Implications of dispersion in connecting capillaries for separation systems involving post-column flow splitting

• Published in: Journal of Chromatography A 2021-02, Vol.1639, p.461893, Article 461893
• Main Authors: Gunnarson, Caden, Lauer, Thomas, Willenbring, Harrison, Larson, Eli, Dittmann, Monika, Broeckhoven, Ken, Stoll, Dwight R.
• Format: Article
• Language: English
• ISSN: 0021-9673
• DOI: 10.1016/j.chroma.2021.461893

•Post-column flow splitting is commonly used for coupling fast LC analyses to MS.•Flow splitting can lead to very small volumetric variance peaks in the detector.•A framework is developed to help choose connecting capillaries when using splitting.•At high split ratio the capillary has a large effect even for larger volume columns.•Narrow capillaries are advised for high split ratios to avoid losses in efficiency. It is common practice in liquid chromatography to split the flow of the effluent exiting the analytical column into two or more parts, either to enable parallel detection (e.g., coupling the separation to two destructive detectors such as light scattering and mass spectrometry (MS)), or to accommodate flow rate limitations of a detector (e.g., electrospray ionization mass spectrometry). In these instances the user must make choices about split ratio and dimensions of connecting tubing that is used between the split point and the detector, however these details are frequently not mentioned in the literature, and rarely justified. In our own work we often split the effluent following the second dimension (2D) column in two-dimensional liquid chromatography systems coupled to MS detection, and we have frequently observed post 2D column peak broadening that is larger than we would expect to result from dispersion in the MS ionization source itself. For the present paper we describe a series of experiments aimed at understanding the impact of the split ratio and post-split connecting tubing dimensions on dispersion of peaks exiting an analytical column. We start with the simple idea – based on the principle of conservation of mass – that analyte peaks entering the split point are split into two parts such that the analyte mass (and thus peak volume) entering and exiting the split point is conserved, and directly related to the ratio of flow rates entering and exiting the split point. Measurements of peak width and variance after the split point show that this simple view of the splitting process – along with estimates of additional dispersion in the post-split tubing - is sufficient to predict peak variances at the detector with accuracy that is sufficient to guide experimental work (median error of about 10% over a wide range of conditions). We feel it is most impactful to recognize that flow splitting impacts apparent post-column dispersion not because anything unexpected happens in the splitting process, but because the split dramatically reduces th