The early ATR and FMIR attachments had vertical sampling surfaces, which are compatible with dispersive spectrometers with their rectangular slits.  Horizontal sampling surfaces are more convenient however because the sample rests naturally on them.  Tony Gilbey, Wilks Scientific's Chief Engineer, designed a horizontal (HATR) unit with a complex optical system that rotated the beam 90° - and lined up the spectrometer beam with the entrance face of the MIR plate.  This soon became unnecessary with the advent of FTIR instruments because, with their circular beams, entrance face orientation beam orientation makes no difference.  Not all of the circular FTIR beam gets into the rectangular face of the MIR plate however but this is rectified with the use of circular internal reflection (CIR) optical elements.

In a CIR system, radiation is focused on the cone-shaped end of a cylinder and is then reflected down the cylinder to the other cone-shaped end, where it exits and is collected and refocused onto detectors.  Jim Harrick, in a letter to Applied Spectroscopy, demonstrated mathematically that CIR wouldn't work.  Pragmatically, however, CIR does work as demonstrated by close to 2000 beverage monitors made by General Analysis Corporation that continually monitor sugar and carbonation levels in soft drinks on beverage lines with a CIR sensor.  And both Spectra Tech and Axxiom have introduced CIR sampling attachments.

I recall a most delightful event concerning ATR that happened totally by chance.  I was in Amsterdam on business and had a free day, so I called Jacques Fahrenfort at Shell and asked if I could pay him a visit.  He was most cordial and said my timing was propitious because Jim Harrick was also to visit that afternoon.  Needless to say, we had much to talk about as our discussions continued well into the night after having dinner at Jacques suburban Amsterdam home.

And what is the status of ATR today?  It's safe to say that at least 50% of the mid-IR measurements made in the laboratory on liquid and solid samples utilize ATR in some form.  Many of the mid-IR microscope attachments use single- or multi-reflection diamond crystals as ATR elements.  And with the increasing interest in using the mid-IR for process monitoring and control, ATR probes are preferred over transmission because of their short effective pathlength and freedom from fouling - a problem with transmission process cells.  Furthermore, there is a whole new family of mid-IR sensors on the horizon using trapezoidal and hemispherical optical elements that produce the ATR effect.  These sensors will plug into the process stream wherever needed discussion to monitor concentrations of one or several components in the stream.  Mid-IR has the distinct advantage over single valued sensors such as conductivity or pH indicators and density and refractive index meters, in that mid-IR sensors are multi-valued.  They can determine concentrations of individual components in a multi-component stream, while the others cannot indicate which component in a mixture causes change in, for example, conductivity or refractive index.

Thus, the infrared measuring technology introduced forty years ago by Fahrenfort and Harrick will see increasingly broad utilization in the future in such unrelated locations as food and household product processing, vehicle lubrication and hydraulic fluid monitoring, the manufacture of chemicals and pharmaceuticals and certainly in medical diagnostic applications.  ATR might even find its way into the household bar to measure the alcoholic content of myladies' drink!   Fahrenfort and Harrick have made a valuable and lasting contribution to the field of materials analysis.