Structure and Moisture Permeabilitv of Film-Forming olymers J
Textile Fibers D e p a r t m e n t , Pioneering Research Division, E'. I . d u P a n t de ,Ventours & Co., I n c . , W i l m i n g t o n , Del.
T""
nioisture permeability of a wide variety of polymeiic film-forming materials has been determined a t 39.5" C. and a vapor pressure difference across the film equivalent t o 53 mm. of mercury. The data indicate t h a t most polymers possessing low moisture permeability have (1) .I saturated or nearly saturated carbon chain (2) A minimum of chain branching (3) h high degree of lateral symmetry (4) -4f d r degree of longitudinal symmetry (5) A very high proportion of relatively small, ~ionhvrlrophilic substituent8 Physical factors, such as orientation and crystallinity, may also contribute to reducing permeability. A high degree of conipliance with most of these requirements may mask the 1:tcb of conformity in some one respect, such as unsaturation. S a t u r a l and synthetic high polymeric substances aie much used as protective barriers in the form of self-supporting films, coatings on fahrics, and paint and lacquer bases. I n these forms they protect objects against corrosive gases and moisture, or they may prevent the loss of moisture from a packaged substance or even the loss of the substance itself. Ideally most food packages must bar the rapid transmission of moisture in both directions, prevent the loss of natural taste and odor characteristics, and keep out foreign odors and tastes Today there are a multitude of polymeric materialq, many of which are p m n t i a l l y self-supporting films. Because of the availability of numei ous polymerizable monomers, the possibility of obtaining in a polymeric material many of the desirable mechanical characteristics, together with specific permeability characteristics, is readily visualized. I n order to do this in the most direct manner, a n extcnsion of our knowledge of the permeability characteristics of polymers in relation to their chemical and physical structure is required. Although there have been many reports on methods of testing permeability (8, 12, 36) and many discussions of the mechanism of permeability through filmP (6,16, 17, 24, SO, 34, 38, 40, 45), attempts to relate the degree of moisture permeability with thc chemical structure have been based upon a limited number of polymera. It has long been kno.ivn t h a t hydrophilic polymers, such as regenerated cellulose, are very permeable to moisture (13, 23, S I ) while hydrophobic materials, such as chlorinated rubber (7), are much less permeable. Metal foils, which are highly crystalline substances, are almost impermeable t o water vapor. Measurements have shown t h a t films of gutta percha are less permeable than natural rubber (5, 44). X-ray study has revealed t h a t these substances have different chain configurations and that gutta percha is crystalline a t ordinary temperatures while rubber is not (11, 68). The permeability of gutta percha to gases rises at the melting point (45' C.) t o that of rubber (44). Aiken, Doty, and Mark (6, 16) have studied the moisture permeability of a variety of polymers and have concluded t h a t the degree of crystallinity and the hydrophilic or hydrophobic character of a polymer determine its ultimate water vapor permeability. Reitlinger ($3) has related the hydrogen permeability of films t o their chemical structure and consequent microstructure. Molecular weight did not affect permeability. Brubaker and Kammermeyer (10) likewise did not find any correlation of the molecular weight of polyethylene with permeability to carbon
dioxide or hydrogen. Hydrophilic character of polar groups ( 3 5 ) decreased permeability. The latter increased crystallinity. Both irreffularity in the primary valence chain and large side groups or branching increased permeability by holding the chains apart and decreasing intermolecular forces. Simril and Hershberger (38) have studied the permeabilitj- of 10 polymeric films to organic vapors and have correlated the results with the polymer structure. The least permeable wei'e found to be those whose molecular structure was such as to permit close packing and strong intermolecular bonding and whose adsorption or solution of vapor was low. I n another study ( 3 7 ) they reported the permeability oi 21 films tu gases. Side chairis, plasticizers, and water were found t o increase permeability. The present work was undertaken to define further the rules by which one may predict thc moisture permeabilit,y of a given polymer in film form and t o diacover new polymeric materials with very low moisture permeability, EYPERIMENTA L
Preparation of Films. The majority of polymers studied M C L C converted into thin films by solvent casting. The polymer flake was dissolved with stirring and heating in a suitable solvent to make a 10 t o 20% solution by weight. The concentration used depended upon the solubility of the polymer and the viscosity of the resulting solution. When a uniform solution was obtained, it was stored until free of air bubbles in a closed bottle in an oven, Shich was kept at least 15a below the boiling point of the solvent. The deaerated solution was spread onto a heated chromiumplated surface with a doctor knife of such a gage to give a final dried film approximately 0.0015 inch t o 0.0020 inch in thicknew. Initial drying was carried out in still air with the temperature of the plate above the gelling point of the solution but at least 10 below the boiling point of the solvent. When the surface of the film was firm, the drying was completed with a current of hot ail, I n general, the films were stripped from a cold plate. No qtripping agents were incorporated in the films, For many materials special mixing and casting techniques weie necessary. Rubber solutions, for example, were prepared by cutting the gum stock into small pieces, soaking them overnight in the solvent, and then beating the swollen mass for 10 minutes in a Waring Blendor. This type of mixer was found t o be very efficient for quickly making small amounts of polymer solutions. Polymers which produced rubbery and tacky films required light talcing on both surfaces during the stripping operation to prevent the surfaces from sticking together. Some polymers which were either too plastic or too brittle t o remove from the plate as selfsupporting films were cast on top of a base sheet of unplasticized polyvinyl alcohol or uncoated cellophane (69). A few of these were then stripped from the base sheet. Measurement of Moisture Permeability. The test method described by Charch and Scroggie ( I S ) was used; it consists, essentially, of the following steps: Upon a sheet of gasket rubber covered with several layeis of paper, is placed a 4-inch square film sample. I n the center of this a 1-inch square of chip board and an aluminum crucible (2.5em. 0.d. X 3-cm. deep) containing 12 t o 15 ml. of water is placed. The rim of an aluminum permeability cup (5-cm. deep and 7cm. i.d.) is dipped in beeswax a t 140 t o 170" C. and placed over the crucible of water so that a tight seal is made with the film sample. The cup is held in place for 1 minute with a 1-pound weight. The cup and film are then inverted quickly without wetting the film, and the excess film is trimmed off. The cup is cooled t o room temperature, weighed using a counterpoise sealed with an aluminum foil, and placed with other samples in the permeability cabinet. This cabinet is maintained at 39.5 f 0.5" C., and substantially dry air (1% relative humidity) is circulated about the cups, 2296
October 1953
INDUSTRIAL AND ENGINEERING CHEMISTRY
+he loss in weight from the cups is determined every 24 hours after cooling t o room temperature. The permeability of the film may be calculated as follows: Observed loss in milligrams X 1000 = permeability in (Time in hours) X (area exposed in grams/100 m.2/hour For cups described and a 24-hour test, this result is obtained by multiplying 1.25 times the weight lost in milligrams.
I \
100
. IL 0-
0 X
I
I
0.001
FOUND CALCO. FROM O.DOI*
-CALCD. FROM 0.0016‘
0 002
THICKNESS IN
0 003
INCHES
Figure 1. Moisture Permeability versus Thickness of Vinyl Chloride-Diethyl Fumarate (95/5) Film This test provides a vapor pressure difference across the film equivalent to 53 mm. of mercury. The tests were run for 72 hours to obtain equilibrium conditions, and the results are reported as averages of two tests on the third 24-hour period. The initial permeability value indicates the permeability of unhandled film, and a second value was obtained upon films crumpled once by hand in a rather standard manner. The latter test indicates how easily damaged the film is, and a rise in permeability is usually a reflection of stiffness and brittleness
2297
been disregarded. Reitlinger (33) has shown that the niolecular weight of polystyrene and polyisobutylene has no bearing on hydrogen permeability. Very low molecular weight fractions, however, would be expected to behave like plasticizer molecules. The films have been regarded as homogeneous in microstructure though this is shown t o be false in some cases. The existence of chain branching has been recognized but has been ignored for the most part because it is not a readily determined property Mechanism of Permeation. Permeation of moisture vapor through organic polymeric films is generally assumed to take place by diffusion through micropores in the p o l p e r structure, by solution at one face of the film and evaporation at the other face, or b y a combination of both processes (Pd, 18, f7, 40). When the polymer is hydrophilic, the bonding forces between polymer molecules and water must be satisfied by a saturating layer or layers of water before vapor can escape from the dry side. No attempt will be made here to divide permeation into factors ot diffusion and sorption (16,34,4l). Many polymeric materials have the ability to crystallize, but this crystallization is usually quite incomplete (19, R8). The remainder of the structure is a tangled amorphous mass of molecular chains. A single molecule may be part of both the crystalline and amorphous regions. Amorphous polymers also may be ordered or oriented by mechanical drawing of the structure. When polymers are deposited from an evaporating solvent a t elevated temperatures and the films are cooled slowly, a random arrangement of crystallites interspersed with amorphous regions results. The degree to which the molecular “straw-pile” collapses upon itself in a dense structure will depend upon the strength of the forces which cause it t o organize into crystalline areas, the flexibility of the polymer chains, and the presence of interfering substituents and branrhes. j 1
RESULTS
P
The moisture permeabilities of over 100 film-forming polymers were determined at thicknesses in the range of 0.001 t o 0.003 inches. The data are shown in Table I1 under a number ok headings based on polymer structure. For the purpose of comparison the initial permeability values were calculated by inverse proportion into permeability values a t a common thickness (0.0020 inch). The average thickness of the films tested was 0.00175 inch. ,Such a calculation gives an approximately correct permeability value for most films, since the plot of permeability versus film thickness is hyperbolic and is therefore a simple inverse proportion (See Figures 1 and 2). This relationship holds best for thicknesses between 0.0013 and 0.0025 inches and for nonhydrophilic polymers (34, 37). For thinner films the permeabilities calculated for 0.002 inch are too low, and for thicker films they are too high (Figure 1). This should be borne in mind in considering the calculated values given in the tables. The results reported for a film thickness of 0.002 inch may be converted to the permeability constant ( P X 108)used by Aiken, Doty, and Mark ( 2 ) by simply multiplying by a factor of 0.0332. This so-called constant will be in terms of milliliters of vapor a t standard conditions transmitted per square centimeter per second per millimeter of thickness per centimeter of mercury pressure. DISCUSSION
For the purpose of discussion certain simplifying assumptions have been made. The individual molecules of polymers and copolymers are assumed to be chemically uniform as t o the distribution and arrangement of the monomeric units. Possible differences in molecular weight and weight distribution Lave
t
\
z IO00
I
.-9 THICKNESS
IN
I
I
POLYVINYL BUTYRAL (11% FREE HYOROXYLS)
I
-1
INCHES
Figure 2. Moisture Permeability versus Thickness of Polyvinyl Alcohol and Polyvinyl Butyral Films The permeability of the resulting mass will depend partly upon the density and organization of the polymer molecules and partly upon the size of the small permeant molecule and the attractive forces between it and the polymer molecule. Permeation will take place through small and changing pathways in the continually agitated amorphous portions in the polymer. Spacings in crystal regions are too small for the passage of a permeant molecule such as water, and the crystal forces will not be overcome unless the permeating substance is a solvent for the polymer under the conditions of the test. The relation between permeability a t a given temperature and other factors affecting it is often expressed by some form of the equation Q = -PAtAp
1 where Q is the amount of water passing through a film of area, A , and thickness, I, for a time, t , and vapor pressure difference
INDUSTRIAL A N D E N G I N E E R I N G CHEMISTRY
2298
across the film of A p . P , a proportionality constant, is called the permeability constant and is directly proportional to the eaSe with which gaqes and vapors pass through the film. There are many factors affecting permeability which make it more coinplex than this simple relationship indicates. I n this work pernieabilities have been calculated to a conimon thickness, area, time, and vapor pressure difference. These values are related to the permeability or diffusion constants calculated by other n-orkers from the above equation by a simple factor which depend. upon the units in which P is expressed. In the follo~vingdiscussion quoted numerical permeahilitie.; are for a film thickness of 0.002 inch unless otheri3ise indicated. Permeability Test Method. The test method used is a severe and practical one. A temperature of 39" C. is not unduly high, but the permeabilities obtained a t this temperature will be higher than those reported by others a t loner temperatures for permeability is proportional to temperature (1,2, 16, 40, 41). The high humidity and vapor pressure difference used are higher than are likely to be found in most film uqes. Hydrophilic materials show up to great disadvantage under these conditions. However, many tests under milder conditions give a false impression oi the relative permeability of such mateiials. The cup method of testing has been criticized frequently, because of faulty and irregular seals. By the method used uniform seals 2 to 3 mm. in width are obtained. A skilled operator obtains very reproducible areas of exposed film. The width of the seal is so much greater than the thickness of the film that there is no edge effect for nonlaminated films. Errors due to occasional poor seals and pin holes in the films were practically eliminated by rejecting all data which vias inordinately high. Polymeric Films Having Low Moisture Permeabilities. Of all the materials examined only a few had moisture permeabilities below 50. These materials are listed in Table I. The compounds listed in Table I are made up of six principal atiuctures illustrated in Figure 3,
TABLE
I.
HAYIYG I,OK PERVEABILITY
ORG.4NIC b 1 A T E R I A L S
Polymer baranb Type 11 Type A4 recast B-115 B-118 B-130 F-120 1-inylidene chloride-vinyl chloride 92/8 80/20 66/34 60/40
50150
Vinylidene chloride-acrylonitrile 9218 80/20 T'inylidene chloride-acrylonitrile-vinyl chloride (75-80/10/10-15) Vinylidene ohloride-isobutylene (70/30) Polyethylene Tetrafluoroethylene Chlorotrifluroethylene Polyisobutylene (medium Vistanex) Butvl rubber (GR-I) Rugher hydrochloride
Thiqkness in Inches 0.00232 0.00188 0.00196 0.00184 0.00285 0 00169
Initial PermeabilityQ
15 28 30
6 8
Tetrafluoroethylene
Chlorotrifluoroethylene
/\ H M
/-\
H
H
H
Psobutylene
c'1
C1 M
H
CI
C1 H
H
10 5
8 14 40 6
Rubber hydrochloride
Figure 3. ...
0.00268 0.00203 0.00170 0,00179 0.00213
12 29 34 41
0,0012 0.00181
26 10
...
0.00246
18
...
22
0.0018
34
,..
26
0.00206 0.0050 0.0040
35
4
36
...
36
, , .
10
0.0018 0.00265 0.0036
50 30
...
38 40 44
14 31
36 42 24
... 25
13
Schematic Structural Formulas of Polymers Having Low Moisture Permeability
12
25
30
44
16
9
To this list must be added one other structure, poly-2,3-dichlorobutadiene, described in a patent to Barney (6) as copol~7iners containing a small proportion of other dienes. The major component has the structure
10
Initial permeability value in grams/100 rn.z/hr. b Saran is a generic name for a variety of copolymer consisting principally of 7-inylidene chloride units. a
H F T
Vinylidene chloride
15
25
H
PermeCalcuabilitv I ated of " Initial CrumPermepled ability a t Film 0.0020 I n . 10
10
H
H
hIOISTI-RE
9 5 8 28 7.5
H
VOl. 45, T\'o.,io
and the permeability value under the conditions of the test used here was a minimum of 18. Other patents (26) describe copolymers composed primarily of this monomer as being highly crybtailine
October 1953
INDUSTRIAL AND ENGINEERING CHEMISTRY
All t h e e polymers have several common characteristics. The basic structure is a hydrocarbon skeleton. There are no hydrophilic substituents or side chains. The substituents are relatively small or of short length. Polyvinyl chloride has all these characteristics and yet has a permeability value of 185. The outstanding structural characteristic which the majority of the low moisture permeable materials have which is not possessed by polymers like polyvinyl chloride is lateral symmetry on each carbon atom of the chain. In rubber hydrochloride the lateral symmetry is nearly the same as in the other polymers, since a chlorine atom is balanced by a methyl group and these groups are nearly equal in &e. In a polymer such as polyvinyl chloride chlorine atoms are opposed by hydrogen atoms which differ greatly in size. Polychlorotrifluoroethylene is somewhat unsymmetrical, but is much more symmetrical than polyvinyl chloride. Polychlorotrifluoroethylene (18)might be considered as polyvinyl chloride in which all the hydrogen atoms are replaced by the larger fluorine atoms. I n this way the outward contour of the molecule is made more uniform. The apparent outstanding exception in the group is poly-2,3dichlorobutadiene. This polymer, which is presumably largely trans in configuration, has a fixed lateral symmetry about the double bond, whereas no such rigidity or uniformity is present in polyvinyl chloride. In addition to the lateral symmetry mentioned,, the following several variations in longitudinal symmetry are present in these structures. 1. Polyethylene and polytetrafluoroethylene have like substituents on every carbon atom. This type of structure is probably most favorable t o packing of polymer chains into a crystal lattice. 2. I n polychlorotrifluoroethylene there is slightly unsymmetrical substitution on alternate carbon atoms, but the variation in the size of substituents along the chain is small. 3. Polyisobutylene and polyvinylidene chloride have like laterally symmetrical substituents on alternate carbon atoms, although X-ray diffraction studies have shown that the carbon chain is not arranged in the same way. 4. Rubber hydrochloride has a somewhat unsymmetrical substitution on every fourth carbon atom. 5 . Poly-2,3-dichlorobutadiene has a trans unsaturated unit alternating with a n ethylene unit.
sa
.h
X-ray diffraction studies of polyvinylidene chloride have shown that the polymer chain is in one plane and has a “tub” arrangement as in Figure 3 (82,32). Polyethylene has a zigzag configuration, and polyisobutylene also has a zigzag structure, but the substituent groups cause the chain to turn in a spiral (20, 28). The polymer chains of the other materials of low moisture permeability probably have zigzag structures though there may be a tendency t o spiral in some (2f). The regularity and compactness of these structures permit their packing closely into many crystalline areas which are impermeable. The noncrystalline areas probably are composed of fairly close networks of polymer chains because of the relatively small lateral dimensions of the molecules. The exact effect of variations in the type of lateral and longitudinal symmetry on the packing of polymer chains in the solid polymer can only be shown by a detailed study of the X-ray diagrams of a considerable number of individual polymers. Qualitatively, however, it can be stated that the lateral and longitudinal symmetry of the polymer chain is a very important factor in obtaining a material of low moisture permeability. Longitudinal symmetry of the types found in polyethylene, polyisobutylene, and polyvinylidene chloride are probably t o be preferred. Effects of Copolymerization in Symmetrical Polymers. When two laterally symmetrical monomers are copolymerized, the polymer gives a film of low moisture permeability. Such combinations can be made from isobutylene and vinylidene chloride, for example. The introduction of nonhydrophilic, unsymmetrical units into a laterally symmetrical polymer raises the moisture permeability only slightly a t first and then more rapidly as the proportion of
2299
unsymmetrical constituent is increased in the copolymer. For example, a vinylidene chloride-vinyl chloride (92/8) polymer ha5 a permeability value of 13 and a 60/40polymer has a permeability value of 30. As the polyvinyl chloride content is further increased, the values rise quite rapidly to the permeability value of polyvinyl chloride, which is 185 (Figure 4). This curve should probably have a second inflection as it approaches 100% polyvinyl chloride, giving it an S-shape. Sufficient data upon this mixture of components was not obtained to give such a curve, but experience with other vinyl chloride copolymers showed that the effect of the addition of small amounts of another monomer is masked by the major component.
100
%
5 I60 a
3
5120 -1
5
$00
40
10
20
10
40
PERCENT
50
60
TO
80
e0
100
VINYL CHLORIDE
Figure 4. Moisture Permeability versus Thickness of Vinylidene Chloride-Vinyl Chloride Copolymer Film
If the unsymmetrical comonomer is also hydrophilic, the pernieability curve frequently rises more rapidly. The introduction of 6 parts of polyvinyl acetate into polyethylene raised the permeability from 36 to 260. The effect of a n odd substituent, whether hydrophilic or otherwise, is dependent t o a large extent upon its size in relation to the major substituent present. Thus, polystyrene very markedly raises the permeability of polyvinylidene chloride while polyacrylonitrile can be present in considerable proportion before the permeability changes greatly. Vinylidene chloride-acrylonitrile copolymers often have high permeabilities as cast. Heating the free films at 100’ C. or above for a short time drives out traces of solvent and promotes crystallization in some cases. This effect is shown by the data in Table 11. Since this effect was not discovered until near t h e close of this study, the data on unbaked films of this series and those of the vinyl chloride-acrylonitrile copolymer series d o not represent minimum values. Effect of Copolymerization in Unsymmetrical Polymers. I n general the effects of copolymerization on the permeability of unsymmetrical polymers, such as polyvinyl chloride, are very much like those found for the symmetrical polymers. The principal differences are that the starting level of permeability iq higher and there is always the pessibility of going downward b y the use of a considerable amount of symmetrical monomer. Vinyl chloride-acrylonitrile polymers have the range of values expected, between 200 and 400, but the values d o not form a smooth curve when plotted against composition. It is quite probable t h a t this series of polymers behaves like the vinylidene chloride-acrylonitrile series in retaining solvent and that to obtain minimum permeability values the films must be baked at an elevated temperature. Vinyl chloride-diethyl fumarate (95/5) copolymer has a pernieability of about 375 at 0.0010 inch. Vinyl chloridemethyl chloroacrylate copolymer (90/10) has a value of 248 at a similar thickness. The latter contains a somewhat higher percentage of ester oxygen, but the comonomer is more symmetrical in structwre and of smaller dimensions.
INDUSTRIAL AND ENGINEERING CHEMISTRY
2 m
H
I
C1
-c-c- 1
c=o
C=O
I
I
CH, Dipthy1 fumarate unit Vinyl acetate unit
OC&
c=o
Methyl Ethyl Butyl
It seems quite likely that an unbranched t ~ a n s - l , ~ p o l y h u t a ~ i ~ ~ r l r \vould exhibit relatively low water vapor permeability. Cis and trans configurations in polymer chains, which are posfible because of the presence of unsaturation, have a marked effect on permeability. Xatural rubber ha,r a cis configuration, while the majority oi synthetic diene polymers are presumabl. trans (11, 28)
I
OCH3 Methyl chloroacrylat~ unit
Vinyl chloride-vinyl acetate (90/10) copolymer has a prrmcability value above these two. A series of films composed of vinyl chloride-chloracrylatr estei (95/5) copolymers was tested in which the ester group varied from methyl to normal amyl. Very little change in permeability values was found. This s h o w that the size of a minor constituent has little influence on moisture permeability when the permeability value is of the order of 150 or more. Minor effecb are more likely to be noticeable as the permeability is reduced toward a minimum. Effect of the Size of Substituents. An increase in the size of a large proportion of the substituents increases moisture permeability. This is illustrated by a group of three methacrylate polymers below.
R
Vol. 45, No, 10
Permeability Value 1150 1580 1790
Similar examples could be drawn among nonhydrophilic substituents. Effect of Reduction of Branching in the Polymer Chain. I n the preparation of synthetic addition polymers from monomerir unsaturated compounds, straight chainsare not always formed (39, 4g), Frequently a monomeric unit adds to the chain in an abnormal manner and provides a position for the growth of a side chain or branch. The presence of a considerable number of side chains would probably increase the moisture permeability by preventing the close packing and crystallization of polymer molecules. This effect would be similar to that obtained with large substituents on the polymer chain. Certain materials, known as chain transfer agents, when added in very small quantity to a polymerization mixture, rcact with aome of the active positions in the forming polymeric chain and stop polymerization in those particular directions. It ip believed that the over-all effect of the introduction of such materials into a polymerization reaction is to reduce chain branching. The acrylonitrile section of Table I1 contains a series of acrylonitrile polymers in which chain branching has presumably been decreased by the addition of small amounts of modifying agents. The permeability of these films was uniformly around a25 compared t o 380 for unmodified polymer. Effect of Unsaturation-Natural and Synthetic Rubbers. Unsaturation in the skeletal chain gives high permeability values. The permeability of polybutadiene is much higher than that of polyethylene. Effects due to chain branching (3, 39) may have a bearing on this comparison, and so for a rigid proof of the degree to which unsaturation affects permeability, a butadiene polymer would have to be hydrogenated under nondegradative conditions.
cis
?‘~f//lS
Gutta peicha, vhich is the ttans isoniei 01 natural rubber, had lox er moisture permeability. The trans configuration permits a closer parking of the chains. This result fits in with the relative hydrogen permeabilities of rubbei and gutta percha determined by Reitlinger (35, 44). Neoprene GK (polychloroprene) m hich ha= a t i uns configuration similar to gutta percha (21, g5, 4.9), hac; a Ion- moisture permeability value in relation to other synthetic polyene-. Polg-2,3-dichlorobutadiene( 6 ) , already discussed, is trans in structure and more symmetrical than Neoprene. As a result it is not a rubber a t loom temperature, but crystalline, and pomessew Jow permeability to water. The permeabilities of both natural rubber and polychloroprene were greatly reduced by chlorination. The effect was somewhat greater in the case of chloroprene. In the latter case, the outline of the molecule would be more eymmetrical because some position- 7%ould bear two chlorine atoms instead of the combination of methyl and chlorine groups in natural rubber. Such chlorination reactions are not simple addition reactions, but include substitutive and evclization reactionc; (9). Simple produrts might he reprevnted hv the following formulw:
Chlorinated natural rubhei
Chlorinated neoprene The reaction of natural rubber with hydrogen chloride according to Markownikoff’s rule ( 2 1 ) removes unsaturation and produces a more symmetrical structure. The resulting structure is that of a one-to-one polymer of polymethylvinyl chloride and polymethylene, and it has very low moisture permeability. An attempt was made to prepare gutta percha hydrochloride, which Rhould have a similar structure. The permeability of the product was not as low as that of rubber hydrochloride. Effect of After-Chlorination of Saturated Polymers. The after-chlorination of vinyl chloride polymers produced no marked changes in permeability values, though the values were usually
INDUSTRIAL AND ENGINEERING CHEMISTRY
October 1953
2301
TABLE 11. MOISTURE PERVEABILITIES or POLYMERIC MATERIALS IN FILM FORM
Solvent
Thickness
Initial Permeability=
Dioxane Cyclohexanone Cyclohexanone Dioxane Methyl ethyl ketone Methyl ethyl ketone
0.00151 0 00232 0.00188 0.00196 0.00184 0,00285 0.00146 0.00169
8 9 5 8 15 28 79 7.5
Methyl Methyl iMethyl Methyl .Methyl
ketone ketone ketone ketone ketone
0.00134 O.OW34 0.00121 0.00131 0.00127
Dioxane Dioxane Dioxane Dioxane Meth 1ethyl ketone-toluene (111) C ycloXexanone Methyl ethyl ketone-toluene 1/1) Methyl ethyl ketone-toluene [1/1)
Polymer Vinylidene Chloride Copolymers Saran Type M
Calculated PermeInitial ability Permeof ability Crumpled a t 0.0020 Filmb In.
10 6 8 15 28 73 30
ti 10 5 8 14 40 38 ti
10 99 88 10 11
17 104 77 36 11
7 66 53 7
0.00268 0.00203 0.00170 0.00179 0.00213 0.00144 0.00200 0.00170
10 12 29 34 41 150 141 149
..
14 31 36 42 154 132 153
13 12 25 30 44 108 141
Dioxane Cyclohexanone Cyclohexanone Cyclohexanone Cyclohexanone Acetone Dimethylformamide Dimethylformamide Dimethylformamide
0.0012 0.00189 0.00187 0 00208 0.00181 0.0012 0.00162 0.0016 0.0016
388
..
Methyl ethyl ketone Methyl ethyl ketone Dimethylformamide
0.00285 0.00246 0.0018
18 34
Cyclohexanone Cyclohexanone
0.00197 0.00289
144 132
132
142 191
Dimethylformainide Cyclohexanone Methyl ethyl ketone-toluene (1/1)
0.00168 0.0023 0.0016
219 319 172
219 338 185
184 367 138
Cyclohexanone
0.0021
125
122
131
Dimethvlformaniide Diniet hGlforinamide Dimet hylfortnarnide Dimethylforniamide
0.00157 0.00203 0.00157 0.00206
235
282 144
188
250 219 282 144
185 191 221 148
Dimethylformamide Dimethylformamide Dimethylformamide Dimethylformamide Dimethylformamide Dimethylformamide Dimethylformamide
0.00126 0.00194 0.0016 0.00149 0.00144 0.00147 0.00154
363 363 406 438 438 500 507
363 375 438 438 438 500 500
229 352 325 326 316 367 390
Vinyl chloride-vinyl aeetatr Vinyljte VYVW (95/5) Vinylite V Y N S (00/10)
Methyl ethyl ketone-toluene 1/1) Methyl ethyl ketone-toluene {1/1)
0.00125 0,00192
222 219
22 l 219
139 210
Vinyl chloride-diethyl fumarate (95/5)
Methyl ethyl ketone-toluene (1/1)
0.00070 0.0010 0.00151 0.00322 0.0012
438 375 244 144 375
438 350 257 136 375
153 188 184 232 225
Methyl ethyl ketont -toluene (t,,'l) llethyl ethyl ketone-toluene 1/1) Methyl ethyl ketone-toluene [1/1) Methyl ethyl ketone-toluene (1/1)
3.00142 3.001 13 0.00117 0.00122
219 219 250 188
219 219 219 2 19
156 124 146 115
Vinylidene chloride-acrylonitrile 92/8 90/10 SO/ZO (3 days a t 25' C.) S0/20 (1 hr. a t 70" C.) 80/20 (1 hr. at 110° C.) 70/30 60/40 29/71 12/88 Vinylidene chloride-vin 1 chlorideacrylonitrile (75-80,&0-15/10) 6 days a t 25O C. 1 hr. a t looo C. Vinylidene chloride-isobutylene (70/3 0)
Vinylidene chloride-styrene 601'40 50/50 Vinyl Chloride Copolymers Vinyl chloride Vinyl chloride, chlorinated Vinyl chloride-vinylidene chloride (chlorinated! (95/5) Vinyl chlonde-vinylidene chloride (90/10) Koon 102 202 203 204
No emulsifying agent in polymer Vinyl chloride-methyl a-chloroacrylate 95/5 90/10
so/zo
Vinyl chloride-ethyl rr.chloroacrylate (95/5) Vinyl ohloride-propyl' a-chloroacrylate (95/5) Vinyl chloride-butyl or-ahloroacrylate (95/5) Vinyl chloride-amyl a-chloroacrylate (95/5) Acrylonitrile Polymers Acrylonitrile Aorylonitrile-lauryl mercaptan (99.8/0.4)
ethyl ethyl ethyl ethyl ethyl
68 10 263 219 363 344 84
0
24 24
..
.. ..
.. ..
..
..
Reinarks Saran is a generic name for a series of vinylidene chloride copol~-mere.
7
127 16 22 82 71 9
168 178 290 275 99 22
26
Methyl ethyl ketone-toluene (1/1)
0,00104
2 19
250
114
Methyl ethyl ketone-toluene (1/1)
0.00132
219
219
145
Methyl ethyl ketone-toluene (1/1)
0.00132
203
219
134
Dimethylformamide Dimethylformamide
0.00203 0.00155
376 283
375 375
381 204
Cast upon plain uncoated celloohane of 0.00088-inch thiaknem .. -___ hontaining 16% glycerol as' plasticizer. Permeability m e a s ured with cellophane toward water in cup.
Koron is the trademark of B. F Gpodrich Cp., Akron, Ohio, to; vinyl chloride-vinylidene chloride copolymers.
Vinylite is the trademark of Carbide and Carbon Corp. for verious vinyl chloride copolymere.
(Continued on next page)
INDUSTRIAL A N D ENGINEERING CHEMISTRY
2302
TABLE 11.
RfOISTURE PERhIEABILITIES OF POLYJIERIC ;\lATERIALS IN
Polymer Acrylonitrile-dithioglycidol (99/1) Acrylonitrile-ethylene oxide (99/1) Acrylonitrile-trichloroethylene (98.5/1.5) Acrylonitrile-methacrylonitrile (99/1) Acrylonitrile-vinyl acetate (98.4/1.6) Acrylonitrile-methyl vinyl ether 96/4 93/7 Acrylonitrile-maleic anhydride (95/5) Acrylonitrile-diethyl fumarat6 (91/9) Acrylonitrile-vinyl acetate (50/50) Acrylonitrile-Styrene 85/15
Initial Permeability= 313 338 319
FILMFORM (Continued)
Permeability of crumpled Film6
Calculated Initial Permeability a t 0.0020 In. 219 220 204
Solvent Dimethylformamide Dimethylformamide Dimethylformamide
Thickness 0.0014 0.0013 0.00128
Dimethylformamide Dimethylformamide
0 0016 0.0014
288 332
Dimethylformamide Dimethylformamide Dimethylformamide Dimethylformamide Acetone
0.0016 0.0017 0.0016 0.0015 0.0023
263 275 363 344 1188
1725
210 234 290 258 1370
Dii~iethylforrnamide Dimethylformamide
0.0015 0.0016
275 406
i38
206 325
Fluorine-Containing Polymers Tetrafluoroethylene Chlorotrifluoroethylene
Pressed Pressed
0.0050 0.0040
4 5
.. ..
Polyethylenes Polyethylene Peroxide catalyst Lithium butyl catalyst Carbon monoxide (98/2)
Xylene Toluene Toluene
0.00206 0.0010 0.0010
35 93 153
.. ..
36
36 47 77
60/40
Polyethylene-vinyl acetate (94,’s) Chlorinated polyethylene 33.3% Chlorine 46.6% Chlorine 64.0% Chlorine
Vol. 45, No. 10
500 625 344
..
.. .. ..
Remarks
230 232
10 10
Toluene
0.00180
288
275
259
Toluene T o1u en e Toluene
0.00210 0.00172 0.00178
250 147 188
282 157
262 126 167
Toluene
0 0013
400
..
260
Toluene
0 00209 0 00137
360 500
688
376 342
0.00192
363
500
349
Dimethylformamide Diinet hylformamide
0.00200 0.00234
500 438
.. ..
500 513
Toluene
0.0010
570
570
285
Toluene
0 00131
875
..
572
00178 00162 003s 0005 00144
1125 119 25 675 144
..
1OM)
.. ..
44 169 104
Chloroform Chloroform Chloroform
0.00183 0.00209 0.00186
469
500
..
457
781
..
490
726
Chlorinated G N (61.0%) Chlorinated GN (60.4%) Buna S
Toluene Toluene Dioxane
0.00260 0.00161 0.00134
83 89 1030
*...
..
108 72 690
Hycar OR
Ethylene diohloride
0.00223
400
..
446
Copolymer of butadiene-acrylonitrile (62/38) produced by t h e Hycar Chemical Co., Akron, 0.
Butyl rubber GR-I
Toluene
0.00285
30
..
40
Butyl Rubber GR-I is a polymer of isobutylene with a small proportion of a diene such as butadiene or isoprene produced by the Standard Oil Co. of N. J., Piew York, N. Y.
Chlorinated butyl rubber (12% C1)
Toluene
0.00128
225
..
144
Cast upon plain, uncoated cellophane Of 0.00088-in. thickness, containing 16% glycerol as plaaticizer. Permeability measured with cellophane toward water in cup.
Chlorosulfonated polyethylene (29% chlorine; 1.4% sulfur)
Polystyrenes Polystyrene
Styrene-isobutylene
so/zo 50/50
Styrene-Butadiene (90/10) Elastomers and Related Polymers Katural crepe rubber
Polyisoprene. synthetic Chlorinated natural rubber Rubber hydrochloride G u t t a percha Gutta percha hydrochloride Keoprene GN GG
FR
Toluene RIethyl ethyl ketone-toluene (1/1) Toluene Toluene
..........
0 0 0 0 0
..
25
..
Cast upon plain, uncoated cellophane of 0.00088-in. thicknesa, containing 16% glvcerol as plasticizer. Perrneabtlity measured with cellophane toward water in cup. Polyflex film produced by the Plax Carp. Brittle films
Cast upon plain uncoated ceilophane pf 0.00688-in. thickness, containing 16% glycerol as plasticizer. Permeability measured with cellophane toward water in cup.
96
Cast upon plain uncoated cellophane of 0.00688-in. thickness. containing 16% glycerol as plasticizer. Permeability measured with cellophane toward water in cup. i’ieoprene is a generic namt for polymers and copolymers or chloroprene. Cast upon plain, uncoated cellaphane of 0.0008-in. thickness, containing 16% glycerol as plasticizer. Permeability measured with cellophane toward water in cup. Buna S is a butadiene-styrene couolymer produced by Standard Oil Co. of S.J., S e w York, X. Y.
(Continued on next p a g e )
I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
October 1953
TABLE 11.
AhfOISTURE PERMEABILITIES OF POLYMERIC MATERIALS IN
Thickness 0.00111
Initial Permeability5 2500
FILMFORM(Concluded)
Calculated PermeInitial ability Permeof ability Crumpled a t 0,0020 Filmb In. Remarks 1390 Cast upon plain, uncoated cellophane of 0.00088-in. thickness containing 16% lycerol as plas: ticiser. Permeatility measured with oellophane toward water in cup. Thiokol is the trademark of the Thiokol Corp., Trenton, N. J., for a series of organic polysulfides.
Thiokol FA
Solvent Ethylene dichloride
Vistanex, medium
Toluene
0.00123
50
Polybutadiene
Toluene
0.00206
1380
0.00167 0.00340 0.00262
438 152 172
500
.. ..
365 258 225
Aqueous alcohol
0,00174 0.00209
1060 938
1060 938
920 980
Norelac
Toluene
0.00179
813
Paracon
Toluene
0.00125
1645
..
Ester rubber
Ethylene diohloride
0.00250
2200
Water
0,00049 0.00128 0.00266 0.00347
Polyvin 1 butyral 45% g e e hydroxyla 20% free hydroxyls
2-Propanol-water (80/20) Toluene-2-propanol (70/30)
11% free hydroxyls
Toluene-2-propanol (70/30)
Polymer
2303
..
31
..
1420
Cast upon plain, uncoated cellophane of 0.00088-ln. thickness containing 16% lycerol as plas: ticiser. Permea%ility measured with cellophane toward water in CUD. Vistanex is trademark for polyisobut lene produced- bi. Standard &I Co. of N. J., New York, N. Y.
Polyamides and Polyesters Stretched 350% Stretched 450%
66/610/6 (40/30/30)
Melt cast
..........
..........
730
Cast upon plain uncoated cellophane of 0.00088-in. thickness containing 16% glycerol as plas: ticizer. Permeability measured with cellophane toward water in CUD. Norelac is a oolvamidr resin produced from-etLylene: diamine and dimerized or trimerized linoleic and linolenic acids by the U. S.Dept. of Agr., Northern Regional Labs., New York, N. Y.
1030
Cast upon plain uncoated celloDhane of 0.0688-in. thickness. hontaining 16% lycerol as plasl ticiser. Permeatility measured with cellophane toward water in cup. Paracon is a polyester elastomer which was produced by Bell Telephone Labs., Kew York, N. Y.
2150
2750
Cast upon plain uncoated cellophane of 0.0008-in. thickness containing 16% lycerol as plas: ticizer. Permeaklity measured with cellophane toward water in cup. Ester rubber is a polyesteramide produced by Imperial Chemical Industries, Ltd., London.
3800 3100 2500 2500
3800 3100 2500 2500
980 1980 3320 4360
0.0020 0.0010 0 00159 0.00043 0.00173 0.00272 0.00341
1000 1200 750 2500 657 438 369
1000 1200 750 3 100 688 438 375
1000 600 596 1070 568 596 629
0.0020 0.0020 0 . 0020 0.0010
1150 1580 i7~n 1530
1150 1580 1790 1530
1150 1580 1790 765
0.00144
3 100
3100
2230
. Polyvinyl Alcohol Derivatives Polyvinyl alcohol *)
Methacrylate Ester Polymers Methyl methacrylate Ethyl methacrylate Butyl methacrylate Methyl methacrylate-butadiene (84/16) Vinyl Ketone Polymers Methyl vinyl ketone
.......... .......... .......... .......... Methyl ethyl ketone
First sets of numbers indicate number of carbon atoms in amine and acid, respectively, of salts from which pqlyamides are made. Numbers in parentheses. y e weight ratios of salts. Single 6” in code is omega aminocaproic acid.
Cast upon plain uncoated cellophane of 0.00088-in. thickness, containin 16% glycerol as plasticizer. Sermeability measured with cellophane toward water in cup.
5 Initial permeability value is in grams of water transmitted through the film/lOO rn.Vhr. a t B vapor pressure difference equal to 53 mm. of mercury and a temperature of 39.5O C. 6 Values on manually crumpled film.
2304
INDUSTRIAL AND ENGINEERING CHEMISTRY
higher. Such chlorination is supposed not to give rise to vinylidene chloride units until most of the methylene groups bear one chlorine atom. The chlorination of polyethylene produces polymers which are ielated to polyvinyl chloride. The permeability values of films of several of these polymers were obtained. Chlorinated polyethylene (33% chlorine) has one chlorine atom for every five carbon atoms in the chain. I t s permeability value was 262. Chlorinated polyethylene containing 46.6% and 64% chlorine had permeability values of 126 and 167, respectively. The chlorine content of polyvinyl chloride is 56.8% and that of polyvinylidene chloride or a structure with one chlorine atom on every carbon atom in the chain is 73.2%. The chlorination of polyethylene to a point approaching complete replacement of the hydrogen atoms should produce a polymer which would give a film having a very low permeability value and a crystalline X-ray diagram Effect of Hydrophilic Substituents. -4s has already been pointed out, the presence of a small number of hydrophilic groups has only a minor effect on the permeability value as in the case of vinylidene chloride-acrylonitrile (92/8) copolymer with a permeability value of 16. However, a polymer containing a substantially high number of hydrophilic groupq, such as hydroxyl, ester, acetal, carbonyl, ether, and nitrile, always has a high moisture permeability. To illustrate this, polyvinyl alcohol has an initial permeability value of about 2500; polymethyl methacrylate, 1150; polymethyl vinyl ketone, 2230; and polyacrylonitrile, 380. The reduction of moisture permeability by the conversion of hydroxyl groups t o a less hydrophilic derivative is illustrated by the series of polyvinyl butyrals in Table I1 Effect of Nonhydrocarbon Chain in the Polymer. Polymeric chains frequently contain such groups as ester, ether, acetal, and amide. I n all cases examined in this study where these groups were part of the polymeric chain, films made from the polymer had high moisture permeability. Polyamidw, polyesters, polysulfides, and cellulose derivatives ( S I , 34) all have quite high permeabilities. Polyethylene terephthalate (Mylar, Du Porit Co.’s polyester film) is reported by Amborski and Flier1 ( 4 )to have a permeability value of 160 in 1 mil thickness under the conditions of the present test. This would extrapolate to 80 for a 2 4 1 film. This low value may be partly attributed to crystallinity, for noncrystalline films of this material have higher permeabilitic 3. However, the factors of stiffness and consequent higher second order transition temperature (T,)provided by the aromatic ring certainly contribute to setting this polymer apart from aliphatic polyesters. Effect of End Groups in Polymers. It hay been seen that small amounts of hydrophilic groupings do not exert a profound effect on permeability values. However, in attempting to achieve a minimum permeability, as many small deleterious effects as possible should be eliminated to diminish their additive effect. One such factor is the possible effect of hydrophilic end groups on the polymeric chains. These groups romr from portions of catalyst molecules and materials which add into the polymer and stop polymerization. Ordinary polyethylene contains some oxygen-containing groups. A sample of polyethylene prepared with lithium butyl catalyst was obtained. This suppowdly contained hydrocarbon end groups. No decrease in moisture permeability was observed. Effect of Crystallinity-X-Ray Study. Thtx structural pioperties which have been found in the preceding discussion to accompany low moisture permeability are those which favor crystallinity. X-ray diffraction patterns were prepared of most of the materials having low moisture permeability (Table 111). A11 these polymers showed some degree of crystallinity, but this was not an exclusive characteristic of materials of low permeability. Polyamides, polyacrylonitrile, and regenerated cellulose films are highly crystalline and yet have high moisture permeability values. These materials should be quite impermeable to nonhydrophilic and nonsolvent vapors or gases.
Vol. 45, No. 10
TABLE 111. CRYSTALLINITY OF POLYMERIC F I L YAC. ~ DETERMIKED BY X-RAYDIFFRACFION PATTERN*
Polymer Saran (Type M)
Calculated Permeability Value a t 0.0020 In. 10
3 8
ene Butyl rubber (GR-I) Polyethylene Chlorinated . Dolvethvl. ene 46.6% Chlorine 64% Chlorine
Degree Considerable High
Low
Crystallinity Size of Crystals Small
Orientation Slight
Very Large small
Sone Sone
13 12 25 30 44 184
High High Moderate Moderate Low Low
Large Large Large Large Smal I Small
Sone Sone Nono None None None
142
Moderate
Large
None
16 44 10
Moderate Very High high
Largo Large Large
Very high
10
High LOW High
Small Small Large
None None
... ...
...
40
3G 126 167
350
Aniorphous Amorphous Very low
None None
None
...
. . I
204
Xoderate
Large
None
188
Amorp hous
...
, . .
Within the sensitivity of the permeability test no di5tinction could be made between the permeabilities of films of hydrophobic polymers containing large and small crystals Which condition would promote the lower pernicability probably would dcpcnd upon the degree of organization of the crystals in relation to one another. d l materials showing a completely aniorphouq structuic probably will always be found to have rather high moisturp pcrmeability. Polystyrene, pol) vinj 1 chloride, and chlorinated polyethylene are examples. Effect of Other Physical Properties- Orientation and Durability. The physical state of a film and its method of prepai ation are factors in obtaining minimum moieture permeabilities fiom a given chemical structure. Theoretieally, polyethylene should be one of the materials least permeable to moisture. To a large extent mechanical defects probably prevent the realization of the lowest permeabilities in this case as well as in others, Orientation of polymer crystal. by drawing should promote the organization of both crystalline and amorphous areas and thereby reduce the number and size of openings through which pcrmeant molecules may pass. An appreciable reduction in rnoistuie permeability was obtained by stretching a polyamide film 300 to 450y0 (Table 11). On the other hand, two commercially available films, which were prc-umahly oriented, polystyrene and a vinylidene chloride copal> mer, had no lower permeabili fie? than plate cast films of the same materialq. Dalfsen ( 1 4 ) found that highly stretched unvu1can;zc.d lubber had a lower permeability than unstretched rubber. This process crystallizes the rubber as well as orients it. Tenacity, elongation, stiffness, etc., appear to have little bearing on moisture permeability of unhandled films provided they are not so weak and brittle that they develop cracks upon a minimum of handling. Only the tough and elastic films, however, will withstand crumpling without a rise in permeability, as shown by the values in Table 11. Permeability of Physical Mixtures. Although no such data are presented here, the addition of relatively small plasticizing molecules to a polymer mass may either raise or lower the permeability (2, 5, 15, 37). Since a majority of plasticizing substances contain hydrophilic oxygen groups, the more common effect is an increase in permeability. Plasticizing agents increase the mobility of chain segments and may provide fluid pools through which
I N D U S T R I A L A N D E N G I N E E R I N G 'C H E M I S T R Y
October 1953
moisture may permeate. If the plasticizer is more hydrophobic than the polymer, it may block the passage of moisture through openings between polymer molecules. When mixtures of two compatible polymeric substances are cast into films, the moisture permeability of the resulting structure is intermediate between the permeabilities of the two polymers alone, b u t a t the extremes of the composition range it is not in proportion to the weight ratio of the two substances (Table I V and Figure 6 ) . An S-shaped curve is obtained by a masking effect of the predominant material. 200
2305
the more dense and has fewer faults and pores, which may be left in the air surface by evaporating solvent. Melt-extruded films are probably more uniform in cross-section than solvent-cast films. One experiment was carried out to test the additive effect of thin films (Table V). The lamination of two thin films of vinyl chloride-diethyl fumarate (95/5) copolymer by means of a 0.00001-inch layer of a highly permeable modified polyamide produced a film having a permeability much lower than that of a single cast film of equal thickness. The measured permeability of the lamination is in good agreement with the value calculated from that of a single ply by inverse proportion. This shows that the physical structure of a film may play an important part in permeability relationships.
TABLE v. COMPARISON O F PERME.4BILITY O F LAMINATED AND SINGLEFILMS OF VINYLCHLORIDE-DIETHYL FUMARATE (95/5)
20
10
30
40
10
70
SO
110
90
100
POLY(VINYL CHLORIDE-DIETHYL FUMARATE) 9 5 h
PERCENT
Figure 5. Moisture Permeability versus Thickness of Physical Mixtures of Poly(Viny1 Chloride-Diethyl Fumarate) with Butyl Rubber GR-I Effect of Thickness on Permeability. That permeability through nonsorbing films is an inverse function of thickness seems to be generally accepted (8, 10, 11, 34, 40). The introduction of sorption phenomena and also variations in the testing method upset this relationship ( l a , 41). Aiken, Doty, and Mark ( 3 ) reported that thin films sometimes had higher diffusion constants than thicker films. Halls ( 2 3 ) found thick films to have the higher permeabilities per unit of thickness, and there have been suggestions of the existence of structural differences in different parts of the films, which may account in part for these anomolous relationships (27,37,QO).
TABLEIV. PERMEABILITY OF PHYSICAL MIXTURES OF POLYMERS w m VINYLCHLORIDE-DIETHYL FUMARATE (95/5) COPOLYMER
11
Admixed Per Polymer Cent Saran B-115 50 Hycar OR
*
50
Solvent Cyclohexanone Methyl ethyl ketonet o 1u ene
Thiokneas
Initial Permeability
Permeabilitv of
Crumpled Film
Calculated Initial Permeability a t
0.0020 In.
0.0021
93
99
98
0.0013
500
563
366
0 00156 0 00188 0.00203 0,00245
250 188 113 57
250 188 135 61
195 177 115 70
(50/50)
P
Butyl rubber (GR-I) 10
20 33 50
Initial Structure Thickness, In. Permeability Single ply 0.00095 313 163" Two ply 0.00205 l98b Single ply a Calculated value by inverse proportion 157. Varying permeability values for this pblymer in the several tables are the result of measurements on different batches of polymer.
Tetrahydrofuran
For many film-forming polymers the moisture permeabilitythickness curve approximates a hyperbola, as in Figures 1 and 2. Materials which produce this type of permeability curve often have a lower permeability per unit of thickness in thin films than in thicker films, provided the meastirements are made a t equilibrium. This could mean that one or both of the surface regions of the film have a structure different from the internal regions of the film. I n a macro sense the plate and air sides of solvent cast films are often clearly distinguishable. The plate side is probably
SUMMARY
The moisture permeabilities of over 100 polymers in film form were determined a t 39.5" C. and a vapor pressure differential equivalent to 53 mm. of mercury. Of these materials only 20 had permeabilities below 50 grams of water vapor per 100 m.* per hour a t a thickness of 0.002 inch. The film-forming materials having low permeabilities were polyethylene, rubber hydrochloride, vinylidene chloride copolymers, polytetrafluoroethylene, polychlorotrifluoroethylene, and isobutylene-isoprene copolymers. The basic structural units present in these highly moistureproof materials are ethylene, isobutylene, vinylidene chloride, methyl vinyl chloride and ethylene in a 1 to 1 ratio, tetrafluoroethylene, and chlorotrifluoroethylene. These units are characterized by a nonhydrophilic nature and a high degree of lateral symmetry. X-ray diffraction patterns showed that all of the materials having low moisture permeability were crystalline to some extent, but crystallinity did not assure low permeability because it may be outweighed by hydrophilic properties of the polymer as in cellulose or polyamides. Films from polymers made up of amide, ester, and acetal chains were very permeable to moisture. An exception to this is polyethylene terephthalate, which is a crystalline, aromatic polyester. Polymers having double bonds in the chain, as in many rubbers from dienes, had relatively high permeabilities. A marked exception t o this, appearing in the literature, is poly-2,3-dichlorobutadiene, which is highly crystalline. That the polymer chain should also have a minimum of branching was shown by the reduction in permeability produced in acrylonitrile polymers by the addition of agents to the polymerimtion mixture which reduce branching. Polymers having nonhydrophilic, but unsymmetrical, substitution on a hydrocarbon skeleton were moderately impermeable to moisture. Polyvinyl chloride and polystyrene are good examples Decreasing the symmetry of a polymer by introducing successively larger substituents caused an increase in moisture permeability. Polymethyl, ethyl, and butyl methacrylates are one such series. The addition of an increasing proportion of a small unsymmetrical unit, such as vinyl chloride to a symmetrical polymer, such as polyvinylidene chloride, increased the moisture permeability, although a considerable proportion of such a comonomer was tolerated without a large effect on the permeability value.
INDUSTRIALAND ENGINEERING CHEMISTRY
2306
Polyniers containing a large number of hydrophilic side groups, such as hydroxyl, ester, acetal, amide, or nitrile, were highly permeable to moisture. The effect of a small proportion of the smaller of these groups, such as nitrile groups, was buried within highly crystalline and impermeable sti uctures of the vinylidene chloride type. After-chlorinatio~iof vinyl chloride polymers did not greatly affect the moisture permeability. This was probably due to the fact that chlorination took place on the methylene groups and PO did not develop the symmetrical vinylidene chloride units. Chlorination of unsaturated rubbers, on the other hand, produced a very appreciable reduction in moisture permeability. On the basis of the present study, it is concluded that in older t o obtain films having very low moisture permeabilities, the film forming polymers should have the following structural characteristics:
A4saturated or nearly saturated carbon chain, unless the has symmetry permitting high crystallinity A minimum of chain branching A high degree of lateral symmetry 4. A very high proportion of relatively small, nonhydrophilic substituents 5 . .4 fair degree of longitudinal ordei or symmetry I chain 2. 3.
4CKNOWLEDGRIEYT
The author expresses his appreciation to A. Hershberger of the Filni Department for the suggeqtion of this study and continuing advice and counsel during the investigation, to R. A. Scheiderbauer of the Acetate Research Division and to members of the Jackson Laboratory, Experimental Station, Electrochemicals Department and Ammonia Department of the D u Pont Co. for generously supplying samples of polymers for this work, and to L. W. Brant for laboratory assistance. The X-ray diffraction patterns were prepared and interpreted by D. T. Meloon, Film Department. LITERATURE CITED
1) Aiken, W.H., P a p e r Trade J , , 125, No 14, 31-5 (1947). 2) Siken, W. H., Doty, P. M., and Mark, H , M o d e r n Packagzng.
18, NO. 12,137-40,166,168 (1945). ( 3 ) Alekseeva, E. N., and Belitzkaya, R. AI., Rubber Chem. Techno/.,
15,693-7 (1942).
(4) Amborski, L. E., and Flierl, D. It7,,ISD. ENG.CHEM.,45, 2290
(1953). ( 5 ) Badum, E., Kautschulc, 14,231-2 (1938).
$6)Barney, A. L. (to E. I. du Pont de Kemours & Co , Inc ), C Patent 2,496,976 (Feb. 7,1950). 17) Baxter, J. P., and -Moore, J. G.. J . Soc. Cheri I n d . , 57, 336 (1938). ,8) Berry, W., Plastics ( L o n d o n ) , 9, 33-5 (1945). (9) Bloomfield,G. F., J . C h e m . Soc., 1943, 289-96; 1944. 114. < l o )Brubaker, D. W., and Kammermeyer, I< , IKD. ENG.CHEY., 45, 1148 (1953).
Vol. 45, No. 10
!ll) Bunn, C. W., Rubber C h e w . Teohnol., 15, 709 (1942). (12) Carson, F. T., Natl. B u r . Standards ( U . S . ) , Misc. P u b . M127, (1937); P a p e r T r a d e J., 125, No. 19, 120 (1947). (13) Charch, W. H., and Scroggie, d. G., P a p e r Trade J . , 101, No. 14,31-9 (1935). 14) Dalfsen, J. W. van, Rubber Chem. Technol., 16, 388-99 (1943). 16) Doty, P. W., J . Chem. Phus., 14, 244 (1946). 16) Doty, P. M., Aiken, IT. H., and Mark, H., IND. ENQ.CHEM., ANAL.ED.,16,686-90 (1944). &,17)Edwards, J. D., and Pickering, 9. F., Natl. B u r . S t a n d a i d s (U.S.), Sei. P a p e r s 387,327-62 (1920); Chem. Met. Eng., 23, 71-5 (1920). ,1S) Frey, S. E., Gibson, J. D.. and Lafferty, R. H., Jr., IND. ENG. C ~ ~ ~ . , 4 2 , 2 3 1 4(1950). -17 9) Fuller, C. S., Chem. Reus., 26, 143 (1940). 0 ) Fuller, C. S., and Baker, IT-. O., .I. Chem. E d . , 20, 3-10 (1943). (21) Gehman, S.D., Field, J. E., and Dinsmore, R. P., Rubber Chem. Technol., 12,210-23 (1939). 22) Goggin, W. C., and Lwory. R. D., IND. ENG.CHEM.,34, 327-32 (1942). :23) Halls, E. E., Plastics ( L o n d o n ) , 5 , 257-60 (1942); 6, 9-13, 62-4 (1942). (24) Houwink, J. R., I n d . des Plnstiques. 3, 409-14 (1947); V w f kroniek, 20,172-6 (1947). 25) Kebansky, A,, and Chevychalova, J . Gen. Chem. (U.S.S.R.), 17, (79) 941-56 (1947). :26) Kuhn, L. B. (to Firestone Tire and Rubber Co.), U. S. Patent 2,514,195 (July 4,1950). 127) Lishmund, R. E., and Siddle, F. J., J . Oil & Colour Chemists’ ASsoC., 24,122-37 (1941). :28) hleyer, K. H., “Satural and Synthetic High Polymers,” Xem York, Interscience Publishers, Inc., 1942. (29) Morgan, P. W., M o d e r n Packaging, 22, No. 1, 159 (1948). :30) Muller, F. H., Kolloid Z., 100,355-61 (1942) (31) Oswin, C. R., J . Soc. Chem. h i d . , 62, 45-8 (1943). (32) Reinhardt, R. C., IND. EKG.CHEM., 35, 422-8 (1943). /33) Reitlinger, S. A., J . Gen. Chem. (U.S.S.R.), 14, 420-6 (1944); Rubber Chem. Technol., 19, 385-91 (1946). ;34) Rouse, P. E., Jr., J . Am. Chem. Soc., 69, 1068 (1947). c.35) Sager, T. P., J . Res. Nail. B u r . Standards, 13, 879-85 (1934). (36) Sarge, T . TV., Anal. Cheni., 22, 1541 (1950). ;37) Simril, V. L., and Hershberger, A., M o d e r n Plastics, 27, No. 11, 95 (1950). ,38) I b i d . , No. 10, p. 97. 39) Staudinger, H., and Fischer, K., Rubber Chem. Technol., 15,52334 (1942). :40) Taylor, R. L., Herrmann, D. B., and Kemp, A. R., IND.EKG. CHEM.,28,1255 (1936). (41) Thomas, A. hl., J . A p p l . Chem., 1, 141-57 (1951). :42) Thompson, H. M., and Torkington, P., T r a n s . Faradaay Soc., 41,246 (1945). ;43) Walker, H. W., and Nochel, W. E., Proe. Rubber Technol. Conj.,Wnd Conj., 1948,67-78. ;44) Van ilmerongen, G. J., J . P o l y i i e r Sei., 2, 381-6 (1947). (45) Van Amerongen, G. J.. Ibid., 5, 307-32 (1950); Rubber Chem. Technol., 24, 109-31 (1951); Rubber Chem. Technol., 23, 652 (1950). -4CCEPTED July 29, 1953. RECEIVED for review May 1, 1953. Presented at the Fifth Chemical Symposium of the Delaware Section of the. AMERICAX CHEMICAL SOCIETT at Sewark, Del., January 17. 1953.
