Near-field Optical Microscopy
Collaboration with the University of Tokyo in the Area of Near-field Optical Microscopy
We visited prof. Ohtsu’s group at the University of Tokyo, Japan for 3 weeks between February 22nd and March 15th in the framework of ADOPT mini-project. During our visit at Ohtsu research group’s laboratories, we were trained in near-field probe manufacturing and advanced measurement techniques. First, we were confronted with near-field probe manufacturing for UV range. The probes were manufactured from pure silica core fibers by means of heat-and-pull technique followed by etching in a water solution of hydrofluoric acid and ammonium fluoride. For pulling, we employed a commercial micro-pipette puller. Every pulled fiber was subsequently inspected with the optical microscope. Furthermore, the final quality of the probes after the etching step was assessed by scanning electron microscope (SEM) imaging (depicted in Fig. 1).
We performed near-field photoluminescence (PL) measurements on a proton-implanted GaN epitaxial layer. For this task, we used two different setups. The first setup was a home-made scanning near-field microscope (SNOM), very similar to the one currently used in our laboratory at KTH. However, differently from our setup, the oscillation of the tuning fork is induced electrically rather than mechanically. Because of the thickness of our GaN epitaxial layer, we used SNOM in the illumination-collection mode. In such a case, both excitation and resulting PL are transmitted through the fiber probe. This requires the use of a 3 dB fiber coupler which provides separate arms for excitation light input and PL extraction. We performed a preliminary measurement using a photomultiplier tube as detector. However, we were not able to conduct more systematic studies since the piezoelectric stack used for distance control got broken. Therefore we continued measurements only with commercial SNOM system by Jasco, Inc.
In the commercial setup, we used metal-coated probes provided by the same company. The fiber probes were mounted in the vacuum and low-temperature capable chamber and the infrared laser beam used for the optical feedback was aligned. In this setup, we also employed illumination-collection mode. The collected PL was measured by a spectrometer coupled to a liquid nitrogen cooled silicon detector. Unfortunately, measurements were impaired by the strong self-luminescence produced in the fiber by the UV excitation at 325 nm. The self-luminescence noise overwhelmed the weak PL signal from the sample.
Although we did not manage to collect any scientifically significant data from our sample, our visit was certainly an invaluable experience since we could work with various combinations of tools and techniques for specific SNOM measurements. In particular, the following list describes what we learnt and could make use of in our laboratory:
- We practiced with the fiber pulling technique and assessed its performance. This experience would be of great value if we decide to switch to probe manufacturing by heat-and-pull.
- It is preferable to attach tips to tuning forks using UV-curable adhesive which results in both less hectic workflow and higher quality factor of the resonator.
- The tuning fork can be driven electrically as opposed to mechanically which leads to less complicated assembly. This idea was not new to us but it was beneficial to see its practical implementation.
- The illumination-collection setup can be conveniently implemented using a 3 dB fiber coupler. We subsequently decided to opt for this scheme in our laboratory.
- We learned how to efficiently couple light from the fiber to the spectrometer by using multiple lenses and observing the 2D image produced by the detector matrix.
Moreover, we discovered that many of the problems that we routinely face in our laboratory are also shared by our colleagues at the University of Tokyo, one of the leading groups in the optical near-field technologies. Even though careful attention to details is essential, it does not guarantee successful outcome.