Raman Spectroscopy, Microscopy and Imaging

Q&A
Raman spectroscopy is now well established as one of the The second section introduces the basics of Raman
most powerful and versatile techniques for a diverse range scattering and Raman spectroscopy. It considers the
of applications in both research and analytical laboratories. relationship between the properties of molecules and their
Integrating the Raman spectrometer with a light microscope Raman activity, points out the complementary nature of
offers a unique ability to noninvasively characterize Raman and FTIR spectroscopies, and summarizes the major
chemically complex and spatially inhomogeneous samples benefits of Raman spectroscopy. Special attention is drawn
with a sub-micron spatial resolution. Modern confocal to the phenomenon of fluorescence and how it interferes
Raman microscopy and imaging, which allows one to with Raman spectra.
obtain two- and three-dimensional spectrochemical images
of samples in various states and forms, has become an In the third section, the key components of a Raman
indispensable research method for scientific communities, spectrometer are specified, the basic principles of their
from physicists to chemists, to criminalists, to geologists, to operations are described, and their main parameters are
biologists, to medics, etc. characterized. The importance of choosing the most optimal
excitation (laser) wavelength and finely controlling its power
The Q&A for Raman Spectroscopy, Microscopy and is emphasized in order to obtain the most informative Raman
Imaging aims to familiarize a wide circle of experimenters spectral data for a specific application.
with the fundamentals of modern Raman spectroscopy and
microscopy, in particular. The Q&A consists of four sections: The final fourth section introduces the fundamentals and
1) Introduction to Raman Spectroscopy, 2) Basic Theory and practical aspects of confocal Raman microscopy. The role
Terminology, 3) Raman Spectrometer and Instrumentation, 4) of confocal aperture, the importance of regular alignment
Raman Microscopy and Instrumentation. and calibration procedures, as well as essential relations
between the microscope objective, aperture setting
The first section outlines the physics of the interaction of and mapping step are emphasized to understand and
light and molecules and the experimental discovery of the successfully operate a Raman microscope. The recent
Raman effect. advances in fast Raman imaging technique capable of
obtaining both traditional optical and spectrochemical
images with a spatial resolution at the diffraction limit of
light are outlined.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Introduction to Raman Spectroscopy

What are Raman scattering and Raman spectroscopy? In this first work, Raman and Krishnan used a simple yet When the light propagates through a matter, the light
The interaction of light, a form of electromagnetic elegant experimental setup in which sunlight was focused photons tend not to interact with the molecules of
radiation, with matter results in the absorption, by a telescope on scattering materials, thoroughly the matter but rather pass through. Only one photon
transmission, reflection, and scattering of light. purified liquids or dust-free gases, and the method of out of about a thousand incident photons is usually
complementary light filters was applied to visually observe scattered by the molecules. Of the scattered photons, an
The scattering of light by molecules first considered
“a modified scattered radiation”: overwhelming majority is elastically scattered, keeping
and explained in the works of Rayleigh occurs without
the same energy as the incident photons. Only one out of
changing the frequency of the light––the frequency of the Raman and Krishnan examined more than sixty common1
a million scattered photons changes the energy through
scattered light is equal to the frequency of the incident liquids and vapors and detected the modified scattered
an inelastic interaction with the molecules. Therefore, in
one. Such scattering is called elastic since the state of the radiation with a changed frequency in each of the liquids
Raman scattering, the number of photons is about one
molecule after scattering is the same as before. This is to a greater or lesser degree.
photon of a billion incident photons.
the dominant scattering of light by particles much smaller
Sir Chandrasekhara Violet
than the wavelength of the radiation, and this process is
Venkata Raman was filter
called Rayleigh scattering. Rayleigh established that the
awarded the Nobel
amount of the scattered light I was inversely proportional
Prize for Physics in Telescope
to the fourth power of wavelength λ of the light for the
1930 for his remarkable Sunlight
particles up to about a tenth of the wavelength of light:
experimental discovery (white)
and proof of the Scatter
Rayleigh scattered light liquid
(1) universal character
of this effect by
The effect of inelastic scattering in liquids and vapors
investigating a large Raman scattered
was first experimentally discovered by Chandrasekhara
number of solids
Venkata Raman and Kariamanickam Srinivasa Krishnan
and liquids and the
and reported in March of 1928 in “Nature”.
first publication of a Green
Sir Chandrasekhara Venkata Raman.
spectrum of scattered filter
light with changing Observer
frequencies. Raman scattering was named after him, and
his discovery laid the foundation for the modern use of
Raman spectroscopy and Raman microscopy.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Despite the challenge in the practical detection of

the weak Raman effect, the discovery of this physical
phenomenon formed the basis of the whole direction of
vibrational spectroscopy––Raman spectroscopy.

Vibrational spectroscopy is the most significant

portion of molecular spectroscopy because it delivers
comprehensive information about the structure and
properties of molecules by studying the vibrations and
interactions of chemical bonds that associate atoms
into molecules. Infrared (IR) and Raman spectroscopies
are two major types of vibrational spectroscopy. IR and
Raman represent universal tools for the determination
and identification of molecular structure in virtually any
type of sample and environment. Both methods are
complementary, as they provide complete information on
molecular vibrations.

In general, a Raman spectrum is composed of a number

of spectral lines or bands (the groups of closely spaced
spectral lines) displaying their intensity and wavelength
position. Each spectral line or band corresponds to a
specific molecular bond vibration, including individual
bonds such as C-C, C=C, C=O, N-O, C-H, etc., and
groups of bonds such as benzene ring breathing mode,
polymer chain vibrations, crystal lattice modes, etc.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Basic Theory and Terminology

What are Stokes and anti-Stokes Raman scattering? For the sake of simplicity, consider a ground electronic The incident photon excites the molecule from an initial
According to quantum mechanics, an isolated molecule state E0 with only two corresponding vibrational levels energy state to a virtual Evirt state, as shown in Figure 1(b).
can only exist in quantized energy states that, in of energy denoted by quantum numbers 0 and 1 for the A virtual state is a very short-lived intermediate quantum
the general case, can be represented by electronic, ground and excited level, respectively. The rotational state that is not defined by a strict molecular energy
vibrational and rotational levels of energy. In the quantum- energy levels are ignored since their energy is significantly value. From the virtual state, the molecule instantaneously
mechanical interpretation, the scattering of a photon by a lower than the energy of vibrations, and the rotation relaxes back to a vibrational level in the ground E0
vibrating molecule is illustrated in Figure 1. motion is hindered in condensed matter. electronic state and emits a photon of light. When
relaxation occurs, there are three possible situations.
First, the molecule is excited from and then relaxes down
to the same ground vibrational level 0 and reemits a
photon of light with the energy equal to that of the incident
photon. As we know, this is an elastic process that
corresponds to Rayleigh scattering with the unchanged
frequency of the light. The other two situations result in
the inelastic scattering of the incident photon that alters
its energy. If the molecule is excited by a photon from
level 0 and then relaxes to level 1, then the molecule gains
energy and reemits a photon of lesser energy (referred to
as being “red”-shifted) that corresponds to Stokes Raman
scattering. The third possible situation realizes when the
molecule is excited from level 1 and then relaxes down to
level 0, thus losing its energy and reemitting a photon of
light with higher energy than the incident photon.

Figure 1. Schematic representation of quantum-mechanical interpretation of Rayleigh and Raman scattering. (a) The incident photons of light
with the energy hν 0 are scattered off by the vibrating molecule: the green corpuscles symbolize photons with unchanged hν 0 , whereas the red
and blue ones imply photons in Stokes and anti-Stokes Raman scattering, respectively. (b) The energy level diagrams illustrate the transition
between the vibration levels accompanied by the absorption of an incident photon and emission of a scattered photon: Rayleigh scattering
with unchanged energy hν 0 , Stokes Raman scattering with red-shifted photon h(ν 0 – νvib) and anti-Stokes Raman scattering with blue-shifted
photon h(ν 0 + νvib). The left inset sketches the population of the vibrational levels given by the exponential Boltzmann distribution. (c) The
illustration of the higher intensity of Stokes Raman (red) scattering with regards to anti-Stokes Raman (blue) scattering as follows from the
Boltzmann distribution. (The figure is reproduced by permission from Alexander Rzhevskii, Modern Raman Microscopy: Technique and
Practice, Cambridge Scholars Publishing, 2021, 392.)
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

This situation with the “blue”-shifted photon corresponds Accordingly, there is a much higher probability that The second-order Raman band corresponding to the first
to anti-Stokes Raman scattering. Thus, Rayleigh (ν₀), photons undergo the Stokes transition, and the intensity overtone is much weaker and is observed at ~ 960 cm-1.
Stokes (ν S), and anti-Stokes (νaS) spectral components can ratio of Stokes (IS) to anti-Stokes (IaS) Raman scattering In the anti-Stokes region, the only first-order band at
be observed in scattered light depends on the separation between the vibration energy -520.7 cm-1 is evident, whose intensity is about an order
levels and the temperature in accordance with Boltzmann of magnitude lower than that of the peak at 520.7 cm-1.
distribution: The example of Si demonstrates that, at ordinary
(2)
temperatures, the measurement of Raman spectra in the
Stokes region has the most practical importance. Indeed,
where ν₀ is the frequency of incident light and νvib is the
(3)
the intensities of spectral bands and, thus, signal-to-noise
frequency of the molecular vibration.
ratios (S/N) in the Stokes region are much higher than
A quantum mechanical approach correctly predicts the At normal temperatures, this ratio is very large, and those in the anti-Stokes, as follows from (3).
difference in intensity between Stokes and anti-Stokes anti-Stokes Raman scattering is very weak compared
Raman scattering, as illustrated in Figure 1(c). Indeed, to Stokes Raman scattering. At higher temperatures, What is Raman shift?
the intensity of Raman scattering is proportional to the the ratio decreases, and anti-Stokes Raman scattering The frequencies in Raman spectra are usually calculated
number of molecules being illuminated. Therefore, the increases noticeably. as a Raman shift, i.e., the difference between the
intensity of Stokes Raman scattering is proportional frequencies of excitation light and Raman transitions.
to the number of molecules N0 occupying the ground 520.7
Following (2), in Stokes (ν 0 - ν S = ν 0 - (ν 0 - νvib) = νvib) and
level 0, whereas the intensity of anti-Stokes Raman is anti-Stokes (ν 0 - νaS = ν 0 - (ν 0 + νvib) = -νvib) regions, the
proportional to the number of molecules N1 occupying vibrational frequencies result in quantities with the sign
Intensity (a.u.)
the excited level 1. At thermal equilibrium, the population “plus” and “minus”, respectively (see Figure 2). Thus, in the
of excited vibrational levels decreases exponentially with Raman spectrum, the intensities along the ordinate axis
-520.7
the difference between the energy of excited and ground 960 are plotted vs. the Raman shift along the abscissa axis,
levels following the Boltzmann distribution which, in case 1500 1000 500 0 -500 -1000 -1500 and the shift appears to be independent of the frequency
Raman shift (cm-1)
of h(ν1 - ν 0) = hνvib for levels 1 and 0, is given by of excitation light. This is why the spectral search against
commercial or user-built Raman libraries often used for
Figure 2. Raman spectrum of a Si wafer in Stokes and anti-Stokes
regions. (The figure is reproduced by permission from Alexander the identification of unknown compounds can generally
Rzhevskii, Modern Raman Microscopy: Technique and Practice,
be performed for experimental and reference spectra
Cambridge Scholars Publishing, 2021, 392.)
measured with the use of different excitation wavelengths.
Figure 2 shows the Raman spectrum of a silicon (Si) wafer
where h is Planck’s constant, k is Boltzmann’s constant,
in Stokes and anti-Stokes regions. In the Stokes region,
and T is the temperature in Kelvin. At room temperature,
the so-called first-order Raman spectrum manifests a
the ground vibrational energy level is always more
strong band at 520.7 cm-1 arising from the fundamental
populated than the excited one.
transition assigned to the triply degenerate, long-
wavelength transverse optical phonon (TO).
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

What does Raman intensity depend on? characterizes the ability of a molecule or chemical bond What are the selection rules in IR and Raman
As follows from (1), I∝(ν 0 ) . It can be shown that the
4
to scatter light at a given frequency of vibration νvib and is spectroscopy?
intensity of Stokes Raman component scattered by N called a Raman scattering cross-section. A simple classical approach that takes the changes
molecules of a sample can be written as in the dipole moment and the polarizability during the
The absorptivity ε and Raman cross-section σ play a
molecular vibration into consideration can be used to
somewhat similar role in both methods of vibrational
explain if the vibration results in a corresponding IR or/and
spectroscopy. These coefficients are a measure of
Raman spectral band. The dipole moment arises from
(4) the probability of a specific transition to occur and
the difference in electronegativity of atoms composing
characterize the intrinsic properties of molecules or
where I0 and ν 0 are the intensity and the frequency of the chemical bonds and due to non-uniform distributions of
chemical bonds to absorb and scatter light. However, the
excitation light, respectively, and the derivative (∂α/∂q) is positive and negative charges on the various atoms in
Raman cross-section σ strongly depends on the excitation
the rate of change of the polarizability α of a molecule or a molecule. The magnitude of the dipole moment of an
frequency being proportional to its fourth power. That
chemical bond upon the displacement of nuclei q from the individual chemical bond composed of atoms that have
is, for a given sample, replacing a 1064 nm laser with
equilibrium position. This dependence can be presented different electronegativity is given by
a 532 nm laser (both of which are popular choices for
in the form to some extent analogous to the Lambert-Beer
Raman spectroscopy) increases the scattering cross-
law that establishes the relationship between the sample
section by a factor of 16 (see Figure 3). Thus, an important
absorbance A, the extinction coefficient (absorptivity) ε,
outcome of expressions (5) and (6) is that the intensities
the concentration of the molecules C, and the pathlength where Q is the charge and l is the distance between the
of Raman bands can be increased by increasing the
that light travels through the sample d in IR spectroscopy: atoms. Thus, the dipole moment is a measure of the
intensity I0 and/or the frequency ν 0 of excitation light.
A=εCd. Thus, the relationship (4) for Raman scattering can charge asymmetry of a molecule. The total dipole moment
Nevertheless, as in IR, the intensities of spectral bands
be expressed as of a polyatomic molecule is a vector sum of the individual
are proportional to the concentration of an analyte C,
dipole moments of chemical bonds. The dipole moment
making Raman spectroscopy well suitable for quantitative
oscillates during molecular vibration with the frequency of
(5) analysis.
the vibration νvib:
where K is a parameter mostly determined by
experimental conditions of spectral measurement, C is
the concentration of the molecules in the sample, and V
is the sampling volume, i.e., the volume from which the where µ 0 is the permanent dipole moment in the
scattered light is collected and measured. However, in equilibrium state of the molecule and t is time.
contrast to the Lambert-Beer law, the intensities of Raman
bands linearly depend on the intensity I0 of excitation light.
The coefficient σ defined as

Figure 3. The difference in the absolute intensity of spectral

bands in the Raman spectrum of Sulphur obtained with 532 and
1064 nm excitation wavelengths. Note that the relative intensities
(6) of the Raman bands remain unchanged and do not depend on the
excitation wavelength.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The classical electromagnetic theory requires that for a

c
molecule to absorb the light, the change in the dipole
b
moment with respect to displacement during the vibration (10)

(the first derivative) must be non-zero:

1⁄ α"
and the vibration is said to be Raman-active. By contrast,
1⁄ α!
if the molecular vibration does not cause a variation in the a
polarizability α, then there is no amplitude modulation of
(7)
the dipole moment, and there is no Raman Stokes and
anti-Stokes radiation. Such vibration is called Raman-
and the corresponding vibration is said to be IR-active.
inactive.
(a)
The oscillating electric field of the light wave perturbs the
The necessary conditions for an active vibration
cloud of electrons surrounding the nuclei and displaces
expressed by formulas (7) and (10) are often termed
the center of the electron cloud relative to the center of
the gross or major selection rules, respectively, for
the nuclei. The displacement results in an induced dipole
IR absorption and Raman scattering. In a polyatomic
moment p:
molecule, the vibrations that bring on a net change in
a dipole moment are IR-active, and those that induce
polarizability changes are Raman-active. Obviously, some
where α is the polarizability of the molecule and E is the vibrations can be both Raman- and IR-active. (b)
electric field of the light wave with the amplitude E0 that
Figure 4. The ellipsoid of polarizability. (a) Half the lengths of the
oscillates with time t as a cosine function What is the ellipsoid of polarizability? principal axes are inversely proportional to the square roots of
The polarizability α of the molecule is an important the components of the polarizability in the orthogonal axes a,
b and c. (b) The change of the ellipsoid of polarizability for the
(8) parameter in the classical theoretical consideration of homonuclear diatomic molecule O₂ upon its vibration. (The figure
is reproduced by permission from Alexander Rzhevskii, Modern
the Raman effect. Polarizability can be described as a Raman Microscopy: Technique and Practice, Cambridge Scholars
Thus, the oscillation of the induced dipole moment p measure of the deformability of the electron cloud in the Publishing, 2021, 392.)

occurs with the frequency ν₀ of the light wave: molecule or, in other words, how easily the electron cloud The ellipsoid of polarizability is shown in Figure 4(a).
can be distorted by the electric field of the light wave The axes a, b and c are termed “principal” axes that
(9) interacting with the molecule. The polarizability α is not can often be chosen so that they are related to the
constant because the vibrations of atoms in the molecule symmetry axes of a molecule. Note that the axes of the
The interaction of light with a molecule leads to the can cause it to alter. The polarizability of a molecule ellipsoid are inversely proportional to the square roots of
appearance of an induced dipole moment, even if there can be represented in the form of an ellipsoid called the the components α a, α b and α c of the polarizability in the
may be no permanent dipole moment. An important ellipsoid of polarizability that helps to visualize how the orthogonal axes a, b and c. Since the electrons are more
consequence of equations (8) and (9) is that for Raman polarizability of a molecule changes during its vibrations polarizable in the direction of a chemical bond (a larger α),
scattering to occur, the vibration that changes the and to figure out whether or not a given vibration is the shortest axes of the polarizability ellipsoid correspond
polarizability α with the displacement of nuclei q must be Raman-active. to the directions of the easiest polarization, as shown in
non-zero: Figure 4(b) for a homonuclear diatomic molecule.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

A vibration is Raman-active if the size, shape or polarization. In contrast, samples such as solid or liquid vibrational. A Raman spectral band with a depolarization
orientation of the polarizability ellipsoid changes during crystals, carbon nanomaterials, strained polymer films ratio in the range of 0 ≤ ρ ≤ 3/4 is called a polarized band,
the vibration. In a homonuclear diatomic molecule, such and fibers are typically very sensitive to sample orientation and a band with a depolarization ratio ρ ≥ 3/4 is called a
as O₂, H₂ or N₂, there is no permanent dipole moment and are called anisotropic. Oriented or anisotropic depolarized band.
in their equilibrium states. The vibration of the only truly samples have an axis of symmetry or an optical axis.
In a typical polarization experiment, a Raman spectrum
covalent nonpolar bond does not change the dipole The orientation of this axis relative to the incident laser
is collected with the light linearly polarized in one
moment such that the derivative of the dipole moment polarization can result in quite different spectra.
direction, and then another spectrum is collected with
(7) is zero, and the vibrations of these molecules are not
Hence, Raman spectra measured using the naturally the light polarized perpendicular to the first direction.
IR-active. However, the polarizability of the molecules
polarized light may not be well reproducible if the relative Normally, the polarization experiment is performed
changes with the nuclear displacement, and the vibration
orientation of the anisotropic samples and the direction without a polarization analyzer so that the scattered
results in a non-zero derivative in (10). As shown for the O₂
of the laser polarization is not taken into consideration. light of both polarizations is measured at its maximum
molecule, the polarizability ellipsoid noticeably changes
For this reason, the laser excitation beam is purposely intensity. However, polarizers and analyzers are the parts
alongside the bond, and, therefore, the vibration is
depolarized in general-purpose instruments to ensure of Raman polarization optics in which the analyzer acts
Raman-active. The vibrational bands of O₂ and N₂ can be
consistent results regardless of the sample orientation. as a second polarizer. The main functional purpose of
correspondingly observed at about 1556 and 2329 cm-1 in
The Raman spectra obtained using the depolarized the analyzer is that it can be used to accurately check
the Raman spectrum of ambient air.
laser excitation are referred to as depolarized Raman whether the polarization of the light has been changed
spectra. Advanced Raman instruments can be configured after the interaction with a sample. The analyzer can
What is Raman polarization?
to perform both depolarized and polarized Raman usually be set to either parallel or perpendicular to the
In modern Raman instruments, the lasers used for
measurements. polarization of the excitation light, but the intensity of
the excitation of Raman scattering are usually linearly
Raman scattering decreases especially noticeably if the
polarized by design. It means that the electric field The polarization of the laser beam can be achieved using
analyzer and polarizer are oriented orthogonally (crossed).
vector of the light wave is oriented along a single axis a polarizer or polarization filter. The polarizer is oriented
Nevertheless, there are situations in which the use of the
perpendicular to the direction of the light beam generated accordingly to obtain the required direction of polarization
analyzer can be helpful in the characterization of materials
by the laser. When exposed to linearly polarized light, of the laser beam. The scattered light may consist of the
and assignment of the spectral bands in their Raman
a sample may interact differently with light resulting in components with the directions of polarization parallel and
spectra.
dissimilar Raman spectra depending on how the sample perpendicular to the electric field vector in the incident
is oriented relative to the direction of the polarization axis. light. The ratio of the intensities of the perpendicular to the When making polarization measurements, it is important
However, not all materials are sensitive to the orientation parallel component is called the depolarization ratio ρ: to keep track of the directions of polarization, i.e., the
of polarized light. The term isotropic means the properties orientation of the polarizer and analyzer, so that the
of the sample (e.g., its Raman spectrum) are not sensitive results can be unambiguously interpreted. In polarization-
to sample orientation. Examples include liquids, powders, dependent Raman spectroscopy, it is usual to designate
and randomly oriented polymers. While these samples the directions of incident and collected light and the
are not sensitive to orientation, they may still interact with This ratio represents an important value measured directions of polarization using the so-called Porto
polarized light and cause the polarization of the scattered in Raman spectroscopy because it depends on the notation.
Raman beam to be changed relative to the incident laser symmetry of the molecule and a given molecular
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Porto notation describes the position of a sample with • Z(XY)Z - laser is polarized parallel to an X-axis; and carbon tetrachloride can be found in many
respect to the orientations of the polarizer and analyzer. analyzer is set to pass the light polarized in Y-axis textbooks and original articles. In general, polarization-
The designation consists of four terms written in the form of (polarized and analyzer are crossed); dependent Raman spectroscopy provides information
• Z(YX)Z - laser is polarized parallel to a Y-axis; analyzer about preferentially oriented molecules or materials. For
is set to pass the light polarized in X-axis (polarized example, a drug may crystallize in different polymorphic
and analyzer are crossed). forms which have identical spectra when measured
and defined as follows: with depolarized light. When examined under linearly
polarized light, the drug spectra may change providing
a. direction of propagation of the incident laser beam
confirmation of a specific crystalline structure. Polarization
b. direction of polarization of the incident laser beam measurements are also commonly used to reveal strain in
materials such as stretched polymer films, doped silicon
c. direction of polarization of scattered light
structures or cast materials.
d. direction of observations of scattered light
What are the benefits of Raman spectroscopy?
In the majority of modern Raman instruments, including
As mentioned above, Raman spectroscopy provides
DXR series of Raman spectrometers and microscopes,
chemical and structural information complementary to
the so-called 180-degree backscattering sampling
Fourier-transform Infrared Spectroscopy (FTIR). In general,
configuration is implemented by design. It means that the Figure 5. Orientation polarization axes at the sample in DXR2(3)
Raman microscope and the corresponding Porto notation.
chemical compounds manifest distinctive spectral bands
excitation laser beam propagates along a certain axis,
in their Raman and FTIR spectra, but they may not
and the Raman scattering is collected along the same For an anisotropic sample, it is recommended to make
coincide due to different selection rules. Which technique
axis but in the opposite direction. By convention, the measurements for all four positions of the polarizer and
provides better results depends on the samples under
Z-axis of the Cartesian coordinate system is chosen to analyzer relative to the sample’s orientation to describe
study, the concentration of analytes, the surrounding
be vertical. The convention is also to use an overbar on the sample by means of polarization-dependent Raman
matrix or solution, the presence of impurities, the desired
the Z-axis notation to designate the opposite direction Z. spectroscopy. Polarized Raman with angle resolution
sampling method, and whether the sample is under
Thus, as follows from Figure 5, if the excitation laser beam offered in DXR2 and DXR3 series of Raman microscopes,
temperature or pressure. For many applications, Raman
incidents along the Z-axis, there are only four principal when the analyzer can be set between 0 and 90 degrees
spectroscopy offers indisputable advantages; the major
polarization orientations that can be analyzed with a with an increment of one degree, provides additional
ones are as follows:
Raman instrument: flexibility in Raman polarization experiments. It is also a
good practice to measure the spectrum with the non- 1. No sample preparation is required, or the sample
• Z(XX)Z - laser is polarized parallel to an X-axis; preparation is rather trivial. In contrast to FTIR,
polarized incident laser beam to characterize the sample
analyzer is set to pass the light polarized in X-axis; which often requires dissolution, diluting or pressing
as completely as possible.
samples into pellets, no physical or chemical
• Z(YY)Z - laser is polarized parallel to a Y-axis; analyzer
Historically, polarization-dependent Raman spectroscopy treatment of the sample is typically necessary for
is set to pass the light polarized in Y-axis;
has helped to determine the symmetry of the vibrations Raman measurements. In principle, if a sample under
study fits into the sample compartment of a Raman
and correctly assign the observed Raman bands. Classic
instrument, the Raman spectrum of the sample can be
examples of the polarized Raman spectra of cyclohexane
obtained.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

2. The technique is nondestructive and saves the 6. Since water is a weak Raman scatter and does 11. Both Raman and FTIR methods can be used with
sample integrity, provided that careful control of the not manifest much in Raman spectra, Raman the microscopic technique. However, combining a
laser radiation density is carried out to ensure that spectroscopy is ideal for the study of analytes, e.g., Raman spectrometer with a microscope results in
there is no laser-induced modification to the sample. biological compounds, in aqueous solutions. There one of the major advantages over FTIR: an order of
Then, after the Raman measurements, the sample is no need for time-consuming sample extraction or magnitude better spatial resolution. Spatial resolution
can be treated and analyzed by other methods. The drying that may also alter the chemistry of the sample. is improved because Raman spectroscopy uses
nondestructive aspect is of paramount importance for This is a significant advantage over FTIR spectroscopy, an excitation wavelength in the visible region of the
forensic analyses where the samples collected during which capabilities are limited by the strong absorption spectrum, whereas FTIR uses longer wavelengths in
an investigation should be kept intact. of water when the analysis of aqueous solutions is the infrared region. The shorter excitation wavelengths
required. of visible light result in a smaller diffraction limit and
3. The non-contact method of examination of samples
can, therefore, be focused onto a smaller spot on the
by light provides safe procedures for both samples 7. Raman spectroscopy is applicable to a wide range of
sample than the longer wavelengths of infrared light.
and operators. On the one hand, the procedure substances and materials. Raman spectra of liquids,
Raman microscopy is currently capable of probing
ensures that no contamination is introduced into the solids, gases (even those composed of homonuclear
objects as small as about 0.5 μm, whereas an FTIR
sample. On the other, hazardous, toxic or unstable molecules), vapors and aerosols can be obtained.
microscope can resolve structures of 5–10 μm at best.
to air or moisture samples can be measured through The only exception is pure metals and alloys, which
Thus, individual particles, defects, trace amounts of
a protective sampling insertion or placed in sealed just reflect light. However, metal oxides, carbides and
substances, biological cells and microorganisms can
glass containers. Samples can be analyzed through nitrides are Raman-active. Samples can be in the
be analyzed by Raman microscopy on a sub-micron
transparent windows, cuvettes, vials, bottles or form of a solution, slurry, gel, powder, chunk, crystal,
level.
translucent polymer packaging. wafer, film or fiber with various sizes, shapes and
thicknesses. 12. A confocal analysis is another unique advantage of
4. Remote analysis can be conveniently realized in
Raman microscopy. A confocal Raman microscope
Raman spectrometers. Fiberoptic probes transmit 8. During Raman spectral measurements, the full
allows the measurement of a minuscule discrete
laser light and collect Raman scattering over long spectral region of fundamental vibrations, usually
volume in a transparent sample and, thus, can
distances, up to hundreds of meters away from the from 50 to 3500 cm-1, can be recorded in a single provide the analysis at different depths and even 3D
mainframe of the Raman instrument. exposure. Thus, Raman spectra are appropriate spectrochemical rendering without physical slicing
for the characterization of organic and inorganic of the sample. The confocal capability is particularly
5. The nondestructive, noninvasive and non-contact
compounds, such as those used in polymer useful for the nondestructive analysis of layered
nature of the Raman measurement makes it very
composites. In contrast, FTIR instruments typically samples (e.g., multi-layered polymer films), volume
suitable for in situ, in vitro and in vivo analysis. It
require reconfiguration (a change of beamsplitters, inclusions and defects (in glass materials, plastics
can be used to monitor chemical reactions in situ
detectors or sources) in order to cover the spectral or minerals), semiconductor thin films, polymer
within a reaction chamber, often at cryogenic or
very high temperatures and pressures, including the region below 400 cm-1. composites, and many others.
identification and quantitation of combustion products 9. Beyond general material characterization and
in flames and plasmas. Raman spectroscopy can be In fairness, it should be noted that Raman spectroscopy
identification, Raman spectroscopy also provides
used for in vitro and in vivo characterization of the also has known disadvantages. The major one is
information on more subtle chemical effects, such
chemical and morphologic structure of biological cells, as the extent of crystallinity, polymorphism, phase certainly the fluorescence interference that is considered
tissue and microorganisms. transformation, strain in materials and protein below. Also, the laser power aimed at the sample needs
secondary structure. to be thoroughly controlled to avoid local heating or

10. Often, due to the higher likelihood of the presence of the photodecomposition, especially in resonance Raman
Raman bands that are well resolved and do not overlap experiments where the laser frequency is intentionally
with their neighbors, Raman spectra are better suitable tuned into the absorption band of the compound under
for material identification by a spectral library search. study.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

What is fluorescence, and how does it interfere with As mentioned above, the low quantum yield (QY) of the Aromatic hydrocarbons, such as fluorescein,
Raman spectra? Raman effect is the major challenge in the practical phenolphthalein, naphthalene, anthracene and perylene,
As it is known, vibrational spectra are observed as a detection of Raman scattering. Another challenge to generate a strong fluorescence upon excitation with
result of transitions between quantized energy levels of Raman scattering comes from fluorescence emission that visible light. The intrinsic fluorescence of biological tissues
molecules. Figure 6 shows the transitions in the cases has a similar origin but a much higher QY. Fluorescence is often related to the presence of proteins, especially
of IR absorption, Raman scattering and fluorescence, may overwhelm or mask the relatively weak Raman signal, hemoglobin. In spectroscopy, this type of fluorescence is
along with the approximate quantum yields (QY) and the problem of circumventing fluorescence is of great called autofluorescence. However, fluorescence may also
of the corresponding processes. QY is defined as importance in practical Raman spectroscopy. originate from molecules of impurities of different nature
the number of times the acts of photon absorption, and trace elements in the materials under study.
For the process of fluorescence to arise, a molecule
scattering or reemission occur divided by the number of
must first absorb a photon with enough energy (typically In many practical cases, fluorescence background
incident photons in IR absorption, Raman scattering or
in the visible or ultraviolet range of the optical spectrum) is a major problem when measuring Raman spectra.
fluorescence, respectively. For the sake of simplicity, only
to excite it from the ground to a higher electronic energy Indeed, the intensity of the fluorescence band is
two vibrational levels, the ground v 0 and the first excited
state. Then, the molecule quickly relaxes back down proportional to the amount of the absorbed light, and
v1, are shown for the ground E0 and the first excited E1
to the more stable ground state, releasing the excess the maximum possible QY of fluorescence is 1.0 (100%)
electronic states, and arrows indicate transitions between
energy. The release of the extra energy often occurs when each photon absorbed results in a photon emitted.
these levels. Evirt denotes the short-lived intermediate
through a non-radial transition to a lower vibrational level Fluorophores that provide QY of 0.1 are considered to be
quantum state participating in Raman scattering. The
of the same electronic state, followed by the spontaneous strong fluorescent compounds. Thus, the fluorescence
Raman transitions are shown for Stokes scattering only.
reemission of a photon having less energy than the background can be up to 1010 times more intense than
Typical excitation frequencies in NIR, visible and UV
absorbed photon. The so-called non-radial decay is Raman scattering. Even weak fluorescence from the
regions are marked in the energy (eV) and wavelength
shown by the wavy arrow between v’1 and v’0 vibrational analyte or impurities in the sample is often much stronger
(nm) vertical axes.
levels in Figure 6(e). than the Raman scattering and, therefore, can obscure
the Raman bands.
Fluorescence originates from fluorophores, the chemical
compounds whose molecules typically contain several As it follows from Figure 6, fluorescence only occurs
conjugated aromatic groups or heterocyclic systems if a molecule is excited up to a higher electronic state.
with π-bonds. These parts of the molecules are called Fluorescence can by significantly reduced by using longer
chromophores, and they are responsible for the colors of excitation wavelengths, such as the visible 785 nm or
the chemical compounds. near-infrared 830 and 1064 nm, where the photon does
not have enough energy to excite most of the molecules
to higher electronic states. Another way to minimize
fluorescence is photobleaching by prolonged sample
exposure to excitation light.
Figure 6. The transitions between the vibrational energy levels of a diatomic molecule as the
excitation energy increases from left to right and the approximate quantum yields (QY) of the
corresponding processes: (a) IR absorption; (b) Raman with excitation in the near-infrared region
(NIR-Raman); (c) Raman with excitation in the visible region (Vis-Raman); (d) resonance Raman;
(e) fluorescence; and (f) Raman with excitation in the UV region (UV-Raman). (The figure is
reproduced by permission from Alexander Rzhevskii, Modern Raman Microscopy: Technique and
Practice, Cambridge Scholars Publishing, 2021, 392.)
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Raman Spectrometer

What are the major components of the Raman The common layout of a dispersive Raman microscope An optical microscope (1) is used to observe a sample and
instrument? with all its major parts and optical components is depicted focus the laser beam on it. In the Raman microscope, the
in Figure 7. excitation of the sample and the collection of scattered
Raman instruments are all built around the principles
light are performed in a so-called back-scattering or
developed by the persons who first discovered and
180-degree configuration. The laser (3) emits excitation
observed Raman scattering. Soon after the pioneering
light (the green beam) that is focused by the microscope
experiment with sunlight and light filters, C.V. Raman used
objective (4) onto the sample placed on the sample
a mercury arc lamp as the excitation source and a quartz
stage (5). The light is then scattered by the sample,
spectrograph that allowed the first Raman spectra to be
collected by the same objective and directed through
photographed.
the plate beamsplitter (6) into the spectrograph (2). Thus,
A modern Raman instrument is an integrated system that after being scattered by the sample and collected by
generally combines an excitation source, a spectrometer, the objective, the Raman and Rayleigh components of
a sampling compartment or accessory, and a controlling scattered light share a common path. These components
computer with the software for the acquisition and are shown by the yellow and green dotted beams,
numerical processing of spectral data. Many different respectively. The Rayleigh scattered component is then
configurations of Raman instruments exist depending blocked by the Rayleigh rejection filter (7). An optical
on the application, and these configurations may require interface (optical coupling) between the microscope and
additional parts and optical components. For example, Figure 7: Schematic diagram of a dispersive Raman microscope: spectrometer (8) is used to transfer Raman scattering to
1 – optical microscope; 2 – spectrograph; 3 – laser source; 4
in a Raman microscope, a laser excitation source and – microscope objective; 5 – sampling stage with a sample; 6 – the spectrograph. Raman scattering passes through the
a spectrometer are combined with an optical (light) beamsplitter or dichroic mirror; 7 – Rayleigh filter; 8 – microscope-
spectrograph’s entrance aperture (a pinhole as shown
to-spectrograph coupling optics; 9 – spectrograph entrance
microscope for non-destructive spectrochemical aperture; 10 – collimating mirror; 11 – diffraction grating; 12 – or a slit) (9) and is collected by the collimating mirror (10).
focusing mirror; 13 – array detector. (The figure is reproduced by
analysis of a variety of materials and objects at the permission from Alexander Rzhevskii, Modern Raman Microscopy: The mirror directs the collimated beam onto the diffraction
microscopic level. Technique and Practice, Cambridge Scholars Publishing, 2021, 392.)
grating (11). The grating decomposes Raman scattering
into its constituent spectral components, which are
focused by a second mirror (12) onto the multi-element
array detector (13).
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

How to select the appropriate excitation (laser) For measurements of highly fluorescent materials, it is Also, a combination of a 785 nm laser with a back-
wavelength? common to use NIR laser excitation with a 1064 nm illuminated CCD detector in the Raman spectrograph, as
Interest in Raman spectroscopy increased dramatically in wavelength. Although 1064 nm excitation ensures explained below, may result in an unwanted interference
the late 1960s after the invention and commercialization virtually fluorescence-free Raman spectra, it results in a pattern in the Raman spectrum.
of lasers as new monochromatic and powerful sources theoretical decrease in scattering efficiency by a factor of
Depending on the application and the material under
of excitation light. Lasers have completely replaced the 16 compared to the green 532 nm wavelength.
investigation, using several excitation wavelengths and
mercury lamps that were used as light sources in the early
Excitation in the UV spectral region is attractive for switching between different wavelengths in a Raman
days.
resonance Raman spectroscopy on biomolecules, instrument are considered to be a substantial advantage.
One of the main advantages of Raman spectroscopy such as proteins, DNA and RNA. It also provides a Figure 8(a) provides an example of the application of
is that the appearance of the measured spectrum is unique opportunity to avoid both the fluorescence and different excitation wavelengths to characterize the
independent of the excitation wavelength (see Figure 3). interference of thermal emission at high temperatures diameter distribution of single-wall carbon nanotubes
Nevertheless, the appropriate choice of laser excitation in studying catalysts and catalytic processes in situ. (SWCNT) in a mixture using the position and intensity
wavelengths is necessary to obtain the most informative Moreover, because of the 1/λ⁴ dependence, Raman of their Raman radial breathing modes (RBMs). Further
Raman spectrum of a given sample. scattering efficiency is higher in the UV region. analyzing their Raman spectra can predict the chirality
Nevertheless, UV excitations in Raman spectroscopy and semiconducting or metallic behavior of the SWCNTs.
As we know from equation (1), the scattering intensity is
need to be applied with care since thermal decomposition Figure 8(b) shows the Raman spectra of graphene grown
inversely proportional to the fourth power of the excitation
or other laser-induced modifications of samples are by the chemical vapor deposition (CVD) process on
wavelength, 1/λ⁴. Therefore, Raman spectroscopists
possible. a copper foil. Note that green 532 nm laser excitation
prefer shorter laser wavelengths, such as blue 488 nm,
generates a significant baseline humpback due to the
green 514 and 532 nm or red 633 nm, to obtain a It is well recognized that the best trade-off between
fluorescence of the copper substrate, whereas blue
stronger Raman signal and, respectively, a better signal- Raman scattering efficiency, fluorescence avoidance,
455 nm excitation avoids fluorescence resulting in a flat
to-noise ratio (S/N) in the spectrum provided that all laser-induced effects and compatibility with the detector’s
baseline with well observable Raman bands.
the other experimental conditions are equal. Blue and sensitivity can be achieved with the dark red 785 nm laser
green wavelengths are typically reasonably suitable for wavelength. It should be noted that 785 nm excitation
transparent liquids and polymers, non-colored inorganic causes a significant fluorescence in glass commonly used
materials and some resonance Raman experiments. for laboratory glassware. A broad fluorescence band
However, as the Raman effect is an extremely weak centered at about 1385 cm-1 originates from rare earth
phenomenon, a Raman signal in the visible optical range elements and certain impurities in the glass matrices.
can easily be overwhelmed by autofluorescence from an Therefore, it is recommended to avoid cuvettes, vials,
analyte or even fluorescence from trace impurities in the microscope slides, coverslips and substrates made of
sample under study. glass when using 785 nm excitation, as the fluorescence
may significantly distort the measured Raman spectrum.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Figure 9: The effect of increasing 532 nm laser power on C₆₀.

DXR Raman instruments are equipped with variable

neutral density filters set at the laser output. The
continuously variable opacity of the filters allows
Figure 8: The influence of different excitation wavelengths on Raman spectra. (a) Raman spectra of a mixture of SWCNTs in the region of
manifestation of RBMs measured with 455 (blue), 532 (green) and 633 (red) nm excitation wavelengths (the spectra are offset for clarity). (b) accurate and reproducible light power with 0.1 mW
Raman spectra of multi-layered graphene on a copper foil measured using 532 (green) and 455 (blue) nm wavelengths. The principal bands of increments to be delivered to the sample. It may be of
graphene are marked as G, D and 2D. The bands in the region below 800 cm-1 belong to copper oxide formed on the surface of the
copper foil. critical importance when light-sensitive samples such as
biological cells and tissues, carbon nanomaterials, and
In practice, when working with various applications and Why is fine control of laser power at the sample many others are being measured.
materials, Raman spectroscopists always want access to important?
as many excitation wavelengths as possible to perform Since a laser beam creates a very high density of light How does spectrograph work?
Raman experiments successfully and optimally. Therefore, power in focus, it is important to protect sensitive samples An optical spectrometer performs a spectral (harmonic)
flexibility and optimization of Raman experiments from any thermal or photochemical damage. For some analysis of light. This analysis is most often carried out
are achieved in high-end Raman spectrometers and materials, it is possible to damage or alter the sample using a dispersing element (usually, a diffraction grating),
microscopes with multiple lasers, interchangeable with laser radiation. This damage can be very obvious thereby deflecting the light rays of different wavelengths at
Rayleigh filters and diffraction gratings. in extreme cases where the laser radiation burns a hole different angles. A spectrometer with a dispersing element
in the sample. In other cases, the damage can be more is called dispersive, while a spectrometer that performs
DXR Raman instruments can be equipped with 455, 532,
subtle, and if care has not been taken to avoid damaging spectral analysis based on the principle of optical
633 and 785 nm laser sources.
the sample, it may result in Raman spectra that do not interferometry, such as the Michelson interferometer in the
represent the actual sample. Such spectral data could FT-Raman instrument, is called non-dispersive.
easily be misinterpreted. Figure 9 provides an example of
A typical dispersive spectrometer consists of five basic
one such situation with a sample of C₆₀ fullerene. Here
components: 1) an entrance aperture, 2) collimating
we can see that the C₆₀ begins to breakdown into other
optics, 3) a diffraction grating, 4) focusing optics, and 5) a
structures, probably amorphous carbon, with as little as
detector.
0.5 mW of the 532 nm laser power at the sample.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The most widely used spectrograph configuration for

Raman spectroscopy is based on the Czerny–Turner
design outlined in Figure 10(a). The entrance aperture
may be a slit, a pinhole or the output end of an optical
fiber, and the spectrum can be essentially regarded
as a discrete number of monochromatic images of the
entrance aperture in the focal plane of the spectrometer.
However, as will be considered below, the aperture
confines an element of the sample from which the
spectrum is measured. Thus, the aperture performs two
main functions: it provides both the necessary spectral
and spatial resolution.

The Czerny–Turner spectrograph employs concave

collimating and focusing mirrors and a planar reflective
diffraction grating. The collimating mirror collects the
light diverging from the input aperture, forms a collimated
Figure 10. Czerny-Turner (a) and triplet (b) spectrographs: A – pinhole aperture, M1 – collimating mirror, M2 – focusing mirror, M3 – parabolic
beam of light and directs it to the diffraction grating. The mirror for aberration correction, GR – rotational turret with diffraction gratings.
collimating optics ensure that the full area of the diffraction
grating is illuminated, which is necessary for the optimal The monochromator makes one measurement of the The triplet mirrors spectrograph used in DXR Raman
operation of the grating. The diffraction grating separates light intensity at a particular wavelength at a time. If a instruments and shown in Figure 10(b) employs an
the light rays in space as a function of wavelength so that multichannel (array) detector is placed in the focal plane, additional focusing parabolic mirror to minimize
each constituent wavelength component is deflected then all the parallel rays of different wavelengths can be aberrations. The focusing optics with two mirrors also
under a certain angle, but all rays of the same wavelength registered simultaneously, and the instrument works as make the spectral resolution less dependent on the
travel under the same angle. The dispersed rays are then a spectrograph. The spectrograph measures the light excitation wavelength, resulting in a configuration superior
collected by the focusing mirror, which forms an image intensities at every wavelength and records the entire to the Czerny-Turner spectrograph with a single focusing
of the entrance aperture by focusing the light rays of spectrum in a single snapshot. mirror.
different wavelengths at the exit focal plane. The detector
The Czerny–Turner design offers flexible and easily There are two important merit figures that allow the
placed in the exit focal plane generates an analog
configurable setups to fit specific instrument and evaluation and comparison of the spectrometers’
electrical signal proportional to the intensity of the light
application requirements. However, the primary issue for performance: optical throughput and spectral resolution.
that hits the detector. If the rays of different wavelengths
a classical Czerny–Turner spectrograph with collecting The presence of an entrance aperture, which restricts the
are scanned over the detector through an exit slit
and focusing mirrors placed symmetrically relative to light flux passing through the spectrograph, results in a
sequentially by rotating the diffraction grating, then the
the grating is astigmatism which results in the spectral well-known trade-off between the spectral resolution and
instrument operates as a monochromator.
resolution varying across the spectrum and depending on optical throughput of the dispersive spectrograph.
the excitation wavelength.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

What is the resolving power of diffraction grating? Thus, in order to increase the resolving power of the The steps are sloped with respect to the grating surface
The most important component of a dispersive grating, it is necessary to increase N. Accordingly, for the at the so-called blaze angle, which determines the
spectrograph is the diffraction grating that provides grating of a given width, the grating constant needs to direction in which the maximum efficiency is reached.
angular dispersion of analyzed light, i.e., spatially be increased. It is important to note that the collimating Therefore, the grating appears to “blaze” when viewed
separates its constituent wavelengths. mirror in the spectrometer should be aligned to illuminate from that direction. The purpose of the slope of the steps
the grating’s full width to provide the theoretically is that the blaze angle θB can be chosen in such a way
A typical diffraction grating consists of a substrate
achievable resolving power. that the light rays diffracted by the grating, and the rays
made of an optical material, with a large number of fine,
reflected from the facets of the steps are both deflected
equidistant, parallel lines (grooves) created on its surface.
How does blazed grating increase its diffraction in the same direction. Consequently, the angle of
In a reflective grating, the grooves are created on a
efficiency? diffraction coincides with the angle of specular reflection
reflective surface of the usually metal-coated glass so that
The diffraction grating efficiency can be maximized for a that significantly increases the diffraction efficiency at a
light reflects only between the grooves.
particular wavelength such that most of the light power designated wavelength. The wavelength for which the
The number of grooves per unit length (grove density), is concentrated into the first diffraction order k =1 for that blazed grating is most efficient is therefore specified
usually given in gr/mm, is called the grating constant. wavelength and reduced for all other wavelengths. The as the blaze wavelength. As a result, the efficiency can
The gratings used for applications in the visible spectral efficiency optimization can be achieved by appropriate reach over 80% for the specified wavelength. Figure 11
range typically have from 600 to 2400 gr/mm. For longer groove shaping, i.e., by making a sawtooth groove profile shows the reflective blazing grating principles and typical
wavelengths, the grating constant should be about to form a step structure. efficiency curves vs. wavelength for ruled and holographic
300 gr/mm or less. blazed gratings.

The grating constant affects both the wavelength region in

which the grating operates and the dispersion properties
of the grating. The higher grating constant results in
higher resolving power and greater dispersion of the
grating. The resolving power R of the grating is defined
in terms of wavelengths λ and the spectral resolution δλ,
the minimal difference between the wavelengths that can
be resolved. The resolving power R is a dimensionless
number. It can be theoretically shown that R is a product
of the order of diffraction k in which the grating is used
(typically, k =1), and the number of grooves N illuminated
on its surface by the incident light:

Figure 11. Simplified cross-sections of blazed ruled (a) and holographic (b) diffraction gratings and the typical efficiency curves (c) for ruled
blazed at 500 nm (green), holographic blazed at 500 nm (orange) and ruled non-blazed (red) gratings. (The figure is reproduced by permission
from Alexander Rzhevskii, Modern Raman Microscopy: Technique and Practice, Cambridge Scholars Publishing, 2021, 392.)
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The corresponding blazed gratings accompany all the This is why silicon CCD sensors are not useful detectors In a typical dispersive Raman spectrograph, constituent
lasers in DXR Raman instruments to increase the optical for Raman spectroscopy using long wavelength NIR wavelength components of analyzed light are projected
throughput and sensitivity of the spectrometers. lasers. The conversion rate of incident photons to electron onto the CCD array’s long axis. Thus, the longer horizontal
charge is known as quantum efficiency (QE). The higher direction along the rows is ascribed to the dispersion or
How do detector options impact Raman spectra? the QE, the stronger the signal produced by a given spectral axis, and the vertical direction along the columns
The focusing mirror ultimately collects the light diffracted number of photons. If noise is assumed to be constant, - to the spatial axis, as shown in Figure 12(a).
by the grating onto a detector mounted in the focal plane detectors with higher QE will produce Raman spectra with
The major configurations of the CCD detectors for Raman
of the spectrograph. A charge-coupled device (CCD) higher signal-to-noise (S/N) values.
spectroscopy include front-illuminated (FI) and back-
is now one of the most commonly used detectors in
The pixels are arranged in rows and columns. The illuminated (BI) ones. The principal cross-sections of these
Raman spectroscopy. The CCD’s major advantage is
matrices of CCD detectors for Raman spectrographs have CCDs and their typical QE characteristics are illustrated in
its high sensitivity to light, which makes it very suitable
a rectangular shape and typically comprise 1024, 2048 Figures 12(b,c) and (d), respectively.
for detecting inherently weak Raman scattering, and a
or 4086 vertical columns and from 128 to 512 horizontal
multichannel principle of operation, which allows the full
rows with a pixel size ranging from 13.5 μm to 26 μm.
Raman spectrum to be obtained by the simultaneous
measurement of its spectral components. By the full
Raman spectrum, we mean the spectrum in the range
of wavenumbers corresponding to the fundamental
vibrations of the main chemical groups.

A CCD detector consists of a large number of

photosensitive elements arranged in a two-dimensional
array on the silicon substrate. The discrete elements of
the array are called pixels, which is a word that is derived
from “PICture ELement”. The pixels are defined in the
silicon matrix by a grid of current-conducting polysilicon
pads called gates. Electrical electrodes are connected
to the gates. The entire structure is composed of gates,
metal electrodes, an insulating silicon oxide layer and a
silicon semiconductor substrate. This structure is known
as a metal-oxide-semiconductor (MOS) capacitor, with a
photoactive region, also known as the depletion region, to
convert photons to electron charge. Shorter wavelengths
of light are absorbed close to the surface, while longer
wavelengths travel deeper before being absorbed. If the Figure 12. Two-dimensional matrix of CCD detector (a) and simplified cross-sections of front-illuminated (b) and back-illuminated (c)
CCD detectors: 1 - gates and electrodes, 2 –insulating layer made of SiO₂, 3 - photosensitive region. The incident photons of different
wavelengths of light are long enough (>1050 nm) they do wavelengths and their depths of penetration are indicated by the corresponding colors’ arrows. The étaloning effect in BI CCD created by
deep penetration of long light wavelengths (dark red) through the photosensitive region is shown as multiple reflections from the boundaries
not interact with the depletion region and do not produce
of the region. Typical QE characteristics of BI and FI CCD detectors (d): To illustrate how the variation of QE curves with wavelength would
a signal. modulate the intensities of Raman bands, the spectra of polystyrene in the range of approximately 50–3300 cm -1 for the Stokes region are
shown for the typical excitation wavelengths of 455, 532, 633 and 785 nm.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The differences between FI and BI CCD sensors arise parallel and reflective surfaces. Thus, the reflected light When calculating the total transmission of a dispersive
from how the incident photons get to the photoactive creates the étaloning effect, generating an interference spectrometer, it is necessary to consider all its optical
region of the sensor. fringe pattern. Respectively, the light intensity modulation components’ efficiency. In the dispersive spectrometers,
in the fringes produces the corresponding charge the total number of surfaces on which reflection or
In a FI CCD, as shown in Figure 12(b), light photons
modulation in the photosensitive region. This effect can transmission occurs often reaches 14. If we consider the
impinge on the sensor’s front surface, which, as
lead to interference fringes superimposed on a normal loss on one surface to be 5% on average, then the total
mentioned above, is overlaid with a grid of metalized
Raman spectrum, particularly noticeable in the red and transmission will be
electrodes and polysilicon gates. In this design, the
NIR regions. Thus, if a Raman instrument is equipped with
electrodes and gates reflect or absorb some of the
BI CCD, then the use of excitation wavelengths of 785 nm
photons and thus limit the number of photons that actually
and longer is not recommended.
reach the photosensitive region. Since fewer photons get
The optical throughput of a spectrometer has a direct
to the photosensitive region, this typically limits the QE to DXR Raman spectrometers can be configured with either
impact on the sensitivity of the Raman instrument and the
around 50–60%. FI or BI CCD.
amount of time needed to acquire Raman spectra with
Alternately, the photons can impinge upon the back an adequate S/N ratio. When evaluating the final optical
What does the optical throughput of a spectrograph
side of the sensor (hence, back-illuminated CCD). Since throughput of a Raman spectrometer, it is essential to
depend on?
the back of the CCD matrix has no electrodes on it, the take into account the efficiency of the CCD detector that
The ratio of the light flux passing through the entrance
surface is clear of any obstacles for the photons to enter was discussed above, as well as the quality of all the
aperture of the spectrograph to the corresponding flux
the photosensitive region. However, in order to make this spectrograph’s components.
incident on the detector is called the optical throughput
work effectively, it is necessary to remove, mechanically
of the spectrograph. The practical optical throughput
or via chemical etching, some of the bulk silicon substrate
is determined by the losses of light energy during
to provide access to the photoactive region. This is why
absorption, reflection and scattering by the optical
these types of CCD sensors are also known as back-
elements of the spectrograph.
thinned devices. BI CCDs can attain QEs up to 95%.
Optical materials used in Raman spectrographs in the
The higher QE of BI CCDs would appear to provide
wavelength range for which they are intended usually have
a clear advantage. However, there is a drawback.
low absorption coefficients so that the absorption losses,
In contrast to the shorter light wavelengths, which
for example, in lenses, do not exceed a few %. Reflection
are absorbed close to the silicon surface, the longer
losses are mainly associated with the reflection from the
wavelengths penetrate much deeper into the silicon
surface of optical mirrors used in spectrographs.
before being absorbed. This deeper penetration can
result in longer wavelengths of light passing all the way
through the thinned region and causing constructive and
destructive interference by multiple internal reflections at
the boundaries of the photosensitive region. This effect
is analogous to one in a Fabry-Pérot étalon consisting of
a thin, transparent optical medium with two flat, highly
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

What does the spectral resolution of the spectrograph When both broadening mechanisms are presented What is the difference between the spectral and digital
depend on, and what is the resolution practically simultaneously, which often occurs, the line shape can resolution of a spectrograph?
required? be given by a convolution of the Lorentzian and Gaussian Spectrograph manufacturers often prefer to specify the
The spectral resolution characterizes the ability of an functions resulting in a Voigt profile. resolution in cm-1 per pixel of the CCD detector. This
instrument to separate two closely spaced spectral parameter is termed the digital or pixel resolution and
In general, spectrographs introduce distortion into the
lines. The practical spectral resolution of the dispersive usually corresponds to the point spacing in the digitized
intrinsic spectral profiles. If the width of the intrinsic
spectrometer is determined by the instrumental function spectrogram. The digital resolution is related to the
spectral line significantly exceeds that of the instrumental
(profile). The narrower the instrumental function, the Nyquist-Shannon sampling theorem that states that the
function, then the shape and width of the line recorded by
higher the spectral resolution that can be achieved. The sampling frequency should be greater than twice the
the spectrometer practically coincide with the parameters
instrumental function depends on the following major highest frequency in the spectrum. The highest frequency
of the intrinsic line. If the width of the instrument function
parameters: that the spectrograph can detect is determined by its
constitutes no more than ¼ of that of the intrinsic spectral
spectral resolution. Consequently, at least two CCD pixels
• the resolving power of the grating: the higher the line, then the broadening of the experimentally registered
must fall within the spectral resolution value. Otherwise,
groove density, the higher the spectral resolution; line does not exceed 5%. A spectrograph with a practical
the size of the pixels will limit the resolution of the
• the focal length of the spectrometer: the longer the resolution of about 4 cm-1 is capable of registering
spectrograph. Therefore, the digital resolution is two times
focal length, the higher the spectral resolution; an intrinsic linewidth of at least 20 cm-1 without any
higher than the spectral resolution.
noticeable distortion.
• the width of the entrance aperture of the
spectrometer: the narrower the aperture, the higher Since the half-widths of the bands in the spectra of
the spectral resolution; solids rarely appear to be narrower than 10 cm-1 and,
• the pixel size of the CCD detector: the smaller the in the spectra of liquids, mostly exceeding 20 cm-1,
pixels, the higher the spectral resolution. spectrographs that provide spectral resolution in the
range of 2–5 cm-1 are generally well suited for analyzing
The vibrational spectral lines of materials in a condensed
samples in a condensed phase. As a rule, compact
phase (liquids and solids) are broadened significantly
Raman spectrometers equipped with spectrographs
due to different mechanisms of interactions between the
having a focal length between 0.2 m and 0.3 m contain
neighboring chemical groups and molecules and their
fewer reflective and refractive optical components,
strength (for example, the Van der Waals mechanism or
increasing their optical throughput. The choice of high
hydrogen bonding). As a result, the typical vibrational
spectral resolution is justified for specialized applications
spectra of substances in a condensed state consist of
or the analysis of gas samples, which is a rather rare task Figure 13. Two closely spaced spectral lines (red and green)
individual spectral lines or closely spaced overlapped projected onto the detector array in analog form (a) and converted to
for Raman spectroscopy. digital form (b and c). The position of the maxima of the lines differs
lines that form spectral bands with bandwidths ranging by less than one pixel. The pixel filling histograms shown in (b) and
from several to a few hundred cm-1. All the broadening (c) for the red and green lines, correspondingly, demonstrate the
coincidence of the histograms for the lines fitting a single vertical
mechanisms form the so-called intrinsic line shape. pixel column (left) and their mismatch as the lines become broader
(middle and right). (The figure is reproduced by permission from
These mechanisms are divided into homogeneous and Alexander Rzhevskii, Modern Raman Microscopy: Technique and
inhomogeneous ones and are fairly well described by Practice, Cambridge Scholars Publishing, 2021, 392.)

the Lorentzian and Gaussian functions, respectively.

 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Figure 13 illustrates how two closely spaced spectral What is the benefit of the spectrograph with fixed-
lines are projected onto the detector array in analog form position diffraction grating?
(a) and conversed to digital form (b–c). The position of Several diffraction gratings, often with different groove
the maxima of the lines differs within one pixel. Suppose density and, consequently, resolving power, is usually
both the lines are so narrow that they fit entirely in one mounted on a rotational turret. A triple-grating turret is a
vertical column of pixels. In that case, such lines cannot standard option available in many Raman spectrographs.
be resolved because, regardless of the exact position of The drive mechanism rotates the turret to select a grating
the line’s maximum within the pixel, the total charge is and turns the selected grating around its pivot axis to set
only accumulated and read out from this column of pixels. an operational wavelength. As an example, Figures 14(a)
If the full width at half-maximum (FWHM) of these lines and (b) illustrate the process of acquisition of a Raman
spans two pixels, as required by the Nyquist-Shannon spectrum of polystyrene by sequentially projecting two
theorem, these spectral lines are resolved. Indeed, the halves of the spectrum onto CCD by turning a diffraction
digital projections of these lines on the detector, as shown grating GR1. When the turret rotates and replaces the
by pixel-filling histograms in the columns, do not coincide. grating GR1 by GR2 with half the dispersion, the full
When the spectral lines get broader so that their FWHMs spectrum is projected onto the CCD and acquired in one
occupy several pixels, their maxima’s position can pass, but with half the spectral resolution (c).
nevertheless be determined with adequate accuracy, as
This practice of turning the grating, acquiring the Figure 14. Acquisition of a Raman spectrum using diffraction
long as the corresponding histograms do not completely gratings on a rotational turret. In (a) and (b), two halves of a
spectrum in multiple spectral regions and stitching the spectrum are sequentially projected onto CCD and acquired using
coincide with each other. Thus, in contrast to the often- the diffracting grating GR1. In (c), the full spectrum is projected
regions together using spectroscopic software is often
repeated misleading statement, determining an accurate and acquired using GR2 grating with half the dispersion and,
applied in Raman instruments. Despite the significant respectively, half the spectral resolution. The Raman spectrum of
position of the band and its shift in the wavenumbers polystyrene is shown as an example. (The figure is reproduced by
advantages of multi-grating turrets, moving the gratings permission from Alexander Rzhevskii, Modern Raman Microscopy:
scale does not require spectral resolution equal to or Technique and Practice, Cambridge Scholars Publishing, 2021, 392.)
has a few shortcomings. First, it takes time to move
better than the value of the assumed shift. The position
between grating positions and, if a broad spectral These shortcomings are addressed by using a fixed
of a spectral band can be found quite accurately by
coverage is required, collecting multiple regions may position grating that directs the full spectral range of
calculating the so-called center of gravity of the band or
be too time-consuming for fast spectral imaging or most interest in Raman spectroscopy (for example, from
from a curve-fitting procedure.
reaction monitoring applications. Second, as the different 50 to 3500 cm-1) to the array detector while maintaining
DXR Raman spectrographs offer a practical spectral regions of the spectrum are acquired at different times, adequate spectral resolution.
resolution of better than 5 cm (2 cm /pixel) for the
-1 -1
continuous reaction kinetics measurements appear
All DXR Raman spectrographs contain no rotating turret
full Raman spectrum and 2 cm-1 (1 cm-1/pixel) for the to be problematic. Third, the grating mechanical drive
but fixed-position interchangeable blazed-angle gratings.
half-spectral range. This resolution is sufficient for the inevitably yields errors in the accuracy and repeatability
qualitative and quantitative spectral analyses of the of the grating position when turning the grating that,
majority of liquid and solid materials. correspondingly, introduces inaccuracies in optical
performance and the Raman spectra.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Raman Microscopy

What is Raman microscopy?

In a Raman microscope, a research-grade optical microscope is coupled to
a laser excitation source and spectrometer, thereby constituting a scientific
instrument capable of obtaining both traditional optical and spectrochemical
images with a spatial resolution at the diffraction limit of light (<1 micron). The
Raman microscope enables the measurement of chemically and spatially
inhomogeneous samples to reveal their chemical composition, morphology,
molecular orientation, conformation, polymorphism, crystallinity, material
deformation, local temperature, etc. A 3D spectrochemical representation of
the sample can be constructed by acquiring 2D area images at different depths
in a confocal mode by sequentially shifting the laser focus inside a transparent
or translucent sample. This method of 3D spectrochemical imaging preserves
the sample integrity and avoids possible artifacts that may result from physically
cutting or sectioning the sample. No other microscopic technique offers such a
combination of spatial resolution and “information-reach” spectral content within
a workable acquisition time.

The principle diagram and a description of a dispersive Raman microscope are

given above (see Figure 7).

What is the role of a confocal aperture (pinhole)?

The paths of the light rays in the Raman microscope and the function of a
confocal aperture are illustrated in Figure 15. The light rays from the laser and
the rays collected by the microscope objective from the focused laser beam are
denoted in green, while the rays entering the spectrometer are yellow. As the
circular aperture (pinhole) is placed in the image plane of the microscope, the
Figure 15. The role of a pinhole in a confocal Raman microscope. The paths of light rays emitted by the
pinhole is confocal with the focal point. The principal role of the confocal pinhole laser and collected by microscope objective (green) and the rays “observed” by the spectrograph through
is to filter the light rays collected by the microscope objective spatially and to a confocal pinhole (yellow) are shown in the center, where BS is a beamsplitter, and L1 and L2 denote
the focusing lens and the laser beam expander lens system, respectively. (The figure is reproduced by
pass the selected rays through the pinhole to the spectrograph. permission from Alexander Rzhevskii, Modern Raman Microscopy: Technique and Practice, Cambridge
Scholars Publishing, 2021, 392.)
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The laser beam focused by the objective converges in In a non-confocal mode, when the aperture is a slit, A proper confocal configuration of a Raman microscope
the form of a cone in the focal plane and then diverges the Raman spectrum measured by the spectrometer means the spheroid is well centered within the hourglass
in the same way. The volume illuminated by the focused represents Raman signals emanating from the entire shape, as illustrated above. The process of bringing
laser beam can be depicted as a green hourglass hourglass shape. Suppose this shape comprises the laser focus and the volume element “observed” by
shape, as shown in Figure 15(c). The objective collects chemically different and spatially separated the spectrometer into a perfect coincidence is known
the light scattered from this hourglass shape; it is then microdomains. In that case, the recorded Raman as alignment. The regular alignment compensates for
magnified and projected onto the image plane. The light spectrum is the superposition of Raman spectra of all variations caused by the inevitable drift over time. A
rays emanating from the converging and diverging cones these microdomains, and they will not be discriminated correctly aligned confocal Raman microscope provides
above and below the focal plane are combined to form spatially. A significant improvement in the spatial the best spatial discrimination at a high intensity of the
a converging green cone in the image plane. When a discrimination is obtained in both the lateral (perpendicular Raman spectrum.
pinhole is placed in the image plane, it blocks all the rays to the optical axis of the objective, i.e., in the horizontal XY
Calibration is another procedure that ensures the
emanating from the out-of-focus regions and allows only plane) and axial (along the optical axis, i.e., in the vertical
accuracy of the experimentally measured Raman
the rays from a small volume around the focal point to Z-axis) directions in the confocal mode of operation.
spectrum with the correct Raman shift and spectral
pass through. The small volume can be represented by
By varying the diameter of the confocal aperture, intensity. Unlike FT-IR instruments with a built-in “internal
the prolate spheroid “spread” out along the optical axis.
operators of Raman microscopes can select the volume standard,” a laser that monitors the optical path difference
This prolate spheroid is highlighted in yellow in Figure
element to measure at a given objective magnification in the interferometer and provides accurate wavenumber
15(a). Since the image and sample plane are conjugated,
and optimize the intensity in the Raman spectra of the measurements, dispersive Raman instruments must
the spheroid can then be projected back onto the sample
materials under study. A confocal aperture with the ability use samples or standards that have spectral peaks with
plane to visualize the small volume element from which
to continuously vary its size is often called a confocal known Raman shifts for calibration. Raman spectra
the spectrograph measures the scattered light. Thus, the
diaphragm. The confocal diaphragm is advantageous for obtained with different lasers may have significant
confocal aperture ensures that only the light emanating
precise confocal Raman microscopy. variations in the relative peak intensities, mainly due
from the spatially discriminated volume depicted by
to wavelength-dependent differences in the quantum
the yellow spheroid inside the green hourglass shape
How important are optical alignment and calibration? efficiency of the CCD. The intensity difference can
in Figure 15(b) passes from the laser focus to the
Optical alignment and calibration are two important be adjusted using the so-called white-light correction
spectrometer.
practices that must be routinely executed for the optimal technique. The white-light correction is an intensity
The effect of the attenuation of the out-of-focus regions performance of any high-end Raman microscope. normalization procedure that may utilize a broadband
represents spatial filtering and can be described in terms white-light radiation standard applicable to multiple
In the course of time, all Raman microscopes are subject
of the point spread function (PSF). The volume element excitation wavelengths. This procedure establishes a
to alignment and calibration drifts. These drifts may be
measured in the confocal mode is determined by the scaling factor to correct the intensities depending on the
caused by laboratory temperature fluctuations, external
so-called effective optical PSF. It is a product of the wavelength along the Raman shift axis.
disturbances of the instrument, normal wear of system
hourglass shape illuminated by the laser and the volume
components occurring during regular operation, and
of the spheroid “observed” by the spectrograph through
many other factors.
the pinhole. An effective PSF constitutes one of the
essential properties of the confocal Raman microscope.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Manually performed alignment and calibration routines What does spatial resolution in confocal Raman
usually require a certain level of qualification for the microscopy depend on?
maintenance personnel and additional materials and When a microscope objective focuses the laser beam,
sources; it also often increases the downtime of the the illuminated volume can be determined by the space
instrument during the operation. enclosed by an hourglass shape shown above in Figure
15(c). This volume is characterized by its “waist” or
the diameter of the laser spot in the XY focal plane
approximated by the Airy disk:

(11)

where λ is a laser excitation wavelength, and NA is the Figure 17. Diffraction-limited focused laser spot produced by
Figure 16. The alignment and calibration tool to be placed on the numerical aperture of the microscope objective. objective lenses with low (a) and high (b) NA: D is the diameter of the
microscope stage (a), and the bright pinhole of the tool in the optical laser beam that illuminates the lens; f1 and f2 – focal lengths;
focus centered on the microscope eyepiece crosshair (b). α1 and α 2 – the half-angles of the cone of light that exits the lens;
Hence, the higher NA, the smaller the spot that can be d1xy and d 2xy – diameters of the laser spots for low and high NA,
produced by the objective passing the light of a given respectively. The diffraction pattern of focused light in the plane
perpendicular to the focal plane with the central Airy disk and a
In DXR Raman microscopes, alignment and calibration wavelength. The spot size is directly related to the lateral few sidelobes is shown in dark green. (The figure is reproduced by
permission from Alexander Rzhevskii, Modern Raman Microscopy:
are highly automated procedures, simplifying their spatial resolution in the Raman microscope since the Technique and Practice, Cambridge Scholars Publishing, 2021, 392.)
implementation and eliminating user error or subjectivity. incident light is focused in, and the scattered light is
Assuming that the image plane coincides with the
These procedures are performed in a software- collected from this spot. The lateral spatial resolution is
spectrograph entrance, the beam diameter will be 72 μm
controllable manner using a dedicated matchbox-sized generally defined as the shortest distance between two
at the spectrograph entrance. If a confocal pinhole
tool placed on the microscope stage. The user handles objects in the focal plane, which can be resolved. The
with a diameter of 50 μm is set at the entrance of the
this tool as a sample and focuses on a small pinhole in the spatial resolution is evidently better and is termed “higher”
spectrograph, then only part of the light from the 72 μm
center of the box, as shown in Figure 16. when the laser spot is smaller. A sub-micron lateral
beam spot can pass through the pinhole and reach out
spatial resolution can be achieved for visible excitation
to the detector. Thus, a pinhole size that is less than the
wavelengths using objectives with a higher NA. The
diameter of the laser spot projected onto the image plane
relationship between the laser spot diameter and NA is
will improve the lateral spatial discrimination by a factor of
illustrated in Figure 17.
about 1.4. However, the improvement in the lateral spatial
Consider an example of the diffraction-limited illumination discrimination is accompanied by a significant loss in the
of the sample with a 532 nm excitation light using an intensity of the Raman signal, and factor 1.4 is considered
objective with NA= 0.9 and 100× magnification (100×/0.9). to be an optimally achievable improvement in practice.
According to equation (11), the diameter of the focused
laser spot will be 0.72 μm at the sample plane. The spot
will be projected onto the image plane with magnification
by a factor of 100.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The depth of the region illuminated by a laser in the axial Z Nevertheless, the confocal depth determined by the major
direction can be estimated using the following equation: axis of the prolate spheroid shown in Figure 15(b) and
Figure 18 may be approximated by the following equation:

(12)

Dividing (12) by (11), the ratio of d z to dxy is derived as

(13)

where DPH represents the diameter of the pinhole

projected back onto the sample plane, which, in general,
When the laser beam is focused in air, the depth of focus can be calculated in micrometers if one knows the total
appears to be more than twice the size of the laser spot magnification factor of the microscope with a given
since the maximum NA of an objective operating in the air objective.
does not exceed 0.95:
As it follows from equation (13), for the excitation
wavelength λ and the objective with a given NA, the first
term under the root is constant, and the confocal depth
Figure 18. Diameter d xy and depth of focus d z of the focused laser
is influenced exclusively by the pinhole diameter in the beam (a) and the confocal depth d psf (b). The height d z is determined
as two z r (Rayleigh range), the distance between the two planes at
second term. each side of the focused spot where the width d of the beam is a
However, the illuminated region in the Z direction from factor of √2 larger than it is at the focus and, consequently, the area
which the Raman spectrum is collected in a condense- The depth discrimination improves (the confocal depth of the beam cross-section is doubled, and the brightness is taken
to be equal to one-half of the maximal brightness of the beam at
phase sample has to include the refractive index of the decreases) as the excitation wavelength shortens and its focus. (The figure is reproduced by permission from Alexander
Rzhevskii, Modern Raman Microscopy: Technique and Practice,
sample n so that: the pinhole diameter reduces. It is logical to assume that Cambridge Scholars Publishing, 2021, 392.)
making the pinhole as small as possible is the best option
Thus, using the confocal pinhole improves both the
for depth discrimination. However, as the pinhole size
lateral and axial spatial discrimination and particularly
reduces, the amount of light collected from the sample
accounts for the ability of the confocal Raman microscope
and reaching the detector also decreases. Because the
to provide optical sectioning of a sample. In order to
Raman scattering is intrinsically weak, pinhole apertures
achieve higher spatial discrimination, a shorter excitation
In the confocal arrangement, the height of the sampling with diameters smaller than 25 µm are rarely used and
wavelength, an objective with higher NA and a smaller
volume element, called the confocal depth, can be are only practical for materials featuring strong Raman
confocal pinhole should be used.
estimated using a PSF. Practical calculation of an effective scattering cross-sections σ.
PSF is a complicated task that requires taking the It needs to be emphasized that, in practice, the confocal
details of the optical design of a given confocal Raman aperture does not restrict the sampling volume element
microscope, the quality of its optical components, the to a prolate spheroid with a sharp boundary depicted in
characteristics of the laser beam and many other factors Figures 15(b) and 18(b).
into account.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The measured Raman intensities may fall off fairly slowly This technique is referred to as Raman optical sectioning The second step is spectrochemical image generation.
on either side of the focal plane in the axial direction. The because it does not require mechanical cutting or The image is constructed by processing the array of
relatively slow decay in the Z-axis is in marked contrast sectioning of the sample. Collecting maps successively spectral data in the corresponding maps in various
to the relatively tight confinement in the XY plane. This at different probing distances with typically a regular ways, e.g., as parameters of an individual spectral peak,
slow decay means that in certain circumstances, one can interval along the vertical Z-axis, a 3D representation of the ratio of peak intensities, correlation with a reference
detect significant out-of-focus Raman signals a long way the sample may be constructed. In general, the maps may spectrum or based on more complex chemometric and
above and below the nominal focal plane, which are well be collected along a line or from an area in both horizontal statistical methods of the analysis commonly used in
outside the calculated confocal depths. and vertical planes or a sequence of areas along the optical spectroscopy. A meaningful number of Raman
vertical axis to render a volume of the sample. images generated from a map is determined by the
What is the difference between a Raman map and a distinguishable spectral features associated with the
Raman image? different sample constituents or properties.
Confocal Raman microscopy is widely used to measure
and image the chemical and physical properties of
samples in one, two and three dimensions.

Raman spectral image acquisition is essentially a two-

step process. The first step is mapping, which is the
process of measuring Raman spectra at predetermined
locations in a sample. The mapping is typically carried out
by means of a laser beam focused at a point sequentially
scanned across the selected area. The area and the
measurement locations are determined by the operator of
a Raman microscope and controlled by proper software.

One of the most attractive features of confocal Raman

microscopy is the ability to perform a nondestructive
analysis of a transparent or translucent sample at different
depths. Assuming the confocal Raman microscope is
properly aligned, the focused laser beam may be stepped
through the sample perpendicular to its surface. At the
same time, Raman spectra are recorded in the so-called
depth profiling mode. Collecting an area (XY) map in a
plane parallel to the sample surface at a distance Z below
the surface (probing distance), a spectrochemical image Figure 19. (a) The depth profiles of an individual polystyrene spherical particle from the cluster of the particles with diameters of about 10
deep within the sample can be obtained by analogy with μm embedded in a glass matrix measured at the point indicated by the red dot in (b). In the depth profiles, the integrated intensity I of the
polystyrene band at 1001 cm -1 (the left inset) and the intensities of the Raman bands of polystyrene in the spectral range of 100-3200 cm -1
the image produced by optical sectioning microscopy. (the right inset) are shown as a function of the sample stage displacement in Z-axis. Rendering the Raman spectral image obtained from a
cluster: (b) optical image of the sample surface superimposed on the 3D Raman image; (c) the 3D Raman image shown without the optical
image; (d) the XY section at a probing distance of 8 μm; (e) an XZ section of the 3D image; (f) a YZ section of the 3D image. The red color
shows the highest intensity of the band at 1001 cm -1 in the Raman spectrum of polystyrene and, consequently, highlights the polystyrene
particles, while the blue color corresponds to the glass matrix.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Theoretically, the maximal number of spectrochemical

images that may be generated from a single map is
equal to the number of resolved spectral elements.
Since humans can easily recognize visual targets based
on colors, the processed data outcomes are usually
visualized in pseudo-colored Raman images with different
color schemes or patterns. In the most commonly used
pattern designed around the three primary colors and
categorized as warm and cool, the red and blue colors
are designated, respectively, to the highest and lowest
values of the outcomes.

Thus, depending on the map’s geometry, Raman

spectrochemical images can be represented by color-
coded lines, areas or volume structures in which the
colors indicate the spatial distribution of chemical
components or sample properties. As an example, Figure
19(b–f) shows a 3D Raman spectral image acquired Figure 20. (a) When a metallurgical objective is used to focus the laser beam and make spectral measurements below the sample surface, the
point of focus is shifted into the sample by distance Δtrue , which is deeper than the sample displacement Δ nom designates due to the refraction
within a rectangular box of a sample composed of 10 at the interface between the sample (refractive index n > 1) and air (n = 1). Optical and Raman spectral images of a multi-layer polymer film:
(b) optical image of a mechanical cross-section cut, (c) Raman image of the cross-section cut, (d) Raman optical section obtained with
μm polystyrene beads embedded in a glass matrix. The 100×/1.3 oil immersion objective, (e) Raman optical section obtained with 100×/0.9 metallurgical objective,
3D Raman image is generated as the integrated intensity
of the spectral band of polystyrene at 1001 cm-1. The Immersion objectives with a fluid that matches the A common application of Raman confocal depth analysis
depth profile of an individual particle from the cluster is refractive index of the sample may be used to minimize is the investigation of multi-layer polymer composites.
represented in Figure 19(a). the problem of laser beam refraction at the air/sample 2D Raman images of a layered polymer composite
interface. Filling the gap between the objective’s lens approximately 124 μm thick and consisting of 5 distinct
What is the advantage of oil immersion over a “dry” and a sample with an index-matching fluid removes the layers obtained using both a metallurgical objective
metallurgical objective? bending of the light at the surface of the sample caused and an oil immersion objective are shown in Figure 20.
It should be emphasized that when performing spectral by refraction. An oil immersion objective can be used for The Raman images are generated using a multivariate
measurements using a metallurgical (“dry”) objective (the confocal Raman measurements within polymers since curve resolution (MCR) algorithm. MCR analysis is a
most common commercial choice) beneath the surface of the refractive index of oil (n = 1.52) is close to that of many powerful analytical tool because it allows spectrally
a sample with a refractive index n, the effect of refraction polymer materials. discriminating sample components without requiring any
at the air/sample interface shifts the point of focus much direct knowledge of the Raman spectral features of the
deeper into the sample than the sample displacement in components prior to analysis.
the vertical Z direction designates. As a result, the depth
scale appears to be falsely compressed, as depicted in
Figure 20(a).
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

As a reference, Figure 20(c) shows the results of

imaging of a physical cross-section of the same
polymer composite. When using the metallurgical
objective, there is clearly a significant compression of
the layers compared to what is observed in both the
Raman image obtained from the cross-section and
using the oil immersion objective. In contrast, the layer
thicknesses determined from the image collected with
the oil immersion objective and the image of the cross-
section are in close agreement. It is also evident that
the layers in the confocal image using the metallurgical
objective are less well defined due to a loss of spatial
resolution, leading to increased mixing of the spectral
features at the interfaces between the layers. This
issue also contributes to additional uncertainty in the
layer’s thickness determination. Using oil instead of air Figure 21. Schematic of EM-CCD detector with the electron multiplying register.
between the objective and the sample restores the spatial
discrimination, eliminates the depth scale compression In this way, an additional charge is created, effectively The advantage of the EM-CCD camera is lost as longer
and increases the amount of scattered light collected at amplifying the signal. The gain is controlled by adjusting and longer exposure times are used. Longer exposure
the probing depth. the applied voltage, which affects the probability that times might be required because the samples of interest
impact ionizations will occur. The multiplication register are weak Raman scatter. Longer exposure times result
What is the difference between CCD and EM-CCD amplifies any charge and thus amplifies charges in an increased signal because the camera is exposed
detectors? associated with the signal and charges associated with to impinging photons for longer. In those situations, the
CCD and electron multiplying (EM-CCD) are the most different types of noise. However, since the read noise is camera is read less often, and the relative effect on read
common choices for detectors in Raman microscopes. not multiplied, it becomes essentially insignificant when noise is diminished while the noise factor associated with
The detectors operate on the same basic principles using an EM-CCD and thus is not a limiting factor of the the amplification process remains. It should be noted that
considered above. However, EM-CCD detectors S/N of the camera. EM-CCD detectors excel in situations an EM-CCD can be operated with the amplification turned
are different from CCD ones in that they employ a that benefit from faster read-out speeds (short exposures off (gain =1) in which case it operates essentially like a
multiplication register to amplify charges, as shown in and multiple exposures). Faster read out also reduces standard CCD camera.
Figure 21. As the charge is being transferred for read out, dark current noise. Thus, EM-CCD detectors are capable
it passes through a multiplication register. The register of achieving relatively high S/N when operating at these
consists of a large number of cells where high voltage is faster rates.
applied to impart additional energy to the electrons. When
an electron has sufficient energy, impact ionization can
occur, and this creates an additional electron-hole pair.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

How to select the appropriate step upon mapping? The speed of the stage movement vstage is usually In Raman microscopy, the smallest object that can be
The design and characteristics of the sample scanning controlled by software based on the sampling step equal captured and measured is determined by the focused
mechanisms and techniques are important as they to the linear segment ls divided by the exposure time texp laser spot size and the distance between the points of the
determine the way and the rate at which the Raman map defined by the operator: mapping grid, usually referred to as the sampling step.
is collected. Conventional instruments are equipped with There are three principal mapping conditions that set
mechanical stages driven by stepper motors, while high- the relationship between the laser spot and the sampling
speed imaging systems use sampling mechanisms based step:
on the raster scanning principle. The main difference is
• Par-sampling: the laser spot is
that the former scans the samples in a discrete point- This method allows the map of a relatively large sample
equal to the step size;
by-point, also called stop-and-go mode, while the latter area to be acquired with high spatial precision and a fast
allows scanning continuously without stopping, which is a speed. The measurement of the linear segment results in • Under-sampling: the laser spot is
smaller than the step size;
critical factor for fast mapping. a better opportunity to detect Raman scattering from a
small object. • Over-sampling: the laser spot is
This mode offers the same size of the laser spot at each
larger than the step size.
point and adequate precision of their positioning on the
map. However, the motion of the mechanical stages in
the intermittent stop-and-go regime results in a longer
acquisition time than other mapping techniques.

DXR Raman Microscope is provided with an automated

XYZ mechanical stage, while DXRxi Raman Imaging
Microscope is equipped with a magnetic linear motor
stage and electron-multiplying EM-CCD detector. This
stage moves continuously in a zigzag manner while the
EM-CCD is read out at certain intervals determined by
the sampling step and exposure time. The spectrum
accumulates while the stage is moving so that the
resultant Raman spectrum is obtained from a linear
segment of the sample rather than a discrete point. The
area of the segment is determined by the laser spot size
multiplied by the sampling step: that is, the distance that
the stage passes for a predetermined exposure time.

Figure 22. The relationship between the sizes of the sample, laser spot and the sampling step. The green and red circles represent the laser
spot and a small sample, respectively, in the schematic of the conditions of under-sampling (a), par-sampling (b) and over-sampling (c). The
Raman images of individual polystyrene (PS) bead (d) and the adjacent PS and polymethyl methacrylate (PMMA) beads (e) are constructed from
the maps acquired using a 532 nm laser for excitation, 100×/0.9 objective and different sampling steps. The Raman images are generated as
spectral components extracted by the MCR method with the red and yellow colour highlighting PS and PMMA, respectively. From left to right:
optical images of the beads observed using 100×/0.9 objective, semitransparent Raman images (0.1 μm mapping step) superimposed on the
optical images of the beads and Raman images constructed from the map collected at 0.1, 0.3, 0.6 and 1.2 μm steps. (The figure is reproduced by
permission from Alexander Rzhevskii, Modern Raman Microscopy: Technique and Practice, Cambridge Scholars Publishing, 2021, 392.)
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

Under-sampling is generally employed when a sample although it still allows the PS and PMMS parts to be The imaging microscopes allow the composition and
needs to be evaluated, especially at a preliminary phase discriminated. Par-sampling with the step of 0.6 µm morphology of a sample to be measured and visualized
of the spectroscopic experiment. If there is enough a approximately matching the size of the laser spot in a detailed Raman spectral image in real time. They
priori knowledge about the sample composition, as well provides a satisfactory representation of the shape of principally bring new analytical capabilities to the
as the distribution of its components and their sizes and the individual PS bead, as well as the adjacent PS and spectrochemical characterization of large samples on a
shapes, then par-sampling or properly selected under- PMMA beads. Over-sampling at the 0.3 µm step results microscopic scale. These instruments offer extraordinary
sampling conditions can be used. Over-sampling can in the Raman image that more accurately depicts the productivity combined with information-reach spectral
improve the precision and contrast in Raman images, but shape of the polymer beads. Mapping the adjacent PS and spatial content. Due to the high scanning speed and,
it requires longer acquisition times and does not improve and PMMA beads at a 0.3 µm step means approximately as a consequence, the rapid dissipation of heat, imaging
spatial resolution. three sampling points per the diameter of the beads. Raman microscopes also ensure less laser-induced
As a rule of thumb, three sampling points per object’s damage to sensitive samples, especially biological ones.
Figure 22 schematically illustrates the relationship
aspects are recommended if the shape and dimensions
between the laser spot and step size in an area map
of closely situated objects need to be well discriminated What is terrain mapping?
when the laser spot size and dimensions of the object
and accurately represented in the Raman image. Over- It is important to position the sample surface in focus if
are comparable. Obviously, under-sampling may result in
sampling by further reducing the step size leads to the Raman spectra of the sample at the surface need
missing the microscopic object (red circle) present in the
a minor improvement of the contrast in the Raman to be obtained. When the sample tilt or its surface
sample, while par-sampling and over-sampling ensure the
images but a longer acquisition time. Obviously, this roughness exceeds the confocal depth required for the
object is captured.
recommendation applies to both the lateral and axial measurements, then at least part of the sample may go
The examples of mapping an individual polystyrene measurements. Note that the Raman images in Figure completely out of focus, and no meaningful spectral data
(PS) bead and a PS bead adjacent to a polymethyl 22(d) and (e) slightly overestimate the sizes of the beads will be collected from this part, as illustrated in Figure 23
methacrylate (PMMA) bead using different sampling steps because the images are formed as a convolution of the (a). It may lead to the fact that the obtained spectral data
are shown in Figures 22(d) and (e), respectively. Polymer original object with a laser spot of a similar size. The will be insufficient or not representative of the sample.
beads of 1 µm nominal diameters were deposited on a larger the object, the lesser the effect of the size of the
Typically, flat samples are the most ideal for microscopic
glass slide, and the maps were acquired using a 532 nm laser spot on the object size estimated from the Raman
Raman mapping because the sample is always in focus
excitation wavelength, a 100×/0.9 objective and a spectral image.
upon sample scanning. However, the physical nature of
continuously moving XY stage. According to equation (11),
some samples might prevent preparing a microscopically
the laser spot of about 0.7 µm was close to the diameter What is the difference between imaging and mapping
flat sample. For example, the samples, such as thin
of the polymer beads. The Raman images in Figure Raman microscopes?
films or coatings on substrates, could be damaged or
22(d) and (e) were generated using the MCR method The Raman microscopes equipped with fast scanning
completely removed when trying to produce a flat surface.
that extracts spectral components corresponding to mechanisms and usually EM-CCD detectors are often
In other samples, such as natural stones and minerals,
PS, PMMS and glass. These compounds are indicated, called imaging Raman microscopes. Manufacturers
medicinal tablets and capsules, semiconductor devices,
respectively, by red, yellow and black, visualizing the of Raman microscopes use this term for marketing
or those composed of powders or microparticles of
shape of the polymer beads in Raman images. Clearly, purposes to emphasize speed and other advantages
different sizes, the preparation can result in irrevocable
under-sampling at a 1.2 µm step causes image blurring over conventional stop-and-go mode systems, which
modifications and, consequently, the appearance of
and does not display the actual shape of the beads, are accordingly called mapping Raman microscopes.
spectral artifacts.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

It is also possible that irrevocable modifications of a

sample are not acceptable because of the need to keep
the sample intact and unaltered. Even if there are no
restrictions in terms of what can be done to a sample,
the methods required for producing microscopically flat
samples are too complicated and time-consuming to
be practical. It is, therefore, necessary to have a way
of imaging rough, sloped, or uneven samples. Terrain
mapping provides a way to manage these types of
samples.

Terrain mapping is a technique where the vertical position

of the microscope stage is adjusted to keep the sample
surface in focus while acquiring Raman maps across the
sample, as shown in Figure 23(b). A tightly focused laser
spot provides higher laser power density, giving more
Raman intensity from the analyzed volume. Maintaining a
tight laser focus while moving across a sample ensures
good spatial resolution and Raman intensity.

In DXR3xi Raman imaging microscope, the topography

of the surface is first recorded. It is performed using the
so-called optical autofocus by increasing contrast within
the field of view of an objective (frame). The position of
the sample in focus can be adjusted until the maximal
contrast in the optical image is achieved. Suppose the
area of the surface that needs to be measured exceeds
the size of the single frame. In that case, the sample is
scanned in the Z direction, and multiple frames at different
focal points are stitched together to montage a multi-
focus optical image of the sample, as outlined in Figure
Figure 23. Mapping a part of a medicinal tablet with an uneven surface: (a) scanning the tablet at a constant Z-position of the stage; (b)
23(c). scanning in Z positions adjusted to the surface profile of the tablet; (c) Mosaic tool automatically acquires the objective’s field-of-view in 3D
for retrieving topography of the surface; (d) the recorded topography of the tablet surface with the superimposed Raman image.
 Introduction Basic theory and terminology Raman Spectrometer Raman Microscopy

The individual frames with maximal contrast are then The maps were acquired using a 50×/0.50 metallurgical Raman mapping, performed by changing the Z position
extracted to synthesize the topographical image in objective and a 785 nm laser. When mapping the of the sample stage to adjust focus to the tablet
which the sample surface is maintained in focus, and tablet at a constant Z position, the API distribution is surface profile, provides confocal measurements on
the Z positions of the well-focused frames are recorded. adequately tracked in that part of the tablet that is well the uneven surface and, accordingly, adequate API
In the second step, the Raman map is acquired at the in focus. In the area of embossing, the depth of which tracking throughout the analyzed area, as illustrated in
predefined X and Y positions of the mapping grid, while significantly exceeds the depth of focus of the objective, (b) and (d). The Raman image generated from the map
the Z position at every point is retrieved from the recorded the surface goes out of focus. Accordingly, the quality of obtained when the stage Z position matched the surface
topography of the surface and dynamically adjusted the measured spectra deteriorates so that the API cannot topography demonstrates that the API highlighted in
during the sample scanning. be recognized in the out-of-focus area and distinguished red is distributed over the entire mapped area. The
from excipients (a). Raman images were generated as a correlation with
Figure 23 shows an example of the confocal analysis of
the API’s reference spectrum and superimposed on the
an active pharmaceutical ingredient’s (API) distribution in
corresponding optical images.
an uncoated medicinal tablet with an identification mark
embossed on its surface.

Learn more at [Link]/raman

For research use only. Not for use in diagnostic procedures. For current certifications, visit [Link]/certifications
© 2023 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of Thermo Fisher Scientific and its subsidiaries unless otherwise specified. BK56397_E 03/23M