HeNe interferogram

Figure 8.11. Expanded plot of the spectra shown in Figure 8.10, showing apparent spectral artifacts in the region where the spectral intensity should be zero. The artifact at about 1500 cm has a similar structure to the absorption band of water at 7000cm and somehow results from folding, possibly because the interferogram was sampled at every zero crossing when the HeNe interferogram was not exactly centered at zero.

The response time of the detector should also be borne in mind when the time resolution is to be reduced well below 1 s. If the firequency of the HeNe interferogram is raised much above 10 kHz, the response of the DTGS detector is too slow and a faster detector must be used. In the mid-infrared, this is not a major problem, since MCT detectors operate optimally for modulation frequencies above 1 kHz. For near-infrared measurements, however, while InSb has a very fast response time, other quantum detectors, such as InGaAs, cannot be used at data acquisition speeds much above 5 kHz (see Section 18.2.5). 

Figure 10.11—Optical arrangement of a Fourier transform IR spectrometer, a) A 90c Michelson interferometer including the details of the beam splitter (expanded view) b) optical diagram of a single beam spectrometer (based on a Nicolet model). A weak intensity HeNe laser (632.8 nm) is used as an internal standard to measure precisely the position of the moving mirror using an interference method (a simple sinusoidal interferogram caused by the laser is produced within the device). According to the Nyquist theorem, at least two points per period are needed to calculate the wavelength within the given spectrum.

Atom-radical reactions have recently been investigated using the CS time-resolved instrument (with interleaved sampling) in Leone s laboratory [51,86]. Infrared emission from HC1 was observed in the chain chlorination of C2H6, initiated by 351-nm photolysis of Cl2 in the presence of the alkane [86]. Interferograms were recorded at 34-/is time intervals (fixed by the sampling rate of the HeNe crossings) at a resolution of 0.36 cm -1 as shown in... 

The digitized (ADC) discrete interferogram is Fourier transformed (DFT) by a PC to yield a wavenumber dependent spectrum. Sampling points are determined by the interference pattern of a monochromatic HeNe laser beam which is transferred collinearly with the IR beam. The resulting high wavenumber accuracy constitutes the third advantage of FTIR. 

Most near-infrared (NIR) spectra are measured from 2500 to llOOnm (4000 to 9100 cm ). To record a spectrum over this range, the interferogram must be sampled twice per wavelength of the HeNe laser interferogram to give a bandpass... 

For a Michelson interferometer, the rate of change of retardation is twice the mechanical velocity of the mirror. This parameter is often known as the optical velocity, V, and has the units of centimeters (retardation) per second. Some FT-IR vendors specify the velocity of the moving mirror in their interferometer in terms of its mechanical velocity and some in terms of its optical velocity. Note Users should make sure which of these two parameters is in fact specified.) Still others specify the optical velocity in terms of the corresponding modulation frequency of the laser interferogram although a few commercial interferometers no longer use HeNe lasers, we will denote the frequency of the laser interferogram as /neNe-... 

The SNR of spectra measured interferometrically is determined in part by how accurately the position of the moving mirror is known (see Eq. 7.12). Many measurements made using a step-scan interferometer require the OPD to be held constant to better than 1 nm (just a few atomic diameters ). To achieve this goal, the interference record from the HeNe laser must be measured. The points that correspond to the zero crossings of the laser interferogram are the points where the slope... 

Figure 8.5. Typical glitch on the baseline of a transmission spectrum. In this case the glitch is at 1020cm f Since the spectrum was measured with a data acquisition frequency of 5 kHz using a HeNe laser interferogram to trigger data acquisition, the glitch was caused by interference by a sinusoid of frequency 320Hz, possibly caused by a mirror vibrating at this frequency.

This problem can be illustrated by the data given in a paper by Hinsmann et al. [ 1], who used a Bruker Equinox 55 interferometer (see Figure 5.17) for stopped-flow measurements (see Section 19.2). The mirror velocity on this instrument was set so that data were acquired with a HeNe laser frequency of 280 kHz. For a spectral resolution of 32 cm , the time for each 1000-point interferogram was, therefore, 1024/ (280 X 10 ) s, or 3.66 ms. The time between successive spectra was reported as 45 ms, so that the duty-cycle efficiency was 8.1%. Clearly, the time resolution for this measurement was limited by the turnaround time and not the active scanning time. 

---