WO1996030109A1

WO1996030109A1 - Refrigerant separation

Info

Publication number: WO1996030109A1
Application number: PCT/AU1996/000163
Authority: WO
Inventors: Neville Charles Stephenson; Barry Branscombe Walker; Colin William Spencer
Original assignee: Colcard Pty. Limited
Priority date: 1995-03-24
Filing date: 1996-03-22
Publication date: 1996-10-03
Also published as: AUPN195295A0

Abstract

This invention relates to a method for separating a chlorofluorocarbon, a hydrochlorofluorocarbon or a hydrofluorocarbon from a gaseous mixture of chlorofluorocarbons, hydrochlorofluorocarbons and/or hydrofluorocarbons comprising passing the mixture through a bed of adsorbent (2), the adsorbent being selected so as to preferentially adsorb the chlorofluorocarbon, hydrochlorofluorocarbon or hydrofluorocarbon from the mixture to thereby produce a gaseous mixture depleted of the chlorofluorocarbon, hydrochlorofluorocarbon or hydrofluorocarbon, optionally followed by the step of desorbing or recovering the adsorbed chlorofluorocarbon, hydrochlorofluorocarbon or hydrofluorocarbon.

Description

TITLE: REFRIGERANT SEPARATION

FIELD OF THE INVENTION

This invention relates to the separation of chlorofluorocarbons (CFC^*s) and

hydrochlorofluorocarbons (HCFC's) and hydrofluorocarbons (HFC's) from mixtures

of various components using selective adsorbents.

BACKGROUND OF THE INVENTION

Certain refrigerant mixes are difficult to separate. For example, the boiling

point of R12 (CC1₂F₂) is -29.8°C at atmospheric pressure and the boiling point of R22

(CHC1F₂) is -40.8°C. An azeotrope forms with composition 97.9% R12 and 2.1 %

R22 by weight and complete separation of the two by distillation is extremely

difficult. Further, fractionating columns typically required for separation are long and

expensive. Accordingly, mixtures of R12 and R22 are usually regarded as

"unprocessable". DISCLOSURE OF THE INVENTION

There is therefore a need for a method for separating chlorofluorocarbons

and/or hydrochlorofluorocarbons from mixtures thereof.

According to a broad aspect of the present invention there is provided a

method for separating a chlorofluorocarbon, a hydrochlorofluorocarbon or a

hydrofluorocarbon from a gaseous mixture of chlorofluorocarbons,

hydrochlorofluorocarbons and/or hydrofluorocarbons comprising the steps of passing

the mixture through a bed of adsorbent, the adsorbent being selected so as to

preferentially adsorb the chlorofluorocarbon, hydrochlorofluorocarbon or

hydrofluorocarbon from the mixture to thereby produce a gaseous mixture depleted of

the chlorofluorocarbon, hydrochlorofluorocarbon or hydrofluorocarbon, optionally

followed by the step of desorbing and recovering the adsorbed chlorofluorocarbon,

hydrochlorofluorocarbon or hydrofluorocarbon.

PREFERRED EMBODIMENT OF THE INVENTION

Suitable adsorbents include aluminosilicate clays, activated silica, activated

alumina, activated carbon, selective zeolites or molecular sieves. The particular

adsorbent used is typically selected on the basis of molecular size and differing dipole

moments of the components to be separated.

When molecular sieves are used, preferably the molecular sieve is a

crystalline, metal aluminosilicate with a three-dimensional interconnecting network of

silica and alumina tetrahedra having the structural formula:

M_x/n[(A10₂)_x(Si0₂)_y].wH₂0 wherein n is the valence of the cation M which can be sodium, potassium, lithium,

calcium, barium, strontium, a rare earth metal such as cesium or lanthanum, ammonia

or hydrogen, w is the number of water molecules per unit cell, x and y are the total

number of tetrahedra per unit cell, and y/x have a value of from 1 to 200 or greater.

The size of the entrance to the molecular sieve cages can be varied by using

different cations. For example, large cavities can be formed using small divalent

cations and small cavities result when large monovalent cations, such as potassium,

are used. A Type A molecular sieve having the structure Na₁₂[(A10₂)₁₂(Siθ₂)₁₂] H₂θ

will have cavities of approximately 4 Angstrom unit in diameter. Substitution of a

portion of the sodium for another metal such as potassium results in a molecular sieve

having a cavities of approximately 3 Angstrom unit diameter. Conversely,

substitution of a portion of the sodium for calcium results in a molecular sieve having

cavities of approximately 5 Angstrom unit diameter. Suitable molecular sieves for use

in the invention can be obtained, for example, from the Davison Chemical Division of

W. R. Grace & Co of Baltimore, Maryland. United States of America and sold as 3 A.

4A, 5 A and 13X sieves.

In the present invention, the molecular sieves generally behave as a selective

physical adsorbent such that a chlorofluorocarbon. hydrochlorofluorocarbon or

hydrofluorocarbon smaller than the cavity entrance enters the internal structure and is

held there by physical forces of the Van der Waals type.

For activated carbon, alumina and silica, because of their large surface area

(500 grams of carbon providing about 500,000m of surface), adsoφtion occurs on the

surface by weak bonding of the Van der Waals type between the chlorofluorocarbon. hydrochlorofluorocarbon or hydrofluorocarbon and the adsorbent surface, the surface

of the adsorbent being non-polar and having an affinity for non-polar adsorbents.

Where the molecular size of the chlorofluorocarbon, hydrochlorofluorocarbon or

hydrofluorocarbon is small enough, they are also capable of penetrating the

micropores of the activated adsorbent.

With any adsorbent of the invention, as chlorofluorocarbon,

hydrochlorofluorocarbon and hydrofluorocarbon compounds are inherently polar

because of the asymmetric distribution of atoms which have differing

electronegativities, the Van der Waals forces are augmented by an ion dipole

interaction and the adsorbing force is appreciably strengthened.

Preferably, the method of the invention is used to separate R22 from

R12/R22 mixtures using a 5 Angstrom molecular sieve to trap the R22 and result in

pure R12 product or is used to separate R12 from R12 R22 mixtures using an

activated carbon adsorbent to adsorb the R12 and result in a pure R22 product.

Preferably, when carrying out the invention, a mixture of refrigerants is

typically first analysed (usually by chromatographic methods) to determine its

components. The molecular size and configuration as well as the polarity for each

molecular component is then calculated from the available listings of covalent and

Van der Waals radii and dipole moments and adsorbents, including molecular sieves

with the appropriate cavity sizes and Al :Si ratios, are then chosen for selective

adsoφtion. The adsorbent is usually in the form of small beads (molecular sieves), or

in the form of a powder (activated adsorbents), which is packed in a fixed bed through which the mixture is passed under a pressure drop between the top and the bottom of

the column.

It may be necessary to use more than one adsorbent to separate numerous

components thus giving rise to a number of fixed bed columns which can be run in

series or in parallel. Different adsorbing media can also be placed in the same column.

The number of columns used will depend on the molecular parameters of the

component mixture. Preferably a number of columns are run in parallel so as to

enable a continuous operation whereby columns alternate between adsoφtion and

desoφtion i.e. two or more adsorbent beds are employed and operate on similar cycles

but are sequenced to be out of phase with one another, so that when one bed is on its

adsoφtion step another bed is on its desoφtion step, and vice versa.

Preferably the invention is conducted under vacuum or at normal pressure

and temperature. The appropriate conditions used is however dependent on the

effectiveness of the adsorbent at the relevant pressure and temperature, the

concentration of the chlorofluorocarbon, hydrochlorofluorocarbon or

hydrofluorocarbon to be separated and the actual retention time in the adsorbent. The

adsorbed refrigerant can be desorbed or regenerated by heating and the use of a stream

of purge gas. The preferred method of regeneration is collection, typically under

vacuum, using radiofrequency excitation of the polar adsorbed molecules.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be described by way of example only with reference

to the accompanying drawing and examples in which the separation of R12 and R22

from mixtures thereof was analysed and in which: Figure 1 is a schematic flow diagram of an apparatus for performing the method.

EXAMPLE 1

Prior to performing the method of the invention, an experiment was

conducted to ascertain what size of molecular sieve should be used in order to separate

a mixture of R12 and R22. A 3 Angstrom sieve was found to have a cavity too small

to accept either component. A 10 Angstrom sieve was found to adsorb both

components. A 5 Angstrom sieve was found to selectively adsorb R22 as indicated by

the following Examples 2, 3 and 4.

EXAMPLE 2

A mixture consisting of 16.59% of R12 and 83.32% of R22 was passed

through a column of 5 A zeolite beads and the emission gases were analysed by gas

chromatography. There was an instant reduction in the concentration of R22 present

in the mixture exiting the column, from 83.32% to 0.08%. The gas emitted from the

top end of the column consisted of 99.51% R12. A small amount (0.069%) of Rl 15

(CC 1 F₂CF₃) was evident within the Rl 2. The adsoφtion rate of the R22 followed the

Langmuir isotherm adsoφtion curve.

The unexpected adsoφtion of the R22 molecules in the 5 A cavities of the

zeolite can be partially explained in terms of the higher polarity of the smaller

CHC1F₂ molecule. Both CC1₂F₂ (R12) and CC1F₂CF₃ molecules are larger and less

polar than CHC1F₂ (R22). The dipole moments are 0.61, 0.52 and 1.42 Debye units

respectively. EXAMPLE 3

400g of pure R22 and 4Kg of pure R12 were mixed in an evacuated lOKg gas

bottle to give a final weight of 14.09Kg. The weight/weight ratio in the liquid was

initially 10:1 (R12.R22) but a G.C. run showed the gaseous composition to be

58.38%R12 and 41.62%R22 due to the greater volatility of R22. An apparatus as

shown in Figure 1 was used and comprised a stainless steel tube 1 (96cms long and 8

cms in diameter) packed with 1.44Kg of 5 Angstrom zeolite beads 2 obtained from

Davison Chemical Division of W.R. Grace & Co. The apparatus was evacuated and

the mixture was passed from the gas bottle 3 into the bottom of the tube through a

filter 4. The mixture was allowed to travel through the zeolite bed 2 and the gases

exiting the column 1 were condensed in a specimen container 5 immersed in solid

carbon dioxide. The weight of the gas bottle was monitored 6 throughout so that the

amount of gas entering the tube could be measured. The components of the gas

exiting the tube was monitored using a gas chromatograph 7.

After 170g of the mixture had passed through the bed, the gas exiting the bed

contained 100% R12. After 400g of the mixture had passed through the bed, the gas

exiting the bed contained 99.98% R12 and 0.02% R22. The specimen container was

removed and replaced with a second specimen container. The experiment was

continued and after 800g of the mixture had passed through the bed, the gas exiting

the bed contained 98.46%R12 and 1.54%R22. Finally after 91 Og of the mixture has

passed through the bed, the gas exiting the bed contained 90.229%R12 and

9.771%R22. It is clear that at this point the molecular sieve bed had adsorbed its

capacity of R22 (approximately 130g i.e. approximately 10% w/w adsoφtion). The first specimen was weighed and analysed and contained 1 lOg of gas consisting of

99.875%R12 and 0.125%R22. The second specimen was also weighed and analysed

and contained 11 Og of gas consisting of 94%R12 and 6%R22. As 130g of R22 was

adsorbed, by calculation it can be determined that 182g of pure R12 exited the column

before saturation of the bed and consequently the total weight of processable gas in

order to obtain a 100% separation, was 312g.

EXAMPLE 4

An experiment was conducted in the same apparatus as in Example 3 except

that the mixture contained 33.63 %R 12 and 66.36%R22. Accordingly it was expected

that the zeolite would adsorb 130g of R22 and produce 60g of R12. The condensed

gas was weighed as it was collected into the cooled specimen holder and when 80g of

gas had been collected the contents were analysed using a gas chromatograph and

found to contain 99.9%R12 and 0.08%R22. A further 60g of gas was collected in a

second specimen tube and found to contain 99.668%R12 and 0.332%R22.

The experimental design of the apparatus could be improved by filling the

entire tube with zeolite and sampling the gas exiting the top of the tube thereby

providing immediate detection of the exiting gases and preventing the build up of a

buffer zone of a R12/R22 mix. Further noting a fixed adsoφtion rate for the zeolite,

the amount of exiting gas can be controlled so as to ensure a particular purity of

gaseous product and would enable one to conduct the method in a number of parallel

beds, the beds being alternated when the molecule sieve bed becomes saturated. EXAMPLE 5

Experiments were conducted in the same apparatus as in Example 3 except

that instead of the column being packed with zeolite beads, the column was packed

with a bituminous based activated charcoal. A first experiment was conducted to

determine whether preferential adsoφtion occurs between R12 and R22 and the

surface of the activated carbon. Firstly R22 was passed through the column at a

pressure of 200kPa. A vigorous exothermic reaction was observed during which the

2% water previously adsorbed in the column was emitted as steam and R22 was

adsorbed to an extent of 24.21% w/w. R12 was then fed into the column, and pure

R22 began to emerge from the top of the column. This continued until all the R22 in

the column was quantitatively displaced by the R12, at which point pure R12 emerged

from the top of the column. R12 was also adsorbed in the column to an extent of

44.7% w/w.

In a second experiment the activated carbon was saturated with R12, and R22

was fed into the bottom of the column. Pure R12 emerged from the top of the column

decreasing with time to 0% with a concomitant increase in R22 concentration to

100%. At the conclusion of the experiment, not all of the R/12 was displaced by the

R22, the column containing at the conclusion of the experiment 489g of R12 and 21 lg

of R22.

These experiments indicate that activated carbon surfaces have a preferential

adsoφtion for R12 in R12/R22 mixes. However, the adsoφtion is dynamic, rather

than the static capture mechanism that occurs with zeolites. It is thought that the Van

der Waal's forces may be due to overlap between the vacant 3d orbitals of σ bonded chlorine atoms and the partially filled 2p orbitals of graphitic carbon. This interaction

cannot occur with bonded fluorine atoms due to the absence of d orbitals in the second

shell. The experiments indicate that the selectivity shown by activated carbon is

related to the degree of chlorination of the CFC. an observation in keeping with the pi

bonding theory.

The invention is useful for separation of mixtures of CFC^'s, HCFC^"s and/or

HFC's.

Although the invention has been described with reference to separation of

dichlorodifluoromethane R-12 and hydrochlorodifluoromethane R-22 it should be

appreciated that the invention can be embodied in many other forms and can be used

for the separation of other mixtures of CFCs, HCFC's and HFC's such as mixtures of

any of R-12, R-22, chlorotrifluoromethane R-13, fluorotrichloromethane R-l 1,

trichlorotrifluoroethane R-l 13, dichlorotetrafluoroethane R-l 14,

dichlorotrifluoroethane R-l 23, dichlorofluoroethane R-141B, tetrafluoroethane R-

134A, pentafluoroethane R-125 and heptafluoropropane R-227. The other

halogenated alkane based refrigerants, such as HFC's and iodo substituted derivates

are also embraced by this invention.

Claims

CLAIMS:

1. A method for separating a chlorofluorocarbon, a hydrochlorofluorocarbon or

hydrofluorocarbon from a gaseous mixture of chlorofluorocarbons,

hydrochlorofluorocarbons and/or hydrofluorocarbons comprising passing the mixture

through a bed of adsorbent, the adsorbent being selected so as to preferentially adsorb

the chlorofluorocarbon, hydrochlorofluorocarbon or hydrofluorocarbon from the

mixture to thereby produce a gaseous mixture depleted of the chlorofluorocarbon,

hydrochlorofluorocarbon or hydrofluorocarbon, optionally followed by the step of

desorbing and recovering the adsorbed chlorofluorocarbon, hydrochlorofluorocarbon

or hydrofluorocarbon.

2. The method of claim 1 wherein the adsorbent is selected from aluminosilicate

clays, activated silica, activated alumina, activated carbon, selective zeolites or

molecular sieves.

3. The method of claim 1 or 2 wherein the adsorbent is a molecular sieve and is

a crystalline, metal aluminosilicate with a three-dimensional interconnecting network

of silica and alumina tetrahedra and having the structural formula:

M_{x n}[(A10₂)_x(Siθ₂)_v].wH₂0

wherein n is the valence of the cation M which can be sodium, potassium, lithium,

calcium, barium, strontium, a rare earth metal, ammonia or hydrogen, w is the number

of water molecules per unit cell, x and y are the total number of tetrahedra per unit

cell, and y/x have a value of from 1 to 200 or greater.

4. The method of any one of claims 1 to 3 wherein the adsorbent is selected on

the basis of the molecular size and differing dipole moments of the components to be

separated.

5. The method according to any one of the proceeding claims wherein the

adsorbent is a 5 Angstrom molecular sieve or activated carbon and R12 R22 mixtures

are separated.

6. The method according to any one of the preceding claims wherein prior to

separating the halocarbons, the mixture is analysed to determine its components, the

adsorbent being selected on the basis of the molecular size and configuration as well

as the polarity for each molecular component.

7. A method of anyone of claims 1 to 6 comprising conducting the method

under a pressure gradient.

8. A method of anyone of the preceding claims wherein more than one

adsorbent is used either in a single column or in separate columns, the separate

columns being run in series or in parallel.

9. The method of any one of the preceding claims wherein the halocarbon is

desorbed by heating and passing a purge gas through the bed.

10. The method of any one of the preceding claims wherein halocarbon is

desorbed by radiofrequency excitation under vacuum.