Peak width, chart recorder

The mixture is identical in each example. The peaks are shown separated by 2, 3, 4, 5 and 6 (a) and it is clear that a separation of 6a would appear to be ideal for accurate quantitative results. Such a resolution, however, will often require very high efficiencies which will be accompanied by very long analysis times. Furthermore, a separation of 6o is not necessary for accurate quantitative analysis. Even with manual measurements made directly on the chromatogram from a strip chart recorder, accurate quantitative results can be obtained with a separation of only 4a. That is to say that duplicate measurements of peak area or peak height should not differ by more than 2%. (A separation of 4a means that the distance between the maxima of the two peaks is equal to twice the peak widths). If the chromatographic data is acquired and processed by a computer, then with modem software, a separation of 4a is quite adequate. 

When the baseline varies randomly over times similar to the peak width, it is called wander (Figure 5.2). Wander interferes most with interpreting a chart recording or the operation of an electronic integrator. It can come either from the detector or from extraneous sources, such as a small leak at a column fitting. 

A major disadvantage of manual methods is that the accuracy of each measurement relies on the individual performing the calculations and on where the tangents or baselines are drawn, which may differ from one peak to the next within an analysis. If strip-chart recorders are used, all the peaks must be contained on the chart paper. This severely limits the dynamic range of solute composition that can be analyzed. In addition, the chart speed must be fast enough so that, for measurements based on peak width, narrow peaks are wide enough for accurate measurement. 

If peaks are symmetrical, as is frequently the case with modern capillary columns, only the peak height must be converted. When recording the chromatograms on a strip chart recorder, the chart speed on the recorder should be increased to obtain an accurate measurement of the peak width. The area can then be obtained by triangulation. Other required data are the flow rate through the detector and the moles of sample in the detector. For samples that are split the measurement of the split ratio is also required to obtain the number of moles injected. The... 

Table 5.1 presents data taken to characterize a high-temperature Ni-63 detector using Q,A series of solutions were prepared at concentrations of 5 x 10-5, 5 x 10 6, and 5 x 10 7 moles of C6F6/mL heptane. The samples were injected into the chromatograph with a 10-pL syringe without splitting. The total flow rate was 150 mL/min at 300 K. The data were recorded on a strip chart recorder and the peak width converted to time using the chart speed of the recorder. 

In the calibration curve, Aris plotted on the ordinate versus log Cao or log Cs on the abscissa At (min) is directly read off the recorder paper as the peak width b (mm)—measuring the peak width of all peaks at the same level, approximately halfway between the baseline and the top of the peak—that is, b = At Mr, where Ur is the recorder chart speed (mm/ min). 

During precise determinations of m.r.f. values. Girling and co-workers " " observed that the repeatability of the results was governed by the peak widths in successive chromatograms. These variations were attributed to irregularities in the chart movement, and, of seven recorders tested, the worst showed a variation of 3.65%, and the best, 0.033%. 

Determination of plate number is simple from a detector chart recording, since tangents to such a curve at the point of inflexion intercept the baseUne at a distance of 4o apart. Assuming that peak shapes are Gaussian the peak width is therefore given by 4ct. This can... 

Frequently, it is desirable to use the window mode in conjunction with a pulse-height-selector scanner in order to measure the spectrum of Fig. 4.16(a) on a strip-chart recorder. If the resolution of Fig. 4.16(a) is to be faithfully reproduced in Fig. 4.16(b), the window width must be small compared to the peak width in Fig. 4.16(a). That is, the window width must be small compared to the pulse height resolution inherent in the amplifier pulse height spectrum. A good practical choice is to make the window width less than one-fourth of the width of the narrowest peak in the amplifier pulse height spectrum—the width being taken as the width of the peak at half the peak amplitude above background. 

Duplicate samples are placed on the column, the sample solution is diluted by a factor of 3 and duplicate samples are again placed on the column. This procedure is repeated, increasing the detector sensitivity setting where necessary until the height of the eluted peak is commensurate with the noise level. If the detector has no data acquisition and processing facilities, then the peaks from the chart recorder can be used. The width of each peak at 0.607 of the peak height is measured and the peak volume can be calculated from the chart speed and the mobile-phase flow rate. Now,... 

All manual methods of quantifying peak size make use of the recorder tracing. Some consideration has. already been given to the peak height and width as determined by recorder chart width and chart speed. Both should be maximized for the size measurement technique used. In addition the recorder may have a limiting time constant as far as response to rapid peaks are concerned. This possibility should be considered along with the detector when time constant problems are suspected. 

---