Proceedings Article | 1 February 2008
KEYWORDS: Photodetectors, Optical imaging, Diffraction, Electrodes, Signal processing, Sensors, Spatial resolution, Image quality, Scanning electron microscopy, Fourier transforms
With the advance of nano-lithography and nano-fabrication, individual sizes of electronic,
photonic, and mechanical components, as well as their integration densities, have progressed
steadily towards the sub-100 nm regime. Therefore, being able to image such feature sizes
becomes imperative. Many conventional high-resolution imaging tools such as SEM, STM,
AFM, and NSOM either require operation under high vacuum or slow scanning across the
sample. A far-field optical imaging instrument would thus be highly desirable. Optical imaging,
however, is subject to the diffraction limit, which limits the size of the smallest resolvable
feature to be ~ λ/2, where λ is the wavelength of the imaging light. Recently, negative-index materials and super lens have been proposed to overcome this limit and
achieve high-resolution optical imaging [1-4]. In this paper, we propose a different approach to
achieve sub-diffraction optical imaging with far-field microscopy. The technology builds on a
high-spatial resolution quantum-dot (QD) photodetector with high sensitivity that we have
demonstrated [5]. The photodetector consists of several nanocrystal QDs between a pair of
electrodes with 50-nm width spaced ~ 25 nm apart. An optically effective area of 13515 nm2 was
determined by modeling the electric field distribution in-between and around the electrodes using
FEMLab. High-sensitivity photodetection has been demonstrated by measuring the tunneling
photocurrent through the QDs, with a detection limit of 62 pW of the input optical power. The
proposed sub-diffraction optical imaging system consists of an array of such photodetectors. We
performed theoretical simulations assuming a two slit source and then pixilated the far-field
diffraction pattern to simulate the photodetector array. A Fourier transform of the detector signal
is then performed to determine how much of the original aperture information remains. Using a
wavelength of 500 nm and a screen distance of 10 cm, we found that, as expected, the quality of
the resultant image generally degraded with larger pixilation size. With 50-nm one-dimensional
spatial resolution at the detection plane, it appears that the original slit image with 100-nm width
and 300-nm spacing can still be restored.
