Supplementary dataset and selection record for: Selective polyphenol recovery from grape-processing residues using natural deep eutectic solvents: Coupling extraction with downstream separation
Annotatsiya
Supplementary dataset and selection record for the critical review "Selective polyphenol recovery from grape-processing residues using natural deep eutectic solvents: Coupling extraction with downstream separation" (Zaynab A. Odilova, Sergey N. Shlykov; submitted to Separation and Purification Technology, 2026). Contents. Dataset_S1 (Excel workbook with a README sheet): 957 audited within-report response ratios from 15 reports, the 861-row exclusion ledger with reasons, 776 centred panel-enrichment values, 504 profile fractions, 67 legacy ratios and 46 legacy contrasts, 240 experimental arms, 6 cascade cases, 481 registry property records, PRISMA and data-flow counts, and Tables S1-S5 (evidence table of the 67 included reports; eight-domain appraisal with written bases; solvent-property measurement conditions; text-search counts for LCA, TEA, scale-up and solvent reuse; stage balance for the linked pair Frontini 2024/2025 with the 192 verified recovery cells). Also included: the working protocol with its dated amendments and coding rules, the selection record (prisma_cumulative_routes.json), the frozen-analysis run logs with input hashes, the coding maps with the reviewers' confirmation log, the Crossref check of the reference DOIs, the two-pass extraction and the completed personal verification form of the Frontini 2025 recovery tables, and the scripts that regenerate the figures and supplementary tables from the project files (build_figures.py, build_figure6_stage_balance.py, build_supplementary.py; Python 3 with pandas, openpyxl and matplotlib). Interpretation. Rows of the response-ratio tables are dependent contrasts within reports (shared controls, batches and analytical runs), not independent experiments; no pooled effect, weight, confidence interval or significance test is implied. Censored values (nd, <LOD, <LOQ) were never replaced by zero. Ethylene-glycol systems are flagged and are not natural solvents. Experimental values are printed in the sources (page and table given); calculated values are RR, log2 RR, centred enrichment, profile fractions and the stage-balance columns of Table S5 (product yield = extract yield x recovery; enrichments relative to catechin); coded solvent families and matrix types were reconstructed from the reports and checked against the PDFs by both reviewers. Table S5 combines two linked reports that measured the same extracts; values from unrelated reports are never multiplied. Version 1.1 (2026-10-28). Adds the scripts of the revised Fig. 5 (build_figure5_cascade.py, selectivity-guided cascade architecture) and of the graphical abstract (build_graphical_abstract.py); replaces Dataset_S1, Table_S4_text_search_snippets.csv and build_supplementary.py after removing 87 U+FFFE characters (line-break markers carried over from PDF text) from the Table_S4_snippets sheet. No numerical value changed. All other files are identical to version 1.0.
Hali tarjima qilinmagan