Chemistry 663 Supplementary Material

1.) What is the "pooled blank standard deviation"? Right now I have entered in the tables I'm writing for the report the standard deviation of the 10 blank measurements we did for each element. [Is that the "pooled" blank s.d?]

Ans. The idea was to take 5 blank measurements 3 times or 10 blank measurements 2 times and than find an "average" standard deviation which is also denoted as the "pooled" standard deviation. It is determined as specified on page 6 of the report under part 8. Basically, the pooled standard deviation is the square root of [the sum of the squares of the individual standard deviations divided by the number of standard deviations pooled].

2.). Do you expect us also to show in our tables the calculated concentration for each standard? (the computer printouts from the ICP-AES software show the calculated concentration of each standard by plugging in the signal of each standard into the standard curve equation). I don't think it's necessary to do this, as I don't see any usefulness with that data, but maybe I'm overlooking something here.

Ans. You do not have to report the calculated concentration for the standards in a separate table. This information is useful especially at the time you are taking data. The calculated concentrations should be near the actual concentrations of the standards. If not, there may be a blank problem or an incorrect standard and the calibration curve should be run again.

3. I am uncertain by the meaning of RSD. (Is it relative standard deviation as in 1/S ?).

Ans. The RSD is the standard deviation divided by the mean and is usually given as a percent in the printer output.

4. With the measurements under the diagnostic mode, the signal does not vary with the slit width as expected.

Ans. First, note that the entrance and exit slits are different widths in this instrument. With a monochromatic source and an ideal monochromator, the image of the source in the focal plane at the monitored wavelength should be the size of the entrance slit. Once the exit slit is wider than the entrance slit, the signal should not increase with further increases in the exit slit width. Your data does follow this idea behavior. There are at least two possible explanations. First, the line source is not monochromatic and has a finite width of a few picometers which is on the order of the spectral bandpass. Even with an ideal monochromator, the image of the source in the focal would be a set of overlapping imaging with an overall width greater than the entrance slit. Second, the monochromator is not ideal and is being pushed to its resolution limit. The spectral bandpass is no longer R_dW but it determined by aberration in the monochromators optics. Even with a monochromatic source, the image in the focal plane will be blurred and wider than the exit slit.

Note that for the background radiation, we assume a continuum, and the background emission signal should be proportional to the exit slit width under ideal conditions.

5. Are measurements blank corrected in the diagnostic mode.?Ans. No. The signals you observed are unadulterated. When a standard is measured, the signal is the total emission signal which includes analyte emission, background emission, and dark signal.

Two different spectrometers are used for the two parts of this experiment. Thus, we will form two groups which will use the instruments on different days. The experiment will be conducted over three weeks with the following schedule.

Chemistry 663 Homepage

E-mail Dr. Ingle