I think the question is probably a good one. Metal oxides are an interesting area of semiconductors, and can be either crystalline or amorphous. In both cases they are used for a variety of applications ranging from gas sensing to more recently transparent transistors. Of course there are lots of metal oxide, especially if you consider binary or ternary oxides. In general:

1)Like others have mentioned, the spacing between atoms, i.e. the lattice constant in a crystal will change the band gap. if temperature or pressure changes the spacing the gap will change.

2) disorder in the crystal can change the apparent band gap i.e. band tailing in optical band gap measurements

3) the band gap will change some with the conductivity of the metal oxide. This can be due to a deficit of oxygen for example. In the extreme you can often deposit a grayish metallic like thin film. Before that if you increase the conductivity the optical band gap can increase because of the Moss Burstein effect.

4) changing the composition can be used to tune the bandgap, i.e. add MG to ZnO the band gap will increase, Add Cd to ZnO the band gap will decrease

5) Also as mentioned above if the particles are made very small quantum size effects can change the band gap. A lot of times in metal oxides this is not spectacular for example ZnO since the exciton tends to be tightly bound and the side of the particle needs to be on the size scale of the exciton radius.

There are other classes of metal oxides for example where the material can undergo a phase transition. For example Vanadium Dioxide undergoes a transition between an insulator and metal. When this happens the band gap changes.

Hello Raja Viswanathan,

I assume that you mean thin films of semiconductors and insulators; metals do not have bandgap. The answer of your question lies in the very mechanism why bandgaps occur. Basically, the interatomic distance in a material determines its bandgap . In crystalline materials it is the lattice parameter. Any factor which influences the interatomic distance in a material can cause bandgap shifts, eg. temperature, pressure, strain, concentration of impurities.

One of reasons can be also the degradation of organic material.

http://dx.doi.org/10.1016/j.synthmet.2012.11.002

One possible reason might be oxidation of thin film surface

To Kanad,

My understanding is that there is no close correlation of band gap and interatomic spacing. More temperature dependent as well as strain, impurities etc.

To Raja,

Depending on materials (i.e. organic or inorganic semiconductor), optical band gap and electrical band gap may not be the same. In case of organic semiconductor, they are not identical. So depending on your sample, this aspect should be taken care of.

To Kyung-Min Lee

If you refer to standard solid state physics texts to learn the physics of why bandgaps form at all, you will appreciate the basic role played by interatomic spacing and its periodic nature. This is then influenced, of course, by a large number of factors involving bonding, various kinds of quasi-particle couplings and so forth and the correlation becomes complicated and highly material-specific. In fact, if you try to explain why bandgaps depend so much on temperature and strain (which you agree), I think, you will have to come basically to lattice spacing (for crystalline semiconductors, of course). It will be intersting to know your alternative explanations.

Also, you must have noticed that although slight (part per million) amounts of impurities (eg. B or P in Si) change electronic conduction properties enourmously, the bandgap is not affected as the lattice properties are not affected at this stage. It is only when the doping is very high and tends to the host atomic density that the bandgap is affected.

Organic semiconductors are far more complicated as their structures are often quasi-periodic and bonding is more complicated. The HOMO-LUMO gap may then be the more appropriate term than the bandgap.

Anyway, it looks like the basic fact remains that the atomic neighbourhood in a material determines its band gap, and the interatomic distance is the primary parameter to quantify the atomic neighbourhood. Hence anything that affects it is likely to influence the bandgap.

To Kanad,

The energy band structure is basically formed by the periodic nature of crystalline structure. There are a lot of bands and gaps in the dispersion relation. I think

you referred to this nature.

The aforementioned band gap is the gap between lowest unoccupied and highest occupied states of electron. It's material dependence and also doping and temperature relying on Fermi-Dirac and Pauli principle. Not lattice spacing.

Dear Kyung-Min,

Yes, indeed, by definition, bandgap is the energy gap between the highest occupied and the lowest unoccupied bands in a semiconductor and that is what the question and my answers were about.

From your last statement, I think, you are confusing the energy band gap (unit: eV) with electron occupation or density in a band (unit: cm^-3). Electron density in a band is certainly decided by the density of states (material as well as temperature dependent), the Fermi-Dirac statistics (temperature dependent), and the degeneracy factor, which comes from the Pauli exclusion principle (allows two electrons of opposite spins to occupy the same state) and also on the concentration of ionized dopants.

Hope this clarifies the situation that bandgap does depend on interatomic spacings.

I suggest to have a look at the book

Roald Hoffmann, Solids und Surfaces. A Chemist's View of Bonding in Extended Structures

It nicely addresses the local origins of energy bands from the atomic orbitals in close analogy to the formation of molecular orbitals.

The energy band gap decreases as the temperature increases in the semiconductors, It can be understood if you consider the inter atomic spacing increases when the amplitude of the atomic vibrations increased due to increased in thermal energy. It is the linear expansion of the material.As increased in the inter atomic distance , the potential seen by the electron in the material is decreases, so that reduces the energy band gap.

Dear Raja, Bandgap in bulk materials were once their signatures. But the advent of nanomaterials have made it a size dependent property. Bandgap depends on the nature of material forming the thin films, its particle size, its thickness in Nanothinfilms, the temperature, the concentration and nature of impurities added and by the preparation conditions. You can see many papers on bandgap tailoring. For ternery systems with good mixibility they also obey Vegard's Law.

I think the question is probably a good one. Metal oxides are an interesting area of semiconductors, and can be either crystalline or amorphous. In both cases they are used for a variety of applications ranging from gas sensing to more recently transparent transistors. Of course there are lots of metal oxide, especially if you consider binary or ternary oxides. In general:

1)Like others have mentioned, the spacing between atoms, i.e. the lattice constant in a crystal will change the band gap. if temperature or pressure changes the spacing the gap will change.

2) disorder in the crystal can change the apparent band gap i.e. band tailing in optical band gap measurements

3) the band gap will change some with the conductivity of the metal oxide. This can be due to a deficit of oxygen for example. In the extreme you can often deposit a grayish metallic like thin film. Before that if you increase the conductivity the optical band gap can increase because of the Moss Burstein effect.

4) changing the composition can be used to tune the bandgap, i.e. add MG to ZnO the band gap will increase, Add Cd to ZnO the band gap will decrease

5) Also as mentioned above if the particles are made very small quantum size effects can change the band gap. A lot of times in metal oxides this is not spectacular for example ZnO since the exciton tends to be tightly bound and the side of the particle needs to be on the size scale of the exciton radius.

There are other classes of metal oxides for example where the material can undergo a phase transition. For example Vanadium Dioxide undergoes a transition between an insulator and metal. When this happens the band gap changes.

By reading all these I have a doubt on the amorphous meta oxides. Whether amorphous metal oxide possess a band gap? or we can explain the energy gap in terms of HOMO -LUMO?

Amorphous metal oxides definitely have a band gap by the normal definitions of band gap. Experimentally you see strong absorption where the material is no longer transparent. You also see the changes in band edge as you change the alloy composition etc. Depending on your training you can look at the idea of band gap from different perspectives. From a molecular orbital perspective even though the material is amorphous there can be very short range ordering, and an pretty uniform average bond length between atoms. So one could probably think of it from a homo lumo perspective.

Theorists don't usually compute the band structures of amorphous materials because a lot of the tricks of using the symmetry of the crystal don't apply, but the same principles should hold. For example, you could look at

J. TAUC et al.

Optical Properties and Electronic Structure

of Amorphous Germanium

BY

J. TAUC (a), R. GRIGOROVICI (b), and A. VANCU (b)

As an example of an amorphous materials with a band gap. Of course you see a different band gap than the crystalline material, but the idea of a band structure is still there and valid.

Dear Prof John, Thanks a lot for the information. It was very much useful for me and trying to understand the concept. 

it also depend on the partial pressure of oxygen.

One of the main cause of the change of the energy band structure including the bandgap is the crystallite size of the formed thin film. As the crystallite size decreases the bandgap increases which means upward shift toward lower wavelengths. This is because of the quantum mechanical confinement of electrons in the grains of the material. Technologically one benefits from this observation by trimming the bandgap of nano material structures.

Best wishes