,1 '3} Wu :1; I 11%3'1 ; 1 M ‘1: 1, 4:11P}? ii 1 ‘4 ‘51:“), . 1-1,," '41 151,?4'11' I1‘ .1'“ 11 "1‘1 _ 41g, 11 x1 ,1. 1 ' 1. “@2111“ 1* ,1 1 1 1 11 1113111 [1 ‘ | l 14 141131415 ,5 1 g}, 1%?23’4 1 "u . I 1: 11,13 ’5 : l ' 115% . ' #3323113?” W 1191??" m} P .11. 1 :1. 111111.‘ 321%}; a 7 1"} fs‘ . A??? v . “ L 4' 1412.111 .4 . . , A h .4 I. 1 . ' K . m a t ‘ ‘3 ’ ‘, ., I a 381‘! 1'7 ‘ 7 1 1 . - ‘ , I 1}” @111, 51,41, (111411;;3‘ it” Nils” 7' -_.‘1.‘....,, 4. ’1 1.11%.1- 11119,: I, fin}: J‘EYI ‘ I, 121 11 22114;; 11111- :fi'gzfiy 1 11:1 ‘ r1 {mess I" r I 1 3{.:1 1.111111. '1 bit}' 1 ‘w‘lfl 111‘: L». 1:15:44! .3391 *7: Ax "“1131: ’/ THE 013303102: 0F 3mm , moon , NITROGEN , mm m macaw ELmTaom'Ic NICKEL 33' Russell H. Fay 1 THaSIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requiramenta for the degree of DOCTOR OF PHILOSQPHI Department of Chamistry 19511 Q1198] ACKHWS mac author wishes to apnea his “new thanks to Dr. Might T. Ewing undsr who” muatun, constant mum, tad unfailing interest mesa investigation m mom-taken. Be 18 also dnaply indebted to Mr. Bun-rd 6. Case for many helpful suggestion! as well as to It. Otto Ear-dos. The assistance or Dr. Rustin J. Smith 13 muadly twining the than is many appreciated. A grant from the Hanson-Van Hinkle-meing Company under finch than studies Hare made is also gainfully acknow- has“. mam “ TABLE OF CONTENTS INTROflUCTIGH........1............................................ HISTORICAL BACKGROUED.............1..... ........ ................. EXPERIHERTAL.............................1....................... RESUEffl...........................................1.............o DISCUsslfifl....................................................... 00HC$USIGES...................................................... BITEENTHRE1CITED................1................................ Page 1 2 7 15' ho L1? h? MODUCXI‘IOH mama» or pores in electrodepotited metals have been had responsible for the failure of these metals to protect adequately the hue mm which they cover. In the present work this porosity m “nailed by measuring the rate of diffusion of gases flrough thin IlOOWO nickel 10:11: at omrpreasures of one ahaosphera or less. Tho «fact of different gases, air, hydrogen, helium, and nitrogen was studied and their rate of diffusion max-related with cram: Law. Corrosion of electrolytic foils had been previously reported to gently increase the pemeahility of these foils to gases after a brie! initial period in which the pmegbility was affected very little. This effect of corrosion: was evaluated in two corrosive media. The diffusion of ions in litiuid media through electrolytic nickel mill !” invastigated to attempt to determine the size of the pans which were present in the electrodepcait. HISTOEICAL BACKWND Diffusion of Gases Through Hotels The first recorded observation or the diffusion of gases through mm- on that of Csillotot (1) in 1863. a. town that nascent. man:- can, producsd by immersing an iron vessel in dilute sulfuric acid would, in put, pass through the iron vessel and appear on the inside of the vessel. If the vessel was made cathodic in an electrolytic cell the rate of diffusion was increased and the diffusion would proceed em with a pressure inside the vessel of twenty ataosphcree. However, if the vessel use surrounded by molecular hydrogen there was no notice- able diffusion at. room twparature, it 350°C. diffusion could be ”noted and the rate of diffusion increased rapidly with further turn-store increases. Subcequont imeetigations (2,3) have show that. the rats of diffusion of molecular game through metals is proportional to the square root of the overpressure at. high and intermediate over- prams. Kowsvsr those extrapolated isotherms would not pass through tho origin, but rather, intersect. the abscissa at a value which is loss, the meter tho temperature. Careful measurements indicate that. at low mrprosoures the rate of diffusion departs from its dependency on the sum root or the overpressure and the values do pass through the origin. this indicates that there is no "threshold overpressure" Mary to initiate the diffusion process. Tho autos-m mobilities, which various metal: exhibit when oxpoood so gases, In: oooribod by Fast (h) to tho property at tho gas to for. 1 ow am $13 metal, the diffusion rate of the gas being high 12 oompouui formation were possiblo. Rhino: (S) postulated that a ”I mid am through a metal in than instants” in which 11'. was ”1&1. in tho mm. The behavior of the halogen gases which react radii: with lost metals , but do not. diffuse through any metal to any put extent, disprove Post's statements. The rm gases cannot, at course, for: compounds with any metal. It ha: boon reported (6) that the flu of diffusion of helium through copper, nickel, or molybdenum, 11' it proceeds at :11, mt be at least. 10‘ times slower than the rate of diffusion of common gases. Characteristics of Eleomdeposited Metals It was recognized as early as 1887 , that. one metal deposited won o noon! natal, did not cover the second metal oontimomly until it had rooohod o conlidarabla thicknooo. Oberbeok (7) investigated the thick- nou o! _a metal necessary before a platinum aloctrode would assume the olootmtiva potontial of the deposited metal. This thickness was computed from the might of the deposit and found to be two to three :11me for nine, one to two mfllimiorono for oadnixm, and one lfllhioron for copper; home it appears that court]. atomic layers aunt be built up before the ban metal in completely coma. mandamus! metallic coatings can be dividod into too typos «pacing on thcir reactive position in tho elects-motive mica, in relation to tho base metal vhich they on protecting. It the coating it loco m1. than the base metal, the protection in of c oacrificial “my tho has until in protected because the coating is protonticlly mud, hence the hm metal in pmtcctod doopitc minor discontinuities in the coating. 3mm, in tho case 01' coatings more noble than the base metal, it sum or. am diocontimitioo in tho coating, the proacmc of the «am my actually be detrimontal, since this will root-riot the modio may to tho» crass where discontinuities cadet and thus greatly in:— oum the corrosion of the bozo metal at those locations. The two dic- linilu- uctalo tom a galvarac cell. The potential existing between than totals io dependant only upon the two metals and tho elcotrolyte with which they are in contact. This potential is, or course, indo- pcndont of the relative areas of the two metals. Any current flow, as I unit of thia potential, mt obey Chan's Law. If the modic- reaction in the diuolution of the base metal, with large onodic areas the loss of total night not be objectiontl. In the once of a restricted anode urea, in which me some mount of current must flow and hence the one mount ot‘notol removed but from a smaller ma, 3 rapid failure of the but lots). at this particular- point may occur. The“ cloctrochmiocl mum m be altorod if polarization occurs. Such diacontinuitieo have been held accountable for the failure of olooirodcpooitod nickel coatings. It is characteristic or such coatings that tiny fail in looalizad areas, the base metal than oorroding rapidly tro- thia position while the larger portion of the electro- dapoaitad niokal coating 1- not visibly affected. It has not boon proud that Inch localized failure is related to discontinuities in tho original dopolit but it is not unreasonable to oipoot that they Ira. Attupts to link the two have been carriod on for a considerable lust): of time. Tho early attempts were mainly by the use of some Mull reagent , considered nomorrosivo to the metallic coating, but IMO): gm a characteristic color reaction with the base metal. the): Wool havo boon severely criticized (6} since it cannot bo proved that all pom may be than identified or that tho reagent is non- corroaivo to the metallic ooating. Amt-bur suggested method is that of photographing nonodhomt dopedtlw) . Obviously this method is limited by the fact that tortuous pore! may not allow light to pass, and by the size of the pore which on: be dotominod. Thou and Solomon (10) have reported an apparatus for the measure:- Iaat of the porosity of nonadhoront deposits by means or gas perme- ability. I. ainilor apparatus was used in this work. A omidoration of the manner in which an electrodepoait is formed Ink", it appaar altogothor reasonable that these deposits ahauld be more pol-m than aimilar shoots obtained from the some metal cast and than rolled to similar dimensions. It is characteristic of an olectrodoposit that it does not form. unifomly over the base metal but appears to start at diureto points and than grows laterally from those positions. The won for thio is not clear; it may be due to tho presence of ”activ- csnters“ in the. base metal , to the presence or points or pro- Jsotim (perhaps on a mioroscoyio ooalo) in the base metal, or to the mo of otmio planes having a preferred orientation such that it 1| Ellis! for the electrons to move to the surface of the base metal It M point. Sime the formation of the deposit is in a direction toil-rd the anode it would be expected that dioc-ontimities would de- velop men these laterally growing grains meet. mootrodoposita are toned at tuporaturos considerably below the freezing point of the "til am the nobility of the metallic atom at the moment the solid in mm “It be considerably less than when the solid metal is formed {an s nolt; hence it is loss likoly that the olectrodoposited metal will be tamed with the atoms in thoir most stable configuration. flio oooopoaition of morogon has been considered as a factor in tin {motion of pores . Wood (11) found from x-roy diffraction patterns that electrodepositod nickel has; the some lattice constants: as metallurgi- cal (1.0., out and worked) nickel. He observed that the lines were W for electrod-opoaited nic kol arxi attributed this to the concen- trail-on of hydrogen at the interfaces between crystallitcs . Ho colon.- um those omstollitoa to be 10"" to .10" cm in size . Other investi- altars (12,13,110 attributed this diffusenooo of the lines to the crystal -3 .o 8180. Grystal sizes ranging from 10 to 10 om were reported. mama W3: nick-l tom m producod frm 3 Hats typo nickel solution contacted at 1 pH of 2.2. Four liters of the solution we propane: having the following comentratiom nickel mm also grun/litor flicks]. chimide L5 gram/liter Bea-1c acid 30 yams/liter Tho Willa m dissolved in distilled water. Sui'ficiant nickel Wm momma solutiontoraiaathcpflto aboutS.2, mm milliliters of 30% Wagon peroxide per liter of solution was added and the solution allowd to stand overnight. The aolution was rmmd Waugh #2 Whom tiltar paper and the pH lowered to 2.2 by the addition of concentrated aulhn-io acid. Seven and ans-half was 93 Haunted charcoal (mm) per liter of solution was mu, the solution kept omight with agitation and again rum through #2 Man tutu- papa. Emcee wdrogen peroxide was destroyed by Mating th- Ialuticn. Tho aolution was cloctrolyzed ten ampere—hours per lite to ram metallic inpu-itica. A corrugated cathode with an average current density of 0.5 more: pec- some decimater was used. The solu- tion was stored in sealed bottles holding two liters each. Foils were formed alcctrolytically on panels of rolled steel, the M83. surfac- mc being about one-tenth of a square root. The first lilo panel uud measured two by three and one-half inch“. Strip mm tron than pimple gum three tolls, each large mung}: for the to» porn-ability quatu. The toil: were taken, one above the «m, In. the untu- ef the strip deposit. The panel. :12. m lab- umfly We! to We and noun-sixteenth: by three ond om-oifleenth mm. with this also pea-1, 33.x toils, cash large enough for the m pmwnlty apparatus, oonld be obtain-rd. An entire “up do- mo of w; size was and by the second permeability Ippn‘Btul. lath sides of the panel to be used too- producing strip «posit: m sound with a depoait of electrolytic nickal from the one solo- 15103 Iran which strip deposits were subsequently to be produced. In In. instances the when vaa buffed, in other: it was left as deposited. A for foils mo electrotomad directly on the surface of stainless steel (13-8) pm. without. an intmodiata deposit or ohcttolytlc nickel. Panels which had been buffed were wipoé off with carbon tetra- enema. to remove any adhering buffing mound. rho pamla were 30w electrolytically using an alkaline clamor and given an laid up. The alkaline clamor was prepared. daily with the following eon- mouom ' . Sodium hydrande 21 gram/liter Sodizm metanilicate 15' grams/liter Trilodim phosphate 16 guns/liter Sodium sax-boasts 6 gram/liter Tm liters of thia cleaning solution were tamed. The panel to be cleaned m undo noodle , the cathodo employed being of rolled steel designed to Juli. fit within the two liter bower holding the cleaning solution. The clawing solution was operated at 90 to 95°C. thoalold, also preparod daily; oonoistod or 20 per cent (voluno) .hrflfllihlorlo laid and awn used at room tonperatnre, The timing of the W cycle was varied go produce nonadhoront deposits. I: typical anal-wuofi16.hot 20 aooonds in tho alkaline oleanars the pano1.boing alfldio flith n ourrent denaifiy of he amperon par square foot, rinse in running'untor, 29 seconds aoid dip, rinse in running water, 20 seconds in the alkaline ole one? (medic , he more: per square foot.) , rinse in running utter, S aooonds acid flip, rinse in distillod water-and dopoai- tiln immediately atarted. For thinner foils and foils from unbuffed sunrises it'wao nooesoany to increase tho time or the electrolytic cleaning and/or the current density (80 to 100 emporoa per square root) and to reduce the time of the acifi dit. Tho doposition of the nickel foil took place in a one liter rootlagular glass Jar. Deposition was carried out at a pa of 2.2 i 0.05, tapas-stare 55 : 2°C. The pg was one-okod with a Beckman model a In later which was tested daily against a stamiard buffer. jet. no time was 1% found necessary to raise the #3; when neoessnny to lower the pH additions or concentrated acid wore made near the anode. For all solu— tion: except “D“ oonoontratod hydrochlorio acid was used. For solution ”B” a mixture of com entrater': hydrochloric maid and concentrated sulfuric acid Itih‘tho same ratio of chloride to sulfate ions as existed in the oriafinn1~tolution was used. The panel was bold in place by an alligator type 011p attachod to a_ptooo of nickel wire. The panel was placed against the back or the 3!:- le that. the deposition of mo Ecol took place prodondor antly on one 10 aifie of the panel. The salution mm agitated with a glass stirrer having twa blades , each 1 canthneter by 0 .7375 centimeter, set £56 to the horizontal (909 with respect to mach other) . The: stirrer was drivan at 525 2 25 revolutions per minute by a; variable speed stirring motor. The atimr was set near the level of tha betim of the:- pawl and nearer the cathode than the made; Butairilas which fameé on the panel but we net moved by the stirrer war as dialed g'mi by jarring the panel . I A graph 9f the efficianay of the plating solution was prepared fixing the time necessary to depoait a given finial/mass at nickel. E21331 sufficient "time had elapsed tr.) give the amine“. thickness of dapaait the panel was removed 1‘er :33 solution mm rimad in distilled water . The deposit- mus then cut. with a rmar blade about aaquarter of as: inch tram the edge 01" the ;-_::-arra‘ia If the) s-:.u'fa£:e oi“ the 033$ panel was preperly passivated the swig) fiepoeit. conic: be remand vi in as difficulty . For deposits of about five micrens or less the Strip deposit- was removed mad-er distilled water to reducza the li‘ételilwssc‘; of teaming. The strip daposit was then dried on 3 $1934.31; 01“ absaréaent papar, marks; in om: comer with its idantifflng umber {pencil} , placed in an camelope , and stored in a deaiccator war calcium chloride .v The thickness of the “rip deposit. was determined. win» a micrmeter calipm‘ equipped with a ball attachment. and re wing tea the naamai: tan-athausandth of an. inch. Five Gr 51:; measurements; were; taken for each atriga 69330311; and the re~ lulu Wieated that the geometry 0:? the plating 3911 gave quite unifm M“. For this reason the thinner fails (sight microns or less) were named by mighing the foil amt coazputing their tlficknesa. Each foil was identified by an individual group of characters The first character (a letter) indicated the liter of solution, as dram train the stock solution, frag which the strip deposit.- was produced , 0.3., £1, B , C, D. be sewn-s. 01113131813111.1131 (a number} i: 1dica‘1.aci the particular panel used as a base: plate. Tia-11 latter ”3" preceding the number indicated that 131113 panel “1:103? SbflL’lléfis 81113131. The; bifird charmtm’ (a letter) indicate-.3 1.11.1.1 strip depoait from a. particular panel in the sequence in which 111.13; were 53111311112611, 51.3 . , , B, C , ate . Far the second pamaability 31535313319111.1113 aa- azuir; 31.511531311111051 . 1.1 13 11513.. at once. With the firai pama 21.33.13. 3.15.1 apparatus six beat 3313113 1101.13.11 be out from one strip (1131.10.33.13. ‘311339318113 5311111111321 11112111131111. 1.35-- tha aids of the deposit from which 13.111351 war: 111—553,, 1.13., with 151113 3:11:11 of $116 deposit. which was 333111011311 firm the 131.1119. 53113312 away from the 013331-3733 , the left. side aft 1112 anrip deposit was daaignataci a1: 1 aha right. 3111 as 2. 11 final letter indicated the var-11.11.31.111 position, 1311:: foil taken nearest. the bottam of 11111.1 a trip :1. 531.3131. L 1313:1115; , the one {1313.111 the center B, mad the one ran: the 4.1.353 1’3. The atrip deposits. were 1311111311135 11.18.111.113- aging; is ‘131131‘15111513-3 war binocular microscope ana photograpmfi. +30 13133113 21:15 553933 imparf 1113 "Lions . Th5 £0113 were photography; 5311711111113 1113. printing; frame. .1: 535.11.112.13 132’? glass between the strip 1-.1asit. 13:111. .1111: 53111130111135»... iilm sawed 13:3 diffuse any light which passed Lira-ough the atria 111353133511. , thus. making any @038 pares more easily identified. fluminua foil 1488 11541311 to mask the rmoinder of the phategraphic film . Beth Kodak Spectm analysis Ho. 1 plates exposed 15 minutes and Kodak ortho contrast. film exposed l2 1139‘m1nntas, twelve inches from a 130 watt filamenu lamp gave good results. The permeability apparatus consisted of a vacuum tight systam which could be divided inta twa portians by an alactrolytic nickel foil in a special holderg If9 than, an overpressure was . plied to one side of the foil, thg rate at which the gas passed through the fail could be measured an the opposita side 5 the fail by'aIHcLeofl gauge. A numbar 9f foil hnldera were eonstruc ted befora one was abtatned which was satisfactory, The first heléer (Figmr92 29f) was nouu+W°QJLd ofi’mild stealo fl cfircular foil five~eighta cf an inch in 4%; netgr was held betwaen the awn piecas of the holder in such a'mannar tham an expoaed area of fihreeaeightfi of an ineh in diamatar separated fihe tWD partians of the Systemd Thfi fiwn Mi- 2:25 of the holdar were joined by a threadad collar? .Sach pie’e tha holder was joinej to ‘she we””~- abilitgr apparnt~s by‘meang of a ground glass jointo Thfi graunfl glass joint flag as 8133 t9 the holder with plicené cement. '?he ground glass jnint connbuaefl to this aystam'wafl attached to glass arma with doubla benda to give sufficient flexibility to the qystam an that the holder could be inserted anfi ramoved (Figura 1). The foils placed in this holder soon develapad cancentric cracks under the bearing surfaca of the holder“ Polishing the ends of the holder did not remove the difficulty. A variahy of gasketing:materials were tried including copparg lead, and nylon, but the_s&me trouble persisted The holder had been conatructefi with twa graphited ringa to yravent twisting of the two parts when the holder was tightened; hawevm’, it o 11-. O u ‘0 U K) . r” ~ a a — — —.. __a. FIGURE I. FIRST PERMEABILITY APPARATUS I-' [/(Illlllllllll/I7/11»1\\\\\\\\\\\\\\\\\\\\\\\\>\_S WWW/Iy/«W Warm F IGURE 2-A FIGURE 2-8 13 was felt that there might still be enough twisting action between the two parts- to produce the concentric cracks. A second holder (Figure 2-3) was eonstmatad with a keyway so that all twisting of the We pieces could be eliminated. ifter a few measurements on a foil the concentric cracks still developed. .61 third holder (Figure 3%) was constructed of glasso The foil was sealed to the and of a glass adapter. Silicone high vacuum stopccck game was used between the foil and the glass adaptero s4: Incite ring placed over the foil, was sealedg under presmzra , to the glass adapter. A number of cements were tried; the one finally selected. was glgptal, which did not require heating, that might change the characteristics of the foil, and dic'i not. meet the foil , thua it would mat creep into the capillaries of the foil and seal them. This holder was later discarded to avoid thsa use of stopccck grease on the foils. The first holder was then redesigmd, enlargening the cavity into which the foil was placed. The foils were sealed batween tam lucite rings which had been ground flat. The rings were wales} under pressure using glyptel cement, This method warhead satisfactorily am. another similar helder (Figure 3~B) was constructed of alvminum. Instead. of sealing with a twisting action the two pieces of this 3:191er were jaineci by six cap screws . The second lacuna—ability apparatue (Figure u) emplOyed. ball and musket joints rather than the flafible glam arms for inserting and removing the foil holder 0 Hamsters were added for measuring the over- pressure used ,, FIGURE 3-A. \\ ‘ \\\\\ Illlllll IIIHH FIGURE 3-8 AAAAAAAAAAAAAAAAAAAAAAAAAAA 11: the foil holder m of a typo described by Dr. Then of Princeton Invent: in W work (Figure 5) . Rubber «on rings scratched th! 2911 taut, clininlting my winning at the foil. The vacuum tight and. in: applied by the raised collar of the aluminum holder. No W «I encountered with comencric cracks under the bearing surfaces, probably duo to the alumina being a much cotter metal. Huh the fix-ct tppcrctua, measurements or the pressure increase van lads every five minutes for fifteen minutes . After observing an initial lag in the rate of diffusion , the frequency of readings and the inter-"1 of the no increased. 1cm the accord apparatus was cued, finding- are taken every minute for tmntybfivc minutes . The results were plotted on a ycph (increase in pressure veraca time) and the proc- we increase from the five minute reading to the twenty minute reading and to calculate the permeability constant "k”. Thin constant was calculated tron the formula developed by Then and Kelccan (15): k -V/Fxl/APxp/t where in the volme of the low pressure side of the ”not in liters 3.3 the area of the exposed foil in can“ is the overpressure across the foil in m of Hg is take increase in pressure (mm of Hg) on the low prcuurc side of the foil in time interval t for which a 15 minute interval was used "'6; D'fl fl '9 the units of thin “k” m liters-cm'a-min-l. V FIGURE 5. FIGURE 6 APPARATUS PERMEABILITY FIRST FIGURE 7 PERMEABHJTY APPARATUS SECOND 15 RIMS The first electrodepaaita obtained from a nickel aolntion ware row to have a much higher permeability than these which were produced Manama); during an; cantinwms perior} cf electrolysis . It was foam: that was to four hours of continuous eleomlysis at 14...} aparaa 9e:- mare decimatar cf cathode 82:93:: was neces sat-3r before the solu‘uicn would produce damaita with a permeability constant as"? about 1.0"6 311131: was considered to be about the limit of the sensitivity of cm first. appuatus; walkers "34's" mm in sea-sue instmcea detemine but. since they represented a pressure ci'xange af the order of one Enema of mercury in fifteen mimtes their acclzrémy was questionable . 1.1 typical :3th»: is illustratad in Figure 8 . The sate affact was mated subsecmemly when the; saluticn was again used to produce foils. However, each time the solutisn was used to prodmse fails the initial foil was (31‘ a lower permeability than the initial foil electrofonnec! the pmvious time the solution wan used , and a shorter partied 0f electrolysis was: required m rec-111cc the permeability. fitter nearly a lmnc‘tred hours of electrolysis at Eu} wiper-es pear square deemeter (cathmza) uniformly 3006 (k a 10") £011: could be pmduced frm the first . Figure 9 (Table I) illustrates the effect. of two types of corrosive media on the permeability of electrodepoeited foils. 1211 of these 2011: were 25 I 1 microns. in thickness and were cut. from the ewe strip no“ I PMABLITYOONSTANTX I l0 I l n\ln ‘ 3 I00 200 TIME IN MINUTES FIGURE 8. EFFECT OF ELECTROLYSIS ON THE PERMEABILITY TIBLE I ‘— A__-.._. A“ 16 EFFECT OF CQRfiOSIVE £5331}. 01$ -IIWEHI 80%.: ER? 1139 11.5 N HUI Vapog L L 10%A38394 1 x _ (Abra) -B_7 . (2-1:)fl 11-?) (2-5) (1-9) a 1.0mm“ 1.0x10'5 5 x10"? 2.02m“8 7 30.0"" h 1.0x10’6 5 310'? 8 5 x10“? 9 x10” 10 h.7x10‘9 2.6x10“° 12 2 .1x10"° 3 .axlo“ “ 13 ‘6 1 21:41:)" Lama‘s 2,030.0“ 114 h .2310" your) _3 a w 18 1 .oxlofld 1 .leo m 3 .0313“ 20 2.61m) 1.8110.” 7.02am ° 22 5.8110“ .2 “12:10“ 34 “34310.5 mm mm x no' —c l l 1 IO 20 HOURS. IN coaaosws MEDIUM" c- F’IGURE 9.,EFFECT OF CORROSIVE MEDIUM ON PERMEABILITY 17 deposit (6-31-44), pmdnced on 3. buffed stainless steel (18-8) panel The main media wars-=2 kept; in a canstant twparaturs bath (25.0 1’ 05°C.). To manta corrosion products, which might clog paras, the foil! were rinseai with 4.431.443? 411:2?1 22-2341 been adj-1131.311 tn 3 p11 31’ 3.1) by the wmsn of 1miroz'21 11342113 4.2:. - . mean. 154; the 43349:" of 11.5 21 hydrochlarie acid was fauzr‘. 1'33 .242 about. 12414343 33 eamoaire 9.3 imam-31041 in ten per ca 1'7. {40111443} 3 “a. .11; 33:16 . ‘33.. sharp-.4 3.1" 1.113 3111‘4'443 4433 abstractsriatic as? the eff-4:41.41 4.4: .. scram-33143 and... an 1.119 3.231344411211115“. The initial ems-am: pry-1121122311 4.41:; 1:11. 1.1.1. 1.4.4153 3.3123 314.414 1.... 4.131114% 2422442 2.4.2.4 :. 423412.131. 321...: waf- mp3 2.3: :43- 3456 311 F13 lifts? 41 24224 114mm @444». the pmeability. Aft-2'1 ' "'1. 13.: 13:331...- 133:1}. started it naraaaaa 1" 1.1313113! to a point Wham the. gem-3332111124 242-23 1.322..- we—ai. to be 13341314. $3 1113 @para'ms . Vixen this-z.- gmrtios. 11:1” ‘112-3 441112-44.- , 441...? 1-. .1124 raw-.4: 3......1 "144-... Tammi ixwreeanng, 441;; 2421.13.45.44)? 34.43 :3 it. the poi. 31'. Wan-31123 permeability was 113"“3 , the corn. 312011;.- 3.11:4 222211243 :13; 1.5.4; corrosive-3 323111.43 143$ refs red ’69 as the "brew-31091.: times” fur the: 12:11. If the 4231*; 1024 results are rejected 31-:- due to 1": £13.21... in; 1.1223 £311,124; relatimwhip :13 nearly linear m tbs breakdomz time for 1.3 f 1.1.11.3 :13 "'OL’LLJ again... 1111““ tit-s, thicimss. These results (Figure- 18 , 151.21.23.3- II} ware. 012.143.3434}. wing 3.1.5.7 11' hydra» chlaric 3.1.521 <1 as 1.1 £2 narroaire media: izfter 34421219541211; a mercury avapcr ciifi‘usion pump , in addition to the rotary oil pump , it was 41421.44: that the pressura increase during the first five minutes uf iron was 1433.... ”- than that during any" succaading five situate interval. It. 114.. -<.1 been the practice to evacuate both sides of the system to about 0.0305" m of mercury or less, than to adrait an TABLE II RM‘I‘ICNQIP 2’)? THIC- K!" "8 TO THE} BREEKDG' 3‘1 TIME 73.?" mmmnlmr 310m FOILS IN 11.5 H HG]. VEPOR $111014!!!” wire akdown This: knees firfleakdown‘ SnimmlA ‘1‘ ins { hrs) (microns) Time (hrs) 10 3 ’J' 23 i . J 10 3 .5 23 7 .0 18 5.5 2} 7 .0 18 3 .S 4.3.1 11 f 1 10 .5 25 8 .u 20 9 .0 25 3 .0 20 9 .b 25' ll .11 2 S 12 O III IIII IIII [III III .0 ° / BREAKDONNTIENHOLRS O O x, k h) 1111 11111111 1111 llll O 5 I0 I5 20 THICKNESS IN MICRONS FIGURE l0. RELATIONsr-nr BETWEEN BREAKDOWN TIME AND mu THICkN'ESo‘ 19 am of overpressure to one side and measure the subsequent 1mm in pressure on the apposite side of the system. To determine if this initial lag was caused by the adaorption 91' the gas on the MD. at the foil, an overpressure of one atmosphere was left on one "IMO of the toil while the opposite side was avaeuatad to as 191: a value as possible: (usually a few microns of mercury) , the vacmm pumps were than that off and the increase in pressm'e noted. Ths increase in 9mm m, in this case ,. limar with reapect to time ,I the slape of the curve being abaut the same as that for the previous mm (in which both sides of the system were emanated) , after the initial lag Wing the first few minutes. Typical results are illustrated in Figure 11 and Table: III, IV am V. A change in the permeability of the foils over a period of time is observed; for £011 C-Sl-D-léB tha permeability 1mm :3st , for £011 C-SlloD-l-A the permeability decreased. For foil C-Sl-Dol-Bg Runs 1', II, III and IV were made the name day, Ran V was made faur days later, Run VI twemynsix days after the first. four runs , and Run VII twenty- eight. days after the first four runs. Runs I and II for foil scam-1.5 were made the same clay. Runs III and IV mm made eighteen days later, and Ram V and VI were made thirty days after the first. two runs. Both m for £011 G-Sl-D-sZ-C ware made am same flay. £11 £01.13 were frm the seas strip deposit, 25 micrcns 1n fixicémees. 1:3hen originally pro- dneud, about six months previous in: this series of tests, all foils had an: initial permeability of about. two microns of mercury in fifteen amines. All permeability taste were made with one atmosphere TLBLE III 20 IIITIEL LfiG'IH THE DIFFUSIGN OF AIR THROUGE FOIL C-SI-DaleB W ‘¥:~ ~~ a 1 2 3 h S .003 .015 .016 £32; .001: 6 .005» 7 .022 .00? 10 .00? .031 .030 .1313 .011 12 3i; .025 .023 .OM .025 .021 20 .036 .335 22 Run I Both sides of systam evacuated Run II Evacuated Ofily one sifie 0f system Run III Evacuatad only oae side of system Run IV Both sides of systm evacuated Hun V Evacuated only one side 9f system Run. VI Both sides of 33mm evacuated Run VII Evacuated only ans aide cf system PRESSLRE AT OUTGOING SIDE OF FOIL IN MM. OF HG .030 20 TIME IN MINUTES FIGURE II. INITIAL LAG IN SETTING UP STEADY STATE OF FLOW THROUGH FOL C-Sl~D-I-B EVACUATED BOTH SIDES I. 2 ONE SIDE 3. I I. II 4. " BOTH SIDES 5 " ONE SIDE 6 " BOTH SIDES ONE: SIDE 21 TABLE IV mun. L3G 13 THE DIFE'JSISH OF AIR TEEEOUGE FOIL 0-51-13-1-1; Yin. Pressure at. Out . . Side Foil 1mm of H 7 ‘ . . , . . “a“ 1 ‘31 0 0.000 0.011 O .002 0 .005 O .091 0 .601 1 .002 .001: 2 .005 .022 .007 .015 .602 .003 3 .010 II .015 5 .020 .050 .018 .026 .0014 .007 7 .031 .016 .025 .031: .006: .010 10 .0141; .072 338 .ohé .0 .013 12 .060 .050 .Ohé 3% .016 15 .078 .100 .051; .066 .018 .020 17 .088 .1113 .062 .072 $320 .023 20 .108 .130 .071; .023 .026 22 .120 .136 .0533. .026 .036 Run I Both sides of system evacuated Run II awacuated only one side of 838m Run III Bath sides of syStam evacuated Rm IV Evacuated only one side of system Run V Both aide-a of syntmu Iavasuated Ema VI Evacuated only one side of system; 22 TEBLE V mm. MG BI THE DIFFUSIDfi OF AIR TEROUG! FOIL C-Sl-D-Q-C :33:isqu E? Pressure at Out in Side of Foil in m at H .532 5 ELEM 55 0,009 0.001; .001 .012 .ooh .023 .067 .029 .013 .OIIO .018 .ohé .025 .031 006‘) .0140 .072 .Ohé .080 Run I Both sides of system evacuated Run II Evacuated 0 3; one side of system 23 amt“. (TI-I1 i 5 m of Hg) . There was no correlation between the mum: in overpressure and the variation in permeability. Foil O-fl-B-l-II had bean Stored exposed to laboratory atnoaphare but pran- usual 11‘:- duat particles by cheesecloth. Foils CuSl-D-luB and 0-31-9443 had been stored in a dessicator ova:- calcium chloricie. um“ were made. to determine if lens in solution as Well as gas“ would pass tm‘aug‘n the foils. Using; a foil tc separate: sclutions man contaimd, respectively, chloride ions on one side and. silver ions on the athsr; ahrmate ions on one side and lea/.1 ions on the other; in neither instance was there am indication of any precipitation on cam side of the foil after twenty-ofour haura . It was possible, tmaver, to measure the diffusion of hyflrogen ions threugh a £011 25 microns thick urxler similar eonfiitiom. The foil was usual to “para-he a solution of ten per cent (volume) sulfuric acid. fram dintmed water. The wl‘ma of :iiatillacl water was approximately twenty man, the volume of ten per cent sulfuric acid was Shows 2.75 milliliters for trials 1 and 2, mi 2.0 milliliters far trials 3 and 1;. In trill 3 a little (‘2 .1 25 sodium hydroxide was adders to the distilled Inter in an attempt to get a longer patio-:3 of diffusion before: an equi- 11m was reached. The pre‘dously datemineé ”k“ for this foil was a x 10". The manna are illustrated in Figure 12 and Table VI. The rmainder of the data was obtaineci using the second pameability was” . The rate of diffusion at. different overpressures for foil 0-1-3 (13 morons thick, deposited on a buffed electrelytic nicks]. mm- fm) is shown in. Figure 1.3 and. Table VII. There was a linear increase 2h mam v: cum or p8 DUE TO THE nmsrozz or macaw ma meow mmmnlc mom. Trial 1 Trial 2 Trial 3 Trial h 7.9? 6.148 9.88 8.60 10,03 6.31; 10.26 3.?2 7.68 7.61 7.50 6.32 10.35 ?.32 7.33 _ ? .25 0.3%; 1-0 .15 7 .10 7.08 10.35 6.93 6.70 10.57 6.85 all; 6.86 6,30 6.86 6.2.3 6.19 10 50 6.06 6.80 6.79 «5.80 @322 5.85 5070 5.60 5.75 5.75 PH OF WATER SOLUTION I-I J: l’ IO f_TRlAL- 7; 9 8 7 ‘0 O TRIAL 4 lbU’D‘ a 6 >\ TRIAL 2 \ l TRIAL I I) 40 80 l20 TIME IN MINUTES FIGURE l2. DIFFUSION OF HYDROGEN IONS THROUGH ELECTROLYTIC NICKEL ’3. BLT: VIE EFFECT 0? avg: aasl‘ag on TEE R£T£ GF DIFFUJIQK a? g ; THROUGH FOIL 0-1.? Lat? 91" Diff‘usiqg_ ___ l. E' .., .__ .. n Hg mm EEK; min ccfemzfl5 min I»: x 1'3"" I; ,0 O .4335 ‘3 .001 l“ 23 .021 .003h 7.5 37 .03h .0056 7.5 105 .098 .016 .5 300 .323 .0531 “.9 7&1 c.3aw 1.0L {1 293 3. J” .239 3.1 L W .371 .0938 5+3 l 51 .258 .ahzb L2 13 .131 .0215 60 " 9.9 .123 .020); 110 14.0 .07!» .olzi. 163 3 ,0 “OILIJ, .00?2 2'?) "* Extrapolated re 51.41"" mg I INCREASE IN CC./CM.’/I5MIN. PRESSURE I .000 0.750r 0.500 200 400 600 800 'OVERPRESSURE IN MM. OF Ho FIGURE I3. EFFECT OF OVERPRESSIRE ON TI-E RATE OF DFFUSION OF AR THROUGH FOl. C-I-F /-__ .. 26 in thc rate of diffusion as the oval-pressure vs: increased up to 300 n of mm. When the overpressure was further increased to one chocphm (m m at mercury) the rats of diffuicn increased sharply. Th- rm a: diffusion increased enact meaty times while the over- mm was increased by a teeter of about two and ans-hall. When the at: at diffusion was again tested at successively lower overpresmea the use of diffusion was consistently greater than it. had been previous- ly for a similar overpressure. When the foil was rwoved from the lulder and examined visually a small crack was agparen’s where the holder Ind gripped the foil. Since it was apparent that. the permeability characteristics of the £911 could be changed by high warpreamea it was decided w work with overpressures of the order of 109 m of mercury or less. Foil (3-92-5- (13 microns thick, deposited on a buffed electrolytic mama muse) had a nearly- linear increase in the rats of diffusion la the overpreccure was max-camel frm one to one hundred m of mercury as shown by Figure 11; and. Table VIII. flowever, in View of subsequent results, it should be acted that time was only one reading taken between twenty-three and sixtyetive mm of mercury overpressure and that this raiding m below the straight lice. The rate: at diffusion of hydrogen through foil 3-02-42 (8 microns thick, deposited on a. buffed electrolytic nickel surface) is illustrated in Figurea 15 and 16 (Table Ix). Preview to this dry air had been used as the diffusing gas. it very low overprcssrums the rate of dif- fusion increased very little as the overpressure was increased, at 2? i TABLE VIII max 0? ovmammg on: THE ma. 0? szam: OF AIR THRG-JGI FOIL 0-2-3 Ovarpmom , , Rate of Diffusion my fiL_ ' "am 5 min w oofmflfi’“; 3:- min k x 19“ A 3.0 0.061 00010 170 .0 .119 .0193). 250 9.9 0258 .0h23 220 314.0 .720 .115 180 98.6 2.66 .1536 229 i. 65.1; 1.75:3 .290 22;;- 22 . 580 ,0953 210 7.8 .20? .oaho 22-3 3.0 .071 .011 200 2.22 .th .0079 13:.) 1.10 .028 who 1&2 .88 .018 . {1.730 170 I'mssane "GREASE N comm/15m. I l if 0 2 5 50 75 OVERPRESSUE IN MM. OF H6 FIGLRE I4. EFFECT OF OVERPRESSLRE ON THE RATE OF DIFFUSION OF AIR THROUGH FOIL 0'2-8 28 TABLE Ix mm a? 07me ON THE RATE or DIFFUSIOII or 3113me 1mm FOIL n—oz-c carpi-«sure Rate at wmm a 33 no figfi; m oo7ou'753 I11: I: x 106 3 .S O .0022 0 .00036 S .2 6.8 .0071 .0011 8.7 3.3.5 .0100 .0017 6.1 38.1 .0150 .0025 3.3 58.1 .0270 .0016 3.9 6L8 . O .0072 5.6 BL? 0570 .0091: S .7 76.7 .mIBO .0079 5.2 69.9 90 .0032 2.3 51.2 .0330 .0051: 5.1» 36.? . 80 .oohb 6.3 22.1 .0150 .0025 5.6 32.? .01h0 .0023 3.6 .011 .0010 .0001? 760 .033 .0012 .0001? 300 .12 0010 .0001? 66 1 SB .0031; .00055 18 . 5 0020 .00032 26 . mos .0001) 90 . 0003 .0000; 300 .023 0005 .00008 180 1.95 .0016 .00026 6.1; , .0016 .ooozé 10 1 .313 .0010 .0001? 6 .2 RATE OF DIFFUSION IN chcMFIIs MIN. o A 0.00m / /A 00050 0 A 0.002 <———c o 25 3‘0 75 OVERPRES'SURE IN MM. OF H6 FIGURE I5. EFFECT OF OVERFRESSURE ON THE RATE OF DIFFUSION OF H2 THROUGH FOIL o-oa—c 0 FIRST RUN ASECOND RUN QOTHER READINGS I? RAIE OF DIFFUSION IN oc/cN'nsNN o 0.000I OVERPRESSIRE IN MM. (f H6 FIGURE I6. EFFECT a: OVERPRESSURE (III THE 'RATE a: DIFFUSION OF H2 TI-ROUGH FOIL 0-02-0 WW-fimthroetoonelnmkedmofmrythermo! Winn increased more rapid]; and linearly. The amino intimate the first readings, taken in the order or increasing ourpreuoreo, the triangles represent readings taken in the order of decreasing values of the ourmeeure, hexagon: are other readings taken. Readings which in 101: are webably due to contamination or the hydrogen by air. The reading. below two we of mercury overpressure :an be divided into two ”to, one set indicating a. rate of diffueien about twice that indicated by the etha- set. The data. from foil 9-02-15 (15 microns thick, deposited on buffed electrolytic nickel) had the least variance from a smooth curve or any foil toned. There was a sharp break at 36 an of mercury ovorpmme. When the rate of diffusion of hydrogen is plotted against. the ammo- are only tuna points fall off a smooth curve. The point at 65 m of weary overpressure could be interpreted as another break (Figure 17, mm 1) . I: the rate of diffusion is plotted against the square root or the overpressure two straight lines are obtained with a break at. 36 In of mercury overpressure. The first gas to be used with foil D-OB-F‘. (8 microns thick, do- MM on an unburied electrolytic nickel surface) as hydrogen. The first two series of readings were made starting at low overpressure: and increasing the overpressure for mooeaeive readings as indicated by. the circles and squares. The third series of reading: was made starting with the highest overpressure and decreasing the overpressure for Imo- omive reading-e as indicated by the triangles. Rangoon are , again, TAKE I mm (3' WWW 0! THE RATE G DIFWSIOI 3 3mm WWW FOIL 13-02-31? “ V w Rate of Diffunion e u I; II 3353 min m7n'7i") min 14 x 1'3 0.0!: 0.0092 0.000041; 1w 3.3 0310 .00016 2 h .1 .0016 .00026 1.9 .7 .0023 .0038 1.1 30.5 0030 3001;? .82 .7 .0059 .0039? .93 ‘ 65 .5 .0070 .0011 .91 1 85.5 .3090 .0015 .87 ‘1 109.9 .0110 .0016 .8} “ 3, .0080 .0013 90 ‘ .2 .DOM .000?2 .69 26.6 .‘30h2 .00069 1.3 38 .9 .0038 . . o .0032 .00053 .7h 1.38 .0005 .omos 3 1. .0001; .ozmé 2 .150 .0003 .0000; 2 RATE OF DIFFUSION IN cc./CM?/I5 MIN. /0 0.0055 /0/ A O 0.000 d / A f OW‘” (/IO/ 0 25 50 75 I00 WERPRESSJRE IN MM. 0F HG FIGURE I7. EFFECT OF OVERPRESStRE ON THE RATE OF DIFFUSION OF H 2 OFIRST RUN Asecouo nun THROUGH FOIL 0-02-E U OTHER READNGS 31 mt row. The data (Table XI) as plotud in Figure 18 ohm a bran): st 16 us of mercury wax-prams. Figure 19 (Tables III, m and I!!!) empire: the results a! diffusing hydrogen, helim, and nitrogen my: m1 MB-A. fislim dinghy: the am pronounced bruk as hydrogen; nitrogen, whm plotted to the cane scale as the other we gm. docs not. have nearly so promuncad a break and might be inter- proud a a linear function. Foil. 3503-8 (8 microns thick, deposited on an erd electro- lytic niokal surface) was first. tested with nitrogen. The first ”we of readings gave a typical curve with a break at. about. 53 m of mm mm. After the last (highest) reading was made a atopocck was inadvertently turned in the wrong direction and an overpream of 250 u at weary was mentarily placad on the toil. Eben subn'quont read- 111g; were made, the rate of diffusion had increased and the mend curve ma obtained. M. this time, due to a leak in the connection to the rotary oil pump, uhieh was not. discovered until later, it was necessary ta use the mercury vapor pump to obtain panama-a5 low enough to make permeability maswenta . In this instance the mercury vapor pump was operated eontimoualy for five hours after which the rate a: dif- Minn was found to have increasezi again and the third curve was obtained. The” remit- :re shown in Figure 20 and Table m. The foil as also tested using Ivdrogan and helim, bath gases Whiting a break be‘tween 1:0 mad 50 In of mercury overpressure (Figure 21, Tables XIV, IV and 17!). will: mined visually there was no appmnt «image to the fail Tl-SLE XI mm W OVERPRESSIRE ON THE RATE 0? DIFFUSION OF HYDROGEN TBROUCH FOIL MB-I '1; n A‘ M Onrpmmo Me of Diffusion n R; m FEZE min cahm‘fig min I: x 10' 1.02 0.0056 0.00088 h6 1.30 .0055 .00088 35 2.15 .0063 .0011 2h .90 .0739 .0121 130 .00 .0670 .0110 93 10.00 .0590 .00968 148 11.70 .132 .0217 62 2 .1 .190 .0312 60 C .308 .0506 h? 79.30 . .0820 52 o .620 .102 66.00 .1335 .0713 58 62.10 A60 .0755 61 83.50 .601; .0991 60 3.15 .0213 00398 63 10.30 .0870 .0113 72 22.20 . 69 .02 63 32 70 .220 .0362 56 £3.30 .285 .0h68 55 50.90 .327 .0536 63 58.50 .1420 . 60 70.30 .507 .0832 60 86.50 .590 .0965 .0010 .0001? 85 £2.30 .212 .0398 ha 62.80 .261 .0628 51 71.00 .375 .0615 M 97.00 . .101: Sh 81.10 .0896 55' 11.10 .héls .0761 Sh 61.80 .h28 .0702 57 52 .70 .382 .0626 60 0.110 .269 3th 55 33.2w .265 0105 66 an 70 .160 .0263 51: .50 .125 .0295 63 6.70 .0590 .00968 73 RATE or DIFFUSION IN oa/cu'nsm o 2/ a/ o / 0.0% /A O ,0 0‘20 O 25 30 5 OVERPRESSURE IN MM. OF He FIGURE I8. EFFECT OF OVERPRESSURE 0N TI-E RATE OF DIFFUSION 0F H2 THROUGH FO!L 0-03-A OFIRST RUN DSECOND RUN Ammo RUN O OTHER READINGS 33 $631.15: XII mm: OF OVEPRESSJRE 031 THE RATE OF DIFFUSIOB 0.? Km mm P011: 9-034 Onrpmnum Rate of Intrusion an H; m 06 an n k x 10‘ 0.060 0.0009 9.0002 100 1.13 .0071: .0012 5h 3. .0321 .0053 76 7.50 0571; .0091: 6h . .0900 .0111? 63 23.7 .131 .0215 £6 36 .2 .205 .033? 14? 5 .236 .0387 ha 52.9 .281; .01465 US .3 .th 0565 U6 75.1 .100 .0651 92 h .188 0720 39 3h TABLE 1111 rpm CI OVERPRESSILIL; 0): TE: MTL 0F DIFFUSER (F HITRDEN mm FOIL 13-034. Owrpmasuro Rate af 131115151011 6 m Hg Hg/lg min colm'fifi min k x 10 3.20 0.0086 0.00111 22 10.1 .0291; .001482 211 32.14 .0550 .00903 9 3h.3 .0660 .0132 19 51.3 .118 .019 19 62.14 .11“? .0230 19 72.6 .160 .0263 18 92.0 .216 .0351; L9 RATE OF_ DIFFUSION N swam/mun o: 0/ . 0/. .— 3/° /Z / °" / 7°} nm o J./ N. 0 2‘5 3‘0 OVERPRESSURE IN MM. 0F He FIGURE l9 EFFECT OF OVERPRESSURE ON TEE fiATE OF ' DIFFUSION 0F GASES THROJGH FOIL 0-03-A 35 new m mm: or WWW on rm: am; 3 amsron or unnom 1mm F011 9—03—13 Gurpnum Rate or Diffusion In fig m 3 so an is at. 10° 3.20 0.0970 0.0511 18 10.1: .0169 .00277 13 21.2 .0162 .0075? 18 31.2 .0620 .0102 16 Ms .0800 .6132 15 53.1 .0950 .0156 15 62.0 .121 .0198 16 70.5 .lhé .0210 17 81.0 .171 .0281 15 91.6 169 .0311 17 56.14. . 2 .0233 2 1.111 .0035 .0095? 15 6.140 .0171. .00286 23 16.5 331430 .0070? 22 93.7 .131 .0215 21 76.3 .175 .0287 19 Ml; 09? .0159 20 33.9 .139 .0228 3a 3.0 .0166 .00256 143 25.7 .102 .0168 33 36.3 .1 .0252 35 112.2 .175 .0292 35 111.1 .172 .0282 32 116.7 .2 .0328 36 51.3 .216 .0351: 3; 51.6 .212 mm 55.7 .217 .0356 32 56.9 .225 .0363 33 60.2 .2131 .0391: 33 62.0 .261: mm 33 65 .1 .252 .01113 32 72.5 .285 .0668 33 88.5 .355 .0583 33 RATE OF’OFFUSION N OCJOAFIISMN. 1/ m/flf 5 50 75 OVERPRESSURE IN MM. OF H6 0 FIGURE 20 EFFECT OF OVERPRESSURE ON TI-E RATE OF DIFFUSION OF N2 THROUGH FOIL 0-03-3 oFIRST RUN ASEOOND RUN a THRD Rm 0 OTHER READNGS TAELE I? W38? (F WE‘RPBESSURS on THE HTS; OF DIFE'USIOH 0? m THRWG'I FOIL D—O3-B Overpruaure Rate of Diffusion n: Hg "'m? E753 mm'" ' ' 6676971? m" ' k x 10' 1.9? 0.0160 0.0026 68 9.00 0960 .01 89 27.0 .281 .0669 86 1.3.7 .1.le .0696 81 In.) .hho .0721 77 50.9 .3478 .0785 78 511.2 .500 .0 77 61.8 ."76 7 71.9 .655 108 76 . 766 125 75 37 TRBLE m EFFECT 01‘ vaPRESERIs OR THE EMS 0F DIFFUSION 01" 311330633 MODE FOIL 13-63-13 Overpream Rate 91‘ Dixfusion mm Hg m‘ cc m k x 10' 3.06 O .200 0.0328 1.70 .1146 .0239 120 270 91.1.3 110 .315 .0513 100 .395 .GélIB 99 ~ .0755 98 .530 .0869 98 3‘55 .0913 100 . .103 100 .693 .111. 100 .830 .136 100 .916) .1534 100 1.12 .181. 110 \o on i 0/ § // /A “E g A/ z 0 c» A// / E f“ as N2 g o. - P] ‘ / o / O / / O 25 50 75 OVERPESSURE IN MM. OF He FIGIRE 2|, EFFECT OF OVERPRESSWE (N THE RATE OF DIFFUSION OF GASES THROUG-i FOIL D-O3-B to account for the increase in the rate of diffusion found. when the foil was being tested with nitrogen. The last. mph (Figure 22)- and Table XVII illustrates how closely the behaviar of fitrogen and welt-033:1 carnapoxfied to Graham 1.31: when diffusing through the electrolytic mic 1ch £01.13 over a range at over- pressures. La comparad ’00 1161qu, nitrogen diffused faster than would be expected while the hydrogen dii‘fnsaé shiver than would be expected. The dotteti lines indie ate the theoretical rate (as compared to heli‘tm) that. the gases should diffuse to car-respomi ta firahm Law. 39 TABLE XVII CWTIOE OF THE REES W DIE’imS-IGH TO W53. Liz?! Phil Ovarprasaure Kitro ‘n Helium H e n V mm Hg Rate of' Ratio to Rate of ate of Katie to Diffusion Helium Diffusion Biffuaion Helium D S .0 C: .0026 G .339 O .0078 O .9078 l .00 ’ 10.0 .00h3 .360 .0118 .0129 1.09 0 20.0 .0080 .1400 .0200 .ozho 1.20 3 33.9 .0117 .h30 .0275 .03h7 1.26 ‘ hfi.0 .0152 .hhfl .03b6 .OhbS 1.29 E 50.0 .0191 .hSh .0h20 .0555 1.32 60.0 .0229 .hSh .0505 .0660 1.31 70.0 .0268 .hSS .0590 .0770 1.31 80.0 .0305 .hSS .0670 .0590 1.33 90.0 .03h2 .hSG .0760 .0990 1.30 D 5.0 .00385 .h33 .0089 ' 10.0 .007h0 .h33 .0171 .0237 1.39 0 20 .0 .0116 .hOO .0350 .Oth l .26 3 323.0 .0206 .101; .oh75 .0595 1.25 ' h0.0 .0279 .h33 .0625 .0765 1.22 B 59.0 .0335 ‘ 35 .0770 .0990 1.29 60.0 .0390 .h29 .0910 .121 1.33 70.9 .Ohéfl .h3fi .107 .lh3 1.3? 80 .0 .0520 .53? .119 .167 1 .140 90.0 .0590 .hhh .133 .190 1.h3 Brahma Law .371; 1 .hl LOO IN CC./GM.2/l5MIN., RATIO TO He 0 Z 9 (D 3 IL I: 0 I n. O E 0.50 g Mf_8_3__u;fi 0’ N2. 0 25 50 75 'OVERPRESSLRE IN MM. OF HG FIGURE 22 CORRELATION TO GRAHAM‘S LAW 0 FOIL 0-03—A III FOIL D-O3-B L50 DISCUSSIOR m abaermim that an initial period of electrolysis is moon? nry before a nickel solution will promos olootrodopooit- ot opium: quality has been was before (12) . In this study it. was round mt after the optima condition had been reached, this character of the solution could be lost on standing and enticement electrolysis was menu? to again bring the nickel solution to a condition whore it would produce optima: quality {loos permeable) deposits. The effect of corrosion on the pomability of electrolytic nickel foil: was characteristic of that m1: had been previously reported (15). Both typos of corrodzmt had the sme ultimate affect; a rapid inns-ease in the permeability o: the £011 before am visible sign a: failure (8.8., visible holes) appear“. Host of the owerhontation on the corrodibflity of nickel £01.15 at performed using 11.5 N deroohlorio acid vapor as tho oorrosivo Indium. The low results are probably due to defects in the £0113. The effect or increasing the thiclo’xoso of the toll is newly limar. The effect of doubling the tlfickneoa of the foil 15 to slightly note than double the breakdown time. This is in accord with other reports (16) which indicate that. the rate of corrosion of nickel (both oleotro- We and metallunga) in linoar with respect to time in 11.5 E malaria sold vapor. m initial 133 in setting up a steady state of flow of gases emu pom- material has been noted in other atudiee (17) . This was believed to be due either to the penetration of the gas into the material , whioh included some blind pores , or adsorption of the gee on the surface of the foil. It was observed repeatedly that evacuating both sides of the system with a mercury vapor pump (to lose than {3.5 aim of mercury) then edeitting en amoephere of overpreeeure to one side , would produce an initial lag in the rate of diffusion. If, an overpressure of one atmosphere wee left on one side, while the opposite side was evacuated, it was noted (with one exception) , that the rate of diffueion wee linear, as soon as evacuation was stopped. The initial lag amounted to about. 0.02 t .01 me of mercury. Since the foil area for these runs was 0.375 em” and the volume of the apparatus into which the gee wee- diffusing, 1.115 liters , this momted to 1.9 x 10"4 liter-emoepheree/m’ of foil area. The ‘e'atte type solution 13 lmown (154) to yield deposits such that the 100 plane in parallel to the substrate metal. The closest interatomic distance in this plane is 2.348 x 10" cm, hence 3 one centimeter row would con-- tan 1; x 107 nickel atoms. A square centimeter would not contain over 1.6 x 10“ atoms. Therefore, even if the gaseous molecules (oxygen or nitrogen) were adsorbed in a one to one ratio the geometric surface of the foil could not account for more than one-thousandth of the observed 1m. Measurements of the true curfece area of nickel foils by the gas adsorption method (18) indictte that the true surface area of electro- lytic nickel fella is at least twice the geometric area. The surface ha up w, in not, be much higher, time on tons in this study (18) pm housed at 150°C. tor thirty to forty hm, shaming may have noon-rod, selling or: small capillaries, When the ooooni permeability apparatus was 00th the toil holder was made much lager so that this phomonon could be studied more closely. ‘1‘}er was no instanoo , while using the second pome- obmty appuatus, in which there was an initial lag in the rate of Gamma. The phenomenon may have been due to the surface on whale}: the toil had been deposited... fan initial 183 was observed only with tolls deposited on a buffed stainless steel (18-8) panel. The first. livers of an electrodoposit are known to be mat]; innuenood by tho mots-ate upon which they are deposited (19,20) . The only ion which was fomd to diffuse through the electrolytic 111.6131 101.1 in liquid media was hydrogen. It will be observed that tho dittusion of Wagon ion: reached on oqnilibrim value utter a abort period of diffusion. An attempt to extend this period of mm- lion by aiding 0.1 fl oodim hydroxide to the solution into which the ova-03m ions were diffusing m macaw. The oouilibfim value (fl) morons-ad for successive runs. Tho failure to detect the diffusion of ion- other than hydro-con may how been due to the precipitation of film chloride or load chroma“ within the pore: bloom fur-tho:- mmuon. It ha been reported (15) that the "to of diffusion of gun throogh electrolytic nickel varies directly with tho amt-saw. Raul“ obtained with tho first ”mobility opparatuc indicated that 343 this was true m amoophoric pressure (7130 m of warm) to about 50 u of moroury overpressure, Below thin out-pronoun tho rate of diffusion appeared to decrease more slowly‘thon the overpressure de- creased , lmovor , those road lugs at. low overpressure: were noar the limit of accuracy of the first permeability apparatus. The second ooparaxuo was designed to increase the accuracy of these readings at lower overpressure“. The one of the foil, exposed for pomoation by gases, woo increased about sixty times. The lower limit of the readings possible with the McLeod gauge was decreased ten times by adding on additional scale to the capillary. Eruploying the second penne ability apparatus it was again observed that the rate of diffusion varies; linearly with the ovarproo sure only of. overprooouros of about. fifty mm of mercury or over. High overpres- sure: (one atmosphoro) were found to damage the foil. This was probably due to the increased: foil also since no trouble was encountered with the first apparatus . If the data is plotted employing log—log coordi- nates that. portion of the curve above fifty m of mercury had a slope of unity (it tho overpressure were doubled, the rate of diffusion was doublod) which is indicative of molecular flow; Since molecular flow oocurs only when the mean free path of the gas molecule is large com- pared to the diameter of the capillary this would infiioato that the size of the pore: was of the order of magnitude of about 10.: cm. or about 0.1 microns . At higher ovorpmsouros the gases diffuse through carillon“ by whom flow. In this type of flow the rate of diffusion is a linear formation of the overpressure squared. hh Ldzuni (21) found that the rate of flow of gases through porous diaphrama could be interpreted by the formula: K *Iiwar ‘(BF where K is the rate of IIOW’Of gas in mm-cc/eeconfi £ is 5.230 x 103(1 4) £\\ is the viscosity of the 383) E is equal to (r4 ) where r is the radius of the capillary, 1 the length in cm. P is the mean pressure across the iiaphram K is the coefficient og/slip assumed to be 0.9 B is 3.0h8 x 104(T/H)- 3 where T is the absolute temperature, H the molecular might of the gee F is equal to (ra/l) Any data on the diffusion of gases which give a linear function when the rate of diffusion is plotted against the overpressure could be similarly interpreted. If the pores are assume! to be of uniform radius and normal to the aurface'the ratio of 3/? will give the pore radius. By this method szumi oztcrminol that the pore Size in a series of earthenware plates were of the order of magnitude of 10-0 cm. Lpplying this equation to the dots of diffueion of gases through electrolytic nickel gives less plaueiolc results. Foil D-CZ~L is illustrative of thin fact. he gee usoc was hydrogen, the temperature about 25°C. The ordinate intercept in equal to BF, if the best straight line is drawn through the points the ordinate inter~ cept is about 0.3001 03/15 minutes or 6.L x 10.5 cc~mm/eecon£3 hercc F is eQual to 3.5 x 10“. it an overpressure of 100 mm of mercury (mean pressure so mm of mercury) the rate of diffusion is 0.0017 cc/lr; minutes from which AB? is equal to It x 1.0"4 cc-mm/eecond, The viscosity of lwdrOgen at 25°C. is 8.92 x 13-5 hence the value of i-.- is 24.6 x 10-“. The calculated radius of the pores is the ratio E/F or the order of one £5 Moron, obviouslyuuch too highsime upon or such sizewouldbe readily visible than the tell we examined under a We». Celene latlona Ira the Me of other foils gave similar results. In the study or mama films of genes Hu'kinl (22) observed that these films undergo pm changes. If the pressure of the gas is plotted against the values or gee adsorbed a cecond order piano change will produce c cherectcrietio ”kink“ cleaner to the break noted in the plot a: rate of diffusion versus overpressure. If each a mechanic: were responsible for the dittoeion of game: through the electrolytic nickel foils it would explain why the calculated also or the pores is much too large . It would not be necessary for the gas molecule to strike a pore openlng in order to cum-e through the toil. It the molecule struck the toll and was adsorbed, it could then diffuse over the surface of. the foil, through a pore, and evaporate from the other side of the foil which is under negligible pressure . The ans-two diffu- lien or nimgen and hydrogen was reported (21,22,23,2h) but helm was reported as not undergoing aux-face diffusion. The surface was found to be a very critical rector in this diffusion. The materials reported included glue spheres , elminmn oxide, calcium carbonate, and silica. Throughout these studies it was apparent that the permeability of the toils was affected by some tectorh) of much greater influence than the toil thickness. Often foil: 25 microns thick and having no visible or photoyephio porosity had much greater pemeabmtice than similer foil. 8 microns thick. The permeability was independent of the direction hé in mien the gas passed through the toil. Three or the toil: produced hm solution l'D" , 11-024, D-O3-B, and 9-03-42, were of similar thick- ness (8 microns); foil J's-024:; was about twice as thick (15 31cm) . m on from an Wind surface (D-OH, D-OB-B) m a much water Minty than the other two which were from a buffed mince. The correlation of the permeability of D-OB-A and D-OB-B was very good until D-OB-B use dmegec by high overpreaenre. Moore and Smith (25) round that rolled. nickel foils could be cathodically impregnated with a poster volume of hydrogen if the our-- face of the foil consisted of e buffed layer of metal than if this layer was etched away before the foil was charged with hydrogen. They postu- lated that the buffed layer of disturbed metal noted an semipermeable mmbrano , per-hope by allowing the hydrogen ions to diffuse two-ugh, but preventing hydmgen atone fran difficeing out. 131ch deposited on a mechanically buffed nickel surface has a random orientation until considerable thickness of metal (up to 0.1 Morons) has been formed. Such a layer could have a mentor effect on the permeability than the total thickets-:3 of the deposit. This is prob- ably the reason that depoeito from unbui‘fed surfaces gave more reproduc- ible results and a more permeable deposit than deposits from an unburied surface . is? f 603101351033 rho maxim of gm: through olootmlytic nickel differ: from the diffusion of 343ch through 39wa (1.0., out and who) nickel in at last three respects: 1. Biffnsion of gases through electrolytic nichl is initiated at much lam temporctw and aromas-mu. 2. The effect of WNW on the rate of diffusion of sales through electrolytic nickel in more nearly a. linear fumtion of the overpressure, while with metal- lurgical nickel the rate of diffusion is more nearly a linear function or the scum root of the overpressure. 3. Rare gases (a.g., helium} diffuse my: mama-ma minke]. very much clover than ordinary gases or not at all; hydrogen, helim, and nitrogen all diffuse “rough electrolytic nickel at relative rates comparable to thaw predicted by Graham: Low. The diffusion of gases through metallurgical nickel is recognized as 'latticc diffusion" «Motivated difhaoion“ . The diffusion of gases Wong; cloctrolytic nickcl corresponds mom nearly to molecular flow through capillaries . However, this corrospondance is not exact. Particularly at low overpmsurm there is considerable departure from Granu- Lav. Calculations Iran the rate of flow indicate paras of a motor cine than could be present. Thu rate of corrosion of nickal 1:8 mwm uidvaporinknmtobonnmuthnspao’stomo. Th: rat. or molecular 21m! harm as the who at tin radius of the capillaries. let if. m observed that the initill exposure to warn- m um vapor had littla affect on the rate at diffusion. It appears that than initial pomaabflity 18 due ta 3 cmbinauan of mloaullr flow through capillaries and surface flow. The mun corrosion of the tons due: not attack as max-an]. when a th- ponu but rather starts at an outer surface and corrodes Waugh the fail. This explains the appmzdmatoly linear relationahip of breakdown the to thickness of the foil and the sharp increase in pomeabiflty at m 1911 at breakdam tine. At breakdown time the £011 has ecu-040d W to that the internal Macon of the pores may be attacked. The sham: increase in permeability in due to th. rats of flow than max-using as theme of the capillary rm and tin foot that” pores are continually corroding through. £19 LHMTURE CITED (1) amount, 1.. r., com. "136., 5:}, am (1863); Snithelh, c. 3., am. and am, John (any and Sons, Inc” sow rm, 7? (1937). (2) Slithclll, c. J., ease. and fistula, John may and Sun, 1:10., How rm, 95 (3537). (3) sum-or, a. 3., nut-um: ‘In and Through scum, Cubridgo University Press, Londcn, 169 (191d). (is) Flat, J. 1)., 911m” Tech. Ken, 6, 365 (191a) %, 71. (193.2); supp, c. 12., Progress in final mica, g, 105 1 53). (5) mums, r. 11., Trans. Am. Soc. Metals, 93;, 17:. (1951). (6) munfm,)c. 4., leey, c. E., Proc. Roy. Soc. (London), 3.52, 112 1935 . (7) Goerbeck, 1-," Ta‘ied. 1mm, 2;, 337 (1887); Than, 3., Addison, E. '1'. Jr., Ian. flectroplators Soc. Research Report, 2, 11 (19M). (8) Thon,,fl., Addison, E. '1‘. Jr., Am. Electroplatora Soc. Rosetta man, 2, 15 (19147). (9) Ogburn, F., Honda-1y, r... Plating, 9;, 61 (195k). (10) Then, 3., Xelenen, D., Proc. Am. Eleotmplaux'n $00., 35, 105, (191(8). (11) Wood, (3. 11., Phil. Hag., (7), 1, 9634 (1935). (12) Brenner, A., zentnor,‘ 1., Jennings, c. 14., Plating, 2, 865 (1952). (13) Clark, (3. L., summon, S. H., J. Elmtroehm. Soc. fl, 110 (1951). (11;) Dame, 3., Loidheiaer, H. Jr., J. Elactrochen. 806., g, ((90 (1953). (15) Thou, 21., Islam, 1)., Plating, 6, 362 (19M). (16) Thou, 1a., 1mg, L., Kelmn, 1)., Plating, 11, 7h? (1950). (17) Ban-er, R. 14., J. Pm. Chem, 21, 3S (1953). (18) Then, 3., Yang, L., bag, 5., mung, 59,, 1135 (1953). (19) Flash, 6. 1., Sun, 6. 3., Trans. Faraday 306., 3, 852 (1936) (20) Finch, G. 1. 1m, 21., Yang, L., Discuss-1m, Faraday Soc” 3;, 11)., (16)) ). (21) mm, 3., Bull. Chem. Sam, Japan. 33. Jan (1937). (22) aux-um, w. 1)., “musical shown-y of Surface Fm," Ramon Pub. 00., Rev Ian-k, 1952, p. 2117. (23) Barman, I’. 6., Pm. Roy. Soc. (harden), m, 55 (1950). (2h) Carmen P. 0., Ealhar‘be P. 19R, Pro. Roy. Soc. (Landon) 20 1&5, (1950) . ’ ’ “'2’ G. 1%., Smith, D. 21., Trans. Electrocheu. $05., Q. 5115, (1937) . (25) Home MAY 31'55 - “I 2% 1")?-