基于B样条的拉曼光谱基线校正方法

基线漂移是光谱检测仪器普遍存在的现象,会对光谱信号的特征提取带来十分不利的影响,而基线校正是解决该问题的重要手段,也是拉曼光谱信号预处理的重要组成部分。基线校正的原理一般是通过拟合基线的方法来去除光谱信号中的基线漂移。传统的基线校正方法是利用多项式拟合的方法对拉曼光谱信号的基线进行拟合的,但是该方法容易出现过拟合和欠拟合现象,且拟合阶数难以确定。针对传统方法的缺点进行了改进,利用B样条方法对拉曼光谱信号的基线进行拟合,发挥B样条低阶光滑的优点,能够有效地克服多项式方法的缺陷。在实验中,利用该方法对孔雀石绿、罗丹明B的拉曼光谱信号进行了基线校正,观察并比较该方法和传统方法的校正结果。实验结果表明,该方法能够有效地消除拉曼光谱信号的基线漂移,在基线漂移较小和较大的位置,可以采用相同的拟合阶数,不会出现欠拟合和过拟合的现象,获得了良好的基线校正效果,为进一步分析光谱数据提供准确可靠的信息。     

分段式线性拟合校正拉曼光谱基线漂移

基线漂移是目前光谱仪收集拉曼光谱时难以避免的现象,基线校正是光谱数据处理中一个不可或缺的重要步骤。针对传统的基于多项式拟合的基线校正方法存在的不足,本工作采用线性拟合的方法分段拟合背景基线,与计算机相结合,实现快速、准确、自动的基线校正;经聚苯乙烯小球、红细胞和酿酒酵母等多批次拉曼光谱数据处理的验证,表明该改进方法可以有效地对拉曼光谱进行基线校正,为进一步分析光谱数据提供更准确的信息,是一种可行的基线校正方法。     

基于自动线性拟合的快速拉曼基线校正算法

近年来拉曼光谱以其无创、灵敏度高等众多优点在化学表征、生物医药、材料等领域引起广泛关注,而基线漂移的存在为后续的定性定量分析带来严重困扰,因此设计高性能的基线校准算法以提高分析结果的有效性及准确性具有重要意义。针对传统算法在批量拉曼光谱数据基线校正方面的不足,基于自动线性拟合算法提出一种快速基线校正算法以校正具有相似背景的批量拉曼光谱数据并详细阐述了该算法的核心思想以及算法实现流程。该算法首先从批量拉曼光谱数据中自动选择一条拉曼光谱数据作为基准光谱,使用自动线性拟合算法对其进行基线校准得到其基线以及分段标记点,然后利用标记点快速计算出组内其他与基准光谱具有较高相关性的拉曼光谱数据的基线,对于组内与基准光谱相关性不满足阈值要求的拉曼光谱则使用自动线性拟合算法对其进行单独基线校正,这使得算法具有具有较强的鲁棒性,可以适应复杂的拉曼光谱基线校正情形。分别使用快速基线校正算法与单独基线校正算法对多组实际拉曼光谱数据进行基线校正以对比分析算法基线校正效果,结果表明该算法可以实现对批量拉曼光谱数据的快速校正,基线校正效果良好,并且相较于单独进行基线校正算法耗时减少了30%以上,算法无参,简单易行,无需额外人工干预,是一种切实可行的批量拉曼数据自动基线校正算法。     

化学计量学结合拉曼光谱在生物材料检测中的应用

综述化学计量学结合拉曼光谱在生物材料检测中的应用。报道利用基线校正方法(导数、曲线拟合、小波变换)、归一化方法(矢量归一化、峰高归一化)、复合预处理等方法,可以提高信噪比,恢复失真的信号,实现谱图峰位的正确识别;基于主成分分析、偏最小二乘、聚类分析、目标因子分析、人工神经网络、支持向量机等算法构建化学计量模型,利用模型对拉曼光谱定性或定量分析,分别得到可信的结果;提出未来化学计量建模的方向。     

基于卷积神经网络特征提取的拉曼光谱分类研究

拉曼光谱物质定性识别已被广泛的应用于化工、安防、缉毒等行业和研究领域,但是传统的拉曼光谱分析技术依赖于光谱数据库,通过光谱特征提取进行识别。特征提取是拉曼识别的关键处理步骤,通常利用主成分分析,因子分析等方法进行特征提取,而后通过KNN,SVM和随机森林等方法进行光谱特征定性识别,当拉曼数据库不存在待定性物质时,易造成待检测物质的错误分类。针对此问题,提出一种基于卷积神经网络的对数据库缺少物质光谱识别方法。在实验过程中,采用九类,200余种精神类药品拉曼光谱作为测试对象,通过搭建卷积神经网络自动特征提取并利用Softmax分类器将200余种物质,按照Amphetamine, cathinone, cannabinoids等九种类别进行定性分析。通过与传统机器学习方法如K近邻,支持向量机等方法进行比较,基于卷积神经网络的模型识别准确性有显著提高,该方法可为拉曼光谱数据库的光谱识别检索提供一种新的识别方法。     

基于PCA和LS-SVM的傅里叶变换近红外光谱的黄酒酒龄的鉴别模型

利用傅里叶变换近红外光谱技术,对黄酒酒龄鉴别的模型进行研究。对绍兴黄酒样本光谱主成分进行提取与分析,并发现前3个主成分具有较明显的聚类特征。其次,利用LS-SVM模型对不同主成分数进行分类和寻优,发现当主成分数为6时达到判别的正确率是100%,此时模型的两个参数γ和2σ分别为61.890和1.769。研究表明,利用傅里叶变换近红外光谱技术并结合主成分分析(PCA)和最小二乘法支持向量机(LS-SVM)可作为一种可靠、准确、快速的检测方法用于黄酒酒龄定性鉴别分析。     

小波包熵和Fisher判别在近红外光谱法鉴别中药大黄真伪中的应用

利用傅里叶变换近红外光谱仪采集了中药大黄的近红外漫反射光谱,提取光谱的主成分和小波包熵等特征信息,再以特征信息为依据,利用Fisher分类器对中药大黄的真伪进行了鉴别。通过比较得出：采用小波包熵特征信息建模和预测误判率比采用主成分低。用小波包熵进行特征提取和Fisher分类器相结合对中药大黄真伪进行鉴别,其建模集交叉验证的误判率为6.52%,预测集的误判率是2.04%,为中药大黄的近红外快速真伪鉴别提供了参考。     

基于自适应加窗spline曲线拟合的拉曼光谱去基线方法

拉曼光谱是一种无损快速检测技术,可以提供材料的定性和定量信息,因而在医药、化工等诸多领域得到了广泛的应用。但是,由于样品荧光背景噪声的影响,造成拉曼光谱信号出现基线漂移现象,这给拉曼光谱的特征峰识别和拉曼成像带来十分严重的影响。目前,改进实验方法和数值处理是解决该问题的两种重要手段。改进实验方法上,有偏振调制法和高频调制法等,但存在实验设备复杂,检测技术难度大等缺点;数值处理上,有多项式拟合和小波变换等,但容易出现欠拟合和过拟合等现象。本文在不改换高精密设备的前提下,针对传统基线校正的方法进行了改进,提出一种基于自适应加窗spline曲线拟合的拉曼光谱去基线方法。首先,基于谱峰识别算法和初始搜索步长求得谷值的最优搜索间距,并利用谱谷识别算法完成谷值曲线的拟合;其次,利用最优搜索间距和谱峰识别算法,求得谷值曲线峰值位置,并在该位置处对称添加自适应矩形窗函数去除峰值,重新划分整个区间,拟合谷值曲线;再次,逐点比较拟合曲线与原拉曼光谱信号,取较小值,拟合曲线;最后,重复加窗去除峰值操作,直至自适应窗函数宽度低于阈值,完成拉曼光谱信号的基线拟合。在实验中,选用乙酸丁酯、聚甲基丙烯酸甲酯(polymethyl methacrylate, PMMA)作为实验样品,利用该方法对其拉曼光谱信号进行了基线校正,观察并比较该方法和传统方法的校正结果。实验结果表明,该方法能够有效地消除拉曼光谱信号的基线漂移,较好的保留一些较弱的拉曼特征峰,且不易出现欠拟合和过拟合的现象,获得了良好的基线校正效果,为进一步分析光谱数据和实现拉曼成像提供准确可靠的信息。     

TLC-SERS法快速检测降压类中药中非法添加的四种化学成分

建立了薄层色谱(TLC)与表面增强拉曼光谱(SERS)联用方法对降压类中药中四种非法添加的化学成分,即盐酸尼卡地平、甲磺酸多沙唑嗪、盐酸普萘洛尔及氢氯噻嗪进行快速检测的新方法。利用TLC法将掺杂成分与中药基质进行简单分离,采用SERS法对薄层板上的微量物质进行定性检测。实验中通过优化银胶浓度和薄层展开体系,并考察掺杂成分的最低检测限,初步建立可用于降压类中药非法添加化学成分的TLC-SERS检测方法。结果表明,该方法用于检测这四种化学成分时专属性较好,检测限低可至0.005 μg,最终检测出十种降压类中药中有两种非法添加了化学成分。本方法简便快速,能实现降压类中药中非法添加化学成分的快速检测,有望成为中药掺杂化学成分的现场检测的新方法。     

FTIR和Raman用于盐酸曲马多的结构分析

采用AVATAR 360型傅里叶变换红外光谱仪和RM-1000型激光共聚焦拉曼光谱仪测定盐酸曲马多的红外光谱和拉曼光谱。拉曼光谱和红外光谱中均显示出了盐酸曲马多的特征峰,都能够用于其结构鉴别。采用红外、拉曼光谱这两种方法互相印证,互相补充,可增强鉴定的准确性、可靠性。     

半胱氨酸作为耦联剂制备的纳米银膜上的SERS研究

在直流10V电压下电解聚乙烯醇和硝酸银的混合液3h制备纳米银胶,将经过半胱氨酸修饰后的载玻片浸入银胶24h制得纳米银膜。用紫外可见分光光度计对银胶进行了观测,由于其吸收峰半高宽较窄可知银胶中纳米银颗粒粒径分布较为均匀。同时,使用扫描电镜对银膜进行了表征。通过对半胱氨酸分子SERS信号的分析得出了纳米银粒子在玻璃表面上可能的组装方式。以结晶紫(Crystal Violet)和孔雀石绿(Malachite Green)作为探测分子,采用便携式拉曼光谱仪测得两种分子不同浓度下的SERS谱。发现该方法所制得的纳米银膜有很好的表面增强效果。最后分析了半胱氨酸分子SERS信号对探针分子光谱的影响。     

孔雀石绿在自组装纳米银膜上的SERS

电解法制备了自组装纳米银膜。用紫外可见分光光度计和透射电镜对电解出的银胶进行了观测,并且用扫描电镜对制备好的银膜进行了表征,发现银膜上的纳米银颗粒是银胶中颗粒的二聚体和三聚体。我们得到了孔雀石绿的SERS谱,发现自组装纳米银膜对于用SERS检测孔雀石绿具有很好的增强效果,检测的下限是10-10M。另外,我们对所测不同浓度的孔雀石绿的指纹峰面积进行了分析,发现峰面积随浓度的变化有很好的线性关系,这对以后进行定量分析打下了基础。本实验说明用电解法制备的自组装纳米银膜对于检测孔雀石绿具有很好的应用前景。     

FT-Surface-Enhanced Raman Scattering of Phenylalanine Using Silver-Coated Glass Fiber Filters

A simple and facile method has been developed to produce surface-enhanced Raman scattering (SERS) active surfaces. Tollen's reagent was used to coat silver onto the surface of glass fiber filters. Using a Fourier transform (FT) Raman instrument, strong surface-enhanced Raman signals have been observed for L-phenylalanine on these surfaces. The use of an FT-Raman instrument combined with this new SERS method allows for a method with both high sensitivity and selectivity. A linear dependence of the Raman signal on L-phenylalanine concentration (0.01–1.0 mM) has been demonstrated. The SERS technique presented here is both novel and promising for biochemical analysis.     

Surface‐enhanced Raman scattering of protoberberine alkaloids

Quaternary protoberberine alkaloids are a class of natural dyes characterized by bright colors ranging from yellow to orange. As they present a strong fluorescence emission, their analysis by Raman spectroscopy is limited to specific techniques such as Fourier transform (FT)‐Raman and spectral shift Raman techniques such as shifted subtracted Raman difference spectroscopy (SSRDS) and shifted excitation Raman difference spectroscopy (SERDS). In a previous article, we successfully used surface‐enhanced Raman scattering (SERS) in the analysis of the alkaloid dye berberine in an ancient textile. The examination of the Raman and SERS spectra of berberine in combination with density functional theory (DFT) calculations indicated a flat adsorption geometry of the molecule on the Ag surface. In this article we extend that work to the study of related protoberberine alkaloids, palmatine, jatrorrhizine, and coptisine. The same adsorption geometry as in berberine was deduced. We found that the four alkaloids, although minimally different in their chemical structures, could be differentiated by the position of marker bands. Those bands are the most enhanced ones in the SERS spectra, which appear in the 700–800 cm⁻¹ region. Copyright © 2008 John Wiley & Sons, Ltd.     

Label‐free optical detection of type II diabetes based on surface‐enhanced Raman spectroscopy and multivariate analysis

Surface‐enhanced Raman scattering (SERS) spectroscopy was first employed to detect oxyhemoglobin (OxyHb, the common type of hemoglobin) variation in type II diabetic development without using exogenous reagents. Using silver nanoparticles as SERS‐active substrate, high‐quality SERS spectra are obtained from blood OxyHb samples of 49 diabetic patients and 40 healthy volunteers. Tentative assignment of the observed SERS bands indicates specific structural changes of OxyHb molecule in diabetes, including heme transformation and globin variation. Furthermore, partial least squares and principal component analysis combined with linear discriminate analysis diagnostic algorithms are employed to analyze and classify the SERS spectra acquired from diabetic and healthy OxyHb, yielding the diagnostic accuracies of 90.0% and 95.5%, respectively. This exploratory work suggests that the silver nanoparticles‐based OxyHb SERS method in combination with multivariate statistical analysis has great potential for the label‐free detection of type II diabetes. Copyright © 2014 John Wiley & Sons, Ltd.     

拉曼光谱法鉴定文物及艺术品中染料的研究进展

中国是丝绸之路的发源地,染色历史源远流长。文物及艺术品中的有机染料蕴含着制作年代、原料来源、染色工艺及保存状况等丰富的信息,但因染料成分复杂、含量低、易于降解,其鉴定一直是文物分析领域的难点。拉曼光谱作为一种指纹光谱在文物及艺术品中染料分析方面具有独特的优势。系统阐述了显微拉曼光谱法(MRS/μ-Raman)、近红外傅里叶变换拉曼光谱法(NIR-FT-Raman)、表面增强拉曼光谱法(SERS)、共振拉曼光谱(RRS)法的原理、特点、局限性以及在染料鉴定中的研究进展和发展方向; 分类归纳总结出了黄栀子、姜黄等染料的拉曼光谱特性; 对比分析了紫茜素在采用不同波长激光器激发下的荧光现象; 介绍了以自制的灰绿色银胶颗粒为基底,联合薄层色谱(TLC) 分离技术及SERS鉴定茜草中的色素,表明银溶胶微粒产生的表面增强效应可显著降低拉曼光谱荧光背底。指出拉曼光谱作为一种快速、便捷的结构定性测试手段在文物和艺术品染料分析中应用前景广阔,提出多种拉曼光谱技术相结合、拉曼光谱技术与分离技术相结合以及拉曼光谱技术与远距离传导技术相结合是拉曼光谱技术的发展趋势。     

Fourier‐transform surface‐enhanced Raman spectroscopy (FT‐SERS) applied to the identification of natural dyes in textile fibers: an extractionless approach to the analysis

In the present study, an application of a silver colloid substrate in order to obtain Fourier‐transform surface‐enhanced Raman (FT‐SER) spectra of natural historical dyes is presented. In detail, we collected a spectral database from solutions of pure dyes and then we carried out extractionless both hydrolysis and non‐hydrolysis FT‐SERS analyses on wool fibers previously dyed in our laboratory and on ancient textiles. The term ‘extractionless’ refers to a method of SERS analysis applied directly on the fiber, thus avoiding the extraction of dyes from textile samples. The combination of a low‐energy source of radiation, as in the FT‐Raman technique, with SER spectroscopy can bring the important advantage of reducing the fluorescence typical of ancient samples and organic dyes. In some historical textile samples, for which SER spectra by use of visible excitation could not be obtained, the FT‐SER spectrum of an iron‐gall dye was recorded without hydrolysis, while, with an HF hydrolysis pre‐treatment on ancient fibers, madder, lac dye and brazilwood were clearly recognized. Copyright © 2014 John Wiley & Sons, Ltd.     

银纳米粒子的SERS效应和SEF效应在微流控芯片光学检测中的应用

银纳米离子的SERS技术和SEF技术的信号检测灵敏度非常高,可以用在微流控芯片的定量分析中。为了提高微流控芯片光学检测技术的检测精度,提出一种在微通道中添加银纳米粒子来增强SYBR GreenⅠ拉曼和荧光信号的方法,并对该方法的原理和增强效果进行了研究。首先,利用准分子激光器在PMMA基板上直写刻蚀出宽200 μm、深68 μm的微通道,接着将制备的银前体溶液加入微通道,通过加热制备出表面增强拉曼（SERS）和表面增强荧光（SEF）基板,接下来对添加银纳米粒子前后的拉曼和荧光信号分别进行对比,进一步研究了微通道中不同浓度银纳米粒子对SYBR GREEN I的拉曼和荧光信号增强效果。添加银纳米粒子后,表面增强拉曼（SERS）实验的增强因子为3.5×10³,添加银纳米粒子的样品的荧光信号强度与不含银纳米粒子样品的荧光信号强度相比,约增加了1倍。结果表明,在微通道中检测SYBR Green I时通过增加银纳米粒子显著地增强了拉曼和荧光信号,这种方法可以用在以SYBR GreenⅠ做染料的微流控芯片检测技术中。     

Surface‐enhanced Raman scattering (SERS) and complementary techniques applied for the investigation of an Italian cultural heritage canvas

A cultural heritage canvas from the early 19^th century, painted by the Vaccaro brothers for the church of Niscemi, province of Caltanissetta, Sicily, was analyzed using Fourier transform (FT)‐Raman, attenuated total reflectance‐FT‐infrared and surface enhanced Raman scattering (SERS) spectroscopy. The painting, still used in religious rites related to the Easter mass (‘la calata da tila’), depicts the scene of the Crucifixion and is executed in a scarce palette, with white, green and blue colors. Analysing vibrational data in conjunction with scanning electron microscopy and solid ‐state ¹³C‐NMR signals of the linen threads, we were able to offer valuable insight into the painting technique, unknown prior to this study. SERS is usually employed in artwork diagnosis for the identification of organic lakes and dyes. Due to its sensitivity, SERS has been successfully applied for the detection of either organic painting materials (indigo) that are usually not resolved by conventional Raman spectroscopy or of inorganic pigments difficult to observe in the presence of highly fluorescent aged organic supports or binders. To the best of our knowledge, this is also the first report on the SERS investigation of flax used in linen from cultural heritage objects using Ag colloidal nanoparticles. Copyright © 2012 John Wiley & Sons, Ltd.     

SERS detection of red organic dyes in Ag‐agar gel

Micro‐Raman spectroscopy has been widely employed in the last few years for the study of artworks, allowing for the characterization of a high class of pictorial materials. However, the detection of organic dyes by conventional Raman spectroscopy is quite difficult, due to the high fluorescence provided by these compounds. Recently, remarkable improvements have been achieved by the introduction of the surface enhanced Raman spectroscopy (SERS) technique for the analysis of organic dyes. In the present work, a new method is presented, based on the use of a SERS probe made of agar‐agar coupled with silver nanoparticles, for a non‐destructive and minimally invasive micro‐extraction of dyes from textiles. Ag‐agar gel has been tested first on textile mock‐ups dyed with alizarin, purpurin and carminic acid. SERS measurements have been performed adopting laser light excitations at 514.5 and 785 nm of a micro‐Raman setup. Highly structured SERS band intensities have been obtained. After having verified the safety of the method by colorimetric, X‐ray fluorescence and attenuated total reflectance Fourier transform infrared techniques, a real case, a pre‐Columbian piece of textile, have been investigated by Ag‐agar gel. This cutting‐edge method offers new possibilities for a sensitive and non‐destructive analysis of fluorescent materials. Copyright © 2012 John Wiley & Sons, Ltd.