Q. Sun, Q. Wang, T.M. Briere, V. Kumar, and Y. Kawazoe

Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan

P. Jena

Physics Department, Virginia Commonwealth University, Richmond, Virginia 23284-2000, U.S.A.

Email: sunq@imr.edu

Carbon and silicon are neighbors in that they are members of the same group in the periodic table. However, their chemical behaviors are quite different. Carbon atoms have very flexible bonding features, being able to form single, double and triple bonds with themselves and other atoms. Therefore, carbon systems can exhibit very rich structures: graphite, diamond, fullerenes, carbon nanotubes, amorphous carbon, porous carbon, graphite intercalation compounds (GIC), and so on. However, the larger number of core electrons in Si makes it much more difficult for two Si atoms to form double and triple bonds. Consequently, Si prefers to form multidirectional single bonds (sp3). Although silicon has great potential for applications in computer chips, micro-electronic devices, chemical catalysts, and new superconducting compounds, the spotlight has been shifted to its neighbor, carbon, since the discovery of C60 and carbon nanotubes. In turn, though, the rich structures and novel properties of fullerenes and carbon nanotubes have fired the imaginations of scientists over the world who have raced to find ways of making similar silicon fullerene structures [1]. To explore the possibility of nano-silicon systems with fullerene cage or tube structures, in this talk we will report our recent studies on nano-structures of silicon, which include: (1) metal stabilized Si20 cages, (2) interaction between two magic Si12W clusters; and (3) interactions of magic Si12W cluster with carbon nanotubes. We find that the Si20 cage can be stabilized by doping with metal atoms [2], and the charge distributions and HOMO-LUMO gaps can be easily tuned by changing the species of the metal atoms. Therefore, metal atom doping provides a new way to increase the number of variables for the purpose of nano-material design and control using Si.

[1] V. Kumar and Y. Kawazoe, Phys. Rev. Lett. 87, 045503 (2001).

[2] Q. Sun, Q. Wang, T. M.Briere, V. Kumar, Y. Kawazoe, and P. Jena, submitted for publication.

G. Vaitheeswaran and M Rajagopalan

Department of Physics, Anna University, Chennai 600025,India

and

Department of Physics, University of Bhopal, Bhopal 462026,India

Email: spsanyal@bom6.vsnl.net.in

The self-consistent band structures of BSb in B3 and B1 structures were obtained using the tight binding linear muffin-tin orbital method. The structural stabilities have been studied in this compound by calculating the total energies in the two structures. The ground state properties are calculated and compared with available results. The pressure at which BSb undergoes insulator to metal transition (IMT) and structural phase transition from B3 to B1 is predicted. From the band dispersion curves a possible reason for IMT is discussed.

Vaitheeswaran G., Kanchana V., Rajagopalan M.

Department of Physics, Anna University, Chennai-25, India

Email: gvaithee@hotmail.com

The present work employs the self-consistent tight binding linear muffin tin orbital method to calculate the relative stabilities of the high pressure Primitive tetragonal and CsCl structure in the series of compounds LaX (X=P, As, Sb, Bi). For compressed volumes these compounds are shown to favour primitive tetragonal phase rather than CsCl phase. This can be seen from the total energies of the two high-pressure phases in which primitive tetragonal phase is energetically minimum. The transition pressures are 18,11.2,8.6,11.2 GPa for LaP, LaAs, LaSb & LaBi respectively.

The bulk modulus in the B1 phase gradually decreases as one goes from Phosphorus to Bismuth, which is in agreement with that of experiment. Moreover the bulk modulus in the PT phase is of two orders of magnitude which may be due to the increased covalent nature in this phase. The equilibrium lattice parameter, bulk modulus is calculated for these systems and are compared with the available experimental and other theoretical results. The band structure and density of states are also calculated in the B1 and Primitive tetragonal phase and are compared with the earlier works.

M. Rajagopalan, G. Vaitheeswaran, and V. Kanchana

Department of Physics, Anna University, Chennai-25, India

Email: gvaithee@hotmail.com

The present work employs the self-consistent tight binding linear muffin tin orbital method to calculate the relative stabilities of the high pressure Primitive tetragonal and CsCl structure in the series of compounds LaX(X=P,As,Sb,Bi).

For compressed volumes these compounds are shown to favour primitive tetragonal phase rather than CsCl phase. This can be seen from the total energies of the two high pressure phases in which primitive tetragonal phase is energitically minimum. The transition pressures are 18,11.2,8.6,11.2 GPa for LaP,LaAs, LaSb & LaBi respectively. The bulk modulus in the B1 phase gradually decreases as one goes from Phosporus to Bismuth which is in agreement with that of experiment. Moreover the bulk modulus in the PT phase is of two orders of magnitude which may be due to the increased covalent nature in this phase. The equilibrium lattice parameter, bulk modulus are calculated for these systems and are compared with the available experimental and other theoretical results. The band structure and density of states are also calculated in the B1 and Primitive tetragonal phase and are compared with the earlier works.

Dept of Chemical Sciences, Tezpur University, Napaam, Tezpur -794 018 Assam, India

Email: ramesh@tezu.ernet.in

Lewis acidity of alkali cation-exchanged zeolite is studies using local reactivity descriptors based on hard-soft acid-base concept (HSAB). The local softness for nucleophilic attack (sk+), local softness for electrophilic attack (sk-) and their ratio, which is called 'relative electrophilicity (sk+/sk-) are important paramters in predictingreactivity of zeolites. The Lewis acidity trend derived from relative electrophilicity values decreases in the order: Li+ > Na+ > K+ > Rb+ > Cs+/ This trend of Lewis acidity is confirmed by studying the interaction of CO molecule with the cationic sites. The calculated blue shift of CO vibrational frequency and the interaction energy of he CO molecule with the alkali cation-exchanged zeolite clusters follow the above Lewis acidity trend. An interesting correlation between frequency shift and sk+/sk- is observed.

L. M. Ramaniah(1), S. M. Sharma(1), K. Kunc(2) and N.Garg(1)

1 Bhabha Atomic Research Centre, 7foombay, Mumbai.

2CNRS and University ofP. and M. Curie, Paris.

Email: lavanya@apsara.barc.ernet.in

Berlinite, a-AIPO4, belongs to a class of iso-structural compounds MXO₄ (M=Si,Ge,AI,Ga.. and X=Si,Ge,P,As...) which crystallise, at ambient pressures, in the trigonal phase (space group P3_l2 1), adopting structures made up of MO₄ and XO₄ tetrahedra. Several of these compounds are believed to amorphise at pressures of around 200 kbar. a-AIPO₄ had long been believed to undergo a reversible amorphisation transition at ~ 150 kbar, with both experimental evidence^l and theoretical support² for this widely-held view. However, some recent experiments³ have suggested that Berlinite transforms to a crystalline phase at these pressures. This was confirmed by recent synchrotron x- ray diffraction experiments⁴, which also identified the crystalline phase as cmcm. Recent molecular dynamics simulations⁵ using classical pair potentials, but a much larger simulation cell than used in previous simulations, have failed to reproduce a transition from the a to the cmcm phase. However, the total energy of the cmcm phase above 120 kbar was found to be lower than that of the, a phase.

In order to better understand and clarify the picture regarding the high-pressure behaviour of a-AIPO₄, we have undertaken an ab-initio study of this problem. In this first report, we present the results of the calculation of the static structural properties, including the equation of state in the two phases, a and cmcm, and their relative stabilities. We work within the framework of density-functional theory, using the generalised gradient corrections to the local density approximation⁶ exchange and correlation energy proposed by Becke⁷ and Lee, Yang and Parr⁸ (BLYP). Only valence electrons are treated explicitly, and the electron-ion interaction is described by norm-conserving pseudopotentials of the Troullier-Martins type⁹. The Kohn-Sham orbitals are expanded in plane-waves, using an energy cut-off of 70 Ryd. Most ground-state properties may be obtained from the total energy and its derivatives. Hence, the total energy of the crystal as a function of different supercell volumes is calculated for each phase, and the E(V) phase diagram is constructed. Optimisation of the atomic positions in the supercell is performed by employing ab-initio molecular dynamics^lo. A preliminary analysis of our results shows that the cmcm phase is stabler than the a phase at high pressures, in agreement with the recent experimental results.

References:

M. B. Kruger and R. Jeanloz, Science 249,647 (1990).

J. S. Tse and D. D. Klug, Science 255, 1559 (1992).

P. Gillet, J. Badro, B. Vasel, P.F. Macmillan, Phys. Rev. B 51, 11262 (1995).

S.M.Sharma, N. Garg and S.K. Sikka, Phys. Rev. B 62,8824 (2000).

N. Garg and S.M Sharma, J. Phys. Cond. Matt. 12,6683 (2000).

D. M. Ceperley and B. J. Alder, Phys. Rev. Lett. 45,566 (1980).

A.D. Becke, Phys. Rev. A 38,3098 (1988).

C. Lee, W. Yang, and R.G. Parr, Phys. Rev. B 37,785-789 (1988).

N. Thoullier and J.L. Martins, Phys. Rev. B 43, 1993 (1991).

CPMD Version 3.4.0, J. Hutter, A.Alavi, T .Deutsch, W .Silvestri,

Max-Planck-Institut fur Festkorperforschung and IBM Research Laboratory.

Arya, G.K. Dey, Vijay K. Vasudevan and S. Banerjee

Bhabha Atomic Research Centre, Trombay, Mumbai, India

Email: aarya@chem.ucla.edu

The ordering behaviour of a Ni-Mo alloy in the presence of a ternary additive viz. Cr has been studied. The sequence of ordering transformations in binary Ni-Mo alloys has been shown earlier to be controlled by a competition between several fcc-based superlattices, viz. Ni2Mo (Pt2Mo type), Ni3Mo (DO22) , Ni4Mo (D1a) and the so-called short range ordered (SRO) structure characterized by the presence of $\left\langle1\frac{1} {2}0\right\rangle$ reflections. The effect of ternary addition of chromium in the selection of the superlattice structure has been examined inthis paper in an alloy of composition Ni-- 24 at\% Mo--6 at\% Cr . The presence of Cr has been experimentally found to favour the formation of Ni2 (Mo,Cr) (Pt2Mo-type) phase in preference to the Ni3Mo DO22 and Ni4Mo (D1a) superlattices, This leads to a sequence of transformation different from that obtained in binary alloys. The effect of Cr on the ground state phase stability is determined, in this alloy, using the first-principles local-density based full-potential augmented plane wave (FP-LAPW) method [1]. We have used generalized gradient approximation (GGA) of Perdew-Burke-Ernzerhof (1996) parameterization scheme for the exchange-correlation functional. Our calculations have been fully converged with respect to the size of the basis set (R$_{MT}$K$_{max}$ = 9.5) and the number of $k$-points. We have used supercell approach to model our ternary superstructures with Cr impurity atoms substituting for Mo atoms in an otherwise perfect binary compound of a given space group symmetry.

From the differences in the electronic structures, total densities of states (DOS) and partial DOS, and total energies of binary Ni2Cr, Ni2Mo, Ni3Mo and Ni4Mo phases and the corresponding ternary suerlattice structures, an attempt has been made to predict the hierarchy of relative phase stabilities. Theoretical predictions based on these electronic structure calculations have been found to be consistent with the microstructural observations of the evolutionary stages of ordering in the ternary Ni--24 at\% Mo--6 at\% Cr alloy.

R. Prasad and R. Benedek

Department of Physics, Indian Institute of Technology, Kanpur 208 016, India

and

Chemical Technology Division, Argonne National Laboratory, Argonne IL 60439, U.S.A.

Email: rprasad@iitk.ac.in

We present first-principles calculations of electronic structure and total energies for various phases of Co doped and undoped lithiated mangnese oxides. We have used the ab initio pseudopotential method as implemented in Vienna ab initio Simulation Package (VASP). The generalized gradient approximation (GGA) and ultrasoft pseudopotentials have been used. We consider monoclinic and rhombohedral structures in paramagnetic, ferromagnetic and antiferromagnetic (AF3) spin configurations. We find that for both the doped and undoped cases magnetism plays an important role and results in significant lowering of total energy. The effect of Co doping on the stability of these phases will be discussed.

D. Balamurugan and R. Prasad

Department of Physics, Indian Institute of Technology, Kanpur 208 016, India

Email: balad@iitk.ac.in

The ground state structures and properties of Si$_3$H$_{n}$(n=1-6) clusters have been calculated using the Car-Parrinello molecular dynamics with simulated annealing and steepest descent optimization methods. We have studied total energy and first excited electronic level gap of the clusters as a function of hydrogenation. To understand the stability of the clusters the lowest energy fragmentation products and their dissociation energies have been calculated. Hydrogenation is done till all dangling bonds of silicon are saturated. Our results show that over coordination of hydrogen is favoured in Si$_{3}$H$_{n}$ clusters and the geometry of Si$_{3}$ cluster does not change due to hydrogenation.

A.K.Verma1, P. Ravindran2, , R.S. Rao1, B.K. Godwal1 and R. Jeanloz3

1High Pressure Physics Division, Bhabha Atomic Research Centre, Mumbai -400 085, India

2 Department of Chemistry, University of Oslo, Blindern N-0315, Oslo, Norway

3Department of Geology & Geophysics, University of California, Berkeley, California

94720, USA

Email: hpps@magnum.barc.ernet.in

State of the art high pressure experiments try to generate pressures of several megabars in diamond anvil cell (DAC),along with temperatures of several thousands 0K with laser heating, to obtain the physical conditions in the laboratory as existing in the interior of earth. As these experiments cover wide ranges of pressure and temperature, the choice of the gasket material in DAC is limited. Apart from the strength considerations, any phase transition in the gasket material may lead to complications in the interpretation of the X-ray diffraction data, as gasket peaks which are present along with the sample peaks have to be identified by knowing the details of the high pressure phase. Rhenium is one of the strongest polycrystalline materials with its bulk and shear moduli being among the highest known for metals. The ratio of its shear stress (t) to the shear modulus (m) at high pressures reaches 0.04 [1]. Thus Re is often used as gasket materiel in high pressure experiments. Hence, it is important to know about the high pressure behavior of Re, especially whether it transforms from the ambient condition hcp phase to some other structure under compression. We have carried out the electronic structure total energy calculations on rhenium up to about 1 TPa pressure by the full potential linear muffin orbital method. No structural transition to other commonly occurring close-packed structures like fcc, bcc, or bct from the ambient condition hcp phase takes place in Re. The axial ratio (c/a) of Re changes by less than 0.33% in the pressure range studied, initially decreasing from the ambient condition value, reaching a minimum value of 1.615 around 2 Mbar and then showing that Re will also be employed in ultra high pressure-temperature experiments.It is thus important to know its melting curve as a function of pressure.In our work we have employed two models (dislocation-mediated and Lindemann) to estimate the melting curve as a function of pressure.

[1] R. Jeanloz, B.K Godwal, and C. Meade, Nature, 349, 687 (1991) .

[2]G.I. Kerley, J.Chem. Phys. 73, 478 (1980)

D. Nguyen-Manh, T. Saha-Dasgupta and O. K. Andersen

Oxford Univ., U K

SNBNCBS, India,

Max-Planck-Institut fuer Festkoerperforschung, Stuttgart

Email:manh.nguyen@materials.oxford.ac.uk

The third-generation LMTO method developed by Andersen et.al. [1] provides a new wave function basis set in which the energy dependence of the interstitial region and inside muffin-tin (MT) spheres is treated on an equal footing. Within the improved method, basis functions in the interstitial are screened spherical waves (SSW-s) with boundary condition defined in terms of a set of "hard" sphere radii a_RL. Energy eigenvalues obtained from the single-particle Schrödinger equation for MT potential is energetically accurate and very useful for predicting a reliable first-principles tight-binding (TB) model of widely different systems. In this study, we investigate a possibility of the new basis set's transferability to different environment which could be crucial for TB applications to very large and complicated systems in realistic materials modeling [2]. For the case of C where the issue of sp2 versus sp3 bonding description is primarily important, we have found that by downfolding the unwanted channels in the basis, the TB electronic structure calculations in real-space (band energies, density of states and charge densities) for both hexagonal graphite and diamond structures are well compared with those obtained from the full self-consistent scheme if we use the same choice of hard sphere radii a_RL and a fixed, arbitrary energy e_u. Moreover, the choice is robus and transferable to various situations, from different forms of graphite to a wide range of coordination. Using the obtained minimal basis set, we have been investigating the TB Hamiltonian and overlap matrices for different structures of Carbon, in particular we have prediced the on-site and hopping parameters (g1, g2....g6) within an orthogonal representation for Slonczewski-Weiss-McClure model of the Bernal structure. Our theoretical values are in excellent agreements with experimental ones from magneto-reflection measurements of Fermi surfaces for hexagonal graphite.

[1] O.K. Andersen, C. Arcangeli, R.W. Tank, T. Saha-Dasgupta, G. Krier, O. Jepsen and I. Dasgupta, in Tight-Binding Approach to Computational Materials Science, Eds. by P.E.A. Turchi et. Al., MRS Symp. Proc., 491, 3, (1998);

O.K. Andersen and T. Saha-Dasgupta, Phys. Rev. B, 62, R16129, (2000).

[2] D. Nguyen-Manh, D.G. Pettifor and V. Vitek, Phys. Rev. Lett., 85, 4136, (2000).

Manoj K. Harbola(1) and K.D. Sen(2)

Department of Physics, Indian Institute of Technology, Kanpur 208016, India

School of Chemistry, University of Hyderabad, Hyderabad 400135, India

Email: mkh@iitk.ac.in

Over the past decade and a half, many new accurate functionals, basedon the generalised gradient approximation, have been proposed, and theygive energies close to chemical accuracy. However, accuracy of theenergy functional does not gurantee that its functional derivative, which gives the corresponding potential, is also accurate all over space. For example, although the Becke88 [Phys. Rev. A {\bf 38}, 3098 (1988)] exchange-energy functional gives very good exchange energies, its functional derivativegoes as $-\frac{1}{r^{2}}$ in comparison the correct$-\frac{1}{r}$ for $r\rightarrow\infty$, where $r$ is the distance of theelectron from a finite system. On the other hand, accuracy of the potential is of prime importance if one is interested in properties other than thetotal energy; Properties such as optical response depend crucially on the potential in the outer regions of a system. In this paper we present a different approach, based the ideas of Harbola and Sahni [Phys. Rev. Lett. {\bf 62}, 489 (1989)], to obtain the potential directly from the energy density of a given approximation, without taking recourse to thefunctional derivative route. This leads to a potential that is as accurate asthe functional itself. We demonstrate the accuracy of our approach by presenting some results obtained from the Becke88 functional.

Kavita Joshi, D. G. Kanhere

Dept of Physics, University of Pune,Pune, India

Email: kavitaj@physics.unipune.ernet.in

Technological usage of Lithium Batteries and Tin-oxide glasses as one of the possible candidates for anode material provides motivation to study Lithium-Tin clusters. Theoretical electrochemical voltage profile of extended systems compares well with experiments for ${\it x < 2.5 }$ in ${\rm Li_x\rm Sn}$. We examine whether those properties can be observed in clusters.

We provide a systematic analysis of bonding properties and eigenvalue structure of large clusters ${({\rm Li_{20}\rm Sn_{20-x}, {\it x}={2}, {4}, {6}, {8}, {10}})}$ along with low lying geometries of small clusters ${(\rm Li_{2}\rm Sn_{5}, \rm Li\rm Sn, \rm Li_{7}\rm S_{3}, \rm Li_{5}\rm Sn_{2}, \rm Li_{7}\rm Sn{2} )}$. In {\rm Li-Sn} clusters it is observed that {\rm s}-electrons of {\rm Sn} do not contribute in bonding also {\rm Sn} mixes homogeneously in {\rm Li} clusters. ${8}, {10}, {20}$ electron {\rm {Li-Sn}} systems have a large {\rm HOMO-LUMO} gap which shows that spherical jellium model is applicable to {\rm {Li-Sn}} clusters. Charge distribution, bonding characteristics are studied and corresponding results will be presented.

C Majumder, V. Kumar, H. Mizuseki, and Y. Kawazoe

Institute for Materials Research, Tohoku University, Sendai, Japan

Email: majumder@imr.edu

The present work is motivated by the recent experimental results of Jarrold and co-workers, [ A. A. Shvartsburg and M. F. Jarrold, Phys. Rev. Lett. 85 (2000) 2530] who showed that the melting temperatures of tin clusters with 15 to 30 atoms are atleast 50 K higher than the bulk value. The structuresand chemical bonding in small clusters are generally different from the bulk and that could be a reason for suchan unusual behavior of tin clusters. In the periodic table tin lies belowgermanium (semiconductor) and above lead (metal). It also has two allotropes. In the form of clusters, there is a narrowing of the band due to reduced coordination. This makes clusters of an element to behave somewhat similar to an element above it in the periodic table. Therefore, understanding the atomic and electronic structures of tin is very important as there could be isomers having different nature of bonding from covalent to metallic. In the present work we have carried out an ab-initio study of small tin clusters (n = 2-20) using ultrasoft pseudopotentials and generalized gradientapproximation for the exchange-correlation energy. The lowest energy isomers upto 8atoms have the same structures as Si clusters but larger clusters (except for 10) seem to favor a different structure. Most significantly the binding energies of clusters with n > 10 are found to be only about 10 % less than the calculated bulk value. The high binding energies of small tin clusters are predicted to be responsible for the higher melting temperatures of these clusters. The adiabatic ionization potentials of tin clusters have been calculated using the optimized geometries of neutral clusters as the initial configuration. The fragmentation behavior and ionization potentials calculated in this work show an excellent agreement with the experimental results and suggest that the lowest energy structures, obtained here, are close to the global minima.

Deepak1, D. Balamurugan2 and K. Nandi1

1 Department of Materials and Metallurgical Engineering

2 Department of Physics, Indian Institute of Technology, Kanpur 208016, India

Email: saboo@iitk.ac.in

There is an abundant literature on calculations of formation and ionisation energies of point defects in GaAs. Since most of these energies, especially the formation energies, are difficult to measure, the calculations are primary means of obtaining their values. However, based on the assumptions of the calculations, the reported values differ by each other by amounts greater than 100%. In this paper we discuss the sources of the errors and their impact on practical predictions valuable in GaAs device fabrication. In particular, we have compared a large set of computed energies and selected the most appropriate values. Then, in the context of GaAs material quality and dopant activation, we investigated the impact of errors in calculation of formation energies on the performance of the GaAs substrate for device fabrication. We find that in spite of the errors inherent in ab initio calculations, it is possible to correctly predict the performance of GaAs substrate.

Modak, R. S. Rao, B. K. Godwal and S. K. Sikka

High Pressure Physics Division, Bhabha Atomic Research Centre, Mumbai -400 085, India

Email: hpps@magnum.barc.ernet.in

Since the discovery of the unexpectedly high superconducting transition temperature (Tc~39 K) in MgB2 by Akimitsu et al [1], extensive research activities are taking place on this binary compound. To understand the origin of this superconductivity in this simple metal compound we carried out ab initio electronic structure calculations using full potential linear augmented plane wave method with generalized gradient approximation to the exchange-correlation term. MgB2 crystallizes in the AlB2 structure, where B atoms form a primitive honeycomb lattice consisting of graphite type sheets and the Mg atoms are in alternate hexagonal plane. From the total energy minimization we estimated the equilibrium volume and c/a, and obtained good agreement with the experimental value. Our calculated bulk modulus 148 GPa is in good agreement with the experimental value 152 GPa. We also calculated density of states and the band structure. The details of band structure show that the high energy part of the valence band of MgB2 is made up predominantly of B 2p states. These bands form a flat zone in the G-A direction and contribute considerably to the total density of states (DOS) at EF. Our calculated DOS at EF (N(EF )) is quite high, the high N(EF ) and hence the 2p states of B are responsible for the enhancement of the superconducting transition temperature. Using the N(EF ) we calculated the electronic specific heat coefficient and from available experimental estimate we calculated the mass enhancement parameter and the electron-phonon coupling constant l. Using this l in McMillan formula for Tc, we estimated Tc as 24.7 K. Our estimate of Tc is rather low compared to the experimental value but agrees well with other theoretical estimates.

J. Akimitsu, Symposium on Transition Metal Oxides, Sendai, January 10, 2001;

J. Nagamatsu et al, Nature 410, 63 (2001).

Naoki Watanabe and Masaru Tsukada

University of Tokyo,Tokyo, Japan

Email: naoki@cms.phys.s.u-tokyo.ac.jp

It will be an important study in the field of Computational Material Scienceto simulate the quantum electrons dynamics from first principlesSo far, we have developed some computational techniques for simulating the time-evolution of wave functions under the following Time-Dependent Kohn-Sham equation (TD-KS) %\[{\rm i} \hbar \frac{{\rm d}}{{\rm d} t}\, \psi_n(t) = {\cal H}[\rho]\, \psi_n(t)\ ; \qquad {\cal H}[\rho] = - \frac{\hbar^2}{2m} \triangle + V_{\rm eff}[\rho]\ ;qquad \rho({\bf r},t) = \sum_{n=1}^N |\psi_n({\bf r},t)|^2 \ . \] % where $\psi_n$ means a set of one-electron wave functionsand $V_{\rm eff}[\rho]$ is an effective potential standing mutual interactions among electrons, which is a functional of the density of electron. The analytical formal solution of this TD-KS equation is given by a time-ordering exponential operator as, % \[\psi_n(t) = {\bf P} \exp{\Bigl -\frac{{\rm i}}{\hbar} \int_{0}^{t} \!\! {\rm d}t^\prime\,\, {\cal H}[\rho(t^\prime)] \Bigr]}\, \psi_n(t=0) \]

The difficulty in numerically solving the TD-KS equation is the treatment of the density-dependent Hamiltonian. The wave functions and the Hamiltonian should always be self-consistent with each other, however it needs a self-consistent loop in every time steps, namely it is a tough and complicated work. We have avoided this problem and have proved that the following simple practical formula is sufficient for solving the TD-KS equation, % \[ \psi_n(t+\Delta{t})\simeq \exp{\Bigl[ -\frac{{\rm i}}{\hbar}\frac{\Delta{t}}{2} V[\rho] \Bigr]} \exp{\Bigl[ \frac{{\rm i}\hbar\Delta{t}}{m} \frac{\triangle}{2} \Bigr]} \exp{\Bigl[ -\frac{{\rm i}}{\hbar}\frac{\Delta{t}}{2} V[\rho] \Bigr]} \psi_n(t)\ . \] %

In this poster, we will show the detail of the computationaltechniques, and present some illustrative simulations of fast quantum electrons dynamics computed with our method, for instance, nonlinear photo responses of crystals, electrons flow in mesoscopic circuits and molecular devices.

N. Watanabe and M. Tsukada, Phys.\ Rev.\ E, Vol 62, No 2, 2914 (2000).

N. Watanabe and M. Tsukada, J.\ Phys.\ Soc.\ Jpn.\, Vol 69-9, 2962 (2000).

http://www-cms.phys.s.u-tokyo.ac.jp/\~{}naoki/

Nitya Nath Shukla and R. Prasad

Department of Physics, Indian Institute of Technology, Kanpur 208016, India

Email: nitya@iitk.ac.in

We present an \emph{ab initio} calculation of interlayer magnetic coupling for Fe/Nbmultilayers using the self-consistent full-potential linearized augmented-plane-wave (FLAPW) method. For this calculation, we have constructed supercells consisting of bcc Fe and Nb multilayers in Fe/Nb/Fe sandwich geometry stacked along (001) direction. In the supercells two Fe layers are separated by Nb layers ranging from 1 to 11 layers. We have calculated the total energy of the system as a function of Nb spacer layer thickness. For each spacer layer thickness, we have done three calculations corresponding to para, ferro and antiferromagnetic ordering of Fe atoms. The interlayer magnetic coupling isobtained from the energy difference of the system in which Fe layers areantiferromagnetically and ferromagnetically ordered. We find that theinterlayer magnetic coupling oscillates with increasing Nb spacer thickness in agreement with theexperimental results. The induced magnetic moment is also found to be oscillating with increasing Nb spacer layer thickness.

Nirupam Chakraborti 1and R.Prasad 2

1Department of Metallurgical and Materials Engineering, Indian Institute of Technology, Kharagpur 721302

2Department of Physics,Indian Insdtitute of Technology, Kanpur

Indian Institute of Technology, Kanpur 208 016

Email: rprasad@iitk.ac.in

We discuss the application of biologically inspired genetic algorithms to determine the ground state structures of a number of Si-H clusters. The total energy of a given configuration of a cluster has been obtained by using a non-orthogonal tight-binding model and the energy minimization has been carried out by using genetic algorithms. Our

results for ground state structures and cohesive energies for Si-H clusters are in good agreement with the earlier work conducted using the simulated annealing technique. We find that the results obtained by genetic algorithm turn out to be comparable and often better than the results obtained by the simulated annealing technique.

Vijay A. Singh and V. Ranjan

Physics Department, Indian Institute of Technology, Kanpur,-208016, INDIA.

Email: vranjan@iitk.ac.in

The strong industrial drive towards smaller and ever smaller devices has fueled interest insemiconductor nanostructures, also popularly known as quantum dots (QDs). The electronic structure of these systems has been the subject of intense investigation over the last decade. Little work, however has been done to investigate the role of defects in these systems. At the same time semiconductor QDs came to be identified as ``artificial atoms'' on account of sophisticated and careful experiments on single electron and Coulomb blockade effects in the last decade. This opened up a possibility of studying many body effects in these systems. In bulk semiconductors it is well-known that an addition of a mere 0.001 \% dopant atom, i.e. one in a hundred thousand atoms, can enhance the transport characteristics by several orders of magnitude. It is thus a moot question to enquire what would happen to a quantum dot which consists of about $10^4$ to $10 ^5$ under similar conditions. The ground-state energy and the binding energy of shallow hydrogenic impurities in spherical quantum dots have been calculated as a function of the radius $R$ of the dot. We observe that the binding energy reaches a peak value as the dot radius decreases and then diminishes. It may even become resonant with the host under some conditions. This shallow-deep (SHADE) transition appears to be hallmark for all semiconductor QDs \cite{sing78,ranj01a}. If one takes into account the diminished value of the the dielectric constant $\epsilon(R)$ and its dependence on the dot size, we observe that the impurity binding energy maybe driven even deeper. Thus carrier ``\textit{freeze out}'' is likely to occur. At the same time we argue that this deepening may result in an an enhancement of luminescence efficiency. We have also carried out a tight-binding (TB) calculation for deep defects in semiconductor quantum dots \cite{ranj00}. It reveals that deep defects do not show any SHADE effect. The binding energy fluctuates as the size of the QD is decreased. We have also studied \cite{ranj01b} the many electron effects in QD within the local density approximation (LDA) and the Harbola-Sahni (HS) scheme. The latter incorporates the exchange effects essentially exactly. Wehave studied the effect of shape of the potential on these systems. The shapes considered are quasi-triangular, quasi-harmonic and varying degrees of quasi-spherical. We examine the level spacing, level degeneracy, and the scaling of the electron-electron interaction with size. We find that the scaling laws have a complex dependence on the size and shape of the potential. Finally, we compare our calculations with experiments on CdSe quantum dot.

Vijay A. Singh, C.Weigel, J.W.Corbett, and L. Roth, Phys. Stat. Sol.(b){\bf 81}, 637 (1978).

V. Ranjan and Vijay A. Singh, J. of App. Phys. \textbf{89}, 6415, (2001).

V. Ranjan, Vijay A. Singh, and M. Kapoor, \textit{Physics of Semiconductor Devices}, edited by V. Kumar and S. K. Agarwal (Allied, New Delhi, 2000).

V. Ranjan, R. K. Pandey, Manoj K. Harbola and Vijay A. Singh, Self Capacitance of a Quantum Dot : Dependence on the Shape of the Confining Potential (submitted).

Bach Thanh Cong, Pham Ngoc Anh Huy

Faculty of physics Hanoi University of Science,334 Nguyen Trai Street , Hanoi, Vietnam.

email : cong@cms.edu.vn

The reentrant ferromagnetism (ferromagnetic order exists in certain temperature range above 0 K) in some substitutedm perovskites is explained using the Ising spin model on the square lattice with mixing ferromagnetic and antiferromagnetic exchange interactions . It was shown by numerical calculation that this effect is strongly affected by the external magnetic field and lattice disorder.

S.N. Bose National Centre for Basic Sciences, Block JD, Sector III, Salt Lake

Calcutta 700091, India

email: sugata@bose.res.in

Orthogonal and non-orthogonal tight-binding molecular dynamics simulation methodwere used to obtained groundstate equlibrium structure and cohesive energies ofSi$_n$ ($n = 2-19$) clusters. Femto-second laser induced melting and fragmentationbehaviour of these clusters will be discussed.

S.Ranganathan and P.Ramachandrarao

National Metallurgical Laboratory, Jamshedpur 831 007, India

Email: raghu10101@yahoo.co.in

Liquid-solid and solid-solid transformations occur during the processing of alloys. Accurate prediction of phase transformation during the processing is crucial to control this and achieve maximum efficiency in modern processes such as continuous casting, thermo-mechanical treatment of steel etc. It is also vital for successful design of alloys and processes to obtain the desired properties in the alloys. Phase transformation is controlled by the thermodynamics of the system as well as by the kinetics of transformation. The common approach to addressing this is to predict the size and composition of the nucleus formed from a parent phase. Routines are developed for predicting the phases formed under given conditions of alloy composition and temperature. Usually the codes for predicting phase transformation during processing are linked to those that predict the phase transformation from equilibrium thermodynamics. When several competing phases are present in a system, the latter codes usually predict th phases that are formed by minimizing the free energy of the system. It is assumed that a nucleus of the predicted phase is formed from the parent phase. Assuming thermodynamic equilibrium at the interface between the parent and transformed phases, the progress of transformation is predicted. Reliable prediction of the liquidus and solidus temperatures of the alloy composition at the advancing front is required for the prediction of transformation. In the case of multi-component alloys, prediction of these temperatures using the phase diagrams is complicated. Hence, empirical methods are adopted for this. An algorithm that is simple in mathematics and is inherently free from any need for elaborate schemes to ensure convergence and for eliminating negative and imaginary solutions will quicken the predictive algorithm. The concept of what may be called as "transformation free energy change" has been extensively used in studying the solidification process. This is given by :

DGLS = XAS*DmA + XBS*DmB (1)

where, Xip represents the atom fraction of the component 'i' in the phase 'p'

and Dmi = mi S - mi L

When XAL is plotted against XAS, satisfying the expression the condition

DGLS =0 (2), a curve is obtained defining the domain within which the transformation from liquid to solid is possible. Outside this domain, such a transformation is prohibited. This principle can be used to explore the reverse process also, i.e. the formation of a liquid from a solid. The liquid-solid transformation and solid-liquid transformation curves help in quickly predicting whether a desired transformation is thermodynamically feasible for the given composition and temperature of the system. The tangent to the DGLS =0 line, parallel to the XAS axis defines the thermodynamic equilibrium, i.e. the equilibrium liquidus and solidus composition of the alloy at the given temperature. Similarly, the tangent to the curve parallel to the XAL axis defines the point at which massive transformation i.e. transformation without change in composition occurs. The liquidus and solidus temperatures of the given alloy composition are other input parameters required for prediction of the transformation. These two can be defined, respectively, as the minimum temperature at which the alloy remains completely a liquid and the maximum temperature at which it remains completely a solid. It is shown in this communication that solving the expression (2) above, provides the simplest algorithm to predict the solidus and liquidus compositions and the solidus and liquidus temperatures. Analysis of the solidification phenomenon employing the expression (2) has been traditionally confined consideration of only one solid phase. The case when there are two or more competing solid phases that can form the liquid are present, remains unexplored. Application of this technique to such cases is demonstrated in this communication. The technique has been extended to predicting solid-solid transformation also. Utilisation of the concept of "transformation free energy change" to predict this transformation remains unexplored hitherto. Application of the new algorithm is demonstrated in the case of a few binary and ternary systems. It is also demonstrated that this algorithm would be simple in application and can be easily adopted in any code used for predicting phase transformation during the processing of alloys.

Physics Department

Institute of Engg. & Technology,

MJP Rohilkhand University, Bareilly - 243 006, India

S. Auluck

Department of Physics, University of Roorkee, Roorkee - 247 667, India

Email: drsudhirkumar_in@yahoo.com

We have calculated the electronic density of states (DOS) and dielectric function for the ThX (X=P, As and Sb) using the linear muffin tin method within atomic sphere approximation (LMTO-ASA) and including the combined correction terms. The calculated electronic DOS have been compared with the available experimental data and we find a good agreement. The calculated DOS reveals that thorium f orbital plays an important role as it lies near the fermi level in all the three monopnictides. It's contribution is small in comparison to other orbitals. The calculated optical conductivity for ThP and ThAs is increasing monotonically, while for ThSb a sharp peak has been found at 6.5 eV. Unfortunately there are no experimental data to compare with calculated optical properties.

Dept of Physics, University of Pune, Pune 411 007, India

Email: anjali@physics.unipune.ernet.in

We have calculated the electron momentum density and Compton profiles of Li using full-potential linear augmented plane wave (LAPW) method. Electron correlations left out within the local density approximation are included with the prescription provided by Lam and Platzman employing the self-consistent charge density. Thermal effects on the electron momentum density and Compton profiles of Li are investigated. These effects are accounted for, by incorporating change in the lattice constant as well as by taking into consideration the vibrational motion of ions. Our results for the difference in Compton profiles at 95 K and 295 K are compared with the available experimental results. Compton defect calculations have been included for the core states.

Dept of Physics, University of Pune, Pune 411 007, India

Email: anjali@physics.unipune.ernet.in

Main focus of the present investigation is to understand the quantum confinement effects in medium sized clusers of various classes of semiconductors. We employ the empirical pseudopotential method to calculate the HOMO - LUMO gap as a function of the shape, size and the phase of the nanoclusters. We try to explain the building up of the bulk band structure when the size of the dot is much larger than the bulk Bohr exciton radius.

We use the known geometry of the atoms in a cluster to calculate the excited states using the time-dependent density functional theory for very small clusters.

The results are compared with the available experimental data for various II-VI and III-V semiconductor quantum dots.

T. M. Briere, M. Sluiter, V. Kumar, and Y. Kawazoe

1Institute for Materials Research, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai, 980-8577, Japan

Email: briere@imr.edu

Recent Stern-Gerlach experiments [1] on medium-sized Mn clusters (Mn11-Mn99) show anomalously low values of the magnetic moments of Mn13 and Mn19 clusters while high values for Mn15 and Mn23-Mn25. The total moment per atom itself is rather small varying from about 0.3 mB/atom to about 1.5 mB/atom, as compared to 5 mB for the Mn atom and 1-2.3 mB per atom for the two bulk phases that have antiferromagnetic coupling. Studies on small clusters Mn3-Mn8) show that these are ferromagnetically coupled and have large magnetic moments of 4-5 mB per atom [2,3]. Therefore, a sudden decrease in the values of the magnetic moments of the 11-99 atom clusters is surprising.

In order to understand these intriguing new results, we have performed first-principles calculations on several Mn clusters, including Mn13, Mn15, Mn19, and Mn23. We have used the supercell method with a 15 Å unit cell, ultra-soft pseudopotentials, a plane-wave basis, and gradient-corrected exchange-correlation functional. We discuss the physical basis of the anomalous values of the magnetic moments as well as the transition from the high-moment ferromagnetic coupling in the small clusters to a more complicated spin structure with lower magnetic moments occurring in these medium-sized clusters.

[1] M. B. Knickelbein, Phys. Rev. Lett. 86 (2001) 5255.

[2] S. K. Nayak, B. K. Rao, P. Jena, J. Phys. Cond. Matter. 10 (1998) 10863.

[3] M. R. Pederson, F. Reuse, S. N. Khanna, Phys. Rev. B 58 (1998) 5632.

C. Ravi and R. Asokamani

Division of Science and Humanities, MIT Campus,

Anna University, Chennai - 600 044, INDIA

Email: vcravi@hotmail.com

This paper reports the site preference of Zr atoms in Ti2ZrAl and its phase stability along with the ground state properties such as equilibrium lattice parameters, heat of formation and bulk modulus calculated using first-principle electronic structure calculations. Total-energy calculations performed for as many as sixteen different possible site occupancy of Ti, Zr and Al atoms conclusively shows that in the D019 phase of Ti2ZrAl, Zr atoms indeed prefer to substitute in the Ti sites, which is in agreement with the experimental reports, and in particular to a specific set of sites among the 16 possibilities. This calculation further shows that among the seven competing phases considered, L12 is the ground state structure of Ti2ZrAl and the D019 structure should be a meta-stable phase. The calculated lattice parameters are in good agreement with experiments. The basis for structural stability and bonding behaviour are analyzed in terms of density of states also.

Kirit N. Lad, K. G. Raval and Arun Pratap

Condensed Matter Physics Laboratory, Applied Physics Department, Faculty of Technology and Engineering, M. S. University of Baroda, Vadodara - 390 001.India

Email: (received in mail)

Amorphous ZrxNi1-x belongs to the metal-metal alloys class. These alloys can be produced in the amorphous state over a wide concentration range and therefore are of great scientific interest. Neutron diffraction and computer simulation are the most widely used techniques to the study the structure of such alloys. The other way to study the structure of the alloys is by solving various integral equations like the Percus-Yevick (PY) and Hypernetted Chain (HNC) equations for liquid structure. The HNC equation for a single component system interacting through potential V(r) can be written as c(r) = exp (-bV(r) + g(r)) - g(r) - 1 Where g(r) = h(r) - c(r), h(r) is the total correlation function and b = 1/kBT. Gillan [1] has developed a powerful technique for the numerical solution of this integral equation. The structure factor, which is the Fourier transform of the pair correlation function, will be obtained by solving the HNC equation using Gillan's algorithm. The Zr-Ni alloy is in transition metal group. The Will and Harrison formulation [2] for the transition metals, therefore, can be generalized to derive the effective pair potential in these binary alloys.

[1] Gillan M.J., Mol. Phys., 38 (1979) 1781.

[2] J. M. Wills and W. A. Harrison, Phys. Rev. B 28 (1983) 4363.

M. D. Deshpande and D. G. Kanhere

Department of Physics, University of Pune, Pune 411007, India

Email: mdd@physics.unipune.ernet.in

The theoretical background for calculating ground state properties of many-electron systems is now well established. One of the most widely used technique employs the first principles pseudopotential formalism based on density-functional theory in local density approximation. Several investigations are also available on electronic structure calculations on small homogeneous clusters, relatively few reports are available on heterogeneous systems. Much insight can be gained by examining the effect of single impurity on the geometries of pure clusters. Most of these studies have been carried out for small clusters with very limited geometry optimization without simulated annealing.

It is interesting to note some systematic of the behavior of different impurities in same host.Another interesting aspect concerns the validity of spherical jellium model. Lithium and sodium are considered as simple prototype systems for such studies. We have carried out systematic investigation on ground state geometries and some low-lying geometries, energetics and stability of $Li_nBe$ ,$Li_nMg$, $Li_nNa$ and $Na_nLi$ (n =1-12) clusters.The calculations have been performed by using standard {\it ab initio} molecular dynamics within the framework of density functional theory (DFT) within the pseudopotential approximation using simulated annealing strategy.These investigation reveal that impurities with smaller ionic radii and strong binding with the host are seen to get trapped in the cluster. Impurities having larger ionic radii remain on the surface of the clusterand may distort the host geometries significantly. The eigenvalue spectrum of these clusters generally confirms to the SJM with minor modification. To see whether a similar behavior is seen in mixed clusters, we have also investigated the clusters of the type $Al_4X_4$.One of the important conclusion is that, in general, the eigenvalue spectrum and their character for mixed clusters do not conform to SJM.

Currently we have employed a density-functional pseudopotential approach for solving the Kohn-Sham equation in real space.The polarizabilities are computed for mixed sodium-lithium clusters using finite field method. Equilibrium geometries and electronic-structure properties of Na$_n$Li and Li$_n$Na (n = 1-12) clusters are obtained using {\it ab initio} molecular dynamics method with the generalized gradient approximation for the exchange-correlation potential.The resulting geometries show that Li atoms become trapped inside the Na cage,while Na prefers to be on the periphery of Li clusters.The comparison of total binding energies indicate a high degree of stability for clusters with 8 atoms.We also report polarizabilities for both series of Na$_n$Li and Li$_n$Na clusters. Our calculations demonstrate that Li impurity reduces polarizabilities of Na$_n$ clusters while the doping of Na in Li$_n$ clusters increases the polarizabilities.We relate remaining differences in the magnitude of the theoretical and experimental polarizabilities to the finite temperature present in the experiments.

We implement the linear response theory within time-dependent local density-functional formalism(TDLDA) to calculate excitation energies and photo absorption spectra of Na$_{8-y}$Li$_y$ (y=1-8) clusters.The changes in the absorption spectra reflect the deviation of the cluster geometry. The calculated ab initio spectra for $Na_8$ and $Li_8$ are in very good agreement with experiment.

Vaishali Shah and David S. Sholl,

Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA 15213, U S A

Email: vaishali@andrew.cmu.edu

The chemical properties of CO on Ni surfaces are important in several contexts, including the corrosion of Ni via metal-carbonyl formation and the deactivation of Ni catalysts by carbide formation during steam reforming. Experiments indicate that low coordination sites on Ni surfaces are much more reactive with respect to CO than atomically flat planes. To study this observation quantitatively, we have used Density functional theory to study the adsorption of CO on flat and high Miller index Ni surfaces. The adsorption energies reported are from ab initio pseudopotential spin polarised plane wave calculations. Both the local density approximation and the generalized gradient approximation have been used for the exchange correlation energy. We have focused on three Ni surfaces namely the flat Ni(110) and stepped (210) and highly kinked (531). The adsorption of CO at three sites, atop, bridge and hollow site has been studied for the flat and stepped surface. On the highly kinked (531) various adsorption sites have been identified and shall be energetically compared to locate the preferred site of adsorption. Our results shall be compared with the existing experimental observations of CO adsorption on Ni surfaces.

D.P. Mondal, S. Das, A.H. Yegneswaran and N. Ramakrishnan

Regional Research Laboratory (CSIR), Bhopal - 462 026

Email: ahyegneswaran@rrlbpl.org

Aluminium metal matrix composites (AMMCs) have emerged as potential materials for automobile, aerospace and defence applications because of their desirable specific strength, specific stiffness, damping properties, coefficient of thermal expansion, resistance to friction, wear and seizure etc. In recent times, several components like cylinder blocks, drive shafts, brake drums etc., made of AMMCs have gone into commercial production in countries like USA, Japan and Europe. However, application oriented research on these composites are still in the embryonic stage in India.

The present paper deals with the use of FEM in predicting the structure insensitive properties such as modulus, coefficient of thermal expansion, thermal conductivity. The analysis has been extended to predict structure sensitive properties using large deformation FEM and then relating the mechanical & thermal properties to tribological properties. A specific attempt has been made to model interfaces for this purpose.

Attempts have also been made to predict the wear rate of composites and alloys with in a selected experimental domain using factorial design of experiments. The linear regression equation developed helps in understanding individual and combined effect of the parameters like load, abrasive size, sliding distance, impingement angle etc. on the wear behaviour of composites. These analyses indicate that the interaction of the experimental parameters are reasonably high and thus significantly control the wear mechanism of the composite under given set of experimental parameters.

Vijay Kumar1,2, M. Sluiter1, and Y. Kawazoe1

1Institute for Materials Research, Tohoku University, 2-1-1 Katahira Aoba-ku, Sendai 980-8577, Japan

2Dr. Vijay Kumar Foundation, K.K. Nagar (West), Chennai-600078, India

Email: kumar@imr.edu

We report results of our extensive studies of phase transitions in crystals of arm-chair and zig-zag type single wall carbon nanotubes of differing diameters under hydrostatic pressure. The calculations have been performed using first principles electronic structure ultrasoft pseudopotential plane wave method and local density approximation (LDA) for the exchange-correlation energy. The latter provides a good description of the structural properties of graphite. It is found that the phase transitions depend on the symmetry of the nanotubes. We discuss the conditions under which polygonization of nanotubes occurs into a hexagonal shape under pressure and the cases where a structural transition occurs from a near triangular lattice of cylindrical nanotubes at zero pressure to a monoclinic structure of oval shaped nanotubes at high pressures. Nanotube crystals that have a non-triangular lattice even at zero pressure are also discussed. These results would clarify the present conflicting experimental reports on pressure effects on nanotube bundles. Effects of pressure on the electronic structure will also be discussed.

C34 Microstructure study of laser surface cladding of bearing materials

G. Phanikumar, P. Dutta* and K. Chattopadhyay

Department of Metallurgy, *Department of Mechanical Engineering

Indian Institute of Science, Bangalore 560 012 INDIA

Email: phani@metalrg.iisc.ernet.in

Laser cladding is a versatile manufacturing process to produce coatings of desired microstructure and properties. In this poster we describe a computational model of the laser caldding of a bearing material aluminium-bismuth alloy on an aluminium substrate.

The process involves heat transfer and fluid flow in the laser melt pool, nucleation, growth and coalescence of Bi particles in the alloy. A multi stage computational model is developed to gain insight in to the process of microstructure development. The first stage involves estimation of realistic thermal profiles of various locations in the clad/remelt region taking in to account, the heat transfer and fluid flow which are solved using a control volume technique. The second stage involves homogeneous nucleation and diffusional growth under the thermal profile that prevails during the solidification of laser molten pool. In the last stage, collision and coalescence of particles due to convection is computed by tracing the particles in the laser meltpool by integrating the velocities of the particles. An order of magnitude of the final size of particles is compared with experimentally observed one.

Theoretical Science Unit

J Nehru Centre for Advanced Scintific Research,

Jakkur, Bangalore, 560 064 India

Email: waghmare@jncasr.ac.in

We present a theoretical approach in which ab initio calculations based on density functional theory are used in developing microscopic models that focus on the fundamental physics of ferroelectric materials and phase transitions. Simulations of these models yield macroscopic behavior of these materials and help understand how it links with the microscopics. Applications to simple perovskite and complex relaxo ferroelectrics are presented for illustration.

Raghani Pushpa and Shobhana Narasimhan

Theoretical Sciences Unit,

Jawaharlal Nehru Centre for Advanced Scientific Research,

Jakkur PO, Bangalore-560 064, India

Email : pushpa@jncasr.ac.in

We have studied the reconstruction of the Pt(111) surface theoretically by using the Frenkel-Kontorova model. The parameters in the model are obtained by performing {\it ab initio} density functional theory calculations.

The Pt(111) surface does not reconstruct under normal conditions but experiments have shown that there are two ways to induce the reconstruction: by increasing the temperature, or by depositing adatoms on the surface. The basic motif of this reconstruction is a ``double stripe" with an increased surface density and alternating HCP and FCC domains, arranged to form a honeycomb pattern with a very large repeat distance of 100-300 $A^0$. In agreement with experiment, we find that it is favourable for the surface to reconstruct in the presence of a large number of adatoms, but not otherwise.

Hem Chandra Kandpal and Ram Seshadri

Solid State and Structural Chemistry Unit

Indian Institute of Science, Bangalore 560 012 India

Email: seshadri@sscu.iisc.ernet.in, FAX: (91) 80 360 1310

While bonding between d^10 atoms and ions in molecular systems has been well studied, less attention has been paid to interactions between such seemingly closed shell species in extended inorganic solids. In this contribution, we present visualizations of the electronic structures of the delafossites AAlO_2 (A = Cu, Ag, Au) with a particular emphasis on the nature of d-d interactions in the close packed plane of the coinage metal ion. We find that on going from Cu to Ag to Au, the net bonding between the A and O decreases while bonding between A and A increases. In keeping with what is seen in molecular species, we ascribe the increase in metal-metal bonding with increasing group number to scalar relativistic effects. A comparison with metallic PdAlO_2 where bonding between d Pd is in little doubt, helps cast light on the different interactions in the title compounds. The question of whether such interactions play a role in the transport and optical properties of this important new class of transparent conducting oxides is examined.

Ashwin S. S, Srikanth Sastry, Umesh Waghmare

Theoretical Sciences Unit,

Jawaharlal Nehru Centre for Advanced Scientific Research,

Jakkur PO, Bangalore-560 064, India

Email : ashwinss@jncasr.ac.in

There is evidence of a first order phase transition in Silicon at around 1400K, this phase transition is accompanied by a semiconductor-metal transition. Recent Classical Molecular Dynamics simulation have shown a first order liquid-liquid phase transition, we have investigated if the structural phase transition involved in this transition is capable of triggering an electronic transition in Silicon. In order to verify this we have generated configurations of Supercooled liquid Silicon from Classical MD calculation and from these we generate the density of states and the participation ratio for wavefunctions using DFT methods.

Theoretical Sciences Unit,

Jawaharlal Nehru Centre for Advanced Scientific Research,

Jakkur PO, Bangalore-560 064, India

Email: pati@jncasr.ac.in

The possibility of making low-dimensional molecular circuits, due toadvances in nanofabrication technology, has greatly stimulated the interest in the transport properties of such low-dimensional systems. We have developed a mean-field model for calculating the I-V characteristics of a tight-binding molecular wire with electron-phonon and electron-electron interactions. It is based on the Pariser-Par-Pople Hamiltonian coupled with static phonons for conducting polymers and the current is computed within a time-independent scattering formalism. Our calculations show that the vibronic coupling induces regions of strong negative differential resistance in the I-V curves and can lead to the effective suppression of the current. Electron correlations on the otherhand can give rise to assymetry in the I-V curves. These effects can be traced back to the level crossing and breakdown of electron-hole symmetry caused by electron-phonon coupling and electron-electron interactions respectively in presence of applied external bias.

JNCASR, Jakkur Bangalore 560 064,

India

and

Stefano de Gironcoli,

SISSA, Trieste, Italy.

Email: shobhana@jncasr.ac.in

Improving the treatment of exchange and correlation effects is the holy grail in the field of electronic structure calculations, and as an aid towards achieving this goal, it is desirable to have a clear picture of the comparative merits of various approximate exchange-correlation functionals in various situations. To this end, we have performed ab initio DFT and DFPT ("linear response") calculations to investigate the thermal properties of bulk Cu, using both the local density approximation (LDA) and a generalized gradient approximation (GGA) for the exchange-correlation functional. Thermal effects are treated within the quasiharmonic approximation. In addition to obtaining better agreement with experiment than earlier, more approximate, calculations, we find a clear trend: the LDA and GGA results are closer to each other and to experiment for anharmonic properties (such as Grueneisen parameters) than for harmonic and static properties (such as phonon frequencies, the bulk modulus, and the lattice constant). We argue that this might be a general feature, and also argue that LDA/GGA errors should increase with temperature.

M. Krishnan and U V Waghmare

JNCASR, Jakkur, Bangalore, 560 064 India

Email: mkrishna@jncasr.ac.in

We perform ab initio calculations based on density functional theory (DFT) using plane wave pseudopotential method for GaN in wurtzite and zincblend structures and study the interaction between d-orbitals of Ga and s-orbital of N. The present analysis uses visualization of electron localization function and its band-by-band decomposition to study the nature of d to s bonding between the Ga and N. In addition, we calculate electron localization length using DFT linear response. The calculated lattice constant, elastic modulus and dielectric constants are in fair agreement with the experimental values.

Srikanth Sastry and C. Austen Angell

JNCASR, Jakkur, Bangalore, 560 064 India

Email: sastry@jncasr.ac.in

Silicon is a technologically important element, and a major constituent of the earth. Accordingly, it has been studied extensively in the past, but predominantly in the solid state. Silicon exhibits very interesting behaviour also in the liquid state. Evidence from experimental and computer simulation studies has been held to support the existence of a liquid to amorphous solid phase transition at roughly 1400 K, where the liquid is in a metastable, supercooled, state. Thermodynamically and structurally, silicon has much in common with other network forming liquids such as water and silica, which suggests a reexamination of the nature of this transition in analogy with these other liquids. The analogy calls for a more stringent test of the existence of a first order phase transition in silicon than is available in the literature, and further suggests that the phase transition (if present) is between two liquids which are distinguished by their densities and local geometry. Extensive simulations of silicon using the empirical potential by Stillinger and Weber are used to demonstrate that a first order phase transition exists between two forms of liquid silicon. Evidence supporting the prediction that the first order liquid-liquid transition also marks a change in the nature of the dynamics of the liquid, from that of a fragile liquid to a strong liquid, is presented. Preliminary results concerning the nucleation of the crystal phase, whose rates are very high in the vicinity of the liquid-liquid phase transition, as well as the possible location of a liquid-liquid critical point, are presented.

H. Ramanarayan and T.A. Abinandanan

Department of Metallurgy, Indian Institute of Science

Bangalore 560 012. INDIA

Email: ram@metalrg.iisc.ernet.in

We have used a phase field model to study spinodal decomposition in polycrystalline materials in which the grain size is of the same order of magnitude as the characteristic wavelength ($\lambda_{SD}$) of the early stages of phase separation. Our results show that microstructural

evolution depends crucially on the difference in the grain boundary energies $\gamma_{gb}$ of A-rich ($\alpha$) and B-rich ($\beta$) phases. If $\gamma_{gb}^{\alpha}$ is lower, decomposition is initiated with an $\alpha$ layer being formed at the grain boundary while normal spinodal decomposition takes place in the interior of the grains. If the grain size is sufficiently small (about the same as $\lambda_{SD}$), the interior of the grain is filled with the $\beta$ phase, and further coarsening takes place through Ostwald ripening of these islands of $\beta$ phase. If the grain size is large (say, about $10\lambda_{SD}$ or greater), the early stage microstructure exhibits an A-rich grain boundary layer followed by a B-rich layer; the grain interior exhibits a spinodally decomposed microstructure, evolving slowly. With increasing time, the microstrcture transforms gradually to one which consists exclusively of a set of alternating layers (concentric rings) of $\alpha$ and $\beta$ phases in each grain. During the decomposition the minimal segretation of the A - rich becomes sufficient to pin the grain boundaries leading to no grain growth.

Saswata Bhattacharyya and T.A.Abinandanan

Department of Metallurgy, Indian Institute of Science

Bangalore 560 012. INDIA

Email: saswata@metalrg.iisc.ernet.in

We have studied the microstructural evolution of quenched homogeneous disordered ternary alloys within the three-phase field using computer simulations based on multicomponent Cahn-Hilliard (CH) equations for $c_A$ and $c_B$, the compositions (in mole fraction) of A and B respectively; we have used a semi-implicit Fourier spectral method for integration of CH relative interfacial energies on temporal evolution of morphologies during spinodal phase separation of an alloy with average composition $c_A = 1/4$, $c_B=1/4$ and $c_C=1/2$; the interfacial energies between the 'A'rich, 'B'rich and 'C'rich phases are varied by changing the gradient energy coefficients. We observe that the phases associated with a higher interfacial energy are smaller. We further find that the kinetic paths (i.e., the history of A-rich, B-rich and C-rich regions in the microstructure) is also affected significantly by the relative interfacial energies of the three phases. These differences, and their consequences on the microstructural evolution and its kinetics will be discussed.

Y. Anusooya Pati, C. Raghu and S. Ramasesha

Solid State Chemistry Unit, Indian Institute of Science

Bangalore 560 012. INDIA

Email: anusooya@sscu.iisc.ernet.in

In the last two years, organic molecules such as pentacene and tetracene have been used in lasers and field-effect transistors. We have studied these organic molecules as well as the longer polyacenes, within the Density Matrix Renormalization Group (DMRG) calculations. Our study includes both the ground and excited states of polyacenes and the effect of doping on the ground state of the system, within Pariser--Parr--Pople (PPP) Hamiltonian. Because of its similarity in structure to polyacetelyene, we have addressed the question of Peierl's instability in longer polyacenes. We have obtained the ground state energy as a function of the dimerization, delta, and various correlation functions and structure factors for delta=0. From our study we find that the {\it cis} form is more stabler than the {\it trans} form and the instability is conditional unlike in the case of polyacetylene. We also present the optical gap, spin gap and charge gap for these systems as a function of system size. From the charge density data we find that the bipolaron is unstable towards the formation of two polarons.

P.S. Mohanty, B.V.R Tata, and M.C Valsakumar

Material Science Division,

Indira Gandhi Centre for Atomic Research,

Kalpakkam –603 102, Tamil Nadu.

Email: preeti@igcar.ernet.in

Constant pressure (NPT ensemble) Monte Carlo (MC) simulations have been performed on a liquid-like ordered charged colloidal suspension to study its phase behavior and dynamics under compression. We report for the first time, a liquid-like ordered monodisperse suspension undergoing glass transition (GT) at very low volume fraction (Φ=0.003) and existence of dynamical heterogeneities near the GT. Dynamical heterogeneities is characterized by identifying the mobile particles near GT. These mobile particles have been identified using the non-Gaussian parameter for the self part of the van Hope correlation function and are found to form clusters. The pressure dependence of mean cluster size and the size distribution is also discussed.

S. A. Blundell(1), D. G. Kanhere(2), A. Vichare(2) and R. Zope(1)

(1) D'{e}partement de Recherche Fondamentale surla Mati\`{e}re Condens\'{e}e CEA-Grenoble, 17 rue des Martyrs, F-38054 Grenoble CEDEX 9,

France

(2) Department of Physics, University of Pune, Pune 411007,

India

Email:

Most statistical simulations of atomic and molecular clusters have been based on semiempirical interatomic potentials such as the Lennard-Jones or embedded-atom model potentials, which have a simple analytic parametrization and are computationally cheap. However, recent experiments measuring the caloric curves of small sodium clusters \cite{Ref-1} show an irregular variation of cluster melting point with respect to cluster size, with pronounced maxima at the puzzling sizes of $N=57$ and $N=142$. This points to a subtle interplay between geometric effects and electronic shell effects, which is difficult to incorporate in a semiempirical potential, and a precise theoretical explanation is so far lacking.

In this paper we address the possibility of making an accurate extraction of statistical quantities such as the ionic entropy and the ionic specific heat using an {\em ab initio} interatomic potential derived from density functional theory. Several novel approaches will be discussed. First, extending earlier work on the temperature dependence of the polarizability, we use an $O(N)$ density-based parametrization of the electron kinetic energy, and sample the ionic phase space by means of microcanonical molecular dynamics, with a Car-Parrinello-type propagation of the electron density. The ionic entropy is extracted by multiple histogram methods. We present statistically well-converged caloric curves for clusters of size up to $N=147$. Second, we consider a Kohn-Sham approach, with both {\em ab initio} and soft, phenomenological pseudopotentials, up to size $N=20$. We find that the smaller clusters prefer a multi-step melting process, proceeding via isomerization and rearrangement processes. The larger clusters $90 < N < 147$ considered show a one-step melting process, with simultaneous melting of all ionic shells, and a corresponding pronounced single peak in the ionic specific-heat curves. Using the density-based approach, we find a pronounced maximum in the melting point at a size of $N=55$, in disagreement with the experimental maximum at $N=57$. Prospects for improving the accuracy of the density functional, and of extending the calculations to larger sizes, will be discussed.

M. Schmidt {\em et al.}, Phys.\ Rev.\ Lett.\ {\bf 79}, 99

(1997); M. Schmidt {\em et al.}, Nature {\bf 393}, 238 (1998).

Buddhapriya Chakrabarti & Chandan Dasgupta

Department of Physics,

Indian Institute of Science,

Bangalore - 560 012, India.

Email : buddho@physics.iisc.ernet.in

We study a dynamical phase transition, in a class of growth models. In these models, pyramidal structures, as has been reported in some recent experiments, are observed. As a function of the parameters entering in the models, these structures become unstable, and a rough self-affine surface results. In this work we explore this phase transition in detail.

N. Arul Murugan¹, R. S. Rao², S. Yashonath¹, S. Ramasesha¹ and B. K. Godwal²

1 Solid State & Structural Chemistry Unit, Indian Institute of Science, Bangalore - 560 012, India.

2 High Pressure Physics Division, Atomic Research Centre, Purnima Labs, Mumbai - 400 085, India.

Email : murugan@sscu.iisc.ernet.in

A study of the crystalline phase of adamantane in the pressure range 0.0001 GPa - 25 GPa is reported. Variable shape Monte Carlo simulations in the canonical ensemble suggest that the experimentally observed variation of unit cell volume is reasonably well reproduced. These calculations assumed a rigid molecule of adamantane and the potential of Gavezzotti and Filippini with some modification. Trends in the c/a ratio variation reproduce the experimentally observed behaviour.

Ahmed Sayeed¹, S. Mitra², A.V. Anil Kumar¹, R. Mukhopadhyay²,

S. Yashonath¹, S. L. Chaplot²

1Solid state and Structural Chemistry Unit, Indian Institute of Science, Bangalore-560012

2Solid State Physics Division, Bhabha Atomic Research Centre, Trombay,

Mumbai-400 085

Email : sayeed@sscu.iisc.ernet.in

We have carried out molecular dynamics(MD) simulations and quasi-elastic neutron scattering (QENS) measurements of the diffusion of propane in NaY zeolite, at temperatures 300, 324 and 350 K and a loading of 4 molecules per alpha-cage. The self-diffusivity D has been obtained from mean squared displacemnet (MSD) as well as self-intermediate scattering function F(Q,t) obtained from the MD simulation trajectories. These values are in good agreement with each other. Also, they agree quite well with the experimental QENS results. The MD results indicate a fixed jump length diffusion process, whereas, the QENS data fits well to a jump diffusion model with a Gaussian distribution of jump lenths. From MD studies as well as experimental studies we obtain some characteristic parameters associated with the jump diffusion of propane molecules, like mean jump lengths and residence times.

Toby Joseph and Chandan Dasgupta

Department of Physics

Indian Institute of Science

Bangalore - 560 012, India.

Email : toby@physics.iisc.ernet.in

The introduction of artificial pinning centres in type II superconductors changes the properties of the superconducting state and leads to interesting new phenomena. We describe here certain studies done both analyticlly and computationally on a superconducting film with a regular square array of pinning centres. The ground states of the vortex systems at different filling ratios are found using a simple geometric argument under the assumption that the penetration depth is much smaller than the spacing of the pinning lattice. The zero-temperature magnetization curve is obtained by finding the ground states for a large number of filling ratios. We also present results of a study of the

melting of the flux lattice in the presence of a regular square array of pinning centres for the case where the filling ratio is one. Our results are obtained from Monte Carlo simulations and a mean field treatment of the problem.

A. Chatterji; R. Pandit

Department of Physics,

Indian Institute of Science,

Bangalore - 560012.

India.

Email : apratim@physics.iisc.ernet.in;

We propose a lattice model for Directed, Semiflexible, Equilibrium, polymers and study its statistical mechanics by extensive Monte Carlo simulations and mappings onto related bosonic models in certain limits. We show that this model has a rich phase diagram including analogs of supersolid, Mott insulator, mass density wave. We also examine metastable states and glass formation in this model.