Entrance -

Field mirrors "M2"

Added ' ^>retroreflecting mirrors


FIGURE 11.1 (a) Schematic diagram of a multipass White cell, (b) sequence of images on filled mirror for White cell design, and (c) sequence of images on field mirror for Horn and Pimentel design (1971). (Adapted from Finlayson-Pitts and Pitts, 1986; and Hanst and Hanst, 1994.)


number of spots on the field mirror) has been made in this case.

The advantage of such a White cell is that the source is reimaged on the field mirror M2 after each double traversal of the cell. This keeps the energy that enters the cell within the mirror system so that energy losses occur mainly through light absorption by the mirrors and, of course, by the gases in the cell. In practice, the loss of light energy through absorption by the mirrors imposes a major limitation on the number of passes that can be used. The fraction of the energy lost after n reflections from a mirror whose reflectivity is R is given by (1 — R"). Thus, if a mirror reflects 98% of the incident light and absorbs 2%, only 36% of the incident intensity will remain after 50 reflections from the mirror. After fOO reflections, only 13% of the incident intensity is left. While the path length and hence absorbance have increased, the energy loss may be so severe that such a large number of reflections becomes impractical.

The number of reflections is also limited by the size of the image striking the entrance and the size of the mirror M2. As seen in Fig. ff.fb, the images that are refocused from Ml and M3 onto the field mirror M2 are "stacked" beside each other. The width of M2 therefore determines how many of these images can be accommodated (i.e., how many reflections are possible).

A practical problem arising when the images are too closely spaced (i.e., at long path length) is one of adjustment; temperature changes, for example, can cause very small changes in the mirror adjustments which result in moving the exit beam away from the exit aperture.

Variations of the White cell are also in use. For example, Horn and Pimentel (1971) added a corner mirror assembly to redirect the beam that would normally exit the cell back into it. This doubles the number of passes, giving four rows of spots on the field mirror. The image pattern for such a design is shown in Fig. 11.1c (Hanst and Hanst, 1994).

More complex multiple-reflection systems that give a much greater number of traversals have also been developed. For example, Tuazon et al. (1980) describe a system using four collecting mirrors that focus the light onto four field mirrors. The advantages and disadvantages of such multiple-mirror cells are discussed by Hanst (1971) and Hanst and Hanst (1994).

An alternate design for folded optics was described by White in 1976. In this design, the light beam is folded back on itself, giving larger path lengths and greater optical stability. The effects of vibration, thermal expansion, and astigmatism are reduced and alignment errors are minimized with this design.

Window IR Beams In/Out

Glass liner

pipe Exhaust

Input mirror

Inlet nozzle

Inlet nozzle

Back mirror

Mirror Separation & tilt adjustment pipe Exhaust

Back mirror

Input mirror

Spherical "herrlot" cell

Astigmatic "herrlot" cell

Spherical "herrlot" cell

Astigmatic "herrlot" cell

FIGURE 11.2 (a) Schematic diagram of multipass cell for infrared spectroscopy using astigmatic Herriott configuration (adapted from McManus et al, 1995), (b) spot configurations for normal Herriott multipass cell, and (c) spot configurations for astigmatic configuration (adapted from Zahniser et al, 1997).

A second multipass cell configuration is the Herriott cell (Herriott et al., 1964; Herriott and Schulte, 1965). This is particularly useful for coherent light sources such as lasers used in tunable diode laser spectroscopy but has also been used with incoherent light sources using optical fibers cemented to a ball lens at the entrance to the cell (Zahniser, personal communication). Two spherical mirrors are separated by a distance close to their radius of curvature, and the light beam enters through a hole in one of them, directed in an off-axis direction. After multiple reflections between the two mirrors, the light beam exits through the same hole as it entered, but at a different angle (Fig. 11.2a). The beam remains collimated throughout, in contrast to the White cell system, and gives the spot pattern shown in Fig. If.2b. The path length is changed by changing the distance between the mirrors; in practice this means that this design is most useful for fixed path length systems.

An astigmatic variant of the Herriott cell designed for use in ambient air studies is shown in Fig. If ,2a and described by McManus et al. (1995) and Zahniser et al. (f997). in this design, the two mirrors have different radii of curvature, giving the spot patterns shown in Fig. ff.2c. The spots more evenly fill the mirror, so that for a given number of passes, the spots are more widely spaced, or conversely, more passes can be obtained without problems of beam overlap (McManus et al., 1995).

Major advantages of such cells are that they are relatively easy to align and folded optical paths can be obtained in small volumes. This is important when small amounts of sample are available, for example, in laboratory studies or when a fast response is needed; cells of smaller volume can be pumped out faster, giving shorter residence times in the cell.

(2) FTIR Fourier transform infrared spectroscopy has been used for many years to measure atmospheric gases. Because FTIR has become such a common analytical method, we do not describe the technique itself here but rather refer the reader to several excellent books and articles on the subject (e.g., see Griffiths and de Haseth, 1986; Wayne, 1987). For reviews of some atmospheric applications, see Tuazon et al. (1978,1980), Marshall et al. (1994), and Hanst and Hanst (1994).

A problem in the application of FTtR to ambient air is that water vapor, C02, and CH4 are all present in significant concentrations and absorb strongly in certain regions of the spectrum. As a result, the spectral regions that are useful for ambient air measurements are 760- to 1300-cm"1, 2000- to 2230-cnT1, and 2390-to 3000-cm"1.

PDT HN03 (ppb)

Was this article helpful?

0 0

Post a comment