The Experimental Search for Anderson Localisation of Light in Three Dimensions


Dateien zu dieser Ressource

Prüfsumme: MD5:7d4a0994a5c28dc87b122524da10be04

SPERLING, Tilo, 2015. The Experimental Search for Anderson Localisation of Light in Three Dimensions

@phdthesis{Sperling2015Exper-31700, title={The Experimental Search for Anderson Localisation of Light in Three Dimensions}, year={2015}, author={Sperling, Tilo}, address={Konstanz}, school={Universität Konstanz} }

2015 2015-09-08T05:41:01Z 2015-09-08T05:41:01Z Sperling, Tilo Sperling, Tilo The Experimental Search for Anderson Localisation of Light in Three Dimensions The goal of this thesis was to continue the work Martin Störzer and Wolfgang Bührer, to refine the measurements and to prove the concept of Anderson localisation.<br /><br />Our samples are made of TiO<sub>2</sub> in the rutile phase, which has a high refractive index and very low absorption. The properties of these powders are well known by previous experiments. From these we know that some powders show deviations in the long time tail, where more photons are observed than expected for pure diffusion. These deviations were interpreted as a sign of localisation. By implementing the new method of transmission profile measurements we were able to observe results consistent with the time of flight measurements. This method uses a pulsed laser and an ultra fast gateable camera system, being able to record the temporal evolution of the photon cloud in transmission. Evaluating the transmission profiles has the great advantage that it is unaffected from absorption signatures. Thus pure deviations from diffusion are visible, allowing a more precise measurement of these deviations that were not observable with time of flights. The results are in agreement with theoretical predictions, and comparable to measurements with ultrasound. kl<sup>*</sup> scaling was shown by using different powders and tuning the laser wavelength. With the measured transmission profiles, an estimation of the localisation transition was possible and is in good agreement with previous measurements.<br /><br />In a second experiment we have redone the measurements of Wolfgang Bührer for better comparability of the data. We prepared a mixture of R104 and anatase and also a pure R700 sample. The mixture is expected to show only a very small deviation in the long time tail, acting therefore as a reference. This was confirmed by the measurement. By changing the incoming intensity of the laser we observed a non-linear increase of the long time tail for both samples, but stronger for R700. We further investigated the inelastic scattered contributions via wavelength resolved measurements with bandpass filters. We were able to measure red and blue shifted contributions. The non-shifted photons appear to be diffusive in the time of flight, but not in the transmission profile. Here the advantage of the transmission profile measurements becomes apparent. We were able to show a full interpretation of our data with a localisation picture involving non-linear effects. The main arguments are that localised modes have increased intensity and additionally the photons path travelled are enormous (up to several meters), increasing the probability of an inelastic event to occur. Therewith the inelastic scattered photon is not resonant in the mode any more and can be detected. The doubts raised by Scheffold and Wiersma could be dismissed in almost every points.<br /><br />We have always seen the destruction of time reversal symmetry using the Faraday effect as an ultimate tool to show that we observe localisation. Therefore we introduced a flow cryostat and a superconducting magnet into our setup. We mixed our powder with a Faraday active material, the ratio based on the estimations of Lukas Schertel. The flow cryostat is required to cool down the samples to liquid Helium temperature (4K) to drastically increase the Faraday effect. By driving the magnetic field up to 18T we expect destruction of localisation in our samples. With a proof of principle setup we were able to measure time of flights at low temperatures and 18T without loosing signal to noise ratio. The time of flights showed that absorption was decreased by cooling, but contrary to our expectations the magnetic field had no observable influence. There are many possible reasons why we did not see the expected behaviour. Thus no conclusion can be drawn from this experiment.<br /><br />The last part is the crux of this work. We observed some features in our data that might indicate that there is a lifetime process involved. This was the main argument of Scheffold and Wiersma. We have systematically investigated probable indications that the localisation or lifetime interpretation becomes inconsistent. Very small samples still show deviations, which is still explainable with both models. Calculations done by Mirco Ackermann involving diffusion only and a lifetime process show very similar features as measured and strengthen the suspicion that we observe such a process. In contrast to the localisation interpretation we were able to observe in very first measurements done by Lukas Schertel on mixtures of water or glycerol with R700 an excess of photons at long times. Further short and long pass measurements underlined an inelastic lifetime process and finally a measurement of the emission spectrum gave evidence of fluorescence in the red. Other measurements suggest that the non-linearity we observe originates from luminescence of two photon absorption. However, this non-linear signal looks similar to saturation of fluorescence, which might indicate a saturation of two photon absorption. We can conclude that it is very unlikely that we observed signatures of localisation, but rather misinterpreted the fluorescence. The process behind the effect of intensity loss due to cooling is still unknown. eng

Dateiabrufe seit 08.09.2015 (Informationen über die Zugriffsstatistik)

Sperling_0-300030.pdf 189

Das Dokument erscheint in:

KOPS Suche


Mein Benutzerkonto