prezentacja programu powerpointszczytko/ldsn/2_ldsn_2015_tight_binding.pdfbloch function has a form:...

Post on 16-Jul-2020

4 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Tight-Binding Approximation

Faculty of Physics UW

Jacek.Szczytko@fuw.edu.pl

A wave packet

2015-11-27 2S. Kryszewski Gdańsk 2010

A wave packet

2015-11-27 3

Note that

is a fourrier transform of

Gaussian packet:

S. Kryszewski Gdańsk 2010

A wave packet

2015-11-27 4

Gaussian packet:

S. Kryszewski Gdańsk 2010

A wave packet

2015-11-27 5

Derivation:

S. Kryszewski Gdańsk 2010

A wave packet

2015-11-27 6S. Kryszewski Gdańsk 2010

A wave packet

2015-11-27 7

Ordering expression:

A wave packet

2015-11-27 8

Ordering expression:

S. Kryszewski Gdańsk 2010

Pakiet falowy

2015-11-27 9

A wave packet

2015-11-27 10

𝑚𝑒 = 9.11 10−31kgℎ = 6.626 10−34 Js = 4.136 10−15 eV sℏ = 1.055 10−34 J s = 6.582 10−16 eV s𝑒 = −1.602 10−19 C

Δ𝑥 Δ𝑝 ≥ℏ

2

A wave packet

2015-11-27 11

A wave packet

2015-11-27 12

foton

elektron

Mass particle (electron) and massless (photon)

The band theory of solids.

2015-11-27 13

2015-11-27 14

The solution of the one-electron Schrödinger equation for a periodic potential has a form of modulated plane wave:

𝑢𝑛,𝑘 Ԧ𝑟 = 𝑢𝑛,𝑘 Ԧ𝑟 + 𝑅

𝜑𝑛,𝑘 Ԧ𝑟 = 𝑢𝑛,𝑘 Ԧ𝑟 𝑒𝑖𝑘 Ԧ𝑟

We introduced coefficient 𝑛 for different solutions corresponding to the same 𝑘 (index). 𝑘-vector is an element of the first Brillouin zone.

Bloch wave,Bloch function

Bloch amplitude,Bloch envelope

𝑢𝑛,𝑘 Ԧ𝑟 =

Ԧ𝐺

𝐶𝑘− Ԧ𝐺𝑒𝑖 Ԧ𝐺 Ԧ𝑟

Bloch theorem

Periodic potential

2015-11-27 15

Brillouin zones

Periodic potential

𝑅 = 𝑛1 Ԧ𝑎1 + 𝑛2 Ԧ𝑎2 + 𝑛3 Ԧ𝑎3, 𝑛𝑖 ∈ ℤԦ𝐺 = 𝑚1 Ԧ𝑎1

∗ +𝑚2 Ԧ𝑎2∗ +𝑚3 Ԧ𝑎3

∗ , 𝑚𝑖 ∈ ℤ

Ԧ𝑎𝑖∗ Ԧ𝑎𝑗 = 2𝜋𝛿𝑖𝑗

http://www.doitpoms.ac.uk/tlplib/brillouin_zones/zone_construction.php

2-D square lattice

2015-11-27 16

𝑅 = 𝑛1 Ԧ𝑎1 + 𝑛2 Ԧ𝑎2 + 𝑛3 Ԧ𝑎3, 𝑛𝑖 ∈ ℤԦ𝐺 = 𝑚1 Ԧ𝑎1

∗ +𝑚2 Ԧ𝑎2∗ +𝑚3 Ԧ𝑎3

∗ , 𝑚𝑖 ∈ ℤ

Ԧ𝑎𝑖∗ Ԧ𝑎𝑗 = 2𝜋𝛿𝑖𝑗

http://www.doitpoms.ac.uk/tlplib/brillouin_zones/zone_construction.php

Brillouin zones

Periodic potential

2-D square lattice

2015-11-27 17

𝑅 = 𝑛1 Ԧ𝑎1 + 𝑛2 Ԧ𝑎2 + 𝑛3 Ԧ𝑎3, 𝑛𝑖 ∈ ℤԦ𝐺 = 𝑚1 Ԧ𝑎1

∗ +𝑚2 Ԧ𝑎2∗ +𝑚3 Ԧ𝑎3

∗ , 𝑚𝑖 ∈ ℤ

Ԧ𝑎𝑖∗ Ԧ𝑎𝑗 = 2𝜋𝛿𝑖𝑗

http://www.doitpoms.ac.uk/tlplib/brillouin_zones/zone_construction.php

Brillouin zones

Periodic potential

2-D square lattice

2015-11-27 18

𝑅 = 𝑛1 Ԧ𝑎1 + 𝑛2 Ԧ𝑎2 + 𝑛3 Ԧ𝑎3, 𝑛𝑖 ∈ ℤԦ𝐺 = 𝑚1 Ԧ𝑎1

∗ +𝑚2 Ԧ𝑎2∗ +𝑚3 Ԧ𝑎3

∗ , 𝑚𝑖 ∈ ℤ

Ԧ𝑎𝑖∗ Ԧ𝑎𝑗 = 2𝜋𝛿𝑖𝑗

http://www.doitpoms.ac.uk/tlplib/brillouin_zones/zone_construction.php

Brillouin zones

Periodic potential

2-D hexagonal lattice

2015-11-27 19

𝑅 = 𝑛1 Ԧ𝑎1 + 𝑛2 Ԧ𝑎2 + 𝑛3 Ԧ𝑎3, 𝑛𝑖 ∈ ℤԦ𝐺 = 𝑚1 Ԧ𝑎1

∗ +𝑚2 Ԧ𝑎2∗ +𝑚3 Ԧ𝑎3

∗ , 𝑚𝑖 ∈ ℤ

Ԧ𝑎𝑖∗ Ԧ𝑎𝑗 = 2𝜋𝛿𝑖𝑗

http://www.doitpoms.ac.uk/tlplib/brillouin_zones/zone_construction.php

Brillouin zones

Periodic potential

http://oen.dydaktyka.agh.edu.pl/dydaktyka/fizyka/c_teoria_pasmowa/2.php

Brillouin zone for face centered cubic (fcc) lattice. The limiting zone walls comes from reciprocal latticepoints (2,0,0) square and (1,1,1) hexagonal.

Brillouin zone in 1D

Brillouin zone in 2D, oblique lattice.

2015-11-27 20

𝑅 = 𝑛1 Ԧ𝑎1 + 𝑛2 Ԧ𝑎2 + 𝑛3 Ԧ𝑎3, 𝑛𝑖 ∈ ℤԦ𝐺 = 𝑚1 Ԧ𝑎1

∗ +𝑚2 Ԧ𝑎2∗ +𝑚3 Ԧ𝑎3

∗ , 𝑚𝑖 ∈ ℤ

Brillouin zones

Periodic potential

Ibach, Luth

2015-11-27 21

bcc lattice

heksagonal lattice

𝑅 = 𝑛1 Ԧ𝑎1 + 𝑛2 Ԧ𝑎2 + 𝑛3 Ԧ𝑎3, 𝑛𝑖 ∈ ℤԦ𝐺 = 𝑚1 Ԧ𝑎1

∗ +𝑚2 Ԧ𝑎2∗ +𝑚3 Ԧ𝑎3

∗ , 𝑚𝑖 ∈ ℤ

Brillouin zones

Periodic potential

Potencjał periodyczny

2015-11-27 22

Yu, Cardona Fundametals of semiconductors

Bloch function has a form:

Periodic function, so-called Bloch factor

Generally non-periodic function

Example: electron in a constant potential

substituting 𝜑𝑛,𝑘 Ԧ𝑟 = 1 𝑒𝑖𝑘 Ԧ𝑟

The solution is

The momentum operator Ƹ𝑝 = −𝑖ℏ𝛻 acting on 𝜑𝑛,𝑘 Ԧ𝑟

Ƹ𝑝𝜑𝑛,𝑘 Ԧ𝑟 = ℏ𝑘 𝜑𝑛,𝑘 Ԧ𝑟 . The solutions of the Schrödinger equation with a constant potential

are eigenfunctions of the momentum operator. The momentum is well defined, the eigenvalue

of the momentum operator is Ƹ𝑝 = ℏ𝑘 (this defines the sense of 𝑘-vector).

2015-11-27 23

𝐻 = −ℏ2

2𝑚Δ + 𝑉

𝜑𝑛,𝑘 Ԧ𝑟 = 𝑢𝑛,𝑘 Ԧ𝑟 𝑒𝑖𝑘 Ԧ𝑟

𝐸 =ℏ2𝑘2

2𝑚+ 𝑉

Bloch theorem

Periodic potential

Przykład:Electron motion in a periodic potential.

Thus:

The solution is:

Applying Ƹ𝑝 = −𝑖ℏ𝛻 we get Ƹ𝑝𝜓 Ԧ𝑟 = −𝑖ℏ 𝑖 𝑘 + 𝛻𝑢𝑛,𝑘 𝑒𝑖𝑘 Ԧ𝑟 ≠ ℏ𝑘𝜓 Ԧ𝑟 .

Momentum of the Bloch function is not well defined!

ℏ𝑘 is so-called quasi-momentum or crystal momentum.

2015-11-27 24

𝑉 Ԧ𝑟 =

Ԧ𝐺

𝑉Ԧ𝐺 exp 𝑖 Ԧ𝐺 Ԧ𝑟

𝜓𝑛,𝑘 Ԧ𝑟 = 𝑢𝑛,𝑘 Ԧ𝑟 𝑒𝑖𝑘 Ԧ𝑟

𝑢𝑛,𝑘 Ԧ𝑟 =

Ԧ𝐺

𝐶𝑘− Ԧ𝐺𝑒𝑖 Ԧ𝐺 Ԧ𝑟

Bloch theorem

Periodic potential

Przykład:Electron motion in a periodic potential.

Thus:

The solution is:

2015-11-27 25

𝑉 Ԧ𝑟 =

Ԧ𝐺

𝑉Ԧ𝐺 exp 𝑖 Ԧ𝐺 Ԧ𝑟

𝜓𝑛,𝑘 Ԧ𝑟 = 𝑢𝑛,𝑘 Ԧ𝑟 𝑒𝑖𝑘 Ԧ𝑟

𝑢𝑛,𝑘 Ԧ𝑟 =

Ԧ𝐺

𝐶𝑘− Ԧ𝐺𝑒𝑖 Ԧ𝐺 Ԧ𝑟

Bloch theorem

Periodic potential

LxLy

Lz

If our crystal has a finite size set of vectors 𝑘 is finite (though enormous!). for instance, we can assume periodic boundary conditions and then:

𝑘𝑖 = 0,±2𝜋

𝐿𝑖, ±

4𝜋

𝐿𝑖, ±

6𝜋

𝐿𝑖, … , ±

2𝜋𝑛𝑖𝐿𝑖

Bloch functions whose wave vectors differ by a reciprocal lattice vector, are the same!

Proof:

What about energies?

2015-11-27 26

Bloch theorem

Periodic potential

𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 = 𝜓𝑛,𝑘 Ԧ𝑟 Ԧ𝐺 = ℎ Ԧ𝑔1 + 𝑘 Ԧ𝑔2 + 𝑙 Ԧ𝑔3

𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 = 𝑢𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 𝑒𝑖(𝑘+ Ԧ𝐺) Ԧ𝑟 =

Ԧ𝐺′

𝐶 𝑘 + Ԧ𝐺 − Ԧ𝐺′ 𝑒−𝑖 Ԧ𝐺′ Ԧ𝑟 𝑒𝑖(𝑘+ Ԧ𝐺) Ԧ𝑟 = ⋯

=

Ԧ𝐺′′

𝐶 𝑘 − Ԧ𝐺′′ 𝑒−𝑖 Ԧ𝐺′′ Ԧ𝑟 𝑒𝑖(𝑘) Ԧ𝑟 = 𝜓𝑛,𝑘 Ԧ𝑟

Ԧ𝑝2

2𝑚0+ 𝑉 Ԧ𝑟 𝜓𝑛,𝑘 Ԧ𝑟 = 𝐸 𝑛, 𝑘 𝜓𝑛,𝑘 Ԧ𝑟

Ԧ𝑝2

2𝑚0+ 𝑉 Ԧ𝑟 𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 = 𝐸 𝑛, 𝑘 + Ԧ𝐺 𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟

Bloch functions whose wave vectors differ by a reciprocal lattice vector, are the same!

Proof:

What about energies?

2015-11-27 27

Bloch theorem

Periodic potential

Ԧ𝑝2

2𝑚0+ 𝑉 Ԧ𝑟 𝜓𝑛,𝑘 Ԧ𝑟 = 𝐸 𝑛, 𝑘 𝜓𝑛,𝑘 Ԧ𝑟

Ԧ𝑝2

2𝑚0+ 𝑉 Ԧ𝑟 𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 = 𝐸 𝑛, 𝑘 + Ԧ𝐺 𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟

𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 = 𝑢𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 𝑒𝑖(𝑘+ Ԧ𝐺) Ԧ𝑟 =

Ԧ𝐺′

𝐶 𝑘 + Ԧ𝐺 − Ԧ𝐺′ 𝑒−𝑖 Ԧ𝐺′ Ԧ𝑟 𝑒𝑖(𝑘+ Ԧ𝐺) Ԧ𝑟 = ⋯

=

Ԧ𝐺′′

𝐶 𝑘 − Ԧ𝐺′′ 𝑒−𝑖 Ԧ𝐺′′ Ԧ𝑟 𝑒𝑖(𝑘) Ԧ𝑟 = 𝜓𝑛,𝑘 Ԧ𝑟

𝜓𝑛,𝑘+ Ԧ𝐺 Ԧ𝑟 = 𝜓𝑛,𝑘 Ԧ𝑟 Ԧ𝐺 = ℎ Ԧ𝑔1 + 𝑘 Ԧ𝑔2 + 𝑙 Ԧ𝑔3

⇒ 𝐸 𝑛, 𝑘 = 𝐸 𝑛, 𝑘 + Ԧ𝐺

Energy eigenvalues are a periodic function of 𝑘 (wave vectors of Bloch function).

Energy of the plane wave in empty space as the function of wave vector:

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 =ℏ2𝑘2

2𝑚

−8𝜋

𝑎

8𝜋

𝑎−6𝜋

𝑎

6𝜋

𝑎−4𝜋

𝑎

4𝜋

𝑎−2𝜋

𝑎

2𝜋

𝑎

282015-11-27

Energy of the plane wave in empty space as the function of wave vector:

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 =ℏ2𝑘2

2𝑚= 𝐸 𝑛, 𝑘 + Ԧ𝐺 =

ℏ2 𝑘 + Ԧ𝐺2

2𝑚

−8𝜋

𝑎

8𝜋

𝑎−6𝜋

𝑎

6𝜋

𝑎−4𝜋

𝑎

4𝜋

𝑎−2𝜋

𝑎

2𝜋

𝑎

292015-11-27

Energy of the plane wave in empty space as the function of wave vector:

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 =ℏ2𝑘2

2𝑚= 𝐸 𝑛, 𝑘 + Ԧ𝐺 =

ℏ2 𝑘 + Ԧ𝐺2

2𝑚

−8𝜋

𝑎

8𝜋

𝑎−6𝜋

𝑎

6𝜋

𝑎−4𝜋

𝑎

4𝜋

𝑎−2𝜋

𝑎

2𝜋

𝑎

302015-11-27

Energy of the plane wave in empty space as the function of wave vector:

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 =ℏ2𝑘2

2𝑚= 𝐸 𝑛, 𝑘 + Ԧ𝐺 =

ℏ2 𝑘 + Ԧ𝐺2

2𝑚

−8𝜋

𝑎

8𝜋

𝑎−6𝜋

𝑎

6𝜋

𝑎−4𝜋

𝑎

4𝜋

𝑎−2𝜋

𝑎

2𝜋

𝑎

312015-11-27

Energy of the plane wave in empty space as the function of wave vector:

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 =ℏ2𝑘2

2𝑚= 𝐸 𝑛, 𝑘 + Ԧ𝐺 =

ℏ2 𝑘 + Ԧ𝐺2

2𝑚

−8𝜋

𝑎

8𝜋

𝑎−6𝜋

𝑎

6𝜋

𝑎−4𝜋

𝑎

4𝜋

𝑎

322015-11-27

−2𝜋

𝑎

2𝜋

𝑎

Energy of the plane wave in empty space as the function of wave vector:

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 =ℏ2𝑘2

2𝑚= 𝐸 𝑛, 𝑘 + Ԧ𝐺 =

ℏ2 𝑘 + Ԧ𝐺2

2𝑚

332015-11-27

Reduced Brilloin zone.On the border of the Brillouin zoneenergies are degenerated

The band structure of nearly free-electron cubic lattice

[hkl]=

000,

100,100, 200, 200,

kx

– –

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 = 𝐸 𝑛, 𝑘 + Ԧ𝐺 =ℏ2 𝑘 + Ԧ𝐺

2

2𝑚

Ԧ𝐺 = ℎ Ԧ𝑔1 + 𝑘 Ԧ𝑔2 + 𝑙 Ԧ𝑔3

𝑔𝑖 =2𝜋

𝑎𝑖

𝜋

𝑎𝑥−𝜋

𝑎𝑥342015-11-27

The band structure of nearly free-electron cubic lattice

kx

– –

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 = 𝐸 𝑛, 𝑘 + Ԧ𝐺 =ℏ2 𝑘 + Ԧ𝐺

2

2𝑚

Ԧ𝐺 = ℎ Ԧ𝑔1 + 𝑘 Ԧ𝑔2 + 𝑙 Ԧ𝑔3

𝑔𝑖 =2𝜋

𝑎𝑖

𝜋

𝑎𝑥−𝜋

𝑎𝑥

[hkl]=

000,

100,100, 200, 200,

010,010,001,001,

352015-11-27

The band structure of nearly free-electron cubic lattice

kx

– –

The nearly free-electron approximation

Periodic potential

𝐸 𝑛, 𝑘 = 𝐸 𝑛, 𝑘 + Ԧ𝐺 =ℏ2 𝑘 + Ԧ𝐺

2

2𝑚

Ԧ𝐺 = ℎ Ԧ𝑔1 + 𝑘 Ԧ𝑔2 + 𝑙 Ԧ𝑔3

𝑔𝑖 =2𝜋

𝑎𝑖

𝜋

𝑎𝑥−𝜋

𝑎𝑥

[hkl]=

000,

100,100, 200, 200,

010,010,001,001,

110,101,110,101,101,110,101,110– –– –– – – –

362015-11-27

kx

The appropriate expresions for a perturbation calculation of the influence of a small potential

„small potetntial”

Small potential inluence on the borders of the Brilloun zone:

hkl = 000, 100,100, 200, 200,– –

(1D)

kx

The nearly free-electron approximation

Periodic potential

𝑉 𝑥 = 𝑉0 cos2𝜋

𝑎𝑥

𝑉 𝑥 = 𝑉0 cos2𝜋

𝑎𝑥 =

𝑉02

𝑒𝑖2𝜋𝑎 𝑥 + 𝑒−𝑖

2𝜋𝑎 𝑥

𝜋

𝑎𝑥−𝜋

𝑎𝑥

𝜋

𝑎𝑥372015-11-27

kx

k

ax

probability density

probability density

The nearly free-electron approximation

Periodic potential

Plane waves of the same 𝑘-vector 𝜋

𝑎𝑥

382015-11-27

𝜆 =2𝜋

𝑘

kx

k

ax

The nearly free-electron approximation

Periodic potential

Plane waves of the same 𝑘-vector 𝜋

𝑎𝑥

probability density

probability density

392015-11-27

kx

k

ax

The nearly free-electron approximation

Periodic potential

Plane waves of the same 𝑘-vector 𝜋

𝑎𝑥

probability density

probability density

402015-11-27

kx

k

ax

The nearly free-electron approximation

Periodic potential

Plane waves of the same 𝑘-vector 𝜋

𝑎𝑥

probability density

probability density

𝐸± =1

2

ℏ2

2𝑚0

𝐺

2− 𝜅

2

+𝐺

2+ 𝜅

2

±

ℏ2

2𝑚0

𝐺

2− 𝜅

2

+𝐺

2+ 𝜅

22

+ 4𝑉0

412015-11-27

kx

k

ax

The nearly free-electron approximation

Periodic potential

Plane waves of the same 𝑘-vector 𝜋

𝑎𝑥

see Ibach, Lutch

422015-11-27

kx

k

ax

432015-11-27

The nearly free-electron approximation

Periodic potential

8𝜋

𝑎−𝜋

𝑎

6𝜋

𝑎

𝜋

𝑎

4𝜋

𝑎−2𝜋

𝑎

2𝜋

𝑎

𝜋

𝑎𝑥

xa

kx

k

ax

442015-11-27

8𝜋

𝑎−𝜋

𝑎

6𝜋

𝑎

𝜋

𝑎

4𝜋

𝑎−2𝜋

𝑎

2𝜋

𝑎

The nearly free-electron approximation

Periodic potential

xa

kx

k

ax

452015-11-27

8𝜋

𝑎−𝜋

𝑎

6𝜋

𝑎

𝜋

𝑎

4𝜋

𝑎−2𝜋

𝑎

2𝜋

𝑎

The nearly free-electron approximation

Periodic potential

band

band

band

462015-11-27

The electronic band structure

• It is convenient to present the energies only in the 1st Brillouin zone.• The electron state in the solid state is given by the wave vector of the 1st Brillouin zone, band number and a spin.

2015-11-27 47

W. R. Fahrner (Editor) Nanotechnology and Nanoelectronics

The electronic band structure

2015-11-27 48

W. R. Fahrner (Editor) Nanotechnology and Nanoelectronics

The electronic band structure

Tight-Binding Approximation

2015-11-27 49

We describe the crystal electrons in terms of a linear superposition of atomic eigenfunctions(LCAO – Linear Combination of Atomic Orbitals) 𝐻 = 𝐻𝐴 + 𝑉′:

𝐻𝐴 Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 = 𝐸𝑗𝜑𝑗 Ԧ𝑟 − 𝑅𝑛

𝐻 = 𝐻𝐴 + 𝑉′ = −ℏ2

2𝑚Δ + 𝑉𝐴 Ԧ𝑟 − 𝑅𝑛 +

𝑚≠𝑛

𝑉𝐴 Ԧ𝑟 − 𝑅𝑚

Perturbation: the influence of atoms in the neighborhood of 𝑅𝑚 :

𝑉′ Ԧ𝑟 − 𝑅𝑛 =

𝑚≠𝑛

𝑉𝐴 Ԧ𝑟 − 𝑅𝑚

Good for valence band of covalent crystals, 𝑑-orbital bands etc.

𝑗-th state 𝑒 Atom in position 𝑅𝑛Equation for the free atoms thatform the crystal

𝐻𝐴 = −ℏ2

2𝑚Δ + 𝑉𝐴 Ԧ𝑟 − 𝑅𝑛

2015-11-27 50

Approximate solution in the form of the Bloch function:

Φ𝑗,𝑘 Ԧ𝑟 =

𝑛

𝑎𝑛𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 =

𝑛

exp 𝑖 𝑘𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛

Check:Φ𝑗,𝑘+ Ԧ𝐺 Ԧ𝑟 = Φ𝑗,𝑘 Ԧ𝑟

Φ𝑗,𝑘 Ԧ𝑟 + 𝑇 = exp 𝑖 𝑘𝑇 Φ𝑗,𝑘 Ԧ𝑟

Energies determined by the variational method:

𝐸 𝑘 ≤Φ𝑗,𝑘 Ԧ𝑟 𝐻 Φ𝑗,𝑘 Ԧ𝑟

Φ𝑗,𝑘 Ԧ𝑟 Φ𝑗,𝑘 Ԧ𝑟

Expression

Φ𝑗,𝑘 Ԧ𝑟 Φ𝑗,𝑘 Ԧ𝑟 =

𝑛,𝑚

exp 𝑖𝑘 𝑅𝑛 − 𝑅𝑚 න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑚 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

can be easily simplify assuming a small overlap of wave functions for 𝑛 ≠ 𝑚

Φ𝑗,𝑘 Ԧ𝑟 Φ𝑗,𝑘 Ԧ𝑟 =

𝑛

න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉 = ⋯

Tight-Binding Approximation

2015-11-27 51

Approximate solution in the form of the Bloch function:

Φ𝑗,𝑘 Ԧ𝑟 =

𝑛

𝑎𝑛𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 =

𝑛

exp 𝑖 𝑘𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛

Check:Φ𝑗,𝑘+ Ԧ𝐺 Ԧ𝑟 = Φ𝑗,𝑘 Ԧ𝑟

Φ𝑗,𝑘 Ԧ𝑟 + 𝑇 = exp 𝑖 𝑘𝑇 Φ𝑗,𝑘 Ԧ𝑟

Energies determined by the variational method:

𝐸 𝑘 ≤Φ𝑗,𝑘 Ԧ𝑟 𝐻 Φ𝑗,𝑘 Ԧ𝑟

Φ𝑗,𝑘 Ԧ𝑟 Φ𝑗,𝑘 Ԧ𝑟

𝐸 𝑘 ≈1

𝑁Φ𝑗,𝑘 Ԧ𝑟 𝐻 Φ𝑗,𝑘 Ԧ𝑟 =

=

𝑛,𝑚

exp 𝑖𝑘 𝑅𝑛 − 𝑅𝑚 න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑚 𝐸𝑗 + 𝑉′ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

Tight-Binding Approximation

2015-11-27 52

Approximate solution in the form of the Bloch function:

Φ𝑗,𝑘 Ԧ𝑟 =

𝑛

𝑎𝑛𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 =

𝑛

exp 𝑖 𝑘𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛

Check:Φ𝑗,𝑘+ Ԧ𝐺 Ԧ𝑟 = Φ𝑗,𝑘 Ԧ𝑟

Φ𝑗,𝑘 Ԧ𝑟 + 𝑇 = exp 𝑖 𝑘𝑇 Φ𝑗,𝑘 Ԧ𝑟

Energies determined by the variational method:

𝐸 𝑘 ≤Φ𝑗,𝑘 Ԧ𝑟 𝐻 Φ𝑗,𝑘 Ԧ𝑟

Φ𝑗,𝑘 Ԧ𝑟 Φ𝑗,𝑘 Ԧ𝑟

𝐸 𝑘 ≈1

𝑁Φ𝑗,𝑘 Ԧ𝑟 𝐻 Φ𝑗,𝑘 Ԧ𝑟 =

=

𝑛,𝑚

exp 𝑖𝑘 𝑅𝑛 − 𝑅𝑚 න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑚 𝐸𝑗 + 𝑉′ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

Tight-Binding Approximation

Only diagonal terms 𝑅𝑛 = 𝑅𝑚 in 𝐸𝑗

Only the vicinity of 𝑅_𝑛

2015-11-27 53

𝐸 𝑘 ≈1

𝑁Φ𝑗,𝑘 Ԧ𝑟 𝐻 Φ𝑗,𝑘 Ԧ𝑟 =

=

𝑛,𝑚

exp 𝑖𝑘 𝑅𝑛 − 𝑅𝑚 න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑚 𝐸𝑗 + 𝑉′ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

When the atomic states 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 are spherically symmetric (𝑠-states), then overlap

integrals depend only on the distance between atoms:

𝐸𝑛 𝑘 ≈ 𝐸𝑗 − 𝐴𝑗 − 𝐵𝑗

𝑚

exp 𝑖𝑘 𝑅𝑛 − 𝑅𝑚

𝐴𝑗 = −න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑛 𝑉′ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

𝐵𝑗 = −න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑚 𝑉′ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

Restricted to only the nearest neighbours of 𝑅𝑛

Tight-Binding Approximation

Only diagonal terms 𝑅𝑛 = 𝑅𝑚 in 𝐸𝑗

Only the vicinity of 𝑅_𝑛

2015-11-27 54

When the atomic states 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 are spherically symmetric (𝑠-states), then overlap

integrals depend only on the distance between atoms:

𝐸𝑛 𝑘 ≈ 𝐸𝑗 − 𝐴𝑗 − 𝐵𝑗

𝑚

exp 𝑖𝑘 𝑅𝑛 − 𝑅𝑚

𝐴𝑗 = −න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑛 𝑉′ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

𝐵𝑗 = −න𝜑𝑗∗ Ԧ𝑟 − 𝑅𝑚 𝑉′ Ԧ𝑟 − 𝑅𝑛 𝜑𝑗 Ԧ𝑟 − 𝑅𝑛 𝑑𝑉

The result of the summation depends on the symmetry of the lattice:

For 𝑠𝑐 structure: 𝑅𝑛 − 𝑅𝑚 = ±𝑎, 0,0 ; 0, ±𝑎, 0 ; 0,0, ±𝑎 ;

𝐸𝑛 𝑘 ≈ 𝐸𝑗 − 𝐴𝑗 − 2𝐵𝑗 cos 𝑘𝑥𝑎 + cos 𝑘𝑦𝑎 + cos 𝑘𝑧𝑎

For 𝑏𝑐𝑐 structure :

𝐸𝑛 𝑘 ≈ 𝐸𝑗 − 𝐴𝑗 − 8𝐵𝑗 cos𝑘𝑥𝑎

2cos

𝑘𝑦𝑎

2cos

𝑘𝑧𝑎

2

For𝑓𝑐𝑐 structure :

𝐸𝑛 𝑘 ≈ 𝐸𝑗 − 𝐴𝑗 − 4𝐵𝑗 cos𝑘𝑥𝑎

2cos

𝑘𝑦𝑎

2+ 𝑐. 𝑝.

Tight-Binding Approximation

2015-11-27 55

For 𝑠𝑐 structure: 𝑅𝑛 − 𝑅𝑚 = ±𝑎, 0,0 ; 0, ±𝑎, 0 ; 0,0, ±𝑎 ;

𝐸𝑛 𝑘 ≈ 𝐸𝑗 − 𝐴𝑗 − 2𝐵𝑗 cos 𝑘𝑥𝑎 + cos 𝑘𝑦𝑎 + cos 𝑘𝑧𝑎

H. I

bac

h, H

. Lü

th, S

olid

-Sta

te P

hys

ics

𝐵𝑗=−න𝜑𝑗∗Ԧ𝑟−𝑅𝑚

𝑉′Ԧ𝑟−𝑅𝑛

𝜑𝑗Ԧ𝑟−𝑅𝑛

𝑑𝑉

Tight-Binding Approximation

2015-11-27 56

For 𝑠𝑐 structure: 𝑅𝑛 − 𝑅𝑚 = ±𝑎, 0,0 ; 0, ±𝑎, 0 ; 0,0, ±𝑎 ;

𝐸𝑛 𝑘 ≈ 𝐸𝑗 − 𝐴𝑗 − 2𝐵𝑗 cos 𝑘𝑥𝑎 + cos 𝑘𝑦𝑎 + cos 𝑘𝑧𝑎

H. I

bac

h, H

. Lü

th, S

olid

-Sta

te P

hys

ics

𝐵𝑗=−න𝜑𝑗∗Ԧ𝑟−𝑅𝑚

𝑉′Ԧ𝑟−𝑅𝑛

𝜑𝑗Ԧ𝑟−𝑅𝑛

𝑑𝑉

Tight-Binding Approximation

2015-11-27 57

Linear dispersion relation in graphene:

Tight binding approach with the only nearest neighbors interaction [P. R. Wallace, „The Band Theory of Graphite”, Physical Review 71, 622 (1947).] gives:

𝐸 𝑘 = ± 𝛾02 1 + 4 cos2

𝑘𝑦𝑎

2+ 4 cos

𝑘𝑦𝑎

2⋅ cos

𝑘𝑥 3𝑎

2≈ ℏ ǁ𝑐 𝑘 − 𝑘𝑖

Tight-Binding Approximation

2015-11-27 58

Linear dispersion relation in graphene:

Tight binding approach with the only nearest neighbors interaction [P. R. Wallace, „The Band Theory of Graphite”, Physical Review 71, 622 (1947).] gives:

𝐸 𝑘 = ± 𝛾02 1 + 4 cos2

𝑘𝑦𝑎

2+ 4 cos

𝑘𝑦𝑎

2⋅ cos

𝑘𝑥 3𝑎

2≈ ℏ ǁ𝑐 𝑘 − 𝑘𝑖

Tight-Binding Approximation

2015-11-27 59

Linear dispersion relation in graphene:

Tight binding approach with the only nearest neighbors interaction [P. R. Wallace, „The Band Theory of Graphite”, Physical Review 71, 622 (1947).] gives:

𝐸 𝑘 = ± 𝛾02 1 + 4 cos2

𝑘𝑦𝑎

2+ 4 cos

𝑘𝑦𝑎

2⋅ cos

𝑘𝑥 3𝑎

2≈ ℏ ǁ𝑐 𝑘 − 𝑘𝑖

Tight-Binding Approximation

2015-11-27 60

The number of states in the volume of 𝜋𝑘2:

𝑁 𝐸 =2

2𝜋 2 𝜋𝑘2 =2

2𝜋 2 𝜋 𝑘 − 𝑘𝑖2=

2

2𝜋 2 𝜋𝐸

ℏ ǁ𝑐

2

𝜌 𝐸 =𝜕𝑁 𝐸

𝜕𝐸=

𝐸

𝜋 ℏ ǁ𝑐 2

Linear dispersion relation in graphene:

Tight binding approach with the only nearest neighbors interaction [P. R. Wallace, „The Band Theory of Graphite”, Physical Review 71, 622 (1947).] gives:

𝐸 𝑘 = ± 𝛾02 1 + 4 cos2

𝑘𝑦𝑎

2+ 4 cos

𝑘𝑦𝑎

2⋅ cos

𝑘𝑥 3𝑎

2≈ ℏ ǁ𝑐 𝑘 − 𝑘𝑖

Tight-Binding Approximation

2015-11-27 61

The existence of the band structure arising from the discrete energy levels of isolated atoms due to the interaction between them. We can classify the electronic states as belonging to the electronic shells 𝑠, 𝑝, 𝑑 etc.

H. I

bac

h, H

. Lü

th, S

olid

-Sta

te P

hys

ics

The existence of a forbidden gap is not tied to the periodicity of the lattice! Amorphous materials can also display a band gap.

If a crystal with a primitive cubic lattice contains 𝑁 atoms and thus 𝑁 primitive unit cells, then an atomic energy level 𝑬𝒊 of the free atom will split into 𝑁 states (due to the interaction with the rest of 𝑁 − 1 atoms).

Each band can be occupied by 2𝑁 electrons.

Tight-Binding Approximation

2015-11-27 62

H. I

bac

h, H

. Lü

th, S

olid

-Sta

te P

hys

ics

An odd number of electrons per cell(metal)

An even number of electrons per cell(non-metal)

An even number of electrons per cell but overlapping bands(metals of the II group, e.g. Be → next slide!)

If a crystal with a primitive cubic lattice contains 𝑁 atoms and thus 𝑁 primitive unit cells, then an atomic energy level 𝑬𝒊 of the free atom will split into 𝑁 states (due to the interaction with the rest of 𝑁 − 1 atoms). Each band can be occupied by 𝟐𝑵 electrons.

Tight-Binding Approximation

2015-11-27 63

H. I

bac

h, H

. Lü

th, S

olid

-Sta

te P

hys

ics

Be

C, Si, Ge

Tight-Binding ApproximationThe states can mix: for instance 𝑠𝑝3 hybridization.

2015-11-27 64

Tight-Binding Approximation

2015-11-27 65

Michał Baj

Tight-Binding Approximation

Fermi surfaces of metals

2015-11-27 66

Ashcroft, Mermin

Cu

Fermi surfaces of metals

2015-11-27 67

http://physics.unl.edu/tsymbal/teaching/SSP-927/Section%2010_Metals-Electron_dynamics_and_Fermi_surfaces.pdf

Fermi surfaces of metals

2015-11-27 68

Michał Baj

Tight-Binding Approximation

2015-11-27 69

2015-11-27 70

Michał BajSzm

ulo

wic

z, F

., S

egal

l, B

.: P

hys

. Rev

. B2

1,

56

28

(1

98

0).

Tight-Binding Approximation

Tight-Binding Approximation

2015-11-27 71

Michał BajSzm

ulo

wic

z, F

., S

egal

l, B

.: P

hys

. Rev

. B2

1,

56

28

(1

98

0).

Density of stateslike for freeelectrons!

Tight-Binding Approximation

2015-11-27 72

Michał Baj

Cu: [1s22s22p63s23p6] 3d104s1

[Ar] 3d104s1

d-band

Tight-Binding Approximation

2015-11-27 73

Ni: [1s22s22p63s23p6] 3d94s1 [Ar] 3d94s1

ferromagnetic ordering, exchange interaction, different energies for different spins, two different densities of states for two spins ↑ and ↓

Δ – Stoner gap

Semiconductors

2015-11-27 74

Semiconductors of the group IV (Si, Ge) –diamond structure

Semiconductors AIIIBV (e.g. GaAs, GaN) i AIIBVI (np. ZnTe, CdSe) – zinc-blend or wurtzitestructure

Semiconductors AIVBVI (np. SnTe, PbSe) –NaCl structure

SiIndirect bandgap, Eg = 1,1 eV

Conduction band minima in Δ point, constant energy surfaces – ellipsoids (6 pieces), m||=0,92 m0, m⊥=0,19 m0,

The maximum of the valence band in Γ point, mlh=0,153 m0, mhh=0,537 m0, mso=0,234 m0, Δso= 0,043 eV

2015-11-27 75

GeIndirect bandgap, Eg = 0,66 eV

The maximum of the valence band in Γ point, mlh=0,04 m0, mhh=0,3 m0, mso=0,09 m0, Δso= 0,29 eV

Conduction band minima in Δ point, constant energy surfaces – ellipsoids (8 pieces), m||=1,6 m0, m⊥=0,08 m0,

SemiconductorsSemiconductors of the group IV (Si, Ge) –diamond structure

Semiconductors AIIIBV (e.g. GaAs, GaN) i AIIBVI (np. ZnTe, CdSe) – zinc-blend or wurtzitestructure

Semiconductors AIVBVI (np. SnTe, PbSe) –NaCl structure

2015-11-27 76

GaAsDirect bandgap, Eg = 1,42 eV

The maximum of the valence band in Γ point, mlh=0,076 m0, mhh=0,5 m0, mso=0,145 m0, Δso= 0,34 eV

The minimum of the conduction band in Γ point, constant energy surfaces – spheres, mc=0,065 m0

SemiconductorsSemiconductors of the group IV (Si, Ge) –diamond structure

Semiconductors AIIIBV (e.g. GaAs, GaN) i AIIBVI (np. ZnTe, CdSe) – zinc-blend or wurtzitestructure

Semiconductors AIVBVI (np. SnTe, PbSe) –NaCl structure

2015-11-27 77

a-SnDiamond structureZero energy gap, Eg = 0 eV (inverted band-structure) The maximum of the valence band in Γ point, mv=0,195 m0, mv2=0,058 m0, Δso=0,8 eV The minimum of the conduction band in Γ point, constant energy surfaces – spheres, mc=0,024 m0

SemiconductorsSemiconductors of the group IV (Si, Ge) –diamond structure

Semiconductors AIIIBV (e.g. GaAs, GaN) i AIIBVI (np. ZnTe, CdSe) – zinc-blend or wurtzitestructure

Semiconductors AIVBVI (np. SnTe, PbSe) –NaCl structure

2015-11-27 78

PbSeDirect bandgap in L-point, Eg = 0,28 eV

The maximum of the valence band in L point, constant energy surfaces – ellipsoids , m||=0,068 m0, m⊥=0,034 m0,

The minimum of the conduction band in L point, constant energy surfaces – ellipsoids m||=0,07 m0, m⊥=0,04 m0,

SemiconductorsSemiconductors of the group IV (Si, Ge) –diamond structure

Semiconductors AIIIBV (e.g. GaAs, GaN) i AIIBVI (np. ZnTe, CdSe) – zinc-blend or wurtzitestructure

Semiconductors AIVBVI (np. SnTe, PbSe) –NaCl structure

Photoemission spectroscopy

2015-11-27 79

H. I

bac

h, H

. Lü

th, S

olid

-Sta

te P

hys

ics

ℏ𝜔 = 𝜙 + 𝐸𝑘𝑖𝑛 + 𝐸𝑏

work function potential barrier

Photoemission spectroscopy

2015-11-27 80

H. I

bac

h, H

. Lü

th, S

olid

-Sta

te P

hys

ics

ℏ𝜔 = 𝜙 + 𝐸𝑘𝑖𝑛 + 𝐸𝑏

work function potential barrier

Photoemission spectroscopy

http://www.physics.berkeley.edu/research/lanzara/research/Graphite.html

Phys. Rev. B 71, 161403 (2005)

2015-11-27 81

Photoemission spectroscopy

http://www.physics.berkeley.edu/research/lanzara/research/Graphite.html

Phys. Rev. B 71, 161403 (2005)

2015-11-27 82

Semiconductor heterostructures

2015-11-27 83

Semiconductor heterostructures

2015-11-27 84

Investigation of high antimony-content gallium arsenic nitride-gallium arsenic antimonide heterostructures for long wavelength application

Bandgap engineering

2015-11-27 85

How can we change the heterostructure band structure:• selecting a material (eg., GaAs / AlAs)• controlling the composition• controlling the stress

Heterostruktury półprzewodnikowe

2015-11-27 86

Investigation of high antimony-content gallium arsenic nitride-gallium arsenic antimonide heterostructures for long wavelength application

Bandgap engineering

2015-11-27

87

Valence band offset (Anderson’s rule)

Valence band offset:

(powinowactwo)

htt

p:/

/en

.wik

iped

ia.o

rg/w

iki/

Het

ero

jun

ctio

n

Bandgap engineering

2015-11-27 88

Valence band offset (Anderson’s rule)

Su-H

uai

Wei

, Co

mp

uta

tio

nal

Mat

eria

ls S

cien

ce, 3

0, 3

37

–34

8 (

20

04

)

Valence band offset:

(powinowactwo)

htt

p:/

/en

.wik

iped

ia.o

rg/w

iki/

Het

ero

jun

ctio

n

Bandgap engineering

2015-11-27 89

Valence band offset

Wal

uki

ewic

zP

hys

ica

B 3

02

–30

3 (

20

01

) 1

23

–13

4

Bandgap engineering

2015-11-27 90

How can we change the heterostructure band structure:• selecting a material (eg., GaAs / AlAs)• controlling the composition• controlling the stress

Vegard’s law:the empirical heuristic that the lattice parameter of a solid solution of two constituents is approximately equal to a rule of mixtures of the two constituents' (A and B) lattice parameters at the same temperature:

𝑎 = 𝑎𝐴 1 − 𝑥 + 𝑎𝐵𝑥

Bandgap engineering

2015-11-27 91

Vegard’s law:Relationship to band gaps of „binarycompound”:

𝐸 = 𝐸𝐴 1 − 𝑥 + 𝐸𝐵𝑥 − 𝑏𝑥(1 − 𝑥)

b – so-called „bowing” (curvature) of the energy gap

Z D

rid

iet

al.

Sem

ico

nd

. Sci

. Tec

hn

ol.

18

No

9(S

epte

mb

er2

00

3)

85

0-8

56

How can we change the heterostructure band structure:• selecting a material (eg., GaAs / AlAs)• controlling the composition• controlling the stress

top related