Description
This invention relates to a process for producing cyanuric
chloride by trimerisation of cyanogen chloride at a
temperature of above 200 °C on an activated carbon
catalyst. The process according to the invention also
results in a decreased specific catalyst consumption.
Cyanuric chloride is produced on a large scale by
chlorination of hydrogen cyanide with the formation of
cyanogen chloride and trimerisation of the cyanogen
chloride to form cyanuric chloride - see Ullmann's
Encyclopedia of Industrial Chemistry Vol. A8, 5th ed.
(1987), 196-197. The trimerisation is carried out in the
vapour phase at a temperature of above 2 00 °C, in
particular in the range of about 300 to 450 °C, on an
activated carbon catalyst. During continuous operation, a
temperature profile develops along the longitudinal axis of
the reactor owing to the exothermicity of the trimerisation
reaction; this results in the formation of a so-called hot-
spot, the temperature maximum of which depends on the flow
rate and rises with increasing flow rate. It is known that
the deactivation of the activated carbon catalyst is
influenced by the operating conditions, the flow rate and
the quality of the activated carbon. The deactivation
becomes apparent from the movement of the reaction zone,
and with it the temperature maximum, along the longitudinal
axis of the catalyst.
Owing to its becoming deactivated, the catalyst has to be
exchanged periodically or otherwise activated. The economic
efficiency of the cyanuric chloride process depends
considerably on the service life of the catalyst, as not
only the cost of the catalyst but also the cost of a plant
standstill have to be taken into account. Moreover, with
increasing deactivation of the catalyst, secondary products
such as, for example, cyameluric chloride, are increasingly-
discharged and hence necessitate increased expenditure on
the purification of the cyanuric chloride.
In view of the problems demonstrated, the experts have for
a long time been interested in finding activated carbon
catalysts which have an increased service life and/or in
varying the operating conditions in such a way that the
service life can be increased.
Accordingly, US Patent 3,312,697 discloses a process for
producing cyanuric chloride using an activated carbon
catalyst having a specific surface of above 1000 m2/g, in
which the activated carbon catalyst was activated by a
treatment with acids and/or alkalies and a downstream
washing with water. As a result of the above-mentioned
treatment, inorganic constituents such as oxides,
hydroxides and salts of metals such as Li, Mg, Ce, Ti, V,
Mn, Fe, Ni, Pt, Cu, Zn, Cd, Sn, Pb and Bi, which diminish
the service life of the catalyst, are dissolved out of the
activated carbon. The service life of the catalyst is
further increased in this process by the addition of 0.5 to
10 wt.% chlorine and/or phosgene to the cyanogen chloride.
!In the process according to US Patent 3,707,544, the
service life is increased by mixing the trimerisation
reactor with a mixture of an activated carbon and a solid
diluent having little or no catalytic activity. The
disadvantage of this process is that the space-time yield
is diminished and the expense of disposing of the
deactivated catalyst is increased, above all if the diluent
is a non-combustible material.
In the process described in US Patent 3,867,382, an
untreated activated carbon produced from coconut shells is
used instead of an acid-washed activated carbon. This
activated carbon has an internal surface area of 1200 to
1500 m2/g, a micropore volume of at least 0.7 cm1 /g and an
ash content of below 4 wt.%. Owing to the vegetable origin
of the raw material used for this activated carbon, it has
a low content of heavy metals and an acid wash is rendered
unnecessary. It cannot be inferred from this document how
the micropores are defined, i. e. whether they comprise all
the internal pores, or micropores having precisely defined
limiting values for the pore diameters. A considerable
disadvantage of the activated carbon used in the examples
is that the bulk density, and hence the quantity required
based on the reactor volume, is very high and thus
diminishes the economic efficiency.
In J. Beijing Inst. Chem. Technol. 20 (1993) 1, 55-58, E.
Wang et al. explain that several factors, namely, the ash
content, the iron content, the specific surface and the
pore-size distribution, have to be taken into account when
selecting the catalysts for the cyanogen chloride
trimerisation. The selection of a suitable activated carbon
is complicated by the fact that these factors may mutually
influence one another. It is to be concluded from this
document that it is advantageous to use a carbon which has
as high a specific surface as possible and therefore
contains numerous small pores. The latter help to enable
the reaction to proceed on a relatively large number of
active centres. From the diagrams of the pore-size
distribution of two different activated carbons, it is
suggested that the pores should have a diameter in
particular of less than 2 nm. However, no information can
be drawn from the document as to how the individual factors
influence the service life of the catalyst in a production
plant designed for continuous operation.
Accordingly, the object of the present invention is to
demonstrate an improved process for producing cyanuric
chloride by trimerisation of cyanogen chloride, the
improvement consisting in a decreased specific catalyst
consumption. A further object is to demonstrate the
criteria whereby the person skilled in the art can select
an activated carbon catalyst having an extended service
life for this type of reaction. Other objects can be
inferred from the following description of the process
according to the invention.
A process for producing cyanuric chloride has been found,
comprising trimerisation of cyanogen chloride in the
presence of a washed activated carbon having a BET surface
area of at least 1000 m2/g and an Fe content (calculated as
Fe2O3) of less than 0.15 wt.% at a temperature of at least
250 °C, which is characterised in that an activated carbon
having an effective pore volume Veff of equal to or greater
than 0.17 ml/g is used, Veff being obtained from pores
having a pore diameter in the range of 0.5 to 7 nm. The
subclaims are directed towards preferred embodiments of the
process.
It was found that the trimerisation of cyanogen chloride
proceeds satisfactorily only in those pores having pore
diameters in the range of 0.5 to 7 nm, in particular 0.5 to
5 nm; the pore volume of these pores are to be at least
0.17 ml/g. Although the pore distribution of activated
carbons can differ very widely depending upon the
conditions of their production, the effective pore volumes
Veff necessary for the reaction can be defined from the sum
of a volume increment for the micropores having a pore
diameter of < 2 nm and a volume increment of the mesopores
having a pore diameter of 2 to 30 nm. The effective pore
volume accordingly can be represented as a linear function:
vef£ = a • Vmicro + b ¦ Vmeso. It was also found that the
function
Ve£f = 0.25 ' 0.50 Vmicro + Vmeso is a suitable selection
criterion for an effective activated carbon having a long
service life. The volumes of the micro- and mesopores are
determined as follows:-
The micropore volume is determined from the nitrogen
adsorption isotherm at the temperature of liquid nitrogen
by comparison with a standard isotherm using the t-plot
process of De Boer (cf. De Boer et al. in J. of Colloid and
Interface Science 21, 405-44 (1966)) in accordance with DIN
66135, Part 2 (Version of April 1998).
The mesopore volume and the pore distribution are
determined from the nitrogen desorption isotherm of Barett,
Joyner and Halenda in accordance with DIN 66134 (February
1998). Prior to the measurement, the sample used for the
determination of Vmicro and Vmeso is treated for 1 h at 200 °C
under vacuum (less than 1.3 Pa). The measurement is carried
out, for example, in an "ASAP 24 00" instrument manufactured
by the firm of Micromeritics, Norcross, Ga. (US). The
definition of Vmeso according to the invention includes only
mesopores having a diameter of 2 to 30 nm.
A particularly large increase in the service life of the
activated carbon in this type of process is achieved if Veff
is at least 0.2 ml/g. From an investigation of numerous
different activated carbons, it was found that a maximum
value of the effective pore volume defined above
corresponds to a minimum value of the specific catalyst
consumption. Both extremely mesoporous activated carbons
and extremely microporous activated carbons have too low a
pore volume in the middle pore range*, that is, in the range
between 0.5 and 5 nm, so that the specific catalyst
consumption is considerably higher than in the catalysts to
be used according to the invention.
Another feature of the activated carbons to be used
according to the invention is the specific surface(BET
surface area), which is at least 1000 m2/g/ preferably at
least 1200 m2/g. A high surface area is consequently
advantageous, but is not a criterion which allows a
conclusion regarding the service life of the catalyst.
Thus, different activated carbons having virtually
identical specific surfaces exhibit very large differences
in their rates of deactivation.
In view of the negative influence of a high iron content on
the activated carbon, the iron content, calculated as Fe203,
should be below 0.15 wt.% and preferably around or below
0.1 wt.%. Although an unwashed activated carbon is also
catalytically active, in the process according to the
invention a washed, in particular an acid-washed, activated
carbon is used, because washing is on the one hand a
possible way of decreasing the content of iron and of the
other heavy metals and hence of minimising the formation of
secondary products and, on the other hand, it increases the
pore volume, which is important for the reaction. With
regard to the minimisation of the specific catalyst
consumption, it is moreover advantageous to use a carbon
having a bulk density of equal to or less than 420 g/1.
Where the activity of the activated carbon catalyst is
adequate and the effective pore volume is > 0.17 ml/g,
preferably equal or > 0.20 ml/g, it is advantageous that
the bulk density be as low as possible. In such cases it is
advisable to use an activated carbon having a bulk density
of equal to or < 420 g/1, preferably < 3 90 g/cm3. Figure 1,
which summarises the results of numerous investigations
- see Examples - clearly shows the unforeseen extent to
which the specific catalyst consumption a (kg catalyst per
t of unreacted cyanogen chloride) is dependent on the
effective pore volume defined according to the invention
when a washed activated carbon having a BET surface area of
at least 1000 m2/g and an Fe content of less than 0.15 wt. %
(calculated as Fe203) is used. The specific catalyst
consumption is low, in particular when both the rate of
deactivation (the method of determination may be found in
the Examples) and at the same time the bulk density of the
catalyst are as low as possible.
Examples
The investigations to determine the specific catalyst
consumption in the reaction zone during the trimerisation
of cyanogen chloride to form cyanuric chloride were carried
out in a tubular reactor filled with the activated carbon
catalyst being examined. The tubular reactor was cooled by
means of a heat-transfer medium; the temperature of the
coolant was maintained at 280 °C. The test reactor was
connected parallel to an operating reactor. The gaseous
cyanuric chloride formed was condensed after having left
the reactor and the liquid product was converted into the
solid aggregate state by being sprayed into cooled
chambers.
The ratio of the length of the reactor to the cross-section
of the reactor was 39. During continuous operation, a
temperature profile developed along the longitudinal axis
of the reactor. This profile comprises a heating zone, a
reaction zone and a cooling zone. The maximum of the
reaction zone, the temperature of which rises with
increasing flow rate, moves forward in the direction of the
flow, with increasing deactivation of the catalyst. The
rate of deactivation (udeact) was determined by constructing
time-dependent temperature profiles from temperature-
measuring points arranged along the reactor.
Figure 2 shows that with increasing operating time, the
hot-spot of the reaction zone moves through the complete
set of measuring points arranged one behind the other. The
actual determination of the rate of deactivation was
commenced by a so-called preliminary deactivation of the
catalyst - at that time, the "hot-spot" developed near to
the inlet to the reactor. The preliminary deactivation of
the catalyst lasts for about 12 hours at a flow rate of
cyanogen chloride of 1.1 kg per hour. Figure 2 shows a
typical progression of the deactivation. The rate of
deactivation in cm/t ClCN can be determined from the
distance of the temperature-measuring points and the
average quantity of cyanogen chloride (measured from
maximum to maximum) . The specific catalyst consumption in
the reaction zone can be determined from the rate of
deactivation (vdeact ) , the reactor geometry (cross-sectional
area F) and the bulk density p, in accordance with the
following equation:
Table 2 shows the rate of deactivation u and the specific
catalyst consumption a in the reaction zone using the
activated carbons given in Table 1, the flow rate of ClCN
being 4.4 kg per hour in all the tests.
Temperature of the heat-transfer medium: 280 °C
*) activated carbon catalyst not according to the invention
The tests show that the specific catalyst consumption in
the reaction zone depends considerably on the effective
pore volume and the bulk density of the catalyst. As a
result of a decreased consumption of catalyst, not only is
the cost of the catalyst decreased, but at the same time
the availability of the plant is increased owing to
decreased standstill times and the economic efficiency of
the process is thereby likewise increased.
WE CLAIM:

1. Process for producing cyanuric chloride, comprising
trimerisation of cyanogen chloride in the presence of a
washed activated carbon having a BET surface area of at
least 1000 m2/g and an Fe content of less than
0.15 wt.%,(calculated as Fe2O3) at a temperature of at
least 250 °C,
characterised in that
an activated carbon having an effective pore volume Veff
of equal to or greater than 0.17 ml/g is used, Veff
being obtained from pores having a pore diameter in the
range of 0.5 to 7 nm.
2. Process according to claim 1,

wherein
an activated carbon is used, whose effective pore
volume Veff is formed from the sum
Ve££ = 0.25 . Vmlcro +0.5 Vmeso, Vmicro comprising pores
having a diameter of less than 2 nm and Vmeso comprising
pores having a diameter of 2 to 30 nm.
3. Process according to claim 1 or 2,
wherein
Veff of the activated carbon used is at least 0.2 ml/g.
4. Process according to one of claims 1 to 3,

wherein
the activated carbon to be used has a bulk density of
equal to or less than 420 g/1.
5. Process according to one of claims 1 to 4,
wherein
the activated carbon to be used has a BET surface area
of at least 1200 m2/g and Ve£f is at least 0.2 ml/g.

The invention relates to a process for producing cyanuric
chloride by trimerisation of cyanogen chloride at a
temperature of at least 250 °C on a washed activated carbon
as catalyst. The service life of the catalyst can be
increased by using an activated carbon having an effective
pore volume Veff of equal to or greater than 0.17 ml/g, Veff
being obtained from pores having a pore diameter in the
range of 0.5 to 7 nm.