It was at the 1960 Eastern Analytical Symposium that I first heard about a new sampling technique that enabled the user to obtain infrared spectra on materials opaque in even the thinnest of transmission cells.  At that time, the Symposium was held in the Fall at the Hotel New Yorker in New York City.  In those early days it attracted mostly infrared spectrometer manufacturers - mid infrared, that is! 

Among the exhibitors was Connecticut Instrument Corporation (CIC), a four-year-old company founded by Charles W. Warren and the author which had begun to build a product line of infrared sampling accessories.  It had introduced "cavity cells" - transmission cells ultrasonically machined from a single block of rock salt and the demountable precision fixed thickness cell that is still the basic design of many of the transmission cells in use today. 

During that 1960 EAS meeting, Don Johnson, an infrared specialist from DuPont, came into our booth and described to me a method of obtaining infrared spectra on a variety of materials not previously amenable to IR analysis without extensive sample preparation.  Don told me that spectroscopists at the Experimental Station were very excited about the new procedure because it greatly expanded the usefulness of infrared spectroscopy.  It was called "Attenuated Total Reflection" and had been conceived of by Jacques Fahrenfort of the Shell Laboratories in Amsterdam.  Don suggested that we explore the technique.  Later in the meeting I gave Abe Savitzky a ride back to Connecticut and he described the new technique in detail to me indicating that Perkin Elmer had become quite interested in it.

Farhenfort had made use of a phenomenon that was well known to optical physicists, the fact that when radiation is totally internally reflected within a transmitting medium, an electric field coupled to the radiation is set up at each reflection point.  He found that if a substance was placed in contact with the reflecting surface, energy would be absorbed from the electric field and hence from the radiation it was coupled to at wavelengths where that particular substance absorbed.  Thus the radiation beam was "attenuated" at these wavelengths.  To us pragmatic spectroscopists, it appeared that the beam actually penetrated slightly beyond the reflecting surface, in the neighborhood of a micron or so - the actual depth depending on a number of factors (angle of incidence, relative refractive indices, number of reflections, etc.).  The net result was that here was a method that would produce the very short effective path length through a material required to record its infrared spectrum no matter what the actual thickness of the sample was. 

The ATR effect turned out to be proportional to wavelength, i.e., the longer the wavelength the greater the beam penetration.  Hence the relative absorption of ATR spectra to transmission spectra increases with wavelength.  For this reason, ATR works better in the mid-IR than the NIR - but this is where the best resolved and separated fundamental absorption bands occur anyway.

This was all very exciting to us at CIC, for it appeared to be an opportunity to create a whole new line of  IR sampling accessories.  With advice and council from workers at DuPont and Shell Development Company in Emeryville, CA we were able to design and introduce a single- reflection, variable-angle attachment for IR spectrometers by mid-1961.  We used mostly silver chloride and KRS-5 as prism materials because they were the only high index IR transmitting crystals then generally available.  In the case of silver chloride, we found that we could impress rows of angled ridges on the back of a two millimeter thick sheet that would cause radiation hitting that side to be transmitted to the smooth front surface, from whence it would reflect back out of the sheet.  This gave us an inexpensive disposable ATR sample holder.

For reasons that are not important to this article, CIC was acquired about this time by Barnes Engineering.  I stayed with the Barnes organization for a year, then decided to strike out on my own again and formed Wilks Scientific Corporation.  The CIC product line was later spun off from Barnes and became the nucleus of Spectra Tech.

I had become aware of the work of the other key figure in the field of total internal reflection, N. James Harrick then of Philips Laboratories in Tarrytown, NY.  Jim's interest in internal reflection was based in his studies of surface chemistry and, instead of a prism he was using relatively thin plates with beveled ends and causing radiation to be reflected back and forth from surface to surface, thus multiplying the absorption between the radiation beam and the coating on the surface.

I realized that this added a whole new dimension to ATR spectroscopy, and introduced a multiple internal reflection attachment through Wilks Scientific in 1963, the basic design of which is still in use today.  Because I had a no-compete restriction with Barnes, we called our device a "Frustrated Multiple Internal Reflection" (FMIR) attachment after Jim Harrick's terminology in order to convince the legal profession that ATR and FMIR were not the same.  We were successful, although we did stir up a certain amount of controversy in the infrared field as to which was the correct terminology.

Jim and I had several conversations about joining forces but he wanted a 50% interest in Wilks Scientific and I was willing to give him 20%.  He finally went off and formed Harrick Scientific, which was probably the best for both of us and the world in general because Harrick Scientific has made and continues to make many significant contributions to our field.  Jim and I in the same small organization would probably not have been compatible anyway!