CORROSIOS O F LE.ID I X A SOX-POL.LR SOLVEST
1101
(6) CASSEL,H., ASD FORJISTECHER, 11.:Iiolloid-Z. 66, 18 (1932). ( i )CASSEL,H . , ASD SALDITT, F . : Z. physik. Chem. 156, 312 (1931). (8) FORCH, C . : ;Inn. Physik 17(4), 744-62 (1905). (9) FREUSDLICH, H . : Colloid and Capillarg CiwiiistrU, p. 67. Methuen and Co, Ltd., London (1026). (10) Reference 9, p. 66. (11) Reference 9, pp. 66 and 195. (12) GTAKI,B. P . : J. P h y s . Chem. 49, 4-l2 (1945). (13) GYASI,B. P . : Thesis (pp. 47, Si), London, 1947. (14) Reference 13, pp. 102, 103. (15) Reference 13, pp. 36. 54, 55. (16) GYAXI,B. P . : Proc. S a t l . Acad. Sci. (India) 14A, 85-94 (1944). (17) IHEDALE, T.: Phil. Mag. 45, 1058 (1923). (18) I i R U Y T , H. R., ASU ~ I O D D E R X J .AGS. ,: I n t e r i i a t i o i d Critical ‘ I ’ U b i e S . 5.01. \-, p . 13Q. NcGran---HillBook Company, Inc. S e w York (1929). (19) I ~ R U Y HT . R, . , ASD lIODDERJf.4iV: J. G . : Chem. Revs. 7, 259-346 (1930). (20) LASGMUIR, I . : J. Chem. SOC.38, 221 (1916). (21) MILLS,H.: J. Cheni. SOC.1932, I, 419-30. (22) OLIPHANT,A I . L . : Phil. Mag. 6, 442 (1928). (23) STRAXSKI, I. S.:Z . Elektrochem. 36, 25 (1930). (24) SZYSZKOU-SKI, 13. vox: Z. phj-sik. Chem. 64, 385 (1908). (25) TAYLOR, H. S., EYRISG, H., ASD SHERRIAK, A . : J. Chem. Phys. 1, 68 (1933). . :!.hn.266, 27 (1891). (26) T R A ~ BIE
THE CORROSIOS O F LEAD BY XYLESE SOLUTIOSS O F L l U R I C ,1CID -1SD p-QUIYOSE C . F. PRUTTOS1
I\U
,J. H. D.I17?
Case Institute o j Technology. C l c t e l a n d . Olizo R e c e i r e d Octobcr 27. 1948 IXTRODUCTIOX
The mechanism and kinetics of the corrosion of pure lead by organic acids and various oxidizing agents in a non-polar solvent have been studied by Prutton et al. (3, 4, 5 ) . Pure fatty acids such as lauric do not appreciably attack lead unless an oxidizing agent is present (1, 2, 3). Denison (1) postulated that the corrosion oi metal bearings proceeds in steps: (a)the formation of peroxides by reaction 01 oxygen with the lubricating oil, ( b ) the reaction of peroxide with metal t o form metal oxide, and (c) solution of the oxide by organic acid t o form a soluble soap. Prutton, Turnbull, and Frey (3) found it unnecessary to postulate the intermediate formation of a peroxide, molecular oxygen being capable of acting as the oxidizing agent. lI’iesent address Director of Research, Mathieson Chemical Coiporntion, K m S e w York *Present address Department of Chemistr) , Ohio University, .Athens, Ohio
1-01I;
1102
C . F. P R L T T O S . I S D J . €1. D.LY
Previous ivorli (4)has intlicatecl that the Inec~hanismof the lead corrosion Tvhen p-quincme i* thc ositlizing agmt may he tlift‘erent from that n-hen osygen or other organic oxidizing agents such as lnuroy1 peroxide ancl trrt-butyl hydroperoside are used. The object of the present inr-cstigation ]\-as t o determine the corrosive effects on pure lead of noli-polar solutions of a typical high-molecular-n-eight organic acid (lauric) and p-quinone in the alisence of osygen, and t o determine as far as possible the meclianism nnd kinetics of such corrosion. .iI’P.IR.lTUS
. I S D SIhTERIALS
The apparatus consisted of a onst st ant-temperature hath, regulated t o &0.05’C’. hy means of a merc.iiry-rcscrvoir-t!-pe make-and-break regulator which actuated a mercury relay -lciie. Thebe. s m i c ~ propertie!: iverc o l x e r ~ e di n the extracted precipitate.
E,fwt of" added lend
lalii.nfe
To determine the possihility of the lead lawate foimed in the roi.ro~ionreaction either catalyzing, or inhihiting. fiirther reacation, tivo i'iiiis \\-?re made i\.itli added
1113
CORROSIOS OF L E l D I S A KOS-POLAR SOLVEST
metal >haft, is open to theoretical objection, since the joining of tn-o different metals (as lead and iron) should form a couple, and the metal shaft itself could be corroded. However, since equivalent results for lead corrosion were obtained using both Bakelite and steel stirring shafts, it is felt that objections t o the rod material are more apparent than real.
Discussion of possible mechanism I n the corrosion of lead by organic acids, where the Oxidizing agent was oxygen or Some oxidizing agent which could lose an oxygen atom, such as hydroperoxides, Prutton c t a?. (3, 4, 5 ) have postulated the follo7yingmechanism:
+ +
Pb 302 + PbO PbO 2HX 4 H20
+ PbA42
This is in agreement n-ith the mechanism postulated by Denison ( I ) . However, in thiq 11-ork, n-ith quinone as the oxidizing agent, it seems unlikely that the quinone would react with the lead t o form lead oxide, and at no time did analysis show the presence of a lead oxide either in the solution or on the surface of the lead piece. Hence the mechanism given above u-ould not fit this case. Quinone is reduced t o hydroquinone. It is suggested that the possible mechanism may be the reaction of quinone with lead to form a quinone-lead complex, and that this complex may then react n-ith lauric acid t o form lead laurate (or perhaps a t first lead laurate quinonate) and hydroquinone. If the laurate quinonate n-ere an intermediate, it n-ould then react further with acid t o form the normal salt.
0
A 1' ~' '\/
+
Pb
+
HA
-+
O
H
H O O O H
+
/ Pb \
O
O
-
*4
0 O Pb
U
O
/ - \
H
+
HA1
-+
-
PbA2
,1
It i. seen from the chemical equation postulated that the rate of reaction should be proportional t o the amount of lead surface offered by the lead test piece. Khere this is coiictant, the area of the lead surface n-ould not enter into the rate equation, except in giving the area 9, so that the rate constant will give a rate per square centimeter of lead. Prutton, Turnbull, and Frey (3) found that in fact the rate of reaction 11as a linear function of the lead surface uncovered. This agrees \yell n-ith the modified first-order reaction equation here set up for the quinone reaction rate constant; the lead surface is here the variable, but since it is a function of the
1114
C . F. PRCTTOS A S D J. H. DAY
quinone concentration (at fixed acid concentrationj, then it may be expressed as a quinone function, as was done in the modified equation derived from the Langmuir adsorption isotherm. From the chemical equation it' appears that the rate of reaction should he a function of the first power of the quinone concentration, and this n-as found to he so, as well as the quinone-lead equivalence of 1:l predicted by t,he chemical equation. The chemical equation calls for 2 moles of acid to 1 mole of lead, and this fits the experimental facts. Hon-ever, it seems that the reaction rate might IIe proportional t o the second poTI-er of the acid concentration. From the experimental data, the evidence of acid reaction rate is obscured by the appearance of an "apparent" zero-order reaction rate, but the fact t'hat doubling the acid concentration doubles the rate of corrosion suggests that the order n i t h respect to the acid is first. I'rutton et al. (4) found a first-order reaction rate for organic acids in analogous tests, even though the product contained variable amounts of acid radical per mole of lead.
Effect of temperature Experimental runs were made using 0.01 III lauric acid and 0.002 A I quinone at 70°C. and 80°C. to determine tmhetemperature coefficient of rate of corrosion, and t o find the apparent' energy of activation. These runs n-ere made using the cont,inuous-sample method. Similar runs, but using the separate-sample technique, \\-ere made using 0.01 JI lauric acid a i d 0.01 JI quinone. 1-alues of the apparent activation energy for the quinone rate-controlled reaction were calculated in the usual fashion. The value of A E calculated from runs by the continuous-sample method ivas 9580 cal. per mole. The value of L E from the separate-sample technique \vas found to be 9340 cal. per mole. For the reaction in 11-hichthe rate is primarily controlled by the acid vunvent r ~ t tion (0.004 -11 lauric acid and 0.004 A/ quinone) the calculated ene tion is 16,800 cal. per mole. This ma;v be compared with the value of 14!000 cal. per mole for the activation energy of the lauric acid reaction n-hen the osidizing agent is oxygen from the air, found by Turnhull and Frey ( 5 ) . The difference in activation energies when the rate is controlled by the quinone concentration and when the rate is controlled by the acid suggests the prohahility of separate, consecutive reactions. If the reaction were a single higher-order reaction, the activation energies \\-auld not vary in this manner x i t h relative concentration.
..lPiscellaneoiis c ~ f e c t s (1) E$ect of stirring speed: The cylindrical lead pieces used in these tec-t s ivere rotated at 1723 R.P.M. to eliminate the possibility of a diffusion layer of product being formed in the solution about the surface of the lead. If any such concentration gradient \\\-ereset up at the reacting surface, then the rate of reaction n-odd be a function of the rate of cliflfusion of reactant'. If this Tvere the case, then the rate of corrosion i~-ouldbe dcpcndent upon the stirring speed of the test piece.
C O R R O S I O S O F LE.ID I S A
SOS-POL.iR
SOLVEST
1115
Prutton, Turnhull, and Frey (4) found that for xylene qolutions, Jvith air as the oxidizing agent, rotation speed5 in excess of 500 R . P x . were sufficient t o eliminate solution diffusion as a rate-deterniining factor. For the corrosion of iron in aqueous solutions of hydrochloric acid, +Abramson and King3 found steady coiio*ion a t peripheral speeds of 3000 cm. sec. and above. For the lead test pieces used in this work, the stirring speed of 1725 R.P.M. corresponds t o a peripheral speed of 3800 cm./sec. Speeds of 2400 and 2700 R.P.M. give the same results as does a *peed of 1725 R.P.M. (-3) E,flect of water on thP corrosion of lead by quinone and lauric acid: I n all cases the reaction flasks and reactants were carefully freed of traces of moisture. That thi- precaution is necessary was demonstrated by adding 1 ml. of distilled water t o the corroding solution and comparing the results with a run identical except for the ivater. approximate 50 per cent increase in corrosion rate was observed in the n e t run. (,3) E.fect of letting xylem stand bffore use: The xylene used in all runs 11-as freshly distilled from metallic sodium not more than a n hour prior to use. Prutton, Turnbull, and Frey (3) recommend that the xylene be used within not more than 24 lir. after distillation. Run4 were made with sylene which had stood for 5 days after distillation in a glass-htoppered bottle. The old xylene gave a corrosion rate nearly 100 per cent greater than n-hen freshly distilled xylene n-as used. I n the old xylene, at 20 min. the total amount of corrosion exceeded the amount of corrohion possible if all the added olidant nere used up. This may be due t o the formation of peroxides in the xylene by reaction with atmospheric oxygen. ( 5 ) E,fect of absence of oxidizing agent: To determine the extent of corrosion of lead hy lauric acid, two tests were made. I n the first a 0.01 JI lauric acid solution containing 4 x 31 acetic acid t o catalyze the corrosion n-as run in the absence of air or any other oxidizing agent. The total n-eight lost by the lead test piece in 30 niin. was 9.8 mg. This is t o be compared n-ith a total weight loss for the same length of time in an identical run open t o the air of 178.5 mg. In 0.04 -11 lauric acid, with no oxidizing agent present, the corrosion was negligilile. Trhereas the same run with 0.002 -11quinone added lost 32.2 mg. in 30 min. I n 0.04 AI lauric acid, n-ith no oxidizing agent present but n-ith 1.0 ml. of di>tilled water, the corrosion v-as only 0.6 mg. in 5 min. ( 3 )E,fect of yiiinone i n absence of acid: T o find if there was any attack of the lead te-t piece v-hen there was no acid present, a run n-as made in 0.004 -11 quinone. The weight loss a t the end of 13 min. was only 2.2 mg. There \vas no Ti-ible film formed on the lead surface. (61E f m t of no stirring o n the corrosion rate: Since of necess'ty the lead test piece as put into the solution for a short time before stirring, and timing, n-as begun, it n-a. necessary t o linon- hon- much the lead might corrode under these circumstance-. I n 0.01 *11lauric acid, catalyzed TI ith Ix A1 acetic acid and open t o the air, at the end of 1 hr. the total n-eight loss was only 8.9 mg., a rate of 0.13 mg. min. This compares u-ith an initial rate of lead loss per minute when stirring A ~mIi s o n arid K i n g . J d m . Cheni SOC.61, 2290 (1939).
1116
C. F. PRGTTON A S D J. H. DAY
proceeds of 9.5 mg. K h e n the corroding solution vas 0.01 X in lauric acid and 0.002 -11 in quinone, there was no measurable weight loss in i min. The hame run, v i t h stirring at 1'72.5 R.P.M., gives an initial per minute weight loss of 32.8 mg. cosCLuBIoss 1. Lead is rapidly corroded by oxygen-free xylene solutions of p-quinone and lauric acid, neither of 11-hich by itself attacks lead appreciably. 2. The reaction is a complex one: I n excess lauric acid, the reaction rate is first order n i t h rehpect to the quinone. One mole of quinone reacts 71-ith 1 mole of iead. The reaction order n-ith respect t o the lauric acid is "apparent" zero order, but is probably in fact first order. TITOmoles of acid renct with 1 mole of lead. K h e n the acid: quinone ratio is less than 3, a film is built up on the lead surface; the fraction of surface covered (when lauric acid is 0.01 -11)is given by the Langmuir adsorption isotherm
'
'73 1[QI
= 1
+ i99[Q]
so that the expression for the rate equation becomes:
x.=
2.303 ( I -y) At log
TI',
1RF-v
K h e n the acid:quinone ratio is near 1, the rate of corrosion of the lead is constant until the reactant concentrations h a r e suffered large change. The final products of the reaction are lead laurate and hydroquinone; a possible intermediate is lead laurate quinonate. The probable mechanism is given hy the equations:
f
Pb
+
HA
--
+
Pb'
\
x
It is postulated that a quinone-lead complex is formed on the lead surface, a t a rate proportional t o the fraction of lead surface uncovered and proportional t o the quinone concentration, and that this complex is
T H E SYSTEM B R O R I I S E - T E T R A M E T H T L A ~ l R f O S I ~BROMIDE ~~l
1117
capable of reacting n-ith lauric acid to give lead laurate, the rate of this reaction being a function of the initial acid concentration. 3. The reaction rate is increased by the presence of water and is unaffected by the presence of lead laurate. REFERESCEB (1) D E S I ~ O SG, . H . : Intl. Eng. Chem. 36, 477 (1944). ( 2 ) Fox: .inalyst 8, 116 (1SS3). (31 PRT-TTOS, C . F . , F R E YD. , R . , TTRSBTLL, D.. .%SD DLOI-HT, G . : Irid. Eng. Chem. 37, BO (1945). (-ti PRT-TTOS. C . F . , TTRSBI-LL, D.. .&SI) FREY,D . I i . : I n d . Eng. Cheni. 37, 9 1 i (1945,. ( 5 ; Tt-RSBI-1.1.. D . , A S D F R E YD. . R . : J . Phys. Colloid Cheni. 51, 681 (1947).
O S T H E PH-ISE: DILIGRAUIOF T H E T K O - C O M P O S E S T SYSTEM BROJIISE-TETRd~IETHkL~I~I~~O BROJIIDE S I ~ - ~ I ASD T H E SOLUBILITIES OF THE C O J I P O S E S T S I S TYA\TER R . BLOCH,' L. FAIRIiAS,2J. S C H S E R B , '
~ Y D F.
WISOGROS3
D e p a r t m e n t of P h y s i c a l Chemistry, Hebrew Cnzversity, J e r u s a l e m , Israel ReceiLied A-otember 3, 1948
Chattan-ay and Hoyle ( 2 ) studied the addition products of halogen and quaternary ammonium salts and found that tetramethylammonium bromide forms a compound n-ith bromine corresponding t o the composition ( C H 3 ) S B r'Brz. The melting point of this compound is 118.S"C. These authors, and later Bon-en and Barnes ( l ) ,referred also to the existence of another addition compound, but its composition was not determined n-ith sufficient accuracy. This paper presents the phase diagram of the tn-o-component system tetramethylammonium bromide-bromine and the solubilities of the components in water at 23°C. E XPERIJIELUTAL
Xaterials The bromine used n-as a product of Palestine Potash Ltd., containing 0.2 per cent chlorine. It n-as purified by filtration through a glass cloth, then boiled n-ith a potassium bromide solution (20 per cent by weight) for 20 hr., and washed bromide-free v i t h distilled n-ater. After separation from the water the bromine was fractionated in an all-glass apparatus by distillation and finally dried over phosphorus pent oxide. The tetramethylammonium bromide n-as prepared by condensation of methyl bromide and trimethylamine in an alcoholic solution. Palestine Potash L t d . , Jerusalem, Israel. Department of Physical Chemistry, Hebrew University, Jerusalem, Israel. 3 This paper is part of a thesis submitted hy F. Winogron t o the Senate of the Hebrew Cniversity, Jerusalem, in partial fulfillment of the requirements for the degree of Doctor of Philosophj . l