Methods for the separation of stable isotopes

Karanam L. Ramakumar
India

(This lecture notes was initially meant for the participants of Workshop on
“Radiochemistry and Application of radioisotopes” regularly organized by Indian
Association of Nuclear Chemists and Allied Scientists (IANCAS), a non-profit
organization based in Mumbai, India. Major portions of this document were
included in the IANCAS lecture notes brought out in 1992 and subsequently in a
book “Principles of Radiochemistry” published by IANCAS in 1993-94. This enlarged
document is uploaded for interested readers. Some figures were downloaded from
Internet.)

Introduction
Although the isotopes of an element have very similar chemical properties, they behave
as completely different substances in nuclear reactions. In addition, many of the stable
isotopes find wide spread applications in chemical, industrial, agricultural and clinical research
to elucidate and understand reaction pathways, mechanisms and kinetics, effect of trace
elements on physico-chemical properties, up-take and plant metabolism studies and the
behaviour of trace elements from toxicological and human metabolism point of view
respectively. Table 1 lists separated isotopes that are being produced on a significant
industrial scale for various applications.

Table 1 Separated isotopes and their uses

Isotope Abundance in nature (at%) Use

235
U 0.7205 Fuel for nuclear fission reactors
2
H 0.015 D2 moderator for natural uranium reactors
Fuel for thermonuclear reactors
6
Li 7.5 Source for tritium
Fuel for thermonuclear reactors
7
Li 92.5 As LiOH, water conditioner for water cooled reactors
As metal, possible high temperature reactor coolant
10
B 19.61 Neutron absorber in control rods
13
C 1.107
15 Isotope tracer in living systems
N 0.366
17 Nuclear magnetic resonance studies of molecular
O 0.037
18 structure
O 0.204
57
F 2.20 Mossbauer studies

In addition to these, separated isotopes of practically all natural elements are being
produced for applications in research. Very small differences that are manifested in the
Physico-chemical properties of isotopes of element, because of different masses, such as the
boiling points, diffusion coefficients, rate of electrolysis, and difference in optical properties are
exploited to achieve their separation. Almost all the methods used for separation in
conventional industries can be used for separation of isotopes also with varying degree of
success. The methods can be broadly classified into physical and chemical methods.
Physical methods are those in which difference in physical properties e.g. inertia,
density, electrical properties etc.
Chemical methods are those in which difference in chemical properties e.g. chemical
potential is harnessed
Table 2 lists some of the processes used to separate isotopes.

Table 2 Different methods employed in isotope separation

Method Applied to
235
Electromagnetic U, all others
Electrolysis D, Li
Distillation D, 10B, 13C, 15N, 18O
Chemical exchange D, Li, 10B, 13C, 15N, 18O
Ion migration Li
235
Gaseous diffusion U
18
Thermal diffusion O, 15N, inert gas isotopes
235
Gas centrifuge U
235
Aerodynamic U
235
Photochemical U, other isotopes
Ion exchange Lighter isotopes, 235U

Chemical methods are more suitable for light isotopes and are extensively used for
separation of hydrogen isotopes e.g. production of heavy water. Physical methods are in
general more suitable for separation of heavy isotopes e.g. uranium enrichment.

Separation Factor in a Typical Enrichment Process

Irrespective of the process employed, the extent of separation of isotopes is given by
“separation factor". Two types of separation factor are generally in vogue; (i) separation factor
(α) and (ii) enrichment factor (ε). The two separation factors are a measure of the efficiency
and capability of the particular process in its basic operation to achieve the separation of
isotopes. Separation factors are different for different elements for the same process and also
different for same elements in different processes. The sample material, which is fed into an
enrichment plant, is known as feed. After enrichment, a portion of the sample material will be
obtained as the enriched fraction in the desired isotope. This fraction is known as product.
Correspondingly there will be another fraction known as tails, which are depleted in the
desired isotope. The two separation factors are given as follows:
y /(1 - y) y /(1 - y)
α= ε=
z /(1 - z) x /(1 - x)
where z is the atom % abundance of the desired isotope in the feed. (1-z) is the
corresponding quantity of the other isotope in the feed. y and (1-y) refer to the corresponding
quantities in the product and x and (1-x) are defined for tails. These separation factors are
derived from different mass dependent physico-chemical properties of the isotopes of the
elements based on which the separation process is selected. Table 3 gives representative
separation factors and the physico-chemical property for different isotope separation
processes.

Table 3 Separation factors in typical enrichment processes

Process Property Separation factors for

14
H2 – HD N – 15N 235
UF6 – 238UF6
Distillation R 1.5 1.033 1.00002
Mono-thermal K 3.6 1.055 1.0016
chemical
exchange
Gaseous 1.225 1.017 1.00429
(m2 /m1 )
diffusion
Gas centrifuge (m2-m1)v2 1.056 1.056 1.162
e 2RT
R = relative volatility, K = equilibrium constant

A value of α close to unity indicates that the separation is difficult, a value far from
unity, easier. For gaseous diffusion of UF6 for example, α is so close to unity (I.00429) that
the processes must be repeated many times for useful degree of separation. This is achieved
by employing stages and cascades.

Separating Unit, Stage and Cascade

The smallest element of an isotope separation plant that effects some separation of the
process material is called the separating unit. Examples of a single separating unit are one gas
centrifuge, or one electrolytic cell etc. No single separating unit can enrich uranium from its
natural concentration of about 0.7 percent uranium-235 to the 3 to 5 percent required to fuel
a nuclear reactor. In addition, the throughput of a unit is very small. To multiply the effects of
the enrichment of one unit and to achieve adequate throughput, large numbers of units are
interconnected in parallel. A group of parallely connected separating units, all fed with material
of same composition and producing partially separated product streams of the same
composition, is known as a stage. A series-connected group or stages is known as a cascade.
Examples of a cascade are a complete distillation column, or a battery of solvent extraction
mixed-settlers, or series of electrolytic cells etc. A cascade that has the same number of units
in all stages or in a group is known as squared-off cascade. A cascade in which number of
units in each stage decreases as the product and waste ends of the cascade are approached is
called a tapered cascade. A single multi-plate distillation column is an example of a squared-
off cascade; a gaseous diffusion plant for uranium separation is an example of a tapered
cascade. The pattern of that connection is determined by the properties of the individual
separating units and the required quantity and concentration of the final product. An
enrichment plant usually holds thousands of cascades.
 The cascade theory of isotope separation was developed by P.A.M.Dirac, and R.Peierls
in England and by K.Cohen and I.Kaplan in the United States. The cascade theory is an
involved one and has been extensively covered in the references (1) and (2). Cascades are
generally of two types.
Simple cascade: a cascade in which no attempt is made to reprocess the partially depleted
waste streams leaving each stage is called a simple cascade. In the simple cascade it is
impossible to obtain high recovery of desired component because of losses in the waste
streams leaving every stage, the recovery falls rapidly as the over all enrichment factor
desired is increasing. A simple cascade has only enrichment section.
Feed
Waste

Waste

Waste

Product

Fig.1 An example of simple cascade

Counter current or recycle cascade: a cascade in which the feed for each stage consists of
heads from the next lower stage and wastes from the next higher stage is called a counter
current or recycle cascade. This is by far the most commonly employed cascade type. In
general as (α-l) < < < 1, in most of the cases, these are also known as close-separation
cascades. A recycling cascade has two sections: the enriching section, consisting of the stages
above the point at which the feed enters the cascade and the stripping section below the feed
point. The stripping section increases the recovery of the material, whereas the enriching
section produces material of increased concentration. In a symmetric counter-current cascade,
the waste stream is recycled back to the immediately preceding stage. In an asymmetric
counter-current cascade, the waste is recycled more than one stage back. A common
asymmetric cascade feeds the product stream into the next stage and sends the tails stream
back two stages back. This is called a two-up, one-down cascade. Asymmetric cascades were
used in aerodynamic nozzle separation processes. While asymmetric cascades are rarely used
for gas centrifuges, some argue that an asymmetric cascade can increase centrifuge cascade
performance by about ten percent relative to a symmetric arrangement.
Feed
Waste

Product
Fig.2 An example of tapered counter-current symmetric cascade

An ideal cascade, a type of tapered cascade, is the most efficient arrangement of

separating units. It minimizes the total inter-stage flow rate and as a result uses the least
number of units to achieve the required separation by maximizing the separative work per
separating element. An ideal cascade has two important characteristics: it has the same
separation factor for each stage and streams of different levels of enrichment are never
mixed. The first condition makes the use of identical units possible throughout the cascade.
The no mixing condition lowers energy consumption because no separative power is wasted by
mixing flows of different concentrations. To achieve this, the cuts of the stages are adjusted
so the waste flow passed down to a stage has exactly the same isotopic concentration as the
product flow being passed up from the preceding stage. The ideal cascade has the shortest
equilibrium time and the smallest inventory.
An ideal cascade is never achieved in practice. Not every machine will perform exactly
the same so the no-mixing rule is broken even within a stage. Theoretically, the cut must be
varied for each stage to keep different concentrations from mixing. For an ideal cascade, the
cut fluctuates very slightly around 0.5 but, in practice, these small fine adjustments in flow
rates cannot be achieved. Moreover, the theoretically optimal flow rate for every stage may
correspond to a non-integer number of machines because units are optimized for a particular
cut and throughput. Obviously, it is not possible to operate a fraction of unit, so separating
units will either have to work at non-optimum conditions to balance flow rates or there will be
mixing of different isotope concentrations and, therefore, wasted separative work. In practice,
engineering tradeoffs among flow rate, cut, and separation factor are made to bring a real
cascade as close as possible to ideal performance.

Feed Waste

Product

Fig.3 An example of squared-off cascade

A square cascade has the same flow rates in all stages and therefore the same number
of machines per stage. Square cascades are rarely used because they are not very efficient;
constant flow rate results in constant cut and mixing of concentrations and therefore loss of
separative work.
The overall performance characteristics of enrichment plant are expressed in terms of
two factors: Separation capacity and separative work. The separation capacity is a measure of
the rate at which the cascade performs separation. In many isotope separation plants the
initial cost of plant is proportional to the separative capacity the plant. Separative capacity has
the dimensions of the flow rate (Kg of material/day). It is also useful to have a measure of the
amount of separation performed by a cascade. This measure is provided by the separative
work units (SWU). In an enrichment plant, the annual operating costs are proportional to the
amount of separative work done per year. Separative work unit, S has the dimensions of the
235
amount of material E. More number of SWUs is required to produce 1 Kg. of 90% UF6 as
235
compared to 1 Kg. of 3% UF6. Accordingly cost of enriched uranium increases with the total
235
number of SWUs and enrichment value. As an example, to get 1 Kg of 90 % enriched UF6
 235
we need 178 Kg of natural uranium in UF6 feed. To get 1Kg of 3% enriched UF6 we need 5.6
Kg of natural uranium in UF6 feed.
Concept of Separative Power, Separative capacity and Separative Work Unit

In conventional industrial separations, where the level of separation is almost

100% (or some standard level), throughput parameter is sufficient to indicate
the capacity of the separating plant e.g. a heavy water production plant, where
the grade is fixed for reactor use. However it may be observed that unlike in
heavy water production, varied degree of separation of isotopes is required in
case of uranium. These varying levels of enrichment required for different
applications ranging from 3% for LWRs to 90% & above for weapon grade is
typical of uranium isotope separation. So when the enrichment levels are
different in different plants then a question arises as to how to compare the
capacities of two different plants. Because only throughput may not be
sufficient to gauge the size of an enrichment plant, particularly when
enrichment levels at which the plants are operating are different.
Therefore an additional parameter, which would be a combined function of
quality & quantity of separation performed by a separating element or a plant
is defined. At the same time it should be independent of the level of
concentration of feed material. This additional parameter is called separative
power or separative capacity particularly for uranium enrichment.
The separative power is defined as a change in the Value effected by a
separating element, i.e. the increase in the value of output over the value of
input. A quantity called value function is defined as a function of the
concentration, x, of the desired isotope by the relation:
 x 
V(x) = (2x -1) ln  
 1-x 
The work WSWU (separative work per unit time) necessary to separate a mass F
of feed of assay xf into a mass P of product assay xp, and tails of mass T and
assay xt is expressed in terms of the number of separative work units needed,
given by the expression:
WSWU = P.V(xp ) + T.V(xt ) - F.V(xf )

It should be noted that SWUs required for enrichment increase with decreasing levels
of desired isotope in the waste streams. But the amount of feed material needed will decrease
with decreasing levels of the desired isotope in the waste streams. Thus they operate in the
opposite direction. For example in uranium enrichment we have the following expressions:
For material balance:
Total U = F = P + W (F = Feed, P = Product, W = Waste all in Kgs)
U-235 = F.xf = P.xp + W.xw (x is atom fraction of U-235)

 xp −xw   xp −xf 
F = P  W = P 
 xf − xw   xf −xw 

WSWU function after substituting for value function is written as

  xp   xw   xf 
WSWU = P(2xp -1)ln   + W(2x w -1)ln   - F(2xf -1)ln  
 1-xp  1-xw   1-xf 

SWU required per kg of product (P) produced is given by

WSWU  xp  W  xw  F  xf 
= (2xp -1)ln   + (2x w -1)ln   - (2x f -1)ln  
P 1-xp  P 1-xw  P 1-xf 
Let us calculate the amount of feed (F in kg) required to produce 1 kg of product and the
number of SWUs needed for this operation in two cases:

Case # F (kg) SWU

Case 1: Xf = 0.00711,
176 229
Xp = 0.9, Xw = 0.002
Case 2: Xf = 0.00711,
218 193
Xp = 0.9, Xw = 0.003

It is seen that feed and SWUs operate in opposite direction. If the availability of feed is
no problem, one can save on energy consumption by allowing larger fraction of desired isotope
in the waste streams.
Various methods are employed for the separation of stable isotopes. Among these,
gaseous diffusion, gas centrifuge, aerodynamic methods, distillation, electrolysis and chemical
exchange methods, electromagnetic separation, thermal diffusion, laser methods of separation
and ion migration are described in this chapter.

Gaseous Diffusion Process

The gaseous diffusion process makes use of the phenomenon of molecular effusion to
effect separation. If a gas is allowed to pass through a porous membrane with pore sizes equal
to the at molecular dimensions, the relative frequency with which molecules of different
species pass through the pores is inversely proportional to the square root of their molecular
weights. For a mixture of two masses M1 and M2 (MI < M2), this ratio, called separation factor,
α, is given by
α = [M2 / M1 ]1 / 2
This ideal value can be realised only at the beginning of the diffusion process. As
diffusion proceeds, the concentration of the lighter constituent in the diffusate increases and
there is a tendency for back diffusion to occur. As a result, the effective value of the
separation factor decreases. On the other hand, if diffusion were allowed to proceed long
enough, the composition of the gas would remain same on both sides of the barrier. In
practice, about half of the gas is allowed to pass through the barrier. As the value of α is so
close to unity. to obtain a useful degree of separation. the process must be repeated many
times in a counter-current cascade of gaseous diffusion stages. Obviously the feed should be
available in the form of gas, which should be compatible with the materials used. Otherwise
special precautions (as in the case of highly corrosive gas UF6) have to be taken.
 Tails Pump
A B C D

Feed Product

Fig.3 Typical diffusion counter-current cascade

Diffusion barrier materials should have pore sizes between 0.01-0.05 µm. They should
be chemically inert towards the operating gas. One way in which such barriers could be made
is by etching a thin sheet of silver-zinc alloy with HCl acid. The acid would then dissolve out
zinc atoms leaving a large number of pores in the sheet of metal. Other materials like sintered
alumina, nickel, anodically oxidised aluminium, teflon powder pressed into nickel gauge etc.
can also be used.

Gas Centrifugal Methods

The use of centrifugal fields for isotope separation was first suggested in 1919; but
efforts in this direction were unsuccessful until 1934, when J.W. Beams and co-workers at the
University of Virginia applied a vacuum ultracentrifuge to the separation of chlorine isotopes.
When a gas or vapour flows into a rapidly rotating centrifuge, the force acting on the
molecules will produce an increased concentration of the heavier isotope at the walls, while
the lighter isotope tends to collect nearer the axis of rotation. If the centrifuge is vertical, a
current of vapour can be made to flow down near the walls and up around the central axis. It
should then be possible to draw off a product richer in the lighter isotope at the top of the
apparatus, near the centre, whereas the heavier species would be removed at the bottom near
the periphery. The separation factor α for centrifugal method along the radius is given by

 (M -M )ω2a2 r2 
α= exp  2 1 (1- 2 )
 2RT a 

where M1 and M2 are the masses of the lighter and heavier isotopes, ω is the peripheral
velocity of the molecules, a is the radius of the centrifuge and r is the radius at any given
location in the centrifuge. R is the molar gas constant and T is the absolute temperature. The
separation factor increases with the length of the rotor, the peripheral speed and also with the
radius.
It may be seen that α depends on the difference between the masses and not on their
235 238
ratio. Thus better separation would be obtainable, in principle, of the U and U isotopes of
uranium than of the isotopes of hydrogen with masses 1 and 2. Also, since the difference in
the atomic masses is always same for a given element, the efficiency is independent of the
molecular weight of the compound whose vapour is being centrifuged. A long, thin vertical
cylinder made of material with high strength and relatively low density is used in the
construction of centrifuges. In a typical centrifuge rotating at a peripheral speed of 500 m/s,
the abundance ratio of heavier isotope to lighter isotope with mass difference = 3 at the outer
radius is greater than at the centre by a factor of 1.162 and the pressure at the outside is
greater than at the centre by a factor of 4.6 x 107. Countercurrent flow between the stream
near the outer radius and the stream near the axis is induced either with the help pf internal
scoops and baffles or by establishing a temperature gradient along the wall. This leads to
higher separating factors resulting in higher enrichment.
The main subsystems of the centrifuge are (1)
rotor and end caps; (2) top and bottom
bearing/suspension system; (3) electric motor and
power supply (frequency changer); (4) center post,
scoops and baffles; (5) vacuum system; and (6)
casing. Because of the corrosive nature of UF6, all
components that come in direct contact with UF6 must
be must be fabricated from, or lined with, corrosion-
resistant materials.
The materials used for constructing the
centrifuge should withstand tremendous centrifugal
forces resulting due to high speed revolutions of the
vertical cylinder (50000 to 100000 rpm). The
peripheral speeds reached under these conditions are
more than 500 m/s. The primary limitation on rotor
wall speed is the strength-to-weight ratio of the rotor
material. The materials therefore should have high
tensile strength, low density and high modulus of
elasticity. Further, the material should be inert to the gas being centrifuged. Aluminium alloys,
high tensile steels, titanium, glass fibres are some of the materials preferred. Another
limitation on rotor speed is the lifetime of the bearings at either end of the rotor. Rotor length
is limited by the vibrations a rotor experiences as it spins. The rotors can undergo vibrations
similar to those of a guitar string, with characteristic frequencies of vibration. Balancing of
rotors to minimize their vibrations is especially critical to avoid early failure of the bearing and
suspension systems. Because perfect balancing is not possible, the suspension system must
be capable of damping some amount of vibration.
The casing is needed both to maintain a vacuum and to contain the rapidly spinning
components in the event of a failure. If the shrapnel from a single centrifuge failure is not
contained, a “domino effect” may result and destroy adjacent centrifuges. A single casing may
enclose one or several rotors.
 One of the key components of a gas centrifuge enrichment plant is the power supply
(frequency converter) for the gas centrifuge machines. The power supply must accept
alternating current (ac) input at the 50- or 60-Hz line frequency available from the electric
power grid and provide an ac output at a much higher frequency (typically 600 Hz or more).
The high-frequency output from the frequency changer is fed to the high-speed gas centrifuge
drive motors (the speed of an ac motor is proportional to the frequency of the supplied
current). The centrifuge power supplies must operate at high efficiency, provide low harmonic
distortion, and provide precise control of the output frequency.
Although the separation factors obtainable from a centrifuge are large compared to
gaseous diffusion [ranging from 1.01 to over 1.10], several cascade stages are still required to
produce even LEU material. Separation factors depend on the absolute mass difference
between isotopes (not the mass ratio) and the square of the peripheral speed. Separation
factors for U-235/238 range from 1.026 for a 250 m/sec centrifuge to over 1.233 for a 600
m/sec centrifuge. The power consumption of a centrifuge plant per unit separative capacity is
less than that of a gaseous diffusion plant. For a given mass difference between the isotopes,
the stage separation factor is more than in gaseous diffusion plant. This means far fewer
stages in series would be needed with the centrifuge to arrive at a given enrichment. For
example, to get 3 % enriched uranium, 13 stages are needed as compared to about 1300
stages required in the gaseous diffusion plant. This advantage is however, partly off set by a
lower yield per stage compared to the process of gaseous diffusion. This means a large
number of centrifuges have to he operated in parallel to multiply the net yield.
With current technology, a single gas centrifuge is capable of about 5 separative work
units [SWU] annually, while advanced gas centrifuge machines can operate at a level of up to
perhaps 40 SWUs annually. About 100-120,000 SWU is required to enrich the annual fuel
loading for a typical 1000 MWe light water reactor.

Advantages of gas centrifuge over gas diffusion processes of enrichment

1. Higher separation factor hence requires less number of stages.

2. Absence of inter-stage gas compressors in centrifuge plant allows it to be squared off
more towards ideality. Whereas in case of gas diffusion plant use of compressors
makes it necessary to go for bigger squaring off (more off from ideality) in order to
avoid use of large number of compressors of different capacities. This makes the
centrifuge cascades more efficient.
3. Gas centrifuge being modular in construction, capacity addition can be done more
easily. The plant can initially be constructed for lower capacity and can subsequently be
expanded without much penalty.
4. Gas diffusion plants must be of large capacity to be economical due to requirement of
large number of supporting systems like captive power plants etc. Whereas the gas
centrifuge plants can be economical in smaller capacities.
 5. Higher material inventory in gas diffusion plant makes it more difficult to switch over
from one level of enrichment to another in an operating plant without a sufficient lead-
time. This reduces flexibility of the plant in catering to different users requiring
different enrichment levels in short delivery periods. In G.C. plants this problem does
not arise due to much lower material inventories.
6. It has low equilibrium time, which reduces time between start up of the plant and start
of withdrawal of product. Gas centrifuge process is considered superior above nozzle
process also because of low separation factor (compared to gas centrifuge) and very
high-energy consumption of nozzle process.

Aerodynamic Process
Processes in which isotopic composition changes are produced when a flowing gas
mixture experiences large linear or centrifugal acceleration are termed aerodynamic
processes. These processes are characterised by high speed flow of gases along a stream lined
curvature. Again, like in a centrifuge, the heavier species tend to concentrate on the outer side
of the curvature. One of the important aerodynamic processes is the jet nozzle separation
process. This process has basically been developed for uranium enrichment.
In the jet nozzle process, a mixture of
about 95% H2 and 5% UF6 at a pressure of
about 1 atm. is allowed to expand through a
narrow nozzle (0.01 mm wide) in to a curved
(0.1 mm radius) wall. The high speed gas
experiences forces 160 million times the
gravitational force in the curved nozzle. The gas
stream coming out of the nozzle is divided into
lighter and heavier fractions by a very sharp
skimmer knife. The separation factor depends
on the configuration of the nozzle, type of
diluent gas and its abundance in relation to UF6, the inlet's absolute pressure and expansion
ratio of the heavy fraction. The separation factor is higher for (i) higher pressure ratios
between the inlet and outlet streams and (ii) lower expansion ratio (flow rate of the light
fraction to that of the feed). Separation factor for the dimensions given above is around 1.015
for U isotopes. About 740 stages are required to produce 3% enriched uranium. Dilution of UF6
with hydrogen has two beneficial effects: it helps to increase the speed of flow and it delays
the establishment of a hypsometric distribution of Uf6 density. This delay reduces the re-
mingling by diffusion of the isotopes of uranium already separated by the centrifugal forces.
The Electromagnetic Method
Electromagnetic separation of isotopes of elements makes use of principle underlined in
mass spectrometry. In short, whenever an ion of charge ne (n is number of charges) and
mass m (amu) is accelerated in an electric potential of V (volts), it gains kinetic energy, which
is given by
1
K.E. = mv2 = neV
2
where v is the velocity of ions. If these accelerated ions pass through a magnetic field of
strength H (Tesla), in a direction normal to the application of magnetic field, then a force
equivalent to Hev is experienced by the ions. The direction of this force is mutually
perpendicular to the direction of the ions as well as magnetic field. This force is balanced by
the centripetal force given by mv2/r. So we have
mv2
Hev =
r
Where r is the radius of the path described by ions under the influence of the magnetic field.
Eliminating velocity (v) term from equations 1 and 2, we get the well known mass
spectrometric equation
m H2r2
=
ne 2V
Thus, ions of different masses but with same kinetic energy and charge describe different radii
in the applied magnetic field and can be resolved according to their mass to charge ratio. For a
given constant magnetic field and kinetic energy,
m
α r2
ne
Electromagnetic separation of isotopes is therefore possible.

Historical Perspective
Electromagnetic separation of uranium isotopes was one of the three major processes
considered by USA under the “Manhattan Project” during World War II period. The other two
were gaseous diffusion and centrifugal separation. Of the many electromagnetic schemes
suggested, three soon were recognized as being the most promising: the “calutron” (a name
representing a contraction of “California University cyclotron”) mass separator, the
magnetron-type separator later developed into the “ionic centrifuge,” and the “isotron”
method of “bunching” a beam of ions. The first two of these approaches were followed at
California and the third at Princeton.
Calutrons are basically 1800 deflection magnetic analysers. The isotron is an
electromagnetic mass separator using extended source of ions, in contrast to the slit sources
used in ordinary mass spectrographs. The ions from the extended source are first accelerated
by a constant, high-intensity, electric field and are then further accelerated by a low-intensity
electric field varying at radio frequency and in “saw tooth” manner. The effect of the constant
electric field is to project a strong beam of ions down a tube with uniform kinetic energy and
therefore with velocities inversely proportional to the square root of the masses of ions. The
varying electric field, on the other hand, introduces small, periodic variations in ion velocity,
and has the effect of causing the ions to “bunch” at a certain distance down the tube. (This
same principle is used in the klystron high frequency oscillator, where the electrons are
“bunched” or “velocity-modulated.”). The bunches of ions of different mass travel with
different velocities and therefore become separated. Separation of isotopes using Magnetron
or ionic centrifuge was based on the fact that the strong centripetal acceleration in a rotating
plasma typically of the order 5 x 1010m/sec2, can be used for separating difference species of
elements and isotopes.

Isotope Separators
Electromagnetic separators are large-radius (e.g., 24 in.), 1800 deflection mass
spectrometers. Ions are normally obtained in low voltage (100-150 V) arc discharge sources.
The arc current is about 3-5 A, while the ion current obtained is of the order of a few hundred
mA. Ionization efficiency is usually of the order of 5%. Gaseous materials, often in the form of
chlorides SnCl4, SiCl4) are continuously fed into the arc chamber through a controlled leak.
The pressure inside the arc chamber must be kept constant to obtain a steady ion beam.
Compounds which can be volatilized in the 100-6000C-temperature range (chlorides, bromides
of many elements) are introduced through a heated stainless steel “charge bottles”.
Temperatures as high as 12000C can be obtained with graphite “charge bottles”.
The pumping systems of large separators are impressive indeed. The main requirement
is pumping capacity rather than ultimate vacuum. In a large instrument, total pumping
capacity as high as 20,000 liters/sec is employed to maintain a vacuum of 10-5 torr.
The collection of the separated isotopes presents many problems. First of all, considerable
cooling is necessary. When an ion beam of 100 mA, accelerated to 35 kV, impinges on a
metallic receiver, the heat generated is of the order of 3.5 kW.
Other problems which must be considered in the design of a collector system are
erosion, sputtering, efficiency, and technique of isotope removal. Graphite and copper are the
most frequently used collector materials. The collected material is often removed from
graphite by simple scraping or by the burning of the entire pocket in oxygen. Dissolution in
nitric acid is a frequent removal method with copper pockets. Mercury isotopes may be
collected in silver pockets from which removal is easily accomplished by heating in vacuum.
Enrichment factor and production rate are two important performance parameters in
electromagnetic separators. The enrichment factor is the ratio of the ratios of the isotopic
concentrations in the product and the starting material, i.e., A/B in the product over A/B in the
feed gas, where A is the “wanted” isotope. Enrichment factors are normally in the 30-50
range; in special applications enrichment may be as high as 1000. The production rate of the
separators greatly depends on the throughput. A current of 1 mA corresponds to 6.3 X 1015
unit charges per second. A current of 100 mA will produce about 10 g of a nuclide of mass 100
in a full day’s operation; about one-half of this amount may be recovered from the collector.
Since the efficiency of ionization, as mentioned, is below 10%, at least 100 g of charge
material is required.
When the “wanted” isotope is a rare nuclide, production rate may be very poor; for
4O
example, K can only be produced at a rate of 0.5 mg/day. In cases like this, other
separation methods such as chemical exchange, thermal diffusion, and distillation should be
used if at all possible.

Principal Considerations of Electromagnetic Separators

They are threefold: First, it is difficult to produce large quantities of gaseous ions.
Second, a sharply limited ion beam is usually employed so that only a fraction of the ions
produced are used. Third, too great densities of ions in a beam can cause space-charge effects
which interfere with the separating action. Electromagnetic methods developed before 1941
had very high separation factors but very low yields and efficiencies. Since that time it has
been shown that the limitations are not insuperable. In fact, the first appreciable-size samples
of pure U-235 were produced by an electromagnetic separator. In addition to the foregoing
factors, one has to consider the ensuing factors:
• The Number of Stages: As in all methods, a compromise must be made between yield
and separation factor. In the electromagnetic system, the separation factor is much
higher than in other systems so that the number of stages required is small. There was
a possibility that a single stage might be sufficient. Early studies indicated that
attempts to push the separation factor so high as to make single-stage operation
feasible cut the yield to an impracticably small figure.
• The size of a unit: As determined by the radius of curvature of the ion path the length
of the source slit, and the arrangement of source and receivers;
• The maximum intensity of magnetic field required;
• Whether or not to use large divergence of ion beams;
• The number of ion sources and receivers per unit;
• Whether the source should be at high potential or at ground potential;
• The number of accelerating electrodes and the maximum potentials to be applied to
them;
• The power requirements tar arcs, accelerating voltages, pumps, etc.;
• Pumping requirement;
• Number of units per pole gap

Calutron Separation of Isotopes

It should be mentioned that much of the work on electromagnetic separation of
isotopes was carried out during Manhattan Project in USA and only in the fall of 1945 it
became possible to examine the electromagnetic isotope separation equipment to see if it
could be adapted to the separation of isotopes of elements other than uranium and which find
wide spread applications in fundamental or applied scientific research. This program is
operated now almost entirely for the isotopes produced and the scientific findings which they
make possible.
Fundamentally, there is little difference between a laboratory mass spectrometer and
its production counterpart, the calutron. The primary difference is that the calutron is designed
to collect useable quantities of desired isotopes, while the laboratory spectrometer is used
almost entirely as an instrument for detection or analysis. Of course, if one desired, very small
quantities of virtually any isotope could be collected in the laboratory instrument, but the
collection rates would be exceedingly small. The electromagnetic separation of isotopes by the
calutron involves the following phases:
• preparation of suitable charge material for the calutron ion source,
• operation of the calutron in separating the isotopes,
• quantifative recovery from the collectors and chemical purification.of the separated
isotopes,
• spectrochemical analysis of refined isotopes, and
• preparation of suitable sample and mass spectrometer analysis of enriched isotopes.

The flexibility which makes possible the electromagnetic separation of isotopes is

derived primarily from the magnetic field which can be set at a predetermined value to focus
ions of mass ranging from lithium to uranium at the collector. Also, ion accelerating and
focusing voltages are variable.

Calutron Ion Sources

Usually these are arc discharges. Calutron ion source units should possess a certain
degree of flexibility as to temperature range, accelerating electrode arrangement, arc number
and size, and cooling system. In processing a variety of elements and compounds having
widely varying vapour pressures, it is desirable to have source units operable over
temperature ranges required for realizing sufficient vapour pressures. Earlier investigations
carried with uranium indicated that the temperature range of uranium-type ion source units
could be extended sufficiently to cover approximately half of the elements having more than
one nuclide if the charge compound of the element under consideration is carefully chosen.

Charge Materials
For charge materials occurring as liquids or gases at room temperature, sample
introduction systems external to the ion source can be employed. The flow of feed material to
the arc region should be controllable with the help of a sensitive needle valve.
In the case of Solid materials, usually they are introduced in a contained inside the ion
source and are heated. Usually it is desirable to have a small flow of gas such as nitrogen into
the arc is beneficial, and with some elements necessary, for stabilizing the arc and promoting
production. The problem of producing ions of all isotopes for calutron operation is not
completely solved; however, methods and equipment now in use are adequate to produce
ions, with varying degrees of efficiency, from virtually any element in the periodic chart.

Magnetic Analysers
A special class of shaped magnetic lenses, used successfully in beta-ray spectroscopy
have found wide spread applications in isotope separations. The fundamental concept of this
focusing system relates to the use of magnetic field that decreases in intensity with increasing
radius. The approximate magnetic field intensity distribution across the mid plane in these
“inhomogeneous systems” is given by the relationship
n
r 
βr = β0  0 
r 
where βr is the mid plane magnetic field as a function of radius, r0 is the radius of the central
orbit, β0 is the mid plane magnetic field strength at r0 and n is the inhomogeneous filed index.
Siegbahn and Svartholm have shown that under these conditions, both radial (φr) and
azimuthal (φz) focusing occurs when φr = φz = √2n= 2550 and at this focusing angle and for n
= 0.5, spherical aberration is minimized. Both the radial and azimuthal or axial trajectories
converge to the same image point. As a result, the transmission of a 2550 analyser can
approach 100%. for reasonable ion source limits relating to angular divergence and ion energy
spread. In addition, increase in mass or energy dispersive power is realized. The increased
dispersion is extremely valuable for it permits the use of a magnet smaller than a
homogeneous type. For a magnet having a basic radius of curvature r0, with the object and
image symmetrically located, the dispersion D is given by
r0 ∆m
D=
m0 (1 − n)
where n is the index of inhomogeneity of the magnetic field. For a 0.5 index the dispersion is
doubled. Such an isotope analyser has been constructed. It had a transmission approaching
100% for several types of ion sources. The mean radius of curvature of this 75 ton
electromagnet is 6 cm, the pole-piece gap width at this radius is 10 cm, and the mass
dispersion at mass 100 is about 1.5 cm.

Isotope Collectors
For each element processed, an isotope collector must be designed, providing properly
spaced receptacles for whatever isotopes may be present. Isotope collector structural
materials, structural design, and cooling, all require considerable attention and have to be
chosen depending on the system being studied. Since one mass unit separation at the
collector decreases as the mass of the ions increase, each element requires especially
designed collectors. In addition to this consideration, which is essentially a problem in pocket
placement, suitable pocket materials and cooling techniques must be chosen in order to retain
the deposited material. The temperature of receiving surfaces, which must necessarily be
bombarded With ions, must be kept sufficiently low that deposited material Will be retained
and not lost to cooler surfaces such as walls of the vacuum chamber
through the vacuum system. The problem of building a collector for any element then consists
of calculating proper pocket spacing for the desired isotopes, and from vapor pressure
considerations, determining suitable pocket material and cooling techniques.
Usually the design of a collector is dictated by the desire for having maximum flexibility
so that collectors could be constructed for any element from a minimum number of parts. The
all-purpose type collector, in which pockets up to 10 can be assembled at predetermined
spacings has been fabricated for this purpose. This basic structure permits the placing of
collectors to receive the isotopes of any element in the periodic chart; pockets may be
replaced easily if desired, and cooling may employed as required.

Collector Materials
These vary with the element being processed, and the receiver parts may or may not
be water-cooled, depending on the energy to be dissipated and the vapor pressure of the
collected material. Furthermore, with some elements, special retention techniques are
required. For example: Sulfur is allowed to combine chemically with copper shavings packed in
the collector pocket. Mercury is retained by allowing it to amalgamate with silver collector
pockets, and by cooling the pockets with a circulating refrigerant. In the processing of
mercury, the problem of high vapor pressure in the calutron and the contamination of
collected isotopes by the condensation of neutral mercury vapor in the collector pockets,
needs to be eliminated by refrigerating the, vacuum chamber.

Charge Materials for Production of Ions

The availability or synthesis of suitable calutron charge materials is essential. Several
requirements must be met, such as suitable vapor pressure, at calutron operating
temperatures, simplicity of molecular structure, stability of the compound and absence of
water of crystallization. If the element itself has a suitable vapor pressure, it is generally used
as charge material since its use greatly reduces" extraneous ions or sidebands. For the same
reason, simple compounds generally are preferred over the more complex compounds; thus,
halides are used widely. Anhydrous compounds are necessary because the water of
crystallization gives rise to increased vacuum chamber base pressure and prolonged start-up
time.
The ultimate choice of the compound best suited for use in the electromagnetic
separation of the isotopes of an element in the calutron is often made only after the
systematic testing of a number of compounds. The criterion of suitability of such compounds is
whether or not they give rise to a satisfactory rate of production of the separated isotopes of
the given element. While a number of properties of the compounds are considered, chief
consideration is usually the vapor pressure-temperature relationship, since such compounds
must be sublimed into the arc chamber where positive ions are formed. Usually laboratory
studies involving mass spectrometer are a more satisfactory method for the evaluation of
compounds. The compound under consideration is introduced into a mass spectrometer, and
the ion current due to each of the fragments or species formed is plotted as a function of the
mass of the fragment. This gives the relative abundance of the ions formed due to dissociation
and ionization by the slow electrons used. The appearance potential is also determined for the
singly charged metal ion involved, It then follows that if the number of singly charged metal
ions formed is a fair proportion of all of the positive ions thus produced from a compound, and
if the appearance potential of the metal ions is not greatly in excess of the ionization potential
of the metal, then the rate of positive ion production, and therefore the rate of production of
the isotopes of the element in question, will be high.
A wealth of basic inorganic chemistry has resulted from these charge material
developments. Suitable charge materials of some 43 elements have been supplied for calutron
separations of isotopes, including the synthesis of many unusual compounds and, the
development of methods of purification to
meet special requirements.

Calutron Operational Procedures

After suitable ion sources, collectors, and charge materials are identified and
assembled, the components are placed in an evacuated volume between the poles of an
electromagnet. Metered ion currents at the collector are necessary to properly monitor the
isotope beams and afford a means for estimating total isotope collection. Thus, for a singly
charged ion, 1 amp.-hr, at the collector represents approximately 0.04 gm.-atom of the
monitored isotope. The length of time allowed for a collection is determined from the metered
production rates on the desired isotopes, with some allowance for collector rejection and
chemical losses.

Chemical Purification of Separated Isotopes

From the collector pockets, the isotope contents are removed by one of several
methods depending on the element and pocket material. Extreme care is exercised during this
phase to prevent isotope contamination and to assure quantitative removal of the product
isotope.
Chemical purification procedures vary widely, depending upon the element being
processed. Reactions must be complete, filtrations or centrifugations must be carefully
performed, and care must be exercised to minimize losses caused by discarding reaction side
products. It may be mentioned that by the beginning of 195I, adequate purification techniques
had been worked out in USA for every element which had been processed in the calutron.
These included radioactive materials such as beryllium-10 and potassium-40, rare earths such
as cerium, neodymium, and samarium, and toxic materials such as selenium and mercury.
Procedures for separation of elements with similar chemical properties have also been worked
out. For example, a method for the production of pure zirconium and hafnium fluorides was
developed. It was found that indium may be separated from gallium by the use of mandelic
acid.
It was also found that mandelic acid precipitates some of the rare earths and is useful
in their separation. The physical and chemical properties of the mandelates of zirconium and
hafnium were thoroughly investigated. Also, it was found that the homogeneous precipitation
of the rare earths by hydrolysis of methyl oxalate is a selective process.

Chemical and Mass Analysis of Separated Isotopes

Upon completion of chemical purification, product isotopes are subjected to chemical
and mass analyses. Methods for handling small samples of precious isotope materials are
being continuously improved. In mass spectrometry, experience and development have made
it possible to mass analyze isotope collections from all elements thus far processed in the
calutron. More than 700 samples of isotopes of elements were prepared as compounds
suitable for mass analysis.
The interest in using certain stable isotopes as tracers has stimulated the development
of microwave methods for isotopic analyses. The resolution of the microwave spectrometer is
so great that theoretically there is room for more than 50 million rotational lines in the present
experimental region. Although considerable progress, has been made, isotopic analysis by
microwave techniques do not yet meet the high precision of a mass spectrometer.

Isotopes Enriched Electromagnetically

By December 31, 1950, a total of 43 elements from lithium to lead comprising 177
isotopes had been separated in the calutron in milligram to multigram quantities.

Thermal Diffusion
When heat flows through a mixture initially of
uniform composition, small diffusion currents are set up,
with one component transported in the direction of heat
flow, and the other in opposite direction. This is known as
thermal diffusion effect. The effect is generally small. For
example when a mixture of 50% hydrogen and 50 %
nitrogen is held in temperature gradient between 260 and
10°C the difference in composition at steady state is only
5%. In isotopic mixtures the effect is even smaller. The
mixture to be separated is confined in a long, vertical tube,
cooled externally and heated internally by a hot wire at the
axis of the tube. Concentric annular type cylinders can also
be used. In both types, the mixture to be separated is
confined in a narrow space between the inner heated and
an outer cooled surface. As a result of thermal diffusion
effect, the lighter isotope usually concentrates in the inner zone at higher temperature. At the
same time, convection currents are set up, with the lighter isotope adjacent to the inner wall
moving upward and the heavier isotope adjacent to the outer wall moving downward. Thus it
is possible to achieve substantial degrees of separation in a practical length of column. It is
preferable to work with gases rather than liquids because the higher diffusion coefficients
result in higher separative capacity. The optimum space between hot and cold surfaces is a
few mm for gases and fraction of mm for liquids. A series of very long columns (tens of
meters) are essential to achieve reasonable degree of enrichment.

The separation factor a is given by

T 
ln α = r ln  1 
T 2 
where r is thermal diffusion factor proportional to (M2-M1)/(M1+M2), T1 = Hot wall temperature
and T2 = cold wall temperature.
The thermal diffusion factor r can be calculated from kinetic theory of gases. If r is
positive, lighter isotope concentrates at the higher temperature. The thermal diffusion factor, r
is strictly not a constant and varies with temperature for certain types of gases. Table 4 gives
examples of the highest reported concentrations of separated isotopes that have been
obtained by thermal diffusion.

Table 4 Isotopes separated by thermal diffusion method

Working fluid (Gas) Isotope separated % in product

35
HCl Cl 99.6
37
Cl 99.4
84
Kr Kr 98.2
86
Kr 99.5
78
Kr Kr 10.0
86
Kr 96.1
17
O2 O 0.5
18
O 99.5
15
N2 N 99.8
134
Xe Xe 1.0
136
Xe 99.0
3
He He 10.0
36
Ar Ar 99.8
38
Ar 23.8
20
Ne Ne 99.99
22
Ne 99.99
13
CH4 C 90.0
12
C 4.4
Laser Isotope Separation
The possibility of using the slight differences that exist in the absorption spectra of
isotopes of an element for isotope separation has been recognised ever since isotopes were
discovered. Initially photochemical separation of isotopes was restricted to very few elements
(Cl, Hg). But the advent of lasers provided the intense, monochromatic, tunable light source
needs to make photochemical isotope separation applicable to all elements, at least on a
laboratory scale. Two general methods have been proposed for separating uranium isotopes:
(i) Laser isotope separation of uranium metal vapour and (ii) Laser isotope separation of UF6
235
Laser isotope separation of uranium: U in uranium metal vapour is ionised selectively and
238
then separated from unionised U by deflection in electric or magnetic fields. The absorption
spectrum of uranium metal vapour is very complex, with over 3,00,000 lines at visible
wavelengths. Many of these absorption lines are sharp, with sufficient displacement between a
238 235
U absorption line and the U line for the corresponding transition and without overlap of
235 238
the U line with the U line for a different transition, to permit selective excitation of the
235
U atoms. In actual practice two lasers, one a narrow frequency laser supplying visible light
235
at 502.72 nm to excite U followed by another supplying ultra violet light at 262.5 nm to
235 238
ionise the excited U atoms, are employed. U does not absorb 502.74nm line and absorb
only at 262.5 nm, thereby achieving the selective ionisation.
235
The selectively ionised U ions are deflected by electric or magnetic fields or collector
235
plates. Even though photon absorption process selectively ionises U, charge exchange
235 238 238
between U ions and neutral U atoms and atomic collisions deflect enough U atoms to
the collector plates limiting the maximum enrichment possible to attain. Pressures of uranium
metal vapours are usually kept below 1 torr, to minimise the collisions.
235
Laser isotope senaration of UF6: In this method, UF6 in UF6 vapour is excited selectively and
caused to react chemically to produce a solid lower fluoride, which is then separated from
238
unreacted UF6 vapour.
The absorption spectrum of UF6 is far more complex than that of uranium metal,
because the spectrum of the UF6 molecule involves transitions between many vibrational and
235
rotational states that are absent in uranium atom. Further, absorption bands of the UF6
238
molecule overlap those of UF6 so that highly selective absorption by one isotope is seldom
235
found. For instance, at room temperature, the peak in the UF6 absorption band is displaced
-1 238
by O.55 cm from the peak in the UF6 absorption band at a wave number of 625 cm-1.
238 235
However, the absorption by UF6 at the peak absorption by UF6 is so great as to preclude
selective absorption under these conditions.
However, it was shown theoretically that if all transitions were to occur from the lowest
235
vibrational levels, the fine structure of the absorption bands would be such that UF6
238
absorption maximum might be found at a UF6 absorption minimum. Then a tuned laser
beam with a frequency spread narrower than line spacing of 0.16 cm-1 might be able to excite
235 238
UF6 to the first vibrational level without exciting UF6. Such selective absorption from the
lowest vibrational level is not possible at room temperature. Further, at room temperature,
 238
even UF6 molecules will be in excited states and many of these would get dissociated atoms
235
with UF6 molecules during the chemical reactions. At sufficiently lower temperature (< 77K)
235
it is thus may be possible to realise the selective excitation of UF6 molecules from the lowest
vibrational levels. But at such low temperatures the vapour pressure of UF6 is expected to be
very low 5 x l0-25 torr. What is normally done is to cool a dilute mixture of UF6 in hydrogen to
30K by expansion to high speed through a hypersonic nozzle. The sub cooled UF6 molecules
can remain uncondensed to assume the low temperature energy distribution and display an
235 238
absorption spectrum in which UF6 do not overlap with the UF6 lines. This high speed sub-
235
cooled gas mixture is first irradiated by 16 µm light of a frequency absorbed by UF6 and not
238 235
by UF6 and then by additional light of sufficient energy to dissociate the excited UF6 but
238 235
insufficient to dissociate the unexcited UF6. The dissociated lower fluoride of UF6 and
238
undissociated UF6 are then separated by chemical means.
Efforts have also been made to separate deuterium isotope from hydrogen containing
species. In general, two methods are being pursued; (i) photon dissociation of formaldehyde
with UV lasers and (ii) infrared photolysis of methane derivatives by multiphoton dissociation.
In UV photon dissociation, HCHO is irradiated with UV laser of about 325 nm when the photon
dissociation reaction HCDO → HD + CO occurs. The reaction is carried out at a temperature of
350 K and a relatively higher pressure of about 100 torr. This is to prevent the polymerisation
of formaldehyde, though at high pressures absorption selectivity is somewhat reduced. In the
second method, freon 123 (CF3DCCl2) or fluoroform (CDF3) is subjected to photolysis with
10.65 or 10.2 µ IR laser. In practice, formaldehyde route is preferred as it involves only a
single photon conversion and also easily separable gaseous products are formed. In addition,
corrosion problems are also less. Table 5 lists some of the isotopes separated using lasers.

Table 5 Isotopes separated using lasers

System Isotopes separated Lasers used

H2CO H, D 319 nm frequency doubled dye 325 nm HeCd laser
10
BCl3 B, 11B 10 µm CO2 + UV flash lamp
10 µm CO2
12
CF2Cl2 C, 13C 10 µm CO2
12
CS2 C, 13C, 32S, 34S 193 nm ArF
16 , 17 , 18
O2 O O O 193 nm ArF
235 , 238
UF6 U U 16 µm IR = UV

Distillation methods
In general, isotopic species have different vapour pressures (and boiling points), so
that partial separation by fractional distillation can be employed for their separation. For
example, the boiling point of D2O is about 1.4° higher than that of H2O. This difference,
though small, should make separation by distillation practical, provided efficient columns are
used. Deuterium can also be separated by fractional distillation of hydrogen gas at low
temperatures (< -250°C). Heavy hydrogen gets concentrated in liquid phase.
The separation factor α, for the desired component is given as

x/(1-x)
α =
y/(1-y)
where x is the fraction of the component in the liquid phase and y is the fraction in the vapour
phase. Deuterium in hydrogen or in water is always in equilibrium according to the reactions
H2 +D2 ↔ 2HD and
H2O + D2O ↔ 2HDO
If the liquid and vapour phases form ideal solutions, the separation factor α can be written as
1/2 1/2
 π(H 2 )   π(H 2 O ) 
α(H /D ) =   α(H 2O /D 2O ) =  
 π (D 2
)   π (D 2
O ) 
where π is the vapour pressure. The separation factors for H2/HD and H2O/D2O at their boiling
points are 1.81 and 1.03 respectively.
The elementary separation factor in the distillation process can be multiplied by inducing a
countercurrent flow in the distillation columns (or towers). These columns are cylindrical
containers within which contact is made between the two phases; (i) the liquid phase that
flows by gravity from top to bottom and (ii) the vapour phase that, produced in a reboiler at
the base of the column, flows from bottom to top. Distillation columns can be divided into two
basic categories; plate columns and packed columns. In the former, there is repeated contact
between the two phases, whereas the latter operates on the principle of continuous contact.

Electrolysis
During the electrolysis, the gas discharged at the cathode of an electrolytic cell is
distinctly poorer in heavier isotope content than in the electrolyte. Separation factor is defined
as
(M 1 /M 2 ) electrolyticgas
α =
(M 1 /M 2 ) electrolyte
It has been found that α depends on the cathode material, electrolyte composition and cell
temperature. In the case of water for example, a high value of 8.1 has been reported.
Separation factor is higher for an alkaline electrolyte than for acid. With KOH at 15°C, a pure
iron cathode gave the highest value of 13.2. Minimum is expected to be about 3.8 as dictated
by the equilibrium constant for the reaction

H2O(l) + HD(g) <= = > HDO(l) + H2(g)

A typical electrolysis set-up consists of a vertical cylindrical cell wherein the central anode is
separated by a porous diaphragm of nickel wire-reinforced asbestos from the outer casing
metal (steel) which serves as cathode. The electrolyte is generally a solution of 25 % by
weight of potassium carbonate. It would appear that a cascade of a number of stages, about
15, would be sufficient to arrive at 99.9 % D2O. But practical difficulties such as loss of a vast
amount of deuterium in the early stages of the cascade come in the way. Stage losses of
hydrogen with sufficient deuterium content can be prevented by incorporating the steam-
hydrogen exchange process. The reaction HD + H2O < = = > H2 + HDO attains equilibrium at
elevated temperatures (> 80°C) over a catalyst (supported nickel or nickel on chromia or
platinum on charcoal). The equilibrium constant of the above reaction is 2.8 at 80°C. Hence
rapid exchange of deuterium occurs with hydrogen in steam and after condensation the
deuterium rich water can be re-electrolysed for further enrichment.

Chemical Exchange Methods

The deuterium exchange reaction between water and hydrogen and other compounds
is one of the group of deuterium exchange reactions that have been extensively studied and
are the basis for the most of the world's heavy water production. Table 6 lists deuterium
separation factors between water and gaseous compounds of hydrogen.
In all the systems deuterium tends to concentrate in the liquid phase except in
ammonia-water at high temperature. Separation factors in chemical exchange are much
higher than separation factors in distillation. The high value of these separation factors and
their strong dependence on temperature give chemical exchange processes their importance
for separation of deuterium and isotopes of other elements.

Table 6 Separation factors for different exchange reactions

Reactants Products Separation factor (α) at different temperatures
Liquid Gas Liquid Gas 00 25° 50° 100° 125° 200
H2O NH2D HDO NH3 1.02 1.00 1.00 0.99 0.99 0.99
H2O PH2D HDO PH3 2.71 2.44 2.27 2.04 1.96 1.78
H2O HDS HDO H2S 2.60 2.37 2.19 1.94 1.84 1.64
H2O HD HDO H2 4.53 3.81 3.30 2.65 2.43 1.99
NH3 HD NH2D H2 4.25 3.62 2.99* 2.55 2.34 5.19#
CH3NH2 HD CH3NHD H2 4.85 3.60@ 7.90$
*at 60° #at -25° @at 40° $at-50°

The deuterium exchange reaction between water and ammonia, water and hydrogen
sulphide proceeds rapidly in the liquid without catalysis, because of ionic dissociation. On the
other hand, deuterium exchange between water and hydrogen, ammonia and hydrogen or
methylanline and hydrogen does not proceed without catalysis. The water-hydrogen reaction
can be catalysed by nickel or platinum metal catalyst, ammonia-hydrogen reaction by
potassium amide dissolved in liquid ammonia and the methylamine-hydrogen reaction by
potassium methylamide.
 Solutions used in the ammonia-water, water-hydrogen and ammonia-hydrogen
processes are relatively non-corrosive and may be handled in ordinary steel equipment.
Solutions used in all of the other processes are relatively corrosive, and require use of
stainless steel or other expensive construction materials. For example, in water-hydrogen
sulphide exchange reaction (also known as Guilder-sulphide (GS) process), H2S is highly
corrosive. Carbon steel coated with iron sulphide and bubble trap trays made of SS-304 are
used.
In all the above processes, to achieve a multiplication effect for the overall separation
factor, the light and isotopic fractions collecting at the two ends of the column are reconverted
to the initial chemical forms and returned to the exchange column to recycle countercurrently.
To illustrate, consider the equilibrium reaction H2O(l)+HD(g) <==> HDO(l)+H2(g), The
hydrogen gas, depleted in deuterium, escaping at the top is burnt with oxygen and the water
formed flows back down the column again while the water, enriched in deuterium, collecting at
the bottom is electrolysed into hydrogen and the gas is made to re-ascend the exchange
column for further exchange with water flowing down.
Alternately, a dual temperature exchange or bi-thermal exchange process can be set
up to avoid the cumbersome exercise of
Feed: Recycle D2S
reconverting the products into initial
Natural water Blower reactants to achieve the multiplication
effect in the separation factor. The dual
Cold tower
T = 320C temperature exchange takes the
α = 2.32 temperature dependent property of the
equilibrium constant for the exchange
D2S flow reaction under consideration. For
Depleted water
example, the values for the K for the
Product D2O flow
reaction
Heat H2O(l) + HD(g) < = = > HDO(I) + H2(g)
exchangers

for the temperatures 250C and 800C are

3.78 and 2.80 respectively. Similarly for
Hot tower
T = 1280C the reaction H2O(l) + HDS(g) < = = >
α = 1.80 HDO(l) + H2S(g), the value of K changes
from 2.32 at 32°C to 1.30 at 138°C. By
carrying out the equilibrium exchange
process in one column (cold tower)
maintained at a lower temperature say at
32°C for water-H2S reaction and
Dual temperature Water – H2S Exchange
Girdler – Spevack Process (GS Process) transferring the enriched products to a
second column maintained at higher
temperature (hot tower) at 138°C, the proportion of the reactants to products change. Thus
by carrying out the exchange reaction alternately in the two columns at different
temperatures, the necessary multiplication effect on the separation factor is realised. Much of
the D2O is produced this way.
Chemical-exchange isotope separation requires segregation of two forms of an element
into separate but contacting streams. Since many contacts are required to achieve the desired
separation, the contacting process must be fast and achieve as much separation as possible.
For heavy elements such as uranium, achieving a suitable separation factor involves contact
between two valence (oxidation state) forms such as hexavalent [U6+ as in uranyl chloride
(UO2Cl2)] and the quadrivalent [U4+ as in uranium tetrachloride (UCl4)]. The 235
U isotope
exhibits a slight preference for the higher valence, for example, the hexavalent over the
quadrivalent in the Asahi process or the quadrivalent over the trivalent (U3+) in the French
solvent-extraction process.
The chemical-exchange process, developed by the French, is commonly referred to as
CHEMEX. It uses the exchange reaction that takes place between two valence states (U3+ and
U4+) of uranium ions in aqueous solution. Isotopic enrichment results from the tendency of
238
U to concentrate in the U3+ compound while 235
U concentrates in the U4+ compound. It is
therefore possible to obtain enriched uranium by removing the U4+ ions with an organic
solvent that is immiscible with the aqueous phase (concentrated hydrochloric acid). Several
possible extractants are available; however, tributyl phosphate (TBP), the choice of the
French, is typically used. TBP is diluted with an aromatic solvent, and this organic phase
moves countercurrent to the aqueous phase through a series of pulsed columns.
In the pulse column, the heavier aqueous phase is fed into the top of the column, and
the lighter organic phase is fed into the bottom of the column. A rapid reciprocating motion is
applied to the contents of the column, providing efficient and intimate contact of the two
phases. In an HEU plant, centrifugal contactors might be employed particularly for the higher
assay sections, since the stage times and corresponding specific uranium inventory could be
reduced significantly. After passing through the column, the enriched and depleted uranium
streams must be chemically treated so that they can be recirculated through the column again
(refluxed) or sent to another column for additional enrichment. This requires complicated
refluxing equipment at both ends of the column.
The ion-exchange process was developed by the Asahi Chemical Company in Japan and
uses the chemical isotope effect between two valences (U4+ and U6+) of uranium. In this
process, the organic phase is replaced by a proprietary ion-exchange resin. The aqueous
phase flows through the stationary resin held in a column, and the net effect of all the
chemical reactions is a “band” of uranium that moves through the ion-exchange column. The
exchange between the unadsorbed uranium flowing through the band and that adsorbed on
235 238
the resin enhances the isotopic separation. In this continuous separation system, U and U
tend to accumulate respectively at the entrance and exit ends of the adsorption band. In this
process, it is economical to regenerate many of the chemicals by reaction with oxygen and
hydrogen in separate equipment.
 The development and manufacture of the appropriate adsorbent beads are based on
technology and know-how gained by Asahi in over 25 years of ion-exchange membrane
development and manufacture. The adsorbent is a spherical bead of porous anion-exchange
resin with a very high separation efficiency and an exchange rate over 1,000 times faster than
the rates obtained in most commercially available resins. The two exchange processes
discussed here are representative of exchange processes now under study in several
countries. At present, no country has built or operated a full-scale uranium enrichment plant
based on an exchange process. The primary proliferation concern is that they are based on
standard chemical engineering technology (except for the proprietary ion-exchange resins).
Chemical exchange processes have also been investigated to separate other isotopes
such as C, N, 0, U. But none could go beyond laboratory scale only. Recently a pilot plant for
235
the enrichment of U using cation exchange resins has been set up by Japanese. Difference
235 238
in the kinetics of redox reaction between the isotopes U and U is exploited for getting
enrichment. The reaction
238
UO2+ + 235
U(IV) → 235
UO2+ + 238
U(IV)
235
with an equilibrium constant of 1.0015 is exploited to enrich U. Rate of exchange is first
order with respect to U(IV). Feed composition employed was U concentration = O.1M and free
acid = 0.24 to 0.4 M. A cation exchange resin has been employed and the process is carried
out in H2SO4 medium. Enrichment is effected by repeated oxidation and reduction of uranium
on the column. It has been reported by Japanese that uranium enriched up to 3% U-235 could
be obtained after 20 days of continuous operation.

Ion Migration
Slight differences in velocities of isotopic ions in solution under an electric field, can
also be exploited for the separation of isotopes. These small differences are due to the
different sizes and masses of the ions. Contributions due to differences in the degree of
dissociation and in complex formation have also to be taken into consideration. The ion
migration can occur not only in aqueous media where the ions are invariably hydrated but also
in fused salt media where the ions are relatively more free from solvation effects. The major
advantage of the fused salt medium is the absence of ion solvation resulting in larger mass
effects in the migration of isotopic ions. The separation factor in ion migration process is
proportional to ∆v/v where ∆v is difference in velocities between the isotopes and v is the
mean velocity. The extent of separation effect between the two isotopes can also be expressed
in terms of relative mass effect given as
∆v / v
-µ =
∆m / m
where m is the mass of the ion. Thus the actual enrichment factor is somewhat less than
expected when only velocities are considered. More over, while electromigration builds up a
concentration gradient along the field direction, the reverse flow of the electrolyte due to
 39 41
diffusion tends to neutralise the effect partially. In a typical example, K and K were
separated by the electromigration of potassium chloride solution in a U-shaped tube using
platinum gauge electrodes, with an arrangement for controlled counter flow of HCl acid from a
reservoir above the cathode compartment. The tube was packed with glass beads or sand held
under pressure. The counterflows were so adjusted that the speed of H+ ions was intermediate
39
between those of K+ and 41
K+ ions. 41
K+ ions were washed back while the faster moving 39
K+
ions moved ahead thereby building an isotopic concentration.

Heavy Water plants in India

Kota
More than 90% of the cumulative world production of heavy water (D20) is by
the Hydrogen-sulphide (H2S) water (H20) exchange process, known as the GS
process, with the major contribution being from the Canadian plants. Heavy
Water Plant at Kota is a solely indigenous effort and is based on the
Bithermal H20-H2S exchange process. The plant is located at a distance of 65
KM from Kota Railway Station, adjacent to Rajasthan Atomic Power Plant
(RAPP). The Heavy Water Plant is integrated with RAPP for its supply of
power and steam. Water from the nearby Rana Pratap Sagar lake, purified of
suspended and dissolved impurities forms the process feed with the D20
enriched from 150 ppm (0.015%) in the feed to 15% D20 by chemical
exchange with H2S and later by vacuum distillation to produce 99.8% D20.

Baroda
Heavy Water Plant at Baroda is the first plant set up in the country for the
production of heavy water by employing Monothermal Ammonia-Hydrogen
exchange process. This facility also has a Potassium metal plant for supply of
Potassium metal for preparing catalyst solution for all Monothermal
Ammonia-Hydrogen exchange plants. In 1999, M/S GSFC commissioned their
new technology based low-pressure Ammonia plant (A4), operating at around
140 kg/cm² pressure, and stopped operation of their old high pressure
(640kg/cm²) ammonia plants. Operation of HWP had to be suspended
simultaneously due to non-availability of synthesis gas at high pressure.
Facing the challenge, HWP is modified to reorient its operation independent of
GSFC. The Ammonia Water Exchange Front-End unit is set up which
eliminates dependency of HWP on Ammonia plants of M/S GSFC for its
deuterium feed stock. It uses water as the source of deuterium in the process
and ammonia works as carrier gas only.

Hazira
Heavy Water Plant at Hazira, employs the ammonia-hydrogen exchange
monothermal process. The plant is located at a distance of about 16 km from
Surat city. Work on HWP (Hazira) commenced in August 1986 and the plant
was commissioned in February 1991.

Thal
The plant comprises of two streams consisting of two separate isotopic
exchange units, final enrichment units, final production units & cracker units,
but a common ammonia synthesis unit. Feed synthesis gas (a mixture of one
part of nitrogen and three parts of hydrogen containing deuterium from the
 Ammonia plant is routed through the plant at a flow rate of about 96 T/Hr. at
a pressure of about 180 kg/cm². The deuterium enriched ammonia from the
bottom of the exchange tower is then fed to the second isotopic exchange
tower where it gets further enriched by coming in contact with the enriched
synthesis gases obtained by cracking of enriched ammonia. A part of the
enriched gas and liquid from the second isotopic exchange tower is then
taken to the final enrichment section where the concentration of deuterium in
the ammonia can be further increased as desired upto 99.8%. Finally, the
enriched ammonia so obtained is made free of the catalyst and is cracked. A
portion of this enriched synthesis gas is burnt to produce heavy water.
However, for reasons of better recovery efficiency the concentration of
deuterium in ammonia in the final enrichment section is kept low so as to
produce heavy water of about 60% which is then vacuum distilled to produce
heavy water of nuclear grade.

Manuguru
The Heavy Water Plant at Manuguru, Andhra Pradesh is based on the
Bithermal Hydrogen Sulphide-Water (H2S-H20) Exchange Process. This plant
with a capacity of 185 MTY is the second plant based on this process, the
earlier one being at Kota, Rajasthan for which the complete technology has
been developed indigenously by BARC and HWB. Water from the nearby
Godavari river, purified from both suspended and dissolved impurities forms
the process feed and the D20 content is enriched from 150 ppm (0.015%) to
15% D20 by chemical exchange with H2S and later vacuum distilled to
produce 99.8% D20.

SUGGESTED READING
1. K. Cohen, The theory of isotope separation, Mc Graw Hill, New York (1951)
2. M. Benedict and T.H. Pigford, Nuclear chemical engineering, Mc Graw Hill, New York
(1965)
3. H. London, Separation of isotopes, George Newnes, London (1961)
4. S. Villani, Isotope separation, Amer. Nucl. Soc., Hinsdale, (1976)
5. H.J. Arnikar, Isotopes in atomic age, Wiley Eastern, New Delhi, (1989)
6. J. Koch(Ed.), Electromagnetic isotope separators and applications of magnetically
enriched isotopes, Interscience, New York (1958)
7. G.M. Murphy(Ed.), Production of heavy water, Mc Graw Hill, New York (1955)

QUESTIONS
1. Mention two important and widely used methods each for the separation of uranium
isotopes and hydrogen isotopes.
2. Bring out clearly the difference between the separation factor and enrichment factor.
3. Mention the physico-chemical property on which the following methods for separation
of isotopes are based: (a) gaseous diffusion (b) distillation (c) thermal diffusion (d)
chemical exchange
4. Define the terms separating unit, stage and cascade.
5. Why do we need to employ series of stages and cascades in the separation of isotopes?
6. Name some of the important applications of stable isotopes?
7. Better separation of uranium isotopes is possible in principle with gas centrifuge
methods than with gaseous diffusion methods. Why?
8. Why is UF6 diluted with hydrogen in jet nozzle process?
9. Can one employ distillation methods for the separation of uranium isotopes? Justify
your answer.
10. Explain briefly the Guilder-Sulphide (GS) process in the separation of hydrogen
isotopes?
11. For deuterium enrichment, chemical exchange methods are preferred to distillation
methods. Why?
12. What is the difference between squared-off cascade and tapered cascade? Give an
example of each.
13. Which type of the cascade is preferred for obtaining high recovery of the product:
Simple cascade or counter current cascade? Justify your answer.
14. Calculate the number of stages required to get enrichment of U-235 from its natural
value of 0.77 % to 3% employing gaseous diffusion method.
15. Can a mass spectrometer be used for the separation of isotopes? Justify your answer.
16. What are the limiting factors deciding the maximum enrichment possible in laser
isotope separation of isotopes, even though very large separation factors are possible?