mid-IR raw spectra from low molecular weight permeate of ultra filtrated spent sulfite liquor (UF-SSL) including sugar reference measurements

Description

Raw mid-IR spectra including reference measurements of up to 5 sugars. The datasets include 2 case studies with indipendant datasets for calibration, validation and in-line usage (including reference measurements at sample times with respective spectra as well as all spectra over ~250h of continous in-line application). Low Complexity (LC) case study consists of 1 batch af UF-SSL diluted to 25% with ultra pure water with the three most abundant sugars in UF-SSL (mannose, xylose, glucose) spiked. High Complexity (HC) case study consists of 2 batches of UF-SSL, diluted to 25/50/75% with ultra pure water with the 5 most abundant sugars in UF-SSL (LC + galactose, arabinose).

Steps to reproduce

The low molecular weight permeate of ultra filtrated spent sulfite liquor (UF-SSL) from softwood pulping (stored at 4°C) was used for all experiments. The UF-SSL used for this study had a composition of 132-143g/L mannose, 56-65g/L xylose, 41-47g/L glucose, 29-34g/L galactose, 13-18g/L arabinose. The FT-MIR spectra were taken by a Fiber MultiplexIR FT-IR system (ReactIR~45m, Mettler~Tolido, USA) equipped with a liquid N2 MCT detector and an optical fiber immersion probe from silver halide, with 9.5mm optical path length and a DiComp diamond probe tip (ReactIR 45m, Mettler~Toledo, USA) which was connected to the detector with a 1.5m long fibre optic cable. Each spectrum ranged from 3000cm-1 to 650~cm-1 and consisted of an average of 256 scans, with a resolution of 4cm-1. The probe was inserted into a lab-scale stirred tank reactor glass vessel containing the UF-SSL to be measured. The equipment and setup were the same during model building and in-line application while only the spectral acquisition methods differed. During model building, each spectrum was taken manually once the media was well mixed, while during in-line application, a spectrum was acquired every 3min. The instrument and data collection was controlled by iC IR 7.0 software (Mettler Toledo, USA). As a reference, Glucose, Xylose, Arabinose, Galactose and Mannose were measured off-line by HPLC (Ultimate 3000, Thermo Fisher Scientific, USA) equipped with an RI detector (RI100, Shodex, USA) using a Pb-column (NUCLEOGEL SUGAR Pb719530, Machery-Nagel, Germany) at 79°C with an isocratic flow of 0.4ml/min ultra pure water with a runtime of 65min. All samples were diluted 1:20 with ultra pure water and filtered using a 0.22µm filter before analysis. Samples were prepared with constant stirring. 500mL of UF-SSL, diluted with water respectively, was filled in the vessel and acted as the first calibration sample. On top of this an enriched concentration of either Glucose, Xylose, Arabinose, Galactose and Mannose was added to spike the 500mL of diluted UF-SSL with the respective sugar. Spiking was performed to increase sugar concentrations from +1\% to +28\% sugar relative to amount of the respective sugar in 100\% UF-SSL. When multiple batches of UF-SSL were used spikes were performed in alternating fashion between both batches. Validation samples were prepared in a similar same way. To increase variability and minimize correlation between the sugars, the different enriched UF-SSL solutions were added together and in random order. When multiple batches of UF-SSL were used a 50/50 mix of both was used for the validation samples. During the duration of the in-line application, a samples of 2mL were taken by an automated sampling system (Numera, Securecell, Switzerland) from inside the feed tank and stored at 4°C. Samples for off-line analysis by HPLC were taken every 1.5h during gradients or before and after steps and every 3h in phases where no changes were applied in between.

Institutions

Categories

Funding

Licence