Surface thermodynamics

Steady gas flow

Experimental Thermodynamics II: Open System

Experimental Thermodynamics I: Closed System

Vapor pressure

Chemical thermodynamics, chemical equilibrium, and chemical reaction rate

Band gap of periodic potential

Puzzle

Puzzle

Thermaldynamic functions

As we know, combining the first and second law of thermaldynamics, one gets

dU=TdS-PdV, (1)

where T,S,P and V are temperature, entropy, pressure and volume.

In addition, we also know that S and V are extended functions, meaning:

U(nS,nV)=nU(S,V),

which leads to Euler’s equation (the deviation can be found on the Internet):

U=TS-PV.

On the other hand, we know there are other three energy related functions, say H (enthalpy), F(free energy) and G(Gibbs free energy). Their relation to internal energy U is the following:

H=U+PV
F=U-TS
G=U-TS+PV.

Therefore, with those functions, one can describe four different process, adiabatic+isometric, adiabatic+isobaric, isothermal+isometric, isothermal+isobaric.

However, if we consider Euler’s equation, we get
H=TS
F=-PV
G=0
what’s wrong here?

Periodic potential problem

When we study solid state physics, we often deal with periodic potential, which may make you think that it is trivial. Normally, we don’t really know the function form of the potential. So approximation is made and perturbation theory is invoked, e.g. in free electron approximation and tight binding approximation.

I was surprised when I tried to solve this Schodinger equation:

d2Ψ/dx2+cos(Kx)Ψ=EΨ, here hbar, m are omitted for simplicity.

We know from Bloch theorem that the wave function takes the form

Ψ=exp(-ikx)u(x),

where u(x)=u(x+2n/K).

But to it takes me only so far. I can not find the exact solution.

Does the analytical solution exist?

RHEED simulation

Periodic function and Fourier expansion

Periodic function and Fourier expansion

The theorem is that if a function f is periodic with frequency w (period a), the function can be expanded as Fourier polynomials. i.e.

If f(x+na)=f(x)

Then f(x)=\sum_{k=nK}f(k)e^{-ikx}

where K=\frac{2\pi}{a}

.

The proof actually relies on common sense.

define:

 f(k)=\int f(x)e^{-ikx}dx

One can see that both f(x) and exp(-ikx) are perodic, where average of exp(-ikx) is zero. So if the two periods are not commensurate, f(k) will be zero.

The only possible nozero f(k) occurs when kx=2npi, where k=nK.

pydao0.979 released

Pydao, a new software for data organize and analysis is released as a trial version.

Data organization is based on hierachical data file (HDF) structure.

Data analysis is based on the plugins designed for special usage.

Data visualization is based on matplotlib and mayavi pakage.

Now I have lattice dynamics as a useful plugin.

There are some build in analysis and visualization tools, not as much as the xpy1.xx. But we will make pydao more and more complete in the future.

Right now, only source code is provided. Win32 compiled will come soon.

Surface preparation of MgO

Surface preparation of MgO

What's the use of MgO

1) Good material for hight temperature processing 2800 C melting temperature

2) low dielectric constant ~10, low loss

Problem of MgO

1) surface reacts with H2O and CO2 after exposing to air

2) impurities like Ca

Preparation steps

1) Cleaving

Normally with charged surface, difficult for scanning

2)Polishing

sub-micron polishing is normally used.

3) Acid etching

This is to remove the additional material left from the polishing. So basically, the sample as received is already etched.

agent: phosphoric and nitric-acid

4) Annealing

Most important part. This is to solve problem 1) and to make atomically flat surface.

Surface orientation	Duration	Temperature	Step height	Terrace width	Comments	Reference
(100)	2 hours	Tanneal>1000 C, the surface changes remarkably The annealing is better for higher temperature until  Tanneal>1350 C		200-300 nm	Step comes from the Ca atoms diffuse to the surface	Ahmed1996
(100)	12 hours	> 700 C can get rid of Mg(OH)2 and MgCO3 >1100C, one can get atomically smooth surface	4 nm	700 nm		Aswal2002
(100) 2 degree miscut	360 min	1000	3-7 nm	120 nm		Benedetti2007
(110)	10 min	1000 in 10-7 torr			facet created because of liquid solid interface in the etching	Giese2000
(111)	30 min	1700 to heal the facet			The surface may reorganize into (332) 1700 C to heal the facet

Reference

[Ahmed1996] Ahmed F. et al J. of Low Temperature v105, p1343 (1996)

[Plass1998] Plass R. Surface Science, v414, p 268 (1998)

[Giese2000] Giese D.R. Surface Science, v 457, p 326 (2000)

[Aswal2002] Aswal D.K et al. Journal of Crystal Growth, v236, p.661. (2002)

[Benedetti2007] Benedetti S et al. Surface Science v601, p 2636. (2007)