Introduction to NIR Spectroscopy

Heading

An introduction to near infrared spectroscopy

A.M.C. Davies (Editor NIR news, Norwich Near Infrared Consultancy, 75 Intwood Road, Cringleford, Norwich
NR4 6AA, UK)

When you hold your hand out to a What is Spectroscopy? will be seen as a stretching. When three
burning fire you “feel” the heat being or more atoms are involved then bonds
emitted by the fire but what is happen- Almost the whole of the electromag- can also bend, giving rise to a whole
ing? The fire gives out light and infrared netic spectrum is used in scientific series of different vibrations. Stretch
(IR) radiation; from a fire most of this is studies. Different regions require differ- vibrations require more energy than
near infrared (NIR) radiation. Some of ent instruments to generate the appro- bending vibrations but there will also be
the NIR radiation is absorbed by water priate energy. An appropriate detec- variation in the energy requirements of
molecules in your skin. This raises the tor is required to measure absorption the bending vibrations. Different chemi-
temperature of the water and results in of energy by test samples compared cal bonds (like O–H, C–H and N–H)
an increase in temperature in the sur- to a reference. Such instruments are vary in strength and hence the amount
rounding tissue which is detected by called spectrometers. The primary of energy required for the bond vibra-
nerves in your skin. This radiation was results from a spectrometer are usu- tion to move from one level to the next.
discovered in 1800 by William Herschel, ally viewed as a graph of absorption This variation in energy will be seen in a
a musician and very successful ama- plotted against wavelength (or equiva- spectrum as a series of absorptions at
teur astronomer (he discovered the lent scales) and this is called a spec- different wavelengths. By looking at the
planet Uranus) because he wanted to trum. On an NIR spectrum you will spectrum we can deduce what vibra-
know if any particular colour was asso- normal find that the absorption scale tions are occurring and hence work out
ciated with heat from sunlight. He found is marked “Log 1/R” where R is the the structure of the molecule (or groups
that the heat maximum was beyond the reflection of the sample and the wave- of atoms present).
red end of the spectrum. Herschel could length scale is measured in nanome- One of the very useful properties of
not believe that light and his “radiant tres (nm). The energy of the radiations mid-IR spectra is that the region from
heat” were related but he was wrong. decrease with increasing wavelength. 8500 to 12,500 nm is very characteris-
By 1835 Ampere had demonstrated that X-rays are much more dangerous than tic for the molecule measured and this
the only difference between light and radio waves. region is known as the “finger-print”
what he named “infrared radiation” was region because it can be used to confirm
their wavelength. Then in 1864 James the identity of many pure substances.
Maxwell wrote “This velocity [of elec- While the study of mid-IR spectroscopy
tromagnetic force] is so nearly that of continued to grow, especially after World
light that it seems we have strong rea- a breakthrough in that researchers were War II, interest in the NIR extended to
son to conclude that light itself (includ- able to relate the character of groups of quantitative measurements of water, a
ing radiant heat and other radiations) is atoms within molecules as being related few simple organic compounds and a
an electromagnetic disturbance in the to specific absorptions in the mid-IR very few studies of specific proteins. No
form of waves propagated through the (2500 to 50,000 nm). These absorptions one considered it useful for characteris-
electromagnetic field according to elec- are the result of interactions with the ing samples and it was considered too
tromagnetic laws”. What we now call the fundamental vibrations of the chemical complex for use in quantitative analysis.
electromagnetic spectrum is shown in bonds associated with the atoms of the
Figure 1. groups. We can think of chemical bonds
as weak springs holding together two or
more atoms, these springs will vibrate Absorptions in the
naturally and when energy is added to
Early history the system then they will vibrate more NIR region
energetically. However, atoms in mol-
of the study ecules are constrained by quantum If chemical bonds behaved exact-
mechanics so that only a few specific ly like weak springs then quantum
of infrared energy levels are allowed. If we have mechanics would restrict their vibration
only two atoms then the only vibration to just two states and there would be
absorptions
The first (near) infrared spectra
were measured in 1881 by Abney and
Festing using photographic plates. Not
only did they produce the first spec-
tra but they also suggested, correctly,
that the absorptions were related to –7 15

the chemical composition of the liquids

they investigated, The most important
pioneer of IR spectroscopy was William
W. Coblentz. In 1905 he published the
result of a large study of compounds
whose spectra he had recorded from
1000 to 16,000 nm. Coblentz’s work was Figure 1. The electromagnetic spectrum.

NIR news Vol. 16 No. 7 (2005) �����

(2005) Page 
Introduction to NIR Spectroscopy�
combinations in the same group. So we
see absorptions due the combination of
O–H stretch with O–H bend and C–H
stretch with C–H bend and these occur
in different positions in the spectrum.
Figure 5 is the NIR spectrum of a sam-
ple of biscuit dough (you will see more
of these in Tom Fearn’s article). Biscuit
dough contains several ingredients each
of which contain many different mol-
ecules so this spectrum contains hun-
dreds if not thousands of absorptions
but we see the integration of them all
and there appear to be just a few broad
absorptions. From their position we can
say in general terms the cause of the
absorption, as indicated on the figure.
When the complexity of NIR absorp-
tion was first realised and compared to
Figure 2. The NIR spectrum of chloroform. the relatively more easily understood
mid-IR spectra, it was thought by most
researchers that there was little to be
very few absorptions in the NIR region. atom but all the absorption in its spec- gained by studying NIR spectroscopy.
Absorptions in the NIR region (780 to trum are caused by this single atom. The region became neglected and stu-
2500 nm) are generated from funda- It is an important generalisation that dents were wrongly instructed that there
mental vibrations by two processes; NIR spectroscopy is dominated by was nothing to be gained by study-
overtones and combinations. Overtones hydrogen. Figure 3 is a spectrum of ing the NIR region. Many students are
can be thought of as harmonics. So methanol, CH3OH, which contains four still being taught the same opinion.
every fundamental will produce a series hydrogen atoms (but three are equiva- The requirements were: very low noise
of absorptions at (approximately inte- lent) and this spectrum is much more spectrometers, the electronic computer,
ger) multiples of the frequency (fre- like a typical NIR spectrum with broad the application of mathematical tech-
quency is the reciprocal of wavelength). peaks. Figure 4 is a spectrum of sucrose, niques (chemometrics) and a genius
Combinations are rather more complex. C12H24O12, which shows very broad to bring it all together. The man was
NIR absorptions are at a higher state areas of absorption but also some quite Karl Norris; an engineer working for the
of excitement so they require more narrow peaks. It is important to realise USDA at Beltsville. He had not been
energy than a fundamental absorption. that all of these broad absorptions are taught spectroscopy so he did not know
Combinations arise from the sharing of caused by multiple narrow, over-lap- that there was nothing to be gained in
NIR energy between two or more funda- ping absorptions. NIR spectra are much the NIR region. So, rather like Herschel
mental absorptions. While the number more complex than they appear. who looked for something where there
of possible overtones from a group of Although NIR spectra are more com- was nothing, Norris developed the
fundamental absorptions in a molecule plicated it is possible to make some instruments and utilised computers to
are limited to a few, a very large number general observations. As a fundamental demonstrate that the NIR region was
of combinations will be observed. The O–H stretching absorption is different very useful for quantitative analysis,
effect of all these absorptions combine to a fundamental C–H stretch then the particularly of agricultural samples. One
to make many NIR spectra to look rath- series of overtones generated by these of the reasons why NIR analysis is so
er uninteresting and to consist of only a absorptions will also be different. The useful is that it can use reflected energy
few rather broad peaks. Figure 2 is an same goes for combination bands. The and this means that NIR analysis can
NIR spectrum of chloroform, CHCl3, the most common (and energetic) combina- be done with little or no sample prepa-
molecule contains only one hydrogen tion bands arise from stretch and bend ration. Reflected energy is complex.

Figure 3. NIR spectrum of methanol.

Figure 4. NIR spectrum of sucrose.

Page 10 NIR news Vol. 16 No. 7 (2005)

 Introduction to NIR Spectroscopy
Heading

Karl H. Norris
Karl Norris is regarded as the
“father” of modern near infrared spec-
troscopic analysis. He invented the
technique while working at the USDA
Instrumentation Research Laboratory,
Beltsville, USA. He has won many
awards and honours including the
American Association of Cereal
Chemists’ Osborne medal (1986),
The New York Society for Applied

Figure 5. Spectrum of biscuit dough with the main areas of absorption identi-
fied.

First, because there are two compo- spectroscopy. Consequently, the major-
nents, specular (or mirror-like) and dif- ity of chemistry students leave univer-
fuse. In the context of NIR spectroscopy, sity with no knowledge of NIR, with the
the specular component does not give possible exception of the old fashioned
any information. The diffuse component view that there is nothing useful to learn
depends on the physical nature of the about the NIR region.
sample; particle size being particular- The following articles describe the
ly important. Variation of the physical range of instruments that are available Karl at Beltsville with the Cary 14 in
parameters of a sample causes chang- today to measure NIR spectra and how the background, photographed in
es in the spectrum so that the observed mathematical techniques are used to 1982.
spectrum is a mixture of chemical and transform these complex measurements
physical information. into simple analytical information.
The use of reflected energy was (as Spectroscopy Gold Medal (2001),
he will explain) forced on Karl Norris. The Royal Society of Chemistry’s Sir
While it makes possible the NIR anal- George Stokes Award (2002) and The
ysis of a very much wider range of Further reading Japanese International Fantec Award
samples, it does add another layer of (2003), he was made a life mem-
complication. A complete mathemati- An excellent lecture by Ian Murray is ber of the Council for Near Infrared
cal theory of reflection spectroscopy is available on the ICNIRS website: www. Spectroscopy (1990) and was elect-
not yet available but it has been found icnirs.org/lecture.html ed the First Fellow of Near Infrared
possible by good experimental prac- A Journal of Near Infrared Spectroscopy by the International
tice and the utilisation of mathemati- Spectroscopy paper provides all the Council for Near Infrared Spectroscopy
cal techniques to use NIR reflection information required to comprehend (1997). Although Karl retired from
spectroscopy for analytical chemistry. absorptions in the NIR region: USDA in 1988 he has continued to be
Because the technique can be applied L. Bokobza, “Near infrared a very active consultant and contribu-
with little or no sample preparation, spectroscopy”, J. Near Infrared tor to NIR theory and application. He
analysis times are reduced from hours Spectrosc. 6, 3 (1998). kindly interrupted his current study of
to minutes and furthermore several ana- The paper by Professor Bokobza can some complex data to write his contri-
lytical results can be obtained from the be found on the NIR Publications web- bution.
same NIR data while the conventional site: www.nirpublications.com/special/
analysis would often require another
technique and more hours of work. It Boundaries of the NIR region Karoly Kaffka in Budapest in 1986. Its
is, however, necessary to develop cali- A paper published by the International main purpose was to regulate inter-
brations which require many samples, Union of Pure and Applied Chemistry national conferences but it has also
many hours of work and thousands (or sets out the boundaries of the differ- been trying to promote near infrared
probably millions) of computer calcula- ent regions. Those for NIR are 780 to spectroscopic education. ICNIRS can
tions. With these sort of attributes it is 2,500 nm (or 12,800 to 4000 cm–1 wave­ be found at www.icnirs.org. The next
not surprising that 40 years after the numbers). Pure and Applied Chemistry ICNIRS conference will be in Sweden in
ground-breaking research, a very wide 57, 105 (1985). June 2007: www.nir2007.com
range of analysis can be achieved by
NIR spectroscopy. The International Council for NIR Publications
What is surprising is that in spite Near Infrared Spectroscopy NIR Publications is “The information
of the success of NIR spectroscopic source for near infrared spectroscopy”.
analysis, world-wide there are very few ICNIRS (originally the “C” was for It publishes NIR news and the Journal
university chemistry departments that Committee) was instigated at an inter- of Near Infrared Spectroscopy, as well
have any programme of research in NIR national NIR meeting organised by Dr as books relevant to NIR spectroscopy.

NIR news Vol. 16 No. 7 (2005) Page 11