Monitoring of critical parameters in thermophilic solid-state fermentation process of soybean meal using NIR spectroscopy and chemometrics

Monitoring of critical process parameters was conducted by combining near infrared (NIR) spectroscopy and chemometrics to achieve real-time measurements and adjustments of thermophilic solid-state fermentation of soybean meal (SBM). Fermentation was conducted by Bacillus licheniformis YYC4 under the conditions of unsterilized SBM 20 g, inoculation 10(7) CFU g(-1) wet basis, ratio of substrate to water 1:1.8 (g mL(-1)), MgSO4 0.12%, and 55 degrees C for 60 h in a 150 mL beaker. During fermentation, pH increased from 6.60 to 9.21 (0-50 h), followed by a slight change to 9.09 (50-60 h). Moisture decreased gradually from 66.62 to 58.01%. Soluble protein decreased slightly from 4.92 to 4.48% (0-2 h) before increasing significantly to 16.26% (30 h) and 18.57% (50 h). Then, it decreased to 17.19% at the end of fermentation (60 h). Trypsin inhibitor (TI) activity remained almost no change within 0-6 h before decreasing from 8.19 to 3.19 mg g(-1) (50 h). After that, a further decrease to 2.15 mg g(-1) (60 h) was observed. Based on offline analytics, synergy interval partial least squares (siPLS) models were established to monitor these variables after spectral pretreatment. Root mean squared errors (RMSEP) and coefficient of determination (R-P) of prediction could achieve 0.169 and 0.9781, 0.313% and 0.9909, 0.681% and 0.9883, 0.236 mg g(-1) and 0.9916 for pH, moisture, soluble protein and TI contents respectively, with an acceptable accuracy. The satisfactory prediction model underpins the potential of NIR spectroscopy in bioprocess monitoring applications.

Automated summary

This study combines near infrared spectroscopy and chemometrics to monitor critical process parameters in the thermophilic solid-state fermentation of soybean meal. The established siPLS models can accurately predict pH, moisture, soluble protein, and trypsin inhibitor content, demonstrating the potential of NIR spectroscopy in bioprocess monitoring applications.