Effect of Inhibitors on the Corrosion of Zinc in Dry-Cell Electrolytes

Home , Dry cell

U . S. Department of Commerce Research Paper RP1863 National Bureau of Standards Volume 40, February 1948 Part of the Journal of Research of the National Bureau of Standards

By Clarence K. Morehouse t 1 Walter J. Harner t and George W . Vinal

This paper deals with a study of substitutes for mercury and chromate films in curtailing t he corrosion of t he zi nc anode of Leclanche dry cells at high temperatures. Organic com pounds containing t he carbonyl group (s uch as furfural), heterocyclic nitrogen-containing compounds like quin aldine, a nd certain commercial products were found to be effective in retarding the cOlTosion of zinc in dry-cell electrolytes. Al t hough these types of mat.erials retarded t he corrosion of zinc, dry cells made with

" them did not have the expected increase in shelf-life or electrical output. The inhibitors r' eit her reacted with t he paste wall to decrease its strength, or formed an insoluble film over the ent ire surface of t,he zine anode that co nt ributed to its internal resistance. No paral lelism was observed between t he electrical output of the cells and t he percentage of inhibitor used in their construction. On the other hand, t he paste wall of the dry ce ll, which is usuall y a starch-flou r gel, was found to have inhibiting properties. Starch was inert, but flour h9d inhibiting proper t i.es. The active inhibiting agent of t he flour is gluten. The constit uents of glu ten, namely, glutenin, and gliadin Bnd mesonin in different states of aggregation were isolated, and the 19.st two were found to be effective in retarding t he cor rosion of zinc. These materials can be incorporated in dry cells and inerease t heir capacity at moderate temperatures.

I. Introduction able electrical energy, and (2) the corrosion reac tion (local action), which is a side reaction, and ? When the Leelanche type of dry cell is subjected from which no electric energy is available. These to high-temperature storage, one of the sources of two reactions may be formulated a follows: failure is the increased rate of corrosion of the zinc 1. The dry-cell reaction [2 , 3] in which no H 2 anode [lJ.2 The zinc anode also corrodes at normal gas is formed: temperatures (18 ° to 30° C), and it is common practice either to amalgamate the zinc or to use Zn+ 2 YInOz= ZnO.Mnz0 3' elU'omate films to curtail local action. At high 2. Corrosion reaction in which H2 gas is formed: temperature neither of these methods is so effec tive as to yield good shelf life. In this paper Reaction at anodic areas: Zn-'Zn+++2€, alternate methods are described, including the use R eaction at cathodic areas: of inhibitors and modifications of the paste wall to increase its inhibiting properties. The theory of corrosion inhibitors is still in the Over-all corrosion reaction: formative stage, and their choice is largely a matter of trial and crror. However, the principles postu laLed for the corrosion of steel may be considered. The corrosion reaction may also be considered In the dry cell the reactions are of two types: (1) more simply as the discharge of hydrogen ions that the cell discharge reaction, which gives the avail- penetrate to the cathodic areas of the zinc can. I Now with Oli n Industries, Western Cartridge Co., East Alton, III . The corrosion and the cell-discharge reactions are 2 Figures in brackets indicate the literature references at the end of this paper. not the same, and it should be possible to retard the

Corrosion of Zinc in Dry-Cell Electrolytes lSI corrosion reaction without materially affecting the ionic volume, very mobile, I1nd highly penetrating. discharge reaction. In order to accomplish this, Dry cells have been made wi th electrolytes com the cathodic areas of the zinc must be made more posed of salts of anions other than chloride but anodic, or the anodic areas must be rendered inef with little success. Chlorides are necessary from fective until current is drawn from the cell. the standpoint of electrical output, because they Amalgamation of the zinc and chromate films preven t the anode from becoming passive when function in this sense. the cell is discharged. Therefore, studies were In the steel industry, organic substances are confined to concentrated aqueous solutions of added to the acid pickling bath to inhibit the cor ammonium chloride with or witbout additions of rosion of the metallic iron. The theory of these zinc chloride. The corrosion rate of the zinc inhibitors is that they form an ionizable salt with anode may be altered somewhat, but not signif the acid, whereby positive inhibitor ions are formed icantly, by changing the concentration of the chlo that are attracted by electrical forces to the ride or by using zinc containing various percent cathodic areas. They are not discharged but ages of other metals. In this work, commercial replace the iron ions in a H elmholtz double layer, zinc containing 0.25 to 0.60 percent of cadmium, and when a state of equilibrium is attained they 0.20 to 0.60 percent of lead, and less than 0.03 form a protective layer over the metallic surface. percent of iron was used. The effectiveness of this layer depends on its ability to prevent hydrogen ions from penetrating II. Experimental Procedure to the cathodic areas where they are discharged. The rate of corrosion of the zinc was deter The size, nature, position on the metallic surface, mined by a method suggested and used by D. N . ' and the closeness of packing of the inhibitor ions Craig, in which the amount of hydrogen that is will determine the effectiveness of this protective evolved under controlled conditions is measured. layer. The apparatus consisted of an inverted test tube \ These types;:>f inhibi tors are classified as cath closed by a rubber stopper through which passed odic inhibitors. Attention was given to this type, a capillary tube (fig. 1). Within the test tube the because in the dry cell it is desirable that the cor sample of zinc and electrolyte were placed, the rosion of the zinc be inhibited by stopping the electrolyte completely filling the test tube and corresponding cathodic reaction and not the capillary tube. In the inverted position, the gas anodic one, which is necessary for the production liberated at the zinc rises in the tube and expels of electrical energy by the cell. An anodic inhib an equal volume of electrolyte through the capil- ( itor, unless used in adjusted amounts, would not lary tube. The tube was weighed from day to I be desirable in that it would not only stop the day, and the volume 01 gas evolved was calculated cell-discharge reaction but might lead to localized from the change in weight and the known density corrosion and cause perforation of the zinc can. of the solutions. Measuremerits were made at Chromates act as anodic inhibitors, and experience 54° C, a temperature reached in storage in the I ' with chromate films indicates that they eiLher tropics during the past war. In some experi become ineffective when current is drawn from ments ther ate of corrosion was determined by a ~ the cell, or that the chromates function differently loss-in-weight method, considerecl later. in the presence of the paste wall (starch-flour gels) . III. Results In many corrosion problems it is possible to re Data given in tables 1 and 2 show the effects of duce and sometimes eliminate the corrosion by amalgamation and of chromate films on the rate of altering the existing environments. With dry corrosion of zinc in ammonium chloride main- l cells, however, the environments are necessary tained at 54°C. AA-size zinc cans (17/32 in. in from the standpoint of electrical output and little diameter and 1% in. in height) were used. Neither changes can be made in them. The electrolyte in amalgamation of the zinc nor use of chromate the usual type of dry cell consists of aqueous solu films was as effective as desired at 54° C. Any t.ions of ammonium and zinc chlorides. There is excess of mercury decreases the mechanical an abundance of chloride ions, which are small in strength of the cans.

152 Journal of Research

L TABLE 1. Effect of amalgamation on the corrosion of zinc electrodes (~ t 54° C "....-wire hook [AA-size zinc caos totally immersed io 20% NH.OI so lu tions]

Hydrogen evolved a t 54° C for- ligCI, per AA-sizc zi nc can

/ 1 day 2 days '/ el ectrolyte / / j / u ml ml solution 0.00 ______8. I 15.6 /~/// .01. ______4.8 II. I .02 ______////// 4.6 14. 4 .04 ______3. I 9.1 ///// .08 ______1.9 8.7 .10 ______1.9 7.2 / //// / .15 ______1. 8 6.3 .20 ______1.7 7. 2

The inhibiting properties of a large number of materials were studied. They may be classed as commercial or industrial by-products, organic compounds, inorganic compounds, and colloidal ru bber materials. The organic compounds may be sub stopper divided into noncyclic and heterocyclic nitrogen-, oxygen-, or sulfur-containing compounds_ I so cycli c compounds would not be expected to be gloss tubing effective inhibitors because of their symmetry. The one tried, sali eyl aldoxime, was not efi'ective, as anticipated. el ectiolyte Results obtained with D-size zinc cans (l}~ in. solution in diameter and 2}~ in. in height) immersed at 54° C in atm-ated aqueous solutions of am monium chloride to whieh certain commercial

FIGURE l. Gao tube for meaS1tl'ing the rate of corrosion of products or organic compounds were added are zinc cans. given in tables 3 and 4_ The data were ex tensive,

TABT,E 2. Effect of chromate films and dichromates on the corrosion of zinc electrodes at 54° C

rAA -s iz~ zi nc rans totally immersed in saturated "R,CI solution,]

Hydrogen evolved at 54 °0 [or- Inhibitor I day 5 days 10 days 15 days 20 days 25 days 30 days 35 days 40 days ------ml ml ml ml ml 'fill ml ml ml Control (no film ) ______3.3 > 13 <') 10-secoud Crona k trcatmont , ______. ______0. 5 3.5 6. I 12. 1 > 13 <') IO-minute Crooak treatment , ______.7 3.5 5.4 6.8 6.7 i.9 8.3 9.3 > 13 0.5% K,CnO, added to solution ______3. 4 6.7 9.3 10.9 12.2 > 13 <') 1.0% K,CnO, added to solutioo ______2.7 6.4 8.9 10. 9 12.2 > 13 (I) 2. 0% K,C..,O, added to solution ______3.9 8.1 10.5 > 13 <')

1 Experiment was discontinued. ' 'Th e CronRk proce,s, denloped by the New Jersey Zinc Co., inyolws the treatment of the ziGC with a solution coniaining sodium dichromate and sul [uric acid.

Corrosion of Zinc in Dry-Cell Electrolytes 153 r I TABLE 3. Effect of certain commercial i:1hibitors on the n ess may be made from the average rates. One I corrosion of zinc electrodes at 51,0 C percent of inhibitor was used in most cases. (D-size zinc cans immersed in saturated solutions of ammonium ch loride to Lesser amounts wer e generally ineffective and which 1 percent of inhibitor was added) larger amounts, as will be shown later , were unde sirable in dry cells. D ur ation Rate of Inhibitor of tcst corrosion Of the commercial inhibitors, only tetramethyl thiuram disulfide, the commercial oils, and cer tain Da ys ?nln ,/da.V of the proprietary compounds of unknown com None ...... 3.5 Zinc dibutyl·d ithi ocarbamatc ...... 5.7 position were more or les's effective. As the com Pipcrdino pcntamethylcne dithio· carbamate ______5.0 position of these materials was unknown, no 2-MercaptobcllZotbi azole ...... 4.3 general principle for making a choice of possible Tetramethyl thiuram disulfide ...... 5 1.7 inhibitors could be deduced from their study. Commercial ailg 1 ______I to 20 1.3 to O. 5 Proprietary compounds of unknown Consequently, known organic and inorganic com composition 2 ______I to 160 l OA to O. 023 pounds were studied next. The rcsults obtained with organic compounds 1 T hree types. show that noncycli c compounds have little or no , Eightccn types; most of thcsc were bclie\'ed to be heterocyclic orga n it compounds of thc mercapto type. inhibiting action with the exception of crotonalde hyde. In fact most of these materials were cor and consequently only an average corrosion rate is rosive. Crotonaldehy cl c exhibits resonance, has listed for each material. This rate was obtained conjugated double bonds, and contains the car by dividing the total volume of gas evolved by the bonyl group (=C=O). Mann [4] has shown that time (in days) of the test. The rates varied some the carbonyl group is effective in preventing cor what with the time, as shown in figure 2, for the rosion of steel. H e explained this by the forma better inhibitors. However, a preliminary choice tion of oxonium compounds, which ionize to give of inhibitors and a companson of their effective- a positively charged ion as fo llows:

R H R H + '" c=o/ --.... '" C= O/ + OR- / ~ / R OR R

7.0 ,8- na phlhoquinollne U 0 I conlrol

'0 crolonaldehyde 5.0 c Q) 0> 0 . ~ 'U 4.0 , >- .c

~ 2 () furfurol () .- ~

() quinaldine

10 20 30 40 50 60 70 80 90 100

Ti me - Days

FIGURE 2. lnhib iting~ action of some materials as a function of time, on the corrosion of zinc in saturated solution of am monium chloride at 54° C.

154 Journal of Research

L H H H H I I , I C-C c-c H H, H H H II II II II I I I I H-C C-C=O H-C C-C-O-H H-C-C=C-C=O, "l ~ "l ~ H

furfural fury' alcohol crotonaldehyde

(good) (poor) (good)

H H H H H H H H ~ 7 I I I c-c 0 0 c-c C-C H H C-C C-CI " 0 0 C-CI I N II II "II II II \I I + I II R \I I I H II H-C c-c-c-c C-H H-C C-C-N=C-C C-H H-C C-C-C-C ,C-H \ 01 \01 \ / I \ / ' "01 H H \01 o H 0 I I , , furil +N C-C II II II furoin I H- C-C C-H (poor) '0/ ( fair) hydrofuramide (good)

FIGU RE 3. Comparison oj inhibiting action oj compounds with and wi/hOltt the carbonyl group (=C= O) on the cor l' osion of zinc in satuTated solutions oj ammonium chloride at 54° C.

T A B LJ; 4, E.f) ect of organic compounds on the cOl'rosion oj zinc electrodes at 54° C

(D-si ze zi nc cans immerscd in saturated N H ,CI solutions to which 1 percent of inhibitor was added unless otherwise indicated)

lleterocycli c 1 Noncycli c I

Comp ound N umber I n ura- I Rate of Compound Dura D ate of of tlon of .'. i of corrosion r inj!s test CO Il os on test I

N ITROGEN-CONTAINING C OMPOUND S

days ",I FJ ,lda,y days ",I U ,jday Control. ______1 3.3 Control. ______1 3.3 8-H ydroxyquinoli ne 2 ______2 1 4.3 Sul fa nilic acid , ______13 Qui nalciinic acici ______2 80 0. 14 Anthrani lic acid ______.______13 II-Naphtboq uinoline. ______3 60 . 12 Benzid inc______13 p·D ipyridy L ______60 . 048 'l'riethyl amine______12.9 Quinoline______2 60 . 025 D ipbenyl thiocarhaziclc 3______12.4 II-Naph thoquinalicl inc ______3 60 . 025 Diphcn yl carbazo11 e ______11. 5 Quinoline hydrochloride _.. ______60 . 020 Anilinc______10.8 Qu ina ld ine ______, 004 Diphen yl carhazide______10. 1 P yriciine 2% ______------13 Aldehyde ammonia ,______10.0 Piperazine hydrate ______- ______6. 4 ELhanol amine______8.2 II-Picolinc ______------6,3 di-o-Tolylurea , ______,______7.0 Bruci ne______6 or 7 5. 9 Dieihanol amine ,______5.7 Salicyl aldoxi me , ______Isocyclic 5,8 Diphcnyl gu anidine ______3.9 a-Picoline______1 4. 2 Piperidine______1 3. 0

I R efers to the n itrogen, oxygen, or sulfur in the ring, 2 Also contains oxygen. 3 Also contains sulfur.

Corrosion of Zinc in Dry-Cell Electrolytes 155 771185-48--5 TABLE 4. Effect of organic componnds on the corrosion of zinc electrodes at 54° C- Continued

Heterocycli c 1 Noncyclic 1

Compound I Compound N umber Dura- Rate of Dura- Date of of I tion of I co rrosion of I corrosion rings test test

OXYGEN-CONTAINING COMPOUNDS

Ascorbic acid __ . __ . ______. ___ . ______. ______. ______13 Formic acid __ . ______. ____ .. ______._ 13 Brucine_. _____ . ______.. _____ . . ______. 6 or 7 5.9 Monocbloracetic acid ______13 FnriL ______. ______. ______.______2 2.2 Aldchyde ammonia • ______._ .. __. ______10.0 Fnryl alcohoL ______.. ______. ___ . ____ _ 1. 8 Hydroquinone ______.. _ 6.0 Fnrion. _. __ . ______.. ______. __ . _. ______. _. _. 2 10 0.53 ResorcinoL ______. ______5. 1 llydrofuramide .______3 100 . 12 Pentane dione_ . . ______. ___ ._. _____ ._. ______4.8 Fnrfural ______. ___ . _. _. _. _. ____ _ 100 . 029 Tannic acid ______. ___ _ 4.6 Bpnzoic ac id ~. ______1.4 I Crotonaldehyde _____ ------.------100 0. 075

SULFUR-CONTAI NI NG COMPOUNDS (ALL NONCYCLIC)

Thiophenol ______o 13 Sulfonated castor oiL _ ------_. ______1 3.8 Snlfanilic acid ______.. ______. __ o 13 Potassium etbyl xanthate ___.- ______I 2.0 Diphenyl thiocarbazide ______. ______o 12.4 AmmonIUm sulfate (morgalllc) ______2.8

• Also contains nitrogen. The positive ion is attracted to the cathodic areas H eterocyclic compounds were the most effective of the steel and held there in a H elmholtz double inhibitors, providing a second or third ring was layer, thereby increasing the hydrogen overvoltage attached to the heterocyclic group. This is shown of the cathodic areas. in figure 4. Pyridinc, piperidine, a-picoline, and The effect of the carbonyl group in retarding iJ-picoline, which are simple heterocyclic com the corrosion of zinc is clearly shown for hetero pounds containing nitrogen in the ring, had little cyclic compounds in figure 3. Furfural contains or no inhibiting action, although these compounds the carbonyl group and is a good inhibitor. may form addition compounds with acids or water When the carbonyl group is converted to an alco and subsequently ionize to positively charged holic group by the addition of strong alkali, a ions. On the other hand, phenyl and naphthalene compound, furyl alcohol, is obtaincd that has no derivatives, such as quinoline, p-dipyridyl, quinal inhibiting properties. It might be expected that dine, iJ-naphthoquinoline, and iJ-naphthoquillal furil would be an excellent inhibitor, as it contains dine do show inhibiting action. Here the posi two carbonyl groups. However, this is not the tively charged ion is attracted to the cathodic case, apparently because the carbonyl groups are areas at the nitrogen, and the large molecule, effectively shielded from the zinc surface by the being unsymmetrical, is inclined at an angle to fur an groups, or close packing of the positively the zinc surface whereby a large area of the zine charged ion on the zinc surface is not possible. surface is effectively covered. p-Dipyriclyl is The ineffectiveness of diphenyl carbazide, di attracted to the zinc at both nitrogens and prob phenyl thiocarbazide, or di-o-tolylurea as inhib ably lies flat on the surface of the zinc and is itors may be explained on the same basis. The therefore a good inhibitor. These conclusions carbonyl or sulfonyl group in these compounds is agree, in the main, with those of Mann [4]. The shielded from the zinc surface by thc phenyl or better proprietary materials probably were hetcro toluyl groups. Furoin inhibits slightly the cor cyclic compounds. rosion of zinc, and therefore the carbonyl group The compound, 8-hydroxy in this compound must be attracted to the zinc quinoline shows no inhibiting properties, as tho surface. Hydrofuramide, which contains thTee OH- group favors orientation of the wator furan groups, probably owcs its inhibiting prop molecule in its vicinity. This effect probably erties to the nitrogen and acts as a positively influences the position of the inhibitor ion in the charged amine ion. Helmholtz double layer.

IS6 Journal of Research + Q 0+ + CO+ CO+ Pyridine Piperid i ne Quinoline p"'dipyr i dyl ft -na phthoquinoli ne (poor) (poor) (good) (good) (good)

+ + + co+ OH DC:-picol ine ,8-pi coline Quinaldine f3 - naphtha quinaldine 8-hydroxy quinoline (poor) (poor) (good) (good) (poor)

FJCURE 4. Comparison of heterocyclic-nitTogen containing compounds showing the effect of molecula1" size on inhibition of c01"rosion of zinc in saturated sol1ltions of ammonium chloride at 54° C.

IV. Electrical Characteristics of Dry Cells with the paste wall and decrea cd its strength [5], Containing Inhibitors or formed a high-resistant fi lm on the zinc. In the latter case, these compounds probably form In the foregoing it was shown that a number of an insoluble precipitate over the entire surface of organic compounds eff ec tively curtail the local the zinc, which stifles not only the undesirable corrosion of zinc in aturated solutions of am corrosion reaction, but also the cell-discharge monium chloride. A few principles were ol~t l in e d reaction. In fact, Bartram and K ent [6] have for making a choice of possi ble inhibitors. T o shown that zinc and quinaldinic acid react to determine whether these inhibitors could be form an insoluble complex precipitate. imilar successfully incorporated in dry cells, a number complex precipitates may be formed between the of D-size cells were made in the usual manner zinc and the other organic compounds. in which some of the better inhibitors were In tables 5 and 6, the res ults of "light industrial incorporated. The inhibitors were introduced tests" of the cells arc given. The " light industrial into the cells in one of three ways: (1 ) by spinning test" gives the time in minutes of discharge to a an alcoholic solution of the inhibitor on the zinc cut-off voltage of 0.90 volt through a resistance of can and removing the alcohol by evaporation, 4 ohms for 4-minute periods beginning at -hourly ~ (2) by adding the inhibitor, if soluble, to the intervals of 8 consecutive hours every day, with electrolyte of the paste wall, or (3) by grinding 16-hour rest periods intervening. At 21 0 C, the the inhibitor with the cereal (starch and flour) electrical output of cells made with or without before the paste wall was gelatinized. inhibitors is approximately the same after 1 week On the whole, dry cells made with inhibitors of storage. At high temperatures, however, the did not give the expected increase in shelf life 01' cells containing the inhibitors gave less output, electrical output. The inhibitors either reacted probably for the reasons cited above.

Corrosion of Zinc in Dry-Cell Electrolytes 157 WI

TABLE 5. Light industrial tests of D-size dl'Y cells containing the proprietary compounds and with Oronak inorganic or commercial inhibitors treatments for various times did not show any correlation between electrical output and per Discharge after storage at- centage of inhibitor. It. is probable that there is Inhibitor · 54° 0 an optimum concentration for each inhibitor. 21 ° C, 1 week However, considerable work would be required I montb 2months 3 mon tb s ------1------to ascertain this optimum concentration, which min min min min may depend on the concentration of the electrolyte Con trol at 21° C, 0% ______788 1 783 1781 1778 and the composition of the paste wall, because the Oontrol at 54° C, 0%______' 788 779 739 700 N H ,HF,, 1.0% ______776 3 ------inhibitors m ay react with the paste wall. K 20"'O,, 2.0%______779 699 Cron ak treatment, 5 sec ______708 614 192 V. Inhibiting Characteristics of the Paste Oronak treatment, 10 sec ______694 640 3240 Oronak trcatm ent, 20 sec ______Wall 706 650 • 280 Oronak treatment, 40 sec ______681 63 7 ~84 As the organic and other types of inhibitors for Cronak treatment, 80 sec_-''-- ______689 733 262 Oronak treatm ent, 180 sec ______718 708 288 zinc did not improve the shelf life or electrical output of dry cells, studies were made of the Zinc dibntyl - dithiocarbamate, 5.0% ______676 Ni6 organic substances commonly used in the paste Commercial oils, 5.0% ______622 488 , 363 wall (or separator) of the dry cell. It. has been Pl'oprietan" compounds of unknown composition, 1 to 10% 6. 754 544 207 49 recognized that sts,l'ch and flour not only provide a satisfactory separator for the dry cell, but also • Inbibit.or was mixed with the electrolyte of tbe paste wall or ground witb inhibit to some extent the local corrosion of the the cereal. 'I'be electrolyte consisted of 22. 78 g TH ,CI, 6.1 g ZnC'b, and 49 .4 zinc electrode, without aff ecting the electrical m l of H 20 . To t his waR addcd 7.15 g eacb of waxy cornstarcb, Buff alo corn starch, and l'iIIsbnry XXXX w heat flour. output of the dry cell. These materials may be 1 Stored at 21 ° O. 'Stored at MO O. classed as water dispersible organic colloids. 3 192 minutes at the end of 4 months. • 133 minutes at tbe end of 4 m on t hs. Most colloidal particles are positively charged in , 105 minutes at the end of 4 mon tbs. acid solutions and negatively charged in alkaline • A verage of 20 test.s on diO'eren! inhibitors. olutions. These positively charged particles in TABLE 6. Light industrial tests of D-size cells containing acid may th en migrate to the cathodically active organic inhibitors spots, which are thereby screened or poisoned. In table 7 are given results obtained at 25° 0 D ischa rge after storage at- with zinc immersed in saturated aqueous solutions Inhibitor 1 MOO of ammonium chloride to which colloidal materials 21 ° 0, 1 wcek were added. The natural starches had little pro I month 2 months 3 montbs ------1------tective value. The only modified starch that min min min min appears to have appreciable inhibiting properties Oontrol at 21 ° 0, 0.0%______788 , 783 ' 781 , 778 Control at 54° C, 0.0%______3 788 779 739 700 was a modified starch prepared by H. S. Isbell of Furfural, 0.5%______802 644 368 34 this Bureau. His method involved a treatment :FurfuJ'al, 1.0%______879 569 <3 H ydrofuramide, 0.5% ______524 411 257 of starch with chlorine dioxide, whereby the aidehyde end groups of the starch are converted Orotonaldebyde, 1.0% ______464 <3 Quinaldi})e, 1.0% ______731 648 ---.------to aldehyde groups. Qui naldine, 2.0% ______871 664 ------.- Oornmeal, zein, and starch glycerite all appear Q uinaldinc. HOI, 1.0%______731 648 < 3 ------to have inhibiting properties. On the other hand, QuinaJdine .. HOI, 10 .0%______382 <3 ------dextrose and the soluble starches, inulin and gly Faity acids of COl'll. 0.02% ______429 Hydroquinone, 0.03% ______< 3 ______• ______cogen, accelerate the corrosion of zinc in saturated solutions of ammonium chloride. The starches 1 The composit ion of t h e paste wall was t be same as gi ve n in the footnote apparently have no protective value, because of table 5. The inhibitor was added to t he electrolyte of the paste wal! or they are no t positively charged. In separating was grotmd w ith t he cereal. , Stored at 21 ° O. 3 Stored at 54° O. the amylose and amylopectin components of starch by til e electrodialytic method, amylose The amount of inhibitor to use in dry cells migrates to the anode, and amylopectin remains re.mams a moot question. Data obtained with stationary. Amylose I S therefore negatively

158 Journal of Research .. :p

charged and should be without effect on the T A BLE 7. Results obtained with colloidal materials at corrosion of zinc. The high corrosive effect of 25° C- Continued amylose may be attributed to the butanol that IS MISCELLANEors MATER[ALS used to keep it in a solu ble condition. Duration Rate oC Material oC test corrosion

T A BLE 7. Results obt ained with colloidalllu1terials at 25° C Days ml R ,/da y Ammonjum algin ate ______~ __ 11. 3 fAA·size zin c cans immersed in saturated NH,Cl solutions to whicb 2 per Carpenter's g lu ~ ______3.2 ce nt oC material was added] R ydl'oxyethyl cellulose ______1 2.5 Carboxymethyl cellulose ______20 0.57 Material Duration Rate of casein ______40 . 24 I oC test co rrosion r Agar agar. ______40 . 21 Yellow corn meaL ______20 . 17 NATURAL STARCHES Hemoglobin ______70 . 16 A midex ______Days onl H,/da y SO .15 Control ______White corn meal. ______100 . 11 25 0. 43 Gelatin ______Insulin (da hlia) ______100 .11 5 2. 20 Powdered acacia ______Glycoge n (animal) ______10 1. 23 100 . 11 AI'l'O\Vroot ______15 0. 70 Sweetpotato ___ -- _------__ ------20 . 64 COrll ______20 • GO The flours, with the exception of potato flour, Maine pota to ______20 . 54 \\iaxy corn ______~ __ _ 20 .53 which is low in protein content, all have inhibiting Sago ______25 .52 Rice starcb ______. ______properties for the corrosion of zinc. It is evident 20 .44 Cas!'ava. ______30 .39 from the data that the effective ingredient in I dabo potato ____ . ______30 .39 flour is gluten, the protein-containing material.

N AT URAL F L OURS Gluten consists of glutenin, and gliadin and mes onin in different states of aggregation. R esults ob Potato ______. ______30 0. 40 Soybcan ______. ______tained 'with the e materials obtained from wh.eat 80 . U Rice. ______100 . l 2 flour by the combined methods of Stockelbach Buckw heat ______. 90 . 086 and Bailey [7] and Haugaard and .Johnson [8J IOU I .085 100 . 081 showed that with the excep tion of glutenin all I ~~e::~;~ ~~;:t:::: ::::::::::::::::::::::::: ::::::::: 100 I . 080 were good inhibitors of the corrosion of zinc. When these materials are incorporated in dry MODIfIED STARCllES cells they increase somewhat the capacity of dry ;, Control 25 0.43 cells at 21 0 C. as hown by the data of table 8. D extrose ______10 1. 27 Corn amylose I . ______Current maintenance is defined a the ratio of the I 1.10 V{axy corn amylopcctin ______• ___ _ 10 0. 91 flash (or short-circuit) current after an interval of Corn amylopectin ______. ______20 . 00 storage to the initial flash current. Many com R - 20 cornstarcb , ______20 . 60 400 1 cornstarch , ______25 . 48 mercial cells give better current maintenance than 3006 cornstarch ' ______25 . 48 the experimental controls given here. However, DeCatted cornstarch ______30 . 44 Zein ______'10 .31 cells with and wi thout proteins (experimen tal con GlyC<'r ite oC starch ______45 .27 trols) were made under the same procedure, and HSt corllstarch 4. ______.. __ _ 50 . 15 comparisons should be made with the experimental NIODIFIED FLO URS controls and not with commercial cells. On hy drolysis the flour proteins give amino acids. Glutcnin __ ------30 0.36 4° gli adin ______30 . 27 Therefore, studies were made of the inhibiting l Mcsonin L ______50 . 14 GUill glu ten ______properties of several amino acids, with the r esults ]00 . 094 G liad in ______100 . 087 given in table 9. The amino acids have no pro Wbeat nour ______100 . 085 M eso nin 11.. ______tective value; in fact most of them accelerate the 60 . 063 corrosion of zinc. Other colloidal materials, including gelatin, 1 T o keep amylose in a soluble form , it is kept mo ist with n-butanol. The agar, acacia, casein, and hemoglobin likewise have high valu e for amylose may be caused by the prese nce of a-butanOl, which may have a cOl' l'os i ve e ffect. inhibiting action. Carboxymethyl cellulose, hy , Supplied by Lhe National Slarch Co. droxymethyl cellulose, and ammonium alginate, 'Supplied by R . W . .K err, Corn Prod ucts Refining Co., Argo, lll. , Prcpared by U . S. Isbell , National Bureau oC Standards. which have been tried in the dry cell as substitutes

Corrosion of Zinc in Dry-Cell Electrolytes 159 I ( for starch and flour, have no inhibiting propcrties. TABLE 10. Loss i n weight of D-size zinc cans containing Amidex, a degradation product of starch, has 30 ml of paste after storage at 54° C some inhibiting properties. Loss in weight aftcr storage at 54° C lor- Gel constituents TABLE 8. Effect of pl'otein films on the current maintenance 1 of D-si ze dry cells at 21° C 43 hr 259 br 284 hr 339 hr ---_·_------1------• g g g g Current maintenance 100% water (distilled).______0. 04 N um- Type of paste wall ber of -- 5 parts NH.CI and 95 parts H20______'. 04 cells 1 3 I 6 1 15 20 parts NH. C1 and 80 parts H 20 ______s. 22 month months, months year month A 30 parts NH,CI and 70 parts H 20 __ . ______'.36 ------40 parts NH, CI and 60 parts H 20 . ______'. 63 C ontrol (c orns tarch Electrolyte mixture I (32.2% NH ,CI; 6 7%, w h ea t flour Percenl Percent Percent Perc ent Perren 4.12% ZnC!,) ______:______0. 04 0. 19 ______.32 33%'l-______------48 91 84 75 6G 5 Cornstarch 67%-gum 21.5 g cornstarch ______. 05 . 26 O. IS gluten 33% ______3 91 87 77 61 51 IG.l g cornstarch and 5,4 g wheat flouL ______~ ______Control+ mesonin film' . 9 94 90 79 81 71 f .04 C ontr ol+ me sonin I film ______9 95 88 84 ------R' llOfl.7o~r c ~~~~ ~ ~~ ~~_ ~_~~ _ ~~ : ~ _ ~ _ ~~~ ~~ ______. 04 Control+m eson in II 5.4 g cornstarch and 16.1 g wheat Oour ______. film ______2 91 91 82 ------. 06 Control+ gliadin film ___ 9 98 100 102 84 ------. 21.5 g whcat flour ______. 05 . 05 .05 Con t rol+ 4°g1i a din film ______5 102 ------88 94 ------21.5 g waxy cornstarch ------. 12 Control+ ( - 100) -gliadin C'. 121. 5 g swcctpotato starch ______------. 13 film ______7 94 ----. --- 92 86 83 2.0 g carboxymethyl ccllulose ______. 06 . 36 Control+ n eu tralized !2.5 g carboxymethyl cellulose . ______. 18 4° gliadin tilm ______6 94 89 87 80 7 1 21. 5 g cornstarch and 21. 5 g wbeat flOUT ______. 07 J Curren t maintenance is the ratio of the fl ash (or short-ci rcuit) current 16.1 g cornstarch and 16 .1 g wheat after storage to the ini tial fl ash currcnt. D' flour ______------.09 , M any commercial cells give hetter current main tenance than these ex 10.7 g cornstarch and 10.7 g wbeat perimental controls, but comparisons must be m ade with the lattcr. flour ______. 08 , All film s covered the zinc can. 5.4 g cornstarch and 5.4 g wheat fl OUT ______. ______. 07 TABLE 9. Effect of amino acids on the corrosion of zinc electrodes at 25° C 21.5 g cornstarch and 2.0 g gliadin ______. 07 21. 5 g cornstarch and 2.0 g gluten ______. 09 7.1 g cornstarch, 7.1 g waxy cornstarcb and 7.1 g wheat flour ______I Acid Duration Ra te . 04 of test E2 2.5 g carboxymcthylcellulose and 2.0 ( g gliadin ______. 07 2.5 g carboxymethylcellulose and 2.0 Days ml EI,jday g gluten ______Control. ______25 0. 43 . 09 1 Leucine. ______> ______5 2. 32 1.3 g carboxymethylcellulose and 10.7 g wheat flou r ______Aspartic acid . ______2. 06 . 05 Glutamic acid ______5 1. 84 Arginine ______. ______10 0. 92 T ryptophanc ______. ______10 .91 J The electrolyte mixtm e consisted of 46.1 g of NH ,CI and 12. 3 g of ZnCIo Con troL ______. ______25 . 43 in 100 ru!. of H 20 . Serine ______10 .86 ' Added to 93 ml of thc elec troylte mixture. Phenyl alanine ______._. ______20 . 75 s 334 hours. Histidine.Hel _._. ______10 . 50 • SatUTated; crystals of NH,CI on can. Valine ______. ___ __ 40 .33 Aanine ______40 . 31 A paste of electrolyte and cereal was poured into a D-size zinc can of known weight, and mineral VI. Zinc-Corrosion Experiments by the oil was poured over the paste to prevent evapora Loss-in-Weight Method tion. After storage in an oven at 54° C, the can was thoroughly cleaned of paste and weighed. As the conditions in the gas tubes are not en Results obtained by this method, given in table tirely similar to conditions in the dry cell, and as lO(A) show "the corrosive effect of the electrolyte solutions of higher concentrations of starch and concentration. Concentrated solutions of am flour could not be conveniently and sati.dactorily monium chloride are more corrosive than dilute, tested in these tubes, a few corrosion experiments and the addition of zinc chloride decreases the by a loss-in-weight method were undertaken. rate of corrOSlOn, as expected. Results of table

160 Journal of Research 10 also show that a flom paste has more inhibiting It was found that the paste wall of the dry cell properties than a Larch paste (B) or a carboxy has inhibiting properties, provided it conLain a methyl cellulose pa te (C), in conformity with the colloidal protein like wheat flour. The active results obtained wiLh the gas experiments. The inhibiting agent of the flour was found to be amount of flour in the paste (D ) that is necessary gluten. The constituents of gluten, namely, to inhibit corrosion need not be very large. The glutenin, and gliadin and mesonin in different results (E ) also show that a starch or a carboxy states of polymerization were isolated. Of these, methyl cellulose paste can be made less corrosive gliadin and mesonin were found to be effective in by the addition of mall amounts of either flour, retarding the corrosion of zinc in dry-cell electro gum gluten, or gliadin. lytes. vVhen these materials are incorporated in dry cells, they increase somewhat the capacity of VII. Conclusions the cells.

The results of this investigation have shown that The authors are indebted to Alfred Douty and the corrosion of zinc in saturated solll tions of George S. Gardner of the American Chemical ammonium chloride may be cm tailed by organic Paint Co. , to C. F. Prutton of the Lubrizol compound containing the carbonyl group, by Corporation, and to D. G. Miller, of the Reilly heterocyclic nitrogen-containing compound with Tal' & Chemical Corporation, for their coopera one or more rings attached to the heterocyclic tion during the cour e of thi investigation and group, by certain commercial products, and by for furnishing many of the commercial inhibitors. colloidal materials. Of these the heterocyclic compounds Wero the best. For example, quin VIII. References a1dine, quinoline, furfural, p-dipyridyl, {3-naph [II W. J. Hamer, J . P. Schrodt, and G. W. Vinal, Trans. thoquinaldine, and ,B-naphthoquinoline were bet Electrochem. oc., Preprint 90- 30,367 (1 946). ter inhibitors of the corrosion of zinc in dry-cell [2] L. C. Copeland and F. S. Griffith, Trans. Electrochem. electrolytes at 54° C than mercury (amalgama Soc. 89, 495 (1946). [3] If. F. 1\1 cl\1urdic, D. ~J. Craig, and G. "l'V. Vinal, Trail::). tion) 01' chromate films. Electrochcm. Soc. Preprint 9 31,383 (1946). Of thoso materials found to retard the corrosion [4] C. A. Mann, Trans. Electrochcm. Soc. 69, 115 (1936). of zinc, only the colloidal material could be [51 W. J. Hamer. Unpublished. incorporated in dry cells. Mo t of the non col [6) J. H . Bartram and P. J. C. Kent, i\Ictallurgia 33, 179 (1946). loidal materials either reacted with the paste wall [7) L. S. Stockelback and C. I-I. Bailey, Cereal Chern. 15, of the dry cell or formed all insoluble film, like a 801 (1938). varnish, over the entire surface of the zinc anode, [8] G. Haugaard and A. H . Johnson, J. BioI. Chem. 93, and contributed to the internal resistance and 677 (1931). reduced the electrical output of the dry cell. WASHINGTON, October 1, 1947.

Corrosion of Zinc in Dry-Cell Electrolytes 161