Current Projects of the CSU High Pressure Physics Group
Acknowledgements: The research is supported by the National Science Foundation under Award no. DMR-93-11772 "Study of structure-physical properties relation of new materials with reduced dimensionality"
Furthermore we acknowledge support of NATO Scientific and Environmental Affairs Division Award no. SA(PST.CLG.978929)6993/FP, Collaborative Linkage Grant "Optical Investigations of Single-wall Carbon Nanotubes under High Pressure"
1. High pressure-low temperature optical studies of the adamantine semiconductors with the chemical formula AIIB2IIIC4IV
Contrary to the chalcopyrite (CP) materials with the chemical formula AIBIIIC2VI the ternary adamantine semiconductors with the chemical formula AIIB2IIIC4VI crystallize in the defect chalcopyrite (DC) tetragonal or in the defect famatinite (DF) structure. Furthermore, the DC and DF structures contain a crystallographically ordered array of vacancies (stoichiometric voids or vacancies) in the cation sublattice, causing a low packing density of the materials. DC semiconductors have a high potential for optoelectronic applications due to their high nonlinear susceptibility, optical activity, intense luminescence, and high photosensitivity. In particular tunable filters on CdGa2S4 and UV photodetectors on CdAl2S4 are already used as devices 1,2.
Pressure induced phase transitions in AIBIIIC2VI chalcopyrites have been studied intensively 3,4 . However, the number of high pressure studies on defective AIIB2IIIC4VI compounds is rather small 3-7. It has been pointed out 4, that further investigations of DC and DF materials under pressure could be of interest for establishing a general systematic of pressure induced phase transitions in these materials.
It has been demonstrated that Raman scattering is an excellent tool to study the sequence of pressure induced phase transitions in these materials 3,4. When pressure is applied both the ordered DC and partially disordered DF structure undergo an order-disorder phase transition in the cation sublattice. This transition occurs in two stages as predicted by Bernhard and Zunger 8. In the first stage A and B cations substitute each other leading to a partially cation disordered phase. In the second stage mutual disorder of both cations and vacancies is established, resulting in a zinc blende-like disordered phase. If pressure is increase further this structure transforms then in a high symmetry rocksalt like structure.
As Ursaki et al. 4 have shown, also compounds with the partially disordered DF structure undergo a two stage transition to the totally disordered state. For example in ZnGa2S4 and ZnGa2Se4 which crystallize in the DF structure the first stage involves the ordered Ga ions at the origin sites and Zn and Ga cations become totally disordered. As in the DC structures stoichiometric vacancies are involved in the process of disorder in the second stage. The phase transition sequence is theoretically well understood and one may ask why one should undertake further high pressure studies of these materials.
The answer to this question is that we would add to the above mentioned general systematic by extending the high pressure studies to HgAl2Se4 and ZnAl2Se4 which to the best of our knowledge have not been studied before under high pressure. Furthermore, all high pressure studies have been done at ambient temperatures. We propose to do these studies also at low temperatures (77 K and 40 K).
As mentioned above the disorder of both types of cations and vacancies leads to the formation of a disordered zinc blende (ZB)-like phase . In this phase there are large anharmonic contributions to the lattice vibrations present. The low temperature measurements would allow to study these anharmonic contributions in more detail 9.
Finally we propose high pressure absorption and luminescence measurements on these materials at various temperatures to obtain detailed information about the pressure dependence of the band gap in these materials. This will in turn help to interpret the pressure dependence of the intensity and the full width at half maximum of the observed Raman modes which have been discussed qualitatively in refs. 3 and 4.
Project in collaboration with:
Dr. Mathias Schubert
and his group of the Institut fuer Experimentelle Physik II at the University of Leipzig, Germany and Dr. Andreas Eifler , Infineon Technologies in Dresden
2. High pressure X-ray diffraction and high pressure-low temperature transport measurements of the quasi-one-dimensional sulfides InV6S8 and TlxV6S8 0.1 # x # 1
The quasi-one-dimensional ternary compounds TlxV6S8 and InV6S8 which we want to study have a structure based on a hexagonal cell of Nb3Te4 -type (space group P63 (C66 )) 1 . The structure is composed of VS6 octahedrons linked together by shared edges and faces to form large hexagonal channels running parallel to the c-axis 1,2. The structure is shown in Fig. 1. The Tl ions are statically disordered within the hexagonal channels 3.

a) b)
Fig 1 Quasi-one dimensional structure based on a hexagonal cell of Nb3Te4 type
a) VS6 octahedrons linked together by shared edges and faces to form large hexagonal channels along z-axis. Tl ions form one dimensional chains inside the channels.
b) V-V zigzag chains along z-axis.
The vanadium ions form V-V zigzag chains running along the c-axis (z-axis in Fig 1) with an intrachain V-V distance of 2.864
D and an interchain distance of 3.150 D 2. These values show that the chains are well separated from each other resulting in a highly anisotropic conduction.Bensch et al. 4 have also observed in resistivity measurements, that TlxV6S8 showed superconductivity at 4.4, 3.7, and 3.7 K for x = 0.8, 0.1, and 0, respectively. However, the superconductivity in these compounds has not yet been confirmed by the Meissner effect. Ohtani et al. 5 have studied the quasi-one-dimensional sulfides AV6S8 ( A = In, Tl, K, Rb, Cs) and confirmed the existence of superconductivity in these materials at Tc (onset) = 3.9, 4.1, 0.71, 0.65 and 0.32 K for A in the order above by performing AC magnetic susceptibility measurements.
Most interesting was the observation of an anomaly in the resistivity upon cooling at about 160 K, which Ohtani et al. 5 suggested to be caused by a charge density (CDW) instability. Such a CDW instability which inherently conflicts with superconductivity has also been observed in the isostructural Nb3Te4 6 and InxNb3Te4 7.
The suggestion of the existence of a CDW instability, which our preliminary band structure calculations on InV6S8 have confirmed, motivated us to start a high pressure study in the temperature range from 300 K to 100 mK of TlxV6S8 and InV6S8 which show the largest anomaly in the resistivity with the objective to study the coexistence and competition of superconductivity and a CDW instability. Furthermore, our studies of the spin density wave state in chromium alloys 8,9 have shown that high pressure is an excellent tool to study such instabilities. Though the commensuration of the wave vector of the instability can be changed by the alloy content, the application of high pressure allows to do this in a more controlled way 8.9. Boswqell and Bennet 10 have also claimed that the q vectors become incommensurate in single crystals of InxNb3Te4 as the In content increases. This provides further motivation of a high pressure study of these materials.
As mentioned above our preliminary band structure calculations on InV6S8 have confirmed the existence of a charge density wave instability.. In Fig. 2 we show the Brillouin zone in order to make Fig 3 better understandable which shows the nesting conditions are fulfilled for at least for different bands.

Fig. 2 Brillouin zone of InV6S8 (taken from ref. 13)

Fig. 3 Fermi surfaces of 5 bands which cross the (taken from ref. 17) Fermi level. 4 ΓKHA planes are shown for clarification. At least 4 prominent nestings are seen (each color corresponds to a different band)
First principle calculations on InV6S8 have run into similar problems as the calculations for Cs2MoS4, since the close packing is again missing in the structure. The calculations carried out so far indicate a much larger dispersion of bands along the c-axis direction. Though the band dispersions are much less in the planes normal to the c-axis they are not negligible, and hence the conductivity properties of InV6S8 can be said to be predominantly quasi-one-dimensional. The next step will be now to verify that the convergence criteria achieved in these calculations are adequate by using computers with a larger RAM. Nevertheless the basic nature of the dispersion of the bands and their Fermi surface nesting characteristics are unlikely to change. It is well known that a Fermi surface nesting can lead to instabilities like a charge density wave 11 or a spin density wave 8,9.
Finally it should be noted that problems arising from disorder effects (like at the In or Tl sites in InxV6S8 and TlxV6S8) could be treated by the virtual crystal approximation 12,13, which is known to be reasonable for broad-band systems. More appropriate methods, like coherent potential approximation 12,14 to treat the disorder effects self consistently exist in the first principles framework, but need more than an order of magnitude higher computational facilities than presently available here. Nevertheless we are confident that we can overcome this problem in cooperation with the MPI in Dresden.
Resistivity measurements at ambient and high pressure
In Figs. 4, 5, 6, and 7 we show the results of our first measurements at ambient pressure. As suggested by Ohtani et al.5 and confirmed by our preliminary band structure calculations the anomaly at Th, defined as the minimum of δρ/δT , is attributed to a CDW instability. Using this definition we find a Th of 171 K independent of Tl concentration (Fig 4).
A clear thermal hysteresis in the resistivity versus temperature can be seen indicating a first order transition at Th. The insert in Fig. 5 shows that the superconducting transition of TlV6S8 occurs in two steps. Fujii et al.15 have shown that the high temperature step is due to intra-grain and the lower one to inter-grain superconductivity. Interesting is also that Tl0.47V6S8 does not show the onset of superconductivity down to 1.8 K (Insert Fig. 6) . This may be related to the fact that Tl0.47V6S8 shows a remarkably enhanced anomaly at Th as shown in Fig 6.
We also observe on overall increase of
ρ(T) after after successiv We also observe on overall increase of ρ(T) ig. e thermal cycling while Th = 171.5 K stays constant. The jump Δρ increases from 32% (1st run) to 36% (4th run). Furthermore, we observe only for this Tl concentration an anomaly at Th in the specific heat during warming up. The anomaly is reduced after each thermal cycling and does not change anymore after the 3rd run (Fig. 7).
Fig. 4 Ambient pressure resistivity of Fig. 5 Ambient pressure resistivity of TlxV6S8 showing the anomaly TlxV6S8 showing the onset of at Th about 160 K superconductivity around 4 K except Tl0.47V6S8. Insert: TlV6S8-two step
transition 15.

Fig. 5 Ambient pressure resistivity of TlxV6S8 showing the onset of superconductivity around 4 K except Tl0.47V6S8. Insert TlV6S8-two step transition 15.

Fig. 6 Resistivity versus temperature of Tl0.47V6S8

Fig.7. Specific heat versus temperature
of Tl0.47V6S8
Fig. 8 Resistivity versus temperature at various constant pressures

Fig. 9. Pressure dependence of Th and Tc
Figs. 4-9 are based on measurements of Corneliu Miclea, a graduate student at the MPI-CPFS in Dresden
Fig. 8 shows the resistivity versus pressure of Tl0.15V6S8 . It can be seen that the CDW transition is suppressed at a critical pressure pc, 1 GPa < pc , 1.9 GPa . Above Th ρ increases with pressure. Below Th , ρ increases with pressure for p < pc. For p > pc, ρ decreases indicating that the CDW gap is closed.. In Fig. 9 we have plotted Th and the superconducting transition temperature, Tc, as function of pressure. Th decreases with increasing pressure while Tc remains almost constant.
Our objective is now to grow single crystals which can be better characterized and continue the measurements to see which influence the sample quality will have. First attempts of growing single crystals have been successful. Furthermore, we will do X-ray diffraction measurements to determine, if structural changes will occur and use the results as input to improve our band structure calculations at high pressure. To the best of our knowledge there is only one high pressure X-ray diffraction study 31 on
Nb3Te4 and InxNb3Te4 channel compounds, which showed that the compression along the c-axis involved the folding of the Nb-Nb zigzag chains. The compression perpendicular to c is entirely due to the reduction of the diameter of the channels. Furthermore the presence of intercalated In atoms was found to have hardly any influence on the compression behavior up to 40 GPa.
Project in collaboration with:
Dr. Guenter Sparn is the head of the High Pressure Competence Group at the Max Planck Institute for Chemical Physics of Solids in Dresden (MPI-CPFS), Germany.
Dr. Tsukio Ohtani , Professor at the Okayama University of Science, Laboratory of Solid State Chemistry, Okayama, Japan..
Also involved in this project are Dr. B. Godwal, Head of the High Pressure Physics Division , Dr. V. Vijayakumar, Mrs. Alka Garg (Experimental) , Dr. R. Rao , Dr. D. Gaitonde, Mr. P. Modak, and Mr. A. Verma (Theoretical) from the Bhabha Atomic Research Center (BARC) in Trombay, Bombay, India, where I have spent 2 months during my Sabbatical in 2002.
3. High pressure study of the metal-insulator transition of CuIr2S4
CuIr2S4 undergoes a metal - insulator (M-I) transition 1 on cooling at the transition temperature Tc of 230 K accompanied by a structural transition from cubic to tetragonal with a volume contraction of 0.7%. The magnetic susceptibility is paramagnetic in the metallic phase and decreases sharply and becomes diamagnetic at Tc on cooling. Magnetic measurements have revealed that the insulating phase is stabilized by high pressure. Several mechanism of the M-I transition have been discussed including the Jahn Teller effect because the transition from cubic to tetragonal in CuIr2S4 is similar to the Jahn teller distortion observed in a number of oxide spinel compounds. Furubayashi et al.1 showed, however, that the explanation is not fully convincing at the moment. Their other suggestion that the M-I transition is induced by the band Jahn Teller effect is supported by the band structure calculations of Oda et al.2., who found that the electronic band structure near EF is noticeably modified by the lattice distortion. They concluded that the strong electron-phonon interaction drives the structural phase transition in CuIr2S4. In most oxides that exhibit M-I transitions the metallic phase is stabilized by high pressure 3. Interestingly that is not true for CuIr2S4 where the insulating state is stabilized by high pressure 1,4.
The band structure calculations of Oda et al.2 could not predict the insulating electronic structure. They speculated, however, that the existence of some superstructure or charge density wave state could be responsible for the insulating behavior. Radaelli et al. 5 showed by performing X-ray and neutron diffraction measurements that CuIr2S4 undergoes a simultaneous charge-ordering and spin-dimerization transition in the insulating phase. This is quite interesting and novel, since such dimerization transitions have been reported to occur almost exclusively in compounds with clearly defined quasi-one-dimensional chains and not in three dimensional type structures 6. These results have been supported by the X-ray absorption measurements across the M-I transition by Croft et al. 7. The authors also pointed out that the M-I transition may involve electron localization due to correlation effects contrary to the common belief that 5d electrons form broad bands. The fact that the electrons have a localized character could also explain that the band structure calculations of Oda et al. 2 could not reproduce the insulating phase. With the resources of this collaborative project (resistivity and specific heat measurements at high pressure with the possibility to apply also a magnetic field) as well as optical measurements we would be able to understand the unusual behavior in the insulating phase even more, yielding an answer why the insulating phase is stabilized by high pressure contrary to the case of most oxides exhibiting a M-I transition, where the metallic phase is stabilized by high pressure 3.
Recently Andreev et al.8 have shown that the electrical conductivity of CuIr2S4 can be described in the temperature range from 4.2 – 50 K by the Mott variable range hopping model 9 , whereas in the range from 50 – 200 K the experimental data are best described in terms of a small polaron model 10. Our high pressure studies of conducting polymers, where the dc conductivity follows also the typical Mott variable range hopping behavior at low temperatures have shown that the application of high pressure can change the parameters which govern the hopping process 11,12 leading to a better understanding of the properties of the materials under high pressure. Since the external pressure can be replaced by internal or "chemical pressure 13 or by built in strain as in the case of semiconductor lasers 14, this understanding can lead to tailoring of materials for special applications.
4. Novel low-dimensional materials.
As mentioned in the introduction we have synthesized novel one- and two-dimensional materials which we expect to show interesting effects at high pressure, in particular pressure induced changes in the dimensionality and thus provide more insight and understanding in the mechanism responsible of these changes. One class of materials has a two dimensional structure and the chemical formula Na8M2II(M2IIIQ6)2 with MII = Sn, Pb, Eu, MIII = Si, Ge, and Q = S, Se, Te,. We expect as in the case of KTbP2Se6 1,2 a charge transfer and possible formation of Q-Q bonds when high pressure is applied. Since charge can be transferred from MII (for the Sn and Eu compound) or from MIII , we expect interesting effects and new insights concerning the conditions for the formation of Q-Q bonds at high pressures. In particular, the fact that we have the Pb-compound where charge transfer will not occur, because the oxidation state of Pb2+ is preferred, will provide information to understand possible charge transfer mechanisms in the other compounds. The structure of these materials is shown in Fig. 1.


a) b)
Fig. 1. Na8Eu2(Si2Se6)2 a) viewed parallel to b-axis b) viewed perpendicular to the Eu2(Si2Se6)8- layers, Na sites are omitted for clarity. Na sites are shaded blue, Eu green, Si red, Se yellow
The other interesting new material is trigonal NaEuGeS4 which contains one-dimensional tubes of EuGeS4 and is therefore structurally similar to another class of quasi-one dimensional materials with the chemical formula AxV6S8 with A = Tl and In and 0.05 $ x #1 , which we want to study. The structure is shown in Fig, 2. As mentioned above to the best of our knowledge there is only one high pressure study of such channel compounds 3. Therefore more high pressure studies seem warranted to obtain a better understanding of the effect of pressure on such compounds.

a) b)
Fig 2. NaEuGeS4 viewed a) along the c-axis, and b) perpendicular to EuGeS4 channels. Red polyhedra are GeS44- units, green polyhedra are EuS7 units, blue spheres are Na atoms, yellow spheres are S atoms.
Project in collaboration with:
Dr. Peter K. Dorhout, Professor of Chemistry in the Department of Chemistry at Colorado State University
Dr. Martin P. Gelfand , Assoc. Professor of Physics in the Physics Department atof Colorado State University
5. New device for the simultaneous application of hydrostatic pressure and uniaxial stress
We have fabricated now all the parts for a modified unique device for the simultaneous application of uniaxial stress and hydrostatic pressure 1. It is worth to mention that the development and fabrication of the new high pressure device for the simultaneous application of hydrostatic pressure and uniaxial stress would not have been possible without the massive financial and technical support of the MPI in Dresden, Germany,. This device will provide a unique new experimental technique for our collaborative research, because it will allow us to control parameters like lattice constant, bond length, and interaction strength in a way not possible before. It will allow us to apply nearly "pure" strain (deformation only in one direction) to anisotropic materials, and thus to tune dimensionality in low dimensional materials in a controlled way. The device allows now optical and conductivity measurements. A schematic drawing of the cell is shown in Fig. 1. Furthermore, the device is now designed in such a way that measurements at low temperature can be done. (we estimate about 40 K)

a)

b)
Fig. 1 New modular device for the simultaneous application of hydrostatic pressure and uniaxial stress at low temperatures.
a) exchangeable high pressure cell b) hydraulic part
The device is designed in a modular way so that the high pressure cell can be easily adapted to various other experimental techniques.
Finally, an important point must be addressed. One could argue, why using the new device for the simultaneous application of hydrostatic pressure and uniaxial stress to obtain pure strain dependence, when the same results could be obtained from uniaxial measurements alone, if sufficient knowledge of the elastic constants is available. Furthermore, constant strain measurements are only possible , if two of the axis have the same pressure dependence different from the one of the third axis, so that we can compensate for the Poisson ratio, limiting the use of the equipment. These are all valid arguments, and we are aware of it. We think, however, that theoretical models and calculations should be backed up by experimental evidence, and as we have shown in the case of semiconductor lasers 1, being able to change certain parameters independently from others, can lead to new and exciting insights and results. This fact cannot be emphasized enough , as the example of uniaxial stress measurements in YBa2Cu3O7-δ by Welp et al. 2 has shown. It turned out , that the small hydrostatic pressure dependence of Tc in YBa2Cu3O7-δ is due to a cancellation of large and opposite effects in the a-b plane. In any layered systems where one has strong couplings in the layer and weak coupling between the layers the newly developed device will be extremely useful, even if an exact compensation of the Poisson ratio will not be possible.
Project in collaboration with:
Dr. Guenter Sparn , Head of the High Pressure Section at the Max Planck Institute for Chemical Physics of Solids in Dresden (MPI-CPFS), Germany and Faculty Affiliate at the Department of Physics of Colorado State University.
Dr. Fernando Rodriguez, Professor of Physics at the Facultad de Ciencias, Universidad Cantabria, Spain
Dr. Mathias Schubert and his group of the Institut fuer Experimentelle Physik II at the University of Leipzig, Germany and Dr. Andreas Eifler , Infineon Technologies in Dresden
Dr. Guenter Sparn is the head of the High Pressure Competence Group at the Max Planck Institute for Chemical Physics of Solids in Dresden (MPI-CPFS), Germany.
Dr. Tsukio Ohtani , Professor at the Okayama University of Science, Laboratory of Solid State Chemistry, Okayama, Japan..
Dr. B. Godwal, Head of the High Pressure Physics Division , Dr. V. Vijayakumar, Mrs. Alka Garg (Experimental) , Dr. R. Rao , Dr. D. Gaitonde, Mr. P. Modak, and Mr. A. Verma (Theoretical) from the Bhabha Atomic Research Center (BARC) in Trombay, Bombay, India, where I have spent 2 months during my Sabbatical in 2002.
Dr. Peter K. Dorhout, Professor of Chemistry in the Department of Chemistry at Colorado State University, Department of Chemistry, Fort Collins, CO 80525
Dr. Martin P. Gelfand , Assoc. Professor of Physics in the Physics Department atof Colorado State University
Dr. Fernando Rodriguez, Professor of Physics at the Facultad de Ciencias, Universidad Cantabria, Spain
Dr. Elena Obraztsova, (Senior Research Scientist), Natural Sciences Center of General Physics Institute, 119991 Moscow, 38 Vavilov Street, Russia Project: "Optical Studies of Carbon Nanotubes under High Pressure"
Dr. Jefferey L. Yarger, Assoc. Prof. of Chemistry, Department of Chemistry, University of Wyoming, Laramie, WY 82071
top