You're advertising your paper here - are you asking whether others in RG agree with your propositions?

You have an interesting point there. Maybe you'd like to take a look on one of our papers (on GaMnAs)

http://www.sciencedirect.com/science/article/pii/S0368204805000745

where we did find twi different paramagnetic Mn species besides the metallic, ferromagnetic one. Do you have an idea whether/how that would fit into your picture?

I would be interested to see your work, but I can not pay $31.50 :-)

Please, add your publication to your ResearchGate. In this case, I would be able to read your results. Unfortunately, the file sent by you have been not opened.

Dear Kai,

Here I presented some thoughts on using optical and photoelectric studies based on the cluster model for investigation of ferromagnetism in III-Y semiconductors.

Best wishes,

Yuriy

Dear Kai, I carefully read you paper. I see that the main results obtained by you is observation two magnetic components : the ferromagnetic and antiferromagnetic. The paramagnetic component is associated with interstitial Mn, which have electronic configuration with nd=5. The ferromagnetic component is caused by the presence of superparamagnetic units with ms >> 5 μB (ms ≈ 20 μB) or the equivalent of four Mn atoms for these superparamagnetic clusters. These results were obtained by the MCXD measurements which provide valuable information on the electronic structure of GaMnAs. At the same time, it is important to investigate how the magnetic atoms are introduced into the host crystal and how the various intrinsic defects affect the occurance of long-range ferromagnetic order in Mn-doped III-V compounds. In addition it is necessary to determine the nature and energy structure of such defects. Here, I will try to present some considerations of these issues.

The random distribution of the Mn on the Ga sites will affect on the magnetic properties of GaMnAs. Mn provides localized magnetic moments and is a hole dopant when substituting a Ga site. The appearance of ferromagnetism in GaMnAs is related to the magnetic exchange interaction between holes and magnetic moments of Mn atoms. The magnitude of Tc cruicially depends on both the amount of Mn and the concentration of holes. Usually, GaMnAs includes some amount of defects, such as interstitials, vacancies or antisites. The most abundant defects are the antisites, i.e., As located on Ga sublattice which is a donor. They compensate the holes. Without antisites, the system may be fully ferromagnetic, while for concentrations of antisites above 2%, when the compensation of holes occurs for χMn≈4%, that is paramagnetic. Thus, the magnetic properties of GaMnAs are controlled by the amount of holes and by the amount of compensating defects as AsGa antisites. In this case it is important to effect on the concentration of these defects. The numbers of antisities can be decreased by annealing of the investigated samples. This allows increase the hole concentration and the value of TC. The AsGa antisites of GaMnAs could also undergo a structural transition to an (Asi – VGa) pair upon illumination. This may be an alternative method to reduce the holes compensation by the AsGa antisites and increase TC. On other hand, hydrogen is known to passivate acceptors. The introduction of hydrogen removes the holes required for the carrier-mediated ferromagnetism. Thus, there are different methods to control the ferromagnetic order by changing the defect structure. The study of the nature and energy structure as well as a control in GaMnAs can be performed using optical and photoelectric measurements.

It is also well known that other model of ferromagnetism in GaMnAs is associated with the formation of the magnetic polarons for bound holes and their percolation {see in detail : A. Kaminski, V. M. Galitski, and S. Das Sarma, Ferromagnetic and random spin ordering in dilute magnetic semiconductors, PHYSICAL REVIEW B 70, 115216 (2004)}. It means that at high enough temperatures spins of holes and magnetic impurities are not correlated. As the temperature decreases, magnetic atoms close enough to hole centers. As a result their spins align in the direction opposite to that of the hole. Thus, the bound magnetic polarons appear. Since the temperature is lowered the localized holes produce magnetic clusters of bound magnetic polarons. The polaron size increase with decreasing temperature. At a certain temperature, the overlap of magnetic clusters and the formation of an infinite cluster through the whole sample occurs. This leads to the ferromagnetic temperature-induced polaron percolation transition. In my opinion such cluster’s scenario of the transition to the ferromagnetic state is qualitatively similar to the scenario of the transition to the spin-glass state in CdMnTe crystals, namely, the formation of magnetic clusters, their percolation and formation of infinite cluster. However, the nature of such clusters is different for these systems. In the case of CdMnTe crystals the magnetic spin clusters are formed, which include only the magnetic ions. While GaMnAs crystals are dealing with clusters bound magnetic polarons with the participation of bound holes. Thus, a bound magnetic polaron is the basic unit of the mechanism of ferromagnetic transition in DMS with localized charge carriers.

It should be noted that in GaMnAs there are the regions of the sample with high local Mn concentration. The Mn-rich regions become magnetically ordered at higher temperatures than the regions with lower Mn concentration, where the holes have lower probability of presence. Thus parts of the sample become ferromagnetic at higher temperature than others. Therefore in this system the bound magnetic polarons are formed with different energies. This allows distinguish them in the optical measurements (photoluminescence and absorption spectra and may be in photodiffusion current spectra). In addition, it is possible to study the excitonic magnetic polarons analogously as it was performed by us for CdMnTe.

In summary, I think that the optical and photoelectrical studies together with the magnetic and structural measurements of magnetically inhomogeneous semiconductor systems will help one to better understand the nature of their magnetic transitions to ferromagnetic or spin-glass states.