Atomic layer deposition of interface control layers for the fabrication of III/V MOS devices

The continued advancement of metal oxide semiconductor field effect transistor (MOSFET) technology has shifted the focus from Si/SiO2 transistors towards high-κ/III-V transistors for high performance, faster devices. This has been necessary due to the limitations associated with the scaling of the SiO2 thickness below ~1 nm and the associated increased leakage current due to direct electron tunnelling through the gate oxide. The use of these materials exhibiting lower effective charge carrier mass in conjunction with the use of a high-κ gate oxide allows for the continuation of device scaling and increases in the associated MOSFET device performance. The high-κ/III-V interface is a critical challenge to the integration of high-κ dielectrics on III-V channels. The interfacial chemistry of the high-κ/III-V system is more complex than Si, due to the nature of the multitude of potential native oxide chemistries at the surface with the resultant interfacial layer showing poor electrical insulating properties when high-κ dielectrics are deposited directly on these oxides. It is necessary to ensure that a good quality interface is formed in order to reduce leakage and interface state defect density to maximise channel mobility and reduce variability and power dissipation. In this work, the ALD growth of aluminium oxide (Al2O3) and hafnium oxide (HfO2) after various surface pre-treatments was carried out, with the aim of improving the high-κ/III-V interface by reducing the Dit – the density of interface defects caused by imperfections such as dangling bonds, dimers and other unsatisfied bonds at the interfaces of materials. A brief investigation was performed into the structural and electrical properties of Al2O3 films deposited on In0.53Ga0.47As at 200 and 300oC via a novel amidinate precursor. Samples were determined to experience a severe nucleation delay when deposited directly on native oxides, leading to diminished functionality as a gate insulator due to largely reduced growth per cycle. Aluminium oxide MOS capacitors were prepared by ALD and the electrical characteristics of GaAs, In0.53Ga0.47As and InP capacitors which had been exposed to pre-pulse treatments from triethyl gallium and trimethyl indium were examined, to determine if self-cleaning reactions similar to those of trimethyl aluminium occur for other alkyl precursors. An improved C-V characteristic was observed for GaAs devices indicating an improved interface possibly indicating an improvement of the surface upon pre-pulsing with TEG, conversely degraded electrical characteristics observed for In0.53Ga0.47As and InP MOS devices after pre-treatment with triethyl gallium and trimethyl indium respectively. The electrical characteristics of Al2O3/In0.53Ga0.47As MOS capacitors after in-situ H2/Ar plasma treatment or in-situ ammonium sulphide passivation were investigated and estimates of interface Dit calculated. The use of plasma reduced the amount of interface defects as evidenced in the improved C-V characteristics. Samples treated with ammonium sulphide in the ALD chamber were found to display no significant improvement of the high-κ/III-V interface. HfO2 MOS capacitors were fabricated using two different precursors comparing the industry standard hafnium chloride process with deposition from amide precursors incorporating a ~1nm interface control layer of aluminium oxide and the structural and electrical properties investigated. Capacitors furnished from the chloride process exhibited lower hysteresis and improved C-V characteristics as compared to that of hafnium dioxide grown from an amide precursor, an indication that no etching of the film takes place using the chloride precursor in conjunction with a 1nm interlayer. Optimisation of the amide process was carried out and scaled samples electrically characterised in order to determine if reduced bilayer structures display improved electrical characteristics. Samples were determined to exhibit good electrical characteristics with a low midgap Dit indicative of an unpinned Fermi level

Bismuth embrittlement of copper is an atomic size effect

Embrittlement by the segregation of impurity elements to grain boundaries is one of a small number of phenomena that can lead to metallurgical failure by fast fracture(1). Here we settle a question that has been debated for over a hundred years(2): how can minute traces of bismuth in copper cause this ductile metal to fail in a brittle manner? Three hypotheses for Bi embrittlement of Cu exist: two assign an electronic effect to either a strengthening(3) or weakening(4) of bonds, the third postulates a simple atomic size effect(5). Here we report first principles quantum mechanical calculations that allow us to reject the electronic hypotheses, while supporting a size effect. We show that upon segregation to the grain boundary, the large Bi atoms weaken the interatomic bonding by pushing apart the Cu atoms at the interface. The resolution of the mechanism underlying grain boundary weakening should be relevant for all cases of embrittlement by oversize impurities.

Positronium scattering by atomic hydrogen with inclusion of target excitation channels

Calculations are reported for positronium (Ps) scattering by atomic hydrogen (H) in the energy range 0-6.5 eV in a coupled- pseudostate approximation in which excitation and ionization channels of both the Ps and the H are taken into account. The approximation contains an accurate representation of the van der Waals coefficient. Results are presented for phase shifts, scattering lengths, effective ranges, and various cross sections including partial wave, total, and ortho-para conversion cross sections. An analysis of the possible spin transitions is provided and the energy of the positronium hydride (PsH) bound state is determined. Substantial differences are found from earlier work within the frozen target approximation, now clearly confirming the importance of target excitation channels. Good agreement is obtained with recent calculations of S-wave phase shifts and scattering lengths using the stabilization method. Convergence to the exact binding energy for PsH appears to be slow. Resonances corresponding to unstable states of the positron orbiting H- are seen in the electronic spin singlet partial waves. The importance of the H- formation channel is discussed.

Effective collision strengths for transitions in Fe XIII

Effective collision strengths for transitions among the lowest 97 fine-structure levels belonging to the (1s(2)2s(2)2p(6)) 3s(2)3p(2), 3s3p(3), 3s(2)3p3d, 3p(4), 3s3p(2)3d and 3s(2)3d(2) configurations of Fe XIII have been calculated using the fully relativistic Dirac Atomic R-matrix Code (DARC) of Norrington & Grant (2004). Resonances have been resolved in the threshold region, and results are reported over a wide electron temperature range up to log T-e = 6.8 K. Comparisons are made with the earlier available R-matrix results of Gupta & Tayal (1998), and the accuracy of the data is assessed.

Effective collision strengths for transitions in Fe X

Collision strengths for 4005 transitions among the lowest 90 levels of the (1s(2)2s(2)2p(6)) 3s(2)3p(5), 3s3p(6), 3s(2)3p(4)3d, 3s3p(5)3d and 3s(2)3p(3)3d(2) configurations of Fe X have been calculated using the Dirac Atomic R-matrix Code (DARC) of Norrington & Grant, over a wide energy range up to 210 Ryd. Resonances have been resolved in the threshold region, and effective collision strengths have been obtained over a wide temperature range up to 107 K. The present calculations should represent a significant improvement ( in both range and accuracy) over the earlier available results of Bhatia & Doschek and Pelan & Berrington. Based on several comparisons, the accuracy of our data is assessed to be better than 20%, for a majority of transitions.

Radiative rates, collision strengths and effective collision strengths for transitions in Mo XXXIV

Radiative rates for electric dipole (E I), electric quadrupole (E2), magnetic dipole (M1), and magnetic quadrupole (M2) transitions among the lowest 60 fine-structure levels of the (1s(2)) 2S(2)2p(5), 2s2p(6), and 2S(2)2p(4)3l configurations of F-like Mo XXXIV have been calculated using the fully relativistic GRASP code. Additionally, collision strengths for transitions among these levels have been computed over a wide energy range below 3200Ry, using the Dirac Atomic R-matrix Code. Resonances have been resolved in a fine energy mesh in order to calculate the effective collision strengths. Results for radiative rates and excitation rates are presented for all transitions, and for collision strengths for transitions from the lowest three levels to the higher lying levels. The accuracy of the present data is assessed to be similar to 20%.

Effective collision strengths for transitions in Fe XI

Collision strengths for transitions among the lowest 48 fine- structure levels belonging to the (1s(2)2s(2)2p(6)) 3s(2)3p(4), 3s3p(5), 3s(2)3p(3)3d and 3p(6) configurations of Fe XI have been calculated using the Dirac Atomic R-matrix Code (DARC) of Norrington & Grant (2003). Results are tabulated at energies above thresholds in the range 10 less than or equal to E less than or equal to 100 Ry, although resonances have been resolved in a fine energy mesh in the thresholds region. Effective collision strengths, obtained after integrating the collision strengths over a Maxwellian distribution of electron velocities, are also tabulated over a wide electron temperature range below 5 x 10(6) K. Comparisons with other available results are made, and the accuracy of the present data is assessed.

Effective collision strengths for transitions within the ground configuration of Fe XXI

Effective collision strengths for transitions among the energetically lowest 46 fine-structure levels belonging to the (1s(2)) 2S(2) 2p(2), 2s2p(3), 2p(4), 2S(2)2p3s, 2s(2)2p3p and 2S(2)2p3d configurations of Fe XXI are computed, over an electron temperature range of 5.6 less than or equal to log T-e less than or equal to 7.4 K, using the recent Dirac Atomic R- matrix Code (DARC) of Norrington and Grant. Results are presented for transitions within the ground configuration only, and are compared with earlier R matrix calculations. Large discrepancies are observed for many transitions, especially at lower temperatures.

Radiative rates, collision strengths and effective collision strengths for transitions in Gd XXXVII

Energy levels and radiative rates for transitions among 107 fine-structure levels belonging to the (1s(2)2S(2)p(6)) 3S(2)3p(6)3d(10), 3S(2)3p(6)3d(9)4e. 3S(2)3p(5)3d(10)4e. and 3s3p(6)3d(10)4e configurations of Ni-like Gd XXXVII have been calculated using the fully relativistic GRASP code. Radiative rates and oscillator strengths are tabulated for all allowed transitions among these levels. Additionally. collision strengths for transitions among the lowest 59 levels have been computed using the Dirac Atomic R-matrix Code. Resonances in the threshold region have been delineated, but results for collision strengths are tabulated only at energies above thresholds in the range 120

Effective collision strengths for transitions in Fe XV

Collision strengths for transitions among the energetically lowest 53 fine-structure levels belonging to the (1s(2)2s(2)2p(6)) 3l(2), 3l3l', 3s4l and 3p4s configurations of Fe XV are computed, over an electron energy range below 160 Ryd, using the Dirac Atomic R-matrix Code (DARC) of Norrington & Grant (2003). Effective collision strengths, obtained after integrating the collision strengths over a Maxwellian distribution of electron energies, have also been calculated. These results of effective collision strengths are tabulated for all 1378 inelastic transitions over a wide temperature range of 10(5) to 10(7) K. Comparisons are also made with other R-matrix calculations and the accuracy of the results is assessed.

Effective collision strengths for transitions among the 3s(2), 3s3p and 3p(2) configurations of Fe xv

Effective collision strengths for transitions among the ten energetically lowest fine-structure levels belonging to the (1s(2)2s(2)2p(6))3s(2), 3s3p and 3p(2) configurations of Fe xv have been calculated in the electron temperature range of 10(5)-10(7) K, using the recent Dirac atomic R-matrix code of Norrington and Grant. The results are compared with the other recently available independent Breit-Pauli R-matrix calculations of Eissner et al (Eissner W, Galavis M E, Mendoza C and Zeippen C J 1999 Astron. Astrophys. Suppl. 137 165) and Griffin et al (Griffin DC, Badnell N R, Pindzola M S and Shaw J A 1999 J. Phys. B: At. Mol. Opt. Phys. 32 2139, 4129). Large differences are observed for many transitions over almost the entire temperature range. These differences are analysed and discussed, and the accuracy of the calculations is assessed.

Effective collision strengths for electron impact excitation of C1III

Effective collision strengths for electron-impact excitation of the phosphorus-like ion Cl III are presented for all fine- structure transitions among the levels arising from the lowest 23 LS states. The collisional cross sections are computed in the multichannel close-coupling R-matrix approximation, where sophisticated configuration-interaction wave functions are used to represent the target states. The 23 LS states are formed from the basis configurations 3s(2)3p(3). 3s3p(4). 3s(2)3p(2)3d, and 3s(2)3p(2)4s, and correspond to 49 fine- structure levels, leading to a total possible 1176 fine- structure transitions. The effective collision strengths. obtained by averaging the electron collision strengths over a Maxwellian distribution of electron velocities. are tabulated in this paper for all 1176 transitions and for electron temperatures in the ranges T(K) = 7500-25.000 and log T(K) = 4.4-5.3. The former range encompasses the temperatures of particular importance for application to gaseous nebulae. while the latter range is more applicable to the study of solar and laboratory-type plasmas. (C) 2001 Academic Press.

Effective collision strengths for electron impact excitation of Si VIII

Effective collision strengths for electron-impact excitation of the nitrogen-like ion Si VIII are presented over the wide range of electron temperatures log T(K) = 4.0-6.5. All 231 fine- structure transitions among the 22 fine-structure levels arising from the lowest 11 LS target states (2s(2)2p(3), 2s2p(4), 2p(5), and 2s(2)2p(2)3s) are considered in the tabulation. The collision strengths are evaluated in a multi- channel R-matrix approach, and the corresponding effective collision strengths are obtained by averaging these over a Maxwellian distribution of electron velocities. Comparisons are made with recent distorted-wave results at high incident electron energies. Differences of up to 20% are found, particularly for some allowed transitions. (C) 2003 Elsevier Inc. All rights reserved.

Dynamical and steady-state requirements for effective entanglement transfer

Entanglement transfer processes from a continuous-variable (CV) to a qubit system have primary importance in quantum information processing due to some practical implications in the realization of a quantum network. A CV system can propagate entanglement while a qubit system is easy to manipulate. We study conditions to entangle two atomic qubits with a two-mode squeezed field driving two cavities containing the atoms.