Magnetic behaviour of spin 1/2 finite chains intercalated into graphite. Example of the copper chloroaluminate complex

HAL Id: jpa-00246730

https://hal.archives-ouvertes.fr/jpa-00246730

Submitted on 1 Jan 1993

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.

Magnetic behaviour of spin 1/2 finite chains intercalated into graphite. Example of the copper chloroaluminate

complex

M. El Hafidi, G. Chouteau, V. Polo, P. Pernot, R. Vangelisti

To cite this version:

M. El Hafidi, G. Chouteau, V. Polo, P. Pernot, R. Vangelisti. Magnetic behaviour of spin 1/2 fi-

nite chains intercalated into graphite. Example of the copper chloroaluminate complex. Journal de

Physique I, EDP Sciences, 1993, 3 (2), pp.371-376. �10.1051/jp1:1993138�. �jpa-00246730�

Classification

Physics

Abstracts

75.30E-75.50E

Magnetic behaviour of spin 1/2 finite chains intercalated into

graphite. Example of the _copper chloroaluminate complex

M. El

Hafidi(~,~),

G.

Chouteau(~),

V.

Polo(~),

P.

Pemot(~)

and R.

Vangelisti(~)

(~)

Service National des

Champs Intenses(*), CNRS,

B-P- 166, F-38042 Grenoble Cedex 9, France

(~)

Facult6 des Sciences 2, B-P- 6621, Casablanca 04, Maroc

(~)

Laboratoire de Chimie Min6rale

Appliqu6e,

Universitd Nancy i, B-P. 239, F-59506

Vandceuvre-les-Nancy Cedex,

France

(Received

it June 1992, accepted in final form 9

July1992)

Rdsumd Des _nouveaux

compos6s

chloroaluminates de cuivre et de cobalt ont 6td

pr6par6s

puis ins6r6s dans le

graphite.

Dans le _cas du compos6 _au cuivre

l'analyse

aux rayons X montre

une structure _en chaines. Les propr16t6s

magn6tiques s'interprbtent

bien dans _un modble de chaines finies de

spins 1/2

port6s _par )es ions

Cu~+.

_A basse temp6rature la contribution des chaines

impaires

est dominante et du type loi de Curie.

Abstract. New chloroaluminate

compounds

with copper and cobalt have been

prepared

and intercalated into

graphite.

For the copper intercalated

compound

X _ray

analysis

shows _a chain structure. The

magnetic properties

_are consistent with _a model of finite chain distribution of spin

1/2 ^Cu~+

ions. The contribution to the

susceptibility

of the short chains

containing

_an

odd number of spins ^is _a

paramagnetic-like

tail which is dominant at low temperatures.

1 Intro ductiou.

The weak Van der Waals

coupling

between the

graphene layers

in

graphite

allows intercalation of _a wide

variety

of chemical

species.

The distance between two

planes

is

changed

in _propor-

tionality

with the

guest

^{size. The} intercalation _process is also driven

by

the amount of

charge

which _can be

exchanged

between the intercalated

specie

and the

graphite.

^The

resulting

_com-

pounds

have _a

layered

structure and

are

good

model

systems

for the

study

of _many

physical

(*)

Laboratoire

asso~i6

I l'Universit6

Joseph

Fourier Grenoble-1.

372 JOURNAL DE

PHYSIQUE

I N°2

properties,

like

transport properties

and

magnetism

_{as a} function of the effective dimension-

ality.

The

magnetic properties

of many transition metal chlorides have been studied in the

pristine compounds

and _more

intensively

in the

graphite

intercalation

compounds (GIC'S).

In

most _cases

they

_can be

interpreted

within the framework of 2D models. A _way for

reducing

the

magnetic dimensionality _parameter jJ'/Jj (inter-

to

intraplane

interaction

ratio)

consists in

increasing

the

layer spacing

while

keeping

_a

large intraplane

J value. This is realized in

GICS in which the

exchange

interaction is much

greater

ⁱⁿ one direction than in the others.

The intrinsic

magnetism

of such low-dimensional lattices continues to

request

a fundamental interest.

Copper (II) compounds usually provide

_a

large variety

of structures which have led to many

interesting

and

significant spin-1/2 Heisenberg magnetic systems, including

lD and 2D

ferromagnets _{[1, 2],}

lD

antiferromagnets _{[3, 4],}

_as well

as more

complicated polymeric systems [5].

Jahn-Teller distortion effects

play probably

_an

important

role in the wide

variety

of

magnetic properties

of these

compounds.

In the

present work,

the _structure and

magnetic properties

of _new GIC with the

complex

molecule

CuA12C18

_are studied and

interpreted

in the

framework of _a

spin-1/2

finite chains

description.

2. Structural

aspect

and

synthesis.

The

CuA12C18 complex

_seems to exist in two different

crystal structures,

monoclinic [6] ^and triclinic

[7].

^{In both}

forms,

the coordinates _of

Cu~+

and

Al~+

are the _same

nearly

_square

plane

for copper ions and tetrahedral for Al ions. The networks

composed

of

Alclj

and

Cucl/

distorted octahedra lie

parallel

to the

bc-plane

of the unit

cell, forming

_a

layered

structure of chains.

The molecule is intercalated into

highly

oriented

pyrolytic graphite (HOPG)

at 300 °C.

Synthesis

and structural characterization will be

given

ⁱⁿ more details elsewhere

[8].

^{The final}

compound

of first

stage(~)

^has _{an average} formula

C22CuA12C18

which shows that the

filling

factor is of about 85

%

when

compared

to the theoretical

composition

_[9]. This

filling

factor is _very

good

and

comparable

to the _one

usually

obtained in the

majority

of GIC'S. It is _now well established that the structure of the intercalated molecule

layers

is far from

being

ideal and exhibits many defects like voids and islands

[10].

The

identity period along

the

graphite

c-axis is

Ic

₌ 9.47

I, _comparable

_{to that of}

_{G-CuC12 (9.37} I).

The three-dimensional

stacking

consists in the

three-layer

_sequence

Cl-M-Cl (M=Cu,

Al with the ratio _:

2)

_as

proved by

the measured electronic

density profile p(z).

The

in-plane

structure of the intercalants is

a

hexagonal

_array with the

parameter

_{a =} 13.8

I.

_The

_in-plane

_coherence

length

for this structure is of the order of 300

I.

3.

Magnetic

behaviour.

The measured

in-plane

static

susceptibility

_{x =}

fit/H

_as

reported

in

figure

I does not show any characteristic _feature of _a

magnetic ordering

neither

discontinuity

_nor

peak.

It is almost

independent

of the

temperature

below

TM"

31K

except

for _a drastic increase at very low

temperature

^{far below}

TM-

In the

region

4 _< T _< 25

K,

_x is

satisfactorily

described

by

the law _{: x = xo +}

C/T

where the second _term is dominant. We evaluated C

= 0.100 K-emu

/mol.

yielding

an effective moment pea =

0.89pB

_{per copper}

(II)

ion. This _moment is smaller than the free ion

Cu~+

_{moment value}

(1.9pB)

The additional term is small _xo " 0.008 _emu

/mot.

(~)

^The stage ^{is defined} as the number of

adjacent graphene layers separating

two successive

planes

of the intercalated

specie.

^~

After substraction of the

paramagnetic term, C/T,

the

remaining susceptibility

also

represented

in

figure

I

displays

a very broad maximum around

TM-

^Such a maximum is

expected

for _an

antiferrromagnetic

linear chain

system [11].

^Broad

susceptibility

maxima could also _occur in

systems

^{of isolated clusters} _or two-dimensional sheets of

exchange coupled spins [12].

G-CuAl~cl~

40

6

~i

'

,

30 ,,

~ » ,a

£ _{20 a'''}

J

I7

~$

Q-I

'°'~li~lc/T

o

0

20

(K)

Fig.

1.

Susceptibility

_versus temperature at H

= 0.5 T

applied

in the

graphite plane.

At low tem- perature, ^the measured data _are fitted with

x(T)

_{= xo}

+C/T.

The extracted

susceptibility corresponds

to

x-C/T.

The insert shows the

plot

of _x

as a function of

I/T

which allows for the determination of C and _xo at low temperatures.

The anomalous

paramagnetic

^behavior observed at low

temperature,

_may

originate

in I) ^{the existence of diluted} or isolated

spins (or

clusters of

spins)

which behave _{as a} localized

paramagnetic

moments

[13], it)

^the _presence of loose

spins

which arises in frustrated

systems [14,

15] or

iii) unpaired spins

at the free ends of

independent

chains

containing

_an odd finite number of

spins [11, 16].

^{This latter}

possibility

is _more consistent with the structure of the

present

compound.

Within this

interpretation

the "substracted

susceptibility" x(T) C/T

would be

essentially

due _at low

temperature,

to _even

spin

^{chains and}

/or

to

long

chains. The

temperature

at which the maximum of

x(T) C/T)

_occurs

gives

_an evaluation of the intra-chain interaction J which is

responsible

for the

short-range

correlation between the

spins.

A

semi-quantitative analysis

within the _one dimensional

magnetic

models is

possible

if the system ^is

considered,

in

a crude

approximation,

_m made of

independent

identical

antiferromagnetic

chains.

Following

this

assumption,

_we have found that the Fisher model

[17] adapted

for finite

antiferromagnetic Heisenberg

chains

by

Smith and

Friedberg _[18] provides

_a

good representation

of _our results with _an

antiferromagnetic exchange integral J/kB

" -40 K. At

high temperatures

far above

TM,

the measured

susceptibility

follows _a Curie-Weiss law _{: x "}

C'/(T 9).

The

assymptotic

Curie-Weiss

temperature

⁹ is

negative (9

m -44

K)

and in

good agreement

^{with the}

quantity

)zS(S

+

I)£, _(S

=

1/2,

_{z =}

2,

^J ₌ -40

K),

which is

predicted

for

Heisenberg

linear finite- chains in

th/ _high temperature

^limit

[16].

^{The effective} moment

(pea m1.64pB)

calculated from

C' (C'

_m 0.334

K.emu/mol)

is consistent with _a free

paramagnetic Cu~+

distribution _at

374 JOURNAL DE

PHYSIQUE

I N°2

high temperature.

^{The interchain}

coupling

becomes

significant only

at very low

temperature

(T

_< 4

It)

where it induces _a three-dimensional

ordering.

In this range the

approximation

of

independent

chains is _{no more} valid.

The _presence of

uncompensated spins

within the odd chains has been confirmed

by

the

measurements of the

magnetization

_versus field.

Representative

^isotherms are

reported

in

figure

2. The

magnetization M(H)

contains two

types

of contributions

I)

a low field

part

which exhibits below 4 K _a

positive

curvature characterisitic of _an

antiferromagnetic contribution, it)

a

high

^field

part corresponding

to the orientation process of individual

spins

in the

magnetic

field direction.

v T=1.85K

~ ~=~_~0~

~~~~~~~~~~

t~

w T=7.35K

~

a T=10.15K

~ ~v ^~

~J 0.3

~ ~ ~v a w

~

~ ~~ ~

~~~~

~

~ ~

£

V

a ~

_ ~ V a ~ W

~

~ ~

~ ^~ ~

~ ~ ~ ~ ~

'~

~

h~ ~ ~

~

~

~~

^~~

@

~

~ ~

~~

_~.~

fl

~

i~ js~

~

e

~

V~

~

~ ~~ _@

~~

~

~ ~ ~~

~ ~ ~ ~

~ ~ ~~ ~

fl~

~~

~ ~

~~

~ ~ ~

2 4 6 8 lo

Magnetic

Field

(tesla)

Fig.

2.

Magnet12ation

_curves at various temperatures as function of the

applied magnetic

field. The insert shows the low field shape of

Jf(H)

at 1.85 1<. Note that the average

magnetic

moment measured in 12 T is

0.33~B/Cu~+

The _average saturation moment obtained

by

_zero field

extrapolation

of the

high

field

part

of

fi~(H)

_curl,es is ps " 0.30

~B/Cu2+ (at

1.85

K).

If all the moments _were

aligned

in the

magnetic

field _one should find the free ion theoretical value ps "

lpB. (for Cu~+).

We _can thus conclude that about 301t of _copper ion

spins

_are

unpaired

and that

a field of12 T is not

strong enough

to break the

ant,iferromagnetic

interaction J.

The small

positive

curvature of

fi~(H)

observed in the low field

region

_may be related to the "thermal

spin flopping"

which arises

usually

in

long

_even

antiferromagnetic

chains. It

coeresponds

to _a

change

in the

susceptibility regime

due to the

lowering

of the energy ^states

by

the

magnetic

field

[16].

^{In the} pure

CuC12 crystal,

_a

spin-flop

transition has been observed at _a field

Hsr

₌ 4.I T

(at

4.2

K) along

the

antiferromagnetic

b-axis. The

jump

of the

magnetization

at this field is _very small

(0.012pB/Cu~+) _[19].

In the _case of the

present compound,

such _a transition could be masked

by

the dominant

paramagnetic

contribution _at low

temperatures

of the odd chains. For the _{same reason an} accurate determination of the eventual

spin-flop

transition from

magnetization

measurements has not been

possible

for the

CUCI2GIC [20].

4. Conclusions.

We have

reported experimental

evidence of

antiferromagnetic spin 1/2

finite chain behaviour in the _new

CuC12.2AlC13 graphite

intercalated

complex, using susceptibility

and

high

field

magnetization

measurements. Our results _are consistent with the extension of the Fisher model

developped by

Smith and

Friedberg

for finite

antiferromagnetic Heisenberg

chains. Full

quantitative analysis

is made difficult because calculations _are

performed

in _zero field and _our results show that the

susceptibility

is _very field sensitive _at least _at low

temperatures.

In

addition,

the structure is disordered and the

compound

must be described _{as an}

assembly

of chains of various sizes. The

susceptibility

must be

averaged,

for _a

given field,

_over the sizes

distribution

:

x(T, J, J')

=

En in (T, J, J') xn(T, J, J')

where

in

is the fraction of chains

containing

n

spins (odd

_or

even),

_xn is the

susceptibility

of the chain with n

spins

and

J(J')

is the

intra-(inter)-chain exchange coupling. Although

the contribution of odd chains is

dominant,

an accurate calculation _must include the statistical distribution of chains due to finite size effects. Parallel to this

study,

cobalt chloraluminate

Cocl? 2AlC13

has also been intercalated into

graphite.

This

compound

is very attractive because its _structure is formed of _very well

decoupled

linear

Co~+

chains in the c-direction

(of

the monoclinir unit

cell) [21, 22].

^{This GIC}

offers the

possibility

of

obtaining

_a canted

spin magnetic

structure which arises _more

commonly

with linear chain

systems

^and

originates

from

antisymmetric spin coupling _[23, 24].

Weak

ferromagnetism

_{may occur} due to _an

incomplete antiferromagnetic alignement

of the

spins

_as observed in

CsCoC13.2H20

and

CoC12.2D20 [25-27].

The

canting

becomes

important

when

strong anisotropy

exists _as for the

Co~+

_{ion. Another}

very

interesting

chloraluminate could be the

Ni-complex.

This

compound

is

structurally isomorphous

to the Co-chloroaluminate and is formed

by Ni++

chains with _an

integer spin

value

(S

=

I). Graphite

intercalation of this molecule is of

particular

interest to check the

special

Haldane

conjecture _[28]

which is

predicted

for S ₌ I

Heisenberg antiferromagnets. Experimentally

such _a

conjecture

_was observed in _a few

systems

_as NENP

[29].

^{It is} characterized

by

_{an energy gap} between the

singlet ground

state and the first excited states.

In

conclusion,

the

advantage

of

graphite intercalation,

is the

possibility

of

reducing

the effective

magnetic dimensionality

and to

explore lD,

^{2D and} 3D

magnetic systems.

Acknowledgements.

We wish to thank Pr. A-A-

Stepanov

for _many discussions and Dr. J.P. Boucher for _a number of useful

suggestions.

References

Ill

DE IONGH L-I- and MIEDEMA

A-R-,

Adv. Pl1yS. 23

(1974

1-

[2j DE GROOT H-I-M-, DE IONGH

L-I-,

WILLET R-D-, REEDIJK I., J. Appl. Pl1yS. 53

(1982)

8038.

[3j BATHIA

S-N-,

O'CONNOR C-I-, CARLIN R-L-, ALGRA H-A- and _DE IONGH

L-1-,

Cllem.

Pllys-

Lett. 50

(1977)

353.

[4j GRIGEREIT T-E-, LIU

Y-,

ZHOU P-, DRUMHLER I-E-, BONOMARTINI-CORRADI A-, BOND

M-R-,

PLACE H. and WILLET R-D-, J-M-M-M- lo4

(19921

831.

[5] CARLIN

R-L-, Magneto-Chemistry (Springer-Verlag, Berlin, Heidelberg, 1986)

_p. 70.

376 JOURNAL DE

PHYSIQUE

I N°2

[6] ^SHAFER

V-H-,

BINNEWIES M., LAUMANNS

R.,

and WACHTER

H.,

Z.

Anorg. Allg.

Cllem. 461

(1980)

31.

[7] KITAJIMA

N.,

SHIMANOUCHI

H.,

ONO Y. and SASADA

Y.,

Bull. Chem. Soc.

Jpn.

55

(1982)

2064.

[8] VANGELISTI R. and POLO

V.,

to be

published.

(9] ^PERNOT P., thbse Universit6

Nancy (1989) unpublished.

[10] ^FLANDROIS

S.,

HEWAT

A-W-,

HAUW C. and BRAGG

R-H-, Syntll.

^Metals 7

(1983)

305.

(iii

BONNER J-C- and FISHER

M., Pllys.

Rev, ₁₃₅

(1964)

640.

[12] ^LINES M-E-,

Pllys.

Rev. 135

(1964)

1336

J.Pllys.

Diem. Sol. 31

(1970)

101.

[13] DATE M. and MOTAKAWA M.,

Pllys.

Rev. Lett. 15

(1965)

854.

[14] ^BRAY

A-J-,

MOORE M-A- and REED

P-J-,

J.

Pllys.

C _: Sol. St.

Pllys.

II

(1978)

i187.

[15] RAMMAL R., SUCHAIL R. and MAYNARD R., Sol. St. Comm. 34

(1979)

487.

[16]

FLtCHE,

thkse doctorat, Universit6 Bordeaux1

(1978) unpublished.

ii?]

FISHER

M-E-,

Am. J.

Pllys.

32

(1964)

343.

(18] SMITH S. and FRIEDBERG A.,

Pllys.

Rev, 176

(1968)

660.

[19] ^{BILLERY D. and TERRIER}

C., Pllys.

Lett. 68A

(1978)

278.

[20] ^RANCOURT

D-G-,

MESCHI C, and FLANDROIS

S., Pllys.

Rev. B 33

(1986)

347 EL HAFIDI

M.,

CHOUTEAU G. and YAzAMI

R., unpublished

results.

(21] IBERS

J.,

Acta

Cryst.

15

(1962)

967.

(22] ^O'REILLY

D-E-,

POOLE Jr C-P-, BELT R-F- and SCOTT H., J.

Polym.

Sci. 2

(1964)

3257.

[23] DzYALOSHINSKY

I.,

J.

Pllys.

Diem. Solids 4

(1958)

241.

[24] MORIYA T.,

Pllys.

Rev. ii?

(1960)

635

Pllys.

Rev. 120

(1960)

91.

[25] ^HERWIJER

A.,

_DE JONGE

W-J-M-,

BOTTERMAN

A-C-,

BONGAARTS A-L-M- and COWEN J-A-

pl1ys.

Rev. B 5

(1972)

4618.

[26] ^COX

D-E-,

FRAzER B-C- and SHIRANE

G.,

Pllys. Lett. 17

(1965)

103.

[27] ^MORIYA T.,

Magnetism,

Vol. i

Chap.

3, G-I- Rado and H. Suhl H. Eds.

(Academic

Press, New

York, 1963).

(28] ^HALDANE F-D-M-,

Pliys.

Lett. A93

(1983)

464,

Pllys.

Lett. 50

(1983)

i153.

(29] ^RENARD

J-P-,

VERDAGUER

M.,

REGNAULT

L-P-,

ERKELENS

W-A-C-,

ROSSAT-MIGNOD J. and STIRLING W-G-,

Europllys.

Lett. 3

(1987)

945

RENARD

J-P-,

GADET

V.,

REGNAULT L-P-, and VERDAGUER

M.,

J-M-M-M- 90

(1990)

213.