1、Data Acquisition and Processing,谱仪的调试,匀场及参数的设定,目的:获得最高的灵敏度和分辨率保障实验的可重复性一般通用实验过程:,进样,调出标准参数,样品的平衡过程,探头的调谐,锁场,匀场(优化磁场的均匀程度),收集信号,探头的调谐和匹配,每个探头对不同频率的信号的灵敏度不同 调谐:调整探头电路(可调电容)以获得相应发射频率最大的灵敏度 匹配:调节探头的线路阻抗以使能量最大限度地被传递到线圈中 调谐和匹配的结果与探头,溶剂和样品体积相关,进行杂核或二维谱测定时,需要调节探头的调谐和匹配,探头的调谐和匹配,bad matching and tuning,bad m
2、atching, good tuning,good matching, bad tuning,good matching and tuning,Tuning and matching affect each other.,探头的调谐和匹配,锁场(locking),这个超导磁体的范围很大,而所有观察的磁场只是其中的一小部分,所以要用某种物质的信号做为标准,锁定欲观察的频率范围。 氘代溶剂里的D信号。,Bruker,Varian,锁场(locking),Lock level的最佳范围在80100之间。 建议相应提高lock gain而使用较低的lock power。 Lock power 太高时,
3、会导致对样品产生影响的能量在驰豫过程中无法被完全消散掉,进而导致溶液的过热及谱图质量的下降,称为saturation。 调节lock power 使锁场信号的强度大约为最高值的80。,匀场(Shimming),磁场的均匀程度重要,不均匀的磁场会严重导致谱线的加宽 。,CH3CHO,Shims是一些小的电磁场(线圈),用以抵消样品中的磁场的非均匀性。 调节这些小电磁场以达到整个样品中磁场的均匀的过程成为shimming。,No shimming, no NMR!,匀场(Shimming),奇数Z:峰宽 偶数Z:对称性 X,Y:旋转边带及对称性 低阶的匀场影响更大,匀场(Shimming),Z3,
4、Z4,Z5,Z2,Z1,匀场(Shimming),各级匀场之间相互影响,Z2 Z1 Z3 Z, (Z2) Z4 Z0, Z2, (Z1, Z3) Z5 Z3, Z1, (Z2, Z4) Z6 Z4, Z2, Z0, (Z1, Z3, Z5) X Y, (Z1) Y X, (Z1) XZ X, (Z1) YZ Y, (Z1) XY X, Y,一般情况下调节Z1, Z2, Z4 旋转边带特别明显或峰型怪异时调节X, Y, XY等,Shimming using lock signal,Lineshape,用标准样品的线形来衡量磁场的均匀程度。1% CHCl3 in (CD3)2CO对称峰型峰宽表示:
5、50高度处,0.55高度处,0.11高度处。以20Hz频率旋转,无旋转边带。,Good shimming,Bad y- shimming,Shimming using lineshape,Shimming using lineshape,Shimming using FID,well shimmed,poorly shimmed rapid decay broad lines,poorly shimmed,Experimental Considerations Data Acquisition,NS: number of scans1H: 16 13C: 1024 SW: spectral w
6、idth1H: 20 ppm 13C: 220 ppm AQ: acquisition time1H: 24sec. 13C: 0.51 sec. RG: receiver gain由仪器自动测定 D1: delay between scans1H: 12 sec. 13C: 12 sec.,用户可修改,脉冲对核磁信号的影响,Dt,当前的核磁共振波谱仪使用射频频率的短脉冲激发原子核. PW: pulse width 较短的脉冲可以激发较宽的频率范围; 较宽而能量低的脉冲,通常用于选择性激发,excitation range,采样点数和谱宽,目的: 准确地确定频率 Nyquist theorem
7、:要准确定义振动频率,至少每个波长内要采样两点以上。 400MHz 每秒至少采样800000000点 SW:spectral width, 谱宽1H:20 ppm 13C:240 ppm DW:dwell time,采样点之间的时间间隔DW = 1/(2SW),信号频率在谱宽范围之内的信号,可以得到足够的采样,确定准确的频率;而频率高于这个范围的,无法被准确采样,因而信号出现在低于它的实际频率的位置而成为错误的信号。(aliased),采样点数和谱宽,correct,correct,aliased,如果谱图中出现多余峰,且相位不同于其它的峰,很可能是谱宽设置的问题。,采样点数和谱宽,在Vari
8、an的谱仪上,aliased峰一般出现在整个谱图的最右端,correct,aliased,采样时间和数字化分辨率 (Acquisition Times and Digital Resolution),Acquisition Time: total sampling period of the FID 要想分辨频率上相差Dn Hz的两条谱线,至少需要采样1/ Dn秒 在AQ内可以分辨的最小峰宽大约为0.6/AQ AQ太小无法区分谱线的精细结构,DR1/AQ,TD:number of time domain data points DR:digital resolution (Hz/pt) 不同于
9、我们所说的谱图的分辨率 谱图的分辨率取决于磁场的均匀程度,而DR之取决于谱宽和采样点数,Experimental Considerations - Acquisition Times and Digital Resolution,DR1/AQ,Experimental Considerations - Acquisition Times and Digital Resolution,high resolution : long acquisition timesmall spectral width large number of data points1H: 2-4 s 13C: 0.5-1
10、 s,Experimental Considerations - Acquisition Times and Digital Resolution,at 太短,at 适当,Truncation of the FID: too short of an acquisition time.,Experimental Considerations - ADC Gain,Fourier transformation: analog-to-digital converter (ADC). Do not want the signal to have the same amplitude as the di
11、gitization noise. The conversion factor is controlled by the parameter receiver gain or gain.,Receiver Gain and FID,The clipped FID as a result of RG set too high,Receiver Gain and FID,RG太高,会造成基线扭曲变形,RG太高,会造成信号太弱,Experimental Considerations - Signal Averaging and Signal-to-noise ratio,当扫描的次数增大n倍时,信噪
12、比增加,当谱仪频率增大n倍时,信噪比增加,Sensitivity of NMR Measurements,Experimental Considerations - Signal Averaging and Signal-to-noise ratio,NS1416,Experimental Considerations - Signal Averaging and Signal-to-noise ratio,NS41664,原子核在吸收电磁波,从低能级跃迁到高能级。而在这种条件下,原子核高能级回到低能级,并保持低能级布居数小于高能级布居数的过程就叫做弛豫过程。如无有效的弛豫过程,高、低能级的布
13、居数很快达到平衡,此时,无核磁共振吸收信号,这种现象叫做饱和。,弛豫(Relaxation),弛豫(Relaxation),弛豫可分为纵向弛豫和横向弛豫。 纵向弛豫 spin-lattice relaxation, or longitudinal relaxation T1:又称自旋 - 晶格弛豫。处于高能级的核将其能量及时转移给周围分子骨架 ( 晶格 ) 中的其它核,从而使自己返回到低能态的现象。 横向弛豫 spin-spin relaxation, or transverse relaxation T2:又称自旋 - 自旋弛豫。当两个相邻的核处于不同能级,但进动频率相同时,高能级核与低能级
14、核通过自旋状态的交换而实现能量转移所发生的弛豫现象。,纵向弛豫时间T1,质子的弛豫时间T1一般在0.52s,因此在常规实验的时间范围内可以得到充分的弛豫,从而得到准确的定量。 13C的弛豫时间T1可以从几秒到几十秒,在一般实验条件下不能得到充分的弛豫,如果需要定量,则需要采用特别的程序。,13C驰豫时间:大多数110 s 长驰豫时间:气相,I1/2重核,季碳,1,2,4,6,15,acetone,纵向弛豫时间T1,NS: number of scans1H: 16 13C: 1024 太小则信噪比不够,太大则用时过长 SW: spectral width1H: 20 ppm 13C: 220
15、ppm 大小则会使某些信号出于区域外而显示错误频率,太大则会降低谱图分辨率 AQ: acquisition time1H: 24sec. 13C: 0.51 sec. 太小会降低谱图分辨率,并导致接近基线部分的峰变形 RG: receiver gain由仪器自动测定 太小会降低灵敏度而使信号减弱,太大会导致接近基线部分的峰变形及基线扭曲 D1: delay between scans1H: 12 sec. 13C: 12 sec. 太小则使某些峰不能定量甚至不显示,太大则用时过长,Experimental Considerations Data Processing,FT,数学处理(窗函数,充
16、零,线型预测),FT,相位调整,窗函数(Window Functions),当采样结束时FID尚未衰减为零(truncated FID),经过傅立叶变换后会在邻近峰的基线处出现比较强的噪音,需要经过窗函数处理解决( Apodization )。,EM - Exponential Multiplication, lb=1,EM - Exponential Multiplication, lb=3,Applies an exponential weighting function to the FID to enhance signal-to-noise,SG - Shifted Gaussian
17、, lb=1,SG - Shifted Gaussian, lb=3,Applies a shifted Gaussian function to the FID to enhance resolution,窗函数的使用,Exponential 增加灵敏度但牺牲分辨率 lb in Bruker or Varian width in JEOL or MestreC 设为半锋宽,Gaussian 提高分辨率但增加噪音牺牲信噪比 GB 00.5 Bruker lb为负值, MestreC lb为正值,窗函数的使用- exponential,Lb 越大则灵敏度越高,但分辨率下降1H: 0 lb 0.4
18、 Hz13C: 1 lb 3 Hz,窗函数的使用- Gaussian,GB 越大而LB越小,则分辨率越高而灵敏度越低 0 GB 1 LB 0,窗函数的使用,无窗函数处理,Lorentzian apodization (lb),Gaussian apodization (gf and gfs),90 shifted sinebell (sb or sbs),Zero-Filling and Linear Prediction,Zero-Filling: 在FID的结尾加零,可以使峰型更平滑在一定程度上增加峰的分辨率缺点: 增加了数学变换过程的计算量,Zero-Filling and Linear
19、 Prediction,一般设定: SI 0.5 TD充零: SI TD,Zero-Filling and Linear Prediction,Linear Prediction:根据现有FID的形状预测变形分的数据点 比单纯的充零更接近真实的FID形状 提高分辨率 要求有比较高的信噪比做为预测的根据,original FID,apodized with cosine squared and zero-filled one time,linear prediction and cosine squared window function,Zero-Filling and Linear Pred
20、iction,Linear Prediction,Linear Prediction,Experimental Considerations - Phasing,Experimental Considerations - Phasing,Zero-order correction: 与频率无关First-order correction: 与频率相关首先进行Zero-order correction,Experimental Considerations - Phasing,Experimental Considerations - Phasing,Good Phasing,Bad first-order correction,Bad zero-order correction,Bad zero- and first-order correction,