Nitty-Gritty Details of DSM CD

The CD measurement begins with the digitization of the voltage signals generated by the photomultiplier tubes (PMTs). These signals represent the time-dependent intensity of the light detected by each phototube. Once digitized, the CD information is extracted from these signals in an entirely digital manner using mathematical manipulations derived by rigorous analysis of the optical system.

The digitized signal consists of a DC level, a 50 kHz signal, and noise. The digital processing derives the CD amplitude from the DC and 50 kHz signals while repressing the noise.

Step one is to measure the DC component of each signal by computing the time average of each signal. This DC level is derived from the same data as the 50 kHz signal, so there is no need to deal with independent amplification of each component and thus no need for a photometric calibration using a chemical standard.

With the DC amplitudes known, second step is to normalize each signal by dividing each by ‘their respective DC level. The resulting signals have DC amplitudes of unity with normalized 50 kHz components. The CD information is derived from these normalized signals; thus, the absolute DC amplitude of either signal before normalization is of no consequence to obtaining the right answer. That is, the two beams of light do not need to be identical (although they happen to be very nearly identical). The absolute DC amplitudes prior to normalization do not contain information about the CD signal¹.

Step three is the subtraction of one normalized signal from the other. The results are (1) the constructive interference between the 50 kHz signals in the two channels and (2) the destructiveinterference of all remaining components. In detail, the 50 kHz modulation which contains the CD information is phase shifted exactly Ð radians between the two channels. This phase difference is a consequence of the orthogonal polarization states produced by the beam splitter and, as such, is absolute and unchanging. The subtraction of the two channels is essentially equivalent to a phase shift followed by an addition. As a consequence, the two ‘out of phase’ signals are brought ‘into phase’ and then added.

The destructive interference is achieved in exactly the same manner, except here, the two signals are originally ‘in phase’ and are brought ‘out of phase’ and then added; their sum is zero. These originally in-phase signals includes noise produced by sources before the PMTs (i.e., lamp fluctuations).

At this point, we have a single data set which is the result of the subtraction of the two normalized signals. These data can be visualized as having the 50 kHz modulation and noise. The time average of these components is zero. If we now apply a 50 kHz rectification to the data, we are effectively multiplying the data by a 50 kHz square wave which varies between +1 and -1 and has the same phase as the 50 kHz modulation in the signal. After this operation, the ‘rectified’ 50 kHz component has a non-zero time average, whereas the time average of the noise is unaffected, remaining zero. The final step is to average the data. Again, the noise will average to zero and the result is the CD signal¹.

W. Curtis Johnson Coined “DSM”

The overall purpose of this document is to help you understand the unique technological advances in our instrument. The Olis Dual Beam CD spectrophotometers are the first and only new CD spectrophotometers in decades. These are not me-too solutions to an old problem. The Olis Dual Beam CD spectrophotometers are fundamentally different. They are CD Done Right.

We close with a quotation from W. Curtis Johnson's chapter in Fasman's 1996 book, CD & Conf Ana of B (Plenum Press):

“There are three methods for measuring the CD of a sample. The most straight-forward is to measure the absorption for each rotation of light and subtract directly the measurement for right circularly polarized light from the measurement for left circularly polarized light. However, this requires making each measurement to great accuracy, which was not practical until the recent availability of powerful and inexpensive computers...Inexpensive, high speed digital computers have made direct subtraction of left and right circularly polarized beams a practical method of measuring CD. This method is pioneered in a commercial instrument for electronic CD by On-Line Instrument Systems (Olis, Bogart, Georgia)...This dual beam collection and direct subtraction method has two advantages in addition to the flat baseline. First...CD signals are measured correctly. Second, the instrument is measuring absorbance directly, so there is no constant of proportionality to calibrate.”[Quoted from pages 640 and 644-5, underlining added]

(To satisfy your curiosity, if you do not have this book, the other two means of measuring CD, according to Dr. Johnson, are “to measure the ellipticity imparted on linearly polarized light that passes through the sample. This method is limited by the quality of the linearly polarized light and the mechanics of measuring the small perpendicular component of the elliptically polarized light. The third method is to modulate the light between the two rotations, and measure the difference at each wavelength. This method is well suited to analog electronics, and is the method used by most commercial instrumentation.”)