Peak Broadening.
As pointed out in the post column void volume section, the width of peaks as they appear in the mass spectrometer is a result of dispersion in injection, the column itself and the detector, or:

σpeak = SQRT(σ2injector+σ2column2esi interface)

Neglecting broadening during injection (peaks adsorb ina narrow band on top of the column) and a σcolumnof 2 sec, than the equation simplifies into:

     σpeak = SQRT(4+σ2esi interface) .

Peak broadening (Vobserved) in  the ESI emitter follows from the peak volume at the end of the column (Vpeak= peak width x flow rate) and the emitter dead volume (Vextra):

Vobserved = Vpeak x SQRT{1+(Vextra/Vpeak)2 }

Hence, the width in which peaks are detected dependents on the column and the quality of the ESI interface. Small particle sizes results in narrow peaks leaving the column but are very susceptible for broadening in the dead volume of the emitter. It follows that the total dispersion occurs predominantly in the emitter and not in the column. Hence, use of small particles is almost useless with badly designed or plumped emitters. The table shows that a gain in separation efficiency by smaller particle sizes is lost by a poor ESI interface

particle size Column ID Flowrate peakwidth 4σ peakwidth 4σ peakwidth 4σ
(um) (um) (nL/min) no post column dead volume  5 nL post column dead volume  50 nL post column dead volume
1.7 50 120 3.5 5.3 (150%) 25.2 (720%)
1.7 75 240 3.5 3.9 (13%) 12.9 (370%)
3 50 120 6 6.5 (8%) 25.7 (430%)
3 75 240 6 6.1 (2%) 13.8 (230%)
5 50 120 10 10.3 (3%) 26.9 (270%)
5 75 240 10 10.1 (1%) 16   (160%)

Table: peak broadening (%) caused by a post column dead volume, i.e the ESI emitter or the connection to this emitter, with 50 and 75 umID columns packed with 1.7, 3 and 5 um particles, resp.