Yes, that's right, one can relate the conduction mechanism to temperature dependence. For example, negative temperature coefficient generally indicates a semiconductor, positive temperature coefficient indicates a metal. Ionic conduction in liquids can be either positive or negative depending on the ionic species. With respect to AC conductivity vs. frequency, it really depends on many things, including the method of measurement, frequency, etc.. For example, if the measurement is isothermal, and the method is surface based (eg: 4-point probe), and the frequency is high, the "skin effect" will play a role along with the bulk conductivity. Under these circumstances, different bulk resisitivities will give different conduction vs. frequency slopes. Remember that one need not actually have a DC conductivity to have an AC conductivity (eg: a capacitor), that is to say, one can push AC power through a dielectric and calculate AC conductivity through Ohms' power law (P=E^2/R). Remember also that while DC conductivity is purely real, AC conductivity is actually a complex number (real+imaginary). So, I guess, in general, the slope of a conductivity vs. frequency plot does not indicate conduction mechanism.

Yes, that's right, one can relate the conduction mechanism to temperature dependence. For example, negative temperature coefficient generally indicates a semiconductor, positive temperature coefficient indicates a metal. Ionic conduction in liquids can be either positive or negative depending on the ionic species. With respect to AC conductivity vs. frequency, it really depends on many things, including the method of measurement, frequency, etc.. For example, if the measurement is isothermal, and the method is surface based (eg: 4-point probe), and the frequency is high, the "skin effect" will play a role along with the bulk conductivity. Under these circumstances, different bulk resisitivities will give different conduction vs. frequency slopes. Remember that one need not actually have a DC conductivity to have an AC conductivity (eg: a capacitor), that is to say, one can push AC power through a dielectric and calculate AC conductivity through Ohms' power law (P=E^2/R). Remember also that while DC conductivity is purely real, AC conductivity is actually a complex number (real+imaginary). So, I guess, in general, the slope of a conductivity vs. frequency plot does not indicate conduction mechanism.

Hi Patrick , thanks for your reply.

Actually i am also investigating the surface effects on silicon nanostructures and hybrid devices.

impedance is measured in frequency range 20 Hz- 3 MHz. and two probe method is used. then we have extracted AC conductivity from impedance measurements.

With different surface passivation material on silicon nanowires i get different slopes of AC conductivity.

i think there should be some conduction mechanism related to range of slop, what do you think?

When you say you've extracted AC conductivity from impedance measurements, are you saying you're taking just the real part? Or, perhaps you're talking about the mag. of the impedance vector?

At any rate, since you're talking about nanowires, a whole pile of different things could be going on. As you probably already know, as we enter the nano-realm, material properties change in un-expcted ways. After all, that's why nanotech is so hot now-a-days. Still, as a physical chemist (and a EE), let me hazard a guess if that's alright...

When you passivate the nanowires in different ways, you're probably also modifying the interior of the nanowire, not just its surface. Remember, the surface area of a nano-object plays a huge part in it's properties. But it's also true that chemistry we normally think of as located on the "surface" is also penetrating into the "bulk" of the nanopartical.

Right now, in our lab, we're doing some research on TiO2 nanoparticles. They're only about 2nm in diameter. When we modify the "surface" of the nanoparticle using Fe or Au, (just a single metal atom per particle now .. ) the entire quantum mechanical structure of the particle changes. This is because the wave functions of all the (only several hundred) atoms overlap and interact. This produces profound changes in the TiO2 nanoparticle properties. The concept of "surface" and "interior" for nanoparticles is rather fuzzy as you can imagine.

I think something like this is going on with the Si nanowires. By changing the passivation, you're doing more than just changing what's happing on the surface of the wire. The "bulk" properties of the nanowire are also changing. This change in impedance vs. frequency slope you're observing might not have anything to do with changes in conduction "mechanism", simply a change in the bulk conductivity.

Now, you're using Si nanowires, so let me ask if they're doped or not and if they're doped, are you making them n-type or p-type? Are you doping them to degeneracy? When you passivate, it is possible that you're changing the doping on the nanoparticle. Remember, if you simply place an electron ritch molecule on the surface to passivate, you can actually change the effective conductivity of the nanowire.

As an asside, let's model the nanowire as a series RL circuit where the RL time constant is quite short (ie: very high frequency cutoff). If you measure the circuit impedance with different R values, and hold the L constant, you'll see that the slope of the impedance vs. frequency plots will change based on the value of R. So, I guess all the previous discussion is about why R might be changing.

Finally, to the point of conduction mechanism in a semiconductor, there's really only one: electrons in the conduction band and holes in the valence band diffuse in opposite directions under the influence of the electric field. (I'm not talking about what happens when the semiconductor is exposed to photons of energy greater than the band gap.). If we were talking about a carbon nanotube (metalic heiix) conduction can also occur via ballistic mechanisms. But, we're talking about Silicon.

phew .. a little long ... sorry about that ... :-)

Thanks Patrick for long answer :) i am fully convinced what you said in your answer.

Nanowires are of p-type and doped. i have already published work related to enhanced DC and AC conductivity from passivation of different materials.

because with some passivation holes content on the surface increases and hence bulk resistance also decreases.

i think i can't convey my question properly.

there are different hopping conduction mechanisms associated with slop of AC conductivity.

different models: correlated-barrier hopping (CBH); overlapping large-polaron

(OLP); (i) for a reduced polaron radius, (ii) for a large polaron radius;

quantum-mechanical tunneling (QMT); small polaron (SP)

i just want to ask if there is some relation of slop range with that specific conduction models even at fixed temperature.

Hi Kamran

What I understand from your question is --You are talking about σ Vs ω data.

Such data are represented as log σ Vs log ω curves and the slope of the curve is estimated. Actually this data is supposed to fit σ =σ0+Aω^n curve and the log σ Vs log ω curves slope gives you the value of exponent n. This value varies from 0 to 2 (or some times greater than 2) in different frequency range. It also depend upon the temperature and many other factors.

In general it is explained as

n=0 ----dc conduction

n=0.5-0.65----correlated ion hopping

n=1 Nearly Constant loss

n>1 Super-linear Power Law

If you search literature for these terms you will certainly get detailed information

The value of n of particular range will depend upon temperature

Hi Neelam thanks for your reply.

can you suggest me any research paper related to that?

i have already checked but i haven't got any paper in which all conduction mechanisms are discussed with different ranges of exponent "n."

getting all information in single paper is difficult everyone discuss the portion of particular interest.

Still i am sending one paper which may give you some information.

Related file

Thanks for your response.