Of course it is possible to predict the peak widths theoretically. Check, for example, section “10.8.3 Peak Width”, pages 283-294, in chapter 10 of L.M. Blumberg, Temperature-Programmed Gas Chromatography, Wiley-VCH, 2010.
Amazon:
https://www.amazon.com/Temperature-Prog ... g=UTF8&me=
It was also possible some time ago to buy a pdf file of a single chapter from the publisher for $35.00:
http://onlinelibrary.wiley.com/doi/10.1 ... 5.ch10/pdf
Depending on specific conditions, expressions for the peak width calculations can be simpler or more complex. Probably the simplest case is the single-ramp temperature program in GC-MS at constant pressure. Suppose that the carrier gas is helium at the flow rate recommended in Agilent GC manuals (for helium, it is 8×ID×mL/min/mm) and heating rate is optimal (10 ºC per void time). Then all peaks eluting at the temperatures that are more than 60 ºC higher than the initial temperature of the heating ramp should have approximately the same widths. The half-height widths (w) of these peaks can be estimated with about 20% accuracy as:
w = (52 ms/m)×L
where L is the column length.
(Taking into account more details (as, for example, a predictable slight peak width increase during the ramp), more accurate theoretical predictions are possible.)
It is interesting to notice that the peak widths in GC-MS are proportional to the column length and do not depend on the column ID (internal diameter).
Example: ID=0.25 mm, L=30 m, initial temperature of the heating ramp is 40 ºC. The flow rate in this case is 8×0.25 mm×mL/min/mm = 2 mL/min resulting in 1 min void time for which the optimal heating rate is 10 ºC/min. Under these conditions, the peaks eluting when the column temperature is higher than 100 ºC have widths:
w = (52 ms/m)×30 m = 1.56 s.
Check it experimentally.
Good luck.
