Page 129 - Vibrational Spectroscopic Imaging for Biomedical Applications

Evanescent W ave Imaging 105

25
to noise. Lewis and Sommer also demonstrated that the spherical
aberrations could be compensated for by collecting a background at
the exact same off-axis position as the sample and, more importantly,
that the penetration depth changed for each off-axis position as a
result of the changing incident angle within the crystal. Should
quantitative information be required, the change in penetration depth
would need to be accounted for. The work of Nakano and Kawata
demonstrated the usefulness of ATR microspectroscopy, but the draw-
back to their system was its complexity. The work of Lewis and Som-
mer showed that the method could be employed on a conventional
system, but with a sacrifice in the theoretical spatial resolution albeit
better than a transmission measurement. Earlier, Sommer and Katon
demonstrated that the main cause for the degradation in spatial
resolution in an infrared microscope was the use of the confocal
aperture employed to isolate the sample of interest. 27
At about the same time these rudimentary microscopic ATR
mapping experiments were conducted, array detectors became avail-
able, principally in the near-infrared region, but not too long after
mid-infrared detectors also became available. Lewis et al. reported on
a near–infrared microscope outfitted with an InSb array detector. 28
Soon thereafter (1996), Biorad introduced the first commercial system
with an InSb array detector interfaced to a microscope, the Stingray
1. In 1997, Kidder et al. reported on a similar system, but with a
mid-infrared mercury cadmium telluride (MCT) detector inter-
29
faced to the microscope. In both, the report of Lewis et al. and
Kidder et al. diffraction-limited performance was not achieved.
However, Lewis et al. anticipated diffraction-limited performance
and Kidder et al. came within a factor of 2. In theory, the degradation
in spatial resolution should be greater in a conventional microscope
rather than an array-based system. In a conventional infrared microscope,
the majority of diffraction occurs from the confocal aperture (high-
contrast edge) located at the primary image plane of the sample.
Radiation diffracted by the aperture then propagates onward to a rela-
tively large area detector (i.e., 100 × 100 μm), where it is detected and
degrades the theoretical spatial resolution. By removing the aperture, the
most significant source of diffraction is eliminated. In the array-based
system, true diffraction-limited performance can be observed so long as
the diffraction-limited beam diameter at the sample, when imaged onto
the detector, is greater than the pixel size on the array (vide infra).
In 1997, Biorad introduced an infrared microscope outfitted with
an MCT array detector, the Stingray. In the spring of 1999, Sommer
attempted to repeat the ATR experiments of Lewis on the Stingray
system in the 3M laboratory of Rebecca Dittmar. However, due to
operational problems with the Stingray, the experiments were
unsuccessful. Later in the spring of 2000, Sommer attempted the same
experiments in the Procter and Gamble laboratory of Curtis Marcott.
These experiments were successful and demonstrated ATR imaging