Document byKyMzpxERe6N97xzen9QaDL6

& & Jt * 1117 Buckman Laboratories, Inc. **1/ Swelling of PJtfft? f8ms in Water X. Rate of Penetration of Water, Permeability to Water Vapor, and Penetrability to Air in Relation to Water Absorption F. L. BROWNE Chemist, Forest Products Laboratory,1 Forest Service, U. S. Department of Agriculture Differing theories exist on resistance of house paints to blister ing. Tests were made to measure the permeability to water and water vapor and the ability to breathe of some paints for which water absorption and swelling had already been determined. Factors are discussed which apparently affect the speed of penetration Into vari ous coatings.' r e v io u s a r t ic l es in this series "breather-type" paints have been sold P(5)2 reported the extent to which commercially, but examination of some free and bound films of various paints of them has failed to confirm that and paint vehicles absorb water and they have greater permeability than do swell when soaked in water. The data ordinary paints. They did prove more suggest strongly that the absorption resistant to blistering and lower in ab and swelling of paints in water may sorption and swelling than most ordi be closely related to their tendency to nary paints, however, but paints may blister and peel under practical condi be highly resistant to blistering ana tions of service on houses. Those house yet low in permeability (34). It re paints that are known from actual ex mains to be determined whether ex perience to blister easily are also high tremely high permeability affords re in absorption and swelling, whereas liable protection against blistering. those paints similarly known to be The differing theories on resistance relatively resistant to blistering and to blistering made it desirable to meas peeling are also low in absorption and ure the permeability and ability to swelling. breathe of some of the paints for Other workers have observed a re which the absorption of water and lation between water absorption and swelling had already been determined. blistering of coatings (30, 35), both It was of interest to learn: 1) how on wood (8, 19) and on metal (1, 17, long it takes for water in contact with 20, 23, 27, 28). To compare water ab the exposed surface of high- and low- sorption with tendency to blister, there swelling coatings to penetrate to the is need for more quantitative data on interface with the substrate, 2) the blistering pressure (18) or other di relative permeability of free films to rect measure of blistering. water vapor, and 3) the penetrability American technologists often em of free films by dry air under a mod phasize another property of coatings, erate difference in pressure. their permeability to water vapor, as a The composition of the paints used measure of sensitiveness to moisture in these tests was reported in papers blistering. As far back as 1934, Gard VI and VII of the series (5). ner (12) suggested that readily per meable paints would resist blistering, Time Required for Water and proposed to make permeable paints to Penetrate by incorporating porous pigments such as diatomaceous silica. It was thought that such paints "breathe" sufficiently to let water escape from the painted surface of wood faster than it can penetrate through boards from be hind (4, 8, 15, 31). So-called Glass microscope slides l^A by 3 inches in size were marked with crossed stripes of an aqueous solution of cobalt chloride containing a very small proportion of methyl cellulose. The markings, % inch wide, were blue in color when dry, and turned 1 Maintained at Madison, Wis.. in cooperation with the University of Wisconsin. * Numbers in parentheses refer to Literature Cited. pink when exposed for a few minutes to air with a relative humidity of 65 per cent or more. They returned The Author: F. L. Browne received B. Checn. degree from Cornell U.. Pb. P. in colloid chem istry from U. of Wisconsin. He joined the For est Products Lab staff in 1918, helped develop casein glues for wood airplanes of World War 1. In 1921-22 he was a National Research FeUow at Wisconsin. Since 1929 Dr. Browne has edited the General and Physical Chemistry Section of Chemical Abstracts. promptly to the original blue color when brought back to air with a rela tive humidity of 30 per cent or less. The weight of the markings was deter mined by weighing the slides before and after the solution was applied. To determine the density of the mark ing material, dried coatings on glass were scraped off with a razor blade and measured by the pycnometric method in kerosene. The thickness of the markings, calculated from the weight, area, and density, was 0.11 to 0.17 mil. The marked slides were coated with paint at the desired thickness with doc tor blade, suction plate, and suitable shims. Five single-pigment, raw lin seed oil paints of pigment volume 0.30, pigmented with basic carbonate white lead, antimony oxide, titanium dioxide, zinc oxide, and magnesium silicate, respectively, were tested. The first two pigments make low-swelling paints, the fourth makes high-swelling pint, and the others make paints of intermediate swelling. For each pint, slides were prepared with coatings that were approximately 2 mils, 4 mils, and 6 mils thick when dry. One complete set of specimens was tested for water penetration when die pint coatings were 8 days old. A sec ond complete set was exposed for 15 days to ultraviolet light m a weatherometer, but without spraying with water. (It was found that water turned the cobalt chloride markings pink within one to two days, and the coat ings became blistered over the mark ings soon afterward.) To observe the rate of penetration of water, the test specimens were placed with the painted face down on glass rods that rested on the bottom of a shallow dish, and the dish was filled with water until about i/2 inch of water covered the specimens. The blue markings were then readily vis ible through the water and glass against the white background of the Kints. The lapse of time was observed tween submergence and the first appearance of change from blue to pink color, between submergence and com plete chance from blue to pink, and between submergence and blistering of the coatings. Blistering could be de tected without lifting the specimens out of the water because formation of a blister allowed the dissolved cobalt chloride to gather in a rounded droplet of solution that was easily recognized Reprinted from the April, 1957, Forest Products Journal (Vol. VII, No. 4), pages 1-43--If Table 1.--TIME REQUIRED FOR WATER TO PENETRATE COATINGS OF UNSEED Oil PXlNTS ON GLASS WHEN IMMERSED IN WATER Ccpoltlci of palat (pi^ent la linseed oil at plgeent Toluae 0.30) 1 Coatlags not weathered ---- ) Cctlu. vwtlwnA artificial nnlUM .... fBUekaeMJ of ; - TUe required for latter .bMrtea:HtlekMM: by free fiU of - ..,Htine required for ...... :Vater absorbed ... by fm film ; coating first Complete :Blisterlag: In 3 dayai : coating Tirtt t estate :ftUoterlag: la 3 days* :penetration:penetration: 2 : penetration s peaetrst ton: 1 mu 5Kr. lr. : : XU t Percent by * mi. < t votene lr. . Si t &. t i Percent by i Toluee Basic carbonate vfclte lead : 8.3 t U t 89 3* 3 8.9 3.8 t w 89 i 89 3 6.9 13 : 18 : 31 X 1-3 7 : X 81 : 9-9 3.0 7 35 i 89 X 6.6 7 35 35 * Actteeny oxide : * 2.1 : 3.6 ; 7. 7* ; 10 s . si t 13 53 7.J t 13 1 31 : 81 I 1.3 X 3.1 f t : 3t 12 3 2t 5 8.5 6.1 ) Si : 18 ftutlle tltanha dioxide : 1.9 : 3-5 t 6.9 2. Ts us 9. 6 13 s u 81 2 13 : aS.j t 2 1-3 1/8 8.9 2 6.1 3 3t 3: 82 5i fib t 18 8.3 line oxide i 1.5 r 1.1 : 7.0 32 3 3s 9.7 Si 2 a 81 2 81 1 83-3 2 2 l.l 3.7 fftt 6.1 3 1ft t 5 X aS.fi 81 t al i 1ft l 35 2 Mtgaesiue silicate s 1.6 3.8 : 6.1 27t 10 ; 9t6 Sk 2l ; 2d 1 5.8 2 t i.i 3.3 2ft . 6.3 3 3X 32 5 2 12 * 12 2 81 2 9.9 hta from srwae, paper VI (J) by the variation in intensity of color fom center to edge of the droplet. Results of the tests of time of pene tration are reported in Table 1. Water penetrated some portions of a paint coating much more rapidly than others. For example, with the unweathered coating of titanium diox ide paint X.9 mils thick, pink spots strongly contrasting wth the rest of the blue stripes could be seen within two hours. The spots slowly increased in size and number. After six hours, about half the total area of the stripes were pink, the other half still blue. Nine hours elapsed before the last of the blue color changed to pink. All other coatings exhibited a similar course of events, although differing in time scale, with an initial appearance of pink in spots that grew slowly in size and number until alt blue changed to pink. Essentially the same behavior has recently been reported by Pragst (26), who allowed an aqueous solu tion of methylene blue to diffuse through coatings of paint on glass. Although more time was required for water to penetrate the thicker coat ings of paint, the time was not always proportional to thickness. Thus with the unweathered coating of white lead paint 2.3 mils thick, first penetration was observed in 11 hours, and the coating 3 times as thick, 6.9 mils, was penetrated in 13 hours. Unweathered coatings of zinc oxide paint 1.8, 4.1, and 7.0 mils thick were all penetrated first in three hours. On the other hand, first penetration of. coatings of anti mony oxide, titanium dioxide, and magnesium silicate paints was always delayed by increased thickness at least as much as proportionality would indi cate. With weathered coatings of all paints, first penetration usually oc curred sooner than with corresponding unweathered coatings, and increased thickness delayed first penetration rel atively little. The relatively short time sometimes required for first penetration, and the lapse of time between first penetration and complete penetration, together with the variable relation between coating thickness and time of penetra tion, suggest that paint coatings have an exceedingly heterogeneous internal structure that is more readily pene trated in some places than in others (21, 29). Bell's (2) discussion of the structure of paint films, particularly his photographs of the formation of IJenard cells and electron photomicrograph of a cross section of a paint film, fur nish adequate explanation of the ob served irregularity of penetration. James (16) measured the time re quired for water to penetrate coatings of red iron oxide paints and the clear vehicles with which they were made by an electrical method. Most of his coat ings were 1.6 mils thick, thinner than any of the unweathered coatings re ported in this paper. Coatings of linseed-tung oil vehicle with or without pigments apparently were penetrated almost immediately, but their water content continued to increase for sev eral days. Coatings of clear and of pigmented phenolic-resin varnish, lin seed-rosin-epoxide resin, and linscedpentaerythritol alkyd vehicle were pen etrated in 15 to 90 minutes, but con tinued to gab water for several days. As reported in Table 1, unweathered coatings of white lead and of antimony oxide paints, which are low in water absorption, always required a longer time for first penetration and nearly always took longer for complete pene tration than coatings of similar thick ness of the high-absorbing titanium dioxide, zinc oxide, and magnesium silicate paints. Among the weathered coatings, white lead paint remained slowest to permit either first or com plete penetration, but antimony oxide paint, despite its low absorption, was penetrated about as rapidly as any of the other paints. For unweathered coatings, high absorption seems to hasten penetration, although the two are by no means exactly parallel. For weathered coatings, other factors ap pear to be more important than water absorption in determining speed of penetration. Blisters usually appeared in the coatings over the stripes of cobalt chloride either somewhat before or at least soon after penetration was com plete. It was difficult to tell when olistering began in the thicker weath ered coatings of zinc oxide and mag-, nesium silicate paints, because the coat ings were too hard to yield readily in blisters, and the blisters, when they became observable, were relatively large and only gently curved. There was seldom any observable blistering where no stripes of water-soluble mate rial were interposed between coating and glass. The pressure responsible for blisterbg arose therefore from osmosis (17). Adhesion may have been as badly impaired by penetration of water around as over the water-soluble stripes, but sufficient pressure to over come tiie mechanical resistance of the film to flexing may have been lacking where there was no soluble material to give rise to osmosis. Permeability to Water Vapor The permeability of free paint films to water vapor was measured by the Payne cup method (22, 25). Distilled water was placed in the cup, and the paint film to be observed was clamped In position. The cup was then placed in a desiccator over calcium chloride in a room held constantly at 70 F., and the cup was weighed daily to 2 GLD38042 determine the loss of water through the paint film. Steady state conditions were reached within one day. The rate of loss during the second and third days, which was nearly constant, was taken as a measure of the permeability of the film. The permeability is ex pressed as the number of milligrams of moisture vapor lost per square cen timeter of film area per day for a film 1 millimeter thick, which is known as the specific permeability, P,. The re lation is in which m is the weight of water lost in milligrams, q is the area of film exposed in square centimeters, l is the thickness of the film in millimeters, and / is the time in days. The calcu lation assumes that permeability is in versely proportional to film thickness. The assumption is not always correct (7, 29), but the error is not serious for the comparatively small variation in thickness of the films used for the study. Each film was weighed immediately before and again immediately after the permeability test. The film was then dried in a desiccator over calcium chloride and weighed once mpre. The difference between the second and third weighings was recorded as the moisture held by the film at the end of the test period. The difference be tween the first and third weighings was recorded as the change in weight (usually negative) of the moisturefree film during test. The results with oil and oleoresinous films are recorded in Table 2, in which the films are arranged in order of decreasing absorption of water by the unweathered films when soaked for three days, as reported in prior papers (5). The films are separated into four groups according to the ab sorption of water by the unweathered films. These groups are: above 49, be tween 20 and 49, between 10 and 20, or below 10 per cent of water by 'vo)- ume. Results with latex films are simi larly recorded in Table 3. The results show that the permea bility of paint films to water vapor is not primarily related to their water absorption. Among the unweathered pigmented oil and oleoresinous films, there was 100-fold variation in water absorption (0.8 to' 86.4), but only 9-'foId variation in specific permeabil ity (0.06 to 0.34). For the group with absorption greater than 49 per cent (average 73.9), the permeability rangerf from 0.15 to 0.30, and aver aged 0.23- For the group with absorp tion less than 10 per cent, the perme ability ranged from 0.06 to 0.33, and averaged 0.16. The next to the lowest permeability, 0.07, and the highest per meability, 0.54, occurred in the same group. Although weathering usually de creased both water absorption and per- Toble 2--SPECIFIC PERMEABILITY TO WATE* VAPOR (AT 70 F.J OF FREE FILMS OF OIL AMD OLEORESINOUS FAINTS, TABULATED IN ORDER OF DECREASING ABSORPTION OF WATER DURING 3-DAY SOAKING fin* Vtwr ahaerWdTfrlw y# af i --vetflt i Helitur* : OtaJp la iftltk-- m atahLa* far:ttsl* maif frw*n*Uny: a*aartad : ealfikt f i #f ) Rayi i --lek 4*U fur i f. t 4vti| : try tUa : fill V*o t A/*. m x *a*--ahilltjsAuri-- teat*. t taka* ()____t __ t teat 1 I ClMtUw at fUa rtNt x Vakicuf* Htoter ikuartatf: taa aaaktij t } day* x k/T, t I. t t WaaUwra* fttaa fftclfic i Mitw t flbift La rftfekwM rtaartsl t w i^l if > a* >, Aarlac t try fite s fll* {ywiimuritwii tm: t Utl I i tetooalla--t kjr Rc-jei- m < tkJPT CL .'SjiEjfiz-' 100 W, j MO ee. j aoarinauA4 66.4 *5J 03-3 2:1 A*5 ,61.6 S56.*) hvwep* T5-9 60.) )1.B 9.6 6666..90 *6.9 9J.3 6). *2.6 *> lnit|i 66.6 16.0 li.fi 1A.J u.y MID..9) 10.) hvai'M* 16.6 M 3t.oJ *?:!* 6.6 ,4.4 3<5.)> *3-.4) ).l 3.1 6..7) ,,1.6 Sun .6? ttn# 6.7 n n n n n nn *1 VI 71 71 7X31 S 71 S 71 n 71 n <ii n n n 71 71 7711 71 71 71 S n n 71 71 71 71 71 71 0.17 )0 .69 .19 .n .69 .19 .63 63 .16 .17 .56 .69 .*5 .19 .07 - -1 *11.) .69 .U .89 .15 .9 .a .11 .19 .16 .14 t(.) *:3> .96 .16 .U .U -.1303 ..S1O6 .Cfi a.J.152! .67 .16 6.9 74..30 13.-0fi fi.6 1.6 4.2 1.1 1.1 10.fi 1-3 l.fi 1.6 1.) 1.1 *:?) 6-3 6.1 l. 9.6 1.3 .7 9..50 1.5 .4 k.V .) S:V 1.3 .7 .5 1.0 1.0 .7 .5 .7 .6 .2 .6 : --01..34 -.1 -.1 -1.5 -.6 -.4 -1.9 -1-9 -1-4.9 -1.9 . -1.) A-i. 1.0 -- .43 -.6 -.2 -.5 M -.9 -.9 *0) -3 *::!* --.91 -.3 -9 5 -.6 -.1 0-.3 3*-..J6> 0 -.3 1 ).3 ' sMttalwslxr, Bm t 4.1 attac Mtfia * 3.6 attaa MU < 9.9 ilaafifilatj IIm * 4.9 iArtiMkr-wrjjfe$ i 4.1 9.9 rttaa aM4a a--A t <|M0* ()fi* 9.) tt talk)!* t MllOt f talkyi ilLO 1 t *0.9 96.0 96.9 90.) *0.* ir. s J.T t s maw na tjt, mmm o t mn t 90.7 : 1.0 4. : 41..19 ). 1 ). t 4.6 t t.9 s * t,T Hhpaif alllaaW tUkc mm ltlfMmmt-Uimtti.imUHmi* f|--alliaai* itlMt* filaxUa mMlM-tla*i m iMwihl *im7 ftn* 1711--It iniMQ. talkytt t sito x t h o .1 tm* (96* 9.) * m* * in* s lD ,, % 1 1 (j) . talk?* 1 iKA (. U.l 14.* 6-3 116*.43 . 4.6 ).* * x x 10.6 MMrm $m m m t ).0 (toll-- atot, Mm 1 3.0 <9&t--im Ala-- alkyl s 1 l.fi ifMtnlM, u m **40 ... 1 1 6.9 if--mol miothMMalatMt *x li) R > ). a--a!-- alltau* t 4.6 aVklfia la-- >iu (36* r.) 1 t 3.6 61-- -- 44.7 3.4 165..67 6.9 10.9 x 4.4 X X 9-6 mem ins a/% uw c m 10 rmcm 1 4,6 tUic a-- x 4 3.1 (941U 1--A t .2(`.4) *0 91--t ).9 ittito 3--A :sfcp 5.4 mta*1-- Aloxlfit r 4.9 ;ni*a Um 4.8 tLaaA.alke, Um t ).? :Htti 1--A 3.9 :Tft--1-- AlaUfia t 4.6 iVhlu 1--A ).4 :TUxl* AlexlAa > 2.6 iMtiaway oxlic : A 4.7 ;MbLU )-- : a*.o) no U-mt 4.9 tiAtUftav exits fi.O lAatljaeay wide m. itUyt x x m* xMU talked x*ka. :IL0 (I--. :tt0 iyfcea. t x t : T : uu (k c t.y. I*U <U0*r.|: iW> t l|--a. NklkjrA t 190--. t--aa. tlUa'a. 1 >lt0 1 5.) M 1.9 4.3 9.5 5A 4)..4) 4.6 t1..49 _ 1.5 ki. 9 1.5 : 4.5 : : t J.9 > 0.U s t .94 x .31 X X .06 X x .12 > . .10 ; ,)4 .19 x .10 t .Ifi tX X1 .IT x ..9336 .16 1 .ifi x 1 .92 x .15 : X 1 .U x .U x .15 * t .15 x .09 X t .15 x t 13 1 x 1 .U r .06 i .10 *:!P ; 1 : .25 -13 x , .be t .0? s t x ..5141 X xx .12 ; .12 X : A .05 : > A.i> .11 : X i *5 x ( .14 : 1-3 x -C5 x 3-0 16..72 xt *a.T.l x1 36 5-7 1-3 t -.7 : 3.0 1.1 X -1.0 X 2.6 .7 x -.1 < 2-9 >3' x -.4 x 3.6 1.5 X *.fi X 2.8 .6 -.3 X 3-9 l.fi ( l1O.4 x -.7 -.7 s 4.5 1^0 9 1 -.7 t 4.9 1.1 x -.3 x 3.0 1.6 x -.5 X 3.6 .6 1 -.4 j 4.6 .fi 1 -.) X 3,6 4.3 4.4 1.9 1 0 1 4.1 .6 .6 1.1 1 -.3 < -.4 x -2.3 x : x 4.9 3.< 3.5 .9 -.6 x 4.1 .4 .9 .7 *:?> .7 .7 1.0 .5 .9 1.5 .6 .4 1.0 ,, .5 k..4i) .4 .7 x *.3 !*::)> t -.1 x -.3 x X -.3 1O : -.3 x -.1 = Sf-s) : -.7 t -.2 : -.3 i 4.2 3.0 41- 4.3 4.5 2.5 4.6 3.5 X 3*4 3-9 46r.S8 . 4.4 4.7 4.0 --anj ran Itaeaet atl; *610* bmm toiled liae--d eUi *al*yi*---- 1 laet-otl xUAyi-rrxU vereixfi; *fttn ." --aaa yktaaH r*mii uraiin. ate *d--esi VDlv-- w 0,30 Me--t ia the eaaaarelal 'treether-ty*** yaiat. All'fUaa w fn*i {Axled) at agpraalaaUlj 75* 7. eacaft thoaa taxied "(36* r.F*a *(160* f.)*. IvaXuat U yamtkitti vr* aot Laeluded le --xUllae ef the var*je vtlwM for tlf prom?. Ak U--a-krutf pel*l mt* nth Lltulia Alcalde, ate*, and aaelt etluate la a Vot kfunoiihly ra>ortK. af w---oil v4ralab sM todiad lumwt oil; yl#w toimt 0.)9. At tnUa-krwaR yalat a4r with tiuaMa 1Loxide ud m-raiia allleet* It a "eopolyawr" vehicle; vaim* O.JO. 3 GLD30O43 nii-ability, water absorption generally was decreased proportionately much more than permeability. Thus for the most absorptive group, weathering les sened the average absorption from 73-9 to 20.7 per cent, whereas it only lessened the average permeability from 0.23 to 0.19. Among the unweathered pigmented latex films in Table 3, water absorption ranged from 7.0 to 75.9 per cent, well within the range for oil and oleoresinous paints, but permeability ranged from 1.44 to 6.45, far above the range of 0.06 to 0.54 for oil and oleoresinous paints. Weathering usually decreased the per meability of latex films proportionately much more than it did that of oil and oleoresinous films, but weather ing was less effective in decreasing the absorption of latex than of oil or oleo resinous films. In fact, weathered latex films often absorbed more water after they were weathered than before. With some latex films, weathering increased water absorption while it decreased permeability. All films absorbed moisture during the permeability test. The moisture absorption varied from 0.1 to 10.6 grams per 100 cubic centimeters of film. In general, those films that were high in water absorption in the soak ing test were also relatively high in moisture absorption during the perme ability test, although the parallelism between* water absorption and mois ture absorption was imperfect. Meas urement of moisture absorption was purely incidental to the measurement of permeability. As a quantitative measure of moisture absorption, the method was crude and subject .to much uncertainty because the films were ex posed to damp air on one face and to dry air on the other face. Also, there was often a small but variable amount of condensed water on the damp face. More reliable measure ments of the hygroscopicity of films of some oil paints were reported in paper II (5), from which it can be seen that the moisture absorbed dur ing the permeability test was only a fraction of the absorption by similar films when both faces were exposed for three days to air at 97 per cent relative humidity. Thus a film of zinc oxide in linseed oil absorbed 15.8 grams of moisture per 100 cubic cen timeters from air of 97 per cent rela tive humidity, but only 6.0 grams dur- ` ing the permeability test. There was nearly always a slight change in the weight of the dry films during the permeability test. Among the unweathered films, 57 lost from 0.1 to 5.3 (only 4 of them over 1.9) grams per 100 cubic centimeters, 4 remained unchanged, and 1 gained 0.5 gram per 100 cubic centimeters. Among the weathered films, 43 lost from 0.1 to 2.3 grams per 100 cubic centimeters, 3 remained unchanged, and 11 gained from 0.1 to 0.6 gram per 100 cubic centimeters. Paint films, of course, generally lose weight stead ily after they have passed the point of maximum weight soon after they have been formed. The small number of positive changes in weight observed may indicate that the downward course in weight may sometimes be inter rupted by temporary additions of oxy gen or combined moisture. To study the characteristic effects of different pigments and of different vehicles on the specific permeability of films, parts of the data of Tables 2 and 3 were regrouped, certain aver table 3 --SPECIFIC PERMEABILITY TO WATER VAPOR (AT 70 F.) OF FREE FILMS OF LATEX FAINTS. TABULATED IN ORDER OF DECREAS ING ABSORPTION OF WATER DURING 3-DAY SOAKING AS REPORTED IN PAPER VII OF THIS SERIES (S) Unweathered films t1 Composition of fila Weathered flUs Water absorbed on soaking for 3 days /T0 Percent by volwe 75.9 3(57.9) 51.5 50.5 L9.7 Average 56.9 30.2 27.7 23.0 22.7 20.6 20.4 Average 24.1 Specific Moisture :Change in: Italck- peraeability: absorbed ;velght of: aes* P, during :dry fils : of :praeabillty.- during : file teat : teat Ptgjsent Latex water : Specific : Moisture .'Change in: ifcick- : vehicles abaorbed :peraeablllty: absorbed .`weight of: ness on soaking: p# : during {dry m* j of tor 3 days: spenaeablUty: during : film ,A/V0 : : test : test : Mg- b b . per : 0*. per Ga. per: Mil ca. per day: 100 oc. : 10O cc.j mwnt by:. . j*r s Go. per : W. per : mil voliae icm.H per day: 130 cc. : 100 cc.-: GROUP HITS A/V0 MORE TBM >9 FEftCOT , 5.98 l.L9) 1(1.33) 3-15 6.03 5.L9 ,1.6 ,k.6) 2(1.2) 4.2 .1 .2 ,-o.8 k-. 4-.L) -.6 0 -.2 , 3.4 :Ccmaercialt T102 silicates :?Va2 3.1):CosMtrclal> Tide, silicates :PVAi tacrylic 12.5 : 1.31. -. 0.8 ; -0.5 : 3-1 .....j...... 4 68.7) : 2(1.31) : 3(*.R> : Sf3-A> 3.9 :White lead :sty.-but. MS 15-0 : l.U : 2.9 t -.9 3.6 3.4 :2ioc oxide 2,7 :Zinc,oxide ;acryllc :PWA 42.3 : ?.W> A : -1.5 2.5 61.7 i 6.60 : .3 : -2.1 : 2*5 5-U 1-5 -.1 3.3 : : 3*9 t 3.61 : 1.0 : -1.2 : 2.9 o r c u p im a /v mum mb 69 mem 2.42 t.l -.1 4.6 5Magnesian silicate :acrylic 1X9 : .62 i .4 : *.l : . 4.J 3.05 2.8 -.i k.k :Ztae oxide :sty.-but. MO *3.T : 94 1 .7 : -.7 : 3.T 3*29 .7 -.9 4.8 :71taalua dioxide :PVA i 13*0 : l.Sk t .2 : a.2 s 4.2 1.44 3.60 1.8 -.4 4.9 :White lead 1.7 -.2 3.8 :(NPw *1v b silicate tacryllc ? jaty.-but. MB I *53 : 1.5 .42 t .6 : -.1 : 4.6 : -.4 s 3.5 4.57 1.2 -.1 5.2 .Zinc oxide :Sty. -but. ie.7 i -85 : 3-6 : *.9 1 w.e 3.09 1.5 *.3 k.f i s 3. : .77 : l.e : -.2 : 4.2 g r o u p w it h A/v. Bnum 10 a x > so menrr 15.3 no U.9 10.9 Average 12.8 302 6.1*5 6.08 5.23 . 5.42 1.2 ..6 .7 -1.5 2.0 l.l : -.2 1.2 : -.6 3.7 :Titaolu* dioxide 5.8 :>MgnMlun silicate 7.3 :White lead 5.0 :Titanl\m dioxide 5.L : :sty.-but. MB: 5.9 : .13 : 1.9 jsty.-but. 10.6 ; .21 9 :aty.-but. L6.7 : .05. : 5.4 :aty.-but. : 15*2 s 1.03 3 1.4 : 12.1 : .35 : 2.4 : -.7 ; -*.,2 : -l.l s -.3 1 : : : = -5. 3.6 4.2 5.0 5.0 4.4 , 8.9 2(8.7) i(8.2> 7.0 Average 7-9 ,, 209 3(8.77) I(2.3<.) 3.&7 3-13 . 20 1(6.9) 1(1.0' .7 1.6 g r o u p v x t b a A,, u s s ib a x l o retem : . -2.5 : 4.4 :Magnelias silicate 4-5.3) 3(5.4) :lfo plgsent 4-S-D : 3(4.0) :Xo plgnent : -.2 : 4.7 jTUaMun dioxide :PVA :sty.-but :PVA tacryllc : -1.3 : 4-5 : : i 18.2 : 1.10 .4 : +.4 : 4.0 : Styrene-butadiene fllas were destroyed by cracking dur log weathering : 2(18.9) : *1.33) : 4-7) = 4*-6) < A3.5) 6.5 : 1.61 : -3 S *.l : 5.0 : 12.3 i 1.35 -3 : *.2 : L-5 JrTrade breed Utew# vere used. "Acrylic* *etn 4 polyscrylic latexi "FVA" polyvinyl acetate; "sty,-but." s styrene-butadiene; "ety.*but. M?" * styrenebutadiene latex with Addition or netfcyl eeHulose. zA trade-bread PVA Latex paint. ^Values la pexanthese* war* not included In computation of the average value* for the group. mm trade-brand pnlat tbe uaweathered file of which v% leached in water nod redried befor* testing* fbe Absorption of water by the leeched fils use not measured quantitatively but was observed qualitatively to be such leas than that of the unleashed file* 4 GLD38044 f .1 - \ age values computed, and the results recorded in Table 4. Thus in the first line of Table 4 are reported the aver age permeability and average water absorption of the four unweathered films and the four weathered films of the white lead paints made with the raw linseed oil, bodied linseed oil, alkyd-resin varnish, and phenolicresin varnish vehicles. In the eighth line of Table 4 are reported the aver age permeability and average water absorption of five unweathered and five weathered films with phenolicresin varnish pigmented with white lead, titanium dioxide, zinc oxide, lead-zinc, and titanium zinc. In oil and oleoresinous vehicles, white lead and zinc oxide made films of lower specific permeability than cor responding films with antimony oxide, magnesium silicate, or titanium diox ide. White lead and 2inc oxide are chemically basic and capable of reac tion with free fatty acids or acid de composition products in paint films, whereas the other pigments are unreactive in this sense. White lead made' less permeable films than zinc oxide. The pigments had very different effects on water absorption. Thus white lead caused low absorption as well as low permeability, but zinc oxide caused highest absorption despite relatively low permeability. Antimony oxide, which produced lowest absorption, stood high in permeability. In latex vehicles, pigments had rela tively less effect on permeability than they did in oil and oleoresinous vehi cles. White lead, magnesium silicate, and titanium dioxide stood in the same relative order, but zinc oxide was out of line in that it gave higher per meability than any other pigment in latex vehicles'. In water absorption by latex films, white lead failed to take the low relative position it did among oil and oleoresinous films. Among oil and oleoresinous vehi cles, phenolic-resin varnish made the least permeable and raw linseed oil the most permeable paint films. Bodied linseed oil and alkyd-resin varnish made films of intermediate and about equal permeability. Phenolic-resin films were also lowest in water absorption, whereas bodied linseed oil was inter mediate, and alkyd resin and raw lin seed oil were relatively high. The unpigmented oil and oleoresinous vehi cles (recorded in Table 2 but not in Table 4) made clear films that were more permeable than the pigmented films with the same vehicles, but in water absorption the clear films were intermediate among the pigmented films with the same vehicles. Among latex vehicles, styrene-buta diene made pigmented films that, be fore weathering, were on the average more permeable but less absorptive of water than the pigmented films of the other latexes. ITie unweathered pig mented films with the other latexes did not differ significantly in either permeabilily or water absorption. Pig mented films of styrene-butadiene and of styrene-butadiene with methyl cel lulose became much less permeable after they were weathered, so much to' that they were little higher in perme ability than some oil paints. On the other hand, weathering decreased the permeability of pigmented polyacrylate and polyvinyl acetate films only mod erately, and they remained far more permeable than any oil paints. Weathering tended to increase the water absorption of the latex films ex cept those made of styrene-butadiene with methyl cellulose, for which the absorption decreased after weathering. Unpigmented films of polyacrylate and txilyvinyl acetate were less perme able before weathering than any of the pigmented films of the same latex, and after weathering were less permeable than corresponding films pigmented with titanium dioxide or zinc oxide. Such behavior stands in strong con trast with that of the clear and pig mented films with oil and oleoresinous vehicles. The unpigmented film of styrene-butadiene was somewhat more permeable than any of its pigmented films, but the unpigmented film could not be tested after weathering because Table 4.--SPECIFIC PERMEABILITY AND WATER ABSORPTION AS AFFECTED BY PIGMENT AND VEHICLE INGREDIENTS OF FILMS Pigment or vehicle ingredient considered Kind and number of vehicle or pigment ingredients averaged Specific permeability, Pg Water absorption, A/V ------- - - r . - - - - a 1 < 1 1 1 UnweatherelkWeathered Unweathered: Weathered films films films : films Mg. mm. per cm^per Pigments: - White iead, L oil and oleoresinous (4) Lead-zinc, LZ25 Zinc oxide, Z oil and oleoresinous (4) oil and oleoresinous (4) Titanium-sine, T225 oil and oleoresinous (4) Antimony oxide, A oil and oleoresinous (4) Magnesium Bilicate, X Titanium dioxide, T Soil and oleoresinous (4) oil and oleoresinous (4) Oil and oleoresinous vehicles Phenolic-resin varnish L, T. Z. LZ25, TZ25 (5) Bodied linseed oil L, T, Z, LZ2c , TZ,, (5): Alkyd-resin varnish Raw linseed oil L. T, Z, LZ25, TZ25 (5) L, T, Z, LZ25, TZ25 (5) Pigments: White lead, L latex vehicles (3) Magnesium silicate, X .latex vehicles (4) Titanium dioxide, T latex vehicles (4) Zinc oxide, Z latex vehicles (4) Latex vehicles; Polyacrylic resin L. T, Z. X (4) Styrene-butadiene and methyl cellulose L, T, Z, X (4) Polyvinyl acetate T, Z, X (3) Styrene-butadiene L, T, Z, X (4) 0.10 .14 .18 .19 .20 .24 .28 .13 .17 .17 .25 3.56 3.76 4.08 4.78 3.44 3.48 3.72 5.58 0.08 .09 .17 .13 .18 .25 .18 .10 i .16 .12 .22 .57 .59 1.15 t 3.30 i 2.G4 \ .65 3.18 :39 Percent by volume 5.9 31.0 46.7 44.6 2.3 34.9 13.1 5.1 13.8 16.4 13.5 1.8 11.4 4.4 5.0 16.3 49.6 42.2 3.7 10.9 13.2 15.5 28.7 18.2 14.0 37.1 26.3 39.0 10.6 36.6 27.6 53.8 28.8 27.2 14.1 13.2 31.6 15.3 GL0380 *i5 it failed to remain intact. The high permeability of the latex vehicles when compared with oil and oleoresinous vehicles, together with the differences in relative permeability of unpigmented and pigmented films and the generally low swelling efficiency of latex paints, even before weathering, that was pointed out in paper Vll (5), ail suggest that the permeability of latex depends largely on a condition of porosity caused by failure of the droplets of emulsified polymers to coalesce to form a continuous matrix (9). This condition is absent in oil and oleoresinous films. It should be noted, however, that latex paints of low permeability comparable with oil paints are said to be obtainable by special formulation with suitable pig ments (14). Variation of Permeability with Temperature Further information about the nature of the permeability of paint films was obtained by studying the effect of tem perature on the permeability of films of 12 paints. The equation given pre viously for specific permeability, P,, as customarily used by paint technolo gists applies at only one temperature, at which the vapor pressures on the two faces of the film differ always by the same amount. A more general re lation for a permeability constant, P, is in which m is the mass of water vapor in grams that passes through a him of thickness / in centimeters and area q in square centimeters during time t in hours under a difference in vapor pressure of p in centimeters of mer cury (6, 7, 11, 15, 24, 29, 32, 33). (Choice of units of measurement var ies among the workers.) If the vapor pressure over calcium chloride is con sidered zero, p may be taken as the vapor pressure of water at the tem perature of measurement, and then: at 38 F., P = 72.3 X 10 P, at 70 F., P = 22.3 X 10 ~T P, at 156 F,, P = 1.88 X 10-7P, If the permeability constant is in versely proportional to the thiacness of film and directly proportional to the difference in vapor pressure, the relation between permeability constant and temperature is P = P. e -E/RT in which P,, is the permeability con stant independent of temperature, E is an activation energy, R is the gas con stant, T is the absolute temperature, and e is the base of natural logarithms.6 6 The logarithm of P should then be proportional to l/T (6,7,11, 24, 32) . Films of 12 paints and clear vehi cles reported in. Tables 2 and 3 were selected for measurement of permea bility at different temperatures. The permeability of each film was deter mined first at 38 F., then at 70 F., and finally at 156 F. Specific perme ability was converted to a permeabil ity constant by the equations cited. The results appear in Fig. t, in which P is plotted on a logarithmic scale against 1/T. The logarithm of the permeability constant was linearly related to 1/T for all films tested. The constant was much larger for latex films than for oil and oleoresinous films, although at 156 F. (1/T = 2.9 X 10 "*) the largest constant for an oil film was nearly as large as the smallest constant for a latex film. The 12 constants spread over a much wider range at 38 F. (1/T = 3 6 X 10 -) than at 156 F. For six of the eight oil or oleoresin ous films, the slope of the line relating log P to 1/T was negative, which in dicates that the activation energy was greater at high than at low tempera ture. Such negative slope has generally been reported in the literature on per meability of films of plastics and paints to water vapor ana gases. But for two oil or oleoresinous paints, white lead in alkyd resin and weathered white lead in raw linseed oil (lines 6 and 8 in Fig. 1), and for all four latex films studied, the slope of log P versus 1/T was positive. It therefore appears possible for the ac tivation energy to be greater at low than at high temperature. Positive slope is not peculiar to white lead, raw linseed oil, the alkyd resin, or weathered films,, because negative slopes were observed for unweathered white lead in raw linseed oil and for weathered lead-zinc in alkyd resin. Pos itive slope, however, may be character istic of latex films. The permeability constant is further related to the diffusion constant, D, and the solubility of water vapor in the film, S, by the relation P = DS. The activation energy is the sum of the activation energy for diffusion and the heat of solution of water vapor in the film. Since a decrease in diffusion with increase in temperature is highly improbable, the positive slope of log P versus 1/T lot latex and some oil films suggests that the solubility of water vapor in such films decreases with ris ing temperature. Experimental data to test the suggestion are not yet avail able. It should be pointed out that water vapor passed through all films more rapidly at 156 F. than at 38 F., whether the permeability constant in- * creased or decreased. Thus for the film of titanium dioxide in polyacrylic latex (line 12 in Fig. 1), for which log P fell off most rapidly with rising tem perature, the Payne specific permeabil ity, P,, was 29-1 at 156 F. and only 2.16 at 38 F. The Payne-cup method was chosen for the measurements of permeability because much of the published data for paint films has been obtained by that method. The method is attractive for its convenience, but it is open to possible objections, especially for de terminations at other than room tem perature. Once the cup has been dosed with a film that is impenetrable by air under a pressure gradient, there can. be no movement of air in or out of the cup to offset the change in volume of enclosed air when the temperature is altered or when moisture vapor es capes through the film. Flexible films arc then stretched, and their exposed area is increased by the difference be tween the internal and external pres sures. On the other hand, if the film is penetrable by air under a pressure gradient, there may also be loss of moisture vapor by penetration in addi tion to the loss by permeation. More over, poor circulation of air inside and outside the cup may result in pres sure of moisture vapor somewhat be low saturation at the film interface within the cup and greater than zero at the film interface outside the cup. Thus it is desirable to test further the relations between the temperature co efficient of the permeability constant and the absolute temperature by some method that is less open to objection than the Payne cup. More recent tests have shown that the logarithm of the permeability con stant may not always be related lin early to the reciprocal of the absolute temperature. Chips of paint from the Washington Mansion at Mount Ver non, Virginia, gave permeability con stants of 8.8 X 10 at 36 F., 29.0 X 10 -1 at 70 F., and 36.1 X 10 at 156 F. The constant at 36 was much too low to comply with a linear relation. The chips were 73 to 110 mils thick, and were made up of white paint containing sand to produce a rough surface. The paint was applied in about 20 paintings from 1773 to 1956. Penetration bv Air under Pressure Difference The term breathing properly applies to a movement of air into or out of a discretely porous or cellular body such as the lungs under a difference of pressure. Gardner (12) apparently so intended, when he introduced the terra, to distinguish some paint films GL038046 (000r Tobta 5.---PENETRATION OF FILMS OF OIL PAINTS fcY AIR UNDER A PRESSURE ORADIENT Competition of ott-coouinln* Aim Commercial "brextW-type*' paint.... Commercial "bHatg-mfrtant* paint... Rate ol paaaife of air u b- der 048 atmoapbere differ* nee of preaeure through Hramt 9.14 quart centimeters of volume Unweatb* Weathered of Aim ered Aim film Cc. per min. Cc. per min. 0.89 .SO 00 0_._6 Raw Unaeed oil potato piemented with: Basie carbonate whitelead...______ Baak carbonate white lead................. .60 .90 0 0 00 Zinc oxide-..-........................... Zinc oxide........................................... . Zinc oxide...___ --........... . .60 66 .90 6.8 .6 0 .26 Titanium dioxide... Titanium dioxide... Titanium dioxide... Titanium dioxide... Titanium dioxide.. Magnesium silicate. 21.9 .66 18.8 .60 .46 .80 0 .90 0 Toble PENETRATION OP FILMS OP LATEX PAINTS 8Y AIR UNDER A PRESSURE GRADIENT Fig. 1.--Permeobility constant, P, plotted on o logarithmic scole os o function of the reciprocal of the absolute temperature for Alms num bered os follows: i. Lead-xinc in atkycf resin, weothered) 2. White lead in bodied linseed oil; 3. White leod In row linseed oil; A. Zinc oxide in row linseed oil; 6. Rodied linseed oil without pigment; 6. White lead in olkyti resin; 7. Titanium dioxide in row linseed oil; 8. White lead In row linseed oil, weothered; 9. Polyvinyl ocelote latex without pigment; 10.- Titanium dioxide in styrene-butadiene latex; IT. White lead in styrene-butadiene iotex; 13. Titanium diox ide in ocryllc iotex. from others. All breathing paint films should therefore be readily penetrated by air, but not all films should breathe, even though all paint films tested so far are permeable to water vapor in at least some degree. It was therefore of interest to learn whether ordinary paint films let air pass through under a pressure gradient, and if not, whether there are paint films that do. Composition of latex-containing Alva Commercial polyvinyl acetate paint... Polyvinyl acetate latex with: No pigment____________ _ Zinc oxide..____ ................ Titanium dioxide...______________ Magnesium silicate............................ Polyacrviate latex with: No pigment..... ................................. White lead.......................... ............... Zinc oxide____ ................... Titanium dioxide............................... Magnesium silicate...................... . Styrene-butadiene latex with: No pigment________________ _____ White lead.......................................... Zinc oxide................ ........... .. Titanium dioxide..-------------- ...... Magnesium silicate............ .. Styrene-butadiene latex with methyl ceUukwe and: White lead.................................... . Zinc oxide...____................ Titanium dioxide...... ................... Magnesium silicate........................... Rate of paaasge of air under 0-88 atmosphere difference of pressure through 9.14 square eeatimeters of Uawmtbered film Cc. per min. Weathered Aim Cc. per min. 00 0 1800 r* OF or 0 OF 1500 T 0 0 0 0 0 46 F 0 0 2000 F OF OF 64 F 78 F 1600 F 1 0 280 0 1600 F 1 OF 920 F 1 65 OF 0 00 A suitable device for the purpose was made with an Asbeck-Van Loo critical pigment volume cell (1). The cell consists of two bell-shaped parts of glass, the wide portions of which fit together in a ground-glass joint. The bottom part, which forms the male piece of the joint, is closed by a fritted glass disk set flush with the glass rim. The narrow portion of each part ends in an outlet tube. The out let of the male part was inserted through a rubber stopper in the mouth of a 1-liter filter flask, the side arm of which was connected through a stopcock to a water aspirator for suc tion. An open-arm mercury manometer was provided to measure the pressure of air in the filter flask. There was also a bleeder tube with stopcock to admit air for precise adjustment of suction at the beginning of a test, prompt release of suction after a test, and admission of air when calibrating the apparatus. The total volume of air at atmospheric pressure contained within the apparatus between the sur face of the fritted glass disk and the aspirator stopcock was 1290 cubic centimeters. The symbol F Indicates Alma that could be seen to contain small vacuoles caused by retention of loam after tie wet Aim hardened. A circular sample of paint film to be tested was cut exactly to the diam eter of the fritted glass disk. A ring made of a short piece of copper tub ing 2 centimeters in diameter inside was centered over the film. A rim of molten paraffin was then formed be tween the outer surface of the copper ring and the glass rim of the cell and allowed to harden. Thus only the por tion of the paint film enclosed within the copper ring was exposed for pos sible penetration by air. The female part of the cell was then placed in position over the sample, and its out let tube was dosed by a short piece of rubber tubing and pinchcock. Air was evacuated from the appa ratus and the suction was adjusted through the bleeder tube until the manometer registered --28.9 centi meters of mercury, a pressure differ ence of approximately 0.38 atmos phere. A constant level in the manom eter for five minutes indicated that the apparatus was free from leaks. The pinchcock at the outlet of the female part of the cell was then removed, and a stop-watch was set in motion at the same time. If the paint film permitted air to pass, the manometer reading was recorded when exactly five minutes had elapsed. Impenetrable films, of course, permitted no change in manometer reading. If a film was so porous that the pressure rose to --7.9 centimeters of mercury within five minutes, *he time required for pressure to rise from --28.9 to --7.9 was recorded, and several successive tests of the film were made. To permit computation of the rate of passage of air through the films, the apparatus was calibrated as fol lows: The fritted glass disk was sealed with a disk of paper entirely covered by a layer of paramn. The system was evacuated to a manometer reading of --28.9 centimeters of mercury. Succes sive volumes of approximately 50 cubic centimeters, accurately measured in a gas burette, were admitted through the bleeder tube, and the manometer reading was recorded after each ad mission until the manometer reached --7.9 centimeters of mercury. A cali- 7 GL036047 bration curve was then drawn to relate manometer readings to volume of air at atmospheric pressure. When the fritted glas disk was left open, air passed through.it at the rate of 3320 cubic centimeters a minute. When an impenetrable paint film was perforated five times with the point of a common pin and then tested, it al lowed 321 cubic centimeters of air to pass a minute. Table 5 reports the results obtained with some oil paints. A commercial "breather-type" paint, said to resist blistering because it is porous enough to breathe, allowed no air to pass through its unweathered film and only 0.6 cubic centimeters a minute through its weathered film. Its specific permea bility to water vapor (Table 2) was also very low. Such paint cannot prop erly be said to breathe, although it may well be reasonably resistant to moisture blistering. Another commer cial paint said to resist blistering but not daimed to breathe allowed no air to pass and was also vety low in spe cific permeability to water vapor (Table 2). Single-pigment paints made with raw linseed oil and white lead, zinc oxide, titanium dioxide, or magnesium silicate at 0.30 pigment volume, which is within the range of pigment vol ume customary for house paints, al lowed no air to pass through their un weathered films. The weathered films were also impenetrable or, in the case of the weathered zinc oxide film, very neatly so. Unweathered films porous enough to pass air, however, were made with ztnc oxide and with titan ium dioxide by raising the pigment volume to a point far above that prac ticable for house paints. Thus the zinc oxide paint film at 0.55 pigment vol ume passed 0.6 cubic centimeters of air a minute, and at 0.60 pigment vol ume it passed 5.3 cubic centimeters a minute. The titanium dioxide paint film at 0.55 pigment volume passed 13.3 and at 0.60 pigment volume 21.9 cubic centimeters a minute. Films of white lead paint, weathered or un weathered, were impenetrable even at 0.60 pigment volume. The critical pig ment volume, the point above which there is no longer enough linseed oil to fill all the spaces between particles of pigment in the dried films (1, 3, 10, 27), was shown in paper IV (5) to lie between 0.55 and 0.60 for white lead paint, 0.45 and 0.50 for zinc oxide paint, and 0.40 and 0.45 for titanium dioxide paint. It may be con cluded, therefore, that films of oil paints breathe in the sense of becom ing penetrable by air under a pressure gradient when the paints are made at a pigment volume somewhat above the critical. The data of Table 5 suggest 8 further that weathering tends to de velop penetrability to air in some films made with effective anti-foaming* agents. Such agents were not used in that before weathering were impene the latex paints made in the laboratory. trable. Sufficient weathering to cause The latex fillms that were tested for paint films to' crack renders them penetration by air were therefore ex readily penetrable to ait, of course. amined by transmitted light under a Results with films of latex paints microscope. Some of the films (indi are reported in Table 6. Most of the cated in Table 6) contained easily ob unweathered films and 8 of the 18 servable vacuoles that presumably were weathered films allowed no air to left by foam in the wet paint after the pass. Four unweathered and 10 film hardened. Some vacuoles clearly weathered films were penetrated by air. afforded passage from one side of the Two unweathered ana four weathered film to the other. Others apparently films were porous enough to let were completely enclosed within the through 920 to 2000 cubic centimeters film. Most of the penetrability of latex a minute. Penetrable weathered films films to air may therefore be attrib always passed more air than their un uted to open pores left by foamy weathered counterparts. Neither the paint. There were some films with composition of the latex nor the kind vacuoles that were impenetrable, how of pigment seems to govern penetra ever, and some apparently free from bility, because both penetrable and im vacuoles that were penetrable, espe penetrable films were obtained with cially the weathered film of white lead each kind of latex and with each pig in styrene-butadiene latex that passed ment. 230 cubic centimeters of air a minute. The behavior of latex paints in But permeability of the films to water tests of permeability to water vajpor vapor apparently was unrelated to the and their low swelling efficiency when presence of vacuoles, because the eight soaked in water were attributed to unweathered latex films with observa porosity caused by failure of droplets ble vacuoles ranged in specific perme of emulsion to coalesce completely ability from 1.44 to 6.03 (average when the films harden. Such porosity 3.87), whereas the 11 films without evidently need not render the films observable vacuoles ranged from 1.33 penetrable by air if the pores are not to 6.77 (average 4.24). connected with one another. Penetra Most films that were penetrable to bility by air may require larger pores air could be tested repeatedly with that join together to afford continu concordant results, which indicates ous passageway. - that the films were unaltered by the Clearly, water vapor permeates most paint films by some mechanism other pressure gradient of 0.38 atmosphere during test. Four weathered films of than passage through interconnected latex paints, three with and one with pores, because penetration by air under out observable vacuoles, were excep- a pressure gradient and permeability . tions. With the film of white lead in to water vapor are not at all parallel. polyaciylic latex, five successive tests* Fifteen unweathered latex films that of air penetrability gave rates of 78, were impenetrable by air ranged in 96, 118, 123, and 145 cubic centi specific permeability to water vapor meters a minute, respectively. Such in from 1.33 to 6.77 (Table 3), com creasing penetrability might be attrib pared with a range of 0.07 to_ 0.45 uted to repeated fracturing of the walls (raw linseed oil without pigment of vacuoles under stress exerted by the 1.39) for films of oil paint (Table 2). pressure gradient on a film rigid Latex films through which air passed enough to resist compression. But the at 1500 or 1800 cubic centimeters a minute were lower in specific perme other three exceptional films exhibited a reverse behavior, a decrease in pene ability to water vapor than some of the trability with each successive test. The latex films through which no air weathered film of unpigmented poly assed. Weathering rendered penetra acrylic latex gave successive rates of te some latex films that were impene ' penetration of 64, 45, and 36 cubic trable to air before weathering, and centimeters of air a minute, the weath increased the penetrability of those ered film of zinc oxide in styrene- latex films that were previously pene butadiene latex containing methyl cel trable. Weathering often had the op lulose gave successive rates of 65, 56, posite effect on specific permeability to 54, 52, and 51 cubic centimeters a water vapor, however. Thus, for ex minute, and the weathered film of ample, when films of titanium dioxide white lead in styrene-butadiene latex, in styrene-butadiene latex were which contained no observable vac weathered, penetrability increased 46 to uoles, gave successive rates of 230 and 1600 cubic centimeters of air a minute, 90 cubic centimeters a minute. De but specific permeability to water va crease in penetrability in successive por decreased from 5.23 to 1.03. tests may indicate that the film be Latex paints have a notable tend comes permanently compressed, and ency to retain foam unless they are the effective cross section of its pores GLD38048 i i i 1 3# aa I I )/ 5 -t I " becomes diminished by application of pressure. Conclusions 1. The penetration of water through coatings of oil paints can be observed by application of the coatings over a thin layer of blue cobalt chloride in methyl cellulose on clean glass.' The coatings may then be weathered before testing by exposure to artificial sun light (without spraying with water). Tests are made by immersing the coated specimens in water and observ ing the time required for the cobalt chloride to change from blue to pink and for the coating to become blistered. 2. Penetration of water occurs earlier in some portions of a given coating than in others. First penetration through unweathered coatings was ob served in 2 to 13 hours, but penetra tion was completed only after 9 to 48 hours. The "wet line" advancing through a coating evidently is of very irregular depth, which suggests a heterogeneous film structure. Although longer time is required in general for water to penetrate thick than thin coat ings of a given paint, the relation be tween time ana thickness is not sharply defined, as might be expected from the observed irregularity in penetration. Blistering does not occur until penetration is well advanced, but may occur before penetration is com plete. 3. Penetration occurs faster in coat ings of high water absorption, such as zinc oxide paint, than in coatings of low absorption, such as white lead paint. Weathered coatings are usually penetrated more rapidly and blistered sooner than are unweathered coatings of similar paint. The relations between time of penetration and thickness of coating and between penetration and water absorption are also more obscure. 4. The mechanism of blistering in these tests is largely osmotic, because the blistering occurs much sooner where the water-soluble cobalt chloride and methyl cellulose are interposed be tween coating and glass than else where. 5. Measurements of the specific per meability of paint films to water vapor (Payne cup method at 70 F.) re vealed that permeability and water ab sorption bear little or no relation to each other. Both high and low permea bilities were found among highly ab sorptive films and also among films of low absorption. As groups, oil and oleoresinous paint films are very much less permeable to water vapor than are latex paints, yet the range in water absorption of the latex paints falls within the range for oil and oleoresin ous paints. If water absorption affords an indication of the resistance of a can be: made with pigment volumes aint to moisture blistering, permea- well above the critical pigment volume. ility to water vapor does not, at least A commercial "breather-type" paint within the range of permeabilities of was found to be impenetrable by ait the oil and oleoresinous paints. until its film had been weathered. Im 6. During permeability tests in penetrable films may become porous Payne cups, paint films absorbed much enough to breathe after they have been less moisture than when they were weathered, but some paint films re soaked in water. The amounts were main impenetrable after 15 days of roughly parallel, however. Such absorp artificial weathering. Films of latex tion in Payne cups may be high when paints made at pigment volumes suit the permeability is low, and may be able for exterior use are impenetrable, low when the permeability is relatively or nearly so, provided that the films high. are free from vacuoles left when 7. In oil and oleoresinous vehicles, foamy paint hardens. Vacuoles may- both white lead and zinc oxide make make the films penetrable or leave paint films of lower permeability than them still impenetrable, depending on is obtained with titanium dioxide, an whether die vacuoles are connected timony oxide, or magnesium silicate. with one another or remain isolated. White lead and antimony oxide impart 10. The permeability of paint films low water absorption, however, where to water vapor is not governed by the as zinc oxide imparts very high water presence of pores through which air absorption. In latex vehicles, permea can pass under a pressure gradient. bility is high with any of these pig Films highly permeable to water vapor ments, and both white lead and zinc may be entirely impenetrable by air. oxide tend to cause higher water ab On the other hand, latex films of very sorption than titanium .dioxide or high penetrability by air because of magnesium silicate. Among the vehi vacuoles may be no more perhieable to cles tested, phenolic-resin varnish water vapor than other latex films made paint films that were usually low that are impenetrable to air. in both permeability and water ab sorption. Raw linseed oil vehicle fa Literature Cited vored high permeability and high ab sorption, bodied linseed oil held an 1. Asbeck, W. K. and M. Van Loo. 1949. Critical pigment volume relationships. intermediate position for both permea Indus, and Eng. Chem. 41: 1470-5. bility and absorption, and alkyd-resin 0 2. Bell, S. H. 1955. The structure of paint vehicle developed intermediate per meability but even greater absorption than raw linseed oil. films. Jour, of the Oil and Colour Chemists' Assoc. 58; 595--618. 3. Blackmore. P. O. 1956. Film formation, film properties, and film deterioration, 8. The specific permeability of all four, of the Oil and Colour Chemists' paint films increases rapidly with tem perature. Study of films of 12 paints, Assoc. 39: 23, 24. 4. Broekex, J. F. 1955. Development of exterior coatings for wood. Paint, Oil however, showed that the logarithm of and Chemical Review 118 (11): 16, the permeability constant, which con 17, 20. .............. ...... siders differences in vapor pressure, is linearly related to the reciprocal of the 5. Browne, F. L 1953-56. Swelling of paint films in water. Paper I. Jour, of For. Prod. Res. Soc. HI (5): 108-24; absolute temperature, which is to be Paper II. Jour, of For. Prod. Res. Soc. expected of films that are not too hy IV (6): 391--400; Paper III. For. Prod. drophilic and do not swell too much. Latex films differ strikingly from most Jour. V (1): 92-6; Paper IV. For. Prod. Jour. V (2): 142-6; Paper V. For. Prod. Jour. V (3): 192-200; oil or oleoresinous films in that their Paper VI. For. Prod. Jour. VI (4): permeability constant js not only much 152-9; Paper VII. For. Prod. Jour. higher, but also decreases as tempera ture rises. There were two oil paints, VI (6): 235-40; Paper VIII. For. Prod Jour. VI (8): 312-118; Paper IX. For. Prod. Jour. VI (10): 453-8. however, for which the permeability 6. Brubaker, D. W. and Kurt Kammer- constant decreased with rising temper meyer. 1952. Separation of gases by ature. For such films, the solubility of water vapor in the film probably falls means of permeable membranes. Indus, and Eng. Chem. 44: 1465-74. Flow of gases through plastic membranes. Indus, as temperature rises, and Eng. Chem. 45: 1148-52. 9. A device was constructed, based 7. Doty, P. M., W. H. Aiken and Her on an Asbeck-Van Loo cell for meas urement of critical pigment volume, to mann Mark. 1944. Water vapor per meability of organic films. Indus, and Eng. Chem., Anal. Ed. 16: 686-90. study the ability of paint films to 8. Dunn, E. J., Jr. 1954. Moisture re "breathe" in the sense of permitting sistance of paint films. Official Digest, air to pass through under a pressure difference of 0.38 atmosphere. Films Federation of Paint and Varnish Pro duction Clubs 26 ( 353): 387-407. 9. Dunn. E. J., Jr. 1956. Physical prop of linseed oil paints within the range erties of latex films. Official digest. of pigment volume practicable for Federation of Paint and Varnish Pro house paints are incapable of such breathing. Films that can be pene trated by air under a pressure'gradient duction Clubs 28 (378): 564-89. 10. Eckhaus, Sigmond, Irvin Wolock and B. L. Harris. 1953- Porosity of paint films. Water vapor adsorption and per- 9 GLD38049 meabiliry. Indus, and Eng. Chem. 45: 426-8. 11. Fabre, G. 1955. The permeability of plastics. Industrie Plastiques Modernes (Paris) 7 (4): 33-5, (5): 41-3, (6): 35-8. 12. Gardner. H. A. 1934. Primers for wood. Indus, and Eng. Chem. 26: 1272-3. 13- Harris, B. L. and A. Bialeclce. 1952. Permeability to water of thin unsup ported films of pure drying oils. Offi cial Digest, Federation of Paint and Varnish Production Clubs No. 335: 884-7. 14. Herzog, P. C. and A. E. Gilchrist. 1955. Emulsion vapor barrier coating com positions. U. S. Patent 2,709,689. 15. Hess, Manfred. 1951. Paint film de fects, their causes and cure, p. 105. Reinhold (New York). 16. James, D. M. 1956. Experiments on paint adhesion under moist conditions. Jour, of the Oil and Colour Chemists' Assoc. 39 : 39-62. 17. Kittelberger, W. W. and A. C. Elm. 1946-47. Water immersion testing of metal protective paints. Indus, and Eng. Chem. 38: 695-9, 39 : 876-81. 18. Kuzmak, J. M. and P. J. Sereda. 1955. The blistering of paint in the presence of water. Canadian Jour, of Technology 33: 67-76. 19. MacGregor, J. R. 1952. The unex plored field of exterior paint. Official Digest, Federation of Paint and Var nish Production Clubs No. 335: 869- 80. 20. Mayne, J. E. O. 1950. The blistering of paint films. Jour, of the Oil and Colour Chemists' Assoc. 33: 312-16, 538-47. 21. Morgan, P. W. 1953. Structure and moisture permeability of film-forming polymers. Indus, and Eng. Chem. 45: 2296-2306. 22. New York Paint and Varnish Produc tion Club. 1937. Scientific Section Cir cular 546, Amer. Paint and Varnish Manufacturers' Assoc. 23. Nicholson, A. and H. A. H. Jenkins. 1955. The blistering of paint films on steel. Paint Manufacturer 23:303. 24. Othmer, D. F. and G. J. Frohlich. 1955. Correlating permeability con stants of gases through plastic mem branes. Indus, and Eng. Chem. 47: 1034-40. 25. Payne, H. F. 1944. Adsorption, per meability. water resistance, and struc ture of organic surface coatings. Chap ter 7 in vol. 4 of Protective and De corative Coatings by J. J. Mattiello, Wiley (New York). 26. Pragst, W. 1956. Report of some ex periments on the question of the humidity blocking action of coatings. Farbe und lack 62: 145-51. 27. Rains, H, G. and G. L. E. Wild. 1955. Passage of water through a paint film. Jour, of the Oil and Colour Chemists' Assoc 38: 800-2. 28. Taylor, J. , 1955. Treated rutile-tjr]* titanium dioxide pigments and their properties. Jour, of the Oil and Coloui Chemists' Assoc. 38: 233-49. 29. Thomas, A. M. and W. 1. Gent. 1945. Permeation and sorption of water va por in varnish films. Proc. Physical Soc. (London) 57: 324. 30. Turner, J. H. W. and S. Hodgson. 1936. Composition-property relation ships in high-polymer emulsions. Jour, of the Oil and Colour Chemists' Assoc. 39: 114-28. 31. Vannoy, W. G. 1952. Blister-resistant house paints. Amer. Paint Jour. 36 (32): 82, 84-7. 32. Waack, Richard, N. H. Alex, H. L. Frisch, Vivian Stannett and Michael Szwarc. 1955. Permeability of polymer films to gases and vapors. Indus, and Eng. Chem. 47: 2524-27. 33. Weinmarra, Kurt. 1955. Water vapor permeability of varnish and synthetic films and! its relation to swelling and stability in dampness. Farbe und Lack 61: 315-23. 34. Wirth, W. R., II. 1956. Study of the permeability and blistering of five ex terior house paints. Amer. Paint Jour. 40 (35): 72-110. 33. Woodbridge, R. J. 1955. Evaluation of - emulsion paints. Jour, of the Oil and Colour Chemists' Assoc. 38 : 285-99. GLD38050