Abstrict Difluoromethane (R-32) is of current interest as a partial replacement
for chlorodifluoromethane (R-22) refrigerant heretofore widely used
in vapor compression refrigeration systems. R-32 has, however, proved
to be more reactive than is desirable with the zeolite A adsorbent-desiccant
compositions used in such systems to prevent corrosion and freeze-up
problems. The potassium cation form of a zeolite A molecular sieve--with
at least 60 percent of the sodium cations replaced with potassium
ions, agglomerated with a clay binder, and pore-reduced to essentially
exclude the adsorption of R-32 having essentially no reactivity
with difluoromethane, and having a surface ratio of silicon to aluminum
of less than about 1.7 mol/mol as determined by X-ray photoelectron
spectroscopy--has been found to be an effective desiccant for refrigerants
comprising difluoromethane.
Claims We claim:
1. A desiccant for use with a refrigerant comprising difluoromethane,
said desiccant comprising a potassium exchanged zeolite A having
greater than about 60 percent of available cations exchanged with
potassium and a clay binder, said desiccant having a water adsorption
capacity greater than about 15 weight percent, having essentially
no difluoromethane adsorption capacity, and having a surface ratio
of silicon to aluminum of less than about 1.7 mol/mol as determined
by X-ray photoelectron spectroscopy.
2. The desiccant of claim 1 wherein the difluoromethane adsorption
capacity is less than about 1.2 weight percent.
3. The desiccant of claim 1 wherein the difluoromethane adsorption
capacity is less than about 1.0 weight percent.
4. The desiccant of claim 1 wherein the difluoromethane adsorption
capacity is less than about 0.5 weight percent.
Description FIELD OF THE INVENTION
The present invention relates in general to the removal of water
from refrigerant mixtures containing difluoromethane (R-32) and
relates more particularly to the treatment of such mixtures employed
as circulating refrigerant streams of refrigeration systems to sequester
water as a means for avoiding freeze-ups and corrosion. The treatment
comprises adsorption of these impurities on a zeolitic molecular
sieve.
BACKGROUND OF THE INVENTION
In view of the now, well-established relationship between chlorofluorocarbons
(CFC's) released into the atmosphere and the depletion of the earth's
ozone layer, considerable attention is being directed to finding
effective substitutes for these once widely used compounds. It appears
that the worst offenders are the fully halogenated CFC's which contain
chlorine. These compounds are relatively unreactive with other compounds
in the lower atmosphere and thus are able to diffuse into the stratosphere
intact and be decomposed by ultraviolet radiation to form inter
alia, chlorine-free radicals which readily react with ozone. On
the premise that it is the chlorine constituent of the CFC's which
ultimately reacts with and destroys the ozone molecules and in the
interest of approximating as closely as possible the physical properties
of the CFC's already in use, the proposed substitutes in general
have been HCFC's containing lesser proportions of chlorine or fluorocarbons
containing no chlorine at all. For example, dichlorodifluoromethane,
widely used under the trademark Freon 12 as a refrigerant in household
refrigerators, in automotive units and in commercial freezers and
display cases, has been replaced in many instances by 1112-tetrafluoroethane
(also known as R-134a) or by chlorodifluoromethane (also known as
R-22 or HCFC-22). Because R-134a is not miscible with many commonly
used lubricants, mixtures of R-134a and R-22 have been proposed
for systems employing lubricants soluble in R-22. See U.S. Pat.
No. 5198139 (Bierschenk et al.) in this regard. In the recent
past, over 90 percent of the chlorodifluoromethane and about a third
of the dichlorodifluoromethane manufactured was utilized in air-conditioning
and refrigeration.
U.S. Pat. No. 3536521 (McKinney et al.) discloses a method of
preventing the adsorption of gases other than water, such as fluorinated
hydrocarbons, by the coating of Type A zeolite molecular sieves
with silicones such as methyl silicone. McKinney et al. further
discloses that the fluorinated hydrocarbons used in refrigerant
systems react at active sites on the surface of the molecular sieve
with subsequent decomposition into halogen acids which react with
the basic structure of the molecular sieve.
U.S. Pat. No. 5514633 (Noguchi et al.) discloses a method for
the production of a desiccant for HFC-32 and blends of HFC-32 which
comprises immersing a formed 3A zeolite which has been 20 to 60%
ion-exchanged with potassium in an aqueous solution of sodium silicate,
potassium silicate, or mixtures thereof to deposit SiO.sub.2 on
the formed 3A zeolite. The resulting formed 3A zeolite with the
SiO.sub.2 coating is dehydrated and subsequently activated. U.S.
Pat. No. 5514633 is hereby incorporated by reference.
U.S. Pat. No. 3625866 (Conde) discloses a process for preparing
composite desiccant materials for applications such as refrigerant
drying wherein the pores of the desiccant are less than about 4.9
angstroms in diameter to permit the inclusion of water molecules
and to exclude the larger halogenated hydrocarbon molecules. Conde
discloses the use of zeolite 3A which he describes as prepared by
replacing at least 65 percent of the sodium cations in zeolite A
with potassium cations by conventional cation exchange techniques
to adsorb molecules having critical diameters up to 3 angstrom units.
Conde hardens the zeolite A into an agglomerate optionally with
clay, silicates or both and then applies to the surface of the agglomerate
a thin coating of diaspore (hydrated alumina) and a clay mineral,
wherein the diaspore is the major component. The coated agglomerate
is then soaked in an aqueous solution of potassium silicate to impregnate
silicate into the agglomerate, dried in air to avoid steaming during
calcination, and finally fired to set the binder and activate the
molecular sieve. The refrigerant decomposition tests were based
on R-22. Conde discloses that clays which may be employed for bonding
molecular sieves without substantially altering the adsorptive properties
of the molecular sieve are attapulgite, kaolin, sepiolite, palygorskite,
kaolinite, plastic ball clays, clays of the attapulgite or kaolin
types, bentonite, montmorillonite, illite, chlorite, and bentonite-type
clay.
U.S. Pat. No. 5347822 (Cannan et al.) discloses the use of a
microporous zeolite molecular sieve having the crystal structure
of zeolite B and a framework silica-to-alumina molar ratio of at
least 2.5 for use in refrigeration systems containing R-32. Cannan
et al. discloses that the modified form of zeolite B has pore openings
small enough to significantly limit the adsorption of R-32 while
retaining a large capacity for water adsorption.
With increasing recognition of the seriousness of atmospheric ozone
depletion, stricter limitations on the future use of any chlorine-containing
refrigerant continue to be imposed. One of the most suitable replacements
for R-22 in stationary refrigeration systems is a non-flammable
mixture of the HFC compound difluoromethane, also known as R-32
with other halocarbons or halohydrocarbons such as R-134a and R-125
(C.sub.2 HF.sub.5). One such mixture known as R-410a has been proposed
and consists of 50 percent R-32 and 50 percent R-125. Another proposed
mixture (R-407c) consists of 23 percent R-32 25 percent R-125
and 52 percent R-134a. A significant problem in making this substitution
arises from the fact that R-32 is more readily adsorbed than R-22
with zeolite A, commonly employed as an adsorbent-desiccant in the
circulating refrigerant stream to protect against freeze-ups and
corrosion of the refrigeration unit. Ideally, a purified and dried
refrigerant fluid, after having been sealed in a refrigeration unit,
would continue to circulate through the compressors, Joule-Thompson
nozzles, cooling coils, etc., for the life of the unit without causing
any corrosion or freeze-up problems. In practice, however, the system
can rarely be so thoroughly sealed or the components so thoroughly
dried before sealing to prevent water and other contaminants from
entering the sealed system. These contaminating materials must be
removed or sequestered to avoid the development of the aforementioned
problems. Conventionally, the contaminants are rendered innocuous
by adsorption thereof on a suitable adsorbent which is inserted
into the sealed system and which is in contact with the circulating
refrigerant stream. In the case of halocarbon refrigerants, the
contaminants of greatest concern, in addition to water, are attributable
to the degradation products of the refrigerant molecules themselves.
Halogen acids, notably HCl, can form and cause corrosion. In some
instances, the adsorbent composition itself can be a reactant in
the chemical reactions which result in the production of corrosive
products. Zeolitic molecular sieves generally exhibit this property.
Accordingly, in view of the physical and chemical properties of
the refrigerant involved, it is necessary to select the particular
zeolite adsorbent to minimize harmful reactions. Since essentially
all of the active sites of a zeolite are reachable only by molecules
which can enter the internal cavities of the crystal structure through
its uniform pore system, it is advantageous to employ a zeolite
whose pore openings admit water and other small impurity molecules
and exclude molecules of the refrigerant. Thus, a commonly used
adsorbent for refrigeration systems is a highly exchanged potassium
cation form of zeolite A having pore diameters of about 3 angstroms.
The effective pore diameters can be further reduced, to a slight
degree, by controlled steaming as disclosed in U.S. Pat. No. 3506593
hereby incorporated by reference. A potassium cation exchanged (40
percent) form of zeolite A, i.e., zeolite 3A, has been found to
be quite effective in drying R-134a and R-22 for example.
R-32 however, is both smaller in molecular size and more polar
than R-22 by virtue of the substitution of a hydrogen atom for the
chlorine atom in chlorodifluoromethane. It is also more reactive
than R-22 with constituents in the lower atmosphere and thus, advantageously,
is less likely to escape unreacted into the stratosphere. It is,
by the same token, more reactive with zeolites, including zeolite
3A, having pores large enough for R-32 to enter. The greater polarity
of R-32 also means that the partial blocking of zeolite pores by
cation exchange techniques is less effective in excluding the R-32
from the inner cavities of the zeolite crystal structure. Adsorbent
aggregates are sought which minimize reactivity with difluoromethane
without reducing the capacity of the adsorbent to adsorb water for
use as desiccant in closed-cycle refrigeration systems.
When a molecular sieve adsorbs the refrigerant molecule, there
is a much greater chance of chemical interaction between the refrigerant
and desiccant, and such interaction will result in the chemical
destruction of both. By the term "destruction of both,"
it is meant that the refrigerant may become decomposed into undesirable
products and that the desiccant can lose significant water adsorption
capacity. Also, when the refrigerant is adsorbed, it is using up
some of the desiccant's capacity for water adsorption. Finally,
the composition of a refrigerant blend will be changed if the smaller,
more polar refrigerant (like R-32) is selectively adsorbed and removed
from the system. All of this is avoided by excluding the refrigerant
from the desiccant.
SUMMARY OF THE INVENTION
The desiccant of the present invention, which excludes difluoromethane
refrigerant (and lubricants and additives) while adsorbing water,
provides the following benefits in refrigeration systems: chemical
compatibility with the R-32 refrigerant, high water capacity (greater
than about 15 weight percent), and maintenance of refrigerant blend
composition.
According to the present invention, in a refrigeration process
wherein a refrigerant fluid comprising difluoromethane (R-32) is
recycled within a closed system and is alternatively vaporized and
condensed in a manner to produce cooling, the improvement comprises
incorporating within the closed system containing the difluoromethane
and in contact therewith as a desiccant an activated potassium cation
form of zeolitic A molecular sieve agglomerated with a clay binder
and pore closed, the desiccant having a water adsorption capacity
of at least 15 weight percent, essentially no reactivity with difluoroethane
and having essentially no capacity for the adsorption of difluoromethane.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows R-32 adsorption at 67 kPa.
FIG. 2 shows water adsorption at about 0.6 kPa.
FIG. 3 shows R-32 adsorption as a function of calcination temperature
of the desiccant.
FIG. 4 shows water adsorption as a function of calcination temperature
of the desiccant.
DETAILED DESCRIPTION OF THE INVENTION
Zeolitic molecular sieves in the calcined form may be represented
by the general formula:
where Me is a cation, n is the valence of the cation, x has a value
from about 2 to infinity, and y has a value of from about 2 to 10.
Typically, well known zeolites include chabazite (also referred
to as zeolite D), clinoptilolite, erionite, faujasite (also referred
to as zeolite X and zeolite Y), ferrierite, mordenite, zeolite A,
and zeolite P. Detailed descriptions of the above-identified zeolites,
as well as others, may be found in D. W. Breck, Zeolite Molecular
Sieves, John Wiley and Sons, New York, 1974 hereby incorporated
by reference.
The preferred zeolite for use in the present invention is a zeolite
A and, more preferably, a zeolite A which has been ion-exchanged
with potassium to reduce the pore opening. Preferably, the zeolite
A of the present invention has been ion-exchanged with potassium
by conventional ion-exchange techniques such as immersing a sodium
zeolite A in an aqueous solution of one or more salts of potassium.
Such potassium salts include halides, especially the chlorides,
nitrates, sulfates, and hydroxides. The aforementioned salts are
by no means the only suitable materials and many other salts of
potassium normally used in ion-exchange procedures are well known
to those skilled in the art.
It is preferred that the zeolite A molecular sieve of the present
invention be exchanged with potassium to a high level. Preferably,
the potassium exchanged zeolite A of the present invention comprises
greater than 40 percent of the available cations replaced with potassium
cations. More preferably, the zeolite A of the present invention
comprises greater than 60 percent of the available cations exchanged
with potassium cations. Most preferably, the zeolite A of the present
invention comprises between 60 percent and 80 percent of the available
cations exchanged with potassium.
Desiccant compatibility is one of the key areas for testing new
refrigerants. A new refrigerant must be compatible with the lubricant,
metals, elastomers, motor insulation, and the desiccant used in
the system. Desiccant compatibility testing comprises placing desiccant,
refrigerant, and often lubricant in a stainless steel vessel for
a given time at an elevated temperature. Chemical, physical, and
adsorptive properties of the desiccant are determined before and
after exposure to the refrigerant. The compatibility of various
desiccants can then be ranked according to the magnitude of the
changes in these properties. Although changes in all properties
have been observed with prolonged exposures of incompatible desiccants
to fluorocarbon refrigerants, it was found that fluoride content
is the most sensitive indicator of reaction between the desiccant
and a fluorocarbon refrigerant. Elevated levels of fluoride (and
chloride) detected after short exposure times tend to predict later
degradation in other desiccant properties. Therefore, fluoride content
is used as a measure of the relative compatibility of various desiccant
samples with alternative refrigerants. Lower fluoride content is
taken as an indication of lower reactivity and greater compatibility
of desiccant with refrigerant.
The compatibility test method consists of several steps. The molecular
sieve sample is usually prepared for testing by thermal activation
to remove adsorbed water. The activated desiccant sample is loaded
into a stainless steel cylinder, the appropriate lubricant is added,
a vacuum (<10 mm Hg [<1.3 kPa]) is pulled, and the refrigerant
is charged. The cylinder is then aged in an oven at a particular
temperature for a particular time. At the end of that time, the
cylinder is removed from the oven and cooled. The refrigerant and
lubricant are removed from the cylinder and then the molecular sieve
is removed. The desiccant sample is then conditioned to prepare
it for subsequent analysis of fluoride and chloride content, adsorption
properties, and physical properties. The sample is ground to a powder,
spread in a thin layer, and exposed to air at ambient conditions
for a period of about 48 hours. The sample is then heated in a programmable
oven with a nitrogen purge from ambient temperature to about 288.degree.
C. to remove volatiles and stabilize the sample. The sample is rehydrated
in air at ambient conditions prior to fluoride testing.
It is believed that the fluoride uptake, i.e., the increase in
inorganic fluoride content of the desiccant after the compatibility
test exposure, is an indication of the reactivity of desiccant with
refrigerant. It is believed that the fluoride thus determined had
reacted with the aluminosilicate framework of the zeolite molecular
sieves to produce silicon and aluminum fluoride salts. In the present
invention, it is preferred that the desiccant exhibit essentially
no reactivity with the R-32 refrigerant. By the term "essentially
no reactivity" with the refrigerant, it is meant that the fluoride
uptake, measured according to the above procedure following exposure
to the refrigerant at 75.degree. C. (167.degree. F.) for seven days,
is less than about 0.1 weight percent, and more preferably, the
term "essentially no reactivity" means that the fluoride
uptake measured at the above conditions is less than about 0.05
weight percent.
In addition, the desiccant of the present invention has essentially
no capacity for the adsorption of R-32. By the term "essentially
no capacity" for the adsorption of R-32 it is meant that the
refrigerant loading on the adsorbent as measured in McBain-Bakr
adsorption apparatus at room temperature and at refrigerant partial
pressure of about 66.5 kPa (500 torr) is less than about 0.5/100
gm, and preferably the refrigerant (R-32) loading of the adsorbent
is less than about 0.3 g/100 gm. It was found that the adsorbent
of the present invention showed essentially no capacity for the
adsorption of CO.sub.2 at room temperature and a pressure of about
33 kPa (250 torr).
For use as a desiccant-adsorbent in R-32 refrigerant streams, the
high potassium exchanged zeolite A crystals are agglomerated into
engineered forms to avoid entrainment in the stream and plugging
of orifices and conduits and abrasive damage to the refrigeration
system. While compaction to create self-bonding of the crystal particles
is possible, it is advantageous to utilize binder materials to create
agglomerates of high attrition resistance. It has been determined,
in this regard, that the choice of binder material and the cation
can be important factors in inhibiting the reactivity of the R-32
with the adsorbent agglomerates and contributing to overall compatibility.
Sepiolite, bentonite, and attapulgite were found to show greater
basicity than halloysite, avery and metakaolin clays. Volclay showed
no basicity. For example, clays exhibiting significant degrees of
basicity react with R-32 in essentially the same manner as the basic
NaA zeolite. Ideally, the clay binder should be neither basic nor
acidic and should require modest calcination temperatures to be
set. Unfortunately, no commercially available clay has been found
to possess all of these properties. Without wanting to be bound
by any particular theory, it is possible that R-32 reacts with basic
zeolitic aluminosilicates, such as zeolite NaA, according to the
following equations:
Dealumination
Presumably, similar reactions occur between CH.sub.2 F.sub.2 and
the various clay compositions commonly used as binder material.
These reactions do not, however, occur with equal facility among
the various clays. For example, in an experimental procedure wherein
liquid CH.sub.2 F.sub.2 was contacted with a sample of avery clay
[ideal formula Al.sub.2 (Si.sub.2 O.sub.5) (OH).sub.4 ] and a sample
of attapulgite clay [ideal formula Mg.sub.5 Si.sub.8 O.sub.20 (OH).sub.2.
8H.sub.2 O ] at 75.degree. C. and the vapor pressure of R-32 at
that temperature (.sup..about. 53 atmospheres) for 7 days, post-treatment
analysis of the avery clay indicated a 0.67 weight percent fluoride
content versus a 1.23 weight percent fluoride content for the attapulgite.
Sepiolite and halloysite appear to resemble attapulgite and avery
clay, respectively, in their reactivity toward CH.sub.2 F.sub.2.
The firing temperatures necessary to achieve adequate bonding exceed
550.degree. C. for halloysite and avery clays. Preferably, the clay
binder of the present invention is selected from the group consisting
of attapulgite, kaolin, volclay, sepiolite, halloysite, palygorskite,
ball clays, bentonite, montmorillonite, illite, and chlorite clays.
More preferably, the clay binder of the present invention is selected
from the group consisting of attapulgite, kaolin, volclay, sepiolite,
and halloysite clays. Sepiolite and attapulgite are from the palygorskite
family of clays; whereas bentonite and volclay are montmorillonites;
and the halloysites and kaolin are kaolinites.
The preferred desiccant of the present invention comprises a high
potassium exchanged 3A zeolite and at least 20 percent clay binder.
The solid adsorbent and binder may be mixed in the slurry in a variety
of proportions, whether as segregated or agglomerated particles,
although it is generally desirable to use only as much binder as
is required to give sufficient strength to the coated surface. It
is preferred that the adsorbent portion comprises about 70 to 95
weight percent of the total weight of solids, i.e, adsorbent and
binder, in the slurry and that the remaining 5 to 30 weight percent
comprises binder. In addition to the adsorbent and binder, the slurry
may contain a dispersing agent or surfactant to aid in suspending
the particles or vary the viscosity of the slurry. Suitable surfactants
include for example, a condensed tannin; Dispex, a salt of a polymeric
carboxylic acid available from Allied Colloids, Suffolk, Va.; and
TSPP, a tetrasodium pyrophosphate available from Monsanto, St. Louis,
Mo. When a surfactant or dispersing agent is used, it is preferred
that its concentration be in the range of about 0.5 to 5.0 weight
percent of the solids content of the slurry.
The zeolite of required chemical composition having been thus attained,
it is an essential step that the zeolite crystals and binder as
a desiccant agglomerate be contacted with water vapor at an elevated
temperature for a time sufficient to reduce its adsorptive capacity
for R-32 to essentially no capacity for adsorption of R-32. Preferably
the desiccant agglomerate of the present invention is pore-closed
to reduce the adsorption of R-32 by contacting the agglomerate in
the presence of steam or water vapor at a temperature greater than
about 550.degree. C., more preferably the desiccant agglomerate
is pore closed in the presence of water vapor at a temperature greater
than about 600.degree. C., and most preferably the desiccant agglomerate
is pore closed in the presence of water vapor at a temperature between
about 600.degree. and about 800.degree. C. A liquid phase R-32 adsorption
test was developed to measure adsorption in refrigerant desiccants.
The test comprised exposing a 5 g (0.011 lb) sample of activated
molecular sieve to 40 g (0.099 lb) of liquid R-32 at 52.degree.
C. (125.degree. F.) for a period of about 16 hours in a 75 ml (0.0198
gal) stainless steel cylinder. The cylinder is then cooled to room
temperature and the refrigerant is removed and the saturated desiccant
is recovered. The saturated desiccant is then analyzed to determine
the adsorption of R-32 using a LOI-KF procedure which measures adsorbed
water and adsorbed volatiles (VOC) simultaneously. In a combined
Loss On Ignition/Karl Fischer (LOI-KF) procedure, an initial weight
of about 1 g (0.0022 lb) of the desiccant sample is weighed on an
analytical balance in a quartz boat. The sample is inserted at room
temperature into a furnace which is purged with dry nitrogen. The
furnace is heated to about 950.degree. C. (1742.degree. F.). Any
volatiles evolved from the desiccant sample are bubbled through
the reaction vessel containing Karl Fischer reagent (a solution
of iodine, sulfur dioxide, and pyridine in methanol) of a volumetric
Karl Fischer titrator for the determination of water. The water
content of the evolved gas is extracted by the methanol in the Karl
Fischer reactor vessel and automatically titrated. When the end
point of the titration is reached, the furnace is switched off and
the sample is allowed to return to about room temperature in the
furnace. The final weight of the sample is then determined. The
difference between the final and initial weights of the desiccant
sample provides the weight Loss On Ignition (LOI), which is a measure
of the total volatiles in the sample. The mass of water in the volatiles
is proportional to the volume of Karl Fischer reagent used by the
titrator. Since the only volatiles on the desiccant are R-32 and
water, the adsorption of R-32 is calculated as the difference between
the total volatiles (LOI) and the Karl Fischer water content.
Desiccants 3A-6 and 3A-9 are commercial molecular sieves (available
from UOP, Des Plaines, Ill.) commonly used in stationary refrigeration
facilities. The R-32 liquid adsorptions for 3A-6 and 3A-9 as shown
in Table 1 are about 15 g/100 g at 125.degree. F. (52.degree. C.).
The desiccant of the present invention effectively excludes R-32
as indicated by its essentially no capacity for R-32 liquid adsorption,
the desiccant of the present invention having adsorbed only about
0.2 g/100 g of R-32.
TABLE 1 R-32 Adsorbed, Desiccant g/100 g (lb/100 lb) 3A-6 15.6
3A-9 14.5 Present Invention 0.2
Although the data in Table 1 was determined for pure R-32 based
on these results, the adsorption of R-32 from blends such as R-407
and R-410 series refrigerants containing substantial concentrations
of R-32 should similarly be at very, very low levels. Thus, the
desiccant of the present invention has essentially no capacity for
the adsorption of R-32.
The temperature of the water vapor (steam) and the time of contact
between the water and the zeolite depend in large measure on such
factors as the crystal size of the zeolite product, the quantity
of zeolite charge, and the effectiveness of the contact. Preferably,
temperatures in excess of about 550.degree. C. are employed over
a period of about 20 to 90 minutes in air at a steam content ranging
from about 10 to about 60 mole percent steam. It is more preferred
that such steaming take place in a furnace or kiln in an air atmosphere
at temperatures ranging from about 600.degree. to about 800.degree.
C. In any case, an unduly rapid loss of water with attendant inadequate
contact time of water with zeolite should be avoided. For any given
physical state and composition of zeolite agglomerate, one skilled
in the art can readily determine by periodic measurement when sufficient
pore closure has been attained.
According to the present invention, it was discovered that by applying
conventional pore-closing techniques, particularly hydrothermal
treatment at elevated temperatures, to an engineered agglomerate
comprising a highly exchanged potassium form of zeolite A and a
select group of clay binders, the resulting modified zeolite A molecular
sieve desiccant surprisingly retained a significant capacity for
the adsorption of water while essentially excluding the adsorption
of difluoromethane. A paper entitled, "Inorganic Fluoride Uptake
As A Measure of Relative Compatibility of Molecular Sieve Desiccants
with Fluorocarbon Refrigerants," by Alan P. Cohen and C. Scott
Blackwell, published in the American Society of Heating, Refrigerating
and Air Conditioning Engineers, Inc. Transactions 1995 Volume 101
and presented on Jun. 26 1995 at the Annual ASHRAE Conference in
San Diego, Calif. discloses a desiccant compatibility test for measuring
the reactivity of zeolite desiccants and R-32 refrigerants. The
above article is hereby incorporated by reference. Surprisingly,
similar compatibility tests carried out on the desiccant of the
present invention indicated that even though the new desiccant was
made from a zeolite A base, the reactivity of the new desiccant
with difluoromethane was essentially nil, and the adsorption of
fluoride from R-32 was less than about 0.5 weight percent and surprisingly,
in some cases, was less than about 0.3 weight percent. It was determined
by conventional means that the water adsorption capacity of the
desiccant of the present invention was in excess of 15 weight percent
EXAMPLES
The invention will be further clarified by a consideration of the
following examples, which are intended to be purely exemplary of
the use of the invention.
R-32 Adsorption Procedure Employed In Examples
In order to verify the suitability of the clay bonded potassium
exchanged Zeolite 3A adsorbent for use in the present process, refrigerant/desiccant
compatibility tests were carried out in which a mixture of liquid
R-32 and a polyolester lubricant were contacted with the adsorbent
in a stainless steel bomb. In carrying out the tests, 10 grams of
1/16" activated pellets of the adsorbent having a water adsorption
capacity greater than about 15 weight percent were first added to
the bomb followed by the injection of 10 grams of the lubricant
and lastly by a charge of 10 grams of the R-32 refrigerant in the
liquid phase. Air was evacuated from the bomb after adding lubricant
but before charging with refrigerant. The temperature of the sealed
bomb was raised to 75.degree. C. and retained in that state for
seven days. Thereafter, the adsorbent was recovered and adsorbed
R-32 refrigerant was removed from the zeolite pores by first grinding
the zeolite particles, permitting the ground solids to hydrate in
open air for a period of about 11 days. The residual fluorine content
of the hydrated solids was determined as described hereinabove as
a measure of the degree which the R-32 reacted with the adsorbent.
Example I
A series of samples of clay bonded high potassium (80 percent)
exchanged zeolite 3A was prepared with a variety of clay binders.
The clay binders included sepiolite, attapulgite, halloysite, halloysite,
volclay, and bentonite clay. The zeolite level was 80 percent and
the binder level was 20 percent. The samples were formed by mulling
the zeolite with the clay in an aqueous medium, extruding the mulled
mixture and drying the extrudate at 100.degree. C. to achieve an
LOI in the range of about 17.4 to 20.1 weight percent at 1000.degree.
C. About 10 grams of each of the clay bonded samples was steam calcined
in a horizontal steamer at temperatures ranging from about 500.degree.
to about 800.degree. C. while 10 percent steam in air was passed
over the sample for about one hour. The R-32 adsorption at about
67 kPa (500 Torr) results are summarized in FIG. 1 and the water
adsorption at about 0.6 kPa (4.6 Torr) results are summarized in
FIG. 2. The clay bonded zeolite samples of sepiolite at 650.degree.
C., volclay at 700.degree. C., halloysite at 725.degree. C. and
bentonite at 750.degree. C. had water capacities ranging from 17
to 19.5 weight percent while effectively excluding adsorption of
R-32 or reducing the amount of R-32 to about 0.99 percent for sepiolite,
0.67 percent for volclay, 0.55 percent for halloysite, and 0.39
percent for bentonite at the above steam calcination temperatures.
The halloysite lost water capacity with increasing steam calcination
temperature.
Example II
A second series of samples of clay bonded 80 percent high potassium
exchanged zeolite 3A with 20 percent clay binder was prepared in
the manner of Example I in 500 gram quantities for sepiolite, attapulgite,
and bentonite clays. The samples were calcined over a range of temperatures
from about 550.degree. to about 750.degree. C. using 20 mole percent
steam in air for about 1 hour. The R-32 and water adsorption results
as a function of calcination temperature are shown in FIGS. 3 and
4. Although the adsorption of R-32 was reduced with increasing hydrothermal
treatment, the sepiolite and attapulgite materials lost water capacity.
Example III
A 500-gram sample of halloysite clay bonded high potassium exchanged
3A zeolite prepared according to the manner of Example I was hydrothermally
treated at 625.degree. C. and 725.degree. C. with 10 mole percent
steam in air in a vertical steamer for about 1 hour. Unexpectedly,
the halloysite-bonded material exhibited the following adsorption
results:
Temp., .degree. C. R-32 wt-% Water, wt-% 625 2.55 17.27 725 0.2
13.80
Example IV
A 500-gram sample of a volclay bonded high potassium exchanged
3A zeolite prepared according to the manner of Example II was hydrothermally
treated with about 10 mole percent and 15 mole percent steam in
air at about 650.degree. C. and about 700.degree. C. for about 1
hour. The volclay-bonded material showed very low adsorption of
R-32 and consistent water capacity as shown below:
Temp., .degree. C. Steam, wt-% R-32 wt-% Water, wt-% 650 10 0
17.96 650 15 0 17.93 700 10 0.35 18.56 700 15 0 18.48
Example V
A potassium exchanged zeolite A (77 parts), having greater than
about 60 percent of the available cations exchanged with potassium,
was combined with a clay binder comprising Koalin clay (18 parts),
Halloysite clay (5 parts) and a dispersant (0.5 parts) and finely
pulverized with a vibrating ball mill and thereto was added water
(as an aqueous medium, 31 parts) and the mixture was well kneaded
with agitation with a kneader to give a potassium exchanged zeolite
A composition. The composition thus obtained was charged to a mixer
agglomerator to form green beads. The green beads were screened
to obtain 6.times.10 beads. The 6.times.10 beads were air dried
for about two days at a temperature of about 120.degree. C., and
thereafter the 6.times.10 beads were charged to a rotary kiln with
the kiln operating at a calcination temperature ranging from about
550.degree. to about 800.degree. C., in the presence of steam. The
steam was introduced at a rate equal to about 40 mole percent of
the air rate. Following calcination, the 6.times.10 bead product
was cooled.
The product thus obtained had the following properties:
1000.degree. C. LOI 1.17 Bulk Density 56.0 Piece Density 1.43 Crush
Strength 15.4 R-32 Adsorption 0.35 wt-% Water Adsorption 18.4 wt-%
Example VI
A sample of the potassium exchanged zeolite A desiccant of the
present invention was compared to desiccants of the prior art which
employed silicate treatment to achieve a reduction in the adsorption
of difluoromethane while maintaining a reasonable water adsorption
capacity. The samples included in this analysis of the desiccant
surface were: A-a silicate coated 2 mm bead prepared according to
U.S. Pat. No. 5514633 B-a silicate coated 2 mm bead consistent
with the method disclosed in U.S. Pat. No. 3625866 (W. R. Grace,
Baltimore, Md.), and C-a 2 mm bead of the present invention. The
following table summarizes the results of an analysis of the surface
of the particle by photoelectron spectra. Photoelectron spectra
were obtained on a Physical Electronics Quantum 2000 Scanning ESCA
Microprobe spectrophotometer (available from Physical Electronics,
Inc. of Eden Prairie, Minn.) using a monochromatic aluminum K-alpha
x-ray source (1486.6 eV) and standard procedures familiar to those
versed in the art. As is well known to those versed in the art,
the method of XPS is sensitive only to roughly the first 50 angstrom
units of the samples and thus only measures surface concentrations
of elements.
Desiccant A B C R-32 Adsorption, wt-%, 16 hr @ 52.degree. C. 5.5
0.4 1.3 Aluminum, mol-% 9.03 6.56 11.48 Silicon, mol-% 15.60 15.58
13.53 Surface Si/Al, mol/mol 1.73 2.38 1.18
The results of the XPS analysis clearly show that the desiccant
C of the present invention has essentially no adsorption capacity
for difluoromethane and has a surface ratio of silicon to aluminum
at the surface of the particle which is less than about 1.7 mol/mol.
Desiccants A and B which were coated to reduce the tendency of the
zeolite to adsorb the difluoromethane were shown to have significantly
higher surface ratio of silicon to aluminum as determined by X-ray
photoelectron spectroscopy. The term "essentially no capacity
for the adsorption of difluoromethane" as used herein means
that preferably, the adsorption capacity for difluoromethane is
less than about 1.2 weight percent, more preferably the adsorption
capacity for difluoromethane is less than about 1.0 weight percent,
and most preferably the adsorption capacity for difluoromethane
is less than about 0.5 weight percent. |