OPTICAL INTERFEROMETERS: PRINCIPLES AND

APPLICATIONS IN TRANSPORT PHENOMENA

Sunil Verma [Corresponding author]
Scientist, Laser Materials Development and Devices Division, Raja Ramanna Centre for
Advanced Technology, Indore 452013, India
Email: sverma1118@[Link], Phone: +91-731-248 8670, Fax: +91-731-248 8650

Yogesh M. Joshi
Associate Professor, Department of Chemical Engineering, Indian Institute of
Technology Kanpur, Kanpur 208106, India (joshi@[Link])

K. Muralidhar
Professor, Department of Mechanical Engineering, Indian Institute of Technology
Kanpur, Kanpur 208016, India (kmurli@[Link])

ABSTRACT

Optical techniques are extensively used for high precision diagnostics and process monitoring in
physical, biological, and engineering sciences. Interferometry falls in one such class of
diagnostics. It relies on changes in the refractive index in the medium arising from variations in
the material density. The physical region in which imaging is being carried out is required to be
transparent. The light source best suited for an interferometer is a laser. Owing to its features
such as greater accuracy, resolution, instantaneous response and non-intrusive nature,
interferometry proves to be advantageous and extensively utilized in broad spectrum of
applications. The present chapter deals with the description of laser interferometers in
visualization and monitoring of processes involving fluid flow, heat transfer, and mass transfer.
The chapter is divided into two sections. In the first section, we discuss the basic
principles of interference and fringe formation. It includes the principles and operations of
various interferometer configurations such as Michelson, Mach-Zehnder, holography, phase-
shifting, speckle, schlieren and dual-wavelength interferometry. Interferometers can provide
vivid images of temperature and solutal concentration fields. Their real utility is in the
quantitative determination of transport properties in addition to heat and mass fluxes. The second
section describes the applications of interferometry in studying transient heat conduction,
buoyancy-driven convection in a rectangular cavity and superposed fluid layers, and crystal
growth from an aqueous solution. These illustrate the utility of interferometry in engineering and
research.

Page 1 of 77
 TABLE OF CONTENTS

1. FUNDAMENTALS
1.1 Optical interference
1.2 Why use optical techniques for imaging transport phenomena?
1.3 Optical imaging techniques
1.4 Imaging fluid flow
1.4.1 Shadowgraph
1.4.2 Schlieren
1.4.3 Interferometric methods
[Link] Schlieren-Interferometer
[Link] Mach-Zehnder interferometer
[Link] Refraction effects
[Link] Evaluation of interferograms
[Link] Infinite fringe interferograms
[Link] Wedge fringe interferograms
1.4.4 Holography and holographic interferometry
1.5 Imaging heat transfer and mass transfer
1.5.1 Mach-Zehnder interferometry
1.5.2 Phase shifting interferometry
1.5.3 Electronic speckle pattern interferometry (ESPI)
1.6 Simultaneous mapping of heat and mass transfer
1.6.1 Dual-wavelength interferometry
1.7 Computerized tomography
1.7.1 Shadowgraphic tomography
1.7.2 Interferometric tomography
1.8 Comparison of optical techniques

2. APPLICATIONS
2.1 Transport phenomena during crystal growth from solution
2.1.1 Importance of mapping solution growth

Page 2 of 77
 2.1.2 Phenomenology of the growth process
2.1.3 Supersaturation: Driving force for crystal growth from solution
2.1.4 Optical imaging of crystal growth from solution
2.2 Heat conduction in a horizontal fluid layer
2.3 Dissolution of sugar in water
2.4 Buoyancy-driven convection in a square cavity
2.5 Convection in a fluid layer with a free surface
2.6 Convection in a two-fluid layer system
2.7 Rayleigh-Benard convection in a rectangular cavity
2.8 Convection in an eccentric air-filled annulus
2.9 Convection in an eccentric annulus filled with silicone oil
2.10 Vortex shedding from a heated cylinder

3. CONCLUSION

ACKNOWLEDGEMENT
NOMENCLATURE
REFERENCES

Page 3 of 77
1. FUNDAMENTALS
The present section is concerned with image formation when a light beam traverses a transparent
medium in which density variation is present.

1.1 Optical interference

Optical methods of measurement are known to have specific advantages in terms of spanning a
field-of-view and being inertia-free. In use for over half a century, optical methods have seen
resurgence over the past decade. The main factors responsible are the twin developments in the
availability of cost-effective lasers along with high performance computers. Laser measurements
in fluid and thermal sciences have been facilitated additionally by the fact that fluid media are
transparent and heat transfer applications in fluids are abundant. Whole-field laser measurements
of flow and heat transfer in fluids can be carried out with a variety of configurations:
shadowgraph, schlieren, interferometry, speckle and PIV, to name a few. In the present chapter,
temperature field measurement in fluids by laser interferometry is addressed.
The ability to record interferograms on a PC using CCD cameras has greatly simplified
image analysis. It is possible to enhance image quality and perform operations such as edge
detection and fringe thinning by manipulating the numbers representing the image. Thus, image
sequences can be collected and spatio-temporal data of the appropriate scalar fields can be
determined.
Interferometric measurements are enabled by the following factors:
i. Light waves that are separated by a phase difference will interfere on superposition
and produce alternating bands of bright and dark fringes.
ii. Phase difference with respect to a reference is created in a variable refractive index
field.
iii. For transparent media, refractive index has a unique relationship with the material
density.
iv. Density scales with variables such as temperature and species concentration.
Hence, spatial as well as temporal changes in temperature and concentration will generate
corresponding changes in refractive index, phase and ultimately lead to fringe formation. It
should be emphasized at this point that interferometry is a differential measurement, namely
measurements of refractive index in a test section against a reference where refractive index is

Page 4 of 77
spatially and temporally a constant. The reference could be vacuum, the ambient, or a liquid bath
of uniform chemical composition and temperature.
Since interferometry relies on the phase of the light waves, it is evident that one should
employ a light source of unique wavelength (with which the phase is associated). Thus, the light
source should be monochromatic. In addition, the phase of the light waves should be stable in
time for meaningful fringe patterns to form. These factors require the light source to be coherent,
a property readily obtainable from lasers, for example, a helium-neon laser. In addition,
refractive index fields create an optical path difference with the reference wave. Accordingly, it
is required that light emerges from a point source and the origin for distance measurement is
unambiguous. These requirements can be comfortably fulfilled when a laser is used.
Light is electromagnetic radiation with wavelengths λ falling in the visible range (400-
700 nm). Light propagates in vacuum at a speed c that is independent of the wavelength.
Interference effects are associated with the modulation of the electric field, magnetic field
playing only a passive role. Electric field is a vector but for near parallel conditions, it is
sufficient to work with its scalar form.
The principle of interference can be illustrated through a simple example. Let a wavefront
move through a spatially homogeneous (reference) environment as
2π
=E1 A sin{ (ct − x)}
λ
The second wavefront moves through the test region where a physical process is in progress.
Through changes in refractive index, the wave undergoes a change of phase φ and the electric
field is obtained as
2π
=E2 A sin{ (ct − x) + φ}
λ
Note that the second wave has the same amplitude A and wavelength λ, requiring that a single
source be used for passage of light through the test and reference media.
When these two wavefronts are superimposed, the resultant electric field can be
calculated as
φ 2π φ
=
E1 + E2 2 A cos { (ct − x) − }
2 λ 2
The amplitude of the resultant wave is

Page 5 of 77
 φ
= 4 A2 cos 2
2
and is perceived as a variation in light intensity. This variation is between zero and 4A2 in a
sinusoidal form. When seen through a detector (or a camera), the image will look dark below a
certain threshold and otherwise, bright. This process is the basis of fringe formation in
interferometry. The superposition of two monochromatic coherence wavefronts results in
addition as well as cancellation of energies and is referred to as the interference phenomenon.
Since one wavelength λ corresponds to a phase difference of 2π, the phase difference φ
can be associated with a path difference δ as per
λ
δ= φ
2π
The relationship between a refractive index field and the phase difference can be established as
follows. Refractive index is defined as
c0
n=
c
where c is the speed of light in the physical region and c0 is that in vacuum. Since n > 1, it is
clear that the physical medium serves to slow down the electromagnetic waves. This effect is
equivalent to increasing the effective distance to be traversed by the wavefronts. In this context,
it is useful to define an optical path length
PL1 = ∫ ndz

Here, the integration is in the z-direction along the passage of the light beam. In general, this
integration is to be performed along the path of the light ray. Further, in vacuum, n=1 and the
path length is simply the geometric length of the apparatus in the viewing direction. In other
contexts, the effective distance to be covered by the light wave (referred to its speed in vacuum)
is greater by a factor that depends on the refractive index.
The reference wave passes through a region of constant refractive index leading to an
optical path length

PL 2 = ∫ n0 dz

On superposition, the path length difference is obtained as

∆PL=PL1 -PL 2 = ∫ (n − n0 ) dz

Page 6 of 77
The equivalent phase difference is calculated from the equation
λ
∆PL = φ
2π
In a refractive index field, n would vary from one point to another. Accordingly, the path length
difference and hence the phase field will be a spatial variable. Lines of constant phase
(specifically those that minimize and maximize intensity) will appear as fringes. These would
also be lines of constant refractive index of the material in the test section, material density, and
ultimately, lines of constant temperature and solute concentration.
The path length calculation is simplified if the refractive index field is two dimensional
(independent of the z-coordinate). For a domain of length L in the direction of propagation of
light

∆PL=PL1 -PL 2 = ∫ (n − n0 ) d =
z (n − n0 ) L

The corresponding phase difference is

2π
=φ (n − n0 ) L
λ
Interferograms contain information on the phase field. The above formula enables a conversion
of the phase information to refractive index. It is now important to comment on the relationship
between refractive index and material density of the region being investigated. The relationship
is unique and is given by the Lorenz-Lorentz formula [Kumar, 2008]
n2 − 1
= constant
ρ (n 2 + 2)
For gases, n~1 and the generalized relationship reduces to the Gladstone-Dale formula
n −1
= constant
ρ
The sensitivity of the measurement depends on the derivative dn/dρ, small values of the
derivative generating a smaller number of fringes in the image plane. The density–temperature/
concentration relationship is material-specific and can be obtained from an equation of state.
Typical values of dn/dT are
air, 0.927 ×10-6 K-1
water, 0.88 × 10-4 K-1

Page 7 of 77
It is clear that the number of fringes in water, as a rule, would be considerably larger when
compared to air.
The rest of the chapter is concerned with specific optical configurations for the
generation of fringe patterns and selected applications.

1.2 Why use optical techniques for imaging transport phenomena?

One could measure fluid properties with external transducers. However, in order to get sufficient
data one has to deploy a large number of such probes, which start interfering with the process
under study. On the other hand, only a few probes provide inadequate data. For example, a single
moving probe in a liquid medium yields temporally inaccurate data if moved slowly, but stirs the
system if moved rapidly. Therefore, external probes are never a satisfactory method for spatially
mapping a physical domain. In contrast, optical techniques are photon probes that do not affect
the process being studied. They map the properties of the process with a spatial resolution of
about a micrometer and a temporal resolution of about a millisecond. Their response is
practically inertia-free and provides a large volume of data of the process under study. Further,
the availability of high speed cameras for image acquisition and powerful computers for data
reduction have helped in revival of optical diagnostics in engineering research in the last decade.

1.3 Optical imaging techniques

Various types of optical imaging techniques have been used for studying transport behavior of
momentum (fluid mechanics), thermal energy (heat transfer) and transport of molecular species
(mass transfer). Overall, these optical techniques can be subdivided into three categories. In the
first category flow marking is carried out by using dyes, bubbles, buoyant seeds, etc. in order to
enable direct visualization of the fluid motion. In the second class of techniques fluid motion is
studied by analyzing the frequency of an incident radiation, which is known to undergo Doppler-
shift after getting scattered from the particles moving with the flow (e.g. laser Doppler
velocimetry). The third category considers all those techniques which analyze behavior of an
index of refraction of the medium through which the light beam propagates due to various
impetuses such as fluid motion, temperature gradient and/or concentration changes (e.g.
shadowgraphy, schlieren, interferometry, etc.). In this chapter we will focus only on the third of
techniques.

Page 8 of 77
 The techniques based on alteration of refractive index n take advantage of unique relation
between n and the density of the medium ρ for transparent mediums. Since transported variables
such as heat (temperature) or mass of chemical component (concentration) directly affects the
density, monitoring of refractive index n directly gives an information about temperature or
concentration gradients in the medium. Although various techniques mentioned in the third
category require variation in refractive index in order to analyze transport behavior, they have
complimentary characteristics. Usually, interferometry is used when the expected changes in the
refractive index are small, whereas shadowgraphy and schlieren are used when the changes in
the refractive index are large. Non-interferometric techniques such as shadowgraph or schlieren
rely on detection of path of light rays as they traverse through the region of interest. On the other
hand interferometric methods detect optical path-length differences. Principally, the three
techniques measure the physical property in an integrated manner, such that the image output is
an integration of the physical quantity of interest along the path of the beam through the process
chamber. Therefore, these methods show the best results when the transport is two-dimensional
and light passes perpendicular to this two dimensional plane. Usually these techniques are
employed when the field is nearly two dimensional, i.e. when variation of a physical quantity
under investigation is negligible in the third dimension. However, if this assumption is not valid,
then variation in the third dimension needs to be obtained by applying analytical techniques such
as computerized tomography. The non-interferometric techniques are simpler to implement and
difficult to interpret quantitatively, while the interferometric techniques are amenable to
quantitative analysis though difficult to set up and maintain.
The application of interferometry are discussed in the context of monitoring the following
transport processes: transient heat conduction, solutal transport, buoyancy-driven convection in a
rectangular cavity, buoyancy-driven convection in superposed fluid layers, and growth of a
crystal from its aqueous solution.

1.4 Imaging fluid flow

Fluid motion and convection phenomena can be visualized by optical techniques through either
(a) refractive index in the form of optical path difference (or its gradient), (b) displacement of
fluid markers as a function of time, or (c) velocities of fluid markers. Flow Visualization by
Merzkirch [Merzkirch, 1987] is a classical text for researchers planning to conduct fluid flow

Page 9 of 77
and convection imaging. Similarly, text books on Optical Interferometry by Steel [Steel, 1983]
and Hariharan [Hariharan, 1985] are a good starting point for familiarizing oneself with the
subject. A survey of holographic interferometry is given by Vest in his book [Vest, 1979].
Optical Measurements: Techniques and Applications, edited by Mayinger [Mayinger, 1994] and
Schlieren and Shadowgraph Techniques, by Settles [Settles, 2001] present excellent compendia
of a variety of optical techniques for fluid flow and convection mapping.

1.4.1 Shadowgraph
Shadowgraphy employs a collimated beam of light from either a white light source or a laser.
The image has uniform contrast if there are no refractive index gradients in the region of interest.
On the other hand, due to the curvature of light rays by the refractive index gradients, the
portions of the image corresponding to the displaced rays appear bright or dark. Shadowgraph
systems are simple and have been extensively used in experimental fluid mechanics research for
flow visualization [Goldstein, 1996; Settles, 2001]. Figure 1 shows the schematic drawing of a
shadowgraph optical set up and the image acquisition system.
Under the assumption of infinitesimal deviation of light rays inside the inhomogeneous
field and also between the exit face of the process chamber and the screen, the linear governing
equation for the shadowgraph process is [Schopf, 1996; Verma, 2005]:

I o ( xi , yi ) − I s ( xs , ys )  ∂2 ∂2 
(
= L × D ) 2 + 2 {
ln n( x, y )}
I s ( xs , ys )  ∂x ∂y 
Here I 0 is the intensity at a particular point ( xi , yi ) on the screen in the absence of the field of

disturbance; I s is the intensity at point ( xs , ys ) due to intensities I 0 ( xi , yi ) of all the points

( xi , yi ) which are mapped onto ( xs , ys ) in the presence of the field of disturbance; D is the

spread of the field of disturbance; L is the distance between the exit surface and the screen; and
n( x, y ) is refractive index at the location ( x, y ) .

1.4.2 Schlieren
Schlieren technique requires a collimated light system and converging lens with a knife-edge that
is introduced halfway into the focal point [Goldstein, 1996; Settles, 2001]. Light rays deflected
(by refractive index gradients) toward the knife edge are blocked resulting in dark spots on the

Page 10 of 77
image; conversely, rays deflected away from the knife edge show as bright spots on the screen.
The knife-edge can be oriented either vertically or horizontally, thus enabling measurement of
refractive index gradients in the two directions. The governing equation for the schlieren process
under the linear approximation is
Io − I s 2 f  ∂ 
=  {ln n( x, y )} 
Is a  ∂y 
where a is the source size and f is the focal length of the lens used for focusing the beam over
the knife-edge. As is evident from the above equation, a single integration is needed to extract
the refractive index field from the schlieren image as compared to double integration required in
case of solving a Poisson equation for shadowgraph image.

Growing
Spatial crystal
filter

He-Ne Laser CCD / Video Video

Collimator Crystallizer Frame grabber
camera monitor + Image proc PC
Screen software

OPTICAL SET UP
SIGNAL ACQUISITION
& PROCESSING INSTRUMENTATION

Figure 1: Schematic drawing of a shadowgraph optical set up and the image acquisition
instrumentation

Growing Frame grabber

card+ image
Crystal
Spatial processing
filter software

FG card +
IP software
He-Ne Laser Collimator Crystal Knife CCD / Video Video
growth Lens edge Camera Monitor
cell Screen

P.C

OPTICAL SET UP SIGNAL ACQUISITION

& ANALYSIS INSTRUMENTATION

Figure 2: Schematic drawing of a schlieren optical technique

Page 11 of 77
1.4.3 Interferometric methods
The physics behind the generation of interferograms is different from shadowgraph and schlieren
techniques. In interferometry, the phase change of the test beam with respect to the reference
beam is the origin of the formation of interferograms. As a rule, interferometry is used when the
expected changes in temperature or concentration are small. If the thermal or solutal changes in
the region of interest are large, they limit the usefulness of interferometry in process monitoring
for the following reasons: (i) the linearity between density (through temperature or
concentration) and refractive index breaks down; and (ii) large gradients in the temperature and
concentration fields result in sharp bending of rays. Both these effects complicate interferometric
data analysis.

[Link] Schlieren-Interferometer
This technique has the characteristics of both non-interferometric and interferometric methods in
that it detects the curvature as well as the path length of the light rays. Since the final
experimental output is an interference pattern, we have classified it as an interferometric
technique. It is also referred to either as a Wollaston-prism shearing interferometer or a
differential interferometer [Mayinger, 1994, pp. 78-83; Settles, 2001, pp. 132-134]. The
Wollaston prism uses double refraction to separate an incident light ray into two rays which are
polarized orthogonally to each other and diverge from each other at an angle of ±θ/2 with respect
to the original direction of the beam (Figure 3). Collimated monochromatic light is passed
through a polarizer, set at an angle of 45° to the direction of separation of the Wollaston prism,
so as to produce e- and o- rays of equal intensity on passing through the prism. The parallel-ray
polarized beam then passes through the solution in the test cell and is then focused by a lens into
the Wollaston prism. As the polarized rays traverse through this prism, they split into e- and o-
rays. Incident ray 1 is separated into two rays 1e and 1o, representing polarization parallel and
perpendicular to the plane of the diagram. On leaving the prism, these two rays have an angle of
θ between them. Adjacent incident rays 2, 3… are similarly split, as shown. Since the rays 1 and
2 have a separation of θ between them, ray 1e coincides with 2o, and similar pairings occur for all
adjacent rays having separation of θ, i.e. the ray 2e coincides with 3o, and so on. Since the
coinciding rays have orthogonal planes of polarization, there is no interference between them.
But when passed through a polarizer oriented at 45° to the separation direction of the Wollaston

Page 12 of 77
prism, the emerging paired wavefronts interfere and an interference pattern is obtained on the
screen. In short, the collimated image from the experimental test cell is split by the Wollaston
prism into two mutually displaced images which then interfere with each other. Therefore each
element of the final image represents the interference between Pxy and Px ( y + k ) , where P is the

optical path length at image coordinates x and y . If the prism is centered upon the focus of the
converging lens, infinite-fringe mode interference results, which resembles a schlieren image.
Displacement of the centre of the Wollaston prism from the focal point results in wedge fringes
to appear in the image. The bends in the wedge fringes are proportional to the refractive index
gradients in the test region and can be used for quantitative evaluation. The principal virtue of
the schlieren interferometer is the flexibility of being able record quasi-schlieren images, which
facilitate qualitative interpretation, to multiple-fringe images for quantitative measurements.

Test region
Spatial
1
filter 2
3

2 o3 o
Laser 1o 3 e Fringe
Polariser 2e
1e pattern
(45° w.r.t.
Collimating Focussing Wollaston
separating
lens lens prism Screen
direction Polariser
of WP) (45° w.r.t.
separating
direction
of WP)

Figure 3: Optical layout of a schlieren (differential) interferometer [Verma, 2008].

[Link] Mach-Zehnder interferometer

The Mach-Zehnder interferometer is a popular configuration for studies in heat and mass transfer
in fluids. The optical components namely beam splitters and mirrors are inclined exactly at an
angle of 45° with respect to the beam direction (Figure 4). The first beam splitter splits the
incoming collimated beam into two equal parts, namely the transmitted and reflected beams. The
transmitted beam is the test beam and the reflected beam is the reference beam. The test beam
passes through the test region, is reflected by the mirror, and recombines with the reference beam
on the plane of the second beam splitter. The reference beam undergoes a reflection at the second
Page 13 of 77
mirror and passes through the reference medium and superposes with the test beam at the second
beam splitter. On superposition, the two beams produce an interference pattern. This pattern
contains information about the variation of refractive index in the test region.
The Mach-Zehnder interferometer can be operated in two modes, namely infinite fringe
setting and wedge fringe setting (Figures 5(a-d)). In the former the test and reference beams are
set to have identical geometrical path lengths, and the fringes form due to refractive index
changes alone. Since each line is a line of constant phase, it is also a line of constant refractive
index. The fringe thickness is an inverse measure of the local refractive index gradient, being
smaller when the gradients are high. This setting is used for high-accuracy refractive index
measurements. In the wedge fringe setting, the mirrors and beam splitters are deliberately
misaligned to produce an initial fringe pattern of straight lines. When a refractive-index
disturbance is introduced in the path of the test beam, these lines deform and represent refractive
index profiles in the fluid. The quantitative methodology of extracting temperature change per
fringe shift from the wedge and infinite settings is described below.

Figure 4: Optical configuration of a Mach-Zehnder interferometer. A compensation chamber

may be required in the path of the reference beam to cancel effects arising from an undisturbed
medium in the test cell.

Page 14 of 77
 (a) (b)

(c) (d)

Figure 5: Mach-Zehnder interferograms. (a) Wedge fringes, (b) Candle flame in the wedge
fringes, (c) Infinite fringe, and (d) Candle flame in the infinite fringe

Let n(x,y,z) and C(x,y,z) be the refractive index and concentration fields respectively in
the physical domain of interest. With reference to the test chamber, coordinates x and y form the
vertical plane, which is perpendicular to the beam propagation direction, and the z-axis is the
horizontal line along the length of the test section. The light beam propagates parallel to the z-
axis. Let no and Co be the reference values, as encountered by the reference beam passing
through a compensating chamber. The interferogram is a fringe pattern arising from the optical
path difference
S
∆PL( x, y, z )= ∫ [ n ( x, y , z ) − n ] d s
0
o

which in terms of concentration is

Page 15 of 77
 S
dn
∆PL( x,=
y, z )
dC ∫ [C ( x, y, z ) − C ] d s.
0
o

The above integral is evaluated along the path of the light ray given by the coordinate s. If we
neglect higher order optical effects such as refraction and scattering, the path of the light ray will
be straight along the length of the test chamber. Since the interference fringes are loci of constant
phase or constant phase difference, the optical path difference ∆PL is constant on a given fringe.
Therefore

∆PL( x, y, z )
L

∫ [ C ( x, y , z ) − C =
0
] dz
o
dn dC
= constant

and it follows that

L

∫ C ( x, y, z )dz
0
− Co L =
constant

for a given fringe.

L
The integral
∫ C ( x, y, z )dz
0
is equal to C L , where C is the average value of the variable

C(x,y,z) along the length L traversed by the laser beam inside the test chamber. Under the
approximation of negligible refraction, the above integral reduces to a line integral of the
function C(x,y,z) along the z-axis. Hence, we get
∆PL
( C − Co ) L= constant= (1)
dn dC
This equation holds for all the fringes in the case of the infinite-fringe setting. Each fringe of the
interferogram represents a locus of points over which the average of the concentration field
evaluated along the length of the test chamber is constant; i.e. the fringes are isoconcentration
contours. In processes governed predominantly by thermal changes, fringes correspond to
isotherms. For a change in path length by one wavelength, one fringe shift occurs. Since this
change in path length per fringe shift is constant, the change in concentration per fringe shift is
also a constant. Defining the function ( C − Co ) L in Equation (1) as f ( C , L) , the concentration
at two successive fringes for a given value of L can be obtained as
∆PL
fringe 1: f1 ( C1 , L) =
dn dC

Page 16 of 77
 (λ + ∆PL)
fringe 2: f 2 ( C2 , L ) =
dn dC
where λ is the wavelength of the laser. From these two equations the concentration change per
fringe shift can be calculated as
λ/L
=
∆CE
1
L
{ f ( C , L) − f=
2 2 ( C , L)}
1 1
dn dC
(2)

The number of fringes expected in a projection can be estimated from the relation
C 2  C1
Number of fringes =
C E
This relation is valid when (a) the refraction effects are negligible and (b) the test beam is
traversing through a single fluid.
Now consider the situation wherein concentration gradients are large and the ray
traverses a different distance through the cell. Let the two light rays traverse paths L1 and L2
respectively inside the test cell, and the corresponding line integrals of the concentration field

C(x,y,z) resulting in the average concentrations be C 1 and C 2 respectively. Using Equation (1)
we can write
(C2 − Co ) L2 = ( C1 − Co ) L1 .

Thus, the line integral value C 2 at a location on the fringe which corresponds to the length L2

can be expressed in terms of the line integral value C 1 at some other location corresponding to
the ray length L1 as
L1
C 2  Co  C  Co  .
L2 1
[Link] Refraction effects
The presence of strongly refracting fields can modify Equation (2) for the concentration
difference per fringe shift. In the present context, a strongly refracting field will arise when there
is large concentration or thermal difference in the vertical direction inside the test cell. Hence,
the light ray will not travel in a horizontal plane; instead it will bend in the vertical direction, the
extent of which depends on the magnitude of the concentration or thermal gradient. Therefore
refraction will introduce an additional path length for the test beam. For the sake of completeness
of the discussion on interferometric imaging, we present below the derivation of the formula for

Page 17 of 77
calculating the change in concentration per fringe shift under the presence of strongly refracting
fields.
Consider the path of the light ray AB through a test cell (Figure 6) when it is affected by
refraction effects. Let α be the bending angle at a location P of the test cell. The optical path
length from A to B is given by
B
AB = ∫ n( x, y, z ) ds
A

L
dz
= ∫ n ( x, y , z ) .
0
cos α

Here y is a coordinate parallel to the gravity vector and z is parallel to the direction of
propagation of light. The length of the test section in the z-direction is indicated as L.

B
ds
α (z)
Light ray
P
A

Figure 6: Bending of the light ray due to refraction effects.

Assuming α to be small, cos(α) can be expressed as

cos   (1  2 )1/ 2 .

Using the first two terms of the binomial expansion, one can approximate
2
cos   1 
2
Hence the optical path length is given by
−1
L
 α2 
=AB ∫
0
n( x, y, z ) 1 −  dz
 2 

Page 18 of 77
 L
 α2 
= ∫
0
n ( x, y , z )  1 +

 dz
2 
(3)

The cumulative bending angle α(z) at any location along the z axis can be easily calculated. It
can be shown that the expression for the bending angle α(z) is (Goldstein, 1996)
1 n(x , y )
(z )  z (4)
n(x , y ) y

where n(x , y ) is the average line integral of n(x , y ) over the test cell dimension L. Similarly the
expression n(x , y ) y represents the average line integral of the transverse derivative of

n(x , y ) over the length L.

Substituting the expression for α(z) from Equation (4) in Equation (3) we get
L  1 1  ∂n  2 
2

=AB ∫ n ( x, y , z )  1 +
 2 2 
n   ∂y
 z  dz

0  
2
1  ∂n  3
= n ( x, y ) L +   L
6n  ∂y 
The optical path of the reference beam is
L
AB = ∫ no dz
0

= no L
Hence the difference in the optical path lengths of the test and reference beams in the presence of
refraction effects is
2
1  ∂n  3
=
∆PL n ( x, y ) L +   L − no L
6n  ∂y 
2
 ∂n  3
= ( )
n ( x, y ) − no L +   L
6 n  ∂y 
1

2
1  ∂n  3
(
=C1 ( x, y ) − Co L
dn
dC
+ )   L
6 n  ∂y 

Page 19 of 77
where C1 ( x, y ) represents the average line integral of the temperature field along the direction of
the ray at a given point on the fringe. The corresponding ray over the next fringe corresponds to
an additional path of λ and can be written as
2
1  ∂n  3
+λ
∆PL= ( )
C 2 ( x, y ) − Co L
dn
dC
+   L
6 n  ∂y 

where C 2 (x , y ) represents the average line integral of the temperature field along the direction
of the ray at a point on the next fringe. The successive temperature difference between the two
fringes is

 dn 
2  2
  ∂C 
2

λ= ( )
C 2 ( x, y ) − C 1 ( x, y ) L
dn
+
1
    ∂C  −   L3
dC  dC    ∂y   ∂y  
 2  1
6n

and the concentration change per fringe shift is

  d n    ∂C 
2 2
  ∂C 
2
 1  L3
λ −     − 
 6 n( x, y )  dC    ∂y   ∂y
2 

1

 .
∆CE =
 dn 
L 
 dC 
Since the gradient in the concentration field is not known before the calculation of the fringe
2 2
 ∂C   ∂C 
concentration, the factor   −  must be calculated from a guessed concentration
 ∂y   ∂y 
 2  1

field. Thus the final calculation of ∆CE relies on a series of iterative steps with improved
estimates of the concentration gradients.

[Link] Evaluation of interferograms

The thinned fringes carry the essential information of the process parameter rather than the thick
fringe bands. Hence, in order to extract quantitative data from the Mach-Zehnder interferograms,
we have to obtain the fringe skeleton. For our process, the thinned fringes contain the
information about the path-integrated concentration field. In the sub-sections below we discuss
the methodology for calculating the concentration field from interferograms recorded in the
infinite- and wedge-fringe settings.

Page 20 of 77
[Link] Infinite fringe interferograms
Let us consider an infinite fringe interferogram as shown in Figure 7. The presentation is in the
context of a concentration field (of salt in water, for example) that varies from Csat to Csupersat.
The reference is to a crystal growth process discussed latter in this chapter. The aim of the
following discussion is to find the absolute concentration corresponding to each fringe curve. We
know that the change in concentration per fringe shift is given by
λ/L
∆CE =
dn dC
and
λ/L
C2 − C1 =
dn dC
At the top boundary the concentration is known to be the saturation concentration Csat, whereas
the bottom is at a concentration equal to the supersaturation value Csupersat. Our aim is to find the
concentrations C1, C2 and C3 at the fringes 1, 2 and 3 respectively in the interferogram (Figure 7)
with the given boundary conditions. Although the concentration values at the lower and the
upper boundaries are known, assigning concentration to the first fringe appearing near the
boundary is not straightforward. This is because the concentration gradients near the boundary
result in several fine fringes that are lost because of the finite resolution of the CCD/video
camera. Also, the first fringe may be distorted during the image processing operations. Hence it
becomes necessary to assign concentration to the first visible fringe by an appropriate analytical
procedure. The methodology of finding the concentration values at the first and the subsequent
fringes is presented below.
Step-1: First identify that region of the interferogram where the fringes are closely packed. The
fringes 1, 2, and 3 are three such fringes in the above interferogram (Figure 7).
2
Step-2: Fitting a second-order polynomial of the type C (y )     y   y to the three

fringes, we get
C 1     y1   y12

C 2     y 2   y 22

C 3     y 3   y 32

Page 21 of 77
where y is the vertical coordinate measured from the lower boundary. An exploded view of the
three fringes and the coordinate axis is shown in Figure 8. Using above equations we obtain two
simultaneous equations for the two unknowns  and  :

C E  C 1  C 2   (y1  y2 )   (y12  y22 )

C E  C 2  C 3   (y2  y 3 )   (y22  y 32 )

These are solved to obtain the constants  and  . Here, C E is the concentration change per
fringe shift in the fluid.
Step-3: The next step in the analysis procedure is to compute the concentration gradient near the
boundary close to the fringes. In the interferogram shown in Figure 9, we are interested in
finding the concentration near to the lower boundary.
The concentration gradient at the boundary can be given as
C
   2  yo .
y y yo

Step-4: Once the concentration and the concentration gradient at the boundary are known, the
concentration at fringe 1 near the lower boundary can be obtained as
C
C 1  Cboundary  ( y1  yo )
y boundary

and so
C 1  C supersat  (   2  y1 )( y1  yo )

Step-5: C E is the quantum of concentration change per fringe shift; thus concentration at
subsequent fringes 2, 3 and others can be computed by adding it to the concentration of the
previous fringe. However, if the concentration gradient is negative in the counting direction, an
equivalent amount has to be subtracted.
Step-6: Since the interferogram is digitized and stored in a computer in the form of a matrix of
integers, the above procedure of computing concentration can be implemented for any element of
the matrix. In our experiment the concentration is a smoothly varying function, and it is expected
that the concentration difference between adjacent rows of the matrix will be small. Hence the
evaluation procedure is applied to the rows coinciding with the fringe locations but not in the gap
between them.

Page 22 of 77
 Csat

Csupersat
Process boundary

1 2 3
Csupersat
Fringes
Figure 7: Infinite fringe interferogram of the concentration field in an experiment that develops a
concentration distribution.

2
y3
y2
y y1 1

Figure 8: An exploded view of the fringes 1, 2 and 3 along with the coordinate system used.

[Link] Wedge fringe interferograms

The Mach-Zehnder interferometer is not always used with the two interfering wavefronts parallel
to each other, as in the infinite-fringe setting discussed above. There is a second mode in which
the two interfering wavefronts have a small angle θ between them, introduced deliberately during
alignment. Upon interference they produce an image consisting of bright and dark fringes,
representing the loci of constructive and destructive interference respectively. These parallel and
equally spaced fringes are referred to as wedge fringes. The spacing between the wedge fringes
is a function of the tilt angle and the wavelength of the laser light used, and is given by
2
d 
sin ( 2) .

For small tilt angles, the above expression becomes

Page 23 of 77
 
d 
.
As θ is decreased to zero, the wedge fringes get farther apart, approaching the infinite-fringe
pattern.
When a thermal or concentration field is introduced in the path of the test beam, the
phase of the test wavefront gets distorted. Upon interference with the reference wavefront, it
manifests itself as a change from straight and parallel fringes to curved fringes. The two
interference patterns are shown schematically in Figure 9 and as an exploded view in Figure 10.

Original undeformed
wedge fringes
Deformed wedge fringes

3
W Growth cell boundary
4 at a concentration of
Csupersat

Crystal-solution interface

Figure 9: Wedge fringe pattern with and without the field of disturbance.

Csat Csupersat

D
1

d 2
x

Figure 10: Computing field concentration in the presence of the wedge fringes.

Page 24 of 77
Let us develop a procedure to the solutal concentration at a point x in the interferogram. The
deviation of fringe 2 near the crystal-solution interface is D, while at point x is d. Thus it can be
written as [Goldstein, 1996]
C (x )  C sat d

C supersat  C sat D

Since all other parameters except C (x ) are known, the local concentration can be easily
calculated. The concentration gradient dC dx is obtained from the slope of the fringes at a point

x. The above methodology is used for calculating the concentration at points which lie on the
wedge fringes. However, if one were to find concentration at a point lying between two wedge
fringes, then one has to locate four nearby points on the wedge fringes and compute their
respective concentrations. It is then followed by using a suitable interpolation technique to find
the concentration at the desired point.

1.4.4 Holography and holographic interferometry

Holography is a form of interferometry and can be used to achieve new modes of measurement,
for example, interference of images of the same object at two different points in time. We
explain the conventional holography and how it differs from the holographic interferometry.
Also various types of holographic interferometry are discussed.
Holography was discovered by Dennis Gabor in 1948 [Gabor, 1948 & 1949] and is based
on the principles of interference and diffraction. Ordinary photography records only the wave
intensity while the phase information is lost. In contrast, both the amplitude of the object wave
and the phase are recorded in holography [Collier, 1971; Hariharan, 1986 & 2002]. Briefly, the
procedure for hologram recording and subsequent reconstruction of the object image is as
follows: A coherent beam of light from a laser is split into an illumination beam and a reference
beam. The former illuminates the object so that the scattered light from the object strikes a high-
resolution photographic plate. The reference beam is also reflected onto the photographic plate,
where it interferes with the light from the object. The photographic plate, or hologram, is then
developed to reveal the interference pattern registered on it.
The object wave can be reconstructed in two different ways, one resulting in a virtual
image and the other in a real image. When the hologram is illuminated by the original reference

Page 25 of 77
beam, the interference pattern on it behaves as a diffraction grating, and diffracts the incident
light. A part of the diffracted beam is a replica of the object-image wavefront. By looking
through the hologram into the direction of the original object, the viewer perceives the waves to
be originating from a virtual image of the object located precisely at the original location (Figure
11a). On the other hand, if the back of the hologram is illuminated by a conjugate of the
reference beam (a beam identical in phase and intensity distribution to the original reference
beam but propagating in the opposite direction), the diffracted beam is directed and shaped by
the hologram into a real image, corresponding spatially to the original object. Viewing in the
direction of the diffracted beam recreates the object through the hologram (Figure 11b).
Holographic interferometry is a refinement of holography [Vest, 1979]. The main difference
between classical interferometry and holographic interferometry is that for ordinary
interferometry, an object wavefront is continuously compared against a reference wavefront, and
the optical path length difference between the two wavefronts is measured to determine changes
in the physical parameter of interest. In holographic interferometry, two types of interference
phenomena occur - the interference between the reference and the object waves to form a
hologram and the interference between two such holograms. This is equivalent to the
interference of two coincident images of the same object at two different times. Thus, in classical
interferometry, the two waves follow different paths at the same time while in holographic
interferometry, they pass through the same path at two different times.
The reader is referred to the edited volume by Mayinger [Mayinger, 1994] and the review by
Verma & Shlichta [Verma, 2008A] for details of different types of holographic interferometry.
The three forms of holographic interferometry are explained schematically in the Figures 12(a-c).
The three holographic interferometric techniques described above have several trade-offs. In the
first, one follows the process in real-time, but provides only averaged parameter values along a
single direction. The second technique is simple in implementation but provides the parameter
values at a particular instant and also provides only averaged values along a single direction. The
third can provide a three-dimensional array of parameter values, but again only at a particular
instant of time.

Page 26 of 77
 ve nce
nt
wa fere
fro
Re
Virtual
Diffracted
image of
wavefront
process Observer

Hologram

(a)

Real Diffracted

e
image of

ref jugat
wavefront

vef nce
t
Observer process

ron
ere
n
Co

wa
Hologram

(b)

Figure 11: (a) Formation of virtual holographic image with original reference beam used for
reconstruction, and formation of a real holographic image with a conjugate reference beam used
for reconstruction [Verma, 2008].

Page 27 of 77
 ro e
ef nc
nt
ro ce

a v re
nt

w ef e
ef n
av re
Develop and fix the

R
w ef e
holographic plate to

R
get the hologram
Dynamic
Step 2. Get hologram New interference
object pattern
Object wavefront
wavefront
Replace the hologram Process occuring Hologram
Process Chamber in the test chamber
(no process taking place) Holographic in its original
plate position

Step 1. First exposure Step 3. Replace hologram Step 4. Holographic interferogram

(a)

ro ce

ro ce
ro ce

nt

nt
ef n

ef n
nt
ef n

av re

av re
av re

w efe

w efe
w efe

R
R

Develop and fix the Single instant

inteference pattern
New holographic plate to between reconstructed
Object get the hologram original & new object
wavefront wavefronts

Holographic Holographic
No process occuring plate plate Hologram
inside the test region Object Process occuring
wavefront inside the test region

Step 1. First exposure Step 2. Second exposure Step 3. Get hologram Step 4. Holographic interferogram
(b)

e
ce t nc t
en n e re ron
er fro f e f
f
Re av
e
N object Re av N new object
w w
Diffraction wavefronts wavefronts
grating

Develop and fix the

holographic plate to
get the hologram

No process occuring Process initiated

inside the test region inside the test region
Holographic Holographic
plate plate

Step 1. First exposure Step 2. Second exposure Step 3. Get hologram

ro ce
nt
ef n
av re
w efe
R

3D Reconstruction
using tomographic
algorithms

Hologram

N interferograms

Step 4. Holographic interferograms Step 5. Tomographic reconstruction

(c)

Figure 12:(a) Single-direction, single-exposure, real-time holographic interferometry, (b) single-

direction, double-exposure, single-instant holographic interferometry, and (c) multi-directional,
double-exposure, single-instant holographic interferometry and tomography [Verma, 2008].
Page 28 of 77
1.5 Imaging heat transfer and mass transfer
Virtually all of the quantitative refractive index mapping techniques discussed thus far apply to
the mapping of both, the heat transfer as well as the mass transfer fields. It is assumed that the
experiment is essentially two-dimensional along the optical axis, so that the refractive index is
constant along each ray of light. Further, an equation that connects refractive index with
temperature (via density) is required to convert refractive index into absolute temperature.
Interferometry can measure temperature in the following cases: (a) pure liquids having an
appreciable dn dT , (b) systems where changes in concentration are zero or negligible, and (c)
applications where a second technique maps the solute concentration.

1.5.1 Mach-Zehnder interferometry

Mach-Zehnder interferometry can be employed for mapping heat and mass transfer. The
quantitative methodology for extracting the field parameter (either temperature or concentration)
from wedge- or infinite-fringe mode Mach-Zehnder images has already been discussed in the
previous section. Mathematically, let n(r) and T(r) be the refractive index and temperature fields
respectively, in the physical domain being studied. Let no and To be their reference values, as
encountered by the reference beam. The interferogram is a fringe pattern arising from the optical
path difference:
δ
= ∫ (n − n )dx
o

which, in terms of temperature is

dn
dT ∫
=δ (T − To )dx .

Here the integration is over the path along the ray direction. Other forms of single-pass
interferometry and holography can also be used and interpreted in an identical manner. The
principal problem with interferometry, for both heat and mass transfer mapping, is calibration.
This will not matter for fluid flow (convection) mapping because one is interested in having an
observable difference between the moving and static parts of the fluid. Even when experiments
require relative temperature or its gradients, the equation relating refractive index to temperature
(or density) is required. Such a relationship is unique and obtainable for transparent media via
the simplified Lorenz-Lorentz formula

Page 29 of 77
 n2 − 1
= constant (k )
ρ (n 2 + 2)
The material density can then be related to solute concentration and temperature. For a
discussion on the generalized relationship, please refer [Kumar, 2008].

1.5.2 Phase shifting interferometry

Conventional interferometric mapping is based on obtaining an interferogram, locating the
centers of the fringes, measuring the difference between the fringe-center position with that of
the reference pattern, and finally interpolating between the fringes to obtain the variable of
interest anywhere in the field. This procedure has the shortcomings of being tedious and time
consuming and entailing a potential inaccuracy in locating the centers of fringes. Moreover, the
data is localized at the fringes, which are often irregular in shape. It necessitates the process of
transferring data to a regular grid by the interpolation procedure. The procedure of interpolation
also can potentially introduce numerical errors.
Phase-shifting interferometry (PSI) can be used to circumvent the difficulties referred above.
First developed by Bruning [Bruning, 1974], PSI involves (a) shifting the phase of one of the two
interfering beams so as to get a sequence of phase-shifted interferograms and (b) data reduction
to obtain the phase information. Several authoritative reviews exist on phase shifting
interferometry [Creath, 1986, 1987 & 1988; Greivenkamp, 1992; Cloud, 1995] and are adopted
here in explaining the evaluation steps involved. In general, phase shifting technique can be
incorporated into any type of interferometer system. The phase of the reference beam is shifted
with respect to that of the object beam in steps of π/4 or π/2 to get a sequence of phase-shifted
interferograms. Various techniques are available to achieve the phase shift [Wyant, 1975 &
1978; Shagam, 1978]. These include mounting the reference mirror on a piezoelectric transducer
that moves linearly in proportion to the applied voltage [Wyant, 1982 & 1985].
There are several algorithms reported in the literature for determining the phase of the
interferogram using PSI [Creath, 1986 & 1988]. A minimum of three phase-shifted
interferograms is required for determining the phase of the object wavefront. This is because the
interference equation
I = I ref + I obj + 2 I ref × I obj cos (φref − φobj )

Page 30 of 77
involves three unknowns, the reference and object beam intensities, and the phase difference
between the two interfering beams. The following three-step technique serves as an example:
Let the phase shift between successive interferograms be π/2 and the absolute phase for the
π 3π 5π
=
three interferograms be θ1 = , θ2 =
and θ3 . The set of three equations representing
4 4 4
the phase-shifted interferograms are solved simultaneously to get the phase at each point as
 I 3 ( x, y ) − I 2 ( x, y ) 
φ ( x, y ) = tan −1  .
 I1 ( x, y ) − I 2 ( x, y ) 
Therefore, by detecting the intensity modulation at each pixel and taking the above ratio, the
phase at each point of the wavefront is determined.
PSI involves an additional procedure to be performed on the phase calculated through the
arctangent equation in order to get the correct final description of the phase. This is necessitated
by the nature of the arctangent function, which lies between –π/2 to +π/2. As a result, phase
values fall within ±π/2 regardless of the actual spread of the phase. Therefore arctangent
calculations result in a wrapped up phase. The numerical process of phase unwrapping is carried
out in two steps:
(1) Obtain the phase values through the arctangent equation that are confined within ± π/2.
Noting the discontinuity at every |π/2| value, extend these values to fall between -π and π.
The phase obtained after this step is referred to as modulo 2π phase, and show discontinuity
after 2π radians.
(2) Remove the 2π discontinuities by staring from end of the image in order to obtain a true and
continuous map of the wavefront phase.
The phase map obtained by the above procedure is displayed as a gray image wherein the
gray value of each pixel corresponds to the phase value at that point. Using 8-bit digitization the
total phase shift (0 to 2nπ) is represented quantitatively by gray values in the range of 0 to 255.
Using higher bit digitization could further enhance the resolution of the phase measurement, and
as a result the resolution of the physical property being measured.
For PSI based on Mach-Zehnder interferometry, the change in refractive index ∆n( x, y )
corresponding to the phase change of ∆φ ( x, y ) is
λ
∆n( x, y ) = ∆φ ( x, y )
2π L

Page 31 of 77
where L is the length of the chamber through which the beam traverses. In order to get an idea
of the resolution of the PSI, a calculation with typical values of parameters in the above
equations will show that for a gray value change of unity between adjacent pixels, the minimum
detectable ∆h and ∆n is 1 nm and 10-5 respectively. They are an order of magnitude higher than
what is possible with conventional interferometry. It can be further increased by an order of
magnitude if 9 or 10 bit digitization is employed.
Finite time is required for translating the reference mirror mechanically in order to obtain
the phase shifted interferograms. Thus, conventional phase shifting interferometry has the
limitation of being unable to follow a process in real time. Nakadate et al. [Nakadate, 1990 &
1995] overcame this limitation ingeniously by employing polarized light to simultaneously
record the required number of phase-shifted interferograms.

1.5.3 Electronic speckle pattern interferometry (ESPI)

When a rough surface is illuminated with a coherent radiation, each facet of the object scatters
radiation towards the observer/detector. Here rough stands for random and microscopic
variations of height on the surface of the object with dimension more than the wavelength of
light. Since the phase of the light scattered will vary from point to point in proportion to the local
surface height, interference occurs between the light scattered from any two facets. Owing to the
superposition of individual interference patterns, each arising from a separate facet on the rough
surface, a random pattern of interference fringes, called speckle is observed. For a detailed
discussion on the speckle phenomena, the reader is referred to [Dainty, 1984].
Electronic speckle pattern interferometry (ESPI) has been used for measuring concentration
gradients. It can also be used to visualize convection in the form of variations in refractive index
in the solution. This technique can be adapted for mapping temperature as well as convection
patterns in fluids [Dupont, 1995; Verga, 1997]. In these applications, the speckle pattern is
obtained by means of a stationary ground glass plate. When a chamber containing the fluids is
interposed between the plate and the CCD camera, the optical path length is altered and a speckle
interference pattern of the experiment is obtained. As the optical properties of the fluid in the
chamber change, for example, due to changes in temperature, the phase of the speckle pattern
changes. It results in changes in the interference pattern between the object and the reference
speckles. This is subtracted from the speckle-interference-pattern recorded at the start of the

Page 32 of 77
experiment to get the phase variation produced during the time interval. Thereby the spatial
change in the field parameters such as refractive index and temperature are determined.

1.6 Simultaneous mapping of heat and mass transfer

In single-wavelength interferometry, the interpretation of the fringe pattern becomes difficult
when two independent quantities (such as temperature and concentration) affect the phase of the
traversing beam. The analysis is then based on the approximation that one of the two parameters
has a negligible influence on phase. The approximation may not be not valid in certain processes
wherein heat (temperature) and mass transfer (concentration) simultaneously affect the refractive
index of the aqueous solution and hence the phase of the beam traversing the crystallizer. Here
we need a diagnostic that can simultaneously map the two parameters. Dual-wavelength
interferometry is capable of detecting temperature and concentration and is described below.

1.6.1 Dual-wavelength interferometry

Dual-wavelength interferometry can separate the influence of the two competing parameters
such as temperature and concentration, provided the sensitivities dn dC to dn dT are
significantly different for the two wavelengths. This problem is similar to solving two equations
for two unknown variables. Using two different wavelengths for mapping the process results in
two independent fringe patterns. Dual-wavelength interferometry was first used by El-Wakil and
Jaeck to study heat and mass transfer in gaseous boundary layers [El-Wakil, 1964]. Later, Ecker
proposed the concept of holography in conjunction with dual-wavelength interferometry for
measuring temperature and concentration fields during alloy solidification [Ecker, 1987 & 1988].
Mehta (1990) and Vikram et al. (1990) reported the applicability of this technique for
determining the temperature and concentration fields inside liquids. The principle of dual-
wavelength interferometry (and holography) as explained in a few reports by Vikram and co-
workers [Vikram, 1992A] is described below.
For processes influenced simultaneously by the temperature and the concentration fields,
the total change in refractive index can be expressed as
∂n ∂n
∆n= ∆C + ∆T
∂C ∂T

Page 33 of 77
where ∆C and ∆T are the changes in the concentration and temperature fields influencing the
refractive index field respectively. Combining the two equations given above, we get
N λ ∂n ∂n
= ∆C + ∆T .
L ∂C ∂T
Extending the above discussion to the dual-wavelength interferometry, we get two independent
equations corresponding to the two source wavelengths, λ 1 and λ 2 , which are then solved for

independent estimates of the thermal and the concentration variations [Vikram, 1992A]. Thus
∂n ∂n
N1λ 1 2 − N 2 λ 2 1
∆C = ∂T ∂T
 ∂n ∂n ∂n ∂n 
L 1 2 − 2 1
 ∂C ∂T ∂C ∂T 
and
∂n ∂n
N1λ 1 2 − N 2 λ 2 1
∆T = ∂C ∂C .
 ∂n ∂ n ∂ n ∂n
L 1 2 − 2 1
 ∂T ∂C ∂T ∂C 
For accurate determination of ∆C and ∆T , we must know precisely the refractive index and its
gradients with respect to temperature and concentration at both the wavelengths and the fringe
order. In the absence of experimental values, a suitable analytical methodology has been put
forth by Vikram and co-workers [Vikram, 1991, 1992B] for determining the needed data. The
refractive index is obtained using the two-constant Cauchy equation. A higher order Cauchy
equation involving more constants may be used for greater accuracy. The refractive index
gradients with respect to temperature and concentration are determined using Murphy-Alpert
[Murphy, 1971] and Lorentz-Lorenz relationships respectively.

1.7 Computerized tomography

Since conventional optical techniques can only measure the average refractive index along the
optical path (e.g. the thickness of the process chamber), quantitatively meaningful results are
possible only in the following situations: Either the experiment is "two-dimensionalized" so that
the parameter of interest is constant along the optical path; or truly three-dimensional imaging
techniques are adopted for experimentation, such as holography [Gabor, 1948 & 1949; Hariharan,
1986 & 2002] for a direct 3-D imaging or computerized tomography for an indirect 3-D

Page 34 of 77
visualization from a set of two-dimensional projection images. The principle of holography and
holographic interferometry has already been discussed in section 1.4.4 of this chapter. The
principle of tomography is briefly presented below.
Computerized tomography is defined mathematically as the process of constructing a
continuous two-dimensional function from its one-dimensional line-integrals obtained along a
finite number of lines at known locations [Ramachandran, 1971; Herman, 1980; Natterer, 2001].
This can be extended to include reconstruction of a three-dimensional function from its two-
dimensional projections taken along a fixed number of known directions. Tomography is
essentially a two-step process: first, collecting the projection data and second, reconstructing the
three-dimensional function using numerical algorithms.
The tomographic reconstruction methods can be broadly classified as (1) transform methods,
(2) series-expansion methods, and (3) optimization methods. The transform methods are
computationally efficient, but require a large number of experimental projections to recover the
process parameter with some satisfactory accuracy [Lewitt, 1983]. In contrast, the series
expansion methods are computationally intensive, but have the advantage of giving satisfactory
result even with limited or partial projection data [Censor, 1983]. The optimization approach
requires the selection of a suitable functional that has to be extremized, under the experimental
constraint [Gull, 1986].
Various forms of optical tomography have numerous potential applications in many diverse
fields of science and engineering. We shall confine our discussion here to the application of
shadowgraphic and interferometric tomography to the study of transport processes. In these cases,
the line integrals of refractive index can be converted to point-by-point maps of refractive index
at selected planes.

1.7.1 Shadowgraphic tomography

A three-dimensional pattern of varying refractive index inside the process chamber results in
light rays getting refracted out of their original path. This caused a spatial modulation of the
intensity distribution with respect to the original intensity and produced a shadowgraph pattern
on the screen. Shadowgraph images recorded from several view angles constitute the
experimental projection data. The intensity of each pixel in the shadowgraph images is processed
numerically to extract the refractive index distribution. This constitutes the numerical input to the

Page 35 of 77
reconstruction algorithms. It should be noted that refractive index distribution has one-to-one
correspondence with the physical parameter, such as temperature, concentration or fluid flow
field in the process chamber. The output of the tomography reconstruction algorithms is a matrix
representing the process parameter values over a particular plane in the region of interest. A
detailed discussion on the application of shadowgraphic tomography to the process of crystal
growth can be found in the work of Verma et al. [Verma, 2006 & 2008B].

1.7.2 Interferometric tomography

In case of interferometric tomography, the projection data is in the form of interferograms
captured from several angular directions. For example, a Mach-Zehnder interferometer can be
used to capture interferograms of the process from several coplanar viewing directions. Such
infinite (or wedge) fringe interferograms recorded nearly simultaneously from different view
angles constitutes one set of 2D projection data. These interferograms are numerically processed
to get the refractive index distribution, which serve as input to the tomography algorithms. It
should be noted that data reduction in case of shadowgraph images is two orders of magnitude
difficult as compared to interferograms due to analytical expression relating intensity with the
refractive index field in the two processes. Finally, the methodology of implementing the
reconstruction algorithms for interferometric tomography is similar to that for shadowgraphic
tomography. Reader is referred to Mishra et al. [Mishra, 1998 & 1999] and Muralidhar [2001]
for in-depth details of interferometric tomography as applied to heat transfer problems.

1.8 Comparison of optical techniques

Each technique described above is accompanied by certain advantages and weaknesses.
Shadowgraph and schlieren have the advantages of low cost, ruggedness and simplicity of
apparatus, and ease of real-time qualitative interpretation. Gradient interferometers provide an
attractive trade-off between simplicity and versatility. Two-beam interferometers are probably
the easiest to setup and permit quantitative information. Reconstructive holography has the
important advantage of providing rich three dimensional detail. With regard to sensitivity, an
analysis provided in pp. 150-155 of [Merzkirch, 1974] indicates that schlieren, gradient
interferometry and two-beam interferometry all have about the same order of magnitude of
sensitivity (about λ 100 ). The trade-offs between conventional and holographic interferometry

Page 36 of 77
are particularly complex. Since holographic interferometry operates in a time-differential, rather
than space-differential, manner, it has several advantages over classical interferometry. Unlike
the latter, it does not depend critically on the alignment of the two interfering beams. Moreover,
the need for high quality optical components is much less critical. Holography also provides
capabilities not possible with conventional schlieren and interferometry. A collimated-beam
hologram can be made of an experiment and then reconstructed into a real image that can then be
processed at leisure to generate schlieren and gradient-interferometric images of variable
orientation and fringe width [Smigielski, 1970; Merzkirch, 1987]. By using stored-image
holography or by making time-lapse double exposures on the given holographic plate,
interferograms which image the change in optical path length with time can be generated
[Heffinger, 1966]. All these advantages, however, must be weighed against the disadvantages of
lack of real-time viewing, labor of reconstruction, and incompatibility with simultaneous
mapping of concentration and temperature fields. Electronic-speckle pattern interferometry is
different from other interferometric and holographic techniques, in that it can be used under
ordinary room lighting conditions. It does not require vibration isolation tables [Lokberg, 1980;
Jones, 1989]. Although it is less sensitive to temperature/concentration changes compared to the
conventional techniques of interferometry and holography, it can provide useful data under non-
ideal laboratory conditions. Since ESPI functions as a time-difference interferometer it is similar
to holographic interferometry. Hence, it does not require accurate adjustments of the interfering
beams. Unlike holographic interferometry, it has the advantage that the information is directly
recordable onto the CCD sensor without any intermediate recording plate. The primary drawback
is the fringe quality, which is quite noisy due to the speckle property itself. As a result, the
spatial resolution of the camera limits the resolution of the measurement.

2. APPLICATIONS
The second part of the chapter describes various applications of the above techniques in studying
transport phenomena as encountered in engineering applications. These are: (i) crystal growth
from solution [Verma, 2007, 2008 & 2009; Dinakaran, 2010], (ii) heat conduction phenomenon
in a differentially heated layer of water [Singh, 2009], (iii) mass transfer from high concentration
region to low concentration (dissolution of sugar) [Bhandari, 2009], (iv) natural convection in a
differentially heated rectangular cavity of square cross-section (hot lower wall and a cold upper

Page 37 of 77
wall) filled with air [Muralidhar, 2005], (v) buoyancy-driven convection in an octagonal cavity
half-filled with 50 cSt silicone oil, the rest being air [Punjabi, 2002], (vi) long-time
interferograms formed in an octagonal cavity containing silicone oil (50 cSt) floating over water
[Punjabi, 2004], (vii) buoyancy-driven convection in a differentially heated cavity that is
rectangular in plan (Rayleigh-Benard configuration) [Mishra, 1998], (viii) convection in an
eccentric annulus with the gap filled with air [Ranjan, 2005], (ix) steady state heat transfer in an
eccentric annulus when the gap is filled with 390 cSt silicone oil [Ranjan, 2005], and (x)
convection and heat transfer behind heated circular cylinder at low Reynolds numbers [Singh,
2007].

2.1 Transport phenomena during crystal growth from solution

Crystal growth from solution is the most widely used method for growing large crystals, several
centimeters in size [Mullin, 2001]. It is applicable to all classes of materials, including inorganic,
organic, metal-organic, macromolecules and complexes. Solution growth offers the convenience
of low operating temperatures, less complicated apparatus, and simple instrumentation. More
importantly, when the process to be imaged uses optical radiation, the medium under study has
to be necessarily transparent at the chosen wavelength of radiation (He-Ne laser wavelength in
our experiments). Among the various methods of growth, only solution growth satisfies this
requirement. For these reasons, the present work is primarily focused on crystal growth from
solution. An ensemble of crystals grown using solution growth are [Paschotta, 2008]: KDP
(potassium dihydrogen phosphate, KH2PO4) and its deuterated analogue DKDP (KD2PO4) for
frequency conversion and electro-optic switching (Q-switches and Pockels cell) respectively;
TGS (triglycine sulphate) for laser-energy measuring devices; KAP (potassium acid phthalate)
for monochromator applications; LiIO3, ZTS [zinc tris (thiourea) sulphate], and many organic
and metal-organic crystals for linear and non-linear optical applications.

2.1.1 Importance of mapping solution growth

The ultimate goal of research on crystal growth is to achieve an in-depth understanding of the
process with the aim of improving the microscopic and macroscopic homogeneity of the grown
crystals, in addition to increasing their size. A crystal growing from solution creates thermal and
concentration gradients in the surrounding solution by releasing the heat of crystallization and

Page 38 of 77
depleting the solute near the growth surface. The resultant temperature and concentration
gradients affect the perfection and stability of the crystal grown [Bunn, 1949]. The change of
solution density with temperature (dρ/dT) is negative and the change of solution density with
solute concentration (dρ/dC) is positive. Therefore, crystal growth in the Earth's gravitational
field is accompanied by a rising buoyant convection current which envelops the crystal, is often
oscillatory and unstable, and therefore drastically modifies the concentration gradient along the
growth interface [Chen, 1979A & 1979B]. Thus, the growth history and defect structure of the
crystal is a function of the time-dependent spatial distribution of the convection patterns, and of
the temperature and concentration profiles in the surrounding solution. Therefore, in order to
learn how to grow large defect-free crystals, it is helpful to map the spatial distribution of
thermal and concentration profile around the growing crystal.

2.1.2 Phenomenology of the growth process

The rate of crystallization is determined by its slowest stage. Two major stages can be identified:
(1) supply of crystallization material from the solution to the crystallization surface; (2)
incorporation of this material into the crystal structure, i.e. growth of the crystal. If the first stage
is the limiting step, the growth is said to proceed in the diffusion regime. If the second stage is
the limiting step, growth is said to be operating in the kinetic regime. Diffusion regime
conditions are prevalent during free convection growth, also referred as buoyancy-driven growth.
In practice, forced flow conditions are adopted. These conditions bring the process into the
kinetic regime of growth and are achieved by employing various modes of stirring. Here the
solute is forced to move from the bulk solution towards the crystal surface. There is increased
probability of occurrence of defects under irregular or turbulent flow conditions. Under such
conditions the growth rate is high. Therefore, the mass transfer and the fluid flow conditions
have to be optimized to avoid deleterious effects such as morphological instability, spurious
nucleation, and inclusions. Thus, of particular importance in crystal growth from solution are
fluid flow, and heat and mass transfer. Before discussing the results of crystal growth imaging, it
is pertinent that a brief description is given of the experimental steps involved in growing a
crystal from its aqueous solution.

Page 39 of 77
2.1.3 Supersaturation: Driving force for crystal growth from solution
The prerequisite for crystallization to take place is the creation of a suitable driving force for the
nucleation to occur. In solution growth, this driving force is referred to as supersaturation. It is a
state of the solution achieved by manipulating its thermal behavior. It can be achieved in the
following three ways: (a) Cooling: Cooling a saturated solution will result in a supersaturated
solution, (b) Evaporation: Evaporation of the solvent leads to an increase in concentration of the
solution, leading to a supersaturated solution, and (c) at times, a combination of cooling and
evaporation is adopted.
The supersaturated state is created by having prior information about the solubility of the
material in an appropriate solvent. Solubility of a material in a solvent decides the amount of the
material that can be dissolved in it at a particular temperature. If the solubility is too high, it is
difficult to grow a large crystal from such a solution. This is because of its sensitivity to thermal
fluctuations. At the same time, if the solubility is low, it restricts the growth of large crystals
since enough solute is not available for crystallization. The solubility of the material is measured
by dissolving it in a continuously stirred solvent; this is done at several temperatures. A
relationship so obtained between concentration and temperature fields is referred to as the
solubility diagram.
A small good quality crystal is used as seed for initiation of the growth. While immersing
the seed crystal, the solution temperature is kept 0.5 degrees above its saturation value so as to
dissolve a few surface layers of the crystal. The dissolution helps in the elimination of the
physical imperfections on the crystal faces, and also of any surface contamination introduced
during crystal processing. After dissolving the crystal for a few minutes, the solution is cooled to
its saturation point. A programmed cooling of the solution with an appropriate cooling rate is
given from this point onwards to continue the growth. The difference in the concentration at
saturation temperature and the temperature at which the experiment is performed gives us the
extent of supersaturation of the solution. This is the driving force for the growth process and is
computed at each stage of imaging experiment.

2.1.4 Optical imaging of crystal growth from solution

Several research groups working in the field of thermal sciences have also used it for mapping
convection during various engineering processes [Rasenat, 1989; Mayinger, 1993; Schopf, 1996;

Page 40 of 77
Goldstein, 1996; Settles, 2001]. However, this technique has not been widely used for crystal
growth studies. Shiomi et al. used shadowgraphy to measure the temperature profile around a
growing Rochelle salt crystal [Shiomi, 1980], while Lenski et al. used a beam deflection
technique to measure convection during vapor crystal growth [Lenski, 1991]. Recently, Verma et
al. [Verma, S., 2003, 2005C & 2006] have used shadowgraphy for visualizing the buoyancy
driven convective patterns around a growing KDP crystal, and studied the gradual build up of
this activity with the dimension of the crystal and the time of the growth. A few representative
images, shown in Figure 13, demonstrate that buoyant convection increases with the increase of
the Grashof number, which is the fluid dynamical dimensionless number representing the
strength of the buoyancy driven convection. With the progress of growth, the convection mode
changes from laminar (Figure 13a) to chaotic (Figure 13f). The influence of free and forced
convection, and the cooling rate on the growth rate and the quality of the grown KDP crystal has
also been studied [Verma, S., 2007]. Recently, Dinakaran et al. studied the polar morphology of
zinc tris (thiourea) sulphate (ZTS) crystals using this technique [Dinakaran, 2010]. The influence
of growth geometry on convection and the morphology of the grown ZTS crystals is shown in
time lapsed shadowgraph images shown in Figure 14.
The schlieren technique was used by Chen et al. measured the magnitude and stability of
convection around growing sodium chlorate crystals [Chen, 1979A & 1979B]. Chen also studied
the relationship between the convection irregularities observed over a growing sodium chlorate
crystal, and formation of defects such as fluid inclusions [Chen, 1977]. Onuma et al. used
schlieren techniques to study the buoyancy driven convection above a growing barium nitrate
crystal [Onuma, 1988 & 1989]. Srivastava et al. used schlieren for visualizing convection around
a growing KDP crystal [Srivastava, 2004].
Rashkovich and co-workers used a Mach-Zehnder interferometer to measure the
diffusion boundary layer thickness under different flow velocities of the solution flowing past the
vicinal facets of ADP and KDP crystals [Rashkovich, 1990A & 1990B]. Duan et al. used the
high sensitivity of phase-shifting Mach-Zehnder interferometry to map minute convective
features around a growing NaClO3 crystal [Duan, 2001]. Recently, Verma et al. have used such
an interferometer in both infinite-fringe-width and wedge fringe modes to study the convective
field around a KDP crystal growing under free and forced convection conditions in both platform
and suspended crystal geometries, and correlated their results with the growth rate and quality of

Page 41 of 77
the crystal [Verma, 2005B & 2007]. Figure 15 shows the concentration gradient inside the
crystallizer during KDP growth under buoyancy driven convection conditions. Due to the growth
of crystal, a low density convection plume rises up from the crystal resulting in accumulation of
low density solution in the upper part of the crystallizer as compared to the lower portion of the
crystallizer. This results in a concentration gradient inside the crystal, which manifests itself in
the form of horizontal fringes in case of infinite fringe Mach-Zehnder interferometric imaging.
This condition is detrimental to the growth of good quality optical crystals and should be avoided.
Bedarida’s group used transmission holographic interferometry as a tool for controlling
NaClO3 crystal growth by monitoring the convection around the growing crystal [Bedarida,
1977]. The Chinese group of Yu Xiling adopted real-time, single-exposure holographic
interferometry for measuring the diffusivities of KDP and DKDP solutions as a function of
temperature and concentration [Xiling, 1992]. The same technique was used to study the
convection driven mass-transport processes at the solid-liquid interface during the growth of
DKDP crystals [Xiling, 1996].
Optical techniques have been extensively used for imaging the process of crystal growth
aboard space flights. For example, Owen and Shlichta constructed a multimode optical
observation system (MOOS) for observing crystal growth process using schlieren, shadowgraphy,
absorption and shearing interferometry aboard NASA space shuttle [Owen, 1975; Shlichta,
1985]. Owen also used a Mach-Zehnder interferometer and a holographic unit for studying the
growth of NH4Cl and TGS crystals under microgravity environment [Owen, 1982 & 1986].
Single exposure holograms were recorded to study the crystal profile microscopically; the
interferometric data was used to obtain concentration and temperature profiles near to the crystal;
double-exposure holographic interferometry was performed to obtain details of the refractive
index changes occurring during growth; and the holographic images were used to observe the
development of crystal morphology, growth rate, and the surface structure. Witherow used a
multipurpose holographic system used for investigating the TGS crystal growth experiment
aboard Spacelab3 [Witherow, 1987]. The methodology of generating shadowgraph, schlieren,
and interferometric images of the growth process in a ground-based laboratory from the
holograms recorded in space was also discussed. Solitro, Gatti, Bedarida, and their co-workers
developed a fiber-optic based multidirectional holographic interferometer for monitoring
convection during crystal growth in space [Solitro, 1989; Gatti, 1989; Bedarida, 1992]. The

Page 42 of 77
special volume containing specific contributions dealing with the development of advanced
optical diagnostics for crystal growth applications in space was compiled by Trolinger and Lal
[Trolinger, 1991]. Experiments on later missions made use of phase-shift interferometry (PSI)
and multicolor holographic interferometry (MCHI), developed at Space Sciences Laboratory of
the Marshall Space Flight Centre (MSFC) for detecting the temperature and concentration field
gradients during crystal growth [Witherow, 1994]. Dubois et al. describe the development of an
integrated optical set-up for fluid-physics experiments, including crystal growth [Dubois, 1999].
The features of the schlieren technique, electronic speckle pattern interferometry (ESPI),
differential interferometry, digital holography, and holographic interferometry were discussed.
The performance of different diagnostics was compared for temporal resolution (the recording
speed, which is dependent on the hologram refreshing rate), spatial resolution (sensitivity), and
dynamic range, using the benchmark experiment of refractive index measurement in Rayleigh-
Bénard instability.
The availability of high quality protein crystals is vitally important for determining their
structure and subsequently for drug development [McPherson, 1982 and 2001; Ducruix, 1992].
Shlichta presented a summary of the optical techniques available for imaging crystal growth
from solution in general, and for protein crystal growth in particular. The European Space
Agency (ESA) has developed a Protein Crystallization Diagnostics Facility (PCDF) equipped
with several optical imaging diagnostics for on-line and in-situ monitoring of the growth
phenomena aboard the International Space Station (ISS) [Pletser, 2001]. Cole et al. mapped
refractive indices around growing protein crystals by Mach-Zehnder interferometry [Cole, 1995].
McPherson et al. used a phase-shift Mach-Zehnder interferometer to observe the fluid
environment around a growing protein crystal in the ground [McPherson, 1999] and in
microgravity [McPherson, 2001]. Yin et al. used a Mach-Zehnder interferometer to study the
influence of magnetic field on the convection, concentration, solubility and temperature during
growth of lysozyme crystals [Yin, 2001 & 2003].
Several research groups have used Mach-Zehnder interferometer for solutal and thermal
field studies of crystal growth process. Onuma et al. measured the surface supersaturation under
different flow velocities of the solution flowing over the face of a growing K-alum crystal
[Onuma, 1989]. Nakadate and Yamaguchi developed a real-time phase shifting Mach-Zehnder
interferometer for concentration measurements [Nakadate, 1990], which was later adopted by

Page 43 of 77
Onuma et al. and Kang et al. for concentration measurements around NaClO3 crystal [Onuma,
1993; Kang, 2001]. Kim et al. used a similar interferometer to measure concentration variations
at the interface between the solution and the crystal of L-arginine phosphate (LAP) [Kim, Y.,
1998A]. Nagashima and Furukawa used this type of interferometer to study the solute
distribution during directional growth of ice crystals, and to understand the origin of
morphological instability leading to the dendritic growth [Nagashima, 2000]. Chen et al. used
the Mach-Zehnder interferometer as an aid to monitor and control the growth of large size
NaClO3 crystals [Chen, 2002].
Verma et al. used a Mach-Zehnder interferometer in infinite and wedge fringe settings for
mapping convection and concentration field [Verma, 2009]. Figure 16 shows infinite fringe and
wedge fringe interferograms of a crystal growing under free convection conditions in top
hanging geometry. The free convection plume and the diffusion boundary layer are clearly
visible in infinite fringe interferograms but not in wedge fringe interferogram. Figure 17 shows
infinite and wedge fringe interferograms of a crystal growing under forced convection conditions
in top hanging geometry. No buoyancy driven convection plumes are visible as the crystal is
rotating at high RPM and the concentration is uniform inside the crystallizer. Minor fringe
distortion can be seen near to the crystal-solution interface. Figure 18 shows infinite fringe and
wedge fringe interferograms of the crystal growing on a platform under free convection
conditions. The concentration gradient is clearly visible in the third interferogram. The free
convection plume and the diffusion boundary layers are visible in the infinite as well as wedge
fringe interferograms.
Onuma et al. used the real-time phase shift interferometer, developed by Nakadate and
Yamaguchi, to measure the concentration field around growing and dissolving NaClO3 crystal
[Onuma, 1993; Nakadate, 1990]. Maruyama and co-workers employed a real-time PSI to study
the thermal and mass diffusion fields of the aqueous solutions of NaClO3 and Ba(NO3)2 under
microgravity conditions [Maruyama, 1999 & 2002]. Duan et al. used real-time PSI to map
convection during NaClO3 crystal growth [Duan, 2001]. McPherson et al. used a phase-shift
Mach-Zehnder interferometer to observe the fluid environment around a growing protein crystal
on the ground [McPherson, 1999] and in microgravity [McPherson, 2001].
Piano and co-workers have used speckle pattern interferometry for measuring the
refractive index variation around a growing KDP crystal [Piano, 2000 & 2001]. The applicability

Page 44 of 77
of dual-wavelength holographic interferometry for studying triglycine sulfate crystal growth has
been discussed by Vikram and co-workers [Vikram, 1991 & 1992A]. Witherow et al. [Witherow,
1994] describes the efforts in combining the dual-wavelength interferometry with the phase-
shifting interferometry for studying crystal growth experiments under microgravity environments.

2.2 Heat conduction in a horizontal fluid layer

Figure 19 provides snapshots of interferograms that reveal the heat conduction phenomenon in a
differentially heated layer of water [Singh, 2009]. The container is a closed cavity, insulated on
the sides, with the top surface maintained at 23oC and the base at 18oC. The room temperature
during the experiments was maintained at 20oC. The temperature difference creates a density
field and in turn, a refractive index field, thus providing a basis for the formation of
interferograms. Since fluids expand with an increase in temperature, the temperature differential
creates a density stratified fluid layer, with hot fluid floating over the cold. The thermal field is
imaged using a Mach-Zehnder interferometer in the infinite fringe setting. To set the
interferometer initially, a compensation chamber filled with water at room temperature is used in
the path of the reference beam. Since the surface temperatures are different from the initial value,
fringes emerge from the top and the bottom till steady state is reached. Steady state is
characterized by straight equi-spaced fringes. The fringe curvature during the initial stages could
be the result of imperfect optical alignment that balances the arms of the interferometer. At later
times, this error in units of optical path difference is small in comparison to the path difference
created by thermal disturbances.

2.3 Dissolution of sugar in water

In Figure 20, a layer of sugar applied at the base of container is dissolving [Bhandari, 2009]. On
evaporation of water, a layer of sugar adheres to the surface. Water is poured in the container till
it gets filled. A compensation chamber is used to balance the arms of the interferometer and
arrive at an infinite fringe setting. With the passage of time, sugar slowly dissolves in water. The
concentration of sugar in water, thus, changes with time. It is accompanied by the appearance of
interferometric fringes. Water adjacent to the lower wall has the highest sugar concentration,
diminishing progressively towards the upper wall. During the transient phase, a layer of clear
(sugar-free) water is obtained in the upper half of the cavity. Though the physical problem is

Page 45 of 77
driven by mass transfer, the overall situation is analogous to heat transfer in a differentially
heated cavity. There is point of difference, however. Sugar at the base creates a region of high
concentration, high density and high refractive index while the other wall is one of zero sugar
concentration, as in the compensation chamber. Thus, fringes originate from the lower wall alone.

2.4 Buoyancy-driven convection in a square cavity

Figure 21 shows natural convection in a rectangular cavity of square cross-section [Muralidhar,
2005]. The fluid medium in the cavity is air and convection is driven by the temperature
differential between a hot lower wall and a cold upper wall. The side walls are insulated. Since
the lower wall is heated, it is a region of low density when compared to the top surface. Thus,
cold fluid is overlaid the hot in a gravitationally unstable configuration. The resulting fluid
motion distorts the temperature distribution. When the temperature difference is large, the
strength of fluid convection is great and large temperature changes occur close to the boundaries.
This phenomenon is akin to the formation of boundary-layers - both hydrodynamic and thermal.
The symbol Ra is Rayleigh number and is a measure of the strength of natural convection in the
cavity. The first column carries interferometric fringes. The second and third columns contain a
re-distribution of light intensity as obtained from schlieren and shadowgraph techniques. In
schlieren, the initial light intensity is rather low but can be suitably adjusted. When the
temperature differential across the fluid layer is turned on, regions of high temperature gradients
are uniformly illuminated, with low gradient regions remaining dark. In a shadowgraph, the
initial intensity distribution is set at a convenient initial value. Temperature gradients can result
in either a local increase in intensity or a reduction all the way to zero. Thus, the intensity field
may brighten or darken about the initial value. In addition, the resulting intensity of shadowgraph
depends on the position of the screen.

2.5 Convection in a fluid layer with a free surface

Figure 22 shows buoyancy-driven convection in an octagonal cavity half-filled with 50 cSt
silicone oil, the rest being air [Punjabi, 2002]. The cavity has eight sides and forms an octagon in
plan. The lower surface is heated while the top surface is cooled. The thermal arrangement
creates cold/heavy fluid at the top floating over hot/light fluid at the bottom, in a gravitationally
unstable arrangement. Thus convection patterns are possible in both fluid layers – air and

Page 46 of 77
silicone oil. However, the thermal conductivity of air is small when compared to silicone oil.
Hence, a considerable fraction of the overall temperature difference occurs in oil and the
temperature difference across air is negligibly small. Neither is air set in motion nor do clear
fringes form in air. In contrast, convection is set up in silicone oil and fringes that are isotherms
are deformed. The omega-shaped fringe (as well as others that are similar but show smaller
deviation) is a result of projecting an axisymmetric thermal field along the viewing direction.
Figure 22 shows interferometric projections of the refractive index distribution when the thermal
field is viewed in various directions – 0, 45, 90, and 135o. Since the figures are practically
identical, one may conclude that the thermal field does not depend on the polar angle and may be
classified as axisymmetric.

2.6 Convection in a two-fluid layer system

Figure 23 collects long-time interferograms formed in an octagonal cavity containing silicone oil
(50 cSt) floating over water [Punjabi, 2004]. Interferometric projections recorded from four
different angles are shown. The flow is driven by an overall temperature difference, the lower
surface being warmer than the top surface. Thus, cold/dense fluid set over warm/light fluid
generates buoyancy forces and drives fluid convection. Water and oil are immiscible and
independent convection patterns are set up in each part of the cavity. Since the fluid
conductivities are comparable, a finite temperature difference is available across each fluid layer
and detectable fluid convection is to be seen in each half. The overall temperature difference
across the cavity is 1.8 K, the top plate being cooler than the one at the base. Of special interest
are the energy and momentum transfer at the oil-water interface. When water is set in motion, it
can drive the oil layer beneath even if buoyancy is absent. The underlying mechanism is
viscosity and the fluid layers then said to be mechanically coupled. The other possibility is that
each fluid layer experiences buoyant convection whose strength scales with the temperature
difference appropriate for each of them. The fluid layers are then said to be thermally coupled. In
the present experiment, interferometric images show that the nature of coupling is thermal in
origin. In addition, the images look different in each view angle, indicating the onset of three
dimensional velocity and temperature fields.

Page 47 of 77
2.7 Rayleigh-Benard convection in a rectangular cavity
Figure 24 reports images and analysis related to buoyancy-driven convection in a differentially
heated cavity that is rectangular in plan [Mishra, 1998]. The height of the cavity is small in
comparison to the linear dimensions of the rectangle (around 10%). The lower surface is warmer
than the top. The overall arrangement is the familiar Rayleigh-Benard configuration. The fluid
medium considered is air. For temperature differences and cavity height considered, three
dimensional fluid motion and temperature distribution are to be expected. The circulatory motion
of the fluid particles creates a roll-type of motion. The roll patterns of the flow field are seen as
fringe displacement in the interferograms. The corresponding interferograms seen from various
viewing directions are shown in the figure. The schematic drawing at the top indicates how a roll
is created within the cavity. Here, a parcel of warm fluid rises as a buoyant plume at a certain
location, spreads on its way upwards, reaches the cold top, and descends downwards at a second
location by gravity. Points where the fluid rises with those where the fluid descends together
form a cellular pattern within the cavity. The geometry of the pattern can be understood by
recording interferograms from various directions. A sample interferogram is shown in the figure.
Here, fringe displacement upwards can be correlated with the rising buoyant plume. Conversely,
fringe displacement downwards shows the descent of the cold fluid. The entire thermal field in
can be reconstructed by a numerical procedure called tomography. Here, temperatures integrated
along the viewing direction are required over the image as input data. An intermediate step that
assists conversion of fringe patterns to temperature is fringe thinning. It replaces the fringe bands
by a skeleton, using image processing techniques. A sample set of thinned interferograms is also
shown in Figure 24, for various view angles.

2.8 Convection in an eccentric air-filled annulus

Figure 25 shows a time sequence of interferograms recorded in an eccentric annulus with the gap
filled with air [Ranjan, 2005]. The axes of the two cylinders are horizontal. The cylinder length
is roughly ten times the outer cylinder diameter. The inner cylinder is heated while the outer
cylinder is kept at the ambient temperature. Heating is by an electric coil wound over a bakelite
rod and covered by mica. The heating assembly is placed within the inner cylinder.
Thermocouples have to be located over the cylinder length to ensure temperature uniformity
along the axis as well as in the angular direction. Initially, the fluid in the annulus is also at the

Page 48 of 77
ambient temperature. The fringes are recorded in the infinite fringe setting. Fringes originate
from the inner heated boundary. For the temperature difference considered, density differences
are large enough for buoyant convection to be set up in the annulus. The fluid is driven upwards
near the inner cylinder and flows around the inner periphery of the outer cylinder during the
descent. Thus, fringes show upward displacement closer to the inner cylinder. Owing to the
eccentricity of the inner cylinder in its placement with respect to the outer, symmetry in the
fringe patterns is broken. The stem attached to the inner cylinder holds it in place and is visible
in the recorded images. The fringe patterns yield information on the distribution of heat flux
around the inner cylinder. Here, the wall heat flux scales with the derivative of temperature in the
radial direction. Since large derivatives will create closely spaced fringes, fringe spacing itself is
indicative of the wall heat flux. The interferograms show that fringe spacing is small on the
lower side of the cylinder and consequently, the wall fluxes are large. On the top side, fringe
spacing is large and the wall fluxes can be expected to be small (in comparison). It is also to be
understood that fringe thickness scales inversely as the local temperature gradient. Thick fringes
are regions of low gradients. Thus, above the inner cylinder, one obtains a region of warm fluid
of nearly uniform temperature.

2.9 Convection in an eccentric annulus filled with silicone oil

In Figure 26, interferograms recorded during steady state heat transfer in an eccentric annulus are
presented when the gap is filled with 390 cSt silicone oil [Ranjan, 2005]. The applied
temperature differences increase from 0.2 to 2 K while going from (a) to (e). When compared to
air, several differences are to be observed in the fringe patterns. These include a substantial
increase in the number of fringes. In addition, strong refraction errors are seen errors near the
inner cylinder where the temperature gradients are large. However, the thermal fields retain
symmetry about the position of the inner cylinder indicative of weak convection effects in the
annulus. For small temperature differences, the buoyancy forces cannot overcome the viscous
and the fluid is stationary. Heat transfer is then by conduction alone, as seen by near circular
fringes for a temperature difference of 0.2 K. In liquids such as silicone oil, experiments with
interferometry are to be carefully designed so that refraction errors are small and the fringe
density is matched to the camera resolution.

Page 49 of 77
2.10 Vortex shedding from a heated cylinder
Figure 27 shows instantaneous schlieren images for flow past a circular cylinder [Singh, 2007].
Wakes behind heated circular cylinders have been experimentally investigated at a low Reynolds
number (100-300). The electrically heated cylinder is mounted in a vertical airflow facility such
that the buoyancy aids the inertia of the main flow. The operating parameters, i.e. Reynolds
number and Richardson number are varied to examine flow behavior over a range of
experimental conditions from the forced to the mixed convection regime. Laser schlieren-
interferometry has been used for visualization and analysis of the flow structures. At the
Reynolds number studied, the unheated cylinder will periodically shed vortices. The complete
vortex shedding sequence has been recorded using a high-speed camera. The results on detailed
dynamical characteristics of the vortical structures i.e. their size, shape and phase, Strouhal
number, power spectra, convection velocity, phase shift, vortex inception length and fluctuations
are reported. On heating, the alteration of organized (coherent) structures with respect to shape,
size and their movement is readily perceived from the instantaneous Schlieren images before
they reduce to a steady plume. For the circular cylinder, Strouhal number shows a slow increase
with an increase in the cylinder temperature, namely the Richardson number. At a critical value,
there is complete disappearance of vortex shedding and a drop in Strouhal number to zero. The
corresponding spectra evolve from being highly peaked at the vortex shedding frequency to a
broadband appearance when vortex shedding is suppressed. The geometry of the vortex
structures transforms to a slender shape before shedding is suppressed. The convection velocity
of vortices increases in the streamwise direction to an asymptotic value and its variation is a
function of Richardson number. The convection speed abruptly falls to zero at the critical
Richardson number. The phase difference of the shed vortices between upstream and
downstream location increases with an increase in Richardson number. Velocity profiles show an
increase in fluid speed and beyond the critical point, buoyancy forces add enough momentum to
cancel the momentum deficit due to the cylinder.

Page 50 of 77
 (a) (b)

(c) (d)

(e) (f)

Figure 13: Shadowgraph images show an increase in the strength of buoyancy-driven convection
with the growth of the KDP crystal. (a) No buoyant plume is seen at the start; (b) A stable plume
start to appear as the growth begins; (c-e) Buoyant convection intensifies as the crystal grows; (f)
The buoyant plume changes from laminar to irregular and finally becoming chaotic. Images
adapted from Imaging techniques for mapping of solution parameters, growth rate, and surface
features during the growth of crystals from solution, Sunil Verma and P.J. Shlichta, Prog. Cryst.
Growth & Charact. Materials, 54 (2008) 1-120.

Page 51 of 77
 (a)

(b)

Figure 14: In order to measure the growth rates of {001} and {010} faces of ZTS crystals and
their influence on the growth morphology, imaging was performed in two different growth
geometries. In the first experiment, the seed was oriented such that (001) face was pointing
upwards (a), whereas in the second the (001) face was pointing downwards (b). Shadowgraph
images were recorded in a time lapsed manner for the entire duration of the two experiments.
The images were used to calculate precisely the growth rates of {001} and {010} faces. Optical
imaging helped in real-time and in-situ observation of the evolution of the growth morphology as
a function of the growth geometry and supersaturation. Images adapted from Optical imaging of
the growth kinetics and polar morphology of zinc tris (thiourea) sulphate (ZTS) single crystals, S.
Dinakaran, Sunil Verma, S. Jerome Das, S. Kar and K.S. Bartwal, Cryst. Res. Technol. 45
(2010) 233-238.

Page 52 of 77
 20 20 20
18 18 18
16 16 16
Y-axis (mm)

14 14 14

Y-axis (mm)
Y - axis (mm)

12 12 12
10 10 10
8 8 8
6 6 6
4 4
4
2 2
2
0 0
0
348.848 348.849 348.850 348.851 348.852 348.853 348.854 347.855 347.860 347.865 347.870 347.875 347.880 347.885
348.520348.522348.524348.526348.528348.530348.532348.534348.536
Concentration (gm/1000 gm) Concentration (gm / 100 gm) Concentration (gm/ 100 gm)

Figure 15: First row - Infinite fringe interferograms of concentration gradient inside the growth
chamber at three stages of growth under conditions of free convection. Second row - the
computed concentration distribution corresponding to the images in the top row are shown. As
the solute-depleted solution accumulates in the upper region of the crystallizer, the concentration
diminishes locally, producing a density stratified, gravitationally stable configuration. The Mach-
Zehnder interferometer proves to be a very sensitive diagnostic to detect and quantify solution
stratification during the crystallization process, a stage that is detrimental for growth of good
quality crystals. Images adapted from Imaging transport phenomena and surface
micromorphology in crystal growth using optical techniques, Sunil Verma and K. Muralidhar,
Nat. Acad. Sci. Letts. 33 (2010) 107-121.

Page 53 of 77
Figure 16: First row - Infinite fringe interferograms of a crystal growing under free convection
conditions in the top hanging geometry. . The free convection plume and the diffusion boundary
layer are clearly visible. Second row - Wedge fringe interferograms of a crystal growing under
free convection conditions in the top hanging geometry. The diffusion boundary layer is not
visible as clearly as is in case of infinite fringe interferogram. Images adapted from Convection,
concentration, and surface features analysis during crystal growth from solution using optical
diagnostics, Sunil Verma and K. Muralidhar, in Recent Research Developments in Crystal
Growth, Vol. 5, Transworld Research Network, India, 2009, pp. 141-314.

Page 54 of 77
Figure 17: First row - Infinite fringe interferograms of a crystal growing under forced convection
conditions in the top hanging geometry. Second row - Wedge fringe interferograms of a crystal
growing under forced convection conditions in the top hanging geometry. No buoyancy driven
convection plumes are visible. Minor fringe distortion is visible near to the crystal-solution
interface. Images adapted from Convection, concentration, and surface features analysis during
crystal growth from solution using optical diagnostics, Sunil Verma and K. Muralidhar in Recent
Research Developments in Crystal Growth, Vol. 5, Transworld Research Network, India, 2009,
pp. 141-314.

Page 55 of 77
Figure 18: First row - Infinite fringe interferograms of the crystal growing on a platform under
free convection conditions. The concentration gradient is clearly visible in the second image.
Second row - Wedge fringe interferograms of crystal growing under free convection conditions
in a platform geometry. The free convection plume and the diffusion boundary layers are visible
in the infinite as well as wedge fringe interferograms. Images adapted from: Convection,
concentration, and surface features analysis during crystal growth from solution using optical
diagnostics, Sunil Verma and K. Muralidhar, Recent Research Developments in Crystal Growth,
Vol. 5, Transworld Research Network, India, 2009, pp. 141-314.

Page 56 of 77
 0 min 5 min 10 min

15 min 20 min 30 min

40 min 50 min 60 min

Figure 19: Heat conduction in a differentially heated layer of water with the top surface
maintained at 23oC and the lower surface at 18oC. The thermal field is images using a Mach-
Zehnder interferometer in the infinite fringe setting. Images adapted from [Link]. dissertation
of Vishnu Singh, Indian Institute of Technology, Kanpur (2009).

Page 57 of 77
 0 min 2 min 8 min 12 min

20 min 30 min 40 min 50 min

1 hr 1 hr 15min 1 hr 30 min 2 hr

Figure 20: A layer of sugar applied at the base slowly dissolves in water. The concentration of
sugar in water changes with time. It is accompanied by the appearance of interferometric
fringes. Images adapted from doctoral dissertation of Susheel Bhandari, Indian Institute of
Technology, Kanpur (2009).

Page 58 of 77
Figure 21: Natural convection in a rectangular cavity of square cross-section. The fluid medium
in the cavity is air and convection is driven by the temperature differential between a hot lower
wall and a cold upper wall. The side walls are insulated. The symbol Ra is Rayleigh number and
is a measure of the strength of natural convection in the cavity. The first column carries
interferometric fringes. The second and third columns contain a re-distribution of light intensity
as obtained from schlieren and shadowgraph techniques. Images adapted from Optical imaging
and control of convection around a crystal growing from its aqueous solution, K. Muralidhar,
Atul Srivastava and P.K. Panigrahi, in New Developments in Crystal Growth Research, Nova
Publishers, 2005, pp. 1-89.

Page 59 of 77
 0o 45o

90o 135o
Figure 22: Buoyancy driven convection in an octagonal cavity half-filled with 50 cSt silicone oil,
the rest being air. The lower surface is heated while the top surface is cooled. The figure shows
interferometric projections when the thermal field is viewed in various directions – 0, 45, 90, and
135o. Images adapted from Interferometric study of convection in superposed gas-liquid layers,
Sunil Punjabi, Doctoral dissertation, Indian Institute of Technology, Kanpur (2002).

Page 60 of 77
Figure 23: Long-time interferograms formed in an octagonal cavity containing silicone oil (50
cSt) floating over water as recorded from four different angles; Overall temperature difference is
1.8 K, the top plate being cooler than the one at the base. Of special interest are the energy and
momentum transfer at the oil-water interface. Images adapted from Buoyancy-driven convection
in two superposed fluid layers in an octagonal cavity, Sunil Punjabi, K. Muralidhar and P.K.
Panigrahi, International Journal of Thermal Sciences, 43 (2004) 849-864.

Page 61 of 77
Figure 24: Buoyancy-driven convection in a differentially heated cavity. The roll patterns of the
flow field are seen as fringe displacement in the interferograms. The flow pattern is three
dimensional and sketched at the top of the figure. The corresponding interferograms seen from
various viewing directions are shown below. Images adapted from Experimental study of
Rayleigh-Benard convection using interferometric tomography, Debasish Mishra, Doctoral
dissertation, Indian Institute of Technology, Kanpur (1998).

Page 62 of 77
Figure 25: Interferograms recorded in an eccentric annulus with the gap filled with air. The inner
cylinder is heated while the outer cylinder is cooled. The images show a time sequence of
interferograms from the initial infinite fringe setting all the way till steady state is reached.
Images adapted from Interferometric study of steady and unsteady convection in cylindrical and
eccentric annuli, Manoj Ranjan, [Link]. dissertation, Indian Institute of Technology, Kanpur,
(2005).

Page 63 of 77
 (a) (b)

(c) (d)

(e) (f)

Figure 26: Interferograms recorded during heat transfer in an eccentric annulus when the gap is
filed with 50 cSt silicone oil. The applied temperature differences increase from (a) to (e) from
0.2 to 2 K. When compared to air, the principal differences observed in the fringe patterns are an
increase in the number of fringes, strong refraction errors near the inner cylinder, symmetric
thermal fields about the position of the inner cylinder and weak convection effects. Images
adapted from Interferometric study of steady and unsteady convection in cylindrical and
eccentric annuli, Manoj Ranjan, [Link]. dissertation, Indian Institute of Technology, Kanpur,
(2005).

Page 64 of 77
 (a)

(b)

(c)

(d)

Figure 27: Instantaneous schlieren images (a-e) for flow past a circular cylinder separated by a
time interval of one eighth of the time period of vortex shedding at Re=110. (a) Ri=0.052, (b)
Ri=0.104, (c) Ri=0.140, and (d) Ri=0.157. For Ri ≥0.157, images show steady patterns. Images
adapted from Effect of buoyancy on the wakes of circular and square cylinders - a schlieren-
interferometric study, S.K. Singh, P.K. Panigrahi and K. Muralidhar, Experiments in Fluids, 43
(2007) 101-123.

Page 65 of 77
3. CONCLUSION
The present chapter provides an introduction to the subject of interferometry as a method of
measurement of temperature in fluids and solute concentration in liquids. As a rule, a coherent
source of light such as a laser is required and the medium is transparent and non-scattering. The
record originates in refractive index variations and is an image of the density field in the form of
fringes. The data contains information related to the line-of-sight averaged density and hence
temperature as well as concentration. Specific advantages of the interferometer can be listed as
follows:
i. The measurement takes place for the entire cross-section and as opposed to point
measurements, it classifies as a field scale technique.
ii. Data is localized at the fringes. Thus, temperature and concentration are available at all
points along a fringe. In practice, the resolution is limited by the number of pixels
available in the camera.
iii. The measurement is instantaneous without any time lag. In experiments, temporal
resolution is limited by that of the camera.
iv. An image, namely the interferogram is a two dimensional projection of the material
density distribution. However, information in three dimensions can be recovered by
reconfiguring the interferometer for holography or analytically using the principles of
tomography.
v. A variety of optical configurations are available. These can be used to adjust the
sensitivity of the measurement. Alternatives such as isotherms, profiles, and gradients can
be explored for the image content.
vi. The breadth of applications reported in this chapter shows that interferometry can play an
important role in industrial processes and applications.

Interferometry may be limited by the need for frequent alignment and a controlled atmosphere.
Ideally, the optics should be computer controlled. A compensation chamber may be required for
sensitive measurements. The presence of a solid phase such as particulates will result in
scattering and diminish the utility of interferometry. These subjects form the focus of future
research.

Page 66 of 77
ACKNOWLEDGEMENT
SV is indebted to Dr. P.J. Shlichta for the encouragement, valuable advice and unflagging
support. His approval to use some of the material from our earlier joint work [Verma, 2008A] for
the this chapter is gratefully acknowledged. SV owes a great deal to his teacher, Dr. Prabhat
Munshi, for being a friend, philosopher and guide, all in one. He is also grateful to Dr. P.K.
Gupta for the support and advice. KMD gratefully acknowledges close collaboration with his
fellow colleagues Dr(s). P.K. Panigrahi and Prabhat Munshi. He also acknowledges the
contributions of his past graduate students, namely Debasish Mishra, Sunil Punjabi, Atul
Srivastava, S.K. Singh, and Manoj Ranjan. KMD and YMJ acknowledge assistance provided by
their current students, Susheel Bhandari and Vishnu Singh.

Page 67 of 77
NOMENCLATURE

Tsupersat Supersaturated (or growth) temperature

Tsat Saturation temperature
ΔT = Tsupersat – Tsat Supercooling
Csupersat Supersaturated concentration
Csat Saturated concentration
ΔC = Csupersat – Csat Driving force for crystal growth (concentration is expressed in
units of grams of solute per 1000 gm of solvent)
Csup ersat − Csat
σ= Relative supersaturation
Csat
NLO Non-linear optics
KDP Potassium Dihydrogen Phosphate
Is Intensity on screen in the presence of disturbance
Io Intensity on screen in the absence of disturbance
n Refractive index of the solution
λ Wavelength of the laser light
t Thickness of film
θ Angle of incidence
r Radius of the cluster
θ sol ,θ qtz , θ air Angle of incidence of the light at the quartz window, the angle of

refraction of the light ray into the quartz window, the angle of
refraction of the light ray in the air, respectively
nsol ,nqtz , nair Refractive index of the KDP solution, the refractive index of the

quartz window and the refractive index of the quartz window

∂ρ (Co − Ci ) 3
Gr = ρ gL Grashof number
∂C µ2
ω L2
Re = Reynolds number
ν
µ
Sc = Schmidt number
ρD

Page 68 of 77
ρ Density of the solution at a certain temperature in kg/m3

∂ρ ∂C Gradient of density with concentration

Co Bulk solute concentration in mole/m3

Ci Interface concentration near to the crystal

g Acceleration due to gravity in m/s2

L Characteristic length in m
D Diffusion coefficient of solute in the solution in m2/s
k Gladstone-Dale constant
∆CE Concentration change per fringe shift

Page 69 of 77
REFERENCES
[Bedarida, 1977] F. Bedarida, L. Zefiro and C. Pontiggia, Holographic control of diffusion
coefficients in water solutions: Crystal growth from solutions, in: "Applications of Holography
and Optical Data Processing", E. Marom and A. A. Friesem (Eds), (Pergamon Press, Oxford,
1977), pp. 259-265.
[Bedarida, 1992] F. Bedarida, G.A. Dall'Aglio, L. Gatti and F. Solitro, The sensitivity of an
optical fiber holographic interferometer, Proc. 8th Euro. Symp. on Materials and Fluid Sciences
in Microgravity, Brussels, Belgium, ESA SP-333 (August 1992), 321-323.
[Bhandari, 2009] Susheel Bhandari, Doctoral dissertation, Indian Institute of Technology,
Kanpur, India (2009).
[Bruning, 1974] J.H. Bruning, D.R. Herriott, J.E. Gallagher, D.P. Rosenfeld, A.D. White and
D.J. Brangaccio, Digital wavefront measuring interferometer for testing optical surfaces and
lenses, Appl. Opt., 13 (11) (1974) 2693-2703.
[Bunn, 1949] C.W. Bunn, Crystal growth from solution. II. Concentration gradients and the rates
of growth of crystals, Discuss. Faraday Society, 5 (1949) 132-144.
[Censor, 1983] Y. Censor, Finite series-expansion reconstruction methods, Proc. IEEE, 71 (3)
(1983) 409-419.
[Chen, 1977] P.S. Chen, Convection irregularities during solution growth and relation to crystal-
defect formation, Ph.D. dissertation, Univ. of Southern California, Los Angeles, CA, USA,
(1977), 213 pp.
[Chen, 1979A] P.S. Chen, W.R. Wilcox and P.J. Shlichta, Free convection about a rectangular
prismatic crystal growing from a solution, Int. J. Heat Mass Transfer, 22 (1979) 1669-1679.
[Chen, 1979B] P.S. Chen, P.J. Shlichta, W.R. Wilcox and R.A. Lefever, Convection phenomena
during the growth of sodium chlorate crystals from solution, J. Cryst. Growth, 47 (1979) 43-60.
[Chen, 2002] W.C. Chen, D.D. Liu, W.Y. Ma, A.Y. Xie and J. Fang, The determination of solute
distribution during growth and dissolution of NaClO3 crystals: the growth of large crystals, J.
Cryst. Growth, 236 (2002) 413-419.
[Cloud, 1995] Gary Cloud (with contributions from K. Creath), Phase shifting to improve
interferometry, Chapter 22, in: "Optical Methods of Engineering Analysis", Gary Cloud,
(Cambridge University Press, 1995), pp. 477-491.
[Cole, 1995] T. Cole, A. Kathman, S. Koszelak, and A. McPherson, Determination of local
refractive index for protein and virus crystals in solution by Mach-Zehnder interferometry,
Analytical Biochem., 231 (1995) 92-98.
[Collier, 1971] R.J. Collier, C.B. Burckhardt and L.H. Lin, "Optical Holography", (Academic
Press, New York, 1971), 605 pp.
[Creath, 1986] K. Creath, Comparison of phase-measurement algorithms, SPIE 680 (1986) 19-
28.
[Creath, 1987] K. Creath, WYKO systems for optical metrology, SPIE 816 (1987) 111-126.

Page 70 of 77
[Creath, 1988] K. Creath, Phase-measurement interferometric techniques, Chapter V, in:
"Progress in Optics", Vol. XXVI, E. Wolf (Ed.), (Elsevier Science Publishers B.V., Netherlands,
1988), pp. 349-393.
[Dainty, 1984] J.C. Dainty (Ed.), "Laser Speckle and Related Phenomena", 2nd ed., (Springer-
Verlag, Berlin, 1984), 342 pp.
[Dinakaran, 2010] S. Dinakaran, Sunil Verma, S. Jerome Das, S. Kar and K.S. Bartwal, Optical
imaging of the growth kinetics and polar morphology of zinc tris (thiourea) sulphate (ZTS)
single crystals, Cryst. Res. Technol. 45 (3) (2010) 233-238
[Duan, 2001] L. Duan and J. Z. Shu, The convection during NaClO3 crystal growth observed by
the phase shift interferometer, J. Cryst. Growth, 223 (2001) 181-188.
[Dubois, 1999] F. Dubois, L. Joannes, O. Dupont, J.L. Dewndel and J.C. Legros, An integrated
optical set-up for fluid-physics experiments under microgravity conditions, Meas. Sci. Technol.,
10 (1999) 934-945.
[Ducruix, 1992] A. Ducruix and R. Giege, Crystallization of Nucleic Acids and Proteins,
(Oxford Univ. Press, 1992), 331 pp.
[Dupont, 1995] O. Dupont, J.L. Dewandel and J.C. Legros, Use of electronic speckle pattern
interferometry for temperature distribution measurements through liquids, Opt. Letts., 20 (1995)
1824-1826.
[Ecker, 1987] A. Ecker, Solidification front dynamics examined by holographic interferometric
measurement of temperature and concentration using transparent model systems, Final Report,
NASA (March, 1987).
[Ecker, 1988] A. Ecker, Two-wavelength holographic measurement of temperature and
concentration during alloy solidification, J. Thermophys. Heat Transfer, 2(3) (1988) 193-196.
[El-Wakil, 1964] M.M. El-Wakil and C.L. Jaeck, A two-wavelength interferometer for the study
of heat and mass transfer, J. Heat Transfer, 79 (1964) 464-466.
[Gabor, 1948] D. Gabor, A new microscopic principle, Nature, 161 (1948) 777.
[Gabor, 1949] D. Gabor, Microscopy by reconstructed wavefronts, Proc. Soc. Roy., A197 (1949)
457-484.
[Gatti, 1989] L. Gatti, F. Solitro, F. Bedarida, P. Boccacci, G.A. Dall'Aglio and L. Zefiro, Three-
dimensional measurements of concentration fields in crystal growth by multidirectional
holographic interferometry, SPIE 1162 (1989) 126-131.
[Goldstein, 1996] R.J. Goldstein and T.H. Kuehn, Optical Systems for Flow Measurement:
Shadowgraph, Schlieren, and Intreferometric Techniques, in: "Fluid Mechanics Measurements",
R. J. Goldstein (Ed.), (Taylor & Francis, New York, 1996), pp. 451-508.
[Greivenkamp, 1992] J.E. Greivenkamp and J.H. Bruning, Phase shifting interferometry, Chapter
14, in: "Optical Shop Testing", 2nd Ed., Daniel Malacara (Ed.), (John Wiley & Sons, Inc., New
York, 1992), pp. 501-598.
[Gull, 1986] S.F. Gull and T.J. Newton, Maximum entropy methods, Appl. Opt., 25 (1986) 156-
160.

Page 71 of 77
[Hariharan, 1985] P. Hariharan, "Optical Interferometry", (Academic Press, Sydney, 1985), 303
pp.
[Hariharan, 1986] P. Hariharan, "Optical Holography: Principles, Techniques and Applications",
(Cambridge University Press, Cambridge, 1986), 319 pp.
[Hariharan, 2002] P. Hariharan, "Basics of Holography", (Cambridge University Press,
Cambridge, 2002), 161 pp.
[Heffinger, 1966] L.O. Heffinger, R.F. Wuerker and R.E. Brooks, Holographic interferometry, J.
Appl. Phys., 37 (1966) 642-649.
[Herman, 1980] G.T. Herman, "Image Reconstruction from Projections: The Fundamentals of
Computerized Tomography", (Academic Press, New York, 1980), 316 pp.
[Jones, 1989] R. Jones and C. Wykes, “Holographic and Speckle Interferometry”, 2nd Ed.,
(Cambridge University Press, 1989), p. 368.
[Kang, 2001] Q. Kang, L. Duan and W.R. Hu, Mass transfer process during the NaClO3 crystal
growth process, Int. J. Heat and Mass Transfer, 44 (2001) 3213-3222.
[Kim, 1998A] Y.K. Kim, B.R. Reddy and R.B. Lal, Laser and Mach-Zehnder interferometer for
in-situ monitoring of crystal growth and concentration variation, SPIE 3479 (1998) 172-180.
[Kumar, 2008] N.V.N. Ravi Kumar, K. Muralidhar, and Y.M. Joshi, On refractive index of
ageing suspensions of laponite, Applied Clay Science, 42 (2008) 326-330.
[Lenski, 1991] H. Lenski and M. Braun, Laser beam deflection: a method to investigate
convection in vapor growth experiments, SPIE 1557, "Crystal Growth in Space and Related
Optical Diagnostics", James D. Trolinger, Ravindra B. Lal (Eds.), (1991) 124-131.
[Lewitt, 1983] R.M. Lewitt, Reconstruction algorithms: Transform methods, Proc. IEEE, 71 (3)
(1983) 390-408.
[Lokberg, 1980] O.J. Lokberg, Electronic speckle pattern interferometry, Phys. Technol., 11
(1980) 16-22.
[Maruyama, 1999] S. Maruyama, T. Shibata and K. Tsukamoto, Measurement of diffusion fields
of solutions using real-time phase-shift interferometer and rapid heat-transfer control system,
Exp. Therm. Fluid Sci., 19 (1999) 34-48.
[Maruyama, 2002] S. Maruyama, K. Ohno, A. Komiya and S. Sakai, Description of the adhesive
crystal growth under normal and micro-gravity conditions employing experimental and
numerical approaches, J. Cryst. Growth, 245 (2002) 278-288.
[Mayinger, 1993] F. Mayinger, Image-forming optical techniques in heat transfer: revival by
computer-aided data processing, Trans. ASME, 115 (1993) 824-834.
[Mayinger, 1994] F. Mayinger (Ed.), "Optical Measurements: Techniques and Applications",
(Springer-Verlag, Berlin, 1994), 463 pp.
[McPherson, 1982] A. McPherson, “Preparation and Ansalysis of Protein Crystals”, (Wiley, New
York, 1982), 371 pp.
[McPherson, 1999] A. McPherson, A.J. Malkin, Y.G. Kuznetsov, S. Koszelak, M. Wells, G.
Jenkins, J. Howard and G. Lawson, The effects of microgravity on protein crystallization:

Page 72 of 77
Evidence for concentration gradients around growing crystals, J. Crystal Growth, 196 (1999)
572-586.
[McPherson, 2001] A. McPherson, “Crystallization of Biological Macromolecules”, 2nd ed.,
(Cold Spring Harbor Laboratory Press, 2001) 586 pgs. (1st ed. 1998).
[Mehta, 1990] J.M. Mehta, Dual wavelength interferometric technique for simultaneous
temperature and concentration measurement in liquids, Appl. Opt., 29 (13) (1990) 1924-1932.
[Merzkirch, 1974] W.F. Merzkirch, "Flow Visualization", (Academic Press, New York, 1974),
250 pp.
[Merzkirch, 1987] W.F. Merzkirch, "Flow Visualization", 2nd Edition, (Academic Press, New
York, 1987), 260 pp.
[Mishra, 1998] Debasish Mishra, Experimental Study of Rayleigh Benard Convection using
Interferometric Tomography, Ph.D. dissertation, Indian Institute of Technology, Kanpur, India
(1998).
[Mishra, 1999] Debasish Mishra, K. Muralidhar and P. Munshi, Interferometric study of
Rayleigh-Benard convection using tomography with limited projection data, Numerical Heat
Transfer, 12 (1999) 117-136.
[Mullin, 2001] J.W. Mullin, Crystallization, 4th ed., (Butterworth-Heinemann, Oxford, 2001).
[Muralidhar, 2001] K. Muralidhar, Temperature field measurement in buoyancy-driven flows
using interferometric tomography, Ann. Rev. Heat Transf., 12 (2001) 265-375.
[Muralidhar, 2005] K. Muralidhar, Atul Srivastava and P.K. Panigrahi, Optical imaging and
control of convection around a crystal growing from its aqueous solution, in New Developments
in Crystal Growth Research ( Nova Publishers, New York, 2005) pp. 1-89.
[Murphy, 1971] C.G. Murphy and S.S. Alpert, Dependence of refractive index temperature
coefficients on the thermal expansivity of liquids, Am. J. Phys., 39 (7) (1971) 834-836.
[Nagashima, 2000] K. Nagashima and Y. Furukawa, Time development of a solute diffusion
field and morphological instability on a planar interface in the directional growth of ice crystals,
J. Cryst. Growth, 209 (2000) 167-174.
[Nakadate, 1990] S. Nakadate and I. Yamaguchi, Japanese patent describing the technique of
real-time phase shifting interferometry, Japanese Patent # H02-287107 (1990).
[Nakadate, 1995] S. Nakadate, Real-time fringe pattern processing and its applications, SPIE
2544 (1995) 74-86.
[Natterer, 2001] F. Natterer, "The Mathematics of Computerized Tomography", (Philadelphia,
PA, SIAM 2001), 222 pp.
[Onuma, 1988] K. Onuma, K. Tuskamoto and I. Sunagawa, Role of buoyancy driven convection
in aqueous soultion; A case study of Ba(NO3)2 crystal, J. Cryst. Growth, 89 (1988) 177-188.
[Onuma, 1989] K. Onuma, K. Tuskamoto and I. Sunagawa, Measurement of surface
supersaturations around a gowing K-alum crystal in aqueous solution, J. Cryst. Growth, 98 (3)
(1989) 377-383.

Page 73 of 77
[Onuma, 1993] K. Onuma, K. Tsukamoto and S. Nakadate, Application of real time phase shift
interferometer to the measurement of concentration field, J. Cryst. Growth, 129 (1993) 706-718.
[Owen, 1975] W.A. Owen et al. Advanced Application Flight Experiment: Effect of Residual
and Transient Convection on Crystallization in a Spaceflight Environment, Semiannual Report,
NASA Code. No. 638-10-00-01-00 (Jet Propulsion Laboratory, Pasadena, CA, USA, March
1975; see also later reports November 1975 and June 1976.
[Owen, 1982] R.B. Owen, Interferometry and holography in a low-gravity environment, Appl.
Opt., 21 (8) (1982) 1349-1355.
[Owen, 1986] R.B. Owen, R.L. Kroes and W.K. Witherow, Results and further experiments
using Spacelab holography, Opt. Lett., 11 (7) (1986) 407-409.
[Paschotta, 2008] Rüdiger Paschotta, Encyclopedia of Laser Physics and Technology (Wiley-
VCH, Berlin, 2008) 844 pp.
[Piano, 2000] E. Piano, G.A. Dall’Aglio, S. Crivello, R. Chittofrati and F. Puppo, New optical
techniques for crystal growth from fluids, Materials Chem. Phys., 66 (2000) 266-269.
[Piano, 2001] E. Piano, G.A. Dall’Aglio, R. Chittofrati, S. Crivello and F. Puppo, A non-
destructive interferometric technique for analysis of crystal growth and fluid dynamics, Ann.
Chim. Sci. Mat., 26 (2001) 23-28.
[Pletser, 2001] V. Pletser, O. Minster, R. Bosch, L. Potthast and J. Stapelmann, The protein
crystallisation diagnostics facility: status of the ESA programme on the fundamentals of protein
crystal growth, J. Cryst. Growth, 232 (2001) 439-449.
[Punjabi, 2002] Sunil Punjabi, Interferometric study of convection in superposed gas-liquid
layers, Doctoral dissertation, Indian Institute of Technology, Kanpur, India (2002).
[Punjabi, 2004] Sunil Punjabi, K. Muralidhar and P.K. Panigrahi, Buoyancy-driven convection
in two superposed fluid layers in an octagonal cavity, , Int. J. Therm. Sci., 43 (2004) 849-864.
[Ramachandran, 1971] G.N. Ramachandran and A.V. Lakshminarayanan, Three-dimensional
reconstruction from radiographs and electron micrographs: Application of convolutions instead
of Fourier transforms, Proc. Nat. Acad. Sci. USA, 68 (9) (1971) 2236-2240.
[Ranjan, 2005] Manoj Ranjan, Interferometric study of steady and unsteady convection in
cylindrical and eccentric annuli, [Link]. dissertation, Indian Institute of Technology, Kanpur,
India, (2005).
[Rasenat, 1989] S. Rasenat, G. Hartung, B.L. Winkler and I. Rehberg, The shadowgraph method
in convection experiments, Exp. Fluids, 7 (1989) 412-420.
[Rashkovich, 1990A] L.N. Rashkovich and [Link]. Shekunov, Hydrodynamic effects in growth of
ADP and KDP crystals in solution. I. Growth kinetics, Sov. Phys. Crystallogr., 35 (1) (1990) 96-
99.
[Rashkovich, 1990B] L.N. Rashkovich and [Link]. Shekunov, Morphology of growing vicinal
surface; Prismatic faces of ADP and KDP crystals in solutions, J. Cryst. Growth, 100 (1990)
133-144.
[Schopf, 1996] W. Schopf, J.C. Patterson and A.M.H. Brooker, Evaluation of the shadowgraph
method for the convective flow in a side heated cavity, Exp. Fluids, 21 (1996) 331-340.

Page 74 of 77
[Settles, 2001] G.S. Settles, "Schlieren and Shadowgraph Techniques", (Springer, Berlin, 2001),
376 pp.
[Shagam, 1978] R.N. Shagam and J.C. Wyant, Optical frequency shifter for heterodyne
interferometers using multiple rotating polarization retarders, Appl. Opt., 17 (1978) 3034-3035.
[Shiomi, 1980] Y. Shiomi, T. Kuroda and T. Ogawa, Thermal analysis of a growing crystal in an
aqueous solution, J. Cryst. Growth, 50 (1980) 397-403.
[Shlichta, 1985] P.J. Shlichta, Crystal growth in a spaceflight environment: final report,
Materials processing in space environment, Experiment # 770100, (September 1985), pp. 1-123,
(Jet Propulsion Laboratory, Pasadena, CA 91109).
[Singh, 2009] Vishnu Singh, [Link]. dissertation, Indian Institute of Technology, Kanpur, India
(2009).
[Singh, 2007] S.K. Singh, P.K. Panigrahi and K. Muralidhar Effect of buoyancy on the wakes of
circular and square cylinders - a schlieren-interferometric study, Experiments in Fluids, 43
(2007) 101-123.
[Smigielski, 1970] P. Smigielski and A. Hirth, Proc. 9th Internat. Congress on High-Speed
Photography, W.G. Hyzer and W.G. Chase (Eds.), (SMPTE, New York, 1970), pp. 321-326.
[Solitro, 1989] F. Solitro, L. Gatti, F. Bedarida, G.A. Dall'Aglio and L. Michetti,
Multidirectional holographic interferometer (MHOI) with fiber optics for study of crystal growth
in microgravity, SPIE 1162 (1989) 62-65.
[Srivastava, 2004] A. Srivastava, K. Muralidhar and P.K. Panigrahi, Comparison of
interferometry, schlieren and shadowgraph for visualizing convection around a KDP crystal, J.
Cryst. Growth, 267 (2004) 348-361.
[Steel, 1983] W.H. Steel, "Interferometry", 2nd ed., (Cambridge University Press, Cambridge,
1983), 320 pp.
[Trolinger, 1991] J.D. Trolinger and R.B. Lal (Eds.), "Crystal Growth in Space and Related
Optical Diagnostics", SPIE 1557 (1991) 1-296.
[Verga, 1997] A. Verga, P. Baglioni, O. Dupont, J.L. Dewandel, T. Beuselinck and J. Bouwen,
Use of electronic speckle pattern interferometers for the analysis of convective states of liquids
in weightlessness, Proc. SPIE Vol. 3172, Optical Technology in Fluid, Thermal, and Combustion
Flow III (1997) 194-210.
[Verma, 2003] Sunil Verma, S.K. Sharma, K. Muralidhar and V.K. Wadhawan, Optical imaging
techniques in crystal growth research, in: "Crystal Growth of Technologically Important
Electronic Materials", K. Byrappa, H. Klapper, T. Ohachi & R. Fornari (Eds.), (Allied
Publishers, Bangalore, 2003), pp. 616-619.
[Verma, 2005A] Sunil Verma, K. Muralidhar and V.K. Wadhawan, Convection during growth of
KDP crystals: Flow visualization and modeling, Ferroelectrics, 323 (2005) 25-37.
[Verma, 2005B] Sunil Verma, K. Muralidhar and V.K. Wadhawan, Convection and
concentration mapping during crystal growth from solution using Mach-Zehnder interferometry
and computerized tomography, Proc. 3rd Asian Conf. Crystal Growth & Crystal Technology
(CGCT-3), Oct. 16-19, 2005, Beijing, China.

Page 75 of 77
[Verma, 2005C] Sunil Verma, A. Srivastava, V. Prabhakar, K. Muralidhar and V.K. Wadhawan,
Simulation and experimental verification of solutal convection in the initial stages of crystal
growth from an aqueous solution, Ind. J. Pure & Appl. Phys., 43 (2005) 24-33.
[Verma, 2006] Sunil Verma, K. Muralidhar and V.K. Wadhawan, Determination of
concentration field around a growing crystal using shadowgraphic tomography (Chap. 14), In:
“Computerized Tomography for Scientists and Engineers”, P. Munshi (Ed.), (CRC Press, New
York, 2006) 158-174
[Verma, 2007] Sunil Verma, Convection, concentration, and surface feature analysis during
crystal growth from solution using shadowgraphy, interferometry and tomography, Ph.D.
dissertation, Indian Institute of Technology, Kanpur, India (2007).
[Verma, 2008A] Sunil Verma and P.J. Shlichta, Imaging techniques for mapping of solution
parameters, growth rate, and surface features during the growth of crystals from solution, Prog.
Cryst. Growth & Charact. Materials, 54 (2008) 1-120
[Verma, 2008B] Sunil Verma and K. Muralidhar, Three-dimensional reconstruction of
convective features during crystal growth from solution using computerized tomography, In:
“CT2008: Tomography Confluence”, P. Munshi (Ed.), American Institute of Physics CP 1050
(2008) 103-114
[Verma, 2009] Sunil Verma and K. Muralidhar, Convection, concentration, and surface features
analysis during crystal growth from solution using optical diagnostics (Chap. 5), in: Recent
Research Developments in Crystal Growth, Vol. 5 (2009) 141-314 (Transworld Research
Network, India)
[Vest, 1979] C.M. Vest, "Holographic Interferometry", (John Wiley, New York, 1979), 465 pp.
[Vikram, 1990] C.S. Vikram, H.J. Caulfield, G.L. Workman, J.D. Trolinger, C.P. Wood, R.L.
Clark, A.D. Kathman and R.M. Ruggiero, Two-color holographic concept (T-CHI), Final Tech.
Rep., Contract No. NAS8-38078, NASA George C. Marshall Space Flight Centre (April 1990).
[Vikram, 1991] C.S. Vikram, W.K. Witherow and J.D. Trolinger, Refractive properties of TGS
aqueous solution for two-color interferometry, SPIE 1557, "Crystal Growth in Space and Related
Optical Diagnostics", J.D. Trolinger and R.B. Lal (Eds.), (1991) 197-201.
[Vikram, 1992A] C.S. Vikram and W.K. Witherow, Critical needs of fringe-order accuracies in
two-color holographic interferometry, Exp. Mech., (March 1992) 74-77.
[Vikram, 1992B] C.S. Vikram, W.K. Witherow and J.D. Trolinger, Determination of refractive
properties of fluids for dual-wavelength interferometry, Appl. Opt., 31 (34) (1992) 7249-7252.
[Witherow, 1987] W.K. Witherow, Reconstruction techniques of holograms from Spacelab-3,
Appl. Opt., 26 (12) (1987) 2465-2473.
[Witherow, 1994] W.K. Witherow, J.R. Rogers, B.R. Facemire, S.D. Armstrong, J.D. Trolinger,
D. Weber and C.S. Vikram, Methods to detect and measure gradients in fluids and materials
processing, Proc. 6th International Symp. on Experimental Methods for Microgravity Materials
Science, San Francisco (USA), Feb. 27- March 3, 1994, R.A. Schiffman and B. Andrews (Eds.),
The Minerals, Metals and Materials Society, 33-37.
[Wyant, 1975] J.C. Wyant, Use of an AC heterodyne lateral shear interferometer with real-time
wavefront correction systems, Appl. Opt., 14 (1975) 2622-26.

Page 76 of 77
[Wyant, 1978] J.C. Wyant and R.N. Shagam, Use of electronic phase measurement techniques in
optical testing, Proc. 11th Congr. of the International Commission for Optics, Madrid, 10-17
September 1978, J. Bescos, A. Hidalgo, L. Plaza and J. Santamaria (eds.) (Sociedad Espanola de
Optica, Madrid) pp. 659.
[Wyant, 1982] J.C. Waynt, Interferometric optical metrology: Basic systems and principles,
Laser Focus, (May 1982) 65-71.
[Wyant, 1985] J.C. Wyant and K. Creath, Laser Focus, (November 1985), pp. 118.
[Xiling, 1992] Yu Xiling and Yue Xuefeng, Holographic studies of diffusivities of KDP and
DKDP in the solutions, Cryst. Res. Technol., 27 (1992) 825-830.
[Xiling, 1996] Yu Xiling, Sun Yi, Jiang Huizhu and Zhnag Shujun, Growth kinetics of the
metastable tetragonal phase of potassium dideutrium phosphate (DKDP) crystals, J. Cryst.
Growth, 166 (1996) 195-200.
[Yin, 2001] D. Yin, Y. Inatomi and K. Kuribayashi, Study of lysozyme crystal growth under a
strong magnetic field using a Mach-Zehnder interferometer, J. Cryst. Growth, 226 (2001) 534-
542.
[Yin, 2003] D.C. Yin, Y. Inatomi, N.I. Wakayama and W.D. Huang, Measurement of
temperature and concentration dependences of refractive index of hen-egg-white lysozyme
solution, Cryst. Res. Technol., 38 (9) (2003) 785-792.

Page 77 of 77