UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

ESR study of DMTM(TCNQ)₂ Kirui, Joseph Kiprono 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1990_A6_7 K57.pdf [ 2.39MB ]
Metadata
JSON: 831-1.0085016.json
JSON-LD: 831-1.0085016-ld.json
RDF/XML (Pretty): 831-1.0085016-rdf.xml
RDF/JSON: 831-1.0085016-rdf.json
Turtle: 831-1.0085016-turtle.txt
N-Triples: 831-1.0085016-rdf-ntriples.txt
Original Record: 831-1.0085016-source.json
Full Text
831-1.0085016-fulltext.txt
Citation
831-1.0085016.ris

Full Text

ESR STUDY OF D M T M ( T C N Q ) 2 By Joseph Kiprono Kirui B. Sc. University of Nairobi, 1984 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S FOR T H E D E G R E E OF M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E STUDIES D E P A R T M E N T OF PHYSICS We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH C O L U M B I A March 1990 ©Joseph Kiprono Kirui, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of P H 7 S I C S The University of British Columbia Vancouver, Canada Date K n g C H l?*^! DE-6 (2/88) A B S T R A C T The ESR g-value and susceptibility measurements for D M T M ( T C N Q ) 2 have been studied as a function of angle made by crystal with magnetic field and temperature. The angular dependence of g-value is fitted to g2 = a+/3co52^ + 75in2^ for three orthogonal directions of crystal rotation. The principal g-values are close to those expected for T C N Q compounds: gl = 2.0034, g2 = 2.0030, g3 = 2.0024. The susceptibility as a function of temperature agrees with bulk susceptibility measurements except that the maximum position occurs at about 30 K. The results of Oostra et al. for bulk susceptibility showed a maximum at around 50 K. The phase tran-sition reported by Visser et al. at 272 K is observed in the ESR data as a 15% decrease in susceptibility. The linewidth is remarkably anisotropic typical of T C N Q salts. The phase transition study is done for two orien-tations of the crystal with the magnet field. In one of the orientations the Hnewidth narrows from 0.15 to 0.11 gauss and in the other it narrows from 0.24 to 0.18 gauss. In the former case there is a growth of a second line due to the twinned stack; transformation twinning takes place at the phase transition. A small level-crossing interaction is inferred from the change in relative intensities of the lines near the crossover. ii Table of Contents ABSTRACT ii LIST OF TABLES v LIST OF FIGURES vi ACKNOWLEDGEMENTS ix 1 INTRODUCTION 1 1.1 Overview 1 1.2 Previous Investigation of the D M T M ( T C N Q ) 2 4 1.3 Transformation Twinning of D M T M ( T C N Q ) 2 5 1.4 The Hubbard Hamiltonian 8 1.5 Thesis Outline 8 2 X-BAND ESR APPARATUS AND MEASUREMENTS 10 2.1 Phase-Sensitive Detection 10 2.2 Recorder Representation 13 2.3 The Magnet 14 2.4 Low Temperature Operation 16 3 T E M P E R A T U R E MEASUREMENTS 18 3.1 The Temperature Measurements with Crystal Installed for Overlap-ping Lines at 77 K 18 iii 3.2 The Temperature Measurements with Crystal Installed for Near Max-imum Separation of Lines at 77 K 25 4 E X P E R I M E N T A L D E T E R M I N A T I O N O F T H E g T E N S O R 29 4.1 The Experiment and the Principal Values of the g Tensor 29 4.2 Evidence of Level-crossing Effects 40 5 C O N C L U S I O N S A N D S U G G E S T I O N S F O R F U R T H E R I N V E S T I G A T I O N S 41 5.1 Suggestions 43 Bibliography 44 Appendices 46 A Program Files used in the Analysis of Data 47 A . l The Main Programs 47 A.2 Auxiliary Programs 47 iv List of Tables 4.1 Principal values and direction cosines for the g tensor of Stack #1 of D M T M ( T C N Q ) 2 at 77 K. The error is ±0 .0001 . 33 4.2 Principal values and direction cosines for the g tensor of Stack #2 of D M T M ( T C N Q ) 2 at 77 K. The error is ±0.0001 33 v Lis t of Figures 1.1 T C N Q Molecule 2 1.2 N-X, N-Y-Morpholinium + ( X Y M + ). The X and Y bonds can stand for hydrogen, ethyl or methyl groups 2 1.3 Dimethyl-thio-Morpholinium molecule 3 1.4 Schematic representation of structure of D M T M and of the energy levels of T C N Q bands (a) above T c space group P 2 i , (b) below T c , space group Pj or P a . The dashed arrows in the upper drawings denote the average S—>N directions of the D M T M groups. A and B refer to inequivalent T C N Q sheets. The arrows between the electron bands in the lower drawings represent the smallest energy excitations. The drawings have been taken from reference [6] 6 2.5 Block diagram of X-band ESR apparatus 11 2.6 TE102 rectangular cavity with mechanism for rotation of sample holder. The drawings were taken from reference [19] 12 2.7 Calibration graph for H-field in the region of use for conical pole tips of radius of about 80 mm 15 3.8 Linewidth variation w'ith temperature near the phase transition. The crystal was in an orientation of overlapping of the two T C N Q lines at 77 K 20 vi 3.9 Susceptibility variation with temperature near the phase transition. The crystal was in an orientation where the two T C N Q lines were overlapping at 77 K 21 3.10 Plots of derivative of absorption curves to depict the remarkable in-crease in linewidth at low temperature. The crystal was in an orien-tation exhibiting overlapping lines at 77 K. The signal on the far left in each scan belongs to the lithium calibration sample 23 3.11 Susceptibility variation with temperature. The crystal was in an orientation exhibiting overlapping lines at 77 K 24 3.12 Temperature variation of susceptibility near the transition region. The crystal is positioned in a direction in which the 2 lines are well resolved at 77 K. • represents the untwinned stack signal and • represents the growing-in "twinned" stack signal. The dashed lined shows the 15% drop of total susceptibility at the phase transition. . 27 3.13 Temperature variation of linewidth at the transition region.The crys-tal was positioned in direction in which the 2 lines were well resolved at 77 K. • represents the untwinned stack signal and • represents the growing-in "twinned" stack signal 28 4.14 Initial crystal orientation for the 3 planes of rotation in the laboratory coordinate system. The cube held on to the sample holder measures 2 mm on a side 30 4.15 g value for stacks #1 and #2 of D M T M ( T C N Q ) 2 for rotation about' a-axis at 77 K. The solid line is a fit to Eq.4.2. The stack directions make an angle of 25° with respect to each other 34 vii 4.16 g value for the two overlapping stacks of D M T M ( T C N Q ) 2 for rotation about b-axis at 77 K. The solid line is a fit to Eq.4.2 35 4.17 g value for the two overlapping stacks of D M T M ( T C N Q ) 2 for rotation about c-axis at 77 K. The solid line is a fit to Eq.4.2 36 4.18 Peak-to-peak linewidth for stacks #1 and #2 of D M T M ( T C N Q ) 2 for rotation about a-axis at 77 K. The symbol • represents stack #1 and 0 represents stack #2 37 4.19 Peak-to-peak linewidth for the two overlapping stacks of D M T M ( T C N Q ) 2 for rotation about b-axis at 77 K 38 4.20 Peak-to-peak linewidth for the two overlapping stacks of D M T M ( T C N Q ) 2 for rotation about c-axis at 77 K 39 viii A C K N O W L E D G E M E N T S I am immensely grateful to my supervisor, Dr. C. F. Schwerdtfeger for the super-vision, assistance and suggestions during the performance of the experiments and the preparation of the thesis. I thank Dr. R.R. Parsons who read the thesis for his suggestions and comments. I have benefited from many discussions with Dr. F. Cabanas and C. Ma; they were also helpful in many aspects of computer data analysis. The samples for this research were provided by Dr. R. J. Visser and Dr. S. Oostra of the the Univer-sity of Groningen. I wish to thank the Kenya Government and CIDA (Canadian International Development Agency) for awarding the scholarship that enabled me to embark on the study program leading to this thesis. Research for this thesis was funded by CIDA and the Natural Sciences and Engineering Research Council of Canada through grants to Dr. Schwerdtfeger. My wonderful family and dear friends have been a source of unfaltering encour-agement to me during the course of my study program; I owe immeasurable debt of gratitude to them all. ix Chapter 1 I N T R O D U C T I O N 1.1 Overview The quasi-one-dimensional morpholinium tetracynoquinodymethane X Y M ( T C N Q ) 2 compounds have been the subject of intense research [1] attributable to the inter-esting electron Peierls and spin Peierls transitions and level-crossing interactions [2] which have been observed in these compounds. It was also anticipated that these compounds might open the way to a successful synthesis of the much sought-after high T c superconductors especially after a thorough introduction by Shchegolev [3]. In the symbolism quoted X and Y stand for hydrogen or simple alkyl groups whereas M stands for morpholinium (as depicted in Fig.1.2). The significant characteristic that leads to the above phenomena is the quasi-one-dimensional behaviour of the electrical conductivity which is so highly anisotropic [3]. The columns of T C N Q molecules are weakly connected with each other. The intermolecular spacing within the columns are shorter than the usual Van der Waals distances leading to overlapping of molecular orbitals and to an electronic conduc-tivity in the direction along the columns. This stacking of the acceptor planar tetracynoquinodimethane (TCNQ) molecules is such that the molecular planes are parallel to each other. In general, the stacking direction is not perpendicular to the plane of the T N C Q molecules. The large variations in the possible stacking arrangements arise from the wide variety of donor molecules. When X = C H 3 and Y = C 2 H 5 in the morpholinium molecule the result gives Methyl-ethyl Morpholinium (MEM); whereas if X = Y = C 2 H 5 we have Diethyl-morpholinium (DEM). With X = Y = C H 3 we have Dimethyl-Morpholinium (DMM) and further with X=H and Y = C H 3 the molecule is H M M . If Oxygen is replaced with Sulphur 1 2 Figure 1.2: N - X , N-Y-MorphoUnium + (XYM+). The X and Y bonds can stand for hydrogen, ethyl or methyl groups. 3 Figure 1.3: Dimethyl-thio-Morpholinium molecule, with Sulphur then we have Dimethyl-thio-Morpholinium (DMTM) molecule. Crystal Synthesis The morpholinium derivatives have all been prepared as XY(T)M-iodides [6]. Crystals of their charge transfer complexes with TCNQ were grown by slowly cooling hot solutions of XY(T)M.I and either Li.TCNQ or neutral TCNQ in acetonitride. The former method yields a simple salt, whereas in the latter case TCNQ" is partially reduced to give a complex salt. The crystals are mostly lath-or needle-shaped and elongated along the TCNQ stacking axis. Their colour varies from reddish purple for the simple salts to dark violet or black for the complex salts. The chemical equation for the synthesis is given by M++I- + 2TCNQ => M+(TCNQ); + \I2 4 1.2 Previous Investigation of the D M T M ( T C N Q ) 2 D M T M ( T C N Q ) 2 is an example of a crystal with inequivalent stacks [6]. The in-equivalence of the stacks corresponds with an inequivalence of T C N Q sheets where each sheet is built up from translationally or symmetrically related stacks. It has monoclinic structure ? 2 J m (a = 7.845A,6 = 26.948.4,c = 7.91lA;a = 90°,(3 = 55.42° ,7 = 90°); meaning that it has a 2-fold rotation symmetry screw-axis above the transition temperature of 272 K. The crystal structure as studied by Visser [6] was done by X-ray diffraction. On cooling the sample, he reported a phase transi-tion at 272 K accompanied by 10-fold increase in electrical conductivity and a drop in activation energy from 0.25 to 0.03 eV. He found a drop in magnetic susceptibil-ity of 15% at the phase transition. However, his low temperature susceptibility was at variance with the Bonner-Fisher behaviour [16] characteristic of most strongly dimerized salts. At the transition twinning starts to take place. The phenomenon is discussed further in Section 1.3. Changes in Cell Parameters The transition does not drastically affect the unit cell (a,c) plane; hence the T C N Q sheets do not undergo essential geometric changes. The length of b also remains almost the same. The essential change is the lowering of symmetry from mono-clinic to triclinic where the b-axis, which is the principal axis, is rotated relative to the (a,c) plane by about 10°. The loss of the monoclinic symmetry causes the disappearance of the mirror planes and the 2j axis. Successive sheets are no longer equivalent crystallographically and have been shifted relative to each other. As a consequence of the different surroundings after the transition the T C N Q sheets are subject to different crystal potentials. When the effective Coulomb repulsion U is large the dimerization of T C N Q stacks gives rise to full "valence" and empty "con-duction" bands [6] rendering the chain non-conducting. Shifts of the resulting bands result in a decrease in the energy gap between the valence band of one sheet and 5 the conduction band of the other. Visser [6] reports a drop in activation energy of between 0.25 and 0.03 eV. If this shift is greater than the gap of the separate chains electrons are transferred to the lowered conduction band creating partially filled conduction bands and resulting in considerable increase in electrical conductivity. The Space Groups P2l and P j , Pi The crystal exhibits space group P 2 l of the monoclinic Bravais lattice, above the transition, which implies that the system has 2-fold screw-axis perpendicular to the mirror plane . P denotes that the lattice is a primitive type. This is a symmetry operation that has a translation component whose effect on the pattern that repeats in one direction only is to produce a rotation through 180° around the axis combined with a translation of half of the repeat distance of the pattern parallel to the axis. Performing the translation twice has the effect equivalent to simple translation by one period of the pattern. The triclinic structure that the crystal acquires on going through the phase tran-sition is the Bravais lattice of minimum symmetry, the 2-fold axis and mirror planes of monoclinic structure having been lost. In the twinned phase D M T M ( T C N Q ) 2 has either P j or P i symmetry structure; earlier attempts by Visser [6] at full struc-tural determination in the new phase were rendered difficult by the transformation twinning at the inversion temperature T c . P j involves centre of inversion (1) such that every face (hkl) is repeated to an opposite face (hkl) to give a pinacoid, a pair of parallel faces [8]. In class 1, the faces (hkl) and (hkl) are unrelated so that in this class the general form contains just one face or the pedion. 1.3 Transformation Twinning of D M T M ( T C N Q ) 2 A twinned crystal contains two or more portions which are physically continuous with one another but which are in different crystallographic orientations. These orientations are symmetrically related to one another. Classification of twinning fall 6 Figure 1.4: Schematic representation of structure of D M T M and of the energy levels of T C N Q bands (a) above T c space group P 2 l , (b) below T c , space group Pj or Pi . The dashed arrows in the upper drawings denote the average S—•N directions of the D M T M groups. A and B refer to inequivalent T C N Q sheets. The arrows between the electron bands in the lower drawings represent the smallest energy excitations. The drawings have been taken from reference [6j. 7 into categories depending on the processes involved in their formation. Examples are: simple growth, shear, transformation and nucleation twinning. Transformation Twinning There are many materials that have different stable crystalline forms (polymorphs) at high and low temperatures; the higher temperature form being of higher symme-try. If the difference in structure of the two forms is slight [8] they may transform easily and reversibly at the inversion temperature. On cooling the higher tempera-ture form through the inversion temperature, domains of the low temperature form may start to grow in different orientations in different parts of the crystal, these different orientations being related by symmetry of the higher temperature form. The domains are in twinned orientation relative to one another but they are along more or less haphazard boundaries depending on the accidents of where they started to form. If transformation started at many different centers simultaneously, there will be many domains. Such a transformation will not in itself lead to new faces on crystals which will initially have the same morphology as the higher symmetry polymorph, but if it continues to grow after transformation then it may develop new faces growing in different directions where different domains happen to emerge at the surface. The foregoing ties in well with our experimental observation on the cooling of D M T M ( T C N Q ) 2 through 272 K. At T c a second T C N Q ESR signal emerges and this would correspond to growth of domains in the twin. Also the transformation is reversible as recycling the crystal to room temperature several times yielded identical results. 8 1.4 The Hubbard Hamiltonian The magnetic properties of the T C N Q salts have been explained in terms of one-dimensional models. The Hubbard Hamiltonian [9] is written as ft = XjM+l(4+l,<r<V + cJ^ C i + i , , ) +  Vi-i ni nJ + 7T H ^ l t . - c r (1-1) where C i i < 7 is the destruction operator for an electron of spin o~ at the site i, rii^ = (clv ci,<r) I S the operator of the state rii — is the total number of electrons at the site i, titi+i is the transfer integral between the sites i and i + 1 [22], U is the interaction between electrons at the same site and Vn = V _ n is the interaction of electrons n nearest neighbour sites apart. The Hubbard Hamiltonian has been used as a starting point for models of the conduction electrons on the T C N Q stacks. The transfer integral t is allowed to vary along the stacks to accommodate dimerized or tetramerized stacks. In the Hamiltonian the ratio U/t is an important parameter [11] and determines the degree of localization of the electrons on the individual sites on the stacks. Huizinga [11] discusses the fact that spin Peierls transition is observed only if U/t ~ 4. For U/t ~ 4 the spin susceptibility, x looks much like that of antiferromagnetic exchange system without an observable change at the transition temperature. In the results of the present experiments given below, it will be noted that no spin Peierls transition was observed. It might well be that the on-site coulomb repulsion U is much smaller than the corresponding value for most other salts for which spin Peierls transitions have been found such as DEM(TCNQ )2 [2]. It is also noteworthy that for D M M ( T C N Q ) 2 , U/t is very small and this crystal has a very low temperature spin Peierls transition [24] at 1.6 K. 1.5 Thesis Outline The investigation that yielded the results for this thesis was started with the moti-vating background as outlined in Sections 1.2 and 1.3 of Chapter 1. The results of 9 Visser [6] and Oostfa [7] reported a remarkable tenfold increase in electrical conduc-tivity accompaned by decrease of activation energy from 0.25 to 0.03 eV on cooling through the phase transition of 272 K. They also reported a drop of magnetic susceptibility of 15% whereas the x(T) at lower temperature does not follow the Bonner-Fisher behaviour expected. D M T M ( T C N Q ) 2 is an example of a strongly dimerized salt with inequivalent stacks. Since Electron Spin Resonance (ESR) can observe the individual inequivalent T C N Q stacks the present investigation was un-dertaken to learn more about this interesting phase transition. Chapter 2 outlines the ESR instrumentation as well as the computer data ac-quisition and analysis whereby the ESR spectra are computer fit to Lorentzian lineshapes. Li:LiF with g-value of 2.002319 was used as a calibration sample. The TE102 rectangular cavity had fixed to it a gear arrangement with which the sample holder in the cavity could be rotated. In Chapter 3, the results of a temperature study of one of the crystals used is presented in the temperature range 298 K (room temperature) to ~ 2.2 K. The experiments were performed for two crystal orientations: one in which the ESR signals were overlapping at 77 K described in Section 3.1 and the other in which the signals were well separated at 77 K. In the latter orientation, the range covered was room temperature to about 200 K only as detailed in Section 3.2. The g tensor determination is presented in Chapter 4. The experimental proce-dure follows the method of Waller and Rogers. The principal values and direction cosines of the g tensor are determined. Chapter 5 outlines the discussion of the results, the conclusions and the sugges-tions for further investigations on D M T M ( T C N Q ) 2 . Finally, the names of some program files used in the data analysis are presented in Appendix A. Chapter 2 X - B A N D ESR A P P A R A T U S A N D M E A S U R E M E N T S The ESR experiments were performed at X-band using a rectangular cavity in the mode TE102 operating at a frequency of about 9.5 GHz in an homodyne arrange-ment. The ESR spectrometer itself was of reflection type, utilizing a circulator between the cavity and the waveguide systems. The microwave source was a Varian V-153/6315 reflex klystron whose reflector voltage was controlled by an Automatic Frequency Control (AFC) circuit operating at 25 kHz. The A F C stabilizes the fre-quency by phase locking to the cavity resonance. The frequency was measured using HP 5245L frequency counter with an HP 52559 frequency converter. The magnetic field sweep was calibrated using a S E N T E C type 1000 NMR Magnetometer. The reflected power change was detected using an HP 1N23H crystal diode of PIN type which was externally biased with current of 200-400/xA. Care was taken to tune the cavity for optimum sensitivity as well as adjusting phase controls of the Lock-in Detector to avoid baseline drift. For most experiments this phase was 60° at 77 K and 240° at room temperature. 2.1 Phase-Sensitive Detection The incident microwave power induces a voltage drop across the crystal detector and causes the circuit current to flow. The resulting impedance will determine the region of operation of the crystal. At microwatt powers the d.c. impedance is fairly constant and the rectified current is proportional to radio frequency power giving a square law detection of the crystal. In the milliwatt region the current becomes proportional to the square root of the microwave power and the crystal is then operating as a linear detector. Crystal detectors of ESR spectrometers usually 10 11 NMR MAGNETOMETER 1 , 1 r S C O P E K L Y S T R O N SUPPLY | K L Y S T R O N A . F . C . — F R E Q U E N C Y C O U N T E R t ISOLATOR VARIABLE A T T E N U A T O R 0-40 dB DIRECTIONAL C O U P L E R D . V . M . (THERMOCOUPLE VOLTAGE) VARIABLE A T T E N U A T O R 0-50 dB C R Y S T A L BIAS P R E -AMPLIFIER TUNED CRYSTAL D E T E C T O R 2.3 KHz MODULATION COILS MAGNET P O L E F A C E P L O T T E R A D / D A C O N V E R T E R MICRO C O M P U T E R L O C K -IN AMPLIFIER X - Y RECORDER M A G N E T I C F I E L D S W E E P Figure 2.5: Block diagram of X-band ESR apparatus. 12 STAINLESS STEEL TUBE BRASS FLANGE BRASS HOLDER FOR GEAR SYSTEM -SOFT SOLDERED TO CAVITY BRASS WAVE GUIDE SECTION BRASS BOTTOM PLATE Figure 2.6: T E 1 0 2 rectangular cavity with mechanism for rotation of sample holder. The drawings were taken from reference [19]. 13 operate in the transition region 10 - 5 to 10~4 watt power range. As the magnetic field was modulated at 2.3 kHz the microwave signal became amplitude modulated at this frequency during passage through resonance. At the crystal the microwave signal is demodulated and the ESR signal enters the receiver of the preamplifier as an intermediate frequency (i.f.) signal. In the process the crystal generates Johnson noise [18] that is proportional to the temperature and the bandwidth. To reduce the noise and improve the signal-to-noise ratio the Lock-in Detector or Phase-Sensitive Detector is employed which compares the ESR signal from the crystal with a reference signal and passes only the components of the former that have the proper frequency and phase. Since the reference voltage comes from the same oscillator that produces the field modulation voltage it ensures coherence and hence causes the ESR signal to pass through while noise is suppressed. A phase-shifter component of the Dynatrac Lock-in amplifier could be adjusted through a total of 270° to ensure that the modulation and reference signal arrived in phase at the Lock-in Detector . 2.2 Recorder Representation The input signal to the Lock-in unit was first amplified by a preamplifier (see Fig.2.5). In most of our experiments the time constant was set at 1.25 sec. This is a measure of the cut-off frequency of the filter component of the unit whose output gave signal to noise ratio of about 1000 to 1. If the time it took to scan through the magnetic field range AiTpp was very short compared to the time constant then a distorted signal would appear while if one waited many time constants to com-plete the scan the recorder would faithfully reproduce the true fine shape. The typical magnetic field scan was about 8 gauss at a sweep speed of approximately 2 gauss/minute. The data collection procedure used a program [5] written in BASIC and com-piled in BASICA Compiler and originally used with Q-band spectrometer. Slight 14 modifications were necessary to adapt the program for the X-band spectrometer data collection and analysis. It was also modified to be compilable with Turbo Basic Compiler in order to generate a more compact form as well as effecting faster subsequent editing. Some of the program files that were required in the analysis are given in Appendix A. 2.3 The Magnet The magnet used was 9—inch Magnion Model L-96 Laboratory Electromagnet which used the Magnion power supply Model HS-1365B. This power supply was designed to be highly regulated and to provide continuously variable d.c. power, 0.1 to 65 amperes at upto 130 volts into a 2 ohm load. It was originally designed to offer long-term ( 8 hours) stability of 1 part in 105 over the operating range of 10 to 65 amperes with peak transient line variation of 20% [20]. The magnet was water-cooled at the flow rate of about 4 gallons/minute and equipped with a water flow switch which would automatically de-energize the power in the event of interruption of the flow of cooling water. With the calibration sample and the T C N Q salt separated in the cavity a study of the fields of resonance was done as a function of distance of the cavity from the centre of the magnet as shown in Fig.2.7. This was used to give an estimate of the homogeneity of the field which was found to be satisfactory for our measurements. Placing the T C N Q sample on top of the LiF:Li ensured that they sat at the position of almost equal static magnetic field H Q . Fitting Data to a Lorentzian Lineshape The digitized ESR data were fitted to Lorentzian lineshape functions because these shapes have been reported for T C N Q salts [21]. The fit was performed by mini-mizing the sum of the square of deviations between data points and the functions 3402.00 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 S A M P L E D I S T A N C E F R O M C E N T R E O F M A G N E T I C F I E L D ( m m ) Figure 2.7: Calibration graph for H-field in the region of use for conical pole tips of radius of about 80 mm. 16 The susceptibilities of each individual curve were then determined by integrating the Lorentzian functions obtained from the fit. The corresponding peak-to-peak linewidths were calculated from the derivative of the functions derived from the fits. The most significant error in the susceptibility calculations arise from the de-termination of the baseline which was corrected for in the data collection program. Cabanas [5] reported distortions in data scans that did not originate from base line problems but that arose due to asymmetry in the derivative curves leading to inaccurate susceptibility determinations. In our case only slight manifestations of asymmetry were observed. Adjusting the coupling appropriately solved the prob-lem. 2.4 Low Temperature Operation For measurements at 77 K or lower the sample was housed in a double dewar system; actually it was found convenient to use the same arrangement even for room-temperature runs. For the g-tensor determination experiments the sample was secured on a teflon sample holder that could be rotated by means of a gear arrangement [19] shown in Fig.2.6. The worm gear part of the gear system was fixed to a stainless steel rod which was manually rotated from the top; this way the sample orientation with the static magnet field could be changed from 0° to 360°; however, it was sufficient to orient the crystal through 180° since rotating it through another 180° would lead to duplication of the data. To cool to 77 K, the outer dewar of the double dewar system held a steady amount of liquid nitrogen which thus acted as a bath to the inner dewar housing the cavity. For operation at liquid helium temperatures helium was contained in the inner dewar surrounded by, but thermally isolated from, the outer dewar of liquid nitrogen. The isolation from the outside of the cavity was effected by use of high vacuum between the walls of the inner dewar. 17 The procedure consisted of pre-cooling with liquid nitrogen in the outer dewar using helium gas in the inner dewar to provide thermal conduction with the cavity. This took at least one hour; liquid helium was then transferred from its storage dewar through a vacuum-insulated double-walled transfer tube. Bubbling of the liquid helium would lead to change of the dielectric constant. This was overcome by pumping over the surface of the refrigerant using a mechanical pump capable of obtaining as low a partial vacuum as 7 /xtorr which brought the liquid helium below the A-point. Chapter 3 T E M P E R A T U R E M E A S U R E M E N T S 3.1 The Temperature Measurements with Crystal Installed for Over-lapping Lines at 77 K In order to investigate the phase transition reported by Oostra et al. [7] ESR experiments were conducted in the temperature range 298 K (room-temp) to about 2.2 K. The ESR Spectrometer was calibrated using a sample of LiF:Li which has (7-value of 2.002319(2) [21]; the calibration sample was placed together with the T C N Q sample in the cavity. The experimental arrangement consisted of two parts: • Temperature range 298 K to about 200 K. • Temperature range 2.2 K to 77 K. The T C N Q sample was placed at the bottom of the cavity whereas the calibra-tion sample was positioned in the cavity in such a way that the ESR lines from the two were well resolved so as to permit detection of any changes in the T C N Q line. For this part of the experiment the LiF:Li was placed midway on one of the vertical walls of the cavity; care was taken in positioning so that the modulation field would still penetrate to the calibration sample. The design of the cavity had already incorporated that provision by way of a thinned portion of the vertical wall of the cavity. The Temperature Range 298 K to about 200 K For the higher temperature part dry-ice mixed with acetone was used as the coolant. The temperature was monitored through a copper/const ant an thermocouple and a 18 19 rnillivoltmeter. The reference point was obtained using an ice-bath, i.e. 0°C. The system was slowly cooled down from room temperature by controlled addition of dry ice to the acetone in the outer dewar (see Section 2.4). Data scans were recorded at about 5 minutes intervals while noting down the temperature corresponding to the T C N Q resonances. The linewidth variation with orientation of the crystal is very anisotropic as can seen in Figs.4.18, 4.19, and 4.20 where the linewidth variation with crystal orientation are given for three orthogonal directions at 77 K. The corresponding values at room temperature were found to be close to those at 77 K. The crystal was taken through the phase transition for two crystal orientations. At 272 K there was a remarkable sudden decrease in the peak-to-peak linewidth from 0.24 to 0.18 gauss as may be seen in Fig.3.8 below. In Section 3.2 the results of a second investigation of the phase transition that led to linewidth drop from 0.15 to 0.11 gauss are presented. On analysing the results, the magnetic susceptibility of D M T M ( T C N Q ) 2 was found to drop by 15% at this temperature as depicted in Fig.3.9. This agrees with the bulk susceptibility results of Oostra et al. [7j. Above and below the transition temperature the susceptibility variation with temperature obeys a power law. It was also noted that the transition is reversible as recycling the crystal to room temperature several times gave identical results. For the susceptibility and linewidth values the error estimation was determined by noting that fitting the lines to the Lorentzian lineshape involved an uncertainty in the linewidth of ±0 .01 gauss. Error bars could then be generated to indicate the estimated error. Low Temperature Measurements: From ~ 2.2 K to 77 K. Crystal in an Orientation of Overlapping Lines at 77 K At temperatures below 77 K the copper-constantan thermocouple was not sensitive enough. It was therefore replaced with a KP versus Au at 0.07% Fe (Chromel/AuFe Fig.3.8: Linewidth variation with temperature near the phase transition. The crystal was in an orientation of overlapping of the two TCNQ lines at 77 K. 7.00 I • 1 1 1 1 1 —I r 160.0 180.0 200.0 220.0 240.0 260.0 280.0 300.0 320. TEMPERATURE ( K ) Fig.3.9: Susceptibility variation with temperature near the phase transition. The crystal was in an orientation where the two TCNQ lines were overlapping at 77 K. 22 0.07%.Fe) [14]. Making such delicate thermocouple junctions required careful weld-ing; spot-welding was found to burn through the would-be junctions and was aban-doned. We used an oxy-hydrogen gas flame instead. The wires had a diameter of 0.005 inches and thus care was needed in handling them while making connections to the millivoltmeter. In our case it was the gold wire that was cut. To remove the teflon coating on it, the ends were abrased [15]. Similar wires usually have enamel coating that is removable by Strip-X enamel remover. Special solder and flux were also required to obtain reliable connections. The low temperature run was investigated by starting from about 2.2 K when system was initially cooled using liquid helium. It was then allowed to warm up (as the helium boiled off below the cavity) while taking data. Even after the helium completely evaporated, the temperature rose slowly enough to enable data to be taken at temperature intervals as short as 0.1 K. There were two lines at 2.2 K one for T C N Q and the other for the lithium calibration sample. The T C N Q signal had a peak-to-peak linewidth of 0.45 gauss at 2.2 K ; on warming to 4 K it had narrowed leading to the resolution of another line. This second line stayed at the same position with respect to the others; thus its g-value remained about constant. At ~68 K the line had merged up with the original one T C N Q line. The nature of this line remains unknown at the moment though we suspected it to be caused by impurities. Fig.3.10 shows a few plots of the derivative of absorption for the temperature range in question. There was a g-value difference of about 0.0004 for fixed orientation of external magnetic field with the crystal between data at 2.2 K and that at 77 K. Results of Schwerdtfeger et al. [4] on the study of D E M ( T C N Q ) 2 gave similar values for this difference. Our results gave the magnetic susceptibility variation with temperature as shown in Fig.3.11. This agrees favourably in the general shape with the theoretical analysis of Bonner and Fisher [16] for antiferromagnetic susceptibility versus temperature for an infinite Heisenberg ring except at low temperature. Nevertheless, the maximum 23 Figure 3.10: Plots of derivative of absorption curves to depict the remarkable increase in line width at low temperature. The crystal was in an orientation exhibiting overlapping lines at 77 K . The signal on the far left belongs to the lithium calibration sample. 60.00 50.00 to \— =5 40.00 >-or < or m < 30.00 CO ft 20.00 O tO to 10.00 i 1 1 1 1 1 1 r f 0.00 0.0 I • I i I I I i L i r i r J L 10.0 20.0 30.0 40.0 50.0 TEMPERATURE ( K ) 60.0 70.0 80.0 Fig.3.11: Susceptibility variation with temperature. The crystal was in an orientation exhibiting overlapping lines at 77 K. 25 susceptibility occurred at around 30 K whereas the results of Oostra et al. [7] gave a maximum at around 50 K where the susceptibility was 2.15 times that at T c . This is not consistent with what is expected for a dimerized chain and usually implies tetramerization; however, a spin Peierls transition was not observed. The susceptibility at low temperature thus tends to increase probably due to Curie impurities. This is a similar feature to that observed by Oostra et al. in their study of MEM(TCNQ) 2 where they suspected the impurities to be possibly solitons or nonmobile stacking faults [22]. 3.2 The Temperature Measurements with Crystal Installed for Near Maximum Separation of Lines at 77 K In the investigation of the previous section on the phase transition, it turned out that the lines were overlapping completely at 77 K; it was thus necessary to carry out the measurements on the crystal when so arranged as to exhibit optimum separation of the lines. This arrangement was made in order to study more about the transition region. As in Section 3.1, acetone and dry ice mixture were used to cool the sample to about 200 K. On cooling through the transition temperature there followed a sudden drop in susceptibility and the linewidth as before. An additional interesting observation was the clear abrupt growth of the second line. It started at the T c , 272 K, with little intensity and then grows while its position with respect to the initial one line remained essentially unchanged. Thus its g-value was constant. This growing seemed to be explained by the phenomenon of transformation twinning [8] as mentioned in Section 1.3. The domains of the low temperature polymorph were growing gradually as the twinning process took place. Fig.3.12 shows the susceptibility as a function of temperature near the phase transition, 272 K, at which the second stack signal "grows in" until about 250 K when the two signals had equal intensities. The "growing in" presumably occurs as the trans-formation twinning proceeds throughout the crystal. The dashed line of Fig.3.12 26 also shows a 15% drop in the spin susceptibility that agrees with bulk susceptibil-ity results of Oostra et al [7]. Although the total susceptibility follows the bulk susceptibility, the increase in the susceptibility of the "twinned" stack signal is ac-companied by a decrease in the susceptibility of the other stack signal as expected. Upon further cooling the spin susceptibilities of both stacks increase; however, the untwinned stack signal increases faster until at 77 K their intensities are in the ratio 2 to 1. This is attributed [17] to the different interactions along the stacks owing to possible differences in their stacking arrangements and confirms the inequivalence of the stacks. Such a difference has been observed in D E M ( T C N Q ) 2 [2]. No further comparisons can be made at this time because the complete X-ray determination for this crystal has not yet been made below the phase transition and hence differences in the stacking directions are not fully known. Fig.3.13 shows the linewidth as a function of temperature. Above the phase transition the one T C N Q signal has a linewidth of 0.15 gauss; at the transition the linewidth drops drastically to 0.11 gauss. At the same time the growing-in signal from the second stack shows a constant linewidth of 0.10 gauss in the temperature range of the study. As noted above, at 77 K the signals are in the ratio 2 to 1. It would be interesting to proceed down to low temperature and observe the behaviour of the two signals. This might enable a decision to be made on whether one signal has a spin Peierls transition and the other remaining visible to liquid helium temperatures as in the case of D E M ( T C N Q ) 2 [4] which also has two inequivalent stacks. 4.00 0.00 180.0 200.0 220.0 240.0 260.0 TEMPERATURE ( K ) 280.0 300.0 Fig.3.12: Temperature variation of susceptibility near the transition region. The crystal is positioned in a direction in which the 2 lines are well resolved at 77 K. * represents the untwinned stack signal and • represents the growing-in "twinned" stack signal. The dashed lined shows the 15% drop of total susceptibility at the phase transition. 0.16 0.06 180.0 200.0 220.0 240.0 260.0 TEMPERATURE ( K ) 280.0 300.0 Fig.3.13: Temperature variation of linewidth at the transition region.The crystal was positioned in direction in which the 2 lines were well resolved at 77 K. * represents the untwinned stack signal and • represents the growing-in "twinned" stack signal. Chapter 4 E X P E R I M E N T A L D E T E R M I N A T I O N O F T H E g T E N S O R To determine the g tensors it is important to bear in mind two main facts. First, the j-anisotropy is not always related trivially to the symmetry of the crystal. Sec-ondly, the coexistence of differently oriented paramagnets often makes the problem particularly challenging [12]. Generally, the procedure consists of finding the di-rections for which the g-value has the maximum and the minimum values. The third value would then be obtained in a direction perpendicular to the plane that contains the maximum and minimum values. After obtaining the orientations, the problem reduces to measuring the energies of the transitions between the Zeeman energy levels by aligning the magnetic field with each canonical orientation in turn. However, the present investigation followed a generalized method that is used in obtaining any tensors encountered in ESR work such as, D- and A-tensor which are the crystal field-splitting and hyperfine-coupling tensors respectively. 4.1 The Experiment and the Principal Values of the g Tensor The crystal was mounted on a lucite cube 2 mm on a side with the help of silicon grease. The cube was then placed at the center of a teflon sample holder [19] in such a way that when installed in the cavity the sample would be located at the position of maximum microwave magnetic field, Hj. To reduce dielectric losses the site of the samples was also to be one with minimum microwave electric-field E. The experiment was then done following the method of Waller and Rogers [25] where three orthogonal directions a, b, c of rotation were chosen. The rotation could be done with an accuracy of = 0.1°. 29 30 Rotation about b Rotation about c Figure 4 14- Initial crystal orientation for the 3 planes of rotation in the laboratory coordinate system. The cube held on to the sample holder measures 2 mm on a side. 31 The method is found to be generally applicable to orthorombic and monoclinic crystal structures. The present experiments on D M T M ( T C N Q ) 2 involves a triclinic structure at 77 K. This might appear at first sight to require a more general treat-ment than the orthorhombic case used for DEM(TCNQ) 2 or MEM(TCNQ) 2 [5]; but owing to the fact that the ESR signals correspond to unpaired electrons hopping along the T C N Q stacks, the same treatment is applicable. The measurements were taken by rotating the crystal about each of the chosen directions with the magnetic field fixed. Three rotations 0 a, #bi 8C corresponding to the orthogonal axes a, b, c were required to form a right-handed cyclical relationship such that the position 9a = 90° is equivalent to 9b = 0°, 9b = 90° is equivalent to 8C = 0°, and 8C = 90° is equivalent 9a = 0° as illustrated in Fig.4.14. If 8 is the angular coordinate spec-ifying the direction of the magnetic field in the plane of measurement then the parametrized form of g-value in the notation of Waller and Rogers becomes g] = oti + /3icos20i + -fisin29i (4.2) where i =a, b, or c. The quantities a, 8, 7 would be most accurately determined by obtaining the best fit of the above equation for g 2 to a large number of g-value measurements at different angles 9. It should be noted too, that should this method prove impractical then the a, 3, and 7 can alternatively be expressed in terms of 3 directly measurable quantities 2a = g\ + g2_ 23 = (g2+ - gl)cos20+ (4.3) 27 = (g+ -gl)sin26. which are the equations for the extrema obtained by differentiation of Eq.4.2 . The quantities g+, g- are the maximum and minimum g-values in the given plot and 6+ is the angle corresponding to g+; 8+ should be taken from the best extremum. The values a, 0, 7 for each of the orientations are then used to calculate the W tensor given by W = g 2 using the equations 32 Wn = aa + Pa W22 = ab + 0b W33 = ac + Bc Wu = aa-0a W22 = a b - B b W33 = ac - Bc (4.4) ^ 1 2 = ^ 2 1 = 7a ^ 2 3 = ^ 3 2 = 7 6 ^ 3 1 = Wl3 = 7 c The diagonal elements, W;;, are overdetermined. The method then involves an it-eration process to calculate the error, Si, in the azimuthal angles. The iterative solution starts with the following error functions A a = (a c - ab - Ba)ha A b = {aa - ac - p b ) / l b (4-5) A c = (ab - a a - Pc)hc Next, a new set of parameters for a, P, 7 are calculated as follows a'i = on 0i — piCosAi + jisinAi (4.6) 7; = jiCosAi — PiSinAi The calculation continues for new sets of A; until the values of the error functions become negligible. The starting angle shifts, Si, are obtained by continual summa-tion of Aj/2 throughout the iteration process. The final values of a,-, Pi and 7^  are then used to obtain the W tensor according to Eqs.4.4. Next, the principal values and the direction cosines are obtained by diagonalizing the W tensor. The principal values of the g tensor are then the square roots of the principal values of the W tensor. The angular variation of the g-values for the three orthogonal axes of rotation a, b, and c are presented in Figs.4.15, 4.16, and 4.17 respectively. The rotation about the c-axis gave two ESR lines that were difficult to resolve. This led to g-value plot that did not fit equation 4.2 as well as the corresponding data for the 33 Principal Values Dire a ction Cos b ines c 2.002455 -0.3952 0.1136 -0.9912 2.002971 -0.4062 -0.9117 0.0624 2.003554 -0.8239 0.3949 0.4065 Table 4.1: Principal values and direction cosines for the g tensor of Stack #1 of D M T M ( T C N Q ) 2 at 77 K. The error is ±0 .0001 . Principal Direction Cosines Values a b c 2.002329 0.2802 0.3756 0.8834 2.003033 -0.1485 -0.8922 0.4264 2.003421 -0.9484 0.2507 0.1942 Table 4.2: Principal values and direction cosines for the g tensor of Stack #2 of D M T M ( T C N Q ) 2 at 77 K. The error is ±0 .0001 . rotation about a and b axes. However, the g-value plot for rotation about c-axis (Fig. 4.17) uses a finer scale than those for a and b axes; thus g-value plot for c-axis rotation is flatter with difference in maximum and minimum g-values being about 0.0003. Furthermore, the convergence of the iteration process for the data due to rotation about c axis was an indication that the data was good enough. The three rotations gave principal values, presented in Tables 4.1 and 4.2, to an accuracy of ±0 .0001 . Linewidth Variation with Crystal Orientation For the three orthogonal directions, the linewidth variation was obtained and the plots of the values are given in Figs.4.18, 4.19, and 4.20. Owing to background broad lines, possibly impurity lines, the uncertainty in linewidth led to some scatter. The anisotropy characterristic of T C N Q linewidths is pronounced. A N G L E WITH H - F I E L D ( D E G ) Fig.4.15: g value for stacks #1 and #2 of D M T M ( T C N Q ) 2 for rotation about a-axis at 77 K. The solid line is a fit to Eq.4.2. The stack directions make an angle of 25° with respect to each other. 2.0032 2.0030 p 2.0028 2.0026 2.0024 2.0022 2.0020 0.0 j I i L I • I I I I I 1 1 1 1 L 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 ANGLE WITH H-FIELD (DEG) Fig.4.16: g value for the two overlapping stacks of D M T M ( T C N Q ) 2 for rotation about b-axis at 77 K. The solid line is a fit to Eq.4.2. 2.0035 2.0034 2.0033 h 2.0032 h CP 2.0031 2.0030 2.0029 2.0028 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180. A N G L E WITH H - F I E L D ( D E G . ) Fig.4.17: g value for the two overlapping stacks of D M T M ( T C N Q ) 2 for rotation about c-axis at 77 K. The solid line is a fit to Eq.4.2. 0.16 0.14 -in to < Q 0.12 JS I o »— I 0.10 -0.08 -0.06 O O • • o • • o o o o o o o o O <§> <§) <§> _L o • o4> o o o JL 0.0 20.0 40.0 60.0 80.0 100.0 120.0 ANGLE WITH H-FIELD (DEG.) 140.0 160.0 180.0 Fig.4.18: Peak-to-peak linewidth for stacks #1 and #2 of DMTM(TCNQ) 2 for rotation about a-axis at 77 K. The symbol • represents stack #1 and Q represents stack #2. 0.06 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 ANGLE WITH H-FIELD (DEG.) Fig.4.19: Peak-to-peak linewidth for the two overlapping stacks of D M T M ( T C N Q ) 2 for about b-axis at 77 K. 0.08 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 ANGLE WITH H-FIELD (DEG.) Fig.4.20: Peak-to-peak linewidth for the two overlapping stacks of D M T M ( T C N Q ) 2 for rotation about c-axis at 77 K. 40 4.2 Evidence of Level-crossing Effects In an effort to find out any crossing level effects, the ratio of intensities of the two lines corresponding to the two stacks were calculated for various angles made by the crystal with the magnetic field in the crystal direction giving two ESR lines for the T C N Q at 77 K. These were the same angular runs that were used to obtain the g-value plots given in Fig.4.15. With two distinct but closely related T C N Q stacks a level-crossing interaction would be expected to occur as has been detailed for DEM(TCNQ )2 [2]. However, owing to the fact that the stack directions are only 25° apart, as can be seen in Fig.4.15, the lines were never separated well enough to observe any repulsion at crossover. There was an indication of interaction in that the relative intensities of the lines changed near the crossover from a ratio of 2 to 1 to a ratio of 1 to 1. Chapter 5 C O N C L U S I O N S A N D S U G G E S T I O N S F O R F U R T H E R I N V E S T I G A T I O N S All the crystals investigated indicated that there was one ESR signal at room tem-perature. This was true in both X-band and Q-band results. There is a pronounced anisotropy in the linewidth as is typical of most T C N Q salts. The phase transition was studied for two different orientations of the crystal and in both cases there was a 15% decrease in susceptibility at T c . In one of these orientations the trans-formational phase transition in D M T M ( T C N Q ) 2 at 272 K was very remarkable in that the ESR line corresponding to the "untwinned" stack that "grows in " does so very abruptly. However, the cooling of the crystal in the orientation exhibiting the pronounced growth of the second line was not taken below about 200 K. Following the two lines shown in Fig.3.13 from the transition temperature re-vealed that while they have equal intensities at 250 K, on cooling to 77 K the untwinned stack signal increased to the ratio of 2 to 1 in intensity. This fact con-firms the inequivalence of the stacks, and it is attributed to the different interactions along the stacks owing to possible differences in their stacking arrangements. This kind of difference is not unique as it has been observed in D E M ( T C N Q ) 2 [2] and for D M T M ( T C N Q ) 2 such a difference might be explained only after complete X-ray determination below the phase transition is available. The preliminary results of an X-ray determination at 264 K [6] reveal a change in crystal symmetry and unit cell dimensions which suggested an ordering of the D M T M cations and a shift of the T C N Q sheets relative to each other. T C N Q complexes with closely related stacks would be expected to exhibit level-crossing interaction as has been detailed in D E M ( T C N Q ) 2 [2]; however, in D M T M ( T C N Q ) 2 41 42 the stack directions are only about 25° apart, the lines were never separated well enough to observe any repulsion at crossover. There was an indication of an inter-action at crossover in the ratio of intensities of the two signals. At 77 K the ratio of intensties was 2 to 1 in directions of large separation of the signals and it decreased to 1 to 1 in the direction corresponding to crossover of the lines. The shapes of the susceptibility curves versus temperature are similar to those for Bonner-Fisher antiferromagnetic chains [16] except at low temperatures. This is not consistent with what is expected for a dimerized chain and would usually imply tetramerization; however, a spin Peierls transition was not observed. The total susceptibility resembled the bulk measurements except that the maximum occurred at about 30 K instead of about 50 K [22]. This may be due to a broad background signal which was observed between 4 K and 68 K. The origin of this line was not determined but it is believed to be an impurity signal although it is possible that broadening caused by the random orientation of the crystallites could be the cause as Korving et al. [26] discusses for the case of D M M ( T C N Q ) 2 . Using the method of Waller and Rogers for the g-tensor determination the prin-cipal g-values are gi = 2.0034, g2 = 2.0030 and g3 = 2.0024 with an accuracy of ±0.0001 due mainly to the overlapping of the lines. The values are typical for T C N Q salts. Case of a symmetrically similar salt D M M ( T C N Q ) 2 Also observed was a Curie-Weiss law behaviour of the susceptibility below 4 K. The similar phemomenon observed by Korving et al. [26] in triclinic D M M ( T C N Q ) 2 would suggest that the behaviour in our crystal might be due to either an impurity or to the DMTM-chains of the triclinic D M T M ( T C N Q ) 2 becoming incommensurate with the lattice as X-ray studies reveal for DMM-chains [27]. This might cause a distribution of exchange values along the T C N Q chain. It is known that such a distribution of J-values can give rise to an increase of the susceptibility at low temperatures according to Clark et al. [28]. This might also explain the Curie-like 43 behaviour at the lowest temperatures. 5.1 Suggestions A more detailed study at low temperatures in the orientation that lead to obser-vation of a growing signal at the phase transition ( depicted in Fig.3.13) would reveal more of the character of the lines. It would appear that the two lines behave differently [2] leading to a possibility of one of the lines displaying a spin Peierls transition while the other does not but instead remains observable down to low temperatures. A complete X-ray determination below the phase transition is highly called for to study more about the possible differences in charges on the inequivalent stacks. The differences also in stacking arrangements would be unravelled by such a determination. It would be useful to obtain exact calculations that would enable determination of the important ratio U/t, U being the on-site repulsion and t the transfer integral, for it is an indicator to the possible occurence of a spin Peierls transition as mentioned in Section 1.4. Purer crystals would help in the clarification of the actual behaviour of the total susceptibility as they might be free from the suspected impurities that gave the broad background signals. Finally, cooling the crystal to much lower temperatures than attainable in the present experimental investigation might reveal some lower temperature transitions that seem to be signalled by the remarkable broadening of the curves at liquid helium temperatures. This correlation is suggested by Schwerdtfeger et al. [24] based on their study of DMM(TCNQ) 2 and D E M ( T C N Q ) 2 . Bibliography [1] HUIZINGA S., K O M M A N D E U R J. , SAWATZKY G.A., T H O L E B.T. , KOPINGA K., D E JONG W . J . M . and ROOS J. , Phys. Rev. B19, 4723 (1979). [2] C A B A N A S F .X. and S C H W E R D T F E G E R C.F. , Phys. Rev. B39, 11241 (1989). [3] S H C H E G O L E V I.F., Phys. Stat. Sol. A12, 9 (1972) [4] S C H W E R D T F E G E R C.F. , W A G N E R H.J. and SAWATZKY G.A., Solid State Comm. 35, 7 (1980). [5] C A B A N A S F.X. , " A study of Level Crossing Effects in T C N Q Salts" Ph.D. Thesis, University of British Columbia, 1988. [6] VISSER R.J.J. , " T C N Q Complexes with Morpholinium Type Cations" Ph.D. Thesis, University of Groningen, 1985. [7] OOSTRA S., D E B O E R J . , and D E L A N G E P., J . Physique Colloq.C3, 1387 (1983). [8] W H I T T A K E R E.J.W., "Crystallography. An Introduction for Earth Science ( and other Solid State) Students", Pergamon International Library (1981). [9] H U B B A R D J . , Phys. Rev. 17, 494 (1977). [10] OOSTRA S., "Electric and Magnetic Properties of the Morpholinium ( T C N Q ) 2 Family" , Ph.D. dissertation, University of Groningen, 1984. 44 45 [11] HUIZINGA S., "2fcF and 4fcF, A Study of Spin and Electronic Peierls Transitions in One-Dimensinal T C N Q salts", Ph.D. dissertation, Uni-versity of Groningen, 1980. [12] M A C M I L L A N J.A., "Electron Paramagnetism", Reinhold Book Corporation, (1968). [13] B L E A N E Y B. and STEVENS K . W . H . , Rep. Progr. Phys. 16, 108 (1953). [14] SPARKS L. and P O W E L L R., Journal of Research of the National Bureau of Standards, 76A, 263 (1972). [15] C A R O L A N J.F. Private Communication. [16] B O N N E R J.C. and FISHER M . E . , Phys. Rev. 135, A640 (1964). [17] KIRUI J.K., M A C.L. , WEIH D., S C H W E R D T F E G E R C.F. , Solid State Comm. (in press). [18] P O O L E C P . , "Electron Spin Resonance : A Comprehensive Treatise on Experimental Techniques " John Wiley & Sons, Inc. (1983). [19] K E R R R.K., Ph.D. Thesis University of British Columbia, 1971. [20] MAGNION, L-96 Laboratory Electromagnet Instruction Manual. [21] PRESSLEY R.J. and B E R K H.L., Bull. American Phys. Soc. 8, 345 (1963). [22] OOSTRA S., D E L A N G E P., and VISSER R.J.J., J.Phys.(France), Col-loq.,C3, 1383 (1983). [23] W E R N E R H. -P., SCHUTZ J.U.von, W O L F H.C. and K R E M E R R.K., G E H R K E M . and A U M U L L E R A., E R K P., HUNIG S., Sol. Stat. comm. 69, 1127 (1989). 46 [24] S C H W E R D T F E G E R C .F. , OOSTRA S., VISSER R.J. , and SAWATZKY G.A., Solid State Comm. 39, 1133 (1981). [25] W A L L E R W . G . and ROGERS M.T . , Journal of Magnetic Resonance 9, 92 (1973). [26] KORVING W.H. , H U M A N S T.W. , B R O M H.B., OOSTRA S., SAWATZKY G.A., and K O M M A N D E U R J. , J.Phsique Colloq., C 3 (1983). [27] VISSER R.R.J, and D E B O E R J.L. , J. Physique Colloq., C 3 , (1983). [28] C L A R K W . G . , "Physics in one Dimension ", Editors BERNASCONI J. and SCHNEIDER T. , Springer (1980) 289, and references therein. Appendix A Program Files used in the Analysis of Data A . l The Main Programs ESR9.BAS: To collect data and manually fit to Lorentzian lineshapes by Cabanas [6]. ESR9D.BAS: Modified version of ESR9.BAS for X-band data collection and manual fitting. ESR9D3.BAS: Version of ESR9D.BAS that takes care of baseline shifts in results files (ESR.CHN). A.2 Auxiliary Programs ESRLIN.PEL: To obtain the linewidths versus temperature or crystal orientation of the various fines separately. ESRSUS.BAS: To obtain the susceptibities of T C N Q lines normalized to Lithium. ESRCOR.COR: To calculate the correct g-values using the results file ESR.RES. The original ESR program in mainframe had no value for frequency and hence wrong g values were put in ESR.RES. 47 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0085016/manifest

Comment

Related Items