in the cell (i.e. the length of the cell) the peak can exhibit various degrees of dispersion and types of distortion. However, it must be emphasized that the curves shown in figure 18 only occur when there is true Newtonian flow through the cell (i.e., all flow rates are below the maximum where turbulent flow is initiated). If, by appropriate design, the parabolic velocity profile of the fluid flowing through the cell can be disrupted, then the dispersion and distortion arising from Newtonian Flow can be virtually eliminated. In practice, the disruption of the normal parabolic velocity profile of the fluid flowing through the cell can be readily achieved by modifying the manner in which the conduits cause the mobile phase to enter and exit the cell. The conduit connections to the cell are oriented  to produce secondary flow in the manner shown in figure 19. Mobile phase enters the cell at an angle so that fluid stream is directed at the cell window. As a result of this, the

having a given extinction coefficient (k), (l) should be increased. However, in a flow-through sensor cell of a GC/spectrometer combination, the extent to which the path length can be increased is limited, as the total volume of the sensor cell must be restricted to ensure minimum peak dispersion and only a small fraction of a peak can be allowed to exist in the cell at any one time (4). To restrict peak dispersion and maintain a small sensor volume, it follows that the radius of the sensor cell must also be reduced as (l) is increased. This results in less light falling on the photoÐcell which, in turn, will reduce the signalÐtoÐnoise ratio and, thus, the sensor sensitivity, or minimum detectable concentration. Consequently, increasing the sensor sensitivity by increasing the path length has limitations and a wellÐdesigned cell involves a careful compromise between cell radius and length to provide the maximum sensitivity. Most modern UV spectrometer sensor's have path lengths

nbsp;   Figure 6 The Design of a Modern Absorption Cell   The Newtonian flow is distorted by the manner in which the inlet and outlet conduits are connected to and from the cell. Mobile phase enters the cell at an angle that is directed at the cell window. It follows, that the mobile phase flow has to virtually reverse its direction to pass through the cell producing a swirling action which introduces strong radial flow and disrupts the Newtonian flow. The effect also occurs at the exit end of the cell. The flow along the axis of the cell now must reverse its direction to pass out of the port which is accomplished by attaching the exit conduit at an angle to the axis of the cell. Employing this type of entry and exit connections eliminates dispersion resulting from viscous flow

Dispersion in the Detector Sensor Volume The finite nature of the detector sensor volume can cause peak dispersion and contribute to the peak variance by two processes. Firstly there will be dispersion resulting from the Newtonian flow of fluid through the cell in much the same manner as the flow of a viscous fluid through an open tube. This will furnish a variance similar in form to that predicted by Golay but, as the tube length is small and the tube length to radius ratio much larger than that from a connecting tube, a different equation is necessary to describe the dispersion effect. Secondly, there will be a peak spreading which results from the finite volume of the sensor. If the sensor has a significant volume, the concentration

The Effect of Viscous Flow on Dispersion in a Detector Sensor Scott and Simpson (19) also measured the overall effect of dispersion in sensor cells (that is both that due to Newtonian Flow and that due to sensor volume capacity) by simulating a UV adsorption cell used in LC from simple short cylindrical tubes. The results they obtained are shown in table 4. Table 4. Dispersion in Sensor Cells   Cell Volume Cell I.D. Cell Length Vol.Variance Equi.Peak Vol. 0.59  (ml) 0.5 mm 3.0 mm 0.45 (ml2) 2.68 (4s) (ml) 1.96  (ml) 0.5 mm 10.0 mm 1.19 (ml2) 4.36 (4s) (ml) 2.75  (ml) 1.0 mm 3.5 mm 1.19 (ml2

used to produced the elution curves in the lower chromatogram was 1 m long, 1 mm I.D. and the same solvent was used at a flow rate of 40 ml/min Benzene was also used a the solute. It is seen that the reduction in cell volume has a dramatic effect on both peak width and peak shape. The large 25 ml cell causes significant peak asymmetry as well as excessive peak dispersion A result which is predicted by the work of Atwood and Golay (11) which is discussed below. It is seen that the large sensor cell has a disastrous effect on the band width of the solute eluted from the microbore column. Clearly, even cell volumes of 3 ml are too large for use with 1 mm I.D. columns and relatively few contemporary detectors have cell volumes less than 3 ml.    J. Chromatogr. 169(1979)51 Figure 17. Peak Profiles from Detector Having Different Cell Volumes