the

amorphous phase.

For

completeness

^we shall mention that there are two other

possibilities

^to

explain

the decreases of the resistance

anomaly

^below

Ta

^{80 K : at} first let us

consider the influence of the mean free

path.

^With

rising Ta

^{the mean free}

path

of the electrons _grows and

thereby

the interaction between the

magnetic

atoms increases. The interaction causes a decrease of the resistance

anomaly

^as may be seen

by

^the

positive slope

below the maximum. The effect of the ^mean free

path depends

upon the _average distance of the magne- tic atoms that means upon the concentration : the resistance

anomaly

^would

disappear

for low Mn concentration at a

larger

^{mean free}

path

^{than for a}

sample

^with

high

^Mn concentration.

Figure

^{7 demons-}

trates that the resistance

anomaly disappears

^{at the}

same mean free

path independent

of concentration.

Therefore the mean free

path

^{is not the}

primary origin

of the decrease of the resistance

anomaly

^below

Ta

^{= 80 K.}

Secondly

it could be

argued

^{that the Mn atoms}

cluster even below

Ta

^{= 80 K, for}

example

^{to form}

pairs.

^The

unlikely

situation would then exist that the

pairing

^{does not} manifest itself in

magnetic

^measure-

ments and that the

pairs

^{do not show a resistance}

anomaly. Additionally

this effect could

only

^{work if}

diffusion were

large enough

^{below 80 K. Even if}

diffusion does exist the

clustering

^should

depend

^on

concentration. This is not observed. The relative decrease of the resistance

anomaly

^is

independent

^of

concentration (Table I).

832

In conclusion, the drastic decrease of the resistance

anomaly

^on

annealing

^{at low} temperatures ^{is not due}

to a change ^{in mean free}

path

^or

clustering

^{but to the}

decrease in disorder of the Sn lattice. This

explanation

is

supported by

^the

comparison

between the

resistivity

of the

amorphous (Ta

^{= 25 K) and the}

crystalline phase (Ta >,

^{40 K in}

Fig.

6). ^{In the}

amorphous

^state

the lattice is the most disordered of all solid

phases.

For this case the

negative slope

^{has a}

large

^value

(Ta

^{= 25 K in}

^{Fig. 6).}

^{Hence disorder} in Sn increases the resistance

anomaly

^{of SnMn.}

The electrical

resistivity

^of

amorphous

^{SnMn as a}

function of temperature ^{resembles to}

p(T )

^{of AIMn}

[16].

^Both systems ^{have in common a}

large negative

temperature coefficient ^even at low temperature.

Glassy alloys

^{without a}

magnetic

component ^{have also}

a

negative

temperature coefficient

[17].

However, ^due

to the temperature

dependence

^{of the structure factor,}

the

negative

temperature coefficient decreases with

falling

temperature

[18].

It is

noteworthy

that in the

amorphous phase

^the

maximum of the electrical

resistivity disappears

(the

negligible

^{convex curvature in}

Fig.

^{6 has its}

origin

^{in a}

small part ^of

crystalline phase

^as

already

present ^at

Ta

^{= 25 K).} The resistance

anomaly

^is

essentially changed by

the transformation

amorphous-crystal-

line

(Figs

^{5 and} 6,

Ta

^{= 25 K and}

Ta

^{= 40} K). ^This

change

in the electrical

properties

^{has to be}

compared

with the behaviour of the

magnetic properties

^{at the}

transformation. The

magnetic

^{moment and the}

spin glass

temperature ^{do not}

change appreciably

^{when the}

phase

transforms from

amorphous

^to crystalline. ^This

is valid not

only

^with respect ^{to SnMn}

[15]

^{but also}

for other systems ^as e.g. AuFe

[19].

^{That means the}

magnetic

and the electrical

properties

^{behave diffe-}

rently

^{at the}

phase

transformation.

References

[1] BOUCAI, E., LECOANET, B., PILON, J., THOLENCE, ^{J. L.}

and TOURNIER, R., Phys. ^{Rev. B 2} (1971) ^3834.

[2] KORN, D., SCHILLING, ^{D. and} ZIBOLD, G., ^J. Phys. ^F.

9 (1979) ^1111.

[3] ^{TOURNIER, R. and BLANDIN,} A., Phys. Rev. Lett. 24

(1970) ^397.

[4] ^{HENGER, U. and} KORN, D., ^J. Phys. ^{F. 11} (1981) ^2575.

[5] HALLER, H., KORN, ^{D. and} ZIBOLD, G., Phys. ^Status

Solidi (a) ⁶⁸ (1981) ^135.

[6] HANSEN, M. and ANDERKO, K., Constitution of Binary Alloys (MacGraw-Hill, ^New York) ^1958.

[7] HAUSER, ^J. J., Solid State Commun. 11 (1972) ^507.

[8] WACHTEL, ^{E. and} ULRICH, R., Z. Metallk. 59 (1968)

227.

[9] COLLINS, ^E. W., Solid State Commun. 8 (1970) ^381.

[10] BUCKEL, W., DIETRICH, ^{M. and} KESSLER, J., ^Z. Physik

245 (1971) ^283.

[11] ^{BUCKEL, W. and} HEIM, G., ^{in :} Applications of ^Ion

Beams to Metals, ^ed. by ^{Picraux, S. T., Eer} Nisse,

E. P. and Vook, ^{F. L.} (New York : Plenum Cor-

poration) 1974, p. 35.

[12] GERRITSEN, A. N., LINDE, J. O., Physica ¹⁷ (1951) ^573.

[13] ^{RAMOS, E. D., J. Low} Temp. Phys. ²⁰ (1975) ^547.

[14] BUCKEL, W., Z. Physik ¹³⁸ (1954) ^136.

[15] ^{HENGER, U. und KORN, D., Verhandl. DPG} (VI) ¹⁸ (1983) 706.

[16] BABIC, E., KRSNIK, R., LEONTÍC, B., VU010CI0106, Z., ZORI0106, ^I.

and RIZZUTO, C., Phys. Rev. Lett. 27 (1971) ^805.

[17] ^{MATSUDA, T. and} MIZUTANI, U., ^J. Phys. ^{F 12} (1982)

1877.

[18] KORN, D., PFEIFLE, ^{H. and} ZIBOLD, G., ^Z. Physik ²⁷⁰ (1974) ^490.

[19] ^{ZIBOLD, G. and KORN, D., J.} Phys. ^{E 12} (1979) ^490.