楊增玲，杜書榮，梅佳琪，李駿寶，劉二偉，韓魯佳

?

FTIR顯微成像表征堿處理后玉米秸稈木質(zhì)素含量及分布

楊增玲，杜書榮，梅佳琪，李駿寶，劉二偉，韓魯佳※

（中國農(nóng)業(yè)大學(xué)工學(xué)院，北京 100083）

探究木質(zhì)素原位表征方法，能為研究農(nóng)作物秸稈去木質(zhì)化過程中木質(zhì)素變化規(guī)律提供幫助。該研究以玉米秸稈為原料，堿處理玉米秸稈橫切片后進(jìn)行傅里葉變換紅外（Fourier transform infrared，F(xiàn)TIR）顯微成像，并結(jié)合快速非負(fù)最小二乘（fast non-negativity-constrained least squares，fast NNLS）擬合算法計(jì)算組織中木質(zhì)素的分布。結(jié)果顯示：1）FTIR顯微圖像結(jié)合fast NNLS算法可對(duì)組織中的木質(zhì)素進(jìn)行定位及定量分析。如擬合計(jì)算得到，原樣組織薄壁細(xì)胞中木質(zhì)素質(zhì)量分?jǐn)?shù)為7.7%，堿處理5、30、60 min后下降至6.0%、4.8%、3.5%；2）基于fast NNLS擬合計(jì)算的組織中木質(zhì)素變化趨勢(shì)同試驗(yàn)測(cè)定的木質(zhì)素變化趨勢(shì)一致。研究結(jié)果表明，基于FTIR顯微成像的表征方法可用于原位分析堿處理過程中秸稈組織的木質(zhì)素分布及含量變化，實(shí)現(xiàn)介觀尺度研究秸稈木質(zhì)素降解規(guī)律。

秸稈；木質(zhì)素；圖像處理；FTIR顯微成像；堿處理；

0 引 言

木質(zhì)素是一種非線性芳香族聚合物，是由羥基或甲氧基取代的苯丙烷單體無序聚合而成的三維復(fù)雜結(jié)構(gòu)，儲(chǔ)量僅次于纖維素。木質(zhì)素是植物細(xì)胞壁的主要組成成分之一，為植物提供足夠的強(qiáng)度和硬度，同時(shí)避免生物侵害和水的侵蝕。農(nóng)作物秸稈等生物質(zhì)資源的木質(zhì)素含量及分布直接影響其轉(zhuǎn)化利用效率。因此，探究木質(zhì)素原位表征方法，對(duì)研究農(nóng)作物秸稈去木質(zhì)化過程中木質(zhì)素的分布及變化規(guī)律有重要意義[1-5]。

近年來對(duì)植物組織的木質(zhì)化過程和去木質(zhì)化反應(yīng)的研究受到學(xué)者們廣泛關(guān)注，目前已有大量研究采用傳統(tǒng)的濕化學(xué)分析法對(duì)反應(yīng)后木質(zhì)素的去除進(jìn)行了測(cè)定[6-8]。在木質(zhì)素定位研究中，使用的方法主要包括組織染色法[9-10]、熒光顯微技術(shù)[11]和共聚焦拉曼顯微技術(shù)[12]等。也有學(xué)者利用光譜[5,13-14]、電鏡結(jié)合能譜[15-16]等技術(shù)對(duì)木質(zhì)素進(jìn)行半定量分析。隨著現(xiàn)代分析儀器的發(fā)展，已經(jīng)有學(xué)者探索開發(fā)了植物不同部位組分分布原位表征手段[17-18]，傅里葉變換紅外（FTIR）顯微成像技術(shù)[19-20]便是其中之一。

FTIR顯微成像技術(shù)將基于紅外光譜的秸稈組分原位分析與基于顯微鏡技術(shù)的直觀成像相結(jié)合，不僅能提供具有豐富空間和化學(xué)信息的圖像，而且還能提供精確到微米級(jí)微區(qū)的具有指紋特性的紅外光譜信息。Dokken等[21]使用FTIR顯微成像研究了向日葵和玉米根不同部位脂質(zhì)、蛋白質(zhì)、纖維素、木質(zhì)素和碳水化合物的空間分布信息。Gorzsás等[22]以歐洲山楊樹和擬南芥木質(zhì)部為研究對(duì)象，利用FTIR顯微紅外結(jié)合隱變量正交投影判別法對(duì)植物器官進(jìn)行了化學(xué)表征。Yang等[23]使用FTIR顯微成像結(jié)合快速非負(fù)最小二乘（fast NNLS）算法探究了表征小麥不同生長期秸稈組分分布及變化規(guī)律的可行性。以上研究均取得了較理想的效果。但尚未檢索到對(duì)預(yù)處理過程中秸稈不同組織的木質(zhì)素變化進(jìn)行表征的研究。秸稈組分復(fù)雜且各組織具有不同的生物特性，使得木質(zhì)素在不同組織中的含量不均勻，預(yù)處理過程中各個(gè)組織的木質(zhì)素降解率可能也會(huì)存在差異，但傳統(tǒng)濕化學(xué)分析方法僅能測(cè)定木質(zhì)素的平均去除率，如文獻(xiàn)[24]指出，堿處理可以去除85%的木質(zhì)素，是一種有效去除木質(zhì)素的化學(xué)預(yù)處理方式。研究預(yù)處理過程中表皮、維管束及薄壁細(xì)胞等不同組織木質(zhì)素去除過程的差異可以為秸稈轉(zhuǎn)化利用和預(yù)處理作用機(jī)理探究提供幫助。本文以堿處理為例，擬探究堿處理過程中秸稈不同組織中木質(zhì)素變化的表征方法。

本研究以玉米秸稈為對(duì)象，探索FTIR顯微成像技術(shù)表征玉米秸稈堿處理過程中不同組織木質(zhì)素分布和含量變化的可行性。

1 材料與方法

1.1 樣品采集與制備

本試驗(yàn)中使用的玉米秸稈為拔節(jié)期新鮮玉米秸稈，于2018年7月底取自中國農(nóng)業(yè)大學(xué)上莊試驗(yàn)站。在每根新鮮秸稈地面以上第6節(jié)間中部位置連續(xù)取厚度約為0.5 cm的2個(gè)莖段，共取4個(gè)莖段置于福爾馬林-乙酸-乙醇（formalin-aceto-alcohol，F(xiàn)AA）固定劑中固定24 h以上。其中1個(gè)莖段直接采用石蠟包埋[25]作為堿處理前的對(duì)照組樣品；另外3個(gè)使用去離子水清洗后于100 ℃預(yù)熱，然后加入約3倍樣品體積已預(yù)熱的2% NaOH溶液，在100 ℃反應(yīng)5、30、60 min時(shí)取出對(duì)應(yīng)樣品，用去離子水清洗，最后進(jìn)行石蠟包埋。包埋后的樣品用切片機(jī)（ESM-150S，ERMA，日本）切取厚度約為18m的切片，置于ZnS玻片上，用于FTIR顯微成像分析。取整株新鮮玉米秸稈地面以上至穗以下部分，切成1～2 mm薄片，混合均勻并同樣固定24 h以上。清洗干燥后分成4份，取1份原樣作為對(duì)照組，將另3份進(jìn)行與莖段一樣的堿處理。然后按照石蠟包埋中脫水透明過程對(duì)4份樣品分別進(jìn)行脫水透明和逆脫水透明，清洗干燥后粉碎過0.069 mm篩，按照NREL/TP-510-42618[26]標(biāo)準(zhǔn)測(cè)定木質(zhì)纖維成分。

此外，為后續(xù)數(shù)據(jù)分析，需獲得玉米秸稈中纖維素、半纖維素和木質(zhì)素的純組分紅外光譜作為參考光譜。將新鮮玉米秸稈自然風(fēng)干后粉碎并過0.069 mm篩，取篩下樣品按照參考文獻(xiàn)[27-28]中的方法，提取玉米秸稈中的纖維素和磨木木質(zhì)素。半纖維素主要成分為木聚糖，以購買于Sigma-Aldrich（美國）的木聚糖代替半纖維素獲取純組分光譜。

1.2 數(shù)據(jù)獲取

采用傅里葉變換紅外顯微成像系統(tǒng)（Spotlight 400，PerkinElmer，英國）采集置于ZnS玻片上的樣品的FTIR顯微圖像。該系統(tǒng)采用液氮冷卻的碲鎘汞（MCT）陣列檢測(cè)器，圖像采集參數(shù)為：圖像空間分辨率6.25m× 6.25m、波長范圍4 000～750 cm–1、光譜分辨率4 cm–1、掃描次數(shù)8。

提取的玉米秸稈纖維素、磨木木質(zhì)素及木聚糖采用傅里葉變換紅外光譜儀（Spectrum 400，PerkinElmer，英國）在透射模式下獲取光譜。將提取的纖維素、木質(zhì)素和木聚糖按質(zhì)量比1：100分別與KBr混合，取大約1 mg制作KBr壓片。紅外光譜獲取參數(shù)為：光譜范圍4 000～750 cm–1、光譜分辨率4 cm–1、掃描次數(shù)32。

1.3 數(shù)據(jù)分析與處理

1.3.1 數(shù)據(jù)預(yù)處理

對(duì)FTIR顯微圖像，首先利用局部閾值法進(jìn)行二值化，將前景（樣品圖像）信息賦值為1，背景信息賦值為0，從而消除背景對(duì)后續(xù)數(shù)據(jù)處理的影響；然后，選取1 800～800 cm–1光譜范圍的矩陣數(shù)據(jù)進(jìn)行分析，并對(duì)選取的光譜數(shù)據(jù)進(jìn)行Savizky?Golay卷積5點(diǎn)平滑、標(biāo)準(zhǔn)正態(tài)變換（SNV）和Whittaker自動(dòng)濾波基線校正等光譜預(yù)處理，從而消除光譜的基線漂移現(xiàn)象。

相同的，對(duì)玉米秸稈纖維素、木質(zhì)素及木聚糖的紅外光譜，也是在1 800～800 cm–1光譜范圍內(nèi)進(jìn)行與FTIR顯微圖像光譜相同的光譜預(yù)處理。

1.3.2 快速非負(fù)最小二乘擬合

考慮到各組分光譜有重疊，通過單一特征峰峰高或峰面積成像很可能受到其他組分的干擾，因此本研究采用快速非負(fù)最小二乘（fast NNLS）擬合算法，獲取木質(zhì)素的分布信息。

NNLS算法是一種基于朗伯-比爾定律的多變量計(jì)算方法。FTIR原始圖像數(shù)據(jù)可以視為的立方體，其中和為空間坐標(biāo)，為波長，如圖1所示。其中每個(gè)像素點(diǎn)的光譜被記錄為一個(gè)行向量，依像素點(diǎn)順序展開后的FTIR圖像數(shù)據(jù)被記作：

式中為原始圖像數(shù)據(jù)矩陣，是組分個(gè)數(shù)，為系數(shù)矩陣，S是階純組分光譜矩陣，是誤差矩陣。

注：和為空間坐標(biāo)，為波長。

Note:andare spatial positions,is wavelength.

圖1 FTIR數(shù)據(jù)結(jié)構(gòu)示意圖[23]

Fig.1 Data structure of FTIR microspectroscopic image[23]

在NNLS算法基礎(chǔ)上，Bro等[29]于1997年提出了基于三線性平行因子分解模型（three-way PARAFAC model）的快速非負(fù)最小二乘算法（fast NNLS），以解決NNLS因計(jì)算量大故而耗時(shí)的缺陷。其算法公式如下：

式中x表示矩陣的第行，第列，第層元素。載荷矩陣的第行，第列記為a。同理，第2和第3載荷矩陣、的元素被記作b和c。e是矩陣第個(gè)元素的殘差。在上式分解中，須令待分析組分個(gè)數(shù)等于。計(jì)算中，載荷矩陣、和的計(jì)算參照條件最小二乘問題，迭代而得。

目標(biāo)組分在每個(gè)像素點(diǎn)的濃度占比a%由公式（3）計(jì)算得到

為使堿處理不同時(shí)間后的樣品間具有可比性，對(duì)a%進(jìn)行校正。是樣品纖維素、半纖維素和木質(zhì)素3種組分的化學(xué)校正值總和（等于試驗(yàn)測(cè)定的樣品3種組分總和與對(duì)應(yīng)堿處理固體回收率的乘積）。按公式（4）計(jì)算

1.3.3 組織提取

為表征堿處理前后各組織中木質(zhì)素的變化，本研究結(jié)合可見光圖像及其RGB/HSV/Lab多空間分量與光譜預(yù)處理后的顯微紅外平均吸光度圖像，利用均值聚類算法提取表皮、維管束和薄壁細(xì)胞等組織。

1.3.4 數(shù)據(jù)處理軟件

2 結(jié)果與討論

2.1 樣品可見光圖像和FTIR顯微圖像

由傅里葉變換紅外（FTIR）顯微成像系統(tǒng)獲取的堿處理前后樣品的可見光圖像和FTIR平均吸光度顯微圖像（原始圖像）如表1所示。為防止掃樣位置影響試驗(yàn)結(jié)果，掃樣前選取位置相似的切片，并在切片上選取相近位置和相同面積（表皮1 325m×525m、維管束2 475m×975m、薄壁細(xì)胞1 525m×775m）的區(qū)域采集紅外顯微圖像。受樣品中雜質(zhì)的影響，樣品顯微圖像中存在異常點(diǎn)，當(dāng)有些異常點(diǎn)紅外響應(yīng)值較大時(shí)，局部閾值法提取的前景中會(huì)包括異常點(diǎn)，需手動(dòng)去除非樣品部分或樣品因受污染而出現(xiàn)的異常點(diǎn)。選取的掃描區(qū)域中樣品結(jié)構(gòu)應(yīng)清晰可辨，以保證提取前景時(shí)完整保留所有結(jié)構(gòu)信息。

表1 樣品可見光圖像及FTIR平均吸光度顯微圖像

Table 1 Visible images and FTIR average absorbance microspectroscopic images

注：樣品可見光圖像從左到右依次為含表皮區(qū)域、維管束區(qū)域和薄壁細(xì)胞區(qū)域，平均吸光度顯微圖像與可見光圖像對(duì)應(yīng)；由上至下依次為玉米秸稈原樣、堿處理5 min、堿處理30 min和堿處理60 min. Abs.為吸光度。

Note: The visible images of the samples contain the epidermis area, vascular bundle and parenchyma cell from left to right. The average absorbance microscopic images correspond to the visible image. From top to bottom, they are raw samples of corn stover, alkali pretreated by 5 min, alkali pretreated by 30 min and alkali pretreated by 60 min. Abs. is absorbance.

2.2 3組分傅里葉變換紅外光譜圖

圖2是截取的纖維素、木聚糖代表的半纖維素和木質(zhì)素3種組分在1 800～800 cm–1范圍的傅里葉變換紅外光譜圖，3種組分在此范圍內(nèi)的紅外光譜主要吸收峰如圖2所示。表2所示是纖維素、半纖維素和木質(zhì)素紅外光譜圖中不同位置的吸收峰對(duì)應(yīng)的官能團(tuán)歸屬。由圖2可知，纖維素、半纖維素和木質(zhì)素很多譜帶相互重疊，僅1 432和1 514 cm–1處的吸收峰受其他物質(zhì)吸收峰的干擾較小，可分別作為纖維素和木質(zhì)素的特征峰，而半纖維素不存在不受干擾的特征峰，本文采用快速非負(fù)最小二乘（fast NNLS）擬合算法避免了各組分光譜特征峰相互影響的 問題。

圖2 纖維素、半纖維素和木質(zhì)素傅里葉變換紅外光譜圖

2.3 組織提取結(jié)果

由均值聚類算法提取的原樣及堿處理樣品的表皮、維管束和薄壁細(xì)胞等組織如圖3所示。表皮、維管束和薄壁細(xì)胞3種組織分別由表1中3列樣品可見光圖像及其對(duì)應(yīng)的FTIR平均吸光度顯微圖像從左至右依次對(duì)應(yīng)提取得到。組織提取時(shí)先將可見光圖像分解成RGB/HSV/ Lab多空間分量，選取組織結(jié)構(gòu)明晰、冗雜信息少的分量與可見光圖像和光譜預(yù)處理后的FTIR顯微圖像進(jìn)行融合，從而突顯不同組織的結(jié)構(gòu)特征，繼而利用均值聚類算法提取圖像中的表皮、維管束和薄壁細(xì)胞3種組織。從圖3中可以看出，隨預(yù)處理時(shí)間增加，樣品組織形態(tài)的完整度逐漸減小，堿處理60 min的樣品最不完整，且薄壁細(xì)胞完整度減小最明顯。

2.4 FTIR顯微成像及fast NNLS擬合計(jì)算結(jié)果

基于試驗(yàn)分析的木質(zhì)素含量（化學(xué)校正值）和基于fast NNLS擬合的3個(gè)組織（表皮、維管束和薄壁細(xì)胞）中木質(zhì)素含量隨堿處理時(shí)間的變化，如圖4所示?；瘜W(xué)校正值是試驗(yàn)測(cè)定的木質(zhì)素含量與對(duì)應(yīng)堿處理固體回收率的乘積。將試驗(yàn)測(cè)定的木質(zhì)素含量轉(zhuǎn)化為木質(zhì)素絕對(duì)含量，使試驗(yàn)測(cè)定的堿處理不同時(shí)間的樣品木質(zhì)素含量具有可比性?；瘜W(xué)校正值隨堿處理時(shí)間增加不斷下降，且反應(yīng)初始階段，木質(zhì)素降解速率最大，隨后木質(zhì)素降解速率逐漸減小。這可能是由樣品中木質(zhì)素含量減少，反應(yīng)困難程度增大而引起的。

由圖4中基于fast NNLS擬合計(jì)算的3個(gè)組織（表皮、維管束和薄壁細(xì)胞）中木質(zhì)素含量可知，玉米秸稈表皮中的木質(zhì)素含量高于維管束中的含量，而薄壁細(xì)胞中木質(zhì)素含量最少。這一計(jì)算結(jié)果與植物學(xué)研究結(jié)果一致，表皮和維管束中木質(zhì)素大量沉積有利于鞏固和支持植物體、保護(hù)植物體及物質(zhì)輸導(dǎo)。擬合計(jì)算得到，原樣組織薄壁細(xì)胞中木質(zhì)素質(zhì)量分?jǐn)?shù)為7.7%，堿處理5、30、60 min后下降至6.0%、4.8%、3.5%，圖中結(jié)果顯示，除原樣（圖中第1個(gè)點(diǎn)）外，堿處理樣的表皮、維管束和薄壁細(xì)胞3個(gè)組織中木質(zhì)素含量均呈現(xiàn)隨著堿處理時(shí)間增加而不斷減少的趨勢(shì)；并且3個(gè)組織中木質(zhì)素變化趨勢(shì)均與化學(xué)校正值的變化趨勢(shì)一致。表皮中木質(zhì)素降解速率快于其他3個(gè)組織，這可能與表皮中木質(zhì)素含量較高有關(guān)，另外2個(gè)組織中木質(zhì)素降解速率則相當(dāng)；組織中木質(zhì)素降解速率逐漸減小，與化學(xué)校正值速率的變化一致。

圖5是基于fast NNLS擬合計(jì)算的表皮、維管束和薄壁細(xì)胞等3個(gè)組織中木質(zhì)素分布的FTIR顯微圖像。FTIR顯微圖像可以直觀看到圖5中的計(jì)算結(jié)果：表皮中木質(zhì)素含量大于維管束和薄壁細(xì)胞中的，隨著堿處理時(shí)間增加，各個(gè)組織中木質(zhì)素含量逐漸減少。同時(shí)，各組織形態(tài)也出現(xiàn)了不同程度的破壞，維管束和薄壁細(xì)胞的破壞程度比較明顯。

圖5中薄壁細(xì)胞的FTIR顯微圖像顯示細(xì)胞形狀隨堿處理時(shí)間的增加愈加不完整，一方面是由于所選區(qū)域的薄壁細(xì)胞經(jīng)堿處理原有形態(tài)被嚴(yán)重破壞；另一方面，由于薄壁細(xì)胞細(xì)胞壁較表皮部分和維管束薄，導(dǎo)致紅外吸收光程短，所以薄壁細(xì)胞的紅外響應(yīng)值較低，引起細(xì)胞形態(tài)部分缺失。

a. 原樣 a. Raw sampleb. 堿處理5 minb. Alkali pretreated by 5 min c. 堿處理30 min c. Alkali pretreated by 30 mind.堿處理60 mind. Alkali pretreated by 60 min

2.5 討論

圖4中原樣的（第1個(gè)點(diǎn)）基于fast NNLS擬合計(jì)算值與整體趨勢(shì)不符，推測(cè)是擬合計(jì)算值出現(xiàn)偏差。NNLS擬合計(jì)算時(shí)須令待分析組分等于，本試驗(yàn)中取3，即假定每個(gè)像素點(diǎn)僅由纖維素、半纖維素和木質(zhì)素3種組分組成，組分占比總和越接近1，擬合值越準(zhǔn)確。但實(shí)際上，如表3所示，玉米秸稈原樣中纖維素、半纖維素和木質(zhì)素三者總和僅為83.36%，原樣中還存在淀粉、果膠、灰分等多種成分，因而擬合值出現(xiàn)偏差。而堿處理不同時(shí)間的樣品中，纖維素、半纖維素和木質(zhì)素三者總和均在90%以上，所以擬合值更準(zhǔn)確。為解決原樣擬合值出現(xiàn)偏差的問題，在后續(xù)研究中可以通過增加參與fast NNLS計(jì)算的組分?jǐn)?shù)，使組分總和盡量接近于1。

圖4 木質(zhì)素含量隨預(yù)處理時(shí)間的變化

注：每個(gè)小標(biāo)題對(duì)應(yīng)的3個(gè)小圖分別為表皮、維管束和薄壁細(xì)胞組織。組織越接近紅色表明木質(zhì)素含量越高，越接近藍(lán)色，表明木質(zhì)素含量越低.

表3 樣品中木質(zhì)纖維成分的含量

3 結(jié) 論

本研究采用傅里葉變換紅外（FTIR）顯微成像系統(tǒng)獲取玉米秸稈原樣切片和堿處理不同時(shí)間的切片的顯微圖像，然后基于快速非負(fù)最小二乘（fast NNLS）擬合算法計(jì)算樣品各個(gè)組織（表皮、維管束和薄壁細(xì)胞）中木質(zhì)素含量并在FTIR顯微圖像中直觀表達(dá)木質(zhì)素含量的變化。對(duì)比基于fast NNLS擬合所得的3個(gè)組織（表皮、維管束和薄壁細(xì)胞）中木質(zhì)素含量和基于試驗(yàn)分析的木質(zhì)素含量（化學(xué)校正值）隨堿處理時(shí)間的變化發(fā)現(xiàn)，各個(gè)組織中木質(zhì)素含量逐漸下降，且與化學(xué)校正值變化趨勢(shì)一致；木質(zhì)素降解速率呈現(xiàn)先快后慢；表皮中木質(zhì)素含量高于其他2個(gè)組織，降解速率也大于其他2個(gè)組織；受算法原理的限制，原樣各組織中木質(zhì)素?cái)M合計(jì)算含量存在偏差，需進(jìn)一步研究。由本文研究結(jié)果可知，F(xiàn)TIR顯微成像結(jié)合fast NNLS算法可以實(shí)現(xiàn)木質(zhì)素的定位和定量分析，能用于表征玉米秸稈堿處理過程中不同組織的木質(zhì)素分布和含量變化。

[1] 楊增玲，梅佳琪，曹聰，等. 基于紅外光譜的不同農(nóng)作物秸稈磨木木質(zhì)素差異表征[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2018，34(19)：219－224. Yang Zengling, Mei Jiaqi, Cao Cong, et al. Traits of milled wood lignin isolated from different crop straw based on FT-IR[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2018, 34(19): 219－224. (in Chinese with English abstract)

[2] 李雅麗，高錦紅，王麗. 堿處理對(duì)農(nóng)作物秸稈中木質(zhì)素含量的影響[J]. 化學(xué)與生物工程，2017，34(10)：58－60.Li Yali, Gao Jinhong, Wang Li. Effect of alkali treatment on lignin content in crop straw[J]. Chemistry & Bioengineering, 2017, 34(10): 58－60. (in Chinese with English abstract)

[3] Cesarino I, Araújo Pedro, Domingues Júnior Adilson Pereira, et al. An overview of lignin metabolism and its effect on biomass recalcitrance[J]. Brazilian Journal of Botany, 2012, 35(4): 303－311.

[4] 路瑤，魏賢勇，宗志敏，等. 木質(zhì)素的結(jié)構(gòu)研究與應(yīng)用[J]. 化學(xué)進(jìn)展，2013，25(5)：838－858. Lu Yao, Wei Xianyong, Zong Zhimin, et al. Structural investigation and application of lignins[J]. Progress in Chemistry, 2013, 25(5): 838－858. (in Chinese with English abstract)

[5] 馬建鋒，楊淑敏，田根林，等. 植物細(xì)胞壁木質(zhì)素區(qū)域化學(xué)紫外顯微光譜研究進(jìn)展[J]. 林產(chǎn)化學(xué)與工業(yè)，2016，36(1)：147－154. Ma Jianfeng, Yang Shumin, Tian Genlin, et al. Application of UV-microspectrophotometry on lignin topochemistry in plant cell wall[J]. Chemistry and Industry of Forest Products, 2016, 36(1): 147－154. (in Chinese with English abstract)

[6] Zhu Shengdong, Wu Yuanxin, Yu Ziniu, et al. Pretreatment by microwave/alkali of rice straw and its enzymic hydrolysis [J]. Process Biochemistry, 2005, 40(9): 3082－3086.

[7] Hu Zhenhu, Wen Zhiyou. Enhancing enzymatic digestibility of switchgrass by microwave-assisted alkali pretreatment[J]. Biochemical Engineering Journal, 2008, 38(3): 369－378.

[8] Li Junbao, Lu Mingsheng, Guo Xiaomiao, et al. Insights into the improvement of alkaline hydrogen peroxide (AHP) pretreatment on the enzymatic hydrolysis of corn stover: chemical and microstructural analyses[J]. Bioresource Technology, 2018, 265: 2－7.

[9] Micco V D, Aronne G. Combined histochemistry and autofluorescence for identifying lignin distribution in cell walls[J]. Biotechnic & Histochemistry, 2007, 82(4/5): 209－216.

[10] Liljegren S. Phloroglucinol stain for lignin[J]. Cold Spring Harbor Protocols, 2010(1): pdb.prot4954. Doi: 10.1101/pdb. prot4954

[11] Coletta V C, Rezende C A, Da Concei??o F R, et al. Mapping the lignin distribution in pretreated sugarcane bagasse by confocal and fluorescence lifetime imaging microscopy[J]. Biotechnology for Biofuels, 2013, 6(1): 43－43.

[12] Ji Zhe, Ma Jianfeng, Zhang Zhiheng, et al. Distribution of lignin and cellulose in compression wood tracheids of Pinus yunnanensis determined by fluorescence microscopy and confocal Raman microscopy[J]. Industrial Crops and Products, 2013, 47: 212－217.

[13] 劉蒼偉，蘇明壘，周賢武，等. FTIR及CLSM對(duì)轉(zhuǎn)基因楊木細(xì)胞壁木質(zhì)素含量及微區(qū)分布研究[J]. 光譜學(xué)與光譜分析，2017(11)：89－93. Liu Cangwei, Su Minglei, Zhou Xianwu, et al. Analysis of content and distribution of lignin in cell wall of transgenic poplar with Fourier infrared spectrun (FTIR) and confocal laser scanning microscopy (CLSM)[J]. Spectroscopy and Spectral Analysis, 2017(11): 89－93. (in Chinese with English abstract)

[14] Miyafuji H, Komai K, Kanbayashi T. Development of quantification method for lignin content in woody biomass by Raman micro-spectroscopy[J]. Vibrational Spectroscopy, 2017, 88: 9－13.

[15] Singh A, Daniel G, Nilsson T. Ultrastructure of the S2 layer in relation to lignin distribution inpinus radiata tracheids[J]. Journal of Wood Science, 2002, 48(2): 95－98.

[16] Westermark U, Lidbrandt O, Eriksson I. Lignin distribution in spruce () determined by mercurization with SEM-EDXA technique[J]. Wood Science & Technology, 1988, 22(3): 243－250.

[17] Lan Sun, Li Chenlin, Xue Zhengjun, et al. Unveiling high-resolution, tissue specific dynamic changes in corn stover during ionic liquid pretreatment[J]. RSC Advances, 2013, 3: 2017－2027.

[18] Ma Jianfeng, Ji Zhe, Zhou Xia, et al. Transmission electron microscopy, fluorescence microscopy, and confocal raman microscopic analysis of ultrastructural and compositional heterogeneity ofL. wood cell wall[J]. Microscopy and Microanalysis, 2013, 19(1): 243－253.

[19] Pelletier M J, Pelletier C C. Spectroscopic Theory for Chemical Imaging[M]// Raman, Infrared, and Near-Infrared Chemical Imaging. John Wiley & Sons Inc, 2010.

[20] Türker-Kaya S, Huck C W. A review of mid-infrared and near-infrared imaging: Principles, concepts and applications in plant tissue analysis[J]. Molecules, 2017, 22(1): 168－187.

[21] Dokken K M, Davis L C. Infrared imaging of sunflower and maize root anatomy[J]. Journal of Agricultural and Food Chemistry, 2007, 55(26): 10517－10530.

[22] Gorzsás A, Stenlund H, Persson P, et al. Cell‐specific chemotyping and multivariate imaging by combined FT‐IR microspectroscopy and orthogonal projections to latent structures (OPLS) analysis reveals the chemical landscape of secondary xylem[J]. The Plant Journal, 2011, 66(5): 12.

[23] Yang Zengling, Mei Jiaqi, Liu Zhiqiang, et al. Visualization and semiquantitative study of the distribution of major components in wheat straw in mesoscopic scale using fourier transform infrared microspectroscopic Imaging[J]. Analytical Chemistry, 2018, 90(12): 7332－7340.

[24] Shih C J, Lupoi J S, Smith E A. Raman spectroscopy measurements of glucose and xylose in hydrolysate: Role of corn stover pretreatment and enzyme composition[J]. Bioresource Technology, 2011, 102(8): 5169－5176.

[25] Rossi S, Menardi R, Anfodillo T. Trephor: A new tool for sampling microcores from tree stems[J]. IAWA Journal, 2006, 27(1): 89－97.

[26] Sluiter A, Hames B, Ruiz R, et al. Determination of Structural Carbohydrates and Lignin in Biomass: NREL/TP-510-42618 [S]. National Renewable Energy Laboratory, Golden, Co, 2011: 1－15.

[27] Sun Xiaofeng, Sun Runcang, Fowler P, et al. Extraction and characterization of original lignin and hemicelluloses from wheat straw[J]. Journal of Agricultural and Food Chemistry, 2005, 53(4): 860－870.

[28] Fang J M, Sun R C, Tomkinson J. Isolation and characterization of hemicelluloses and cellulose from rye straw by alkaline peroxide extraction[J]. Cellulose, 2000, 7(1): 87－107.

[29] Bro R, Jong S D. A fast non-negativity-constrained least squares algorithm[J]. Journal of Chemometrics, 1997, 11(5): 393－401.

[30] 楊增玲，梅佳琪，韓魯佳等莖稈切片顯微紅外“組織-組分”同步分析軟件V1.0[Z]. 證書號(hào)：軟著登字第BI44059號(hào). 登記號(hào)：2018SRBJ0229.

Lignin content and distribution in alkali pretreated corn straw based on Fourier transform infrared microspectroscopic imaging

Yang Zengling, Du Shurong, Mei Jiaqi, Li Junbao, Liu Erwei, Han Lujia※

(100083,)

Lignin is one of the major components of plant cell wall, and it can provide sufficient strength and hardness to plant cells while avoiding biological damage and water erosion. In recent years, studies on the lignification process and delignification of plant tissues have received extensive attention from scholars. A large number of studies have been carried out to determine the removal of lignin after pretreatment by conventional wet chemical analysis. The lignin content and distribution of biomass resources such as crop straw directly affect its conversion and utilization efficiency. It’s meaningful to develop multi-scale and in situ analysis to identify lignin distribution and quantification on delignified plant cell wall for bio-resource commercial utilization. In this paper, we studied an in situ analysis method to visualize changing of lignin distribution in the alkali pretreated corn straw internodal transverse section based on Fourier transform infrared (FTIR) microspectroscopic imaging, with a fast non-negativity-constrained least squares (fast NNLS) fitting. We collected the middle of the 6th node of fresh jointing stage corn straw. Corn straws were pretreated for series of times (0 , 5 , 30, 60 min) by 2% NaOH solution at 100℃. A paraffin embedding method was used to produce 18m-thick transverse sections for each sample. Then, the sections were transferred onto ZnS windows for FTIR microspectroscopic imaging. We acquired FTIR spectra of major components of corn straw for fast NNLS fitting. We also milled whole straws to 0.069 mm and pretreated in the same condition as sections to determine structure carbohydrate. The data of FTIR microspectroscopic images and FTIR spectra were processed by Savizky?Golay smoothing within 5 points, SNV, and automatic Whittaker filter baseline to correct baseline and offset.-means cluster algorithm was used to distinguish various tissues including epidermis, vascular bundles and parenchyma cells. Fast NNLS fitting was carried out to calculate the lignin concentrations in pixels. For comparison between different pretreated times, lignin concentration in pixels was calibrated according to the sum of content of all components based on laboratory analysis. The lignin content by laboratory analysis decreased with the increasing of the pretreated time, and the lignin degradation rate was larger at the initial stage of the reaction. The lignin content in the epidermis, vascular bundle and parenchyma cells of the samples showed a decreasing trend with the increasing of pretreatment time, and the trends of lignin changing in the three tissues were consistent with the trend of laboratory analysis. By comparing laboratory analysis with fast NNLS fitting results, the results show that the FTIR microspectroscopic imaging combined with fast NNLS could locate and quantify the distribution of lignin in various tissues, and the trend of lignin changing calculated by fast NNLS fitting is the same as the laboratory chemical analysis results. The study demonstrated that FTIR microspectroscopic imaging combined with fast NNLS fitting could be successfully applied to in situ visualize lignin distribution within corn straw pretreatment. The characterization method provides reference for the lignin research of corn stover pretreatment and the study of lignin degradation during straw pretreatment.

straw; lignin; image processing; FTIR microspectroscopic imaging; alkali pretreatment

2019-01-09

2019-04-08

國家自然科學(xué)基金（31471407）；教育部創(chuàng)新團(tuán)隊(duì)計(jì)劃（IRT_17R105）

楊增玲，教授，博士生導(dǎo)師，教育部新世紀(jì)優(yōu)秀人才，主要從事光譜、顯微光譜和光譜成像技術(shù)在農(nóng)業(yè)中的應(yīng)用研究。 Email：yangzengling@cau.edu.cn

韓魯佳，教授，博士生導(dǎo)師，長江學(xué)者特聘教授，主要從事農(nóng)業(yè)生物質(zhì)工程研究資源開發(fā)與利用和農(nóng)產(chǎn)品加工工程方向的研究。 Email：hanlj@cau.edu.cn

10.11975/j.issn.1002-6819.2019.08.033

S216.2

A

1002-6819(2019)-08-0280-07

楊增玲，杜書榮，梅佳琪，李駿寶，劉二偉，韓魯佳.FTIR顯微成像表征堿處理后玉米秸稈木質(zhì)素含量及分布[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2019，35(8)：280－286. doi：10.11975/j.issn.1002-6819.2019.08.033 http://www.tcsae.org

Yang Zengling, Du Shurong, Mei Jiaqi, Li Junbao, Liu Erwei, Han Lujia. Lignin content and distribution in alkali pretreated corn straw based on Fourier transform infrared microspectroscopic imaging[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2019, 35(8): 280－286. (in Chinese with English abstract) doi：10.11975/j.issn.1002-6819.2019.08.033 http://www.tcsae.org