Light Constructions - Direct near-field microscopy measures both intensity and phase

Researchers at the Univ. of Bath (UK) have extended scanning near-field optical microscopy (SNOM) so that it can also be used as a high-resolution tool for measuring phase. The system combines a basic SNOM setup with a Mach-Zender interferometer, and the signal is detected so as to allow both intensity and phase to be read out simultaneously. The tool was developed by physicists in the Optoelectronics Group and is intended to help them characterize new photonic-crystal devices they have been working with.¹

In a scanning near-field optical microscope, the object to be measured is illuminated using laser light. The probe tip is an optical fiber with a very fine point -- in the Univ. of Bath setup, less than 50 nm in diameter -- and this is brought to within 100 nm of the object. Some of the light from the object is coupled into this fiber tip, which then travels down the fiber so its intensity can be detected. To measure an area, the end of the tip is scanned around the object surface, with data taken at each spot.

To allow this system to measure phase, there must be some scale for the phase to measured against (Figure 1). This is accomplished by splitting the laser beam into two parts: one to illuminate the object and be detected by the probe tip, the other to act as a reference. These object and reference beams interfere inside a fiber coupler and are subsequently detected. To determine the phase of the object beam, the piezoelectric (PZT) transducer that sets the relative phase between the two arms of the interferometer is controlled via a feedback loop. The detection and feedback electronics ensure that the two beams always interfere to produce a maximum (ie. they always have the same relative phase). The voltage used to control the PZT, therefore, is directly related to the way the phase between the two beams is changing.

Figure 1. Schematic representation of the combined Mach-Zender interferometer and scanning near-field optical microscope. Light from the laser is split into two beams. The first goes into the object to be examined -- either the tapered fiber (top) or prism (bottom) -- and some of it is coupled into the SNOM tip that is being scanned within 100 nm of the object's surface. This light is coupled with that from the reference signal and detected. The resulting readout is used to control the piezoelectric transducer through a feedback loop and, eventually, to determine the relative phase between the beams.

This relative phase can be extracted as long as the PZT's response to a given voltage is precisely known. Bath researchers calibrated their system using a sawtooth wave form of known amplitude. As a result, they could map each applied voltage onto a specific PZT displacement and, therefore, onto a relative phase for the wavelength used -- 670 nm in this experiment (Figure 2).

Figure 2. (Top) The output from the two detectors, before feedback, as the probe is moved across the sample. Using feedback, the changing relative phase can be extracted.

Univ. of Bath researchers performed experiments using two different objects. First, they examined a traveling evanescent wave produced through total internal reflection in a prism (lower part of Figure 1). They used the interferometer to determine the spacing between the fringes (phase maxima and minima), and were then able to compare the result with theory. The error, just 3 percent, is attributed to miscalibration of the PZT.

A more complex experiment involved tapering an optical fiber at one point on its length, so as to strip off all but the fundamental and LP₁₁ modes. In order to measure the component they were interested in, the standing wave rather than the traveling wave, researchers had to make sure the fiber was cleaved exactly perpendicular to its long axis. In addition, the data had to be corrected with a scan of a flat phase front in order to remove any phase-change components created by off-axis scanning (coupling between the two SNOM axes). The resulting measurements were found to agree well with accepted values.

Having shown that the technique works in known cases, the team now wants to use it to characterize various types of photonic crystals with unknown field distributions, including the photonic band gap fiber recently developed within the group. Better understanding of structures at this level will, they believe, lead to faster development of optoelectronic devices.

References:

1. P.L. Phillips, J.C. Knight, J.M. Pottage, G. Kakarantzas, and P.St.J. Russell, Direct measurement of optical phase in the near field, App. Phys. Lett. 76 (5), 31 January 2000.