Provider: ingentaconnect
Database: ingentaconnect
Content: application/x-research-info-systems
TY - ABST
AU - Everall, Neil J.
AU - Priestnall, Ian M.
AU - Clarke, Fiona
AU - Jayes, Linda
AU - Poulter, Graham
AU - Coombs, David
AU - George, Michael W.
TI - Preliminary Investigations into Macroscopic Attenuated Total Reflection–Fourier Transform Infrared Imaging of Intact Spherical Domains: Spatial Resolution and Image Distortion
JO - Applied Spectroscopy
PY - 2009-03-01T00:00:00///
VL - 63
IS - 3
SP - 313
EP - 320
KW - ATTENUATED TOTAL REFLECTION
KW - PSF
KW - SPATIAL RESOLUTION
KW - FOURIER TRANSFORM INFRARED SPECTROSCOPY
KW - ARTIFACTS
KW - GLOBAL IMAGING
KW - FT-IR SPECTROSCOPY
KW - POINT SPREAD FUNCTION
KW - IMAGE DISTORTION
KW - ATR
N2 - This paper describes preliminary investigations into the spatial resolution of macro attenuated total reflection (ATR) Fourier transform infrared (FT-IR) imaging and the distortions that arise when imaging intact, convex domains, using spheres as an extreme example. The competing effects
of shallow evanescent wave penetration and blurring due to finite spatial resolution meant that spheres within the range 20–140 μm all appeared to be approximately the same size (∼30–35 μm) when imaged with a numerical aperture (NA) of ∼0.2. A very simple model was
developed that predicted this extreme insensitivity to particle size. On the basis of these studies, it is anticipated that ATR imaging at this NA will be insensitive to the size of intact highly convex objects. A higher numerical aperture device should give a better estimate of the size of
small spheres, owing to superior spatial resolution, but large spheres should still appear undersized due to the shallow sampling depth. An estimate of the point spread function (PSF) was required in order to develop and apply the model. The PSF was measured by imaging a sharp interface; assuming
an Airy profile, the PSF width (distance from central maximum to first minimum) was estimated to be ∼20 and 30 μm for IR bands at 1600 and 1000 cm^{−1}, respectively. This work has two significant limitations. First, underestimation of domain size only arises when imaging
intact convex objects; if surfaces are prepared that randomly and representatively section through domains, the images can be analyzed to calculate parameters such as domain size, area, and volume. Second, the model ignores reflection and refraction and assumes weak absorption; hence, the
predicted intensity profiles are not expected to be accurate; they merely give a rough estimate of the apparent sphere size. Much further work is required to place the field of quantitative ATR-FT-IR imaging on a sound basis.
UR - http://www.ingentaconnect.com/content/sas/sas/2009/00000063/00000003/art00011
M3 - doi:10.1366/000370209787598951
UR - http://dx.doi.org/10.1366/000370209787598951
ER -