Content uploaded by Augusto Marcelli
Author content
All content in this area was uploaded by Augusto Marcelli on Nov 26, 2017
Content may be subject to copyright.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Research review paper
Biological applications of synchrotron radiation infrared spectromicroscopy
Augusto Marcelli
a,
⁎, Antonio Cricenti
b
, Wojciech M. Kwiatek
c
, Cyril Petibois
d
a
INFN, Laboratori Nazionali di Frascati, Via E. Fermi 40, I-00044 Frascati (Rome), Italy
b
Istituto di Struttura della Materia, CNR, via del Fosso del Cavaliere 100, I-00133 Rome, Italy
c
Institute of Nuclear Physics, Polish Academy of Sciences, ul. Radzikowskiego 152, 31-342 Kraków, Poland
d
Université de Bordeaux, CNRS UMR 5248, CBMN, IECB-2, Rue Robert Escarpit, 33607 Pessac-Cedex, France
abstractarticle info
Available online 28 February 2012
Keywords:
Infrared
Imaging
Microscopy
Synchrotron radiation
Extremely brilliant infrared (IR) beams provided by synchrotron radiation sources are now routinely used in
many facilities with available commercial spectrometers coupled to IR microscopes. Using these intense non-
thermal sources, a brilliance two or three order of magnitude higher than a conventional source is achievable
through small pinholes (b10 μm) with a high signal to-noise ratio. IR spectroscopy is a powerful technique to
investigate biological systems and offers many new imaging opportunities. The field of infrared biological
imaging covers a wide range of fundamental issues and applied researches such as cell imaging or tissue
imaging. Molecular maps with a spatial resolution down to the diffraction limitmay be now obtained with a syn-
chrotron radiation IR source also on thick samples. Moreover, changes of the protein structureare detectable in an
IR spectrum and cellular molecular markers can be identified and used to recognize a pathologi calsta tus of a tissue.
Molecular structure and functions are strongly correlated and this aspect is particularly relevant for imaging. We
will show that the brilliance of synchrotron radiation IR sources may enhance the sensitivity of a molecular signal
obtained from smallbiosamples, e.g., a single cell, containing extremely small amounts of organic matter. We will
also show that SR IR sources allow to study chemical composition and to identify the distribution of organic mole-
cules in cells at submicron resolution is possible with a high signal-to-noise rat io. Moreover, the recent availability of
two-dimensional IR detectors promises to push forward imaging capabilities in the time domain. Indeed, with a high
current synchrotron radiation facility and a Focal Plane Array the chemical imaging of individual cells can be
obtained in a few minutes. Within this framework important results are expected in the next years using synchro-
tron radiation and Free Electron Laser (FEL) sources for spectro-microscopy and spectral-imaging, alone or in com-
bination with Scanning Near-field Optical Microscopy methods to study the molecular composition and dynamic
changes in samples of biomedical interest at micrometric and submicrometric scales, respectively.
© 2012 Elsevier Inc. All rights reserved.
Contents
1. Introduction ............................................................. 1391
2. Vibrational techniques ........................................................ 1391
3. IR synchrotron radiation sources .................................................... 1393
4. Spectral-imaging ........................................................... 1394
5. Biological and biomedical applications ................................................. 1395
5.1. Sample preparation ...................................................... 1398
5.1.1. Tissue sections .................................................... 1398
5.1.2. Cell culture ...................................................... 1398
5.2. Is IR spectromicroscopy of living cells possible? ......................................... 1399
6. Scanning Near-field Optical Microscopy (SNOM) ............................................ 1400
7. THz imaging ............................................................. 1401
8. Conclusions ............................................................. 1402
Acknowledgments ............................................................. 1402
References ................................................................ 1403
Biotechnology Advances 30 (2012) 1390–1404
⁎Corresponding author. Tel./fax: +39 06 94032737.
E-mail address: marcelli@lnf.infn.it (A. Marcelli).
0734-9750/$ –see front matter © 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.biotechadv.2012.02.012
Contents lists available at SciVerse ScienceDirect
Biotechnology Advances
journal homepage: www.elsevier.com/locate/biotechadv
Author's personal copy
1. Introduction
Relatively young if compared to the age of Infrared (IR) spectros-
copy, the field of biological Infrared spectromicroscopy and imaging
covers a wide range of fundamental issues and applied studies such
as cell imaging or tissue imaging. The Infrared spectroscopy is certain-
ly one of the oldest and most versatile analytical chemical techniques.
IR was discovered in the early 1800, however only at the end of the
century the first IR spectra were published (Abney and Festing, 1881;
Nichols, 1893). Later, at the end of 1940s (Barer, 1949; Blout et al.,
1949; Gore, 1949) the great advantage of combining an IR spectrometer
with a microscope has been pointed out, to extract accurate molecular
information from small areas of a sample. With the introduction of the
Fourier Transform InfraRed (FTIR) spectroscopy, the use of the IR ex-
panded continuously for qualitative and quantitative analyses in many
disciplines and in particular for industrial applications. Nowadays, char-
acteristic biomedical analysis with IR microscopy regard pharmaceuti-
cals, biomembranes, biopolymers, micro-biology, cancer diagnostics,
etc…It is noteworthy to remember here that one of the first IR spectros-
copy application has a strong biomedical background being devoted to
the investigation of the IR absorption of neoplastic and normaltissues in
the spectral regions 1–15 μmand8–15 μm(Woernley, 1952).
In 1954 the first commercial IR microscope was made available by
Coates et al. (1953). The original IR spectrometers were significantly im-
proved in the following years with the use of high quality detectors and
the introduction of new electronic components, e.g. microprocessors. It is
important to underline here that in 1988 using a microscope Dong was
able to collect the first IR spectra of an individual erythrocyte in both
H
2
0andD
2
0media(Dong et al., 1988). Finally, in 1990, the first FT-IR mi-
croscope specifically designed for IR radiation with Cassegrain objectives
became available. Nowadays, the high sensitivity of the last generation of
microscopes opens new experimental opportunities. Although many dif-
ferent IR instruments are now available, the information contained in an
IR spectrum are the same, no matter they were obtained with a disper-
sive spectrometer, a FT-IR instrument or an IR microscope. Moreover,
the improved performances such as the enhanced frequency accuracy,
the ultimate signal-to-noise ratios and the high data acquisition speed,
combined with modern computational methods, allow the ultimate in-
vestigation of many bio-chemical systems. In addition, the development
of alternative vibrational spectroscopic techniques, e.g., reflection mea-
surements, Attenuated Total Reflection (ATR) probes, etc., extended the
analysis to tick, opaque, small and/or otherwise “difficult”samples to be
investigated in transmission geometry.
In 2003, the first rapid scan FT-IR imaging system was available
and thanks to the first small linear array detector it allowed increasing
both sensitivity and imaging speed. Later, the availability ofarray detec-
tors opened a new scenario in spectral imaging. The imaging technique
is an effective interdisciplinary method used to visualize the chemical
composition of materials but also useful to investigate a broad range
of phenomena. In contrast to a typical image, which is collected over
the entire wavelength response of the detector, a spectral image com-
bining both spatial and spectral information represents the spatial
(and eventually the time) distribution of a single molecular component,
e.g., the water content, or of a “selected”wavelength range, revealing
trends as well as fine details of the morphology and molecular compo-
sition in the sample. Although it is always possible to record spectral
images with a single-point detector, the IR array detectors, also
known as focal plane arrays (FPAs) offer a great advantage, in particular,
a bidimensional detector with mxmpixels may allow a reduction of the
image acquisition time proportional to m
2
(Colarusso et al., 1998)The
availability of IR FPA detectors and their recent installation in the exist-
ing SR facilities promises to extend the performance and overcome
many existing limitations (Petibois et al., 2010a). Indeed, advances in
this field are associated to the brilliance of IR Synchrotron Radiation
(IRSR) sources now available all around the world (Marcelli and
Cinque, 2011). A synchrotron radiation source may provide mid-IR
radiation 2–3 orders of magnitude brighter than a conventional Globar
through a small aperture allowing to obtain IR spectra with high signal
to noise ratio (SNR) values and images with a spatial resolution down to
the diffraction limit also in thick samples. Noteworthy, high quality im-
ages with a high lateral resolution, well below the sizes of a cell, can be
obtained with no damage in the sample exposed for a few minutes to
the high brilliance of IR SR light. The maintenance of integrity in
biological samples after IR SR spectroscopy represents an essential con-
dition to achieve complementary information in the sample analyzed
by different methods.
In this contribution, peculiarities of different vibrational spectromi-
croscopy techniques potentially useful for biomedical researches and
diagnostic applications will be illustrated, with particular emphasis on
imaging and time-resolved experimental setups optimized at IR syn-
chrotron facilities.
2. Vibrational techniques
Vibrational spectroscopies study the interaction of light with matter,
i.e., photons that are either absorbed or scattered. Photons of specific
energy are absorbed, and the absorption pattern provides qualitative
and quantitative information on the differentmolecules that are detect-
able in the investigated sample, opposite to the relatively high energy
associated with x-ray photons used to probe matter via multiple scat-
tering processes of excited photoelectrons (Koningsberger and Prins,
1988), vibrational spectroscopies directly probe the concentration and
the of a simple molecule or the functional groups of a large molecule
and its conformation.
Two vibrational techniques are very popular for their applications
in both scientific and industrial environments: FTIR and Raman spec-
troscopy. With no chemical or fluorescent labelling FTIR and Raman
spectroscopy probe molecular rotational or vibrational energy levels
from chemical bonds illuminated by IR or, in the case of Raman spec-
troscopy, near infrared (NIR), visible or ultraviolet (UV) radiation.
The principal difference between these two analytical techniques is
that IR spectroscopy detects vibrations during the changes of electrical
dipole moment, while Raman spectroscopy detects vibrations associat-
ed with the deformation of the electrical molecular polarizability.
The IR region is just a small portion of the electromagnetic spec-
trum. It is usually divided in near, mid and far IR regions.
FTIR spectroscopy is based on the absorption of IR radiation by IR
active molecules that absorb energies within the mid-IR region of the
electromagnetic spectrum (see Fig. 1). In the mid-IR range we mainly
probe two types of molecular vibrations: stretching and bending. In-
deed, as summarized in Fig. 2, absorption bands associated with the
stretching and bending vibrations in molecules, typically occur in
the wavenumber range (the wavenumber is a spatial frequency, pro-
portional to the reciprocal of the wavelength) ~500–4000 cm
–1
(i.e.,
from 2.5 μm to ~20 μm).
In pure molecules as well as in their complex mixtures the detec-
tion of stretching and bending vibrations is a direct consequence of
the interferences (positive or negative) occurring when specific fre-
quencies (or wavelengths) of mid-IR radiation and the natural fre-
quencies at which intra-molecular bonds vibrations resonate.
Indeed, when the wavelength of the radiation has the same energy
or it is close to the energy of a vibrational mode, the radiation is
absorbed and we can observe a decreased intensity in the transmis-
sion spectrum. Only vibrations that result in a change of the dipole
moment of the system are observed in an IR spectrum which com-
poses of a certain number of peaks reflecting either the different vi-
brational modes as well as the number of vibrating components
that may contribute to the intensity value and shape. Strong IR ab-
sorption lines correspond to strong dipole moments and reflect the
quantity of absorbing bonds within a single molecule and/or func-
tional groups characterizing molecules within a sample that are in a
1391A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
resonant vibrational status with specific wavelengths of the mid-IR
radiation.
In a molecular system, the bond axis is usually represented with
an imaginary straight line that connects the nuclei of two atoms
bonded to each other. For a tri-atomic linear molecule (e.g., CO
2
), it
is customary to align the molecule-fixed z-axis with that joining the
different nuclei that compose the molecule. The coordinate of a nor-
mal vibration is a combination of changes in the positions of atoms
within the molecule. For instance atoms within molecules of Fig. 3
rock back and forth within the plane, or bend back and forth outside
the plane, respectively. For a bi-atomic molecule only one normal
mode of vibration can occur, whereas in polyatomic molecules differ-
ent parts will be simultaneously involved with a number of potential
vibrational modes reflecting the complexity of molecule (3N-5 in a
linear molecule and 3N-6 in a non-linear molecule, with N = the
number of atoms, respectively). The absorption of the IR radiation
can be detected at specific wavelengths (or frequencies or wavenum-
bers). Only when stretching or bending result in a change of the molec-
ular dipole moment. A stretching vibration (ν) is a correlated movement
along the bond axis such that the interatomic distance increases or de-
creases. A bending vibration (δ), e.g., scissoring, rocking, wagging, and
twisting, consists of a change in the bond angle between bonds. All
the above vibrational modes are typically observed in the mid-IR region.
However, also the far-IR (25–1000 μm) where intermolecular vibra-
tions occur and the near IR (0.75–2.5 μm) regions shown in Fig. 1 can
be utilized to obtain precious information about biosystems.
According to the above, an IR spectrum reflects the sum of detect-
able vibrations in molecules and of their interactions within the probing
sample. The detection of absorbance peaks at specific wavenumbers in
the IR spectrum of a pure substance may be sufficient to its identifica-
tion. Therefore, the IR spectrum of a molecule is a real fingerprint of
that molecule and the mid-IR spectroscopy is an effective analytical
method to investigate and characterize biosamples (Silverstein et al.,
1991).
It is not necessary to identify all the detectable vibrational modes
in molecules. Rather, detecting the vibrational modes of functional
groups characterizing complex molecules allows to identify and as-
sign the different classes of organic molecules, e.g., amide modes in
proteins. When the probing samples have intrinsic complexity, for in-
stance cells and/or tissue samples whit different molecular and struc-
tural components, the assignment of IR bands to particular molecular
components is challenging. However, some IR bands are particularly
relevant: the amide I and the amide II vibrational bands between
1700 and 1500 cm
–1
, representing the amide backbone of peptides
and proteins. Other modes around 1230, 1400, 1740 and 2900 cm
–1
represent (phospho-)lipids while vibrational components below
1000 cm
–1
and around 1000 cm
–1
may reflect nucleic acids and/or
carbohydrates (Miller and Dumas, 2006; Pistorius, 1995).
Fig. 1. The wide IR range covering more then three orders of magnitude in the unit of
wavelengths (or wavenumbers).
Fig. 2. A typical mid-IR transmission spectrum showing in a schematic way typical ab-
sorptions lines associated to vibrational modes of molecules, organic components and
others contributions of biological and non-biological nature.
Fig. 3. Simple layouts of the vibrational modes associated to a molecular dipole mo-
ment change detectable in an IR absorption spectrum. In addition to the two stretching
modes, the four different bending vibrations are showed.
Fig. 4. A simple layout of light scattering processes with a simple molecule: Rayleigh
scattering (green) and anti-Stokes Raman scattering (red).
1392 A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
Different from the IR spectroscopy, the Raman techniques look at
inelastic scattering processes occurring in the components of matter
probed by a monochromatic laser light. Excitation of light may use
monochromatic NIR, visible or ultraviolet (UV) radiation (Dukor,
2002). When a sample is illuminated with a laser beam, the scattering
is produced by the interaction of the light with the sample compo-
nents and these scattering techniques look at the excitation lines,
i.e., at those wavelengths at higher (or lower) energy of the excitation
light of the laser (see Fig. 4).
Radiation from a laser is collected with a lens and sent through a
monochromator. Wavelengths close to the excitation laser line, due
to the elastic Rayleigh scattering, are filtered out while the rest of
the collected light is analyzed by a spectrometer. The Raman effect
occurs when the light hits a molecule, interacts with the electrons
of its bonds and the incident photon excites one of the electrons in
a virtual energy state higher than that induced by mid-IR radiation
(Fig. 5). In addition to the elastic process (Rayleigh scattering —
Fig. 5B), and taking the original excitation energy as reference,
other two types of processes may occur: the Stokes scattering
(Fig. 5C), when the energy is absorbed and the system relaxes into
a higher vibrational excited state, and the anti-Stokes Raman scatter-
ing (Fig. 5D), when atoms or molecules lose energy during the tran-
sitions from higher to lower vibrational energy states. The Raman
effect is a weak phenomenon, i.e., about one photon over 10
6
is scat-
tered, giving a contribution to the detectable signal. Also Raman
spectroscopy is substantially a non-destructive technique that does
not require any sample preparation. However, because it is based
on the illumination with a laser source, a local heating of the sample
may occur and the possible damage of the organic contents or
changes in the morphology of living systems (Notingher et al.,
2002) are the main drawbacks of the method (Krafft et al., 2007).
The weak sensitivity is an additional limitation for a quantitative analy-
sis of biological systems (Berger et al., 1999). Another strong limitation
is the interference with autofluorescent phenomenon that frequently
occurs in biomaterials and may limit the application of Raman-related
techniques, e.g., in the study of cells and tissues with strong light emis-
sion components. As a consequence, in these systems the Raman signal
can be partially or completely masked by fluorescence processes in-
duced by visible excitation lights. To avoid the fluorescence interfer-
ence, it is possible to use UV radiation at wavelengths b270 nm.
However, because the penetration depth of UV radiation is limited to
a few microns UV Raman methods may probeonly surface layers. More-
over, at UV wavelengths, damage of biosamples may occur also at low
doses (Dukor, 2002).
In addition, due to the weak scattering contribution of water,
Raman has the advantage over FTIR to allow analysis of biofluids or
biosystems in living conditions (Dukor, 2002; Lin et al., 2007).
Summarizing, the vibrational spectroscopies offer several advantages,
although drawbacks exist and have to be considered from time to
time. Moreover, these methods allow to investigate many different
types of samples in many different conditions. Spectra of liquid and
solid samples as pellets, powders and films, emulsions, suspensions,
gases and tissues can be collected by FTIR spectroscopy.
In the last few decades, some technological advances enabled
significant improvements in these spectroscopic methods. In the
1960s Michelson interferometers became available followed
some years later by the implementation of fast Fourier transform
algorithms on personal computer. Now is possible to perform on-
line back Fourier transforms, the indispensable mathematical
tools to immediately transform the original interferogram into
the resulting FTIR spectrum allowing to perform the IR analysis in
a broad IR range. Spectroscopic techniques benefited of the intro-
duction of IR microscopes, new array detectors, optical compo-
nents and the availability of high brilliant synchrotron radiaton IR
sources allowing FTIR spectromicroscopy to become a suitable
method to investigate matter and phenomena at the micrometric
scale as it occurs in individual cells. FPA detectors coupled with
SR sources may also reduce data acquisition time from hours to
minutes allowing to monitor processes and phenomena since now
considered impossible (Petibois et al., 2009).As an example, IR synchro-
tron radiation sources coupled with FTIR microscopy and with a Focal
Plane Array (FPA) detector recently allowed to collect the chemical im-
ages of several individual cells in a few minutes only (Petibois et al.,
2010b). Regarding the Raman technique, at present there is no advan-
tage to use the incoherent SR emission. However, it is necessary to un-
derline here that 4th generation sources like Free Electron Lasers (FELs)
in addition to near field applications (Cricenti et al., 2003; Ortega et al.,
2006) will certainly offer unique new opportunities for Raman experi-
ments and, in particular, for time resolved experiments and imaging
(Adar, 2007; Lee et al., 2007).
3. IR synchrotron radiation sources
The first synchrotron radiation sources dedicated to X-ray experi-
ments date back to 1960s. Indeed, the first pioneering observations of
the synchrotron radiation in the IR region were made at Stoughton
(Stevenson et al., 1973) and at Orsay (Meyer and Lagarde, 1976)in
1970s. However, only at the beginning of the 1980s a port dedicated
to the extraction of InfraRed Synchrotron Radiation (IRSR) was built
on the SRS ring at Daresbury (Yarwood et al., 1984). Later, the first
dedicated IR beamline opened at the NSLS ring of the Brookhaven
National Laboratory in 1987 (Williams, 1990). After this first initia-
tive, a rapid development occurred in USA and in Europe where
beamlines were realized at MAX-lab (Lund), SUPERACO (Orsay), at
SRS (Daresbury) and DAΦNE (Frascati). Nowadays there are more
than twenty beamlines dedicated to IR spectroscopy and microscopy
all over the world where a continuously increasing number of users
apply to perform multidisciplinary studies including researches in
life sciences, biomedicine, biophysics, etc. The most common sources
of IRSR are bending magnets. Due to their considerable source size
and the large intrinsic divergence of the SR in the infrared domain
an optical layout with large optics is required to transfer and focus
the beam from a large port of the storage ring to the entrance pupil
of the interferometer or a spectrometer.
Why has the interest in the IRSR and its unique properties
increased in the last years and what are the effective advantages of
a non-thermal SR source in the IR domain?
Actually, a non-thermal SR source is an intense source whose
emission in the long wavelength region is asymptotic and not
depending on the energy of the electron beam for E>0.5 GeV. Its
intensity is proportional to the large horizontal opening angle and
to the current circulating in the storage ring. As a consequence, at IR
wavelengths almost all storage rings are equivalent and both current
Fig. 5. The energylevel diagram showingthe different processesand the states involvedin
the Raman scattering process: IR absorption (A), Rayleigh scattering (B), Stokes Raman
scattering (C), anti-Stokes Raman scattering (D), resonance Raman scattering (E) and
fluorescence (F). The numbers represent different vibrational levels (ν
vib
) within each
electronic state.
1393A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
and stability, are their qualifying parameters. In the mid-IR range the
gain in brilliance of a IRSR source is at least of two orders of magni-
tude (see Fig. 6). At longer wavelengths, i.e., in the far-IR, they are
the most powerful broadband sources optimized for IR spectromicro-
scopy applications in condensed matter physics, materials science,
biophysics and also biomedicine.
The unique features of IRSR allowing to obtain high quality spectra
and spectral data, push forward the limits of IR spectroscopy opening
new fields of application, in particular those requiring extreme exper-
imental conditions that typically prevent the use of the standard
sources. Experiments running at IRSR beamlines are those most de-
manding such as high-pressure studies, high-resolution micro-
spectroscopy and time resolved experiments (Marcelli and Cinque,
2011). Spectromicroscopy and spectral imaging are two research
areas with most benefits deriving from the availability of IRSR facili-
ties. Indeed, the use of IR microscopes coupled to a SR source give
outstanding results because of the high brilliance of the IRSR sources.
They allow to reduce the aperture slits of the IR microscope and to
collect data in the mid-IR region with good SNR from geometrical
areas of a few microns, e.g. at the diffraction limit (Carr, 2001). In-
deed, the main advantage of FTIR microscopy when using a SR source
is the possibility to achieve the highest spatial resolution at the sam-
ple location, i.e., at the micron scale, allowing a true single cell analy-
sis. The typical spatial resolution performance in the IR is limited to
3×3 μm(Jamin et al., 1998). However, because the size of a non-
thermal SR source is naturally small, the radiation is emitted in a
narrow angular range (~mrad), yielding a high throughput at small
aperture sizes. As a consequence, although IR microscopes have not
been designed for SR sources, owing to the higher brilliance of the
source and the lack of thermal noise, SR based FTIR microscopy sys-
tems leads to extreme performances such as the ultimate lateral spa-
tial resolution in the real diffraction limited regime (Carr, 2001;
Dumas et al., 2008).
An efficient way to take advantage from high brilliance is to com-
bine SR and two-dimensional FPA detectors. With this setup, IR sig-
nals are recorded from many different points at the same time and
several high quality FTIR spectra are obtained within acquisition
time reduced from hours to minutes, as compared with a single ele-
ment detector mapping. Moreover, working with a high current
ring, the intense and brilliant IR beam can be shaped to illuminate
only a limited number of pixels of a FPA detector increasing the SNR
ratio of the image achieved with high contrast (Petibois et al.,
2010b). If the focusing of the IRSR source is also optimized to control
the energy distribution on the FPA area, this combination allows the
investigation of single cells. With a circulating current > 1 A it is in-
deed possible to extract the IR absorption distribution of molecules
within a cell down to 1 ×1 μm pixel resolution (Petibois et al., 2010a).
Finally, because of its wide spectral emission and high brilliance
SR offer unique opportunities to analyze samples of biomedical inter-
est also within the far-IR region, better known as the TeraHertz do-
main (1 THz= 33 cm
−1
) where conventional sources are lacking.
4. Spectral-imaging
Imaging methods developed during last decade may represent the
most significant advances in the application of vibrational spectros-
copies. With the word imaging we refer to the large set of computer
based techniques to visually represent an object for different pur-
poses, from the simple collection of data to the characterization of a
surface or of a material and to the the study of complex dynamic phe-
nomena, to medical diagnostics. When imaging is performed by a spec-
troscopic method, the small differences in the composition typically
associated to small differences among spectra collected in different spa-
tial locations, may significantly enhance experimental capabilities and
easily highlights chemical or molecular components, the morphology
and/or the texture of a sample.
Using IR or Raman microscopes, really small samples can be investi-
gated even at high spectral resolution
1
, with spatial resolutions
approaching or below the diffraction limit. However, working in trans-
mission the practical limit in terms of sample thicknesses is between 5
and 30 μm. In a typical IR transmission experiment, an IR source (both a
conventional or SR source) modulated by an interferometer enters in
the optical system of the microscope, and then is focused to the sample
with a catoptrics condenser (Cassegrain or Schwarzschild), collected by
another catoptrics objective, and imaged onto a focal-plane array.
Coupling a vibrational technique, e.g. IR or Raman, with imaging both
the distribution as well as molecular/chemical information within a
sample surface or a thin slice can be obtained. The technique is chosen
according to the experimental requirements and/or sample characteris-
tics. In materials science but more important, for biological/biomedical
researches, spectral-images can be collected without any sample prep-
aration or labeling. This allows to obtain the distribution of various
molecular/chemical components of the investigated sample with a
high spatial resolution.
As an example, Fig. 7 compares a visible (top left) with a IR
image of a 6 μm thick skin sample. While the contrast of the visible
image is determined only by changes and differences in optical
transmission and scattering properties of the sample, in a IR images
the contrast is obtained with false colors (the red and the blue cor-
responding to the highest and the lowest intensity, respectively) by
selecting a given spectral information (integration of a band or in-
tensity at one wavelength) and show the distribution of selected
molecular components such as proteins and lipids. The top right
image represents the intensity distribution of the amide I band
while the bottom right image represents the intensity distribution
of the amide B band. Both may be associated to the protein distribu-
tion throughout the sample. Finally, the bottom left panel points out
the intensity distribution of the carbonyl stretching vibration corre-
sponding to the lipid distribution (Garidel and Boese, 2007). The
images in Fig. 7 have been recorded at the spectral resolution of 4
or 8 cm
−1
within a few minutes and using a 64x64 elements FPA
detector.
Fig. 6. Calculated SR brilliance of a 3 GeV SR source (red) considering a front end with a
solid angle of 50 × 30 mrad
2
. Comparison is made between a Globar (black) and an
ideal bending magnet SR diffraction limited source (blue).
1
The spectral resolution is a measure of the ability of a spectrometer to resolve fea-
tures in the electromagnetic spectrum and it is usually denoted by Δλ where λis the
wavelength λ.
1394 A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
FTIR spectral imaging techniques maintain all advantages of the
FTIR spectroscopy: energy calibration, high spectral resolution and
throughput. FPA detectors add the multichannel advantage and the
highest time-resolution, that could be overcome only by complex dis-
persive systems coupled to array detectors (Snively et al., 2004). Mor-
ever, FTIR images recorded using a FPA detector of adjacent tissue
sections can also be stitched together to form a single data matrix
for multivariate analysis so that 3D-multivariate imaging provides a
new avenue to study important anatomical and histopathological fea-
tures based on the underlying macromolecular chemistry (Wood et
al., 2006).
Also Raman imaging modalities hold significant potential to study
cellular and molecular processes that may be used in many biomedi-
cal researches. In fact Raman spectroscopy at NIR wavelengths has
been applied to obtain non-invasive molecular imaging of deep tis-
sues in small animals (Keren et al., 2008). In particular, Surface-
Enhanced Raman Scattering (SERS) nanoparticles were used to dem-
onstrate a whole-body Raman imaging with a relatively high signal
reproducibility obtained both in the in vitro and in vivo experiments
of imaging with nanoparticles (Keren et al., 2008). Moreover, a detec-
tion sensitivity b10 pM was observed in a liver tissue using nanopar-
ticles. However, as many other optical methods Raman at NIR
wavelengths is limited by the light penetration to a few centimeters.
This work sets the foundation for future studies although other inves-
tigations with Raman nanoparticles will be necessary to understand
the advantage and the limitations of Raman imaging in biomedical
applications (Tollefson et al., 2010).
Within the space limitations of this contribution it is not possible to
summarizeall available contributions in the field. Actually, several over-
views of the different imaging applications in biology and medicine, in-
cluding early medical diagnostics researches have been published
(Bhargava, 2007; Boskey and Mendelsohn, 2005; Carter et al., 2009;
Holman and Martin, 2006; Levin and Bhargava, 2005; Moreira et al.,
2008; Movasaghi et al., 2007; Petibois and Cestelli Guidi, 2008; Petter
et al., 2009; Srinivasan and Bhargava, 2007; Swain and Stevens, 2007).
In this up to date review in addition to milestone contributions in the
field, we will focus in particular on experiments regarding cells and tis-
sues, pointing out the accessible biochemical information.
5. Biological and biomedical applications
The biological applications of FTIR spectroscopy started with
the analysis of simple biomolecules. Later the availability of
Fourier transform methods running on desktop computers, more
sensitive detectors and microscopy setups, made possible the
investigation of more complex biosystems. The increase of possible
applications deeply transformed the community of biospectrosco-
pists stimulated by modern chemical and statistical approaches
(Ali et al., 2008; Fernandez et al., 2005; Petibois and Deleris,
2006). The latter have been introduced to facilitate the analysis
of spectral data as well as the classification of IR spectra represent-
ing different components in a biosample, for instance different
cells within a tissue biopsy. Due to the intrinsic complexity and
dishomogeneity, many absorption modes from different molecules
overlap making the classical approach useless to perform a strin-
gent molecular analysis in biomedical samples, except for few no-
table cases (Fig. 8).
As discussed in Section 2, FTIR spectroscopy is based upon the ab-
sorption of IR light by vibrational transitions in covalent bonds. The IR
spectrum of complex biological samples contains all the contributions
of different vibrating molecules such as proteins, lipids, nucleic
acids, etc, depending on the natural composition of probed samples
(Petibois, 2001, 2006, 2007). IR analysis is performed to extract infor-
mation on specific molecular components and their changes within
the representative spectra.
Fig. 7. Visible and IR images of a skin sample. In the top left panel the contrast of the visible image is determined by changes and differences in optical transmission and scattering
properties of the sample. In the top right image the distribution of the amide I integral band intensity while the bottom right panel shows the distribution of the amide B integral
band intensity. The bottom left IR image represents the distribution of the carbonyl stretching vibration integral band intensity. [Adapted from (Garidel and Boese, 2007)].
1395A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
The first medical application of IR radiation is probably the pio-
neering work of Delpy and colleagues at University College London
(Cope and Delpy, 1988). Their work monitored the oxygenation sta-
tus of brains of premature babies using NIR radiation. The first real
medical application of FTIR spectroscopy came a few years later
when this technique was applied to detect and identify urinary
crystals (Daudon et al., 1991). This work showed that the visible
microscopy was not sufficient. On the contrary, this was achievable
by the application of FTIR microscopy technique which provided
unique information for understanding the etiology.
Mid-IR FTIR microscopy is therefore a suitable and objective ana-
lytical method to investigate the chemical composition of urinary
crystals that can be distinguished from insoluble drug metabolites
also excreted within urine (see Fig. 9).
Fast acquisition capability, high analytical sensitivity and
specificity, high discrimination power and spatial resolution of the
vibrational spectral microscopy has greatly expanded the number
of applications. The above discussed peculiarities of SR IR sources,
for instance allowing to obtain high resolution FTIR images and to
study molecular composition in individual cells (Jamin et al., 1998),
at the end of 1990s suggested the opportunity to develop IR SR
beamlines optimized also for biomedical researches (Marcelli and
Cinque, 2011).
At present, in several synchrotron facilities worldwide dedicated
instrumentations are available taking the benefit of the impressive
performances of synchrotron radiation in the IR domain. Many IR mi-
croscopes are available in SR facilities all around the world and vibra-
tional spectroscopic mapping and imaging of cells and tissues have
become common research methods within life sciences and biomed-
ical researches (Aitken et al., 2010; Eichert et al., 2007; Japing et al.,
2008; Miller et al., 2007; Steiner and Koch, 2009). These benefitof
the large amount of qualitative, quantitative and structural informa-
tion contained in a vibrational spectrum. Therefore, in the last decade
we observed a continuous increase in the number of different FTIR
spectromicroscopy applications, and in particular cells and tissues in-
vestigations. In the near future, we expect that these techniques, in
addition to fluorescence and electron microscopy, will become funda-
mental methods not only for chemical analysis, but also for the mor-
phological and dynamical characterization of biological systems at the
subcellular level, as well as for for the high throughput chemometric
classification of different components within a tissue. Indeed, the ac-
curate differentiation of different components within a biological
sample via chemometrics methods relies to subtle but multiple
changes of bands or spectral structures that are objectively compared,
not just to intensity variations that may be affected by experimental
conditions or systematic errors (Geladi, 2003; Geladi et al., 2004;
Levin and Bhargava, 2005). From Fig. 8, it is clear that the IR absorp-
tion bands of a small molecule, e.g., urea, can be individually de-
scribed. On the contrary in the case of a protein, e.g., albumin, bands
such as the amide I are a combination of many bands and give an
overall response of the system. Similarly, observing the spectrum of
a tissue, the IR bands are representative of the overall absorption of
the sample. The first issue determining the overlap of IR bands is
Fig. 9. (Top) An unusual structure observed in the urine and (bottom) its FTIR s pec-
trum in transmittance that points out the presence of the glaphenic acid. [from
(Daudon et al., 1991)].
Fig. 8. Characteristic FTIR spectra obtained from biosamples. (Top) small biomolecule,
urea, with its main IR absorption bands; (Center) a large molecule, albumin, with
major global absorption bands of its organic functions; (Bottom) a typical complex bio-
sample, brain tissue, with its main global IR absorption bands belonging to the main
molecular families.
1396 A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
the shape, a combination of Gaussian and Lorentzian functions that
broadens all features. Typically, the width of mid-IR bands of biomol-
ecules ranges between 20 and 50 cm
−1
. Considering that FTIR spec-
troscopy probes functional groups of molecules through their
absorption bands, there is no way to discriminate two specimens
characterized by a large number of molecular groups, such as pro-
teins. Actually, the secondary structure conformation of proteins leads
to the formation of eight IR bands in the amide I spectral range
(1700–1600 cm
−1
). However, under appropriate assumptions these
eight components can be extracted with a curve-fit of experimental
data. Nevertheless, the energy among bands does not change enough
to allow discrimination between two proteins in a mixture. The same
phenomenon occurs for the fatty acyl chains absorptions in the
3050–2800 cm
−1
spectral range. The different groups [νC-(=CH),
ν
as
(CH
3
), ν
as
(CH
2
), ν
s
(CH
3
), ν
s
(CH
2
)] fall in the same frequency range
whatever is the fatty acyl chain length. The same holds true for biolog-
ical systems, that are apparently simpler, such as cells. In the case of the
analysis of a cell and its subcellular components, e.g., mitochondria, we
need also a high spatial resolution. Cell analysis would typically require
a1×1μm (or less) spatial resolution, high brilliance SR sources already
allow observation of sub-cellular components (nucleus).
Regarding tissue imaging, the high brilliance of the SR sources is
certainly useful for the analysis by IR methods of hard tissue sections,
which require high photon flux to perform transmission experiments
on sample with a thickness >5 μm(Burghardt et al., 2007). Moreover,
mid-IR microscopy is an analytical technique that may reach the
nanomolar (nM) sensitivity (Petibois et al., 2006a,b) and because of
its non-damaging nature it may be combined with other well
known analytical techniques, such as PIXE or X-ray fluorescence im-
aging (Miller and Tague, 2002).
Actually, correlative imaging is widely used in clinical practice, e.g., for
unravelling abnormal anatomical structure in organs. However, it is
well recognized that manypathological situationscannot be recognized
only on the basis of anatomical disorders or of their diffuse shape that
makes their interpretation very difficult and possibly hazardous. To di-
agnose myopathies, diffuse tumours, autoimmune disorders, brain de-
generative diseases, blood cell dysfunctions…etc. the support of other
analytical methods is clearly needed. In the next years, molecular imag-
ing such IR imaging will certainly increase its role in the identification
and the treatment of diseases. Actually, structural changes may only
point out the presence of a disease, while functional changes are clear
indicators, even at early stages of different pathologies such as epilepsy
(Chwiej et al., 2010) or rheumatoid arthritis (Croxford et al., 2011).
A recent study of the villoglandular adenocarcinoma from a cervi-
cal biopsy has been used as a model system to demonstrate diagnostic
efficacy of 3D multivariate FTIR images obtained in several adjacent
tissue sections. An unsupervised Hierarchical Cluster Analysis
(UHCA) was performed simultaneously on 4 sections stacked togeth-
er and interpolated by software. The resultant 3D-images can be ro-
tated in 3D, sliced and made semitransparent to recognize its
internal structure, i.e., anatomical and histopathological features in-
cluding connective tissue, red blood cells, inflammatory exudate and
glandular cells that could be also correlated with standard stained
sections (Wood et al., 2006). The possibility to manipulate 3D images
can be extremely useful in determining the extent and penetration of
a disease or of an inflammatory tissue. This multivariate spectroscopic
3D imaging method can be also adapted for many other techniques
including Raman imaging.
Hybrid or multimodal imaging techniques may also provide a better
and intuitive integration of information combining structural imaging
techniques and also functional and/or molecular methods. Multi-
modality is indeed an important issue of the future of all vibrational
imaging methods not only in the biological and biomedical areas
(Aitken et al., 2010; Xu et al., 2011).
In addition to multimodality, the next frontier of imaging is the visu-
alization of living systems in in vivo condition. Because IR is a non ioniz-
ing radiation, IR imaging with a sub-wavelength resolution is a
particularly promising method. Imaging can be obtained with an ATR
crystal or with an apertureless”scanning near-field optical microscope,
leading to an extreme spatial resolution at IR wavelengths (Bachelot et
al., 1995; Knoll and Keilmann, 1999; Ortega et al., 2006). However, op-
tical methods measure the index of refraction of a material and while
the real part may give the sample topography, its imaginary part is pro-
portional to the sample absorption. As a consequence when studying
extremely small samples or look at small areas the local absorption is
weak and the signal is too small to be detected or at most is comparable
to the noise. In the latter case, differentialmethods suchas photoacustic
or photothermal detectionare much more powerful. One of thesetech-
niques based on the photothermal effect was recently demonstrated at
the CLIO free-electron laser facility (Dazzi et al., 2007a,b; Ortega et al.,
2006). This new infrared spectromicroscopy technique, called Photo-
Thermal induced Resonance (PTIR) (Bachelot et al., 1995)isbasedon
the coupling between a tunable free-electron IR laser and an Atomic
Force Microscope in the IR range (AFMIR) (Dazzi et al., 2005, 2007a,
b). The use of short pulse lasers allows IR chemical mapping at the nm
scale. Indeed, this technique probes the local thermal expansion of the
Fig. 10. Images of two Candida albicans fungi with different morphologies: a round one called blastospore (right) and a long one called hyphae (left) immersed in water. Photother-
mal induced resonance data are taken by irradiating at 1080 cm
−1
at 31.6 kHz.
1397A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
sample, irradiated at the wavelength of specificabsorptionbands.De-
tection is performed directly by the AFM tip in contact mode, probing
the thermal dilatation of the absorbing part of the sample rather than
the transmitted electric field. As the duration of expansion and relaxa-
tion of the sample is always shorter than the response time of the can-
tilever in contact, by recording the amplitude of the cantilever
oscillations it is possible to measure the corresponding IR absorption
as a function either of space or wavelength. A spatial resolution around
50 nm has been achieved in the mid-IR region at λ=22 μm(Houel et
al., 2009).
This AFM-IR technique has been validated by the comparison of the IR
spectrum of a single E.coli bacterium with the corresponding FTIR spec-
trum and by chemical mappings with a sub-wavelength spatial resolution
as shown in Fig. 10 (Mayetetal.,2008). It may recognize internal struc-
ture of bacteria and map for example the distribution of viruses and in
the near future, the kinetics of synthesis of molecular components such
as polymers in bacteria (Dazzi et al., 2008; Mayet et al., 2010). Because
of the inherent near-field character, the sensitivity of this technique is re-
duced for objects located below the surface. However, recent experi-
ments showed the possibility to map also an exogeneous compound
in single human cells and in the cell nucleus in a non-destructive way
and without the use of nucleus trackers (Policar et al., 2011).
In addition to the spatial resolution, the great advantage of this
spectroscopy method is that it may work in liquid and at a constant
temperature offering the opportunity to overcome one of the main
demands of biological researches: the characterization of living sam-
ples in their natural environment.
As outlined above FTIR spectroscopic imaging with ATR (Attenuated
Total Reflection) crystals is another powerful tool for studying biomedical
materials. One of the key advantages of ATR-FTIR imaging is that it re-
quires minimal or no sample preparation prior to spectral measurements
(Kazarian and Chan, 2006). Many applications have also demonstrated
the power and the applicability of this imaging method with FPA detec-
tors. ATR-FT-IR imaging using 2D detectors allows to image static samples
but also the analysis of processes in real time or to measure many samples
simultaneously. ATR then represents a useful complementary approach to
the use of FTIR imaging in transmission or reflection modes. Biomedical
and pharmaceutical fields will certainly have benefits from this imaging
methodology (Martin et al., 2010; McAuley et al., 2010). Opportunities
also exist for chemical imaging of dynamic aqueous systems, such as
microfluidics, or for imaging of dynamic processes in live cells.
The exciting possibilities of such dynamic imaging will see further
applications to complex chemical systems that change with time. Op-
portunities exist for further developments of ATR imaging approaches
in either the micro and macro modes by using various internal reflec-
tion elements of different sizes and shapes and in different optical
configurations, in combination with focusing or expanding beam op-
tics (Kazarian and Chan, 2010).
5.1. Sample preparation
IR spectroscopy and microscopy do not require an extended sam-
ple preparation. Nevertheless, some contributions to overall analyti-
cal variability can derive from the application of validated protocols
for sample preparation.
5.1.1. Tissue sections
Tissue samples are usually analyzed by two distinct approaches. To
obtain rapid information on the absence/presence/extension of a disease
during surgery, 10 –15 µm thick slices obtained with a cryomicrotome
from frozen tissue biopsies are analyzed by a trained pathologist. To
obtain diagnostic and prognostic informations as well as therapeutic
indications several slices of formalin-fixed and paraffin-embedded tissues
are subsequently analyzed by different techniques in a histopathology
laboratory. At least two adjacent sections are used (Fig. 11): the sample
stained with Hematoxylin and Eosin (H&E) is analyzed using a light mi-
croscope and the unstained sample by IR microspectroscopy, respectively.
In dependence of the purpose the second tissue slice is directly deposited
on IR-transparent windows for transmission measurements, which give
access to quantitative analysis of the tissue since thickness is quite regular
within all over the surface. Tissue samples can be also deposited on glass
coated or not with gold, or KBr pellet, or thin (usually 0.9 μmor1.5μm
thick) Mylar foil spanned over a frame and let to dry at room temperature.
Samples at low temperature are kept in place without any fixing agent
due to adhesive forces (Kwiatek et al., 2009). FTIR spectromicroscopy
has been successfully applied to analyze formalin-fixed and paraffin-
embedded samples and to identify characteristic alterations associated
with the presence of premalignancy in the esophagus of patients with
Barret's disease (Quaroni and Casson, 2009).
5.1.2. Cell culture
Cells growing in adherent monolayer are detached form the plas-
tic flask by the treatment with trypsin/EDTA or with a scraper. In the
first case after a few minutes incubation the action of enzymes is
stopped with the addition of a complete medium containing trypsin
inhibitors in the fetal bovine serum fraction. Cells are centrifuged
and then resuspended in phosphate buffered saline or directly fixed
in 1–4% buffered-formalin for at least 30 minutes. Cells growing in
suspension can be directly fixed in buffered-formalin. The excess is
then eliminated washing formalin-fixed cells in pure water just be-
fore to prepare analytical samples (Bellisola et al., 2010).
On the contrary, cells grown on biocompatible materials such as
CaF
2
, non doped silicon, MirrIR, Mylar, etc...are washed with PBS,
fixed with ethanol, or buffered-paraformaldehyde or glutharaldehyde
for a certain period. Fixative is then removed by washing samples
with water (Gazi et al., 2005). Fig. 12 shows PC-3 cells cultured on a
Mylar foil (Podgórczyk et al., 2009).
Fig. 11. An example of histological view of two adjacent prostate tissue sections: A) unstained and B) stained with H&E (Podgórczyk et al., 2009).
1398 A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
Also live, unfixed cells have been investigated by FTIR spectroscopy
and microscopy. Early investigations carried out with SR IR spectromi-
croscopy (Holman et al., 2002) on living cells demonstrated that high
brilliance mid-IR synchrotron beam does not affect both the short-
and long-term viability, proliferation, and metabolism of human cells.
Recent studies have been also performed selecting dedicated cell culture
conditions for IR imaging. Cells have been grown directly on IR-
transparent windows with a final cryofixation. This dedicated procedure
for IR imaging can be performed without modifying cell morphology and
molecular composition (see next section) as could occur for proteins and
phospholipids in samples fixed with paraformaldehyde (Petibois et al.,
2010b).
5.2. Is IR spectromicroscopy of living cells possible?
Most analytical setup using synchrotron radiaton sources and in
particular those based on ionizing radiation do not consider analysis
of living specimens due to the instantaneous or integrated dose re-
leased to the sample. In fact, many aspects of the molecular dynamics
are determined in biological systems maintained out of their natural
environment. For instance, the ex vivo time-resolved x-ray analysis
of protein folding/conformation is carried out within the ms to the
ns scale in order to reduce the damage to the specimen. Another
relevant issue to consider designing practical imaging experiments in
biological systems is the analytical sensitivity. In this area, it is clear
that synchrotron radiation sources or FELs will play the major role
to push forward the analytical properties of the imaging modalities
that may be implemented. The high energy and brilliance of synchro-
tron radiation offer a unique opportunity for instance to perform the
imaging of some trace elements having natural concentration levels in
samples within the range 10
–9
–10
–12
moles which are below the
minimum detection limit of most methods using conventional sources.
Actually, as showed in Fig. 13 FTIR microscopy with a FPA detector
may now allows routine chemical imaging on individual cryofixed
cells in a few minutes only. The mid-IR energy range contains bands
due to vibrations of the major classes of biological molecules, such
as carbohydrates, organophosphates, phospholipids and proteins,
and provides considerable information about the structures and rela-
tive concentrations of these molecules within a cell. False color im-
ages and/or maps of the content of biological molecules can be
generated by integrating the intensity or area of one or more of
these bands, thereby illustrating the distribution of simple molecules
and/or biochemicals. However, cellular materials are extremely spa-
tially inhomogeneous on a micrometre scale in terms of physical mor-
phology and chemical content. This inhomogeneity is the origin of the
elastic scattering of the incident radiation that determines the Mie scat-
tering: a broad oscillating background in the IR spectra (Romeo and
Diem, 2005) Actually spectral distortions observable in FTIR spectra in
Fig. 13. FTIR imaging of a single cryofixed cell grown on a Si–O substrate at SINBAD. Background and sample images were collected at same nominal current values each. Spectral
image acquisitions have been performed with 32 × 32 pixels of the FPA detector. From left to right, images have been obtained integrating the spectra in the protein
(1715–1600 cm
−1
), lipids (3020–2880 cm
−1
) and polysaccarides (1152–951 cm
−1
) regions, respectively.
Fig. 12. An example of cultured prostate cancercells —line PC-3 (Podgórczyket al., 2009).
1399A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
spatially inhomogeneous samples can be fixed obtaining subcellular in-
formation free from artefacts (Bassan et al., 2009). Moreover, the bril-
liance of SR IR sources may enhance these small molecular signals
originating from small areas containing reduced amount of organic
matter. In particular, synchrotron radiation imaging provides the possi-
bility to perform a qualitative and quantitative analysis at a molecular
level and new opportunities to better understand cell interactions, cell
micro-environment, tissue formation and possibly diagnostic.
6. Scanning Near-field Optical Microscopy (SNOM)
The recent advent of Scanning Near-field Optical Microscopy
(SNOM) methods increased at a microscopic level the usefulness of
the optical spectroscopy. Moreover, two-dimensional imaging of
chemical constituents at high spatial resolution makes this a very at-
tractive and almost unique approach (Brehm et al., 2006). In fact,
practical implementations of subcellular IR imaging have been always
problematic because the conventional IR microscopy is diffraction
limited with a spatial resolution of a few micrometers in the mid-IR
range (Dazzi et al., 2006). A SNOM allows measuring the local reflec-
tivity when the sample is illuminated by an external source either di-
rectly or through a fiber: in both cases the technique picks up the
light reflected by the surface.
A major requirement of any SNOM apparatus is an intense and
tunable photon source. Tunability is mandatory to cover the relevant
absorption bands that are present in the investigated sample. Intensi-
ty is also a critical issue because of the feeble light transmission of the
typical narrow fiber tips. For these layouts Synchrotron Radiation
(SR) and even better, FELs are suitable sources because of their
unique combination of extreme intensity and brilliance and the pos-
sibility to change the excitation energy.
The possibility to tune the wavelength (λ) of the source allows ob-
serving the presence of topographic, chemical or electronic defects in
an optical image with a lateral resolution actually limited by the size
of the fiber-tip aperture (~50 nm), and a vertical resolution that de-
pends both on the light penetration depth and on the matrix element
that governs the interaction between light and sample. Considering a
penetration depth proportional to λ, we can suppose that the ob-
served defect in an optical image is at a distance bλfrom the sample
surface. This kind of investigation turns out to be very important for
the characterization of defects trapped in a layer below a transparent
coating (e.g., films), not accessible to a direct topographic analysis
and/or to the detection and analysis of toxic compounds inside a bio-
logical sample.
A useful example is showed in Fig. 14 where images of human skin
keratinocytes have been collected at three different wavelengths
λ=3.4, 6.45 and 8.05 μm, respectively. They correspond to the exci-
tations of different bonds present inside the cells. The same approach
can be further used to detect chemical modifications induced by an
external parameter or stress. The experiment was performed with a
multi-technique SNOM module (Cricenti et al., 1988) that can collect
shear-force (topographic) images as well as reflectivity SNOM im-
ages. The direct comparison of both topographic and spectroscopic
SNOM images is essential to prove the resolution of these images. Ex-
periments were performed with the Vanderbilt (USA) FEL as the ex-
ternal source and the emitted IR radiation was continuously tunable
over the 2–10 μm wavelength range, extended down to 1 μm via
second-harmonic generation, with a high output power and bright-
ness. For these experiments IR SNOM optical probes were fabricated
starting from single mode arsenic sulfide fibers(withanexternal
diameter in the range 80–140 μmandacoreof~10μm) manufactured
and tested atthe Naval Research Laboratory (Washington) while the tip
apex was further metalized by side Au evaporation (Talley et al., 2000).
Fig. 14. (Panel A) Shear-force topography image of human skin keratinocytes cells; (Panel B, C and D) IR-SNOM optical reflectivity images taken at the wavelengths of 3.4, 6.45 and
8.05 μm, respectively. All images are unfiltered and only a rigid plane subtraction has been applied. The area of each image is 30 × 30 μm
2
. Brighter areas correspond to higher to-
pography values or a higher current in the optical image.
1400 A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
In a SNOM experiment light from an unfocussed FEL beam or from
another IR sources are directed onto the sample surface while the
reflected photons are detected through a narrow-point optical fiber
tip mounted on a SNOM module that may also collect shear-force images.
For the images showed in Fig. 14, the total radiation power of the FEL is
~100 μJ and at this level no detectable sample damage has been ob-
served during experiments.
In the topography image in Fig. 14 (panel A) the dark areas correspond
to local depressions while in the SNOM images (panels B, C and D) they
correspond to stronger absorption. We may point out that the shear-
force topography image does not change with wavelength, but due to
the fact that the different components are differently distributed inside
the cells, SNOM images are sensitive to the excitation wavelength. The
mechanism ruled out the presence of topographical artifacts in optical
images taken at IR wavelengths since the latter should be visible at
any wavelengths and be independent by the optical absorption. In addi-
tion, because the cell size is comparable to the wavelength this image
points out that small changes in the wavelength should not affect the
optical image if absorption contributions are negligible. Images of dif-
ferent samples at different wavelengths support the claiming that the
optical resolution of SNOM experiments is not due to topographic arti-
facts (Cricenti et al., 2003). By comparison of topographic and optical
images it is also possible recognize that the dark areas in panels B, C
and D in Fig. 14 where the absorption is stronger, are not correlated to
topographic structures of the cell shown in the panel A. Reflectivity
SNOM images of all cells revealed features not detectable in the corre-
sponding shear-force (topology) images, due to localized chemical
changes at the sub-cellular level. This result confirms the reliability of
IR-SNOM spectroscopy in order to detect biochemical variation,
brought to light by the distribution of functional groups at nanometer
resolution. In fact, for these images the upper limit of the lateral resolu-
tion can be estimated by looking at the edge slope of the smallest features
detected in different SNOM images. Typical values for the lateral resolu-
tion were ~80 nm for topographic images and ~100 nm for reflectivity
images.
7. THz imaging
Looking at Fig. 1 the far IR range smoothly merge in the THz do-
main. THz radiation lies between the IR and microwave regions of
the electromagnetic spectrum and can be used to excite large ampli-
tude vibrational modes of molecules. In this interval of frequencies
radiation is non-ionizing and it is of particular interest as it excites
intermolecular interactions, for example the librational and vibra-
tional modes in liquid water or probes weak interactions between
molecules (Zelsmann, 1995). In the last years the interest in science
using radiation in the THz region is steadily increasing (Siegel,
2004). In fact the availability of brilliant and fast sources in this low
energy domain such as synchrotron radiation and FELs is very impor-
tant for studying collective excitations in solids, superconductor band
gaps, protein conformational dynamics etc. (see http://thznetwork.
org). Why SR can be important also for experiments in the THz
range? Actually, radiation emitted by energetic bunches in a storage
ring is incoherent, however bunches should also produce a roughly
comparable power output of coherent radio-frequency radiation
(Williams, 2006). In 1989 the first observation of coherent synchro-
tron radiation in the wavelength of 4.5–25 cm
−1
has been reported
(Nakazato et al., 1989). Multiparticle coherent emission may also
occur in the very far IR-THz domain from bunched electrons associated
to the presence of a density modulation within the bunch (Carr et al.,
2001). At BESSY Il in Berlin, the coherent emission was experimentally
verified and a stable emission was produced using a dedicated mode of
operation (Sannibale et al., 2004).
Important studies using THz radiation are spectroscopy experi-
ments, performed in particular on materials of biomedical interest
(Sherwin et al., 2004) and in the field of bio-medical imaging
(Tonouchi, 2007).
As previously mentioned, radiation in the mid-IR is particularly
useful to explore the molecular composition and to study secondary
structures in proteins. This low frequency domain includes the im-
portant contributions from metal-ligand vibrations, protonation de-
pendent vibrations of aromatic rings as well as signals from
hydrogen bonding and collective modes from the overall protein
structure. Although vibrational spectroscopy is a suitable method to
characterize protein structures such as heme centers due to its sensi-
tivity to the changes in oxidation state, coordination number, spin
state and ligand binding, the low frequency contribution of proteins
in the IR domain is not yet well recognized. Moreover, while Raman
spectroscopy plays a key role in the characterization of hemoproteins,
studies in the far-IR and THz are still limited. A recent IR absorption
study of the low-frequency vibrational modes of the porphyrin-
iron-imidazole bonding identified for the first time the metal-ligand
vibrations by isotope labeling of the imidazole ligand. However, for
its sensitivity to anharmonic couplings and to differentiate the differ-
ent motions temperature-dependent spectroscopy was used to char-
acterize these low frequency modes. The latter were found to be
Fig. 15. (Panel A) A thick human tooth slab with a simulated buried caries lesion on the left side. Confocal image (Panel B) and near-field image collected at 0.3 THz (Panel C) of the
tooth sample in the photograph.
1401A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
temperature-dependent of the ring with higher symmetry and
stronger bonds associated to the low temperature form (Dörr et al.,
2007).
Given the intrinsic long wavelength of the THz radiation (n.b., 1
THz is equivalent to 33 cm
−1
or 300 mm) the diffraction plays a
major role and the optical detection method intrinsically limits the
use to sub-mm type of spectroscopy. On the other hand, using new
detection systems such as near field or a non-optical methods such
as photothermal techniques, the spatial resolution limitation above
mentioned can be overcome. The result can be obtained using bril-
liant IR sources such as SR especially in the coherent emission
mode. For example in Fig. 15 a test of feasibility of the intense THz ra-
diation for the diagnosis of tooth decay both via X-ray radiography
and far-IR in different mode are shown for comparison. In contrast
to the confocal imaging investigation both the near-field technique
and the sub-THz radiation have been used to image bulky tooth sam-
ples. by Drilling a cavity of ~1 mm of diameter in the proximal region
of the tooth filled with hydroxyapatite simulated a buried dental cair
lesion. The near-field image was obtained utilizing a 200-μm wire
cone while the confocal imaging was performed with the same opti-
cal set-up without the near-field cone (Schade et al., 2005). In the
confocal imaging geometry the tooth cannot be spatially resolved
and the image is strongly blurred as one would expect from diffrac-
tion due to the long wavelengths involved.
Another area of application is the investigation of tissues. THz
wavelengths are significantly larger than the scattering structures in
a tissue, and thus the scattering effects of THz radiation should be
considerably reduced in comparison with other optical techniques
working with shorter wavelengths such as visible radiation. More-
over, diseased tissues show a change in the properties compared
with normal tissue and regions where a disease is identified in the
image correlate well with histology. Quite recent studies of basal
cell carcinomas pointed out changes in the THz properties when com-
pared with normal tissue with similar levels of contrast in images ac-
quired in vivo and ex vivo. The contrast associated to these images was
sufficientto identify tumor margins when compared with the histology
(Wallace et al., 2004).
In another study using THz, it has been successfully measured the
refractive index and the absorption coefficient spectra of healthy and
diseased breast tissue from human patients. Comparisons of the dif-
ferences between spectra of different tissue types showed that the
contrast seen in the images can be associated to an increase of the re-
fractive index between cancer and healthy tissues and also in part to
an increase in the absorption coefficient. Although a clear explanation
is not yet available, experimental results indicate that healthy adipose
breast tissue, healthy fibrous breast tissue and breast cancer can be
distinguished looking at the differences in the fundamental optical
properties in the THz region (Ashworth et al., 2009).
8. Conclusions
In spite of the reduced interest for the application of vibrational
spectroscopy in some research areas, in principle there is a growing
interest and a general enthusiasm about the possibility to apply vibra-
tional microspectroscopy to biomedical researches. Relevant develop-
ments have certainly occurred in this field during the last decade.
However, no pre-clinical or clinical trials of FTIR spectromicroscopy
applications have been performed to date. As recently suggested,
this poor consideration of IR techniques most probably depends by
the limited background of clinicians and biomedical researchers
with concepts and methods of IR spectromicrosopy as well as on
the delay to translate the results of researches with SR IR sources in
validated applications easy to use in conventional biomedical labora-
tories and for clinical diagnoses/prognoses (Bellisola and Sorio, 2012).
In fact vibrational spectroscopies are considered non-destructive
photonic techniques capable to provide an analytical and rapid
measure of the sample chemistry in a variety of systems of great in-
terest in different fields of the life sciences. In particular, IR microsco-
py techniques for the mapping and imaging of cells and tissues are
undergoing a rapid expansion, in particular sampling procedures, de-
tection methods and sources. The results obtained from IR based
techniques are providing many new insights into biochemical archi-
tectures and processes, and probably in the next years will have a
significant impact on the development of new treatments and
diagnostics.
Thanks to the underlying macromolecular chemistry multivariate
and multimodal IR imaging methods may provide a new avenue for
important biological and biomedical researches (Amigo, 2010). In-
deed, the coupling of vibrational spectroscopy with multivariate pro-
cessing greatly extends the capabilities of this technology. From a
biomedical perspective existing pathological and histochemical pro-
tocols depend on sample morphology and visualization. The ability
to extend the information in 3D maintaining the spatial integrity
and with precise spectroscopic data represents an ideal combination
for many foreseen applications in biology and medicine.
The future availability of hyperspectral libraries enabling the char-
acterization of unknown constituents is an important development
expected in this area. In this respect, a high current SR ring is manda-
tory to collect FTIR images of biosamples with a high contrast within a
few seconds allowing to study living systems such as single cell.
Finally, 2D detectors that have been already successfully used to
extract spatially resolved chemical information from static and
dynamic systems exploring relatively slow phenomena (ranging
from seconds to minutes), thanks to the development of microfluidic
systems can be applied to investigate much faster processes.
Microfluidic systems are devices allowing the control and the
manipulation of fluids geometrically constrained in small volumes
(sub-millimeter range or smaller). There are many areas that can
take benefits from the use of miniaturized devices and many advan-
tages, e.g., decreased cost in manufacture, use, and disposal; faster
time of analysis; reduced consumption of reagents and analytes;
reduced production of potentially harmful by-products; etc. Micro-
fluidic devices have been already tested using FTIR imaging systems
and allowed to improve their typical slow acquisition time (~sec). At
present, these devices promise to be a reliable way for faster and
simultaneous monitoring of multiple analytes. Moreover, complex
physicochemical processes can be more easily controlled and
investigated using micrometer or submicrometer size devices, e.g.,
working with small fluid/liquid volumes. Analysis of biomolecules,
DNA, proteins or bacteria is one of the area that stimulated the
development of microfluidic devices looking at the possibility to
investigate their fast dynamics with a suitable time resolution.
(Holman et al., 2009). The possibility to combine these devices
with brilliant SR sources may additionally increase the image acqui-
sition rate and broadens significantly the number of microfluidic ap-
plications (Chan et al., 2011).
Within this framework there is really a brilliant future for SR IR
microscopy and imaging, and other extraordinary results in life-
science applications are expected in the next years in particular in
the imaging of living systems.
Acknowledgments
This contribution contains information, experience and practices
accumulated in the last two decades running experiments in many
different SR facilities all around the world. It is the results of many
discussions with colleagues and friends. We should then acknowl-
edge too many persons for the discussions we had in these years on
this subject. It is really impossible to make a list without forget some-
one. A special thank you is however deserved to the entire staff of the
LNF laboratory where some of the experiments discussed in this re-
view have been performed.
1402 A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
There are certainly many researches and relevant manuscripts we
did not mention. We apologize with all that have not been cited and
beg all to forgive us.
References
Abney W, Festing ER. On the influence of the atomic grouping in the molecules of or-
ganic bodies on their absorption in the infra-red region of the spectrum. Philos
Trans R Soc London 1881;172:887–918.
Aitken JB, Carter EA, Eastgate H, Hackett MJ, Harris HH, Levina A, et al. Biomedical ap-
plications of X-ray absorption and vibrational spectroscopic microscopies in
obtaining structural information from complex systems. Radiat Phys Chem
2010;79:176–84.
Ali K, Lu Y, Christensen C, May T, Hyett C, Griebel T, et al. Fourier transform infrared
spectromicroscopy and hierarchical cluster analysis of human meningiomas. Int J
Mol Med 2008;21:297–301.
Amigo JM. Practical issues of hyperspectral imaging analysis of solid dosage forms. Anal
Bioanal Chem 2010;398:93-109.
Ashworth PC, Pickwell-MacPherson E, Provenzano E, Pinder SE, Purushotham AD,
Pepper M, et al. Terahertz pulsed spectroscopy of freshly excised human breast
cancer. Opt Express 2009;17:12444–54.
Bachelot R, Gleyzes P, Boccara AC. Near-field optical microscope based on local pertur-
bation of a diffraction spot. Opt Lett 1995;20:1924–6.
Barer R. The reflecting microscope in anatomical research. J Anat 1949;83:73–4.
Bassan P, Byrne HJ, Lee J, Bonnier F, Clarke C, Dumas P, et al. Reflection contributions to
the dispersion artefact in FTIR spectra of single biological cells. Analyst 2009;134:
1171–5.
Bellisola G, Sorio M. Infrared spectroscopy and microscopy in cancer research and diag-
nosis. Am J Cancer Res 2012;2:1-21.
Bellisola G, Della Peruta M, Vezzalini M, Moratti E, Vaccari L, Birarda G, et al. Tracking
InfraRed signatures of drugs in cancer cells by Fourier Transform microspectro-
scopy Analyst 2010;135:86-3077.
Berger AJ, Koo TW, Itzkan I, Horowitz G, Feld MS. Multicomponent blood analysis by
near-infrared Raman spectroscopy. Appl Opt 1999;38:2916–26.
Bhargava R. Towards a practical Fourier transform infrared chemical imaging protocol
for cancer histopathology. Anal Bioanal Chem 2007;389:1155–69.
Blout ER, Bird GR, Grey DS. Infrared microspectroscopy. Anal Chem 1949;21:1585.
Boskey A, Mendelsohn R. Infrared analysis of bone in health and disease. J Biomed Opt
2005;10:031102–6.
Brehm M, Taubner T, Hillenbrand R, Keilmann F. Infrared Spectroscopic Mapping of
Single Nanoparticles and Viruses at Nanoscale Resolution. Nano Lett 2006;6:
1307–10.
Burghardt AJ, Wang Y, Elalieh H, Thibault X, Bikle D, Peyrin F, et al. Evaluation of fetal
bone structure and mineralization in IGF-I deficient mice using synchrotron radia-
tion microtomography and Fourier transform infrared spectroscopy. Bone
2007;40:160–8.
Carr GL. Resolution limits for infrared microspectroscopy explored with synchrotron
radiation. Rev Sci Instrum 2001;72:1613–9.
Carr GL, Kramer SL, Murphy JB, Lobo RPSM, Tanner DB. Observation of coherent syn-
chrotron radiation from the NSLS VUV ring. Nucl Instrum Methods A 2001;463:
387–92.
Carter EA, TamKK, Armstrong RS, Lay PA.Vibrational spectroscopic mapping andimaging
of tissues and cells. Biophys Rev 2009;1:95-103.
Chan KLA, Niu X, deMello Aj, Kazarian SG. Generation of Chemical Movies: FT-IR Spec-
troscopic Imaging of Segmented Flows. Anal Chem 2011;83:3606–9.
Chwiej J, Dulinska J, Janeczko K, Dumas P, Eichert D, Dudala J, et al. Synchrotron FTIR
micro-spectroscopy study of the rat hippocampal formation after
pilocarpine-evoked seizures. J Chem Neuroanat 2010;40:140–7.
Coates VJ, Offner A, Siegler EH. Design and performance of an infrared microscope at-
tachment. J Opt Soc Am 1953;43:984–9.
Colarusso P, Kidder LH, Levin IW, Fraser JC, Arens JF, Lewis EN. Infrared Spectroscopic
Imaging: From Planetary to Cellular Systems. Appl Spectrosc 1998;52:106A–20A.
Cope M, Delpy DT. System for long-term measurement of cerebral blood and tissue oxy-
genation on newborn infants by near infrared transillumination. Med Biol Eng
Comput 1988;26:289–94.
Cricenti A, Generosi R, PerfettiP, Gilligan JM, Tolk NH, Coluzza C, et al. Free-electron-laser
near-field nanospectroscopy. Appl Phys Lett 1988;73:151–3.
Cricenti A, Generosi R, Luce M, Perfetti P, Margaritondo G, Talley D, et al. Chemically re-
solved imaging of biological cells and thin films by infrared scanning near-field
optical microscopy. Biophys J 2003;85:2705–10.
Croxford AM, Nandakumar KS, Holmdahl R, Tobin MJ, McNaughton D, Rowley MJ.
Chemical changes demonstrated in cartilage by synchrotron infrared microspec-
troscopy in an antibody-induced murine model of rheumatoid arthritis. J Biomed
Opt 2011;16:066004-1. (9 pages).
Daudon M, Martial C, Lacour B, Bader C. Investigation of Urinary Crystals by Fourier
Transform Infrared Microscopy. Clin Chem 1991;37:83–7.
Dazzi A, Prazeres R, Glotin F, Ortega JM. Local infrared microspectroscopy with subwa-
velength spatial resolution with an atomic force microscope tip used as a photo-
thermal sensor. Opt Lett 2005;30:2388–90.
Dazzi A, Prazeres R, Glotin F, Ortega JM. Subwavelength infrared spectromicroscopy
using an AFM as a local absorption sensor. Infrared Phys Technol 2006;49:113–21.
Dazzi A, Prazeres R, Glotin F, Ortega JM. Analysis of nano-chemical mapping performed
by an AFM-based ("AFMIR") acousto-optic technique. Ultramicroscopy 2007a;107:
1194–200.
Dazzi A, Kjoller K, Reading M, High-frequency deflection measurement of IR absorp-
tion. U.S. patent 11/803,421 (May 15, 2007).
Dazzi A, Prazeres R, Glotin F, Ortega JM, Al-Sawaftah M, de Frutos M. Chemical mapping
of the distributions of viruses into infected bacteria with a phototermal method.
Ultramicroscopy 2008;108:635–41.
Dong A, Messerschmidt RG, Refnner JA. Infrared spectroscopy of a single cell - the
human erythrocyte. Biochem Biophys Res Commun 1988;156:752–6.
Dörr MS, Schade U, Hellwig P, Ortolani M. Characterization of Temperature-Dependent
Iron-Imidazole Vibrational Modes in Far Infrared. J Phys Chem B 2007;111:
14418–22.
Dukor RK. Vibrational Spectroscopy in the Detection of Cancer. In: Chalmers JM,
Griffiths PR, editors. Handbook of Vibrational Spectroscopy. Chichester: Wiley;
2002. p. 3335–61.
Dumas P, Jamin N, Teillaud JL, Miller LM, Beccard B. Imaging capabilities of synchrotron
infrared microspectroscopy. Faraday Discuss 2008;126:289–302.
Eichert D, Gregoratti L, Kaulich B, Marcello A, Melpignano P, Quaroni L, et al. Imaging
with spectroscopic micro-analysis using synchrotron radiation. Anal Bioanal
Chem 2007;389:1121–32.
Fernandez DC, Bhargava R, Hewitt SM, Levin IW. Infrared spectroscopic imaging for
histopathologic recognition. Nat Biotechnol 2005;23:469–74.
Garidel P, Boese M. Mid infrared microspectroscopic mapping and imaging: a bio-
analytical tool for spatially and chemically resolved tissue characterization and
evaluation of drug permeation within tissues. Microsc Res Tech 2007;70:336–49.
Gazi E, Dwyer J, Miyan J, Gardner P, Hart C, Brown M, et al. Fixation protocols for sub-
cellular imaging by synchrotron based FTIR microspectroscopy. Biopolymers
2005;77:18–30.
Geladi P. Chemometrics in spectroscopy. Part 1. Classical Chemometrics. Spectrochim
Acta B 2003;58:767–82.
Geladi P, Sethson B, Nyström J, Lillhonga T, Lestander T, Burger J. Chemometrics in
spectroscopy. Part 2. Examples. Spectrochim Acta B 2004;59:1347–57.
Gore RC. Infrared spectrometry of small samples with the reflecting microscope. Science
1949;110:7 10–1.
Holman H-YN, Martin MC. Synchrotron radiation infrared spectromicroscopy: a non-
invasive molecular probe for biogeochemical processes. Adv Agron 2006;90:79-127.
Holman H-YN, Bjornstad KA, McNamara MP, Martin MC, McKinney WR, Blakely EA.
Synchrotron Infrared Spectromicroscopy as a Novel Bioanalytical Microprobe for
Individual Living Cells: Cytotoxicity Considerations. J Biomed Opt 2002;7:417–24.
HolmanH-YN, Miles R, HaoZ, Wozei E, Anderson LM, Yang H. Real-time Chemical Imaging
of Bacterial Activity in Biofilms Using Open-Channel Microfluidics and Synchrotron
FTIR Spectromicroscopy. Anal Chem 2009;81:8564–70.
Houel J, Homeyer E, Sauvage S, Boucard P, Dazzi A, Prazeres R, et al. Midinfrared ab-
sorption measured at l/400 resolution with an atomic force microscope. Opt Ex-
press 2009;17:10887–94.
Jamin N, Dumas P, Moncuit J, Fridman W-H, Teillaud J-L, Carr GL, et al. Highly resolved
chemical imaging of living cells by using synchrotron infrared microspectrometry.
Proc Natl Acad Sci U S A 1998;95:4837–40.
Japing Y, Zhengzhong S, Xin C, Yufang H. Application of Synchrotron FTIR techniques in
Biomedical Fields. Prog Chem 2008;20:1768–1778.
Kazarian SG, Chan KL. Applications of ATR-FTIR spectroscopic imaging to biomedical
samples. Biochim Biophys Acta 2006;1758:858–67.
Kazarian SG, Chan KL. Micro- and macro-attenuated total reflection Fourier transform
infrared spectroscopic imaging. Appl Spectrosc 2010;64:135A–52A.
Keren S, Zavaleta C, Cheng Z, de la Zerda A, Gheysens O, Gambhir SS. Noninvasive mo-
lecular imaging of small living subjects using Raman spectroscopy. PNAS
2008;105:5845–9.
Knoll B, Keilmann F. Near-field probing of vibrational absorption for chemical micros-
copy. Nature 1999;399:134–7.
Koningsberger DC, Prins R, editors. X-ray Absorption: Principles, Applications, Tech-
niques of EXAFS, SEXAFS, and XANES. Chemical Analysis: A Series of Monographs
on Analytical Chemistry and Its ApplicationsNew York: John Wiley & Sons; 1988.
Krafft C, Kirsch M, Beleites C, Schackert G, Salzer R. Methodology for fiber-optic Raman
mapping and FTIR imaging of metastases in mouse brains. Anal Bioanal Chem
2007;389:1133–42.
Kwiatek WM, Paluszkiewcz C, Banas A, Kisiel A, Podgorczyk M, Marcelli A, et al. Appli-
cation of μ-FTIR-SR Spectroscopy to Prostate Tissue Analysis. Acta Phys Pol A
2009;115:602–5.
Lee E, Adar F, Whitley A. Multivariare Data Processing of Spectral Images: The Ugly, the
Bad, and the True Spectroscopy; August 2007.
Levin IW, Bhargava R. Fourier transform infrared vibrational spectroscopic imaging: Inte-
grating microscopy and molecular recognition. Annu Rev Phys Chem 2005;56:429–74.
Lin S-Y, Li M-J, Cheng W-T. FT-IR and Raman vibrational microspectroscopies used for
spectral biodiagnosis of human tissues. Spectroscopy 2007;21:1-30.
Marcelli A, Cinque G. Infrared synchrotron radiation beamlines: high brilliance tools for
IR spectromicroscopy. In: Moss D, editor. Biomedical Applications of Synchrotron
Infrared Microspectroscopy. Cambridge: Royal Society of Chemistry; 2011.
p. 67-104. Chapt. 3.
Martin FL, Kelly JG, Llabjani V, Martin-Hirsch PL, Patel II, Trevisan J, Fullwood NJ, Walsh
MJ. Distinguishing cell types or populations based on the computational analysis of
their infrared spectra. Nature Protocols 2010;5:1748–60.
Mayet C, Dazzi A, Prazeres R, Allot F, Glotin F, Ortega JM. Sub-100 nm IR spectromicro-
scopy of living cells. Opt Lett 2008;33:1611–3.
Mayet C, Dazzi A, Prazeres R, Ortega JM, Jaillard D. In situ identification and imaging of bac-
terial polymer nanogranules by infrared nanospectroscopy. Analyst 2010;135:2540–5.
McAuley WJ, Lad MD, Mader KT, Santos P, Tetteh J, Kazarian SG, et al. ATR-FTIR spec-
troscopy and spectroscopic imaging of solvent and permeant diffusion across
model membranes. Eur J Pharm Biopharm 2010;74:413–9.
1403A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
Author's personal copy
Meyer P, Lagarde P. Synchrotron radiation in the infrared. J Phys 1976;37:1387–90.
Miller L, Dumas P. Chemical imaging of biological tissue with synchrotron infrared
light. Biochim Biophys Acta 2006;1758:846–57.
Miller LM, Tague Jr TJ. Development and biomedical applications of fluorescence-
assisted synchrotron infrared micro-spectroscopy. Vib Spectrosc 2002;28:159–65.
Miller LM, Wang Q, Smith RJ, Zhong H, Elliott D, Warren J. A new sample substrate for
imaging and correlating organic and trace metal composition in biological cells and
tissues. Anal Bioanal Chem 2007;387:1705–15.
Moreira LM, Silveira L, Santos FV, Lyon JP, Rocha R, Zângaro RA, et al. Raman spectros-
copy: A powerful technique for biochemical analysis and diagnosis. Spectroscopy
2008;22:1-19.
Movasaghi Z, Rehman S, Rehman IU. Raman spectroscopy of biological tissues. Appl
Spectrosc Rev 2007;42:493–541.
Nakazato T, Oyamada M, Niimura N, Urasawa S, Konno O, Kagaya A, et al. Observation
of coherent synchrotron radiation. Phys Rev Lett 1989;63:1245–8.
Nichols EF. A Study of the Transmission Spectra of Certain Substances in the Infra-Red.
Phys Rev (Series I) 1893;1:1-18.
Notingher I, Verrier S, Romanska H, Bishop A, Polak J, Hench L. In situ characterisation
of living cells by Raman spectroscopy. Spectroscopy 2002;16:43–51.
Ortega JM, Glotin F, Prazeres R. Extension in far-infrared of the CLIO free-electron laser.
Infrared Phys Technol 2006;49:133–8.
Petibois C, Cestelli Guidi M. Bioimaging of cells and tissues using accelerator-based
sources. Anal Bioanal Chem 2008;391:1599–608.
Petibois C, Deleris G. Chemical mapping of tumor progression by FT-IR imaging: to-
wards molecular histopathology. Trends Biotechnol 2006;24:455–62.
Petibois C, Gouspillou G, Wehbe K, Delage JP, Deleris G. Analysis of type I and IV colla-
gens by FT-IR spectroscopy and imaging for a molecular investigation of skeletal
muscle connective tissue. Anal Bioanal Chem 2006a;386:1961–6.
Petibois C, Gionnet K, Goncalves M, Perromat A, Moenner M, Deleris G. Analytical per-
formances of FT-IR spectrometry and imaging for concentration measurements
within biological fluids, cells, and tissues. Analyst 2006b;131:640–7.
Petibois C, Piccinini M, Cestelli Guidi M, Déléris G, Marcelli A. Infrared Synchrotron
sources - a boon for biology. Nat Photonics 2009;3:177.
Petibois C, Piccinini M, Cestelli Guidi M, Marcelli A. Facing the challenge of biosample
imaging by FTIR with a synchrotron radiation source. J Synchrotron Radiat
2010a;17:1-11.
Petibois C, Cestelli Guidi M, Piccinini M, Moenner M, Marcelli A. Synchrotron radiation
FTIR imaging in minutes: a first step towards real time cell imaging. Anal Bioanal
Chem 2010b;397:2123–9.
Petter CH, Heigl N, Rainer M, Bakry R, Pallua J, Bonn GK, et al. Development and appli-
cation of Fourier-transform infrared chemical imaging of tumour in human tissue.
Curr Med Chem 2009;16:318–26.
Pistorius AMA. Biochemical applications of FT-IR spectroscopy. Spectrosc Eur 1995;7/4:8-15.
Podgórczyk M, Kwiatek WM, Zając W, Dulińska-Litewka J, Welter Erolimund D. Zinc in
native tissues and cultured cell lines of human prostate studied by SR-XRF and
XANES. X-ray Spectrom 2009;38:557–62.
Policar C, Waern JB,Plamont M-A,Clede S, Mayet C,Prazeres R, et al.Subcellular IR Imaging
of a Metal–Carbonyl Moiety Using Photothermally Induced Resonance. Angew Chem
Int Ed 2011;50:860–4.
Quaroni L, Casson AG. Characterization of Barrett esophagus and esophageal adenocar-
cinoma by Fourier-transform infrared microscopy. Analyst 2009;134:1240–6.
Romeo M, Diem M. Correction of dispersive line shape artefact observed in diffuse re-
flection infrared spectroscopy and absorbtion/ reflection (transflection) infrared
micro-spectroscopy. Vib Spectrosc 2005;38:129–32.
Sannibale F, Byrd JM, Loftsdóttir A, Venturini M, Abo-Bakr M, Feikes J, et al. A Model De-
scribing Stable Coherent Synchrotron Radiation in Storage Rings. Phys Rev Lett
2004;93:094801. [4 pages].
Schade U, Holldack K, Martin MC, Fried D. THz near-field imaging of biological tissues
employing synchrotron radiation. Proc SPIE 2005;5725:46–52.
Sherwin MS, Schmuttenmaer CA, Bucksbaum PH, editors. Opportunities in THz Science,
DOE-NSF-NIH Workshop; 2004.
Siegel PH. Terahertz Technology in Biology and Medicine. IEEE Trans Microwave Theory
Tech 2004;52:2438–47.
Silverstein RM, Bassler GC, Morrill TC. Spectrometric Identification of Organic Com-
pounds. 5th ed. New York: John Wiley & Sons Inc; 1991. p. 91-164.
Snively CM, Pellerin C, Rabolt JF, Chase DB. Acquisition of Mid-Infrared Spectra from
Non-Repeatable Events with Sub-100 μs Temporal Resolution Using Planar Array
Infrared Spectroscopy. Anal Chem 2004;76:1811–6.
Srinivasan G, Bhargava R. Fourier transform-infrared spectroscopic imaging: The
emerging evolution from a microscopy tool to a cancer imaging modality. Spec-
troscopy 2007;22:30–43.
Steiner G, Koch E. Trends in Fourier transform infrared spectroscopic imaging. Anal
Bioanal Chem 2009;394:671–8.
Stevenson JR, Ellis H, Bartlett R. Synchrotron Radiation as an Infrared Source. Appl Opt
1973;12:2884–9.
Swain RJ, Stevens MM. Raman microspectroscopy for noninvasive biochemical analysis
of single cells. Biochem Soc Trans 2007;35:544–9.
Talley DB, ShawLB, Sanghera JS, Aggarwal ID, Cricenti A, Generosi R, et al. Scanning near
field infrared microscopy using chalcogenide fiber tips. Mater Lett 2000;42:339–44.
Tollefson M, Magera J, Sebo T, Cohen J, Drauch A, Maier J, et al. Raman spectral imaging
of prostate cancer: can Raman molecular imaging be used to augment standard
histopathology? Br J Urol Int 2010;106:484–8.
Tonouchi M. Cutting-edge terahertz technology. Nat Photonics 2007;1:97-105.
Wallace VP, Fitzgerald AJ, Cluff J, Arnone DD. Dermatological Surgery and Lasers Tera-
hertz pulsed imaging of and in vivo. Br J Dermatol 2004;151:424–32.
Williams GP. The initial scientific program at the NSLS infrared beamline. Nucl Instrum
Methods A 1990;291:8-12.
Williams GP. Filling the THz gap - high power sources and applications. Rep Prog Phys
2006;69:301–26.
Woernley DL. Infrared absorption curves for normal and neoplastic tissues and related
biological substances. Cancer Res 1952;12:516–23.
Wood BR, Bambery KR, Evans CJ, Quinn MA, McNaughton D. A three-dimensional mul-
tivariate image processing technique for the analysis of FTIR spectroscopic images
of multiple tissue sections. BMC Med Imaging 2006;6:12. (9 pages).
Xu W, Marcelli A, Hampai D, Malfatti L, Innocenzi P, Schade U, Wu Z. New opportunity
to investigate physico-chemical phenomena: time resolved X-ray and IR concur-
rent analysis. Rend Accad Lincei 2011;22:59–79.
Yarwood J, Shuttleworth T, Hasted JB, Nanba T. A new radiation source for the infrared
region. Nature 1984;317:742–4.
Zelsmann HR. Temperature dependence of the optical constants for liquid H
2
O and D
2
O
in the far IR region. J Mol Struct 1995;350:95-114.
1404 A. Marcelli et al. / Biotechnology Advances 30 (2012) 1390–1404
