A review on the oldest known gas-hydrate – the chemistry of chlorine hydrate

Full text

ABSTRACT
In the last century a lot of new findings were
reported in the field of chlorine hydrate - the
oldest known compound belonging to the so
called gas hydrates. Chlorine hydrate was a very
interesting compound which encages an elemental
chlorine molecule in a water clathrate cage
causing changes in the reactivity of the chlorine
and ensuring storage in solid phase at room
temperature even at mild overpressures (6 atm).
Chlorine hydrate, however, decomposes at
atmospheric pressure around 8°C. It therefore
acts as a solid phase carrier for chlorination with
gaseous chlorine in aqueous medium, e.g. as
battery chlorine electrode material, chlorinating
and disinfecting agents etc. This review
summarizes the investigation results from the
oldest publication of Davy in 1811 up to the
beginning of the decade, and some unpublished
new results as well.
KEYWORDS: chlorine hydrate, non-stoichiometric
clathrate hydrates
1. INTRODUCTION
The chlorine hydrate - a non-stoichiometric clathrate
compound having elemental chlorine molecule
A review on the oldest known gas-hydrate - The chemistry of
chlorine hydrate
encaged in water molecules - was discovered a
little over 200 years ago by Humphry Davy when
he investigated the so-called “oxymuriatic acid
gas” (chlorine) formed by the action of hydrochloric
acid on pyrolusite (MnO2) [1]. Although the
formation of “oxymuriatic gas” (chlorine) was first
observed by Scheele in 1774, who believed that
the gas was a mixture containing oxygen [2], Gay
Lussac and Thénard considered the possibility that
it might be an element, after their attempt to
separate the oxygen by reacting it with charcoal
failed [3, 4]. Davy repeated the experiment in 1811
and declared it an element naming it chlorine.
Since Davy cooled the gas over an aqueous
solution, he was credited with producing the first
synthetic solid gas hydrate at 10°C. Ever since,
chlorine hydrates, popularly called “chlorine ice”
has been the subject of consistent investigations.
Davy reported that the aqueous solution of
chlorine freezes more readily than pure water,
but the pure gas dried over calcium chloride
underwent no change whatever, at temperatures
below -30°C. The intensive development of the
chemistry of chlorine hydrate started from this
point.
2. General description and properties
Thénard [4] described chlorine hydrate as a solid
formed during the cooling of the aqueous solution
of chlorine. The bright yellow coloured crystals
liquefied at slightly increased temperature releasing
1Chemical Research Center, Hungarian Academy of Sciences, Pusztaszeri u. 59-67, Budapest,
H-1025, 2Deuton-X Ltd., Selmeci u. 89., Érd, H-2030, 3Axial-Chem Ltd., Sajóbábony-Gyártelep,
H-3792, Hungary. 4Department of Chemistry, J. N. V. University, Jodhpur 342 005, Rajasthan, India
László Kótai1,2,*,#, István Gács1, Szabolcs Bálint1, Györgyi Lakatos2, Andras Angyal3
and Raj N. Mehrotra4,*,$
*Corresponding authors
# kotail@chemres.hu
$raj_n_mehrotra@yahoo.co.uk
Trends in
Inorganic
Chemistry
Vol. 13, 2012

excess water was placed into one of the arms of a
closed large V-shaped tube which was then heated
to decompose the hydrate. The liquefied chlorine
condensed in the other arm which was kept cold.
Re-formation of a small amount of hydrate in
water could be observed after slow cooling for
24 h. The bulk of the liquefied chlorine remained
in the arm where it had condensed. The small
crystals of hydrate produced in the middle of the
water had a shape of grouped in clusters or
feather-like fern leaves. The materials inside the
tube changed slowly when the branch containing
the mixture of water and hydrate was immersed in
a vase filled with water, while the other part
containing the liquid chlorine was left exposed to
the ambient air. The amount of the hydrate formed
in the water first increased and then after about a
month the hydrate formed cakes and a kind of
crystallised membrane on the surface of the
liquid, twisted and folded. A certain amount of the
hydrate had deposited on the walls of the curved
portion of the V tube. The reason for the
appearance of the hydrate at the curved part
between the two layers of the liquid chlorine was
that this part was in contact with the atmosphere
of chlorine gas at temperatures between 8 and
12°C. Under these conditions, the hydrate
deposition had spread over a length of the tube
(0.20-0.25 part) and the fern leaf gradually
changed into isolated crystals. Most of these
reached a size of >3 mm in length by the end of
one year.
All the crystals, smaller and larger, were found
to be crisp and perfectly transparent, and had
dark greenish yellow colour. The crystals were
relatively refractive because these showed some
iridescent colours when viewed in sunlight; their
34 László Kótai et al.
the chlorine gas in abundance. Faraday studied the
chlorine hydrate in detail, and determined not only
its composition (see section 5) but studied its
properties as well [5, 6, 7]. Although the dry
chlorine hydrate has high chlorine content, it has a
much weaker chlorine smell than a 0.7 per cent
aqueous chlorine solution. This was explained by
the reduced vapour pressure of the chlorine
captured in the hydrate state [8].
Chlorine hydrate was formed as crystallised crust
or dendritic crystalline mass when chlorine gas
was introduced in less than the sufficient amount
of water to convert it into hydrate [5]. Crystals left
below 0°C for a few days sublimed with the
formation of large brilliant crystals which are
delicate prismatic needles extending from 1.25 cm
to 5 cm. The shorter ones were also formed as
flattened octahedral, and the three axes of the
octahedron had different dimensions [5].
Biewend observed formation of chlorine hydrate
from water and chlorine gas at around 0°C as thin
slurry of lamellar crystals [9]. These crystals
decomposed on heating. The chlorine hydrate
formed from liquid chlorine and water at 0°C,
formed feather-like crystals [9]; however, at room
temperature large beautiful crystals were obtained
[9, 10].
Nordenskjöld left a chlorine hydrate filled bottle
on a cold window during a winter period (when
high temperature gradient takes place between the
parts of the bottle). The chlorine hydrate then
gradually deposited well-framed crystals on the
cooler part of the bottle by slow sublimation.
These crystals belonged to the rhombic system,
a:b:c = 1:07038:0.6924. Beside the rhombic crystals,
there were other crystalline forms. Nordenskjöld
found the following crystallographic parameters:
b:x = 134o32’, x:x1 = 90o56’, a:y = 124o42’, and
y:y1 = 110o36’ [11]. The shape of the formed
crystals can be seen in Figure 1.
Rammelsberg repeated Nordenskjöld experiments
and studied the large chlorine hydrate crystals
under polarized light using a microscope. These
crystals were orthorhombic or clinorhombic [12].
The crystallization process of the chlorine hydrate
was studied by Ditte [13] in detail using the
Faraday method [5-7] to prepare liquid chlorine
from the hydrate. Chlorine hydrate containing
Figure 1. The shape of chlorine hydrate crystals found
by Nordenskjöld [11].

Chemistry of chlorine hydrate 35
only the hydrate was formed and dissolved in
water, but above this temperature at normal
pressure only the dissolution of the chlorine gas in
water was observed [20, 21].
The numerical solubility values of the chlorine
hydrate were determined by Winkler [22] and
Roozeboom [20] as a function of temperature.
These results are given in Table 1.
Winkler found [22] that the dissolved amount of
chlorine did not change with the chlorine pressure
above the hydrate crystal containing chlorine
solutions. Thus, the chlorine did not follow the
Henry Dalton law between 0 and 9°C if chlorine
hydrate was present in a closed system. In this
system the excess of the chlorine hydrate
dissolved and decomposed when the pressure was
decreased to supply the solution with fresh
chlorine. On the other hand, chlorine hydrate
crystallised out when the pressure was increased
(see section 3). It meant that the dissolved
shape seemed to have been derived from the
regular system. Perfect octahedral or other
modification of the crystals where the general
shapes have been altered could be observed.
These crystals are plates with very developed
appearance of both sides, which are completed
with a regular hexagonal contour [13].
Depending on the synthesis conditions, Maumene
[14] found cubic chlorine hydrate crystals
prepared directly by introducing chlorine gas into
water or a mixture of liquefied chlorine and water.
However, octahedral but orthorhombic chlorine
hydrate crystals were formed when a warm
chlorine saturated solution was left to evaporate
and condense into a cold part of the apparatus
(reaction of chlorine gas and water vapour). The
faces of these crystals showed that these
developed very irregularly. However, the hydrate
was formed with variable composition when a
warm solution of dissolved chlorine was placed in
a large branch of a tube and was left to escape
slowly into the cooled small branch of the same
tube [14].
Chlorine hydrate induces the crystallizations of
cubic SO2.6H2O, CH3Cl.6H2O and the crystallization
of the chlorine hydrate at 0°C [15]. This shows
their isomorphous character. The polarized light
does not act on chlorine hydrate crystals, but other
crystalline forms often appear which could not be
observed in the case of other gas hydrates [16].
2.1. Solubility of chlorine hydrate
Faraday observed that during the preparation of
chlorine hydrate by cooling the chlorine solution,
a portion of the chlorine remained in the solution
and the crystals formed had dissolved slowly as
well. The latter could be attributed to the higher
solubility of the chlorine hydrate compared to that
of the chlorine [5]. Gay-Lussac [17] observed an
inflection point on the solubility curve of chlorine
at 8°C. This indicated increased dissolution of the
chlorine hydrate in water with rising temperature.
However, above 8°C (the upper limit for the
existence of the hydrate at atmospheric pressure),
it was only the chlorine that dissolved in water
and the solubility decreased with increasing
temperature and became zero at 100°C [18]. This
justified the observations of Isambert [19] and
Roozeboom [20]. At temperatures lower than 9°C
Table 1. The solubility (g/100 g of water) of chlorine
and chlorine hydrate.
Temperature C° Solubility
[22]
g/100gH2O
[20, 21]
0 0.5598 0.507
1 0.6017 -
2 0.6444 -
3 0.6878 0.615
4 0.7320 -
5 0.7770 -
6 0.8228 0.714
7 0.8694 -
8 0.9167 -
9 0.9648 0.908
12.5 1.112 -
20 1.853 -
28.5 3.627 -
Cl2 gas solubility g/100 g of H2O [23]
10.00 1.00
20.00 0.72
25.00 0.64

36 László Kótai et al.
d = 1.220 and 1.237 g cm-3. Based on these
values, the calculated formula [23] proved to
be Cl2.7.24H2O and Cl2.6.97H2O respectively.
This indicated that the water content might have
been less than what Roozeboom had observed
experimentally [20].
The chlorine hydrate prepared in zinc chloride
solutions can be compacted and stabilised with
60-80 atm pressure. If the pressure is less than
60 atm, the density falls below 1 g cm-3, but at
70 atm the value increases to 1.1-1.15 g cm-3 and
becomes 1.1-1.2 g cm-3 at 80 atm. Above 800 atm
the value of the density was d = 1.21-1.22 g cm-3.
Further increase in the pressure does not result in
significant increase in the density value [25]. The
hydrate prepared in water has a higher chlorine
content and its density at 800 atm was found to
be d = 1.23 g cm-3 [25].
2.3. Stability and decomposition of
chlorine hydrate
The stability of the hydrate was strongly dependent
on its composition, the more chlorine fills the
cavities, the more stable was the hydrate formed
[23]. The hydrate was a thermally unstable
compound, it has an upper temperature existence
limit (incongruent melting, see section 3). The
chlorine molecule in the hydrate behaves as
elementary chlorine and was therefore sensitive
to hydrolysis which was an equilibrium reaction
(see section 2.5.1). One of its hydrolysis products
was the unstable hypochlorous acid which easily
decomposed thermally or by light. Therefore
the equilibrium can be shifted to complete
decomposition.
2.3.1. Stability of the chlorine hydrate
The hydrate decomposed when it was in contact
with metal chloride solutions at lower temperatures
than in pure water [25]. The stability can be
increased easily by compressing the hydrate at
high pressures. At around 115 atm not only the
bulk density increased but the stability as well.
A body of chlorine hydrate (86 cm of diameter,
1.6 cm of height and 61 cm2 of surface area)
completely decomposed at room temperature in
6 hours. The decomposition rate was not completely
linear: roughly 50 or 60% of the decomposition
took place after 2.5 or 3.5 h, respectively [25].
chlorine content (between 0 and 9°C) can be
considered to be the chlorine content of the
dissolved chlorine hydrate. The solubility of the
chlorine hydrate in water was expressed by
equation 1 [22].
s = 0.5598 + 0.04154t + 0.0003842t2 (1)
where t is the temperature in °C, s is the solubility
in g of chlorine hydrate in 100 g of water.
2.2. Density of chlorine hydrate
Faraday, by floating the hydrate crystals over
calcium chloride solutions of various concentrations,
found that the density of the plate like hydrate was
d = 1.2 g cm-3 [5]. Since he observed continuous
slight liberation of gas, and the imperceptible
bubbles could have adhered to the hydrate surface,
therefore, the density could be presumed to be less
than d = 1.2 g cm-3. Using the methods of Kótai
et al. [23], the formula of these type of chlorine
hydrate crystals prepared by Faraday should
have Cl2 < 7.63H2O. The highest experimentally
determined density (1.31 g cm-3) of the isolated
chlorine hydrate was reported by Smirnov and
Kleshchunov [24]. This density provided the
composition of the hydrate to be Cl2.6.01H2O
[23], which was close to the Cl2.6.2H2O obtained
by the authors [24].
In capillaries the density of the hydrate showed a
value of 1.37 g cm-3. This value was too large
even in case of complete occupation (8Cl2.46H2O)
which would give d = 1.325 g cm-3. It means that
the lattice of the chlorine hydrate in the capillaries
is to be contracted (a = 1.192 Å) or the difference
between the molar volumes of the empty clathrate
cage and the ice should be diminished from the
average 3.1 to 2.5 cm3/mol (see section 5.2.3)
[23]. These differences in the lattice constant
and molar volume values are too high. No example
for this magnitude of contraction among the
clathrates, even at higher pressures than the
observed 10-12 atm in the capillaries, could be
found. The perfect filling of the cavities should
cause dilatation rather than contraction. Therefore,
the hydrate formed in the capillaries probably
does not belong to the clathrate type [23].
Roozeboom [20] experimentally determined the
density of two samples of crystalline hydrate to be

Chemistry of chlorine hydrate 37
Hence, the main chemical reaction of the hydrate
was definitely the same as the chemical reactions
of the chlorine in the presence of water (aqueous
solution).
2.4.1. Decomposition by hydrolysis
Faraday had observed some negative effect of
the light on the yield of the hydrate [5]. At
atmospheric pressure the hydrate was partially
decomposed by light with the evolution of some
oxygen [10]. There was no diminution in volume
of the liquid chlorine when the hydrate in a sealed
tube was exposed to sunlight for a whole day and
then heated to decompose into water and liquid
chlorine. When the tube was brought back to
ordinary temperatures the liquid chlorine gradually
reunited with the water and crystalline hydrate
was formed [10].
Normally, the dissociation of the chlorine
hydrate in dilute solutions was represented by
equation (2).
Cl2(aq) → (H+ + Cl− + HClO)aq (2)
The electrical conductivity of such solutions at
0°C was equal to that of a water solution
containing the same amount of hydrogen chloride,
because hypochlorous acid was not considered to
be an electrolyte. In more concentrated solution,
the conductivity of the chlorine was much less.
Since the dissociation of the chlorine takes place
with the absorption of heat, dissociation increases
with the temperature. The addition of 0.1-0.5 M
solution of hydrochloric acid prevents the formation
of chlorine hydrate. An addition of 1-2 M HCl,
however, induces a stronger absorption of the
chlorine on account of the formation of HCl3
[26, 27].
2.4.2. Reactions with inorganic substances
Chlorine hydrate reacts with ammonia or
ammonium salts with the formation of nitrogen
gas, hydrochloric acid and a small amount of
nitrogen chlorides [5]. Faraday also observed that
the amount of nitrogen chlorides are less in the
reaction of ammonia than in the reaction of
ammonium salts [5].
The hydrate reacted with elementary mercury
with the formation of a large amount of HgCl2
[8, 28] due to the chlorination effect of the highly
2.3.2. Thermal decomposition of the hydrate
Solid and dried hydrate decomposes on heating to
liquid chlorine containing dissolved water and
liquid water containing dissolved chlorine as was
first observed by Faraday [5-7] and studied by
Biewend [9]. Faraday heated the chlorine hydrate
in a closed ampoule up to 37°C when a bright
yellow gas formed together with two fluid phases,
among which the lower phase was the liquid
chlorine, the amount of which increased as the
mixture was cooled [6, 7]. The reaction was found
to be reversible [6, 7, 9]. The first experiment by
using the hydrate as a chlorine source was
performed in 1838, when Biewend made chlorine
hydrate from water and chlorine gas at 0°C in a
sealed tube, then warmed the tube to 30-40°C
[9]. During the effervescent melting of the
chlorine hydrate, the empty part of the tube was
filled with the dark yellow chlorine gas, while the
crystal scales dissolved with the formation of
dark yellow oily drops of anhydrous condensed
chlorine falling to the bottom as a yellowish
liquid. The chlorine separated from the water
combined again with the water below 0oC to form
chlorine hydrate to a small extent (if mixing was
avoided). This hydrate formed fine feather-shaped
material, thus the remaining chlorine was
enclosed by water. Crystalline chlorine hydrate
can be obtained at ordinary temperatures by using
the same glass tube without heating. Only a small
portion of the chlorine hydrate melts, the rest was
gradually produced as beautiful crystals [9].
The results were confirmed by Wöhler [10], who
observed that in a sealed glass tube the hydrate
maintained itself even at summer temperature due
to the overpressure of the compressed chlorine
liberated by the partial decomposition of the
hydrate. If the tube containing this mixture was
immersed into water at 30 to 40oC, the hydrate
decomposed into water, and the liquid chlorine
was sedimented. If the tube was placed back to
normal air temperature, the hydrate was gradually
regenerated. Even summer turns the liquid chlorine
into crystallised hydrate again [10].
2.4. Chemical reactions of the chlorine hydrate
Depending on the pressure and temperature
relationship, the hydrate decomposes into chlorine
in water or liquefied chlorine containing water.

38 László Kótai et al.
However, it was reformed under pressure at
temperatures higher than 9°C as a ball-like solid
even at 13 or 14°C [19]. Therefore the knowledge
of its dissociation pressure as a function of
temperature becomes essential [19].
3.1. Experimental dissociation pressure of the
chlorine hydrate
The dissociation pressure of the chlorine hydrate
was measured first by Isambert [19], later by
Le Chatelier [34], Roozeboom [20], Bouzat [35],
Tamman and Krige [36] and Bozzo et al. [37, 38,
39], whereas Wilms and Van Haute studied the
decomposition process in detail [40, 41].
3.1.1. The dissociation in pure water
Le Chatelier measured the dissociation tension
of the hydrate around the lower quadruple
point (between -14 and +9°C). During gradual
cooling of the hydrate in contact with chlorine
atmosphere, the tension decreased simultaneously
with the temperatures varied from one experiment
to another but in general it was between -4 and
-7°C. At this point, pressure of the chlorine
increased sharply then dropped quickly and
became constant. Further cooling caused steady
fall in the pressure. This phenomenon resulted in
the sudden freezing of the water that has been
in super cooled condition. The temperature rose
momentarily due to this change of state and
brought a corresponding increase in the pressure.
When the temperature equilibrium was restored,
the tension of the chlorine resumed its normal
value corresponding to the dissociation of the
hydrate with the formation of solid water. However,
this pressure was higher than the corresponding
value in case of formation of water [34]. Roozeboom
also observed the supercooling phenomenon in
a HCl free system, but the lowest temperature
reached was only -3°C [42].
During heating of the hydrate, the pressure
increased steadily until about -1°C and the
temperature near to this point remained stationary
during the melting. No change was observed in
the pressure during this process, indicating that
the transition at the equilibrium temperature did
not lead to appreciable change in either of the
dissociation or the vapour pressures [34]. The
dissociation pressure values are given in Table 2.
condensed chlorine toward primarily formed
mercurous chloride (Hg2Cl2) [28].
Aqueous solution of the hydrate (which was more
concentrated than the chlorine solution) reacted
easily with zinc oxide, equation (3), and zinc
ferrite, equation (4). However, it did not react with
Fe2O3, Al2O3 or MnO2 [29].
2ZnO + 2Cl2.nH2O = 2ZnCl2 + O2 + 2nH2O (3)
2ZnFe2O4 + 2Cl2.nH2O = 2ZnCl2 + 2Fe2O3
+ O2 + 2nH2O (4)
In contrast with Fe2O3, FeO easily reacted with
the hydrate forming FeCl2 [30]. The free enthalpy
change in this reaction was negative (see section
7.2.2), and the primarily formed FeCl2 was easily
chlorinated into iron(III) chloride as in equation
(5).
2FeCl2 + Cl2.nH2O = 2FeCl3 + nH2O (5)
Metallic aluminium and elementary zinc are
oxidized into aluminium oxide, equation (6), and
zinc(II) chloride as in equation (7) [29].
2Al + 3Cl2.nH2O = Al2O3 + 3(n-1)H2O + 6HCl (6)
Zn + Cl2.nH2O= ZnCl2 + nH2O (7)
These reactions proceed easily between 0 and 9°C
in the existence range of the hydrate solution [29].
Its reaction with iodides liberates a quantitative
amount of elementary iodine which can be used
for the determination of the chlorine content of
the hydrate [20, 31, 32, 33].
2.4.3. Reaction with organic substances
Chlorine hydrate reacts with ethyl alcohol when
the temperature of the reaction mixtures is raised
up to 45-65°C. Diethyl ether and hydrochloric acid
are formed together with some chlorinated
hydrocarbon compound. The reaction was strongly
exothermic [5].
3. Phase relations of the chlorine hydrate
The chlorine hydrate neither decomposes nor
dissolves completely below 9°C in a closed tube
at normal pressure. It was, however, destroyed
at ~9°C when the tension of the dissociation
was roughly equal to the atmospheric pressure.

Chemistry of chlorine hydrate 39
3.1.2. The dissociation in solutions
The dissociation pressure of the hydrate in contact
with various solutions containing dissolved salts
or acids changes rapidly [44]. The dissociation
pressure of the hydrate in aqueous NaCl (100 g dm-3)
and HCl (g dm-3) solutions are shown in Table 3.
It can be seen that a dissociation pressure of
770 mm of Hg was reached in water at 9.8°C, and
at 7°C in NaCl solution with a concentration of
100 g dm-3 and at 3.8oC in the more concentrated
(200 g dm-3) NaCl solution [44]. Thus, the
dissociation pressure of the hydrates formed
or kept in these solutions reached the outer
environmental pressure at lower temperature than
in water, which was consistent with the experiences
of Goodwin [45] about the destabilising effect of
metal chloride solutions on the chlorine hydrate.
Similarly, the inhibiting effect of HCl on the
hydrate formation [26, 27] can be the consequence
of the increased vapour pressure and the
decomposition of the hydrate at lower temperature
in HCl solutions [44].
The dissociation pressure of the chlorine hydrate
above 0°C was measured in detail by Roozeboom
[20], who determined the critical decomposition
temperature (9.6°C) in an open vessel [20, 21],
and pointed out that the hydrate can be stabilised
at higher temperatures by pressure. However,
the highest value of temperature that could be
achieved was when the pressures of the hydrate
and the liquid chlorine saturated with water have
become equal with each other [20].
Roozeboom also determined the lowest angular
point as quadruple point at -0.24°C, (Le Chatelier
The analytical expression (see equation 8) for the
measured values was derived by Kass [39], Wilms
and Van Haute [38, 40].
log P = A - B/T (8)
where P was pressure in mm of Hg and T was
in K. The values of coefficients (A = 16.67 and
B = 3899.4) are valid for the entire range of the
measurements [38, 40]:
Cl2.nH2O(s) → Cl2(g) + nH2O(l) (9)
The values of A = 17.78 and B = 4213.7 were
reported for the process described in equation (9)
from 1 atm to the upper quadruple point [38, 40],
whereas the values by Iskenderov and Musaev
were 16.57 and 3834 respectively for A and B
[43].
The log p-1/T relationship for the solubility of
chlorine in water and the vapour pressure of the
water above the solution in contact with the
solid hydrate, as given in (equation 9), was not
completely linear. Therefore, Wilms and Van
Haute suggested another relation expressed in
equation (10) [38, 40].
log p = -107.5476 + 1325.6/T + 43.1318 log T
(10)
Table 2. Experimental dissociation pressure values*
of the hydrate in the region of supercooling.
T, °C Pressure, Torr
water Ice
-1 290 290
-2 230 -
-3 210 -
-3.5 - 262
-4 205 -
-5 146 -
-6 153 -
-7 - 230
-14 - 175
*The vapour pressure data in this temperature range
were measured in water containing HCl and the curves
for ice and water cut each other around -1°C. Thus, at
this point the hydrate, ice, water and chlorine gas are in
equilibrium.
Table 3. The dissociation pressures (in Hgmm) of
the chlorine hydrate in water, 100 g dm-3 NaCl and
200 g dm-3 HCl solutions.
T, °C H2O NaCl
100g dm-3 HCl
36.5 g dm-3
0 247 375 340
2 307 452 405
4 395 570 500
6 495 615 -
7 - 770 -
9.8 770 - -

by Kótai et al. [47]. The complete phase diagram
contains a new quadruple point and was drawn by
using all of the available dissociation pressure
data including the hydrate tension in the presence
of super cooled water. The complete p-T phase
diagram of the hydrate-water-chlorine system is
shown in Figure 2.
There are four coexisting phases in equilibrium
with each other at the A, B and C invariant points.
Point A represents the decomposition temperature
of the hydrate, or the critical temperature
according to Stackelberg [48, 49], or the upper
temperature of existence according to Van der
Waals and Platteeuw [50] if dT/dp < 0. In this
case (dT/dp > 0) the meaning of the two previous
statements can not be the same, because the
melting point of the hydrate increases with
increasing pressure [47]. This phenomenon,
however, might be the result of increasing
occupation and stabilising effect of the chlorine
capped in the cavities (increasing pressure
increases the amount of the capped chlorine in the
hydrate [51, 52]). At point A the four phases in
equilibrium with each other are the hydrate, the
solution of water in liquid chlorine, the solution of
chlorine in water, and the vapour [21, 24, 37, 39,
46, 47, 53, 54].
Point B represents eutectic point of the hydrate,
ice, water containing dissolved chlorine, and the
vapour phases which are in equilibrium with each
other [21, 24, 46, 55].
40 László Kótai et al.
had found -1°C in the HCl containing system
[34]) and the highest quadruple point was given as
28.7°C [21, 42].
3.1.3. Dissociation into ice
Decomposition of solid hydrate into ice and
chlorine gas was evaluated as an univariant solid
hydrate → gas + liquid type system and can be
characterised by the law that the ratio of the
absolute temperatures at two different pressures
was constant. Calculated from the p-T data at 300
and 900 Hg mm, a value of 1.032 for the hydrate
was obtained [35].
The dissociation pressure in the chlorine gas-ice-
chlorine hydrate system was described by Tamman
and Krige and is given in equation (11) [36]:
T = 49.0 (log p + 0.4707) (11)
where T is the temperature in °C, and p is the
pressure in atm. The reaction can be expressed by
equation (12).
Cl2.nH2O(s) → Cl2(g) + nH2O(s) (12)
The coefficients A and B of equation (8)
calculated from the values of Roozeboom were
7.684 and 1446.6 respectively [38, 40]. However,
Wilms and Van Haute reported values of 8.22 and
1595.2 for A and B respectively [38, 40].
Kótai et al. suggested the equation (13) where p
is the pressure in kPa and T is the absolute
temperature and the coefficients A* and B* had
the values 11.43 and 0.046 respectively.
lnP = A* - B*/T (13)
3.2. The phase diagram of the chlorine hydrate
The phase diagram of the hydrate was partially
drawn and evaluated by Le Chatelier [34]
and Roozeboom [21, 46]. They interpreted the
decomposition processes (equations 9 and 12) and
gave the parameters of two invariant points. The
phase diagrams of the chlorine hydrate-aqueous
metal salt solution-chlorine systems are very
similar; the invariant points, however, are
markedly changed (see section 3.2.2).
3.2.1. The phase diagram of the chlorine-water-
hydrate system
The phase diagram of the chlorine hydrate was
evaluated by Roozeboom [21, 46] and supplemented
Figure 2. The phase diagram of the chlorine hydrate-
water-chlorine system.

Chemistry of chlorine hydrate 41
are a bit different from the values given in Table 4.
The values of the parameters A and B, both in
temperature and pressure, corresponding to
various n values given in Table 5 were derived
from experimental densities [23].
Curve AB (line 2 corresponding to the temperature
dependence of the dissociation of the saturated
vapour (dissociation) pressure of the hydrate)
shows the experimental equilibrium of the the
hydrate in the presence of coexisting solution and
Cl2 gas. Point B belongs to the hydrate, and
continuing the cooling along the BD (BC) curve.
Ice could be formed below 0°C. Curve BC (line 3)
represents the equilibrium of the hydrate in the
presence of ice and gaseous Cl2, LBA represents a
supercooled solution and the presence of Cl2 gas
[47].
During decomposition of the hydrate by heating at
a pressure above the critical decomposition point
two liquid phases (Cl2 and H2O) are formed
increasing the volume. This corresponds to a
positive slope of the melting pressure curve AI
(Figure 2) i.e., the decomposition temperature (of
the incongruent melting point) was increasing
with increasing pressure [53]. Declination of line
4 was determined by Byk and Fomina [54]. The
dT/dp value for the hydrate was found to be
0.80×10-2 °C atm-1 [54].
Curve BF represents the pressure of Cl2 gas above
a solution that was in equilibrium with the ice, or
in other words, line 6 shows the change in the
freezing point of the water in the presence of
dissolved chlorine [21, 46, 47]. The concentration
of chlorine in this solution decreases regularly
from B to F. Therefore F is the melting point of
the ice that has zero concentration of Cl2 gas
(p = 0).
Invariant quadruple point C represents the
equilibrium between chlorine gas, liquid chlorine,
ice and the hydrate. The parameters of point C
were calculated by extrapolation of the saturated
vapour pressure curve of the chlorine and the
dissociation pressure curve of the hydrate (dotted
lines in Figure 2) [47]. There is no equilibrium
between the ice, chlorine gas and its aqueous
solution because the equilibrium curve should
start at the point for which the heat of
transformation, ∆H, would pass through zero and
would become negative [21, 46]. Barrer and Edge
experimentally confirmed the possibility of the
formation of the hydrate at 90 K via the reaction
of solid ice and condensed phase chlorine [56].
This invariant point corresponds to this
equilibrium. Parameters of the invariant point are
given in Table 4.
These invariant points belong to the hydrate at a
given composition, thus on the phase diagram of
p-T-x (x means the chlorine mole fraction in the
hydrate x = 1/(1+n)) the curve AB (AC) shifts to
the left with increasing x, consequently, with
increasing water content (decreasing the chlorine
content), the temperatures and pressures of the
invariant points are decreased [40, 41]. The same
situation was observed in the case of aqueous
solutions kept in equilibrium with the chlorine
hydrate [37, 39] (section 3.2.2). This result easily
explains the results of Bjorkman [57] about
the stability of dense hydrate, because during
densification the chlorine content of the hydrate
increased [57].
For example, the parameters of the invariant
points of the hydrate with a composition of
Cl2.6.2H2O [24] (the calculated composition from
the density value was Cl2.6.01H2O [23]) or the
parameters of some other chlorine hydrates
without the knowledge of the exact composition
Table 4. The parameters of the invariant quadruple
point of the chlorine hydrate-water-chlorine system.
Invariant
point T, K P, kPa Ref.
A 301.85 607.95 [47]
B 272.91 32.53 [47]
C 173.16 0.322 [47]
Table 5. Invariant point parameters for various
hydrates.
n Temperature,°C Pressure, bar
A B A B
6.01[24] 28.0 +0.22 8.3 0.35
6.97[20] 28.7 -0.24 6.08 0.33
6.20[37] 28.3 -0.24 8.52 0.32
6.97[40] 28.3 -0.22 8.42 0.32

concentration [58]. The dependence of the T*C
values on the concentration is shown in Table 6
[59].
The Karsten’s formula, equation (15), could be
used for the calculation of the freezing point of
sodium chloride solutions, Tf(NaCl), at different
concentrations W expressed in % (w/w) [59].
Tf(NaCl) = 0.0022244W2 - 0.7663855W (15)
The correlation of the critical temperature Tc and
the concentration of NaCl is expressed by
equation (16).
Tc = 28.3 - 0.4324W - 0.0111W2 (16)
However, for the concentration of the NaCl
between 0-15 %w/w the correlation is expressed
by equation (17) [58].
Tc = 28 - 0.4780W - 0.0209W2 (17)
The dependence of the ∆Tc, Tf and T*c on the
concentration of the sodium chloride in solution is
shown in Table 6.
3.2.2.1. The aqueous sodium chloride-chlorine-the
hydrate system
The formation of the chlorine hydrate in sodium
chloride solution has been studied in detail
[37, 38, 39, 58]. The pressure-temperature-salt
concentration relation-ships are evaluated between
1-8 atm and from 0-15% NaCl concentration in
the temperature range of 12-24°C [38].
By the addition of sodium chloride to water in
contact with the hydrate, line AB of the phase
diagram shifts to the left so that the boundaries of
Table 6. Relationships between the freezing points
of the sodium chloride solutions and the T*C values
calculated from the average TC values of the hydrate,
in °C.
WNaCl,
%w/w ∆TC Tf T*C
2 0.95 1.52 0.57
4 1.95 3.00 1.05
6 3.05 4.51 1.46
8 4.25 5.98 1.73
10 5.50 7.44 1.94
15 9.25 10.80 1.55
42 László Kótai et al.
The heat of transformation reduces from F to B,
and at point F (0°C) it is the heat of fusion of the
pure ice. Beyond this point it was reduced by the
absorption heat of Cl2 which increases as B was
being approached. There is no doubt that the heat
of transformation of the ice was still positive close
to point B. However, its value was low due to the
small amount of chlorine contained in the solution
at the inter