Monitoring the purity of recovered solvents by NIRS

Importance of the Topic

The recovery and reuse of solvents such as dichloromethane are critical for reducing waste, lowering production costs, and minimizing environmental impact in chemical and pharmaceutical manufacturing. Ensuring the purity of reclaimed solvents before reuse safeguards product quality and process safety.

Objectives and Study Overview

This study demonstrates how near-infrared spectroscopy (NIRS) can rapidly monitor the purity of recovered methylene chloride and quantify two major impurities: water and methanol. Calibration models based on NIR spectra were developed and validated against traditional methods (Karl Fischer titration for water and GC for methanol).

Methodology and Instrumentation

Samples were collected directly from a distillation recovery unit without further preparation. Each sample underwent a spectral scan from 400 to 2500 nm using a fiber-optic immersion probe giving a 4 mm path length. Sixteen co-added sample scans were ratioed to 16 reference scans. Spectral regions above 2200 nm and the non-linear zone near 1700 nm were excluded. Second-derivative pretreatment enhanced spectral features and eliminated baseline drift. Partial Least Squares regression was applied to targeted wavelength ranges for water (1800–2000 nm), methanol (2000–2100 nm), and overall solvent purity (1120–1680 nm and 1720–2200 nm).

Used Instrumentation

• NIRS XDS SmartProbe Analyzer with fiber-optic immersion probe
• Gas chromatography (GC) for methanol reference measurements
• Karl Fischer titration for water content reference

Key Results and Discussion

The moisture calibration over 0–0.2% water achieved a standard error of calibration (SEC) of 0.0156% and a prediction error (SEP) of 0.0133%, matching Karl Fischer accuracy. Methanol calibration in the 2000–2100 nm region yielded SEC 0.0031% and SEP 0.0034%, comparable to GC. The overall solvent purity model across all usable spectral regions showed SEC 0.043% and SEP 0.0455% for a purity range of 99–100%, in line with GC results.

Benefits and Practical Applications

NIRS allows simultaneous multi‐component analysis in under 30 seconds with no sample preparation or specialized operators. The method supports batch-mode laboratory testing using probes, cuvettes, or disposable vials and can be adapted for inline process monitoring during distillation, enabling real-time quality control and process optimization.

Future Trends and Opportunities

Advances may include integration with Process Analytical Technology (PAT) frameworks, deployment of rapid liquid analyzers for high-throughput sampling, and coupling with automated process control systems. Emerging chemometric algorithms and miniaturized sensors will further enhance onsite and inline solvent monitoring.

Conclusion

Near-infrared spectroscopy provides a fast, accurate, and cost-effective approach to assess the purity of recovered solvents and detect trace impurities. Validated against Karl Fischer and GC, NIRS offers robust performance for laboratory and inline applications, supporting sustainable solvent management.

Reference

Metrohm. Near-Infrared Spectroscopy Application Note NIR–X1, Version 1.