Photomicrodomain light scattering

When exciting light beam propagates along optical axis of the sample, usual photoinduced light scattering is not excited because the direction of according noise hologram vector and photovoltaic current direction are perpendicular so that charge redistribution does not take place in the direction required for noise hologram record. So if we use a weak (not focused) laser beam, both optical autowaves and photoinduced light scattering are not expected. Nevertheless, we found under that conditions some light scattering does appear.


When the laser beam was directed into the sample at angle of incidence q L (relative to z-axis of the crystal), the special light scattering appeared in a few minutes. The scattering indicatrix was located at surface of the cone with the angle of 2 q L . At closer examination the indicatrix was found to consist of two orthogonally polarized cones enclosed one into another.
Conic indicatrix and temporal delay of the scattering appearance make the scattering look like some kind of photoinduced light scattering taking place with some phase synchronism (remind four wave light scattering for instance). But this conic scattering had some features making it essentially different.

First, unlike scattering kinds of holographic nature, this scattering can be induced by appropriate incoherent light.

Secondly, the scattering cone axis was always directed exactly along z-axis whereas the spatial parameters of holographic scattering cones are determined by interacting beams directions. Furthermore the photoinduced scattering cones disappear unless appropriate wave synchronism conditions are realized. But such conditions are being broken under change of pumping angle of incidence. On the contrary, after stationary state of the conic scattering had been reached, change of pumping angle of incidence up to 30 o only caused non-inertial change of the scattering cone angle without any change of the total scattering intensity.

One more surprising feature was that placing the sample into conductive media made the scattering instantly disappear.
But the most amazing peculiarity of this scattering was that the scattering sources appeared beyond the region of light action.
Let us test the photoexcited crystal by a number of an inactive light beams which do not cause any photoinduced effects in the crystal (for instance a beam of He-Ne laser with wavelength l T = 0.63 m power 1 mW). Under testing the photoexited crystal by such inactive light beams the conic scattering of the testing light was found to appear first far enough (at 5-10 mm length) from the region illuminated by the pump and only a few minutes latter the conic scattering appeared in the crystal region illuminated by the pump.


Such unlocality of the phenomenon testifies that the conic scattering substantially differ from any kinds of photoinduced light scattering having holographic nature. Indeed it is difficult to imagine that holograms.

The conic scattering features can be naturally explained within the supposition that a set of long thin needles oriented along z-axis of the crystals are induced under light influence on the crystal. If the needles are long enough so that their longitudinal dimension is much more than transversal one, the light scattering will take place under the synchronism conditions K ls =(K so) z = (K se) z, where K L, K S  wave vectors of the incident light and the scattering accordingly, “o” and “e” mean ordinary and extraordinary scattering components polarization. The synchronism conditions determine two scattering cones at angles ΘSo, ΘSe. At small ΘL, finding difference between these angles ΔΘ not complicated: ΔΘ ΘSo - ΘSe = 2  Δn oe / (n o + n e). Using the known values of refractive indexes n o = 2.297, n e = 2.208 for lithium niobate crystals we can get  ΔΘ 0.042 ΘL.
The azimuth dependence of scattering intensity is mainly determined by both polarization types of pump and scattering (o - ordinary, e - extraordinary) and appropriate orts eL,Sviz. I s ~ z d P L. Finding the dependence for scattering at z-axis is not complicated: 
is the angle between x-axis and projection of on XY plane). The obtained dependence agrees with experimental data. The correct polarization measurements brought the experimental value of ΔΘ:
ΔΘ= (0.040 0.006) Θ