红外光谱的样品制备 – 第一部分
每年各地红外光谱的实验室制备和利用红外光谱仪分析成千上万个样品。
这些样品范围从商业产品像高聚物颗粒和液体表面活性剂，一直到高纯度有机化合物。为了从这些不同的材料中得到高质量的红外谱图，我们必须采用多种多样的制样技术。这篇文章的旨在与您交流红外制样技术。在这篇文章中，将对基于样品的物理特性进行的技术选择作讨论。
液体
液样的制备是将少量样品涂于两片红外透明的窗片（KBr、NaCl等）之间。窗片的互相挤压形成一个样品薄层，样品的成分决定了选择哪种窗片。对于无水的样品，窗片材料是KBr。对于含水样品,
KRS-5 较为适合。
固体
固体样品对光谱学家提出挑战。样品的熔点为我们指出首先该考虑哪种技术。
对于熔点低于72。C的样品，用适当的溶剂将样品溶解，成膜于KBr窗片上是最先考虑的。如果因为基线不好或是溶解性差而不成功，可以考虑在两片KBr窗片内熔化成膜。如果这也不行，样品可进行KBr压片。
对于熔点高于72。C的样品，首选的技术是KBr压片。对于聚合物样品，成膜法是首选，接着是热熔法和压片法。
对于熔点未知的样品，结晶度的检测将会指明哪种技术将会成功。高结晶度的样品用KBr压片法较好，对于低结晶度的样品，成膜和热熔会得到更好的谱图。
红外光谱的样品制备 – 第二部分
液体样品
液体样品的分析有多种方法。在本文中，我们主要探讨所使用的制样方法及一些有关的潜在问题。
纯样品技术
分析液体样品的最常用方法就是将一滴液体夹在两片盐片中间，过程如下：将一滴样品滴于合适的盐片上，几秒钟后，将另外一块盐片合上，这样液体被夹在两块盐片之间，变成薄膜状。当然，选用的盐片要与分析的液体样品兼容。不含水的样品可采用KBr（32×5mm）盐片，含水样品则采用KRS-5盐片，这几种晶体材料的选用主要是根据它们在红外段的透光范围（优于4000－450cm-1）和稳定性。每次一个样品做好后，用带合适的溶剂的棉花清洗，然后在倒有甲醇的鹿皮或鸡皮上抛光。
KBr盐片需要经常进行抛光，以维持其表面的光洁。由于KRS-5晶体有毒，所有只有当其表面被划伤或污染时才需要抛光，而且要求专业人员来完成。
ATR技术
水平的单反射ATR主要是由一个ZnSe晶体的凹槽组成，尽管ZnSe晶体的截止频率为650－700CM-1，但它比其它宽频带的材料要更加耐用。
在样品分析好后，要用适当的溶剂将样品冲掉，再用棉花球擦洗干净，这种材料不需要经常抛光。
潜在问题：
但它最大的问题就是样品谱图的非线性，主要指峰位的位置和强度不满足Beer－Lambert法则：
A = abc这里, A =吸收值
a =摩尔吸收系数
b =光程
c =浓度
The Beer-Lambert Law is typically discussed in regards to quantitative
analysis. Spectral searching is a type of quantitative analysis, since most
searching algorithms use band intensity as a factor. Linearity due to sampling
typically occurs in transmission analysis and not in ATR, due to the nature of
the ATR technique.
Beer-Lambert定理主要是针对定量分析的，谱图检索是定量分析的一种类型，因为谱图检索是以吸收强度为基础的，非线性主要出现在透射中，在ATR中则没有，这主要是由于ATR的技术特点导致的。
The most common cause of non-linearity is inconsistent path length across
the sampling area. For our instruments at 2 cm-1 resolution, the sampling area
is 6 mm at focus. If there are any air bubbles or unevenness in the sampling
area, the path length of the sample will vary across the sample area.
最常见的导致非线性的原因是透过样品的光程的不连续性。分辨率为2时，样品区红外光的聚焦点仅有6MM，如果此时样品区样品厚度薄厚不均或碰巧有气泡或，就会引起此处光程不同。
These variations will cause inaccuracies in the band intensities in the
spectrum. In other words, the bands will be weaker or stronger than they should
be. This condition will lead to reduced spectral search quality. Figure 1 is a
plot of a spectrum of 3,4-dichlorotoluene analyzed as a neat with a consistent
path length. The peak intensity of the band at 808 cm-1 is 0.39, while the peak
intensity of the adjacent band at 870 cm-1 is 0.24 (A808 cm-1/A870 cm-1 =
1.6).
这些不同将导致谱图在各个波段的吸收强度与实际值产生差异，从而降低谱图检索的质量，图是纯3,4-dichlorotoluene的吸收图，在808
cm-1波段处的吸收度为0.39，邻近870 cm-1处是0.24 (A808 cm-1/A870 cm-1 = 1.6)
Figure 2 is the same sample analyzed with uneven coating of the sample on
the crystals. As you can see, the band intensities are different. For this
spectrum, the peak intensity of the band at 808 cm-1 is 0.76, while the peak
intensity of the adjacent band at 870 cm-1 is 0.62 (A808 cm-1/A870 cm-1 = 1.2).
This represents a change of 25% from the previous spectrum. This could lead to a
poor spectral search match. We 870 cm-1check for unevenness by analyzing the
sample using a 6 mm sampling area, followed by analysis with a 3 mm sampling
area. The two spectra are subtracted from each other and the residual spectrum
is examined.
图二是同一个样品但是通过在晶体上做成一层薄厚不均的膜而得到的谱图，它的吸收率与上图已经有差别了，808 cm-1处的吸收率是0.76而870
cm-1处却为0.62(A808 cm-1/A870 cm-1 =
1.2)，与图一相比，已有25％的差距，这必然会导致谱图检索结果正确率的下降，将样品聚焦斑为6mm时得到的的图与聚焦斑为3mm时得到的图相减，用差谱的结果来进行分析：
Non-linearities due to unevenness would show up as a large residual. We
also minimize unevenness by frequent polishing of the crystals.
由薄厚不均导致的非线性将会使差谱的结果非常明显，我们可以通过定期对晶体进行抛光来降低这种误差。
The second cause of non-linearity in the spectrum of a liquid sample is
sample thickness. A large percentage of spectra of liquid samples submitted to
us for spectral search analysis have been run too thick. Most FT-IR
spectrometers/detectors have a linear response up to approximately 1.2
absorbance units. Any band above 1.2 absorbance units is of questionable
linearity. Non-linearitycan sometimes be seen as “choppiness” on the top of the
band. We only accept spectra from our lab that have a maximum absorbance value
of 1.2 absorbance units. Figure 3 is a plot of the spectrum of the same sample
used previously. However, the sample thickness was much larger than the previous
spectrum.
经常引起液体样品谱图的非线性的另一个原因是样品的厚度。液样太浓将会导致谱图的吸收太强，而多数红外仪器的检测器的线性响应范围是0到1.2个吸收单元，大于1.2时就会引起线性问题。有时非线性会使谱图中吸收峰的头部成平头状，在我们的实验室中只接受吸收单元1.2的谱图，图三也是上面提到的样品的谱图，但样品的厚度却远远大于前者。
The strongest band in the spectrum is over 30 absorbance units. For this
spectrum, the peak intensity of the band at 808 cm-1 is 1.66, while the peak
intensity of the adjacent band at 870 cm-1 is 0.94 (A808 cm-1/A870 cm-1 =
1.76).
This represents a change in relative intensity of 10%, compared to the
normal spectrum in Figure 1. This change in relative band intensities will have
a negative effect on search accuracy.
谱图中最强的吸收单元已经超过了30个吸收单元，808 cm-1处的吸收度为1.66，870 cm-1处为0.94(A808 cm-1/A870
cm-1 = 1.76)，相比而言，产生了10％的误差，这种不同波段的吸收值的相对性的差异将会给谱图检索带来负面影响
Sample Preparation for IR Spectra - Part
III
红外样品的制备 – 第三部分III
成膜技术Film Technique
在
Sadtler的实验室，涂膜技术是用在熔点低于72。C的样品和低结晶度的样品，比如象高聚物，涂膜法也可在其他方法失败后试用。
<
