Stressed Haematococcus Pluvialis cells would accumulate astaxanthin with the changes of lipid, pigment, and starch, which recently has been widely studied. Real-time dynamic monitoring of these components will have important guiding significance for screening high-yield and high-quality strains in the industry. Terahertz spectroscopy is not only suitable for measuring large flux samples, but also widely used for quantitative analysis of multicomponent mixtures. To investigate the potential of using terahertz spectroscopy to monitor and quantify metabolites in the algae, terahertz spectrum of Haematococcus Pluvialis under nitrogen starvation were captured and examined against those standard reagents of β-carotene, astaxanthin, and starch. The correlation between intensities of the characteristic peaks and the metabolite contents which were quantified by chemical analysis was established, and the results showed the correlation coefficient were higher than 0.92, indicating a relatively accurate performance to predict the metabolites concentration. Furthermore, to identify meaningful spectra range for quantification, the partial least squares combined with intensities of the interval characteristic peaks were established, where it demonstrated an accurate model with high prediction correlation coefficients (greater than0.98), low root-mean-square error of prediction (RMSEP), and high residual predictive deviation. The results approve the feasibility of using terahertz spectroscopy for synchronous detection on metabolites concentration in the microalgae production process.

压力很大的红血球菌细胞会随着脂质，色素和淀粉的变化而积聚虾青素，最近已对此进行了广泛的研究。这些成分的实时动态监测对于筛选该行业的高产优质菌株具有重要的指导意义。太赫兹光谱不仅适用于测量大流量样品，还广泛用于多组分混合物的定量分析。要研究使用太赫兹光谱法监测和定量藻类血球菌的太赫兹光谱中藻类代谢物的潜力在氮饥饿状态下，将其捕获并针对β-胡萝卜素，虾青素和淀粉的标准试剂进行检查。建立了特征峰强度与化学分析定量代谢物含量之间的相关性，结果表明相关系数高于0.92，表明预测代谢物浓度具有相对准确的性能。此外，为了确定有意义的光谱范围以进行量化，建立了偏最小二乘结合间隔特征峰的强度，从而证明了该模型具有较高的预测相关系数（大于0.98），均方根误差低的精确模型。预测（RMSEP）和高残留预测偏差。