Astra 6.1.7.17 rc7

The measurements were carried out according to the scheme shown in the Supporting Information. The sample volume injected onto the channel was 20 μL for an injected sample mass of approximately 20 μg. A 3 min focusing step was performed prior to elution with a focusing flow rate of 1.5 mL/min. Different cross flows in range of 2.5−3.5 mL/min were applied. After elution, the channel was flushed without any cross flow for 2 min before the next analysis. Detector flow rate (V_e) was constant at 1.0 mL/min throughout the separation. Carrier liquids were prepared as 50 mM NaNO₃ (Merck, Darmstadt, Germany) dissolved in pure water prepared with a Milli-Q system (Millipore). Data were collected by ASTRA 6.1.7.17 (RC7) (Wyatt Technology Corporation and Chemstation (rev. B.04.03-SP1 [87]). Lyophilized bovine serum albumin (BSA) (Sigma-Aldrich) was reconstituted by dilution with Milli-Q water to a concentration of 1 mg/mL. A measurement was conducted five times at each condition. In addition, measurements of apoferritin under the same conditions were used as reference for the calibration. For the measurement of polystyrene particles (Thermo Fisher Scientific) a 0.1% w/v of sodium dodecyl sulfate (SDS, p.a. grade, Sigma-Aldrich) was used as eluent. Here, the focus time was 4 min and V_c = 0.5 mL/min.

The collected data were processed via ASTRA 6.1.7.17 (RC7) (Wyatt Technology Corporation) by applying the following steps. The baseline was set with the same data ranges for the quintuplicate analysis for all applied cross flows. Alignment and band-broadening correction was carried out using ASTRA. For the evaluation of the data, a software (“FFFEval”) was written in C++ under the use of the frameworks Qt 5.7 and Qwt 6.1.2¹⁷ for the graphical user interface. Four basic functions are provided by the software: (1) creation of a channel profile and calculation of the channel height ω; (2) calculation of the distributions of D and R_h; (3) calculation of the broadening of the Gaussian kernels and conduction of the deconvolution procedures (iterative relaxation or Richardson−Lucy) using the D-distribution from step 2; (4) partial Zimm plot evaluation of either raw or deconvolved data sets.
A second in-house tool was used for the recognition and separation of the peaks in the acquired fractogram and automatic evaluation of all qualification parameters described below. A detailed description of the applied algorithms is given in the Supporting Information. A current state version of the source code is available on request from the authors.