Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~; W094/~2440 ~ 2 g 2 PCT/U~92/0;85t
TITLE
SEPARATION OF TETRAELUOROETHANE ISOMERS
EI~L~ QE~ Y~NTIO~
This invention relates to the separation of
fluorocarbon products, more particularly to the
-separation of the isomers of tetrafluoroethane, CHF2CHF2
(HFC-134) and CF3CH~F ~HFC-134a).
~5~
Isomers of C2H2F4 (HFC-134s) are used as
refrigeration fluids for a number of applications.
HFC-134s can also be used as starting materials for
producing various other halogenated hydxocarbons.
Products containing isomers of C2H2F4 are produced in
various degrees of isomer purity. One method of
producing HFC-134s is by the hydrogenolysis of C2Cl2F4
- isomers ~CFC-114s). In the manufacture of C2Cl2F4 by the
chlorofluorination of perchloroe~hylene the product
typically consists of a mix~ure of the isomers,
CClF2CClF2 (CFC-114) and CF3CCl2F (CFC-114a) ~see e.g.,
U.S. Patent No. 4,605,798). If the CFC-114s are then
used to produce CHF2CClF2 (HCFC-124a~, CF3CHClF
(HCF~-124), HFC-13g or HFC-134a by hydrodehalogenation,
the products often consist of a mixture of C2HClF4 and
C2H2F4 isomers (see e.g., GB 1,578,933).
It has been found that for many applications, the
presence of the second isomer of the isomer pair can
significantly alter the physical and chemical properties
of the desired isomer. For example, variation in the
HFC-134~HFC-134a ratio in the product can result in
dramatic variability in the thermodynamic properties
critical for use in refrigeration applications. For use
as a raw material feed, the presence of the unwanted
isomer can result in yield loss due to increased side
35 reactions. As a result, there has been a continually ;
.. . . .
W094/02~t~ j P~T/US92/0585~-
increasing demand for high isomer purity materials.
Consequently, the separation of HFC-134 isomers
represents a significant aspect of preparing these
compounds for various applications.
Purification of halogenated hydrocarbon pxoducts
has been the subject of considerable research. Of
particular interest are the challenges presented in
separating desired halogenated hydrocarbon products from
materials such as impurities in the starting materials
used to produce the ~alogenated hydrocarbon, excess
reactants, and reaction by-products and/or reaction
co-products which are difficult to remove by standard
separation methods such as distillation. Selectivle
sorbents such as carbons and zeolites have been proposed
for various separations. The effectiveness of
separation using such sorbents varies with the chemical
components and the sorbents involved. The successful
design of sorbent based systems is considered highly
dependent upon experimental determination of whether the
relative sorbencies of the particular compounds are
suitable for such systems.
HFC-134 has a boiling point of -23C and HFC-134a
has a boiling point of -26.5C. Distillation is
consequently relatively inefficient as a means for
separating these two compounds.
SUMMARY OF THE INvE~TION
We have found that mixtures of the isomers of
C2H2F4 (i.e., CHF2CHF2 and CF3CH2F) can be substantially
separated by using a sorbent for CHF2CHF2 selected from
the group consisting of (i) inorganic molecular sieves
(e.g., zeolitesj having greater intermediate
eltectronegativities than Zeolite Na-X, and (ii)
activated carbons. The present invention provides a
process for separating a mixture of CHF2CHF~ and CF3CH2F
to provide a product wherein the mole ratio of CF3CH2F
~. t ~ ~ .
~1~0292 ~ / i
.
relative to CHF2CHF2 is increased which comprises
contacting said mixture with said sorbent at a
temperature within the range of -20C to 300C and a
pressure within the range of lQ kPa to 3000 kPa and for
a period of time sufficient to remove a substantial
amount of the CHF2CHF2. A~ a result, the mole ratio of ~1
CF3CH2F to CXF2CHF2 incre3ses (preferably by 25~ or
more); and a product wherein the mole ra~io of CF3CH2F
relative to CHF2CHF2 is increased, may thus be
recovered.
This invention also provides a process for
separating a mixture of C~F2CHF2 and CF3CH2F to provide a
product wherein the mole ratio of CHF2CHF2 relative to
CF3CH2F is increased which comprises contacting said
mixture with said sorbent as described above to remove a
substantial amount of the CHF2CHF2, and desorbing sorbed
CXF2CHF2 to provide a product which is enriched
therewith.
Said process for producing a CF3CH2F enriched
product and said process for producing a CHF2CHF2
enriched product may be integrated into an overall
process (e.g., a thermal swing cycle process) whereby
both of said products are provided. Said process for
producing a CF3CH2F enriched product and/or said
9~1TUTE ~ET
: 1 ~ C 2 3 2 /~
process for producing a CHF2CHF2 enriched product may
also be used in conjunc~ion with a process for producing
HFC-134 and HFC-134a by the hydrogenolysis of CFC-114
and/or CFC-114a.
DETALL~_nE ~L~ ISQY
The present invention,provides for the separation
of HFC-134 from HFC-134a. Isomer enriched products are
provided in accordance with this invention by contacting
a mixt~re of C2H2F4 isomers with a sorbent for CHF2CHF2 ~ ,
lD selected from the group consisting of acti~ated carbons
and certain inorganic molecular sieves at a temperature
and pressure suitable for sorption, for a period of time
sufficient to remove a substantial amount of said
CHF2CHF2. CF3CH2F enriched product is thereby provided
using CHF2CHF2 sorption. Where CHF2CHF2 enriched product
is desired, the invention also includes a process
involving desorbing sorbed CHF2CHF2 to provide a product -
which is enriched therewith. The process based upon
preferential CHF2CHF2 sorption is particularly useful
0 for purifying CF3CH2F which contains minor amounts of
CHF2CHF2. Where the process is used for such
puri~ication of CF3CH2F, the isomer mix to be purified
by this process generally has a mole ratio of CF3CH2F to
CHF2CHF2 of at least about 9:1, preferably at least
about 19:1, and most ~referably at least about 99:1.
A mix of the C2H2F4 isomers may result, for
example,from a process in~olving the reaction of the
,CFC-114 and~or CF$-114a isomers with hydrogen.
Unreacted starting materials and C2HClF4 isomers may be
recycled and reacted further with hydrogen to produce
additional C2H2F4. Additional impurities may be present
in these products. Distillation is typically used in
order to remove impurities such as HCl, HF, under- and ,
over- chlorinates and fluorinates to produce products
that are at least 90% C2H2F4. Separation of C2H2F4
~rE~
;` W094/02440 ~ ~ i 023 2 PCT/US92/05851
,
isomers in accordance with this invention to provide
prsducts which are enriched in HFC-134 and/or products
which are enriched in HFC-134a then may be
advantageously employed. This invention can thus be
5 adapted for use in connection with production of C2H2F
by hydrogenolysis of materials such as C2C12F4 such that
after removal of a substantial amount of CHF2CHF2 using
the sorbent, either ~i) a product is recovered wherein
the mole ratio of CF3CH2F relative to CHF2CHF2 is
increased, (ii~ sorbed CHF2C~F2 is desorbed to produce a
product wherein the mole ratio of CHF2C~F2 relative to
CF3CH2F is increased, or both (i) and ~ii). j
Some embodiments of this invention use activated
carbon as the sorbent. Commercially available activated
carbon may be used. The effectiveness of the process
can be influenced by the particular activated carbon
employed. Moreover, ~he sorption efficiency and
sorption capacity of an activated carbon bed depends
- upon the particle size of an activated carbon in a
dynamic flow system. Preferably, the activated carbon
has a particle size range of from about 4 to 325 mesh
~from about 0.044 to 4.76 millimeters). More
preferably, the activated carbon has a particle size
range of from about 6 to 100 mesh ~from about 0.149 to
3.36 millimeters). Most preferably, the activated carbon
has a particle size range of from about 10 to 30 mesh
~from about 0.595 to 2.00 millimeters).
An activated carbon obtained having a particle size
rang~ of about 0.074 x 0.297 millimeters (50 x 200 mesh)
is available from the Barneby & Sutcliffe Corp. as
Activated Carbon Type UU (natural grain, coconut shell
based). An activated carbon having a particle size of
0.595 millimeters x 1.68 millimeters ~12 x 30 mesh) is
~ available from the Calgon Corporation as Calgon ~PL
~bituminous coal based) activated granular carbon. An
W094/0~440 ~ 1 4 U 2 9 2 PCT/U592/0~85t _j
activated carbon having a particle size range of about
0.450 x 1.68 millimeters (12 x 38 mesh) is available
from Barnebey & Sutcliffe Corp. as Barneby & Sutcliffe
Corp. Activated Carbon Type PE ~natural grain, coconut
S shell carbon). An activated carbon having a particle
size range of about 0.297 x 0.841 millimeters (2Q X 50
mesh) is available from Westvaco as Microporous
Wood-Base Granular Carbon.
Typically the activated carbon used will have a
total content of from about 0.1 to 10 weight percent of
alkali and alkaline earth metals selected from lithium,
sodium, potassium, rubidium, cesium, magnesium, calcium,
strontium and/or barium. The alkali and alkaline earth
metal content of carbon can be regulated by techniques
known in the art. For example, the metal content of
carbon can be reduced by acid washing; and the metal
content can be increased by standard impregnation
techniques. In a preferred embodiment using
preferential HFC-134 sorption, the activated carbon
contains inherent aikali and/or alkaline earth metal~s)
selected from the group consisting of lithium, sodium,
potassium, rubidium, cesium, magnesium, calcium,
strontium, barium, and combinations thereof. Inherent
alkali metals (typically Na and~or K) are preferred.
The presence of these metals, particularly as inherent
metals in the range of from about 0.5 to 3 percent by
weight, is considered to improve the HFC-134 sorption
efficiency.
Some embodiments of this invention use inorganic
: t , I !
molecular sieves. Molecular sieves are well known in
the art and are defined in R. Szostak, Mol~cular
Sieves-Principles of Synthesis and Identification, Van
Nostrand Reinhold (1989) page 2. The inorganic
molecular sieves used for preferentially sorbing HFC-134
in accordance with this invention include various
~1~Q292 ~ /q ~
.
silicates (e.g., titanosilicates and zeolites such as
Zeolite Y, Zeolite A, Zeolite ZSM-5, and Zeolite ZSM-8),
metalloaluminates and aluminophosphates, as well as
other inorganic molecular sieve materials. The
molecular sieves useful in the invention will typically
have an average pore size of from about 0.~ to 1.5 nm.
The Sanderson electronegati~ity model (see,
R. T. Sanderson, I'Chemical Bonds and Bond En~rgyl', 2nd
ed., Academic Press, New York, 1976) furnishes a useful
method for classifying inorganic molecular sieves based
on their chemical composition. In ~ccordance with this
invention the preferential sorption of tetrafluoroe~hane
isomers by molecular sieves can be correlated with their
intermediate electronegativity (i.e., their "Sint") as
determined by the Sanderson method based on chemical
composition. The Sint for Zeolite Na-X is about 2.38.
Inorganlc molecular sieves with Sints greater than
the Sint for Zeolite Na-X ~i.e., more electronegative or
more acidic~ may be used in accordance with this
invention for increasing the mole ratio of CF3CH2F
relati~e to CHF2CHF2 by remo~ing a substantial amount of
CHF2CHF2; and/or for increasing the mole ratio of
CHF2CFH2 relati~e to CF3CH2F by desorbing sorbed CHF2CHF2
(i.e., CHF2CHF2 is belie~ed to sorb more strongly than
CH2FCF3).
~e~TlTVTE 9~T
~ 1 ~ 0 2 9 2
Example Sint values are provided in Table I.
TABLE I
Intermediate Sanderson Electronegativities
for Selected Molecular Sieves
olecular Sieve ~in~ !
Na-X 2.38
Ca-A 2.56
Na-Y 2.58
ETS 2.60
H-Y . 2.97 .
Na-ZSM-8 3.00
H-ZSM-5 3 04
H-ZSM-8 3-04
Generally, for the inorganic molecular sieves, it
is desirable to occupy acidic sites of the sie~e
material with alkali metal(s) and or alkaline earth
metal~s) so long as the intermediate electronegativity
remains suitable for the desired separation. In a
preferred embodiment using preferential Y.-C-134 sorption
the inorganic molecular sieve is a Zeolite Y which
contains alkali or alkaline earth metal(s) selected from
the group consisting of lithium, sodium, pot~ssium,
rubidium, cesium, magnesium, calcium, strontium and
CTlTUTE 9t~ET
~?, W094/02440 2 1 4 0 2 g 2 PCT/US92/05851
barium or co~binations thereof. Alkali metals are
preferred. It is preferred that alkali metals occupy
from about 50% to 100~ of the accessible exchange sites
in the zeolite. Particularly preferred zeolite
molecular sieves include those having alkali metal to
àluminum ratios of about 1:1, or alkaline earth metal to
aluminum ratios of about 1:2. ;
Suitable temperature ranges for sorption range from
about -20C ~o about 300C. Suitable pressures for
sorption range from about 10 kPa to about 3000 kPa.
Contact with sorbent should be sufficient to
achieve the desired degree of isomer enrichment.
Preferably, the mole ratio of the enriched isomer to the
second isomer is increased by at least about 25%
relati~e to the mole ratio thereof in the initial
mixture, most preferably by at least about 50%.
Where the process is used to purify CF3CH2F from a
mixture of CF3CH2F and CHF2CHF2 using preferential
HFC-134 sorption, preferably at least about 50 mole % of
the CHF2C~F2 is removed. A particularly advanta~eous
embodiment of this invention involves providing
sufficient contact to produce CF3CH2F of at least about
99.99 mole percent purity.
This invention can be practiced with the sorbent
contained in a stationary packed bed through which the
process stream whose components need separation is
passed. Alternatively, it can be practiced with the
sorbent applied as a countercurrent moving bed; or with
a fluidized bed where thelsorbent itself is moving. It
can be applied with the sorbent contained as a
stationary packed bed but the process configured as a
simulated mo~ing bed, where the point of in~roduction to
the bed of the process stream requiring separation is
changed, such as may be effected using appropriate
switching valves.
., .. . , , , , . - - , :
W094/~2440 , PCT/VS92/0~85 ~
21~0~gZ~' '' o
The production of a product enriched with respect
to one C2H2F4 isomer may be accompanied by the production
of other products which are enriched with regard to the
concentration of one or more other components of the
S initial mixture. Indeed, a typical process might
include both a product which is enriched in CF~CH2F
~e.g., essentially pure CF3CH2F) and another product
which is enriched in CHF2CH~2. The production of
product enriched in CHF2CXF2 generally involves
desorption of CHF2CHF2. In any case, whether or not a
CHF2CHF2 enriched product is desired, the sorbent is
typically regenerated following CHF2CHF2 removal by
desorption of sorbent materials. Desorption of
components held by the sorbent may be effected with the
sorbent left in place, or the sorbent may be removed and
the desorption effected remotely from where the sorption
step occurred. These desorbed components may exit the
sorbent section in a direction either co-current (in the
same direction as the orisinal C2H2F4 mixture feed was
fed) or countercurrent (in the opposite direction of the
original stream requiring separation). Desorption may
be e~fected with or without the use of a supplemental
purge liquid or gas flow. Where supplemental purge
material is used, it may be a component of the feed, or
some appropriate alternative material, such as nitrogen.
Such supplemental purge materials may be fed either
co-currently or countercurrently.
In general, desorption can be effected by changing
any thermodynamic variable ~which is effecti~e in
remo~ing the sorbed components from the sorbent. For
example, sorption and desorption may be effected using a
thermal swing cycle, ~e.g., where after a period of
sorption, the sorbent is heated externally through the
wall of the ~essel containing it, and/or by the feeding
of a hot liquid or gas into the sorbent, the hot gas
". . - . , . . - . . ~
W094/02440 ~ ~ 4 ~ '~ 9 ~ PCT/US92/05851
heing either one of the component materials or
alternative materials). Alternatively, sorbed
components can be removed by using a pressure swing
cycle or ~acuum swing cycle ~e.g., where after a period
S of sorption the pressure is suf~iciently reduced, in
some embodiments to a vacuum, such that sorbed
components are desorbed). Alternatively, the sorbed
components can be removed by use of some type of
stripping gas or liguid, fed co-currently or
countercurrently to the original process feed material.
The stripping material may be one of the process feed
materials or another material such as nitrogen.
One or several beds of sorbent may be used. 'Where
several beds are used, they may be combined in series or
in parallel. Also, where several beds are used, the
separation efficiency may be increased by use of cycling
zone sorption, where the pressure and or the
temperatures of the beds are alternately raised and
lowered as the process stream is passed through.
Practlce of the invention will be further apparent
from the following non-limiting Examples.
Metal tubing ~0.l8 inch I.D. x 12 inch, 0.46 cm
I.~. x 30.5 cm) was packed with a carbon sorbent and
installed in a gas chromatograph with a flame ionization
- detector. Helium was fed as a carrier gas at 33 sccm
(5.5 x 10-7 m3/s). Samples of the various compounds were
then injec~ed into the carrier stream at 200C. The
results of these experiments using Barneby & Sutcliffe
Type PE (3.75 gj carbon (Carbon A), Westvaco Microporous
Wood-Based Granular Carbon ~Carbon B), Barneby &
Sutcliffe Type UU (3.85 g) carbon (Carbon C) and Calgon
BPL (2.59g) carbon (Carbon D) are shown in Table l.
These data show that in each case the isomers had
W 0 94/02440 ,~ g ~ PCr/US92/05851'~-
different retention times, and thus may be separated
using the carbons of this Example.
rABLE, 1
Sample Rete~tion Time fmi~l S~paration
A 5 6.6 4.0 1.65 0.13% 1.0~%
- 200 4.36 3.16 1.36 0.13% 1.09%
B S 4.22 2.61 1.62 0.58% 75ppm
209 3.32 2.27 1.46 0.58% 75ppm
C 200 4.79 3.38 1.42 940ppm 0.93%
D 5 2.32 1.75 1.32 660ppm 650ppm
_ _00 2.01 1.59 1.26 _660pum _ 650D m
)Volume of gas sample injected ~microliters)
~)134 - C~F2CHF2
)134a ~ ~F3CH2F
~d)Separation Factor - 134 retention time/134a retention time
(~)sodium content of carbon in weight percent or parts per
million as indicated
~potassium content of carbon in weight percent or parts per
million as indicated
It is evident from Table 1 that the rela~ive
sorption efficiency for HFC-134 is higher in the
presence of the alkali metals Na and K.
~MP~ 2
Metal tubing (0.18 inch I.D. x 12 inch, 0.46 cm
I.D. x 30.5 cm was packed with a carbon sorbent and
installed in a gas chromatograph with a flame ionization
detector. The experiment was repeated using the same
carbon, but washing it with hydrochloric acid before
using it for separations. The sodium content of
Westvaco Microporous Wood-Based Granular Carbon (Not
Acid-Washed, NAW),was 1.29~. After washing with
hydrochloric acid the sodium content was 9 ppm. This
carbon was designated Acid-Washed ~AW). Helium was fed
as a carrier gas at 33 sccm (5.5 x 10-7 m3~s). Samples
of 134 and 134a were then injected into the carrier
stream at 200C. The results of these experiments are
~'~ W094/0~0 ~ 1 ~ O ~ 9 2 PCT/US92/05851
13
shown in Table 2. These data show that a more efficient .
separation was obtained with the carbon containins
alkali metal; in this case sodium.
~L~
Retention Time tmin.) Separation
Carbon Na Content 134 134a Factor(~)
~AW l.29% lO.67 6.63 l.61
AW 9 ppm 6.2 4.3 _ 1.44 _
)Separation Factor Y 134 retention timeJ134a retention time
~L~ I
A packed tube (2 6 cm x 2 .12 cm I.D) containing
Calgon BPL carbon ~46.1 g, 4.8 x 0.59 mm (12 x 30 mesh))
was purged wi~h nitrogen continuously for 24 hours at
2SQC and at l atmosphere pressure. While still being
purged with nitrogen, the bed was cooled and was
maintained at 25C. HFC-134a containing 1 wt% ~FC-134
was then fed to the bed a~ l6~ grams per hour. The
results are shown in Table 3.
W0 94/02~40 ~ PCT/US921058
2 9 ~ 14
~I~
Time 134a 134a 134
o o o - j
~1 0.164 0 0
0.175 0.011 0 ~.
77 0~207 0.043 0
89 0.239 0.075 0.61
100 0.269 0.105 0.88
112 0.301 0.137 O.g6
124 0.334 0.170 1.00
~)134a in represents the total running sum of the moles ~:E
CF3C~2F fed ~o the column.
~)134a out represents the total running sum of the moles of
CF3CH2F exiting the column.
(C~134 out represents the instan~aneou~ concentration of CI~F2CH~2
in the CF3CH2F exiting the column, expressed as a multiple of
the 1 wt.% feed (i.e., O.5 would equal a O.5 w~.% HFC-134
concentration in the HFC-134a effluent). A zero i~ less than
the detection limit of about 10 ppm.
This example shows that carbon will selectively
hold back HFC-134 allowing HFC-134a free of HFC-134
followed by HFC-134a containing reduced HFC-134
concentrations to be obtained.
E~æL~_i
A packed tube ~26 cm x 2.12 cm I.D) containing
Barneby & Sutcliffe Type PE carbon ~50.3 g) was purged
with ni~rogen continuously for 12 hours at 250C and at
l atmosphere pressure. While still being purged with
nitrogen, the bed was cooled and was maintained at 25C.
HFC-139a containing 1 wt% HFC-134 was then fed to the
bed at 16.7 grams per hour. The results are shown in
.
Table 4.
W O 94~02440 ~ 1 ~ æ ~ ~ PCT/US92/0~851
~A~LE 4 t
Time 134a 134a 134
(min) in(a) out(~) out(C)
,
- O O O ~ ~
69 0.186 0 0 ~ .
73 0.196 0.010 0
8~ 0.22g 0.043 0
96 ~.258 0.072 0
108 0.291 0.105 0
120 0.323 0.137 0.76
3~ 0.355 0.169 0.94
144 _ 0.387 0.201 1.~0
~134a in represents the total running sum of the moles of
CF3CH2F fed to the column.
t~)134a out represents the total running ~um of the moles of
CF~CH2F exiti~g the column.
(C)134 out represents the instantaneous concentration of CHF2CHF2
in the CF3CH2F exiting the column, expressed as a multiple of
the 1 wt.% feed (i.e., 0.5 would equal a 0.5 wt.% HFC-134
concentration in the HFC-134a efflue~t)~ A zero is less than
the detection limut of about 10 ppm.
~I~ :
A pac~ed tube (26 cm x 2.12 cm I.D) containing
West~aco Microporous Wood-Based Granular Carbon (46 g)
was purged with nitrogen continuously for 12 hours at
S 25QC and at l atmosphere pressure. While still being
purged with nitrogen, the bed was cooled and was
maintained at 25C. HFC-134a containing 1 wt% HFC-134
was then fed ~o the bed at 16.6 grams per hour. The
results are shown in Table 5.
,
':
s~
,~
: ,t
W094/02440 ~ PCT/US92/0~85 ~
2140292 16
TABL~_~
Time134a 134a 134
(min~ ) out~) out~C)
i
o o o o
74 0 . 199 0 . 003 0
82 0 . 2~1 0 . 025 0
94 0 . 253 0 . 057 0
106 0 . 285 0 . 089 0
118 0 . 317 0 . 121 0 . 16
130 0 . 350 ~ . 154 0 . 65
142 0 .382 0 . 186 0 . 97
155 _ 0 ! 417 0 . 221 _ 1 . 00
(~)134a in represents the total runnin~ ~um of the moles of
CF3CH2F fed to the column.
(b)i34a out represents the total running sum of the mole~ of
CF3CH2F exiting the column.
~C3134 out represents the instantaneous concentration of CHF2CHF2
- in the CF3~H2F exiting the column, expressed as 3 multiple of
the 1 wt.~ feed ~i.e., 0.5 would equal a 0.5 wt.% HFC-134
concentration in the HFC-~34a effluent). A zero i~ less than
the detection limlt of about 10 ppm.
F~xA~p~E 6
A packed tube ~26 cm x 2.12 cm I.D) containing
Westvaco Microporous Wood-Based Granular Carbon (46 g)
was purged with nitrogen continuously for 12 hours at
5 250C and at 1 atmosphere pressure. While still being
purged with nitrogen, the bed was cooled and was
maintained at 25C. HFC-134a containing 1 wt% HFC-134
was then fed to the bed at 16.6 grams per hour and at
4.7 atm. (476 kPa). The results are shown in Table 6.
~~~ W094~02440 ~ 1 4 0 2 ~ 2 PCT/US92/05851
~7 ~;
TA~LE 6
Time 134a 134a 134
(min) in (a) out~b~ out~
~ t
- O O O O
54 0.261 0.013 0
0.3125 0.067 0
77 0.373 0.1~5 0
89 0.431 0.183 0
101 0.489 0.241 0.33
113 0.547 0.299 0.47
125 0.605 0.357 0.55
137 0.663 0.415 0.8
t~)134a in represents the total running sum of the moles oi.
CF3CH2F fed to the column.
)134a out represents the total running sum of the moles of
CF3CH2F exiting the column.
(C)134 out represents the instantaneous concentration of CHF2~HF2
in the CF3CH2F exiting the column, expressed as a multiple of
the 1 wt.~ f~ed ~i.e., 0.5 would equal a 0.5 wt.% HFC-134
concentration i~ the HFC-134a effluent). A zero is less than
the detection limit of about 10 ppm.
Examples 3 through 6 show that these carbon based
sorben~s will selectively sorb HFC-134 allowing
HFC-134a free of HFC-134, followed by HFC-134a
containing reduced HFC~134 concentration to be
5 obtained. Examples 3 through 6 show that process
material other than the componen~s to be separated can
be used to strip HFC-134 ~i.e., in this case, nitrogen
rather than HFC-134a is used to clear the bed of
HFC-134). Also, examples 5 and 6 show that the capacity
or 134 increases with pressure, and illustrates the
presence of pressure swing adsorption.
EX~PLE 7
This is an example of a thermal swing cycle
~~ alternating a sorption step with a desorption step. The
column and carbon pac~ing are the same as that used in
Example 4 above. During the sorption step, HFC-134a
containing 1 wt % 134 was fed to the packed column at
f!
,, .
q
. . .
WO 94/0~0 , . PCT/U~92/0585~
.. . ~ . "
~ 1 4~ ~ 9 2 18
26C and at a feed rate of 16.6 g/hr with a back-
pressure se~ting of 1 atmosphere ~101 kPa) in the
column. When HFC-134 began to break through at the
other end of the column, the flow of feed was stopped,
and the ends of the column were sealed. The column was
then heated to 150C, and gas was vented from the column
in the direction countercurrent to the original
direction of feed, to keep the pressure at 1 atmosphere
(101 kPa). ~hen the temperature reached 150C, HFC-134a
containing less than l ppm of HFC-134 was fed in the
direction countercurrent to the original feed to purge
the bed, at 16.5 g/hr ,and with a back pressure setting
of 1 atmosphere (101 kPa). The column valves were then
closed at both ends, and the column cooled to 26C. The
cooling of the bed caused a partial vacuum. The
pressure was then brought back to 1 atmosphere (101 kPa)
using the high HFC-134 content HFC-134a and the cycle
was started again. The sorption and desorption steps
were then repeated. The results of the second sorption
step are shown in Table 7A.
TABLE 7A
Time Temp HFC-134a HFC-134a HFC-134
~Min) C in (a) out~) out(C)
,
0 26 0 0 0
93 26 0.250 0.124 0
117 26 0.315 0.189 0.77
129 26 0.347 0.221 0.87
140 _ 26 0.377 0.251 0.96
~a)HFC-134a in represents the total running sum of the moles of
HFC-134a~ed to~the col~mn. : ,
~b)HFC-134a out represents the total running sum of the moles of
HFC 134a exiting the column.
)HFC-134 out represents the instantaneous concentration of the
HFC-134 in the HFC-134a exiting the column, expre-~ed aY a
multiple of the 1~ feed. A zero is le~s than the detection
limit of about 10 ppm.
The results of the desorption step which followed
are shown in Table 7B.
WO94/02440 2 1 4 0 2 9 2 PCT/US92/05851
1 9 `
~L~ . ,
Time Temp HFC-134a ~FC-134a HFC-134 ;
(min) ~C in(a) out~b) out(C~ ~
,
0 25 0 ~ 0
7 3S 0 0.049 1.16
12 98 0 0.107 1.~1
24 150 0 0.133 1.63
150 0.032 0.165 1.67
47 150 0.0~2 0.195 1.63
_ 71 150 0.126 0.259 0 _
la)HFC-134a in re~esents ~he total running sum of the mole~ of
HFC-134a fed to the column.
FC-134a out repr~ents the total running sum of the mo:Le~ of
HFC-134a exiting the column.
~)HFC-134 out repre~ents the instantaneous concentration of the
HFC-134 i.n the HFC-134a exiting the column, expre~sed as a
multiple o~ the 1~ feed. A zero is less than the detection
limit of about 10 ppm.
Initially, no HFC-134a was fed, but HFC-134a and
HFC-134 exited the column due to the let down of the
pressure as the ~emperature was raised from 26C to
150C. Beginning at 24 minutes, when the temperature
reached lS0C, HFC-134a containing less than 1 ppm 134
was fed at 16.5 g/hr. At 71 minutes, the HFC-134a flow
was stopped.
This example shows the use of a temperature swing
cycle as a process concept to produce both HFC-134-free
and HFC-134-reduced HFC-134a.
EXAMPLE 8
This is an example of a thermal swing cycle
alternating a sorption step with a desorption step. The
column and carbon packing were the same as that used in
Examples 5 and 6 above. During the sorption step,
HFC-134a containing 1 wt % 134 was fed to the packed
column at 25C and a 134a feed rate of 16.6 g~hr with a
back-pressure setting of 1 atmosphere (101 kPa) in the
~ column. When the outlet HFC-134 concentration matched
the inlet concentration, the flow of feed was stopped,
.
~i
WO 94/0~0 PCT/US92/05851~f'~
2 1 ~ 0 2 9 2 20
and the ends of the column were sealed. The column was
then heated to 150C, and gas was vented from the column
in the direction countercurren~ to the original
direction of feed, to keep the pressure at 1 atmosphere
(101 kPa). When the temperature reached 150C, HFC-134a
containing less than 1 ppm of HFC-134 was fed in the
direction countercurrent to the sriginal feed to purge
the bed, at 16.5 g/hr and with a back pressure setting
of 1 atmosphere ~101 kPa~. The column valves were then
closed at both ends, and cooled to 25C. The cooling of
the bed c~used a partial vacuum. The pressure was then
brought back to 1 atmosphere (101 kPa) usin~ the hi.gh
HFC-134 content HFC-134a and the cycle was started
again. The sorption and desorption steps were then
repeated.
Tne results of the second sorption step are shown
in Table 8A.
TABLE 8A
Time Temp HFC-134a HFC-134a HFC-134
~min) C in ~a) out~b) out(c)
0 25 0 0 0
29 25 0.140 0 0
73 25 0.353 0.213 0
0.411 0.271 0.25
97 25 0.469 0.329 0.80
_ 100; 25 0.532 0.392 1.00
(a)HFC-134a in repreaents the total running sum of tbe mole~ of
HFC-134a fed to the column.
)HFC-134a out represents the total running sum of the moles of
HFC-134a exiting the column
)HFC-134 out represents the instantaneous concentration of
HFC-134 in the HFC-134a exiting the column, expre~-~ed as a
multiple of the 1~ feed. A zero is les-~ than the detection
lim~t of about 10 ppm.
The results of the desorption step which followed
are shown in Table 8B.
~'^WO94/02440 ~ 1 ~ 0 2 3 2 PCT/US92/05851
21' ' '
~BLE 8B
Time Temp HFC-134a HFC 134a' HFC-134
0 25 0 0 0
36 0 0.049 1.09
22 78 0 0.105 1.31
34 ~30 0 0.150 1.54
4~ 150 0.015 0.170 1.65
~8 150 0.073 0.~2~ 1.59
71 15~ 0.136 0.291 1.56
83 150 0.194 0.349 0.0
_ g5 150 0.252 0.407 _ 0
5a)HFC-134a in represents the total running 3um of the mol~as of
XFC-134a fed to the column.
)HFC-134a out represents the total running sum of the mo:Les of
HFC-134a exiting t~e column.
)HFC-134 out represents the instantaneous concentration of the
HFC-134 in the HFC-134a exiting the column, expressed a~ a
multiple of the 1~ feed. A zero is less than the detec~ion
limit of about 10 ppm.
Initially, no HFC-134a was fed, but HFC-134a and
HFC-i34 exited the column due to the let down of the
pressure as the temperature was raised from 25C to
150C. Beginning at 34 minutes, when the temperature
reached 150C, HFC-134a containing less than 1 ppm 134
was fed at 16.5 g/hr. At 95 minutes, the HFC-134a flow
was stopped.
This example shows the use of a temperature swing
cycle as a process concept to produce both HFC-134-free
and HFC-134-red~ced HFC-134a.
~AM~L g
Me~al,tubing 0.18" (4.6 mm) I.D. x 2 ft.~0.51 m)
was packed with zeolite sorbents as indicated in Table
9, and installed in a gas chromatograph with a flame
ionization detector. The columns were heated at 200C
in flowing helium for a minimum of 12 hours. Helium was
fed as a carrier gas at 20 sccm ~3.3 x '0-7 m3/s).
Samples ~25 ~L) of HFC-134 and HFC-134a were then
~.. ~ , . .
WO 94/02440 ~ ~ ~ PCT/US9~/05~5
injected into the carrier stream at different
temperatures. The resul~s of these experiments are
shown in Table 9. Comparison of the 134/134a data for
Na-Y and H-Y show a much enhanced separation on the more
5 basic zeolite. ,
- TABLE 9
- Zeolite Temperature Retention Times Separation
(C) (min) Factor
~
Na-Y 200 408/71.4 5.7
H-Y 100 68.1~46.1 1.5
H-ZSM-sla) 200 470/308 1.5
5A 200 291/about 150 1.9
(a~The flow rate fos the H-ZSM-5 ~un was 35 sccm (5. 8 x 10-7 m3/s
~L~
Metal tubing 0.18" (4.6 mm) I.D. x 2 ft.(0.~1 m)
was packed with zeolite sorbents as indicated in Table
10, and installed in a gas chromatograph with a flame
ionization detector. The columns were heated at 200C
in flowing helium for a minimum of 12 hours. Helium was
fed as a carrier gas at 30 sccm (5.0 x 10-7 m3/s).
Samples (25 to 500 ~L) of HFC-134 and HFC-134a were then
injected into the carrier stream at different
temperatures. Each test was run in duplicate. Methane
(1% in nitrogen) was run as a standard at each
temperature. The results of these experiments are shown
in Table 10.
;
W094/~2440 ~ PCT/US92/0~8~1 ~
23 I `
~L~
Zeolite Temperature Retention Times Separation .
~C) ~min) Factor .
134/134a
Na-Y 230 160/53.2 3.0 -
240 128/45.1 2.8
~50 98.3/36.8 2.7
H-Y 210 2.07/1.19 1.7
220 1.78/1.06 1.7 ~ .
230 l.OlfO.89 1.1
H-ZSM-8 210 110/70.2 1. 6 3 .
220 79.7/51.1 1~6
230 56.7~37.5 1.5
5A 230 74.7/29.4 2.5
240 64.8/24.9 2.6
-
EXAMPLr l 1
This is an example of a ~hermal swing cycle with
countercurrent purge during desorption. A 1 inch
(2.54 cm) diameter tube was packed with 63 grams of the
zeolite H-ZSM-5, and purged with nitrogen at 50 psig
~450 kPa). The ni~rogen was then turned off, and the
column ed with HFC-134a containing 102 mole % HFC-134
at 60 sccm (1.0 x 10-7 m3/s) and 50 psig ~450 kPa). The
results of this test are shown in Table llA.
W094J02440 PCT/US9~/0585 ~
2140~292 24
Time Temp HFC-134
(min) o C out (~)
0 29
29
29 0*
29 0
29 0
29 0
~9 0
~9 o
29 0
29 0.18
100 29 0.34
110 2~ O.S1
~ , '
*~reakthrough cf HFC-134a occurs.
)HFC-134 out represents the ins~antaneous concentration of
~FC-134 in the HFC-134a exiting t~e column, expr~ aed ~5 a
multiple of the original 1% feed. A zero is less than the
detection limlt of about 10 ppm.
When the outlet concentration of the 134 matched
the inlet concentration, the high 134 concentration 134a
flow was stopped and the ends of the column were sealed.
The column was then heated to 150C, and gas was vented
S from ~he column in the direction countereurrent to the
original direction of feed, to keep the pressure at 1
atmosphere (100 kPa). When the temperature reached
150C, HFC-134a containing less than 1 ppm 134 was fed
in the direction countercurrent to the original feed at
50 psig ~450 kPa). The results are summarized in Table
11~}.
'^: W0 94/02440 21~G~2 PCT/US92/058~1 t
TABL~
Time Temp HFC-134
(min) ~~ out~')
0 31 1.06
86 1.19
114 1.37
126 1.56
133 1.63
133 1.67
133 1.75
134 1.61
134 1.25
134 0.78
100 134 0.41
1~0 134 0.18
~a)HFC-134 out represents the instantaneous concentration of the
HFC-134 in the HFC-134a exiting the column, exprec~ed as a
multiple of the orlginal 1% feed. A zero is less than the
detectlon limit of about 10 ppm.
~AMPLE ~2
This is an example of a thermal swing cycle with
countercurrent purge during desorption. A 0.93 inch
(2.3S cm) diameter by 12 inch (30.48) long tube was
pac~ed with 80 grams of LZ-Y52 zeolite ta Na-Y zeolite),
and purged with nitrogen at 50 psig (450 kPa). The
nitrogen was then turned off, and the column fed with
HFC-134a containing 1.5 mole % HFC-134 at 50 psig
. (4S0 kPa) and 30C. The results of this test are shown
in Table 12A.
WO 94/02440~ 1 4 ~ ~ 9 2 PCT/U~92/0585
26
TAB~ 12
HFC-134a HFC 134a HFC-134
O O O
0.~35 0.004 0
O.g58 0.717 0
- 0.965 0.735 0.108
0.987 0.75~ 0.221
1.009 0.781 0.3Q4
1.032 0.8~3 0.379
1.053 0.8~5 0.447
1.071 _ 0.843 0.460
(a)HFC-134a in represents the total running sum of ~he mole~ of
~FC-134a fed to the column.
(~)HFC-134a out represents the total running sum of the mole3 of
HFC-134a exitin~ the column.
(C)HFC-134 represents the instantaneous concentration of HFC-134
in the HFC-134a exiting the column, express~d as a mul~i.ple of
the 1.54 feed.
When the outlet concentration of the 134 reached
70% of the feed concentration, the high 134
concentration 134a flow was stopped and the column
heated to l50~C. The pressure generated from the
heating was vented from the column in the direction
countercurrent to the original direction of feed.
During the temperature ramp, approximately 0.0514 moles
of 134a and 0.0002 moles of 134 were vented. When the
temperature reached 150C, HFC-134 free HFC-134a was fed
in the direction countercurrent to the original feed at
50 psig (450 kPa) and 150C. The results are summarized
in Table 12~.
~'~ W094/02440 ~110 2 3 2 :PCT/US92/05851
27
~L~L2~
HFC-134a HFC-134a HFC-134
in~) out~b) out(c)
O O O
0.033~ 0.0321 3.26
0.0673 0.0639 3.45
- 0.1010 0.0958 3.48
0.1347 0.1277 3.48
0.1795 0.1702 3.26
0.2118 0.2009 3.05
0.2513 0.238Q 2.4B
_ _ Q.3545 _0.3396 _ 0.58
FC-134a in represents the total running sum of the moles of
HFC-134a fed to the column. . :
)HFC-134a out repre~ents the total running ~u~ of the moles of
HFC-134a exiting the column.
(C)~FC-134 represents the instantaneous concentration o:E HFC-134
.~ in the HFC-134a exiting the column, expres~ed a~ a multiple of
the 1.5~ feed.
EXAM
Metal tubing 0.18" (4.6 mm) I.D. x 2 ft. (0.51 m~
was packed with zeolite sorbents as indicated in
Table 13, and in~talled in a gas chromatograph with a
flame ionization detector. The columns were heated at
200C in flowing helium for a minimum of 12 hours.
-~ Helium was fed as a carrier gas at 30 sccm (S.0 x
10-7 m3/s). Samples t5 to 25 ~L) of HFC-134 and HFC-134a
were then injected into the carrier stream at different
;temperatures. Each test was run in duplicate. Methane
- tl% in nitrogen) was run as a standard at each
~- temperature. The results or these experiments are shown
- in Table 13. , ; ~
~- ` 2 1 ~ 0 2 3 2
28
~L~
Retention Time~
Temperature (m1n) Separation
ETS-10~a) 200 262.5/gO.3 2.9
Na-A 200 1.6/G.6 2.7
Clinoptilolite 200 1.0~0.8 1.3
Ferrierite _ 200 0.9/0.5 _ __ _ 1.8
~)Sodium Pota sium Titanosilicate
~EL~
Metal tubing 0.18" (4.6 mm) I.D. x 4.5 in (11.4 cm)
was packed with Zeolite Na-X as indicated in Table 14,
and installed in a gas chromatograph with a flame
ioni~ation detector. The columns were heated at 200C
in flowing helium for a minimum of 12 hours. Xelium was
fed as a carrier gas at 30 sccm (5.0 x 10-7 m3/s).
Samples (25 ~L) of HFC-134 and HFC-134a were then -
injected into the carrier stream at different
temperatures. Each test was run in duplicate. Methane
(1% in nitrogen) was run as a standard at each
temperature. The retention times were difficult to
' 15 measure.
TABLF L4
Temperature
Zeolite ~C)
Na-X 200
_ Na-X _ __ 21 Q
The examples serve to illustrate particular
embodiments of the invention. The invention is not
confined thereto, but embr~ces embodiments which come
within the scope of the claims.
~ UTE ~ET
