Saturday, September 8, 2012
Wednesday, September 5, 2012
Bandstructure and DoS calculation
Some basic understanding that deserved to be mentioned here:
1. Bandstructure calculation requires the system has translational invariance, which means the system should be a infinite system with certain lattice symmetry, otherwise there is now meaning for the definition of momentum . So basically, we should calculate the bandstructure of the studied system in a 'lead' or at least should treat it as a lead.
2 DoS is by the same argument should be calculate inside a lead. And the system should not be 'finalize()' before your calculation of DoS or bandstructure.
1. Bandstructure calculation requires the system has translational invariance, which means the system should be a infinite system with certain lattice symmetry, otherwise there is now meaning for the definition of momentum . So basically, we should calculate the bandstructure of the studied system in a 'lead' or at least should treat it as a lead.
2 DoS is by the same argument should be calculate inside a lead. And the system should not be 'finalize()' before your calculation of DoS or bandstructure.
Tuesday, September 4, 2012
Collection of Questions for Poster section for the windsor summer
Q1: Why gap closing does not observed in your dI/dV curve as those observed in the LDOS?
A1: As the paper"T. Stanescu et al, arXiv:1206.0013v1(2012)", the LDOS cannot fully represent the whole information of the system. And as you see the joining of the peak in the LDOS in the paper mentioned about is parameters dependent. So if we can find the good parameter, we believe we can also see the process of joining of the peaks.
Q2: Disorder open SC gap, and make SC proximity effect large, why?
A2: According to the paper "P. G. de Gennes and D. Saint-James, Phys. Letter 4, 151(1963)" and "C. W. J. Bennakker, Lect. Notes. Phys. 667, 131(2005)", the formula showed in the poster. When there are disorder impurities in the system, they will make the scattering more intense and thus the long rang scattering which has large parallel momentum is suppressed. Only left the scattering that are nearly perpendicular to the SC-NW interfaces. Thus leading to the small T(p||), and thus large E_gap.
Q3: What could it be if we add a SC term directly in H_1D instead of writing a BdG Hamiltonian?
A3: For the SC Hamiltonian, it is always has the form of BdG.
Q4: What makes the Rashba double parabolic band splited?
A4: The Zeemam terms
Q5: Change chemical potential will bring what change?
A5: Chemical potential is always in the gap. However, you can always change chemical potential in your numerical calculation and those change gate voltage in the experiment. If the system has multi-subband, you can tune your chemical potential within each subband (which is formed due to the confinement).
Q6: What are the many other peaks? Why they form? And what does disorder do with them?
A6: The other peaks are due to multi-band, or modes, or channels which will essentially bring to different Andreev-reflections. We conclude that weak disorder will kill the other modes(peaks), and open the gap.
Q7: What is the difference between V_bias, and V_gate?
A7: V_bias are the voltage difference between Normal lead and SC lead. However, V_gate is the applied electric field control the chemical potential in the NW system. They are quiet different concepts.
A1: As the paper"T. Stanescu et al, arXiv:1206.0013v1(2012)", the LDOS cannot fully represent the whole information of the system. And as you see the joining of the peak in the LDOS in the paper mentioned about is parameters dependent. So if we can find the good parameter, we believe we can also see the process of joining of the peaks.
Q2: Disorder open SC gap, and make SC proximity effect large, why?
A2: According to the paper "P. G. de Gennes and D. Saint-James, Phys. Letter 4, 151(1963)" and "C. W. J. Bennakker, Lect. Notes. Phys. 667, 131(2005)", the formula showed in the poster. When there are disorder impurities in the system, they will make the scattering more intense and thus the long rang scattering which has large parallel momentum is suppressed. Only left the scattering that are nearly perpendicular to the SC-NW interfaces. Thus leading to the small T(p||), and thus large E_gap.
Q3: What could it be if we add a SC term directly in H_1D instead of writing a BdG Hamiltonian?
A3: For the SC Hamiltonian, it is always has the form of BdG.
Q4: What makes the Rashba double parabolic band splited?
A4: The Zeemam terms
Q5: Change chemical potential will bring what change?
A5: Chemical potential is always in the gap. However, you can always change chemical potential in your numerical calculation and those change gate voltage in the experiment. If the system has multi-subband, you can tune your chemical potential within each subband (which is formed due to the confinement).
Q6: What are the many other peaks? Why they form? And what does disorder do with them?
A6: The other peaks are due to multi-band, or modes, or channels which will essentially bring to different Andreev-reflections. We conclude that weak disorder will kill the other modes(peaks), and open the gap.
Q7: What is the difference between V_bias, and V_gate?
A7: V_bias are the voltage difference between Normal lead and SC lead. However, V_gate is the applied electric field control the chemical potential in the NW system. They are quiet different concepts.
Subscribe to:
Posts (Atom)