Abstract:
Factors involved in the determination of PAHs (16 priority PAHs as
an example) and PCBs (10 PCB congeners, representing 10 isomeric
groups) by capillary gas chromatography coupled with mass spectrometry
(GC/MS, for PAHs) and electron capture detection (GC/ECD , for PCBs)
were studied, with emphasis on the effect of solvent. Having various
volatilities and different polarities, solvent studied included
dichloromethane, acetonitrile, hexan e, cyclohexane, isooctane, octane,
nonane, dodecane, benzene, toluene, p-xylene, o-xylene, and mesitylene.
Temperatures of the capillary column, the injection port, the GC/MS
interface, the flow rates of carrier gas and make-up gas, and the
injection volume were optimized by one factor at a time method or
simplex optimization method.
Under the optimized conditions, both peak height and peak area of
16 PAHs, especially the late-eluting PAHs, were significantly enhanced
(1 to 500 times) by using relatively higher boiling point solvents such as
p-xylene and nonane, compared with commonly used solvents like benzene
and isooctane. With the improved sensitivity, detection limits of between
4.4 pg for naphthalene and 30.8 pg for benzo[g,h,i]perylene were obtained
when p-xylene was used as an injection solvent.
Effect of solvent on peak shape and peak intensity were found to be
greatly dependent on temperature parameters, especially the initial
temperature of the capillary column. The relationship between initial
temperature and shape of peaks from 16 PAHs and 10 PCBs were studied
and compared when toluene, p-xylene, isooctane, and nonane were used as
injection solvents. If a too low initial temperature was used, fronting or split of peaks was observed. On the other hand, peak tailing occurred at a
too high initial column temperature. The optimum initial temperature, at
which both peak fronting and tailing were avoided and symmetrical peaks
were obtained, depended on both solvents and the stationary phase of the
column used. On a methyl silicone column, the alkane solvents provided
wider optimum ranges of initial temperature than aromatic solvents did,
for achieving well-shaped symmetrical GC peaks. On a 5% diphenyl: 1%
vinyl: 94% dimethyl polysiloxane column, when the aromatic solvents
were used, the optimum initial temperature ranges for solutes to form
symmetrical peaks were improved to a similar degree as those when the
alkanes were used as injection solvents. A mechanism, based on the
properties of and possible interactions among the analyte, the injection
solvent, and the stationary phase of the capillary column, was proposed
to explain these observations.
The effect of initial temperature on peak height and peak area of
the 16 PAHs and the 10 PCBs was also studied. The optimum initial
temperature was found to be dependent on the physical properties of the
solvent used and the amount of the solvent injected. Generally, from the
boiling point of the solvent to 10 0C above its boiling point was an
optimum range of initial temperature at which cthe highest peak height
and peak area were obtained.
