细胞生理溶液中不同离子浓度对近红外光谱测温的影响

膜片钳测量过程中实时监测离体细胞生存环境温度，控制生理溶液温度值对提高测量的准确性，消除温度不确定性具有重要的意义。采用近红外光谱结合化学计量学法来研究生理溶液中不同离子不同浓度对温度模型精度的影响。通过配制CaCl₂，KCl和NaCl各四种浓度的12份溶液样本，分别采集不同溶液样本在20~40 ℃温度范围内的光谱，波数范围为9 615~5 714 cm-1，并将每种溶液不同温度光谱数据按照三种方式划分训练集和预测集，采用间隔偏最小二乘方法选择有效波段，并建立与温度数值之间的定量校正模型。实验结果显示，浓度为0.25 g·mL-1的CaCl₂溶液模型的RMSEP最大，三次实验结果为0.386 3，0.303 7和0.337 2 ℃，浓度为0.005 g·mL-1的NaCl溶液模型的RMSEP最小，实验结果分别为0.220 8，0.155 3和0.145 2 ℃。总体实验结果表明细胞生理溶液中Ca2+对建立温度模型的精度影响最大，K+其次，Na+最小，当三种离子浓度均增大时，各离子对模型精度影响均为增大。因此在建立细胞生理溶液的温度模型时，有必要在合理范围内改变细胞生理溶液中三种主要离子的配比，来校正不同离子浓度对测量生理溶液温度的影响，从而提高近红外光谱温度测量的精度。

The Influence of Different Ionic Concentration in Cell Physiological Solution on Temperature Measurement by Near Infrared

It is important to real-timely monitor and control the temperature of cell physiological solution in patch clamp experiments, which can eliminate the uncertainty due to temperature and improve the measurement accuracy. This paper studies the influence of different ions at different concentrations in the physiological solution on precision of a temperature model by using near infrared spectroscopy and chemometrics method. Firstly, we prepared twelve sample solutions respectively with the solutes of CaCl₂, KCl and NaCl at four kinds of concentrations, and collected the spectra of different solutions at the setting temperature range 20~40 ℃, the range of the spectra is 9 615~5 714 cm-1. Then we divided the spectra of each solution at different temperatures into two parts (a training set and a prediction set) by three methods. Interval partial least squares method was used to select an effective wavelength range and develop calibration models between the spectra in the selected range and temperature velues. The experimental results show that RMSEP of CaCl₂ solution with 0.25 g·mL-1 is maximum, the result of the three tests are 0.386 3, 0.303 7 and 0.337 2℃, RMSEP of NaCl with 0.005 g·mL-1 solution is minimum, the result of the three tests are 0.220 8, 0.155 3 and 0.145 2 ℃. The experimental results indicate that Ca2+ has the greatest influence on the accuracy of the temperature model of the cell physiological solution, then K+, and Na+ has the least influence. And with the ionic concentration increasing, the model accuracy decreases. Therefore, when we build the temperature model of cell physiological solution, it is necessary to change the proportion of the three kinds of main ions in cell physiological solution reasonably in order to correct the effects of different ionic concentrations in physiological solution and improve the accuracy of temperature measurements by near infrared spectroscopy.