Deterministic noise in correlation chromatography; origins and properties of ghost peaks

Linearization of column responses by correlation chromatography. Separation improvement and reduction of undesired adsorption effects

“…The experimental results show that a considerable amount of this correlation noise (including ghost peaks) can be eliminated by subtracting the inverse-bit correlogram from the normal correlogram (Figures 6a3, 6b3). The eliminated part of the correlation noise therefore can be seen as though it comes from non-complementary injection errors [8]. Obviously, for these non-linear processes the theory of origins and properties of ghost peaks, described by Mulder et al [8] is, at least partly, valid.…”

“…The eliminated part of the correlation noise therefore can be seen as though it comes from non-complementary injection errors [8]. Obviously, for these non-linear processes the theory of origins and properties of ghost peaks, described by Mulder et al [8] is, at least partly, valid. This can be illustrated by the three largest ghost peaks in the correlograms of Sample I (Figures 6al-a3).…”

“…This can be illustrated by the three largest ghost peaks in the correlograms of Sample I (Figures 6al-a3). They show up at 54, 82 and 149 s and can be determined respectively as first-order non-complementary (L3), second-order complementary (~7) and second-order non-complementary (L5) ghost peaks, where ;k3, ~,5 and ~7 describe the nature of the ghost peak [8]. The noncomplementary ghost peaks (~.3 and ~.5) have the same sign in both the normal and the inverse-bit correlogram, they are eliminated with subtraction.…”

“…Experiments were performed on a GC, adapted and tested for CC in our laboratory [8]. Figure 3 outlines the configuration used for performing the experiments and Table I lists the equipment.…”

Comparison of sample concentration methods and correlation chromatography for polymer dynamic head-space studies

“…All deviations from perfection generate "ghost" peaks on correlograms, i.e., chromatograms calculated from the detector output. The CC theory can predict the patter0 and intensity of ghost peaks very precisely [8,9]. The absence of commercially available instrumentation with a proper input system and software is, in otff Opinion, the main reason for low usage of CC.…”