Electronic structure of graphene nanoribbons: Tight-binding versus density functional theory methods: Difference between revisions

From phys824
Jump to navigationJump to search
(Created page with ''''Electronic structure of graphene nanoribbons (deadline: 11/23 at midnight):''' The project explores recently discovered graphene nanoribbons (GNRs) by computing their electron…')
 
 
(18 intermediate revisions by the same user not shown)
Line 1: Line 1:
'''Electronic structure of graphene nanoribbons (deadline: 11/23 at midnight):''' The project explores recently discovered graphene nanoribbons (GNRs) by computing their electronic structure as  equilibrium property using simple tight-binding method and more advanced density functional theory codes.
The project explores recently discovered graphene nanoribbons (GNRs) by computing their electronic structure as  equilibrium property using simple tight-binding method (as implemented in [https://kwant-project.org/ KWANT], [http://www.physics.rutgers.edu/pythtb/ PythTB] or your own Matlab script) and more advanced density functional theory codes (as implemented in [http://www.quantum-espresso.org/ Quantum ESPRESSO] or [https://wiki.fysik.dtu.dk/gpaw/ GPAW] packages).


== Subband structure of armchair GNR ==
== Subband structure of armchair GNR ==
Using the nearest-neighbor tight-binding Hamiltonian with single <math> p_z </math> orbital per carbon atom, compute the subband structure of three armchair GNRs whose width is <math> N_a=4,5, 30</math>. The expected result is shown in Lecture 9. Pay attention to select the proper interval of <math> k_x </math> values as the first 1D Brillouin zone.
Using the nearest-neighbor tight-binding Hamiltonian with single <math> p_z </math> orbital per carbon atom, compute the subband structure of three armchair GNRs whose width is <math> N_a=4,5, 30</math>. The expected result is shown in [[Media:gnr_and_cnt.pdf|Lecture 7]]. Pay attention to select the proper interval of <math> k_x </math> values as the first 1D Brillouin zone.


== Subband structure of zigzag GNR ==
== Subband structure of zigzag GNR ==
Using the nearest-neighbor (<math> t_1=2.7 </math> eV) tight-binding Hamiltonian with single <math> p_z </math> orbital per carbon atom, compute the subband structure of three zigzag GNRs whose width is <math> N_z=4,5, 30</math>. The expected result is shown in Lecture 9. Plot the amplitude squared <math> |\chi(y)|^2 </math> across <math> N_z=30</math> ZGNR as the transverse part of a selected eigenfunction  (i.e., conducting channel)  whose eigenergy is close to the Dirac point <math> E=10^{-3} \gamma </math>. This plot should show that probability to find electron peaks around the nanoribbons edges.
Using the nearest-neighbor (<math> t_1=2.7 </math> eV) tight-binding Hamiltonian with single <math> p_z </math> orbital per carbon atom, compute the subband structure of three zigzag GNRs whose width is <math> N_z=4,5, 30</math>. The expected result is shown in [[Media:gnr_and_cnt.pdf|Lecture 7]]. Plot the amplitude squared <math> |\chi(y)|^2 </math> across <math> N_z=30</math> ZGNR as the transverse part of a selected eigenfunction  (i.e., conducting channel)  whose eigenenergy is close to the Dirac point <math> E=10^{-3} t_1 </math>. This plot should show that probability to find electron peaks around the nanoribbons edges.


== Subband structure of AGNR and ZGNR using third-nearest neighbor hoppings ==
== Subband structure of AGNR and ZGNR using third-nearest neighbor hoppings ==
Repeat subband structure calculations for <math> N_a=5 </math> AGNR and <math> N_z=5 </math> ZGNR using the tight-binding Hamiltonian which includes up to third-nearest neighbour hoppings whose values are: <math> t_1=2.7 </math> eV,  <math> t_2 = 0.20 </math> eV, and <math> t_3 = 0.18</math> eV. Comment on the difference between this result and  a) and b).
Repeat subband structure calculations for <math> N_a=5 </math> AGNR and <math> N_z=5 </math> ZGNR using the tight-binding Hamiltonian which includes up to third-nearest neighbour hoppings whose values are: <math> t_1=2.7 </math> eV,  <math> t_2 = 0.20 </math> eV, and <math> t_3 = 0.18</math> eV. Comment on the difference between this result and  a) and b).


== Subband structure of GNRs as topological insulator below 0.5 K==
== Subband structure of GNRs with spin-orbit coupling as topological insulator below 0.5 K==
At low temperatures, the energy band gap <math> \approx  0.5 K </math> due to intrinsic spin-orbit coupling, as well as the chiral spin-filtered edge states whose subbands pass through the gap, should become visible in experiments. This systems, termed topological insulator  
At low temperatures, the energy band gap <math> \approx  0.5 K </math> due to intrinsic spin-orbit coupling of graphene, as well as the chiral spin-filtered edge states whose subbands pass through the gap, should become visible in experiments. This systems, termed topological insulator where  time-reversal invariance ensures the crossing of the energy levels  at special points in the Brillouin zone so that their energy spectrum cannot be adiabatically deformed into topologically trivial insulator without such states, can be studied using the following tight-binding model:
where  time-reversal invariance ensures the crossing of the energy levels  at special points in the Brillouin zone so that their energy spectrum cannot be adiabatically deformed into topologically trivial insulator without such states, can be studied using the following tight-binding model:


<math> \hat{H}_{\mathrm{TI}} = - t_1 \sum_{\langle ij \rangle} \hat{c}_{i}^\dagger \hat{c}_{j}
<math> \hat{H}_{\mathrm{TI}} = - t_1 \sum_{\langle ij \rangle} \hat{c}_{i}^\dagger \hat{c}_{j}
+ \frac{2i}{\sqrt{3}} t_{\mathrm{SO}} \sum_{\langle \langle ij \rangle \rangle} \hat{c}_i^\dagger \hat{\mathbf{\sigma}} \cdot (\mathbf{d}_{kj} \times \mathbf{d}_{ik})\hat{c}_j </math> (1)  
+ \frac{2i}{\sqrt{3}} t_{\mathrm{SO}} \sum_{\langle \langle ij \rangle \rangle} \hat{c}_i^\dagger \hat{\mathbf{\sigma}} \cdot (\mathbf{d}_{kj} \times \mathbf{d}_{ik})\hat{c}_j </math> (1)  


where <math> \hat{c}^\dagger_i = (\hat{c}^\dagger_{i\uparrow}, \hat{c}^\dagger_{i\downarrow}) </math> are electron creation operators and <math> \hat{\mathbf{\sigma}} </math> is the vector of the Pauli matrices. The third sum in Eq.~(1)  introduces the intrinsic SO coupling compatible with the symmetries of the honeycomb lattice. The SO coupling acts as spin-dependent next-nearest neighbor hopping where <math> i </math> and <math> j </math> are two next-nearest neighbor sites, <math> k </math>  is the only common nearest neighbor of <math> i </math> and <math> j </math>, and <math> \mathbf{d}_{ik} </math> is a vector pointing from  <math> k </math> to <math> i </math>. Compute the band structure of <math> N_z=30</math> ZGNR with SO coupling described by Hamiltonian (1) assuming <math> t_1 = 2.7 </math> eV and  <math> t_{\mathrm{SO}}=0.03 t_1 </math>. The value for <math> t_{\mathrm{SO}} </math> is selected to be much larger than the realistic one in order to see clearly opening of the band gap  <math> \Delta_{\rm SO}=6\sqrt{3} t_{\rm SO} </math> in your figure.
where <math> \hat{c}^\dagger_i = (\hat{c}^\dagger_{i\uparrow}, \hat{c}^\dagger_{i\downarrow}) </math> are electron creation operators on the honeycomb lattice of GNR and <math> \hat{\mathbf{\sigma}} </math> is the vector of the Pauli matrices. The second term in Eq. (1)  introduces the intrinsic SO coupling compatible with the symmetries of the honeycomb lattice. The SO coupling acts as spin-dependent next-nearest neighbor hopping where <math> i </math> and <math> j </math> are two next-nearest neighbor sites, <math> k </math>  is the only common nearest neighbor of <math> i </math> and <math> j </math>, and <math> \mathbf{d}_{ik} </math> is a vector pointing from  <math> k </math> to <math> i </math>. Compute the band structure of <math> N_z=30</math> ZGNR with SO coupling described by Hamiltonian (1) assuming <math> t_1 = 2.7 </math> eV and  <math> t_{\mathrm{SO}}=0.03 t_1 </math>. The value for <math> t_{\mathrm{SO}} </math> is selected to be much larger than the realistic one in order to see clearly opening of the band gap  <math> \Delta_{\rm SO}=6\sqrt{3} t_{\rm SO} </math> in your figure. Your result should look the same as Fig. 1 in  [http://link.aps.org/doi/10.1103/PhysRevLett.95.226801 Phys. Rev. Lett. '''95''', 226801 (2005)].


== Subband structure of GNRs using DFT ==
== Subband structure of GNRs using DFT ==
Using DFT code [http://www.icmab.es/siesta/tutorials SIESTA] (installed of fermi), compute subband structure for <math> N_a=5 </math> AGNR and <math> N_z=5 </math> ZGNR and comment on difference or similarities with your results in 1., 2., and 3.
Using DFT code, Quantum Espresso or GPAW introduced in the [[Computer Lab]], compute subband structure for <math> N_a=5 </math> AGNR and <math> N_z=5 </math> ZGNR and comment on difference or similarities with your results in 1., 2., and 3.
 
===Hands-on Instructions===
 
* How to get band structure with SIESTA (on fermi) in three steps:
1. Run the code: siesta < zgnr.fdf | tee zgnr.out
2. Prepare the data for plotting: gnubands < ZGNR.bands | tee bands.dat
3. Plot bands.dat with xmgrace: xmgrace bands.dat &
 
*[http://www.physics.udel.edu/~bnikolic/QTTG/shared/docs/how_to_run_siesta.pdf How to run SIESTA]
*fdf files for bulk silicon:
**[http://www.mcc.uiuc.edu/summerschool/2005/tutorial/SIESTA/Example1.pdf Bulk Silicon Instructions]
**[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/silicon/silicon.long.fdf silicon.long.fdf]
**[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/silicon/Si.psf Si.psf]
*fdf files for bulk graphene and carbon nanotubes:
**[http://www.mcc.uiuc.edu/summerschool/2005/tutorial/SIESTA/Example4.pdf Graphene and Carbon Nanotubes Instructions]
**graphene:
***[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/graphene/graphene.fdf graphene.fdf]
***[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/graphene/C.psf C.psf]
***[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/graphene/basis.fdf basis.fdf]
**(3,3) carbon nanotube
***[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/cnt/tube.fdf tube.fdf]
***[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/cnt/C.psf C.psf]
***[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/cnt/basis.fdf basis.fdf]
*fdf file for hydrogen-terminated <math> N_z=5 </math> zigzag graphene nanoribbon (5-ZGNR):
**[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/gnr/zgnr.fdf zgnr.fdf]
**[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/gnr/C.psf C.psf]
**[http://www.physics.udel.edu/~bnikolic/teaching/phys824/SIESTA/gnr/H.psf H.psf]


== References ==
== References ==


*MAIN REFERENCE: A. Cresti, N. Nemec, B. Biel, G. Niebler, F. Triozon, G. Cuniberti, and S. Roche, ''Charge transport in disordered graphene-based low-dimensional materials'', Nano Research  '''1''', 361 (2008). [http://www-drfmc.cea.fr/Phocea/file.php?class=pisp&reload=1226481222&file=sroche/files/22/22_88_.pdf [PDF]].
*'''Main reference:''' A. Cresti, N. Nemec, B. Biel, G. Niebler, F. Triozon, G. Cuniberti, and S. Roche, ''Charge transport in disordered graphene-based low-dimensional materials'', Nano Research  '''1''', 361 (2008). [http://www-drfmc.cea.fr/Phocea/file.php?class=pisp&reload=1226481222&file=sroche/files/22/22_88_.pdf [PDF]].


*REFERENCE FOR GNR AS TOPOLOGICAL INSULATOR: C. K. Kane and E. J. Mele, ''Quantum spin Hall effect in graphene'', [http://link.aps.org/doi/10.1103/PhysRevLett.95.226801 Phys. Rev. Lett. '''95''', 226801 (2005)].
* '''Reference for GNR as 2D topological insulator:''' C. K. Kane and E. J. Mele, ''Quantum spin Hall effect in graphene'', [http://link.aps.org/doi/10.1103/PhysRevLett.95.226801 Phys. Rev. Lett. '''95''', 226801 (2005)].

Latest revision as of 13:55, 14 November 2016

The project explores recently discovered graphene nanoribbons (GNRs) by computing their electronic structure as equilibrium property using simple tight-binding method (as implemented in KWANT, PythTB or your own Matlab script) and more advanced density functional theory codes (as implemented in Quantum ESPRESSO or GPAW packages).

Subband structure of armchair GNR

Using the nearest-neighbor tight-binding Hamiltonian with single orbital per carbon atom, compute the subband structure of three armchair GNRs whose width is . The expected result is shown in Lecture 7. Pay attention to select the proper interval of values as the first 1D Brillouin zone.

Subband structure of zigzag GNR

Using the nearest-neighbor ( eV) tight-binding Hamiltonian with single orbital per carbon atom, compute the subband structure of three zigzag GNRs whose width is . The expected result is shown in Lecture 7. Plot the amplitude squared across ZGNR as the transverse part of a selected eigenfunction (i.e., conducting channel) whose eigenenergy is close to the Dirac point . This plot should show that probability to find electron peaks around the nanoribbons edges.

Subband structure of AGNR and ZGNR using third-nearest neighbor hoppings

Repeat subband structure calculations for AGNR and ZGNR using the tight-binding Hamiltonian which includes up to third-nearest neighbour hoppings whose values are: eV, eV, and eV. Comment on the difference between this result and a) and b).

Subband structure of GNRs with spin-orbit coupling as topological insulator below 0.5 K

At low temperatures, the energy band gap due to intrinsic spin-orbit coupling of graphene, as well as the chiral spin-filtered edge states whose subbands pass through the gap, should become visible in experiments. This systems, termed topological insulator where time-reversal invariance ensures the crossing of the energy levels at special points in the Brillouin zone so that their energy spectrum cannot be adiabatically deformed into topologically trivial insulator without such states, can be studied using the following tight-binding model:

(1)

where are electron creation operators on the honeycomb lattice of GNR and is the vector of the Pauli matrices. The second term in Eq. (1) introduces the intrinsic SO coupling compatible with the symmetries of the honeycomb lattice. The SO coupling acts as spin-dependent next-nearest neighbor hopping where and are two next-nearest neighbor sites, is the only common nearest neighbor of and , and is a vector pointing from to . Compute the band structure of ZGNR with SO coupling described by Hamiltonian (1) assuming eV and . The value for is selected to be much larger than the realistic one in order to see clearly opening of the band gap in your figure. Your result should look the same as Fig. 1 in Phys. Rev. Lett. 95, 226801 (2005).

Subband structure of GNRs using DFT

Using DFT code, Quantum Espresso or GPAW introduced in the Computer Lab, compute subband structure for AGNR and ZGNR and comment on difference or similarities with your results in 1., 2., and 3.

References

  • Main reference: A. Cresti, N. Nemec, B. Biel, G. Niebler, F. Triozon, G. Cuniberti, and S. Roche, Charge transport in disordered graphene-based low-dimensional materials, Nano Research 1, 361 (2008). [PDF].