This paper describes the first in-depth attempt to characterize thermally induced aging of polypropylene (PP) samples by near infrared (NIR) spectroscopy. Significant levels of variation in NIR bands associated with short (amorphous-dominated) and long (crystalline-dominated) helices was readily captured when PP samples were subjected to thermal aging treatment. Partial least squares (PLS) regression models derived from the NIR spectra indicated significant level of correlation between the actual and predicted elongations of the samples. Analysis of PLS scores and two-dimensional (2 D) correlation spectra derived from the aged PP samples revealed inner working mechanism of the regression model. Namely, the aging treatment essentially induces compositional change in crystalline and amorphous structures of the PP samples, which eventually affect the variation of the PLS scores. Thus, by utilizing the scores, it becomes possible to predict the change in the elongation property of the aged PP sample.

本文描述了通过近红外（NIR）光谱技术对聚丙烯（PP）样品的热诱导老化进行表征的首次深入尝试。当对PP样品进行热时效处理时，很容易捕获到与短（非晶态为主）和长（晶体为主）螺旋相关的NIR带的显着变化水平。从NIR光谱得出的偏最小二乘（PLS）回归模型表明，样品的实际伸长率和预测伸长率之间具有显着的相关性。对来自老化的PP样品的PLS得分和二维（2 D）相关光谱进行分析，揭示了回归模型的内部工作机制。也就是说，时效处理实质上会引起PP样品的晶体和非晶结构的成分变化，最终影响PLS分数的变化。因此，通过利用得分，可以预测老化的PP样品的伸长率特性的变化。