13 Overview of Application of NIR Spectroscopy to Physical …
313
treatment of the inner hydration shell showed an improved accuracy of the simulated
vibrational spectrum. The authors reasoned that application of a higher ab initio level
method such as perturbation theory to the latter scheme would likely yield an excellent agreement with the experimental spectrum. The study indicated the promising
development directions toward the interpretation of the spectra of hydrated molecules
[40].
13.4 Structural Information Derived from NIR Spectra
NIR spectra are a rich source of information on molecular structures, which are
sensitive to the chemical environment and solvent effects. Sophisticated approaches
are often required to effectively elucidate this information. Toyama et al. examined
temperature-dependent spectral changes in the first overtone of the OH groups of
alkane–α,ω–diols in the liquid and solid phase [41]. It appears that the spectra of
alkane–α,ω–diols with an odd number of carbon atoms are similar in the pure liquid
and solid states. On the other hand, the spectra of diols with an even number of
carbon atoms reveal significant changes upon moving from the solid to the liquid
phase. Hence, an analysis of NIR spectra confirmed the presence of the even–odd
alternation in solid alkane–α,ω–diols. Liu et al. applied NIR spectroscopy coupled
with chemometric methods for the examination of polymorphic transformations of
oleic acid [42]. Temperature-dependent NIR spectra were resolved into independent
spectral components by using alternating least squares (ALS) optimization. The
obtained results demonstrate that the γ→α transition is determined by the behavior
of the COOH group, while the α→β transition is due to the conformational changes
of the acyl chain.
A significant advancement in elucidating structural information from NIR spectra
is linked to the progress in the practical applications of anharmonic methods in
computational chemistry [1]. The theory of NIR spectroscopy and selected applications is presented in another chapter (Introduction to Quantum Vibrational Spectroscopy ). Here, a number of examples of using computational chemistry to increase
the chemical structural specificity of NIR spectroscopy are outlined. General anharmonic approaches, such as VSCF or VPT2 (Introduction to Quantum Vibrational
Spectroscopy ), have been implemented in an almost a routine way in popular computational chemistry software packages. The computationally efficient VPT2 method
has mostly been used in practical applications to NIR spectroscopy. Relatively accurate calculations of entire NIR spectra of medium-sized organic molecules have
become feasible over the recent years [3] improving our understanding of this spectral range. In comparison with MIR or Raman spectroscopies, computational studies
are of particular importance for NIR spectroscopy, due to the complex character of
NIR spectra. Even relatively simple molecules have a large number of contributing
bands as demonstrated by the calculated spectra [43]. As an example, we show
the reconstructed spectra of vinylacetic acid. Figure 13.9 displays the individually
modeled bands and the predicted theoretical lineshapes, which result via summation
313
treatment of the inner hydration shell showed an improved accuracy of the simulated
vibrational spectrum. The authors reasoned that application of a higher ab initio level
method such as perturbation theory to the latter scheme would likely yield an excellent agreement with the experimental spectrum. The study indicated the promising
development directions toward the interpretation of the spectra of hydrated molecules
[40].
13.4 Structural Information Derived from NIR Spectra
NIR spectra are a rich source of information on molecular structures, which are
sensitive to the chemical environment and solvent effects. Sophisticated approaches
are often required to effectively elucidate this information. Toyama et al. examined
temperature-dependent spectral changes in the first overtone of the OH groups of
alkane–α,ω–diols in the liquid and solid phase [41]. It appears that the spectra of
alkane–α,ω–diols with an odd number of carbon atoms are similar in the pure liquid
and solid states. On the other hand, the spectra of diols with an even number of
carbon atoms reveal significant changes upon moving from the solid to the liquid
phase. Hence, an analysis of NIR spectra confirmed the presence of the even–odd
alternation in solid alkane–α,ω–diols. Liu et al. applied NIR spectroscopy coupled
with chemometric methods for the examination of polymorphic transformations of
oleic acid [42]. Temperature-dependent NIR spectra were resolved into independent
spectral components by using alternating least squares (ALS) optimization. The
obtained results demonstrate that the γ→α transition is determined by the behavior
of the COOH group, while the α→β transition is due to the conformational changes
of the acyl chain.
A significant advancement in elucidating structural information from NIR spectra
is linked to the progress in the practical applications of anharmonic methods in
computational chemistry [1]. The theory of NIR spectroscopy and selected applications is presented in another chapter (Introduction to Quantum Vibrational Spectroscopy ). Here, a number of examples of using computational chemistry to increase
the chemical structural specificity of NIR spectroscopy are outlined. General anharmonic approaches, such as VSCF or VPT2 (Introduction to Quantum Vibrational
Spectroscopy ), have been implemented in an almost a routine way in popular computational chemistry software packages. The computationally efficient VPT2 method
has mostly been used in practical applications to NIR spectroscopy. Relatively accurate calculations of entire NIR spectra of medium-sized organic molecules have
become feasible over the recent years [3] improving our understanding of this spectral range. In comparison with MIR or Raman spectroscopies, computational studies
are of particular importance for NIR spectroscopy, due to the complex character of
NIR spectra. Even relatively simple molecules have a large number of contributing
bands as demonstrated by the calculated spectra [43]. As an example, we show
the reconstructed spectra of vinylacetic acid. Figure 13.9 displays the individually
modeled bands and the predicted theoretical lineshapes, which result via summation