HAL Id: jpa-00247274
https://hal.archives-ouvertes.fr/jpa-00247274
Submitted on 1 Jan 1996
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.
Method
H. Helberg, M. Dressel
To cite this version:
H. Helberg, M. Dressel. Investigations of Organic Conductors by the Schegolev Method. Journal de Physique I, EDP Sciences, 1996, 6 (12), pp.1683-1695. �10.1051/jp1:1996182�. �jpa-00247274�
Investigations of Organic Conductors by the Schegolev Method
H-W- Helberg (~) and M- Dressel (~,*)
(~) Drittes Physikalisches Institut, Universitàt Gôttingen, Bürgerstrasse 42-44, 37073 Gôttingen, Germany
(~) Institut für Festkôrperphysik, Technische Hochschule Darmstadt, Hochschulstrasse 6,
64289 Darmstadt. Germany
(Received 2 May 1996, accepted 18 June 1996)
PACS.07.57.Yb Other infrared, submilhmeter wave, microwave, and radiowave instruments, equipment, and techniques
PACS.72.80.Le Polymer; organic compounds (including orgamc semiconductors)
PACS.74.70.Kn Organic superconductors
Abstract. The Schegolev method is a well established technique to determme the dielectric properties of organic materials by placing the small samples m a microwave cavity and mea- suring the perturbation. The electronic properties can be studied without applying contacts and the frequency dependence gives information on the transport mechamsm. The versatility of
this unique method is reviewed by highlighting various applications on organic conductors and
semiconductors.
1. Introduction
In tue late sixties and early seventies tue development of low dimensional organic conductors and semiconductors reacued a new level and finally led to tue syntuesis of TTF-TCNQ. Tue
fragility and limited size of tue crystals, as well as uigu anisotropy caused major problems in
cuaracterizing tueir transport properties- At different laboratories a new measurement tech- nique was developed wuicu allows for a contactless determmation of tue complex conductivity by placing tue specimen in a microwave cavity [1,2]. Altuougu similar metuods bave been used for tue cuaracterization of conventional dielectric, ferroelectric, and magnetic materials [3-8],
tue contributions of Scuegolev et ai. [9-1ii were tue most influential in tue field of organic conductors, wuere tue technique was soon adopted widely [12-14] and by now is well known
as tue Scuegolev metuod.
Laboratones all over tue world je-g- Cuemogolovka, Gôttingen, Los Angeles, Sherbrooke)
utilize tuis technique on a regular basis; uowever, due to tue elaborate analysis required and tue discouraging costs of a smgle frequency point, tue microwave cavity perturbation technique
never became a standard metuod. Nevertueless, in tue frequency range from 3 GHz to 300 GHz
no otuer metuod can reacu tue sensitivity of enclosed cavities. By reviewing vanous applications
on organic matenals we want to show tue advantages and versatility of tue Scuegolev metuod.
(*) Author for correspondence (e-mail: [email protected])
Q Les Éditions de Physique 1996
2. Contactless Transport Measurements
If an enclosed resonator is sligutly disturbed by introduction of a small object, tue electrody-
namic properties of tuis object can be evaluated from tue perturbation if tue geometry of tue
sample is known [15-17]- Tue change m tue resonator properties are expressed by a complex quantity lô
= w jr /2
~~
=
~°~ ~ ~~
=
~~ °~°
,
ii)
20 Ldo 2 Mo Ldo 2 Qs Qo
wuere S and 0 indicate tue filled and tue empty cavity- AF
= rs ro is tue change of tue ualf widtu wuicu is connected to tue quality factor by Q " MoIF. AM
= us uJo is tue change of tue center frequency-
In tue quasi-static regime ((a < i, wuere ( is tue wavevector and a is tue sample size, 1e- tue sample is uomogeneously penetrated by tue electromagnetic field), tue complex frequency
suift hé /wo is related to tue complex dielectric constant ê
= e'+ le" by
A2
" 1- ~°~~ (2)
~ + N-
uJo
AUJ AM ~r ~ ~r
~ ~ ~
Mo Mo
~ ~
2wo ~2wo
~
+ N
~°~ ~
+ N2 ~~
~ ~ ~
(~ + N
~~ ~
+ N2 ~~
~ '
L°o 2Ldo L°o 2Ldo
as pointed out by Buravov and Scuegolev for tue first time [10]- In equation (2), N is tue depolarization factor wuicu can be calculated assuming an ellipsoid-suaped sample and ~ is
tue resonator constant wuicu depends on tue volume of tue sample and tue cavity and tue
resonator mode excited.
In Figure i a typical setup is suown similar to tue design used by Buravov and Scuegolev [10].
Tue waveguide couples to tue cylindrical copper cavity; a Teflon disk ensures tuat only tue TEoii is excited. Tue sample is placed in one groove of a Teflon uolder wuicu is mounted on a quartz rod. By rotating tue rod tue specimen can be moved in and out of tue cavity m order
to measure tue resonance parameters witu and wituout tue sample- Tue cavity is placed in tue
He excuange gas of a cryostat and can be ueated for temperature dependent measurements-
Tue common cavity designs are reviewed in [iii; a detailed discussion of tue measurement
technique can be found in [18-20].
From a tecunical point of view, tue main advantage of uigu frequency measurements is tuat
no contacts bave to be attacued, wuicu in particular is a problem for tiny and fragile organic crystals. Tuus no additional stress is applied wuicu often causes microcracks or even tue crystal
to break upon cooling. Air sensitive samples may be sealed in quartz tubes or coated by grease
and epoxy, for instance, and still measured in microwave cavities- Powder can be investigated, too, even wituout pressing it; just tue volume fraction bas to be known to get an absolute value [1,2, ?Ii.
3. Anisotropic Materials
For uigu frequency as well as for de measurements, tue anisotropy of big crystals can easily determined, by cutting specimen in various directions [22-24]. In tue case of tuin needle-like
couplinq Ploie
( -Heoier
ii il
_
Somple
iemperoture
~f/ Holder
$ensor G~~~v~s
Quartz Rod
Teflon Disk
Fig. l. Side view of
a cylindrical cavity used for measurements at 7 to 12 GHz. The sample is placed in either of the grooves on the Teflon tray and in this configuration it will pass through the
maximum electric field inside the cavity.
,~s
ié
"
,
+J
cz io ~~
- ~ +
+
~
O +
+ -
[ %
'Ç O ~
z ~ ++
U ~~3 >~ ~
_g C~ ~
~
O
~~
~ fi ++
u O ~+~
_q +ç
W z j
> i O ~
O io
Î %~_$~
_~ ~
~
+
io~
o SO loo lso 200 250 300 0 50 loo lso 200 2So 300
temperature [K) temperature (K)
a) b)
Fig. 2. (a) Comparison between longitudinal (X,Z,Y') and transversal(D. A,+,o conductivity of
(2,5-DCH3-DCNQI)2Cu. The measurements were done m cavities resonating at 10 GHz where the
sample is placed in the antinode of the electric and of the magnetic field, respectively. Panel (b) shows the amsotropy factor ajj la
i as a function of temperature.
90°
13s° çs°
,' 1',
~ ~
, ', ,
k[ 18Ù° Ù°
1' , ' 1
,
, , /
/ Q '
é
lis° 31s°
170°
a) b)
Fig. 3. (a) Sketch of the split resonator with the sample of thickness d placed in the center.
(b) Angular dependence of the
room temperature conductivity in units of (Qcm)~~ of PdPCI~ films measured at 20.4 GHz in) and 23.8 GHz (A). The data exhibit a cos~ dependence.
crystals it is often not possible to arrange tue contacts in tue way necessary for tue Montgomery-
von der Pauw metuod commonly used to determine tue de-anisotropy. If no crystals witu a
large extension in tue perpendicular direction are available, microwave measurements of tue transverse conductivity in tue maximum of tue electric field are possible by placing a large
number of samples side by side If.g. [25]). Positioning tue sample in tue antinode of tue mag- netic field of a cavity (B [ needle axis), eddy currents are induced in tue plane perpendicular to tue axis and allow to determine ai In Figure 2 tue conductivity of (2,5-DCH3-DCNQI)2Cu in
botu orientations is plotted as a function of temperature as obtained by cavity measurements at 10 GHz [25]. Tue anisitropy significantly increases as tue temperature is reduced mainly
due to tue temperature dependence of tue longitudinal conductivity; tue ratio is agreement witu estimations of tue overlap integral by NMR measurements [26]-' Since tue crystals grow
as tuin libers witu a typical diameter of 5 /tm, tue de resistivity cannot be measured in tue perpendicular direction; tuus tue microwave measurements for tue first time show tuat tue matenal is metallic even perpendicular to tue cuains-
Tue microwave technique gives tue possibility to actually scan tue conductivity of one sample by rotating it witu respect to tue electric field E. Tuis can be done eituer by placing a disk-like
sample in tue axis of a circular cavity resonating in tue TEoio mode [27]. If tue sample is big enougu and bas low losses, tue eutire cross-section of tue cavity can be covered; placing tue tuin sample in tue center of a rectangular TEiop cavity wuicu is spht in two parts as suown
in Figure 3a, tue angular dependence can be easily determined by rotation [27,28]. Figure 3b demonstrates tuat a film of putualocyanine can be grown well onented witu an anisotropy ratio ajj lai m 10 [29]. Witu a similar metuod tue angular dependence of reticulate doped polymers
was determined [28], wuere TTT-(TCNQ)2 crystals grown anisotropically in a polyetuylene
aa.aa'
~~o
~~ a
~
"~
is a
io
T
lKl
0~
~ àà
li
lô~
~y~
EDT-TTFIiIj
g
0GHz llb
à o
o so ioo
a) [K]
103
mm ~
~ 1 z ,j ,>~/ .
<
6 . . "
£
loi
~
f 10°
.b
~fl io-1
fl3 vo = (
C 35
o
CJ 10-z~
adc
" lx
iQ-3
ÎÙ~I ÎÙ°
ÎÙI
b) umber [cm-1]
10~
10~
10'
fi4n Q n ~
Î io3
1 °
C
t 10~ °
'%
Z
g Q
10 g
~w 10 2~
10~~
9
~yi Q
°~
u d~
~~~~ ~
d~=70% 30%
6 h~ d~=50% 50%
~~~0
20 ~0 60 80 100
température lKl
Fig. 5. Microwave conductivity of (DCH3-DCNQI)2Cu in a protonated and two deuterated mod-
ifications. While the h8 compound stays metallic, the d6 specimen becomes insulating; the mixture 70% : 30% shows a reentrant behavior.
rail during casting. Depending on the preparation conditions, systems with an anisotropy ranging tram 10 to 10~ can be obtained which is basically mdependent of temperature and frequency [27]. Since polarization microscopy shows that there are large homogeneous areas,
the anisotropy is intrinsic ta the crystals.
4. Study of Phase Wansitions
Due ta the reduced dimensionahty, organic conductors commonly undergo phase transitions which significantly change their electronic properties. Similar ta the de transport, the high frequency measurements may show a sharp drap m conductivity, but often an interesting frequency dependence is observed in the insulating state. In Figure 4a the temperature depen-
dent conductivity of the two-dimensional organic conductor a-(BEDT-TTF)213 is displayed;
the measurement was conducted at 10 GHz m bath orientation, E
ii a and E
ii b [30]. While the de conductivity rapidly draps below the metal-msulator transition TMI
= 135 K, the
high-frequency data show a plateau of constant conductivity between 30 and 110 K which is
strongly frequency dependent as displayed in Figure 4b. This phenomenon is net well un- derstood yet [31]; hopping conducting and a Spin-Peierls ground state are discussed among others.
10~
DDD ~~~ °°DDD~ ~ ( ~ ~~ô xx x~ ~
~
o n x~ô~~ ~
lÙ~ II(iÙÙl ~x~ ~~ x (
à à ~
à
É 10°
T=300K
z~ o
5 Ô
Î Al
ÀÎ lÙ~~ ~
~c Î
w
~
>
lÙ~~ x
( ~
~ ô
É x x x xx xxx à
~
+~++~ ~x x x(~A § j ~~
lÙ~~
~
ù ~ n ~ ~
~
~=60~
~~-~
0 100 200 300 5 10 20 50 100 200 k00
temperature lKi duration of exposure ta ladre [mini
a) b)
Fig. 6. (a) Microwave conductivity of a-(BEDT-TTF)iI3 for increasing doping concentration ~ersus temperature along E
ii a at 10.3 GHz. Duration of exposure to radine (+) 60 min, (Zi) 90 min, (o)
120 min, (K7) 180 min, (O) 360 min. (b) Microwave conductivity of cv-(BEDT-TTF)213 at 60 K and 300 K for bath crystal directions (Zi for E
ii a, x for E1a) as a function of duration of exposure ta radine.
From pressure studies it is known that DCNQI copper salts are in the vicinity of an insulating phase. Even the substitution of hydrogen by deuterium can dramatically change trie tempera-
ture dependence of trie conductivity (Fig. 2a) [32]. As seen m Figure 5, trie conductivity of trie deuterated derivatives (2,5-DCD3-DCNQI)2Cu (d8) and (3-D,6-D,2,5-DCD3-DCNQI)2Cu (dû) draps almost seven orders of magnitude, while (2,5-DCH3-DCNQI)2Cu (h8) remains outside the insulating phase at any temperature [25,33,34]. Most astonishing is the reentrant behavior if only 30% are substituted bydû 133]. In the low-conductivity region the microwave values are
higher compared ta the de data (sometimes several orders of magnitude) mdicating the exis-
tence of highly conducting areas even m the low conducting region. As their volume fraction
undergoes the percolation threshold, they cannot be detected by de measurements but probed by the high frequency technique.
5. Effects of Special Weatment
Since no contacts have ta be attached and the same sample can be measured several times, mi-
crowave methods are frequently used ta study the change m dielectric and transport properties
of orgamc crystals upon irradiation [35], doping [36,3î], polymerization [38] or thermal treat- ment [21]. In Figure fia the change of the temperature dependence of the 10 GHz conductivity
of a single a-(BEDT-TIF)213 crystal is shown; after doping by radine for more than one heur the low temperature conductivity starts mcreasing slightly without becoming superconducting (Fig. 6b) [37]. If a-(BEDT-TTF)213 is tempered at 70°C it transforms ta the superconducting
lÙ~
Q 10 min x=Ù.Ù
6 ~0min Ù.Ù16
o 1,3h Ù.Ù67
x 2.7h Ù.27
. 5.3h Ù58
~ à lÙ7h Ù6~
/j . 21.3h Ù9Ù
ÎÎ
+ ~2.7h 1-Ù
( ÎIÙ~
à
~ ~%mmé
)Î
'Î
lÙ°
IÙÙ 2ÙÙ 300
temperature lKl
Fig. 7. Microwave conductivity of a pressed sample of a-(BEDT-TTF)213 after stepwise annealing
at 70 °C. Parameters added up annealing duration and volume fraction ~ of the ot-phase. ~ is
determined by fitting the curves from ap and apt using a(T)
= (1- ~)ap(T) + ~apt(T).
at-Phase (Tc m 8 K). Microwaves enable the measurement of the conductivity for stepwise an- nealing after every annealing step in always the same sample (Fig. 7) [21]. Similar experiments
have been conducted on polymer films doped with o-(BEDT-TTF)213 139]. In bath cases the system can be described by a mixture of two components with changing ratio as the thermal
treatment proceeds.
In Figure 8 the real and imaginary parts of the dielectric constant of disubstituted diacetylene
DNP surgie crystals are shown as a function of temperature for diflerent states of the solid state polymerization. The ferroelectric phase transition at Tc
= 46.5 K shifts with advanced polymerization ta lower temperatures and gets weaker [38]. The peak in the dielectric constant
e'(T) is getting flat. The dielectric lasses e"(T) first increase but after three heurs of thermal treatment at 130 °C the peak decreases and finally no lasses are found for fuit polymerization.
This behavior indicates a lattice softening m the boundary between the polymerized part and trie remaining monomer.
i~
Î
11 6 oh
~~ ~
~jj .Î 8 °°°~ 4
~ ~~~
ù
~
e é °u~
Î
~Ù 2Ù ~Ù 6Ù 8Ù IÙÙ
a) temperature [K]
25
o
'~ 20 p 6 oh
~ ( ° 3h
l-S Î$ + lÙh
[ / + n 13h
.i j
Ù-S
~Ù 20 ~0 60 80 IÙÙ
b) temperature lKl
Fig. 8. Temperature dependence of the real (a) and imaginary (b) part of the dielectric constant
= e'+ ie" of
a single DNP crystal measured at 10 GHz with the electric field parallel to the b- direction of the monomer. The variation with duration of thermal polymerization at 130 °C is shown for different periods of time as indicated.
6. Thermal Expansion
Besides the electrical transport, measurements of the resonance frequency allows for determi- nation of the thermal expansion if needle shaped samples with high conductivity are placed in the electric field maximum of the enclosed resonator [40]. Figure 9 shows the relative thermal
expansion of TTF-TCNQi the Debye temperatuie is obtained from the fit of the data [41].
7. Frequency Dependence
In the first review by Schegolev [11] the most interesting point of the data on TCNQ complexes
was the sigmficant deviation of the high frequency data from the de results at low tempera-
tures and the large dielectric constant. While most examples mentioned above just utilize the contactless way of measuring the conductivity, the frequency dependence can give insight in the relevant transport mechanism. For a simple Drude-like metallic conduction, no frequency dependence is expected up to the far mfrared spectral range, however, hopping conduction or collective transport hke charge or spm density waves lead to a distinct frequency dependence
even in the microwave range [42,43]. Since enclosed resonators are usually restricted to a single
resonance frequency, either a large number of cavities have to be employed [18, 44] or a special
Ù3 ~
Îi°
TTF -T[Nù n~
/
~Î
Ù2 u'
n'
ooi
Debye-temperature 6~=38ÙK
o
IÙÙ 2ÙÙ 3ÙÙ
temperature lK]
Fig. 9. Measured relative linear thermal expansion of TTF-TCNQ (symbols) and theoretical be-
havior (fine) with the Debye temperature ÙD as parameter.
cavity design with higher modes is used [45, 46]. The frequency dependence of the conductivity of a-(BEDT-TTF)213 is displayed in Figure 4b; each of the data points at 2.3, 7, 10, 12, 24, 35, 60, and loo GHz are taken in a diflerent cavity setup. While at room temperature there is no
significant frequency dependence, the low-temperature conductivity mcreases with frequency
as a c~ uJ~ with s
= 1.3 [31].
In Figure 10 the real and imagmary part of the room temperature dielectric constant of
(TMTTF)2BF4 are shown as a function of frequency. The measurements are performed in
long rectangular cavities as shown in Figure 3a which are excited in the TEiop-mode with
large p of the order of 40 [46]. The mcrease m e'(uJ) and e"(uJ) together with the large value e' m 600 indicates a strong mode at higher frequencies. The temperature dependence of the dielectric properties is displayed in the insets; at TSF " 40 K (TMTTF)2BF4 undergoes a
Spin-Peierls transition.
à
8. Summary
The Schegolev method is a well established technique to contactless measure small samples by placing them m a microwave cavity. The main applications on orgamc crystals are the determmation of the temperature, frequency, and angular dependence of the conductivity, as
12
Z p
( à 11 D~°a ÎÎ
~
8
Ù Ù-B $ ~~~~~~~ 1°~ i
O '~
~
G o
à Ù.6
, if
ç q~
~ . 1ç~ ~
~~
tl>
IÙÙ 18Ù 26Ù IÙÙ 18Ù 26Ù o
T[K] T[K] O°O 5 É
Go ~
~ O b
~é ~ Î~
(
g" O°§j~ m~
O ° ~
l-S oS
É jé~
2
f ~oO
~ ~OO~~ ~~
h ° g' ; l
çz ~ . .
~ ~ ~ .. . .
~ l-Ù - ~
'
8 lÙ 12 15 2Ù 25 3Ù
frequency lGHz]
Fig. 10. Real part e' (left axis) and imaginary part e" (right axis) of the dielectric constant of
(TMTTF)2BF4
as a function of frequency ai T = 300 K. ai * 0.23 (Qcm)~~ and e' m 600. The insets show the temperature dependence of the conductivity ai and the dielectric constant e' measured at 9 GHz (open squares) and 23 GHz (solid squares).
well as the change upon irradiation, doping, or thermal treatment. Recently the microwave
technique has been used to study the complex conductivity and penetration depth of orgamc superconductors like (BEDT-TTF)2Cu(NCS)2 in order to obtain information on the nature of the superconducting state [4îj. The applications of this unique method are by far net
exhausted yet.
References
Ill Kawabe M.. Masuda K. and Yamaguchi J., J. Phys. Soc. Jpn 24 (1968) 1281.
[2] Dix G., Diploma Thesis, Univ. Gôttingen (19îo).
[3) Spencer E-G-, Le Craw R-C- and Reggia F., Proc. IRE 44 (1956) 790.
[4j Champlin K-S- and Krongard R.R., IRE Trans. Microwave Theory Techn. MTT-9 (1961)
545.
