Irradiation effects on the plasma edge of organic conductors based on TMTSF

HAL Id: jpa-00210022

https://hal.archives-ouvertes.fr/jpa-00210022

Submitted on 1 Jan 1985

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.

Irradiation effects on the plasma edge of organic conductors based on TMTSF

L. Zuppiroli, C.S. Jacobsen, K. Bechgaard

To cite this version:

L. Zuppiroli, C.S. Jacobsen, K. Bechgaard. Irradiation effects on the plasma edge of organic conductors based on TMTSF. Journal de Physique, 1985, 46 (5), pp.799-805. �10.1051/jphys:01985004605079900�.

�jpa-00210022�

Irradiation effects on the plasma edge

of organic conductors based on TMTSF

L. Zuppiroli

Section d’Etude des Solides Irradiés, ^{Centre d’Etudes} Nucléaires, ^B.P. 6, 92260 Fontenay-aux-Roses, ^France

C. S. Jacobsen

The Technical University ^of Denmark, Building ³⁰⁹ C, ^{DK 2800} Lyngby, ^Denmark and K. Bechgaard

H.C. Oersted Institute, Universitetsparken ^{5, DK-2100} Copenhagen, ^Denmark

(Rep le 30 octobre 1984, révisé le 30 novembre, accepté ^le 7 janvier 1985)

Résumé. 2014 Nous présentons ^{des mesures de} réflectivité effectuées, ^à température ordinaire, ^{sur 17} monocristaux irradiés de (TMTSF)2PF6, (TMTSF)2AsF6 ^et (TMTSF) (DMTCNQ). ^{Le domaine de} fréquences ^couvert par ^ces

mesures va de 5 000 a 12 000 cm-1. C’est le proche infrarouge, ^{où l’on trouve} habituellement la fréquence ^de plasma

03C9P ^des métaux organiques. Un modèle de Drude reproduisant ^très convenablement les courbes expérimentales,

nous avons pu ^mesurer l’influence des défauts d’irradiation sur la fréquence ^de plasma 03C9p ^{ainsi que sur le taux de} relaxation optique 03B3.

Pour des niveaux de concentration en défauts de l’ordre de quelques pour cent, alors que la résistivité à fréquence

nulle augmente exponentiellement ^{d’un ordre de} grandeur par pour ^cent de défauts, ^{le taux de} relaxation optique

augmente linéairement ^{avec la} concentration de 15 à 50 % par pour ^cent de défauts.

La fréquence ^de plasma ^{diminue de 5} % par pour ^cent de défauts. Cet effet, correspondant ^{à une} diminution du nombre de porteurs ^{est dû à des effets} secondaires des défauts d’irradiation dont le plus important ^est l’augmentation

des paramètres cristallins.

Abstract

The reflectance of 17 irradiated single crystals ^of (TMTSF)2PF6, (TMTSF)2AsF6 ^and (TMTSF) (DMTCNQ) ^{have been measured at room} temperature ^{in the} frequency range 5 000 ^to 12 000 cm-1 corresponding

to the metallic plasma edge. The reflectance curves have been satisfactorily ^{fitted to a Drude} model in order to measure the influence of radiation induced defects on the plasma frequency ^{and the} optical relaxation rate.

At concentration levels of the order of a few mole % ^of defects, ^{where the room} temperature ^dc resistivity ^is

known to increase exponentially ^{with dose} by ^{an order of} magnitude ^{for each new} percent ^of defects, ^the optical

relaxation rate was found to increase linearly ^{with the} concentration, ^{with a} slope ^{of the order of a few 15-50} %

per percent ^defect

The plasma frequency ^{was observed to decrease at a rate of about 5} % per percent defect. This effect, correspond- ing ^{to a decrease of the number of carriers, is} explained by secondary ^{effects related to} defects, ^the principal ^of

which is the change ^{in lattice} parameters ^{due to} irradiation.

Classification

Physics ^Abstracts

72.15N - 72.30

1. Introductiom

Recent, ^{as well as} early experiments ^{on irradiated} organic conductors, ^have shown that the introduction of strong foreign potentials ^as the level of a few tenths of a percent ^{to a few} percent, ^induces drastic changes

in the resistivity of the metallic state of these conduc- tors [1-4].

In the early days of these studies such an unusual

sensitivity ^was sometimes attributed to particular

aspects ^{of the} electron-phonon scattering (phonon drag, ^libron drag, Frohlich collective modes... [1, 5]).

During ^{the same} period, ^{a more} simple, single-electron picture ^was developed ^to understand the longitudinal

as well as the transverse conductivity dependences ^with

defect concentration and with temperature [6, 7].

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:01985004605079900

800

Recently, this model was completed ^{in order to include}

Coulomb effects [8]. ^{It was assumed that the} strong- irradiation defects simply ^{cut the metallic} chains into

weakly interacting ^metallic segments. As the defect concentration is increasing ^and reaching levels of the order of 1 mole %, the dc and microwave transport,

as well as the electron spin ^{resonance linewidth were}

shown to be more and more dominated by interstrand

hopping [7], the microwave dielectric constant was

shown to exhibit the concentration dependence pre- dicted in an interrupted ^{strand model} [9], and the low temperature magnetism ^was explained ^{in terms of} weakly interacting unpaired spins sitting ^on magnetic segments [10, 11 ].

Thus, the intrinsic behaviour of electrons within the conducting ^{chains was observed to} disappear ^from

the transport ^and magnetic properties ^{of irradiated} samples, ^{even at} moderate concentration levels, ^the statistical properties ^{of the} assembly of metallic _par- ticles dominating ^{over the intrinsic} processes. The aim of the present paper is ^to detect possible changes ^{in the}

intrinsic properties of the electrons due to irradiation.

For this purpose, the frequency ^{of the} excitating ^field

was raised from dc or microwaves to the optical range and the chain axis reflectivities of irradiated samples ^of (TMTSF)2PF6, (TMTSF)2AsF6 ^and (TMTSF) (DMTCNQ) ^were measured in the infrared _range from 5 000 to 12 000 cm-1 i.e. in the region ^{of the}

metallic plasma edge.

A similar experiment, performed ^on (TTF) (TCNQ)

was reported ^{in the} early paper of Gunning ^and Heeger [12], ^{but the} irradiation doses were very low and the optical conductivity ^{was found to be insensi-}

tive to defects in the _range of a few tenths of a percent, whereas the room temperature conductivity, ^{as mea-}

sured at dc or microwave frequencies, ^{was found to}

decrease by ^{a factor of two.}

Optical reflectivity measurements are known to be

an important ^{tool for the} investigation of the electronic structure of organic conductors [13], especially ^{in the} (TMTSF)2X family ^of Bechgaard ^{salts where exten-}

sive infrared reflectance measurements were used to extract information on bandwidths and on charge

transfer bands in the _pure samples [14].

2. Experimental.

Single crystals ^{of the materials were} prepared ^either chemically ^or electrochemically, ^{as described in refe-}

rence [15] or [16], depending ^{on the} compound. They

have been irradiated in three different irradiation

devices, ^{that we describe} shortly, ^{and at two different} temperatures.

i) ^{In the} samples ^of (TMTSF) (DMTCNQ) ^the

radiation induced defects have been produced by

irradiation in the primary ^{beam of a} x-ray tube; ^the

8 keV photon ^{flux was about 660} roentgen/s, ^corres-

ponding ^{to an} absorbed energy of 35 Gy/s (3 ⁵⁰⁰ rad/s) (1).

ii) ^{For one} group of samples ^of (TMTSF)2PF6 (1 ^to 4), the defects have been produced by ^room temperature irradiation in a low _energy electron acce-

lerator (40 ^keV electrons) described elsewhere [ 17, 18 ],

at an electron flux of 0.3 J.IA/cm2, corresponding ^{to an}

absorbed _energy at the surface of the sample ^of

1.7 kGy/s. ^{In order to} avoid surface contamination,

the irradiations have been performed in clean turbo- molecular vacuum.

iii) ^{For the} other group of (TMTSF)2PF6 crystals (5 ^to 8) and for the (TMTSF)2AsF6 samples, ^{we have}

used a van de Graaff accelerator producing ^{2.0 MeV} electrons; ^{the flux was} also of 0.3 IlA/cm2, correspond- ing ^{to an absorbed} energy of 0.45 kGy/s ^{in the} sample.

In order to avoid surface contamination and to check the role of the irradiation temperature ^{on the defect} production, the bombardements have been performed

in a 20 K liquid hydrogen ^bath.

The irradiation parameters ^{of the 17} samples ^mea-

sured are reported ^{in table} I, including ^{the defect con-}

centrations which were determined as follows. In

(TMTSF) (DMTCNQ) the defect concentrations were

deduced from dc transport measurements by ^simul-

taneous measurements of longitudinal ^{and transverse} resistivity. The defects detected by ^this method, ^des-

cribed in reference [19], ^are those which produce

random foreign potentials high enough ^to interrupt ^the conducting chains and change ^{them in an} assembly

of metallic segments. ^Electron spin ^resonance experi-

ments performed ^on irradiated (TMTSF) (DMTCNQ)

have demonstrated that the number of ^« paramagne- tic centres », responsible ^{for the low} temperature Curie-like tail, is twice the number of defects deter- mined from dc transport [20]. Although ^{each method}

is reliable and reproducible, ^{there is some unknown}

factor of the order of one in the absolute number of defects, ^{due to the} distribution between both chains and the percolation character of the conduction _pro-

cess. Furthermore, ^the precise ^{chemical and} crystallo- graphical ^{nature of} irradiation defects in organic

conductors is not yet known, except ⁱⁿ quinolinium (TCNQ)2 ^{where one of} the defects has been identified

recently [21]; ^{in the case of} (TMTSF) (DMTCNQ), ^a speculation suggests ^that 1/4 of irradiation defects are

induced on the TMTSF chains and 3/4 ^{on the} DMTCNQ [20]. Fortunately, (TMTSF)2PF6 ^and (TMTSF)2AsF6 ^are single ^chain compounds; ^but (TMTSF)2AsF6 ^{has never been} irradiated before the present experiment, ^and (TMTSF)2PF6, ^{which was}

irradiated several times, especially ^{at low doses} [22-24],

was not studied seriously ^{from the} point ^{of view of} (1) ¹ Gy ^{is an} energy per unity ^of mass, of 1 joule per kilo- gram of compound, absorbed in the sample by ^electronic

excitation from the incident ionizing radiation, ^{while 1 rad}

is 100 ergs/g. ^In (TMTSF) (DMTCNQ), ¹⁸⁰ kGy (18 Mrad)

correspond ^{to an} absorbed energy of 1 eV per molecular unit

Table I.

Irradiation parameters of ^the different samples ^measured.

damage ^rate determinations. The only compound ^of

the (TMTSF)2X series where the defect production

is known quite well is the ambient pressure supercon- ductor (TMTSF)2CI04. Sanquer and Bouffard have shown that 10.8 MGy ^of energy, absorbed in such a

samples per unit _mass, during ^an irradiation at room

temperature, produce ^{1 mol} % ^of spins ^as observed in the low temperature Curie-like tail [11 ]. ^{Most of the} experimental results in irradiated organic ^conductors

of the (TMTSF)2X series have been interpreted ^with

the idea that the molecular decomposition ^is indepen-

dent of the precise ^nature of the counter-ion X and that the rate in the single ^chain compounds ^{based on}

TMTSF molecules is of the same order as in (TMTSF) (DMTCNQ) [22-24]. ^{As a} starting point, ^{it is conve-}

nient to follow the previous ^authors.

Another important ^{effect on the} damage production

in organic conductors has been discovered recently

and should be mentioned here : it is the temperature effect. Samples ^of Qn(TCNQ)2 have been shown to be

damaged ^two times faster at 300 K than at 20 K [25].

In (TTF) (TCNQ) radiation induced defects are

produced ^three times faster at 300 K than at 20 K [26].

Thus, it is reasonable to correct the defect concentra- tions for the low temperature irradiations by ^{a factor}

of two with respect ^{to the room} temperature ^determi- nation.

The present discussion about the absolute concen-

tration of defects shows clearly that, within the present

state of the art, it is difficult to measure the absolute number of radiation induced defects with an error

lower than 50 to 100 percent ^{and often even more.}

In order to do it more precisely ^{in the} present paper,

we have decided to check the previous ^reasonable hypothesis ^{in the} following ^way.

i) ^The temperature ^{effect on the} damage production

has been checked directly by optical experiments by using samples irradiated at 20 K as well as samples

irradiated at room temperature (see ^Table I).

ii) ^A complementary check of the relative sensiti- vities to irradiation of (TMTSF)2PF6, (TMTSF)2AsF6

and (TMTSF)(DMTCNQ) ^{has been} performed ^{in the}

form of an electron spin ^resonance experiment, ^the

results of which are reported in table II. The irradia- tion conditions (2 ^MeV electrons, temperature ^{of 20} K)

were of course the same for the three compounds.

The results of both these experiments confirm the

previous hypotheses within the experimental ^accuracy,

i.e. the sensitivity ^{of the} compounds ^{of the} (TMTSF)2X

802

Table II.

The ESR lines of five irradiated samples have been recorded from ^3.6 K, ^{and the} susceptibilities

in the Curie-like tails have been measured by ^double integration of these lines. The numbers of «paramagnetic»

centres per TMTSF molecule have been deduced from ^these measurements and reported ^{here in} arbitrary ^{units. The}

main errors in the measurement of ^this quantity ^{are not due to the ESR} lines, ^{which were} pretty good ^Lorentzian signals, ^{but to} the determination of ^{the sizes and masses} of ^the very small samples. ^The present ^table demonstrates that within the experimental ^{errors, the} sensitivities to irradiations performed ^{at 20 K} of (TMTSF)2PF6 ^and (TMTSF)2AsF6 ^{are close to each other.}

series to irradiations at a given temperature ^{are close}

to each other and the sensitivity ^{at 20 K} is lower than the sensitivity ^{at 300 K} by ^{a factor of two. These com-} plementary experiments ^{make us} confident in the values of the defect concentration reported ^{in table I,}

within a factor of 2.

The reflectance measurements have been performed

on a single ^beam spectrometer ^as described else- where [14]. The electric field was always polarized along ^the stacking ^{direction a.}

3. Experimental ^results.

The experimental chain axis reflectance spectra ⁱⁿ the near infrared of three of the (TMTSF)2 AsF6 samples ^{have been} reproduced ^on figure ^{1, the other}

curves being omitted for clarity. Except ^{for the most}

irradiated samples ^of (TMTSF)2AsF6 ^and PF6, ^{all the}

other crystals exhibited reflectance curves which were

successfully ^{fitted with a} Drude dielectric function

where cop ^{is the} plasma frequency, y the relaxation rate, and _E the background ^core polarization, ^which

was kept ^{constant and} equal ^to the pure sample ^value (E.

2.54 for X

AsF6 and 2.42 for X

PF6) [14].

The reasonable quality ^{of the fits} is demonstrated in

figure ^{1 and the} resulting values of the plasma ^fre-

quency cop ^{and of the} relaxation rate expressed ^{in cm -1}

are reported in table III. A natural, disorder induced increase of the relaxation rate with increasing ^defect

concentration is observed in both cases. It is of the order of ^a few hundred cm-1 _per mole % of defects as

shown in figure ^{2. More} unexpected ^{is the} significant

and systematic decrease of the plasma frequency ^with increasing damage. ^{If the} change ⁱⁿ plasma frequency

Fig. ^1.

Reflectance spectra ^{of three} irradiated samples ^of (TMTSF)2AsF6 ^{recorded in the near infrared at room tem-}

perature. ^The reflectance scale is logarithmic. ^{The continuous}

and dotted lines represent ^{the fits of the} experimental ^curves by ^a Drude model.

is converted to change in number of carriers according

to (02

^{4nne2jm*, and if the} concentration levels are

correct, each defect removes 8 electrons from the Fermi

sea in (TMTSF)2AsF6 and 10 in (TMTSF)2PF6

irradiated at the same temperature ^{of 20 K.}

The samples ^of (TMTSF) (DMTCNQ) ^{are not}

included in table III ; ^{it was} impossible ^{to achieve} acceptable fits of their spectra, ^{due to} their bad surface

quality, producing ^diffuse scattering ^of light (a long

300 K x-ray irradiation, ⁱⁿ presence of air, produces ^an

important degradation ^{of the} surface). Nevertheless,

looking carefully ^{at the curves, one can} say that ^a

Table III.

Parameters of ^{the Drude} fits of ^the plasma edges.

Fig. ^2.

^The optical relaxation rate is plotted ^{as a function}

of the defect concentration. The dots correspond ^to samples

irradiated at 300 K ((TMTSF),PF6, ^{1 to} 4) ^{the crosses and} triangles ^to samples irradiated at 20 K((TMTSF)2PF6, ⁵

to 8 and (TMTSF)2AsF 6’ ^{1 to} 6).

decrease of the plasma frequency with irradiation is also qualitatively visible in (TMTSF) (DMTCNQ).

The reflectance curves of the most irradiated samples

of (TMTSF)2AsF6 ^are plotted ^on figure ^{4 in order to}

show the extreme effects of irradiation when the defect

Fig. ^3.

The square of the plasma frequency plotted ^versus

the defect concentration.

Fig. ^4.

^At defect concentrations of the order of 10 %

the Drude edge disappears completely from the reflectance

curves. Notice that the frequency range has been extended towards the low frequencies ^with respect ^to figure ^1.

concentration reaches 10 %. ^{In this} experiment ^the frequency range of the measurement has been some-

what extended towards the low frequencies.

4. Discussion.

The present results demonstrate that nothing ^drastic

occurs to the underlying local electronic structure of

organic conductors, ^when submitted. to levels of

irradiation destroying ^{a few} percent of their molecules.

804

As mentioned above such levels of damage ^increases

the room temperature resistivity by ^{several orders of} magnitude [17]. ^{We shall now} discuss in more detail the moderate changes ⁱⁿ optical parameters.

Within the conventional tight-binding model, ^the

square of the plasma frequency ^is given by [14] :

In this equation, n ^{is the carrier} density, t the transfer

integral, a the lattice constant in the chain direction,

and it has been assumed that there is 1/2 ^carrier per molecule in the pure sample, ^as appropriate ^{for the}

materials discussed here. Typically, ^from figure 3,

w2p ^{decreases about 9 % for 1 %} defects. To account for this, ^{we must} consider each parameter ^of figure ^2.

The subtle point ^{is that} the lattice constants change ^on

irradiation. _x-ray diffraction has shown an increase in a of order 1 % ^{for 1} % defects. This of course leads to a rather strong decrease in t because of the _exponen- tial dependence ^{of the wave function} overlap. ^Values

based ^on the observed blue shift of the plasma edge ^on cooling [14], ^{or on molecules} orbital calculation con-

sistently suggest ^this change ^{to be 5} % ^{for 1} % ^increase

in a. Thus the lattice expansion appears ^to be the dominant factor. This expansion, ^{of course, also has a}

direct influence on the product ^{na2 = 0.5} a21V m’

where V m ^{is the} average volume per molecule. If the relative change ⁱⁿ V. [18] is somewhat larger ^{than the}

relative change ^{in a, na2 could} also decrease a little due to that. Two other possible effects should also be mentioned. Firstly, ^{some carriers} (- ^one per defect)

may be bound to the defects with energies ^{in excess of} ficop. These will not contribute to n. Secondly, ^{the bar-}

riers between the metallic segments prevent ^Drude excitations between them. This gives ^{rise to an effec-}

tive reduction of the oscillator strength, ^{which is} presumably smaller than the defect concentration [27].

Thus, although ^{the shift in} plasma edge position may be due to a combination of several factors, ^{it is} easily

accounted for and the lattice expansion ^{appears to be}

the dominant factor. Similar changes ^{in the} plasma frequency ^can also be obtained by applying ^{an external}

pressure. Welber et al. [28] ^{measured the} optical pro-

perties of (TTF) (TCNQ) ^under pressure and obtained.

ACOI/CO2 - ^{4 % for ðhlb 1’-1 1 %, thus} confirming ^the

order of magnitude ^{of our results. Of course, one} expects ^{a more} homogeneous ^{effect on the lattice} parameter by using ^{an external} pressure than by introducing ^{internal stresses} through radiation induced defects. In a recent experiment ^on (TTF) (TCNQ) [25], ^the x-ray diffraction patterns ^of irradiated _crys- tals were recorded with ^a good angular resolution in order to measure not only ^the parameter change,

but also the width of the Bragg ^{lines. The} parameter change ^{was about 1} % per percent defect while the

changes ^{in the widths of the} Bragg peaks ^{were unde-}

tectable below 1.5 %. ^{Because of the random distri-}

bution of defects, the internal strain inhomogeneity

reflected in the changes in linewidth represents only

a minor effect with respect ^to the uniform parameter change.

Above we attributed the increase in optical ^relaxa-

tion rate to the increasing ^{disorder in the} crystals. ^To

be more specific, ^we consider the origin ^of optical absorption ^at frequencies of the order of the band width (which ⁱⁿ TMTSF2X ^{salts is}

^{1 eV or}

8 000 cm -’ [14]). ^{The most} commonly accepted ^source

is phonon assisted electron-hole excitations. We note that when fico > 4 t, such _processes are only possible by emission of high energy phonons (molecular ^vibra-

tion quanta). ^{At lower} frequencies low energy acoustic and optic modes may participate ^as well. This explains why ^{y is almost} always ^{observed to} decrease with fre- quency in this range, and also why the dc and optical

conductivities are usually different. In the present

context it is important ^{to note} that if the introduction of defects made a strong ^{increase in the local scatter-} ing rate, it should be visible in the lower frequency

range of ^our data. Again, ^no strong ^{effect is seen, and} consequently, ^the change ^{in dc} transport is related to the segment model. The observed increase in _y may have two different origins. The first is simply ^{that the}

increase in disorder will relax the requirement ^for

momentum conservation, ^and thus, ^a greater ^number of absorption processes ^are allowed. For strong disorder direct electron-hole excitations may domi-

nate. Another conceivable effect of disorder is to make

possible ^direct optical excitation of plasmons. ^Present knowledge ^{is not detailed} enough ^to give any quanti-

tative analysis. ^The very different dependences ^on

defect concentrations for samples irradiated at diffe- rent temperatures may hint that the disorder poten- tials have ^a somewhat different character in the two cases.

Finally, ^{a few comments on} the results for the highly

irradiated samples ^{are in order} (Fig. 4). ^The complex disappearance ^{of metallic features} suggests ^{that the} local electronic structure is strongly perturbed. ^{It is} possible ^{that the} segment length has become short

enough, ^{that the} optical transitions should rather be described as being ^from bonding ^to antibonding ^states

in pairs ^{or trimers of} molecules. Typical energies ^for

such transitions would be 2 t + 0.5 eV - 4 000 cm-1.

Another explanation ^could be that the direct optical

excitation of plasmons completely ^{dominates the} spectrum.

Conclusion.

Six years ago, Gunning ^and Heeger [12] ^{have found}

that the Drude edge of irradiated (TTF) (TCNQ) ^was

insensitive to irradiation inducing strong ^{defects at the} level of a few tenths of ^a percent, while the dc and microwave resistivities were increasing by ^{a factor}

of two.

The present study, performed ^{on three} compounds

of the TMTSF family, ^{at room} temperature, ^demons-

trates that this decoupling ^{of the} optical ^{and low} frequency transport processes is simply ^{related to the} inhomogeneous character of the irradiated sample composed of metallic strands interrupted by insulating

defects.

Acknowledgments.

One of us (C. ^S. J.) ^is supported by ^the Royal ^Danish .Academy of Sciences and Letters, through ^{a Niels} Bohr fellowship.

References

[1] CHIANG, ^C. K., CoHEN, ^M. J., NEWMAN, ^{P. R.} and HEEGER, ^A. J., Phys. ^{Rev. B 16} (1977) ^5163.

[2] HOLCZER, K., GRUNER, G., MIHALY, ^{G. and} JANOSSY, A., Solid State Commun. 31 (1979) ^145.

[3] MIHALY, G., BOUFFARD, S., ZUPPIROLI, ^{L. and BECH-}

GAARD, K., ^J. Physique ⁴¹ (1980) ^1495.

[4] BOUFFARD, S., CHIPAUX, R., JEROME, D. and BECH- GAARD, K., Solid State Commun. 37 (1981) ^405.

[5] YESHURUN, ^{Y. and} WEGER, M., Phys. ^{Rev. B 21} (1980) ^5646.

[6] ZUPPIROLI, L., BOUFFARD, S., BECHGAARD, K., HILTY,

B. and MAYER, ^C. W., Phys. ^{Rev. B 22} (1981)

6035.

[7] FORRO, L., BOUFFARD, ^{S. and} ZUPPIROLI, L., ^J. Physique Colloq. ⁴⁴ (1983) ^C3-927.

[8] ZUPPIROLI, L., Berichte der Bunsen-Gesellshaft für Physikalische ^{Chemie 88} (1984) ^304.

[9] JANOSSY, A., HOLCZER, K., HSIEH, ^P. L., JACKSON, ^{C. M.}

and ZETTL, A., Solid State Commun. 43 (1982) ^507.

[10] KORIN-HAMZIC, B., MILJAK, ^{M. and} COOPER, ^J. R., Mol. Cryst. Liq. Cryst. ⁸⁵ (1982) ^177.

[11] SANQUER, ^{M. and} BOUFFARD, S., International Confe-

rence on Synthetic ^Metals, Abano Terme (1984)

to be published ^{in Mol.} Cryst. Liq. Cryst. (1985).

[12] GUNNING, ^{W. J. and} HEEGER, ^A. J., ^Solid State

Commun. 29 (1979) ^585.

[13] TANNER, ^D. B., ^{in Extended Linear Chain} Compounds,

Vol. 2, ed. J. S. Miller (Plenum Press, ^New York and London) 1982, p. 205.

[14] JACOBSEN, ^C. S., TANNER, ^{D. B. and} BECHGAARD, K., Phys. Rev. Lett. 46 (1981) 1142, ^and Phys. ^{Rev. B28} (1983) ^7019.

[15] JACOBSEN, ^C. S., MORTENSEN, K., ANDERSEN, ^{J. R.}

and BECHGAARD, K., Phys. ^{Rev. B 18} (1978) ^905.

[16] BECHGAARD, K., JACOBSEN, ^C. S., MORTENSEN, K., PEDERSEN, ^{H. J. and} THORUP, N., Solid State Commun. 33 (1980) ^1119.

[17] ^BOUFFARD, S., ^Thèse d’Etat, Orsay (1981) ^and Rapport

C.E.A.-R-5015 (1982).

[18] TROUILLOUD, P., BOUFFARD, S., ARDONCEAU, ^{J. and} ZUPPIROLI, L., ^Philos. Mag. ^{B 45} (1982) ^277.

[19] ZUPPIROLI, L., ^Radiat. Effects ⁶³ (1982) ^53.

[20] ZUPPIROLI, L., DELHAES, ^{P. and} AMIELL, J., ^J. Physique

43 (1982) ^1233.

[21] MERMILLIOD, ^{N. and} SELLIER, N., ^J. Physique Colloq.

44 (1983) ^C3-1363.

[22] CHAIKIN, ^P. M., ^MU-YONG CHOI, HAEN, P., ENGLER,

E. M. and GREENE, ^R. L., ^Mol. Cryst. Liq. Cryst.

79 (1982) ^79.

[23] BOUFFARD, S., RIBAULT, M., BRUSETTI, R., JEROME, ^D.

and BECHGAARD, K., ^J. Phys. ^{C : Solid State} Phys. ¹⁵ (1982) ^2951.

[24] FORRO, L., ^Mol. Cryst. Liq. Cryst. ⁸⁵ (1982) ^315.

[25] ZUPPIROLI, ^L., HOUSSEAU, N., FORRO, L., ^GUIL- LOT, J. P. and PELISSIER, J., Fading of the diffraction

patterns ⁱⁿ irradiated organic conductors, to be

published (1985).

[26] SANQUER, M., BOUFFARD, ^{S. and} FORRO, L., private

communication and to be published (1985).

[27] RICE, ^{M. J. and} BERNASCONI, J., J. Phys. ^{F : Metal} Phys. ² (1972) ^905.

[28] WELBER, B., SEIDEN, P. E. and GRANT, ^P. M., Phys.

Rev. B 18 (1978) ^2692.