ANODE CORROSION OF LEAD IN SODIUM HYDROXIDE SOLUTIONS BY 0. W. BROWN, C. 0. HENKE AND
L. T. SMITH
Lead, used as anode in a sodium hydroxide solution, is dissolved and precipitated on the cathode as a sponge, in a very finely divided condition. Glaserl has obtained sponge lead by the electrolysis of feebly acid or neutral solutions of lead nitrate and acetate. However, with a more strongly acid solution of the nitrate or acetate, he obtained brilliant crystalline deposits of metallic lead. The formation of the sponge lead was attributed to the basic salts, which are almost insoluble, and which impede the uniform crystallization of the metal at the cathode. Elbs and Rixon2have also made a study of the electrolysis of lead salts. In their experiments on phosphorous acid with lead electrodes, they observed that, with small current densities, the anode goes into solution and at the same time remains bright, and a crystalline deposit of lead forms a t the cathode. If the current density is increased, the polarization potential increases from a few tenths to about 1.9 volts, and a t the same time the anode becomes covered with a dark gray layer, while the deposit of lead on the cathode becomes spongy. They state that lead is dissolved in the electrolyte forming plumbic salts and on electrolysis sponge lead is formed at the cathode. If plumbic salts had not been present the deposit would have been crystalline. Tommasi3 produced sponge lead by the electrolysis of a solution of the double acetates of lead and sodium, using lead anodes and a rotating disk cathode. Elbs and Forssel14 have investigated the behavior of lead as an anode in sodium hydroxide solutions. They state that Glaser: Zeit. Elektrochemie, 7, 365 (1900). Elbs and Rixon: Ibid., 9 , 267 (1903). 3 Tommasi: Electrochem. Ind., I, 498 (1902). 4 Elbs and Forssell: Zeit. Elektrochemie, 8,760 (1902). 2
368
0. W . Brown, C. 0. Henke and L. T . Smith
a t first the lead anode remains bright and does not gas. Then there appear dark gray spots which spread out over the whole anode, later brown spots appear and soon the anode is covered with a brown coating. A t the same time the anode begins to gas and yellow particles fall off the anode, which soon ceases, and then only gas is given off. These phenomena do not change with increasing concentration of the sodium hydroxide or with increasing current density, except that the single phases cannot be distinguished. They also state that the electrolytic corrosion of the anode is not essentially influenced by the concentration. With increasing temperature the formation of lead peroxide only begins later. High current density and a quiet electrolyte are said to favor the formation of lead peroxide. In our experiments we found that the temperature, concentration of the sodium hydroxide solution and the current density affected very greatly the corrosion of the anode. The lead which we used as anode was purified twice by electrolysis in a Betts bath. The purified lead was then melted and cast into anodes. A new anode was used for each electrolysis so as to have the condition of the anode the same a t the beginning of each experiment. The anodes in the experiments given in the first three tables had a surface of 56 square centimeters, while the anodes used in all the other experiments had a surface of 64 square centimeters, counting both sides. The anode was hung on a glass rod in the middle of the cell, which was a beaker 9 cm in diameter by 12 cm high. The cathode consisted of two thin strips of lead, of the same size as the anode, placed against the sides of the beaker, one on each side of the anode. Enough electrolyte was used to cover the anode, which required about 550 cc. The cells were kept in water baths and within 2' of the indicated temperature. The current was measured by a copper coulometer the current strength being indicated by an ammeter. The quantity of lead corroded was determined by the loss in weight of the anode. Theoretically one ampere hour should dissolve 3.8642 grams of lead.
Anode Corrosion of Lead, Etc.
369
TABLE I Effect of concentration of sodium hydroxide solution on anode corrosion at 20' C. Current strength-1.5 to I .6 amperes. Electrode tension-3 to 3.1 volts after electrolysis had been started for some time. Concent ration Grams NaOHl per liter
Current per sq. dcm in amperes
Ampere hours passed
Lead corroded in grams
Corrosion in percent of theory
38
2.68
12.256
1.93 1.76
43
2.86
0.9102 0.8337
2.587
0.0550
152
2.68
12.256
I74
2.86
2.587
304
2.68
12.256
348
2.86
2.587
0.55 0.93 4.95 4.46 0.93 2.36 I .28 1.45 0.81 0.49
0.0930 2.5430 2.1448 0.0926 0.2356 0.6073 0.6873 0.0814 0.0487
The two figures given for each concentration in each of the last two columns are results from two runs that were carried out under the same conditions. It will be noted that these do not check. Thus in the two runs with a concentration of 174 grams per liter the one result is 2.16 percent, over 21/2 times the other result which is 1.03 percent. Apparently they should be the same but the explanation is apparent upon watching the behavior of the lead anode. When the current is first started there is no gas given off and the lead dissolves quantitatively. Suddenly a gray spot will appear which will rapidly spread over the anode. Then dark brown spots of PbOz appear which likewise spread over the surface of the anode. At the same time gas is liberated and the corrosion of the anode is very materially reduced. The results in Table I show that a t this temperature and current density the corrosion of the lead anode is poor a t best, 1
The sodium hydroxide used contained a trace of chlorine.
370
0. W . Brown, C. 0. Henke and L. T . Smith
the greatest amount of corrosion taking place in a sodium hydroxide solution containing 152 grams per liter. TABLE I1 Effect of concentration of sodium hydroxide solution on anode corrosion a t 60' C. Current strength-1.4 amperes. Electrode tension-about I volt at start. Concentration Grams NaOH per liter
Current per sq. dcm in amperes
Ampere
hours
passed
Lead corroded in grams
Corrosion in percent of theory I
35
2.50
3.628
71
2
.50
3.628
I IO
2.50
152
2.50
3.628
192
2.50
3.824
220
2.50
3.824
304
2.50
3.628
'
3.824
0.6271 0.9975 7.7483 6.0721 14.5169 13.9598 13.7798 13.7854 8.8748 8.4815 9.2573 8.9166 5,2875
4.47 7.12
55 * 28 43.32 98.27 94.50 98.32 98.36 60.08 57.41 62.67 60.36 37 * 73
The increase in temperature very materially increased the corrosion at the anode. Also the effect of concentration is quite marked, the corrosion varying from 98.3 percent where we have 152 grams of sodium hydroxide per liter to 37.7 percent where the concentration is 304 grams per liter and 6 . 4 percent where the concentration is 35 grams per liter. This is directly at variance with the statement of Elbs and Forssell that the corrosion of the anode is not essentially influenced by the concentration of the sodium hydroxide solution. It will be noted that when the corrosion is near IOO percent, the duplicate runs check very well, while when the corrosion is low as in the run where the concentration is 304 grams per liter the duplicate runs do not check within I O percent. T-he results in Tables I, I1 and I11 are plotted in Plate I. The curves show the effect of concentration on the anode corrosion at the temperatures indicated. In each case we have
Anode Corrosion of Lead, Etc.
371
TABLE I11 Effect of concentration of sodium hydroxide solution on anode corrosion a t 75 O C. Current strength-1.4 t o I .5 amperes. Concentration Grams NaOH per liter
Current per sg. dcm in amperes
Electrode tension in volts
Ampere hours passed
Start
End
2.68
1.2
3.2
2.50
.o .o I .o I .o
0.3
2.50 2.50 2.68 2.50 2.68
I I
I
.o
0.2
0.2
0.3 0.2
.o I .o
0.3
-
-
I
0.2
'0
I1 I
Lead corroded ingrams
Corrosion in percent of theory
3.3971 2.7538 I3.9149 13.8476 13.8825 13.8304 13.9085 29.5476 13.9064 I3 * 8940 13.3050 16.4268
11.6 9.4 99.95 99.46 99.71 99.34 99.9 100.8
30
99.88
99.80 45 40 56.06
372
0. W . Brown, C. 0. Henke and L. T . Smith
but little corrosion at low and high concentrations. But at intermediate concentrations, that is from about IIO to 170 grams sodium hydroxide per liter, we have theoretical corrosion of the anode a t 75' C, and nearly theoretical corrosion a t 60" C. At 20' C the corrosion is very low a t all concentrations, though even here it is highest at a concentration of 152 grams per liter. In the foregoing experiments the current density, although not always the same, was usually 2.5 amperes per sq. dcm though in some cases it was 2.68 and 2.86 amperes per sq. dcm. For the following experiments the concentrations 38, 152 and 304 grams sodium hydroxide per liter were chosen and the current density was varied a t each temperature and concentration. Also a few runs were made at 85' C. The results are given in Table IV. It will be noted that with a concentration of 38 grams per liter and a current density of 0.16ampere per sq. dcm. the corrosion was 102.5 percent. This indicated that the lead was soluble in this strength of caustic soda. To prove this an anode of the same size as those used in Table IV was placed in a solution of sodium hydroxide containing 38 grams per liter. After standing, without current passing, a t 20° C for an hour and a half, which was the length of time that the electrolysis was carried out, the anode had lost 0.0114gram, which was equivalent to 1.92percent in the case mentioned. When the anode corroded quantitatively it remained bright, but when it became covered with a coating Qf oxide it did not corrode quantitatively. However, in Table IV, with a concentration of 304, a t 29' and a current density of I , it was noticed that the anode was bright although the corrosion was only 4.8 percent of theory. The same was true with the same concentration and a current density of 4 at 60' and 75" when the anode corrosion was only 6.3and 26.7percent of theory. This does not seem to agree with the statement of Elbs and Forssell that as long as the anode remains bright it corrodes quantitatively. The results in Table IV are plotted in Plates I1 and 111,
Anode Corrosion of Lead, Etc.
Concen- "emperatration Gm NaOH tFre c per liter
Current per sq. dcm in amperes
Ampere
hours passed
-38 38 38 152 304 304 304 38 38 38 152 1.52
152 304 304 304 152 152
60 75
0.16 0.16 0.16 0.16 0.16 0.16 0.16
20
1.0
60 75
1.0 1.0 1.0
20
60 75 20 20
20
60 75
.
1.0'
1.0
20
1.0
60 75 60 75
1.0 1.0
152 304 304 304 152
85
152
85
60 75 85
75
4.0
4.0 4.0 4.0 4.0 4.0 6.6 6.6
.
Electrode tension in volts Start
End
1.2
0.84 0.14 0.14 0.94
373
I . ::~zj:
Lead 'OrCorroded In grams of theory
-0.1533 0.1533 0.1533 0.1533 0.1533 0.1533 0.1533 0.867 0.8714 0.8714 0.8714 0.8714 0.8714 0.867 0.867 0.867 3.3698 3.9003 3.9003 3.3698 3.3698 3.9003 6.3619 6.3619
0.9 0.9 1.0 1.0
0.8
1.0 0.2
0.8
0.24
1.6 1.2
-
0.6
1.3
0.2
1.2
3.0 0.4
1.1
1.3 1.4
0.2
-
1.1
0.5
1.2
0.24
1.5 1.3 1.4 1.4 1.3 1.4 1.35 1.4
-
3.3 3.3
0.6071 0.6047 0.6067 0.5651 0.4358 0.5928 0.5968 0.4391 3.3193 3.3964 0.5415 3,3719 3.3758 0.1601 3.2517 3.3259 0.8507 7.0840 14.9453 0.8227 3.4707 6.4122 2.3888 9.1647
102.5 102.1
102.4
100.4 73.6 100.1
100.7 13.1 98.8 100.8 16.1 100.2 100.2
4.8 97.1 99.3 6.6 47.0 99.2 6.3 26.7 42.6 9.7 37.3
374
0. W . Brown, C. 0. Heizke and
L. T . Smith
T€hfP€RATUR€ /N DECREES C€N?/GRAbE
Fig.
2
Anode Corrosion of Lead, Etc.
3 75
theoretical in nearly every concentration a t any temperature. This table shows how high a current density can be used under any of the given conditions of concentrationand temperature with an anode efficiency of IOO percent. Thus in Curves A, B, C and D, it is seen that with each increase in temperature a rapidly increasing increase in current density is possible without the anode corrosion dropping. The same is also shown by Curves E and F for a concentration of 304 grams per liter and by Curves G and H for 38 grams per liter. The plate also shows the effect of concentration. Thus a t 75' the Curves H, F and C show that a higher current density is permissible at 152 grams sodium hydroxide per liter than at either of the other two concentrations. In Plate IV the current density is plotted against the temperature. The concentra3 tion is 152 grams per liter. The points on this curve represent the highest current density that was used a t the indicated temperatures with ' IOO percent anode corrosion. 0 Thus a t 75 roo percent corro- OO 7agpmh~s cen~~roc,e 80 sion was secured with a curFig. 4 rent density of 2.68, while, when this was increased to 4 the corrosion decreased t o 47 percent. Hence the current density 2.68 is the one used in the curve. Interpreting the graph we see that at any point in Area I less than IOO percent corrosion is secured, while a t all points in Area I1 the corrosion is IOO percent of theory. In Table I the electrode tension was about 3 volts after the electrolysis had been started for some time, while in Table I1 the electrode tension was only about one volt at the beginning of the electrolysis. In Table I11 the electrode tension was measured a t both the start and end of the electrolysis At the start it was about one volt while a t the end it was only 0 . 2 to 0 . 3 volt except where the concentration is 38 grams O,
376
0.W . Brown, C. 0.Henke and L. T . Smith
per liter and here it is 3.2 volts at the end. On examining the figures in the corrosion column of Table 111 it is seen that where we have IOO percent corrosion the bath tension is low, 0.2 and 0.3 volt, while a t 38 grams per liter where we have low corrosion the voltage is high a t the end, being 3.2 volts. A similar condition is found on examining Table IV. In order to examine this further a study of the discharge potentials was made. To get the discharge potentials of each electrode a normal calomel electrode was used as auxiliary, the electromotive force of which was taken as -0.56 volt. The discharge potentials were measured by means of a specially designed make and break switch, which interrupted the current and then instantly made contact with the normal electrode circuit, in which the pressure to be measured was balanced against a pressure of known value, a galvanometer being used as the zero instrument. The results are given in Table V. No measurements were taken until after the current had passed 70 minutes, because on first starting the current hydrogen is liberated at the cathode and hence the voltage is higher than a t a later time when the lead is being deposited. Each time after increasing the current the readings were made as rapidly as possible which was about one minute after increasing the current. In the last column headed voltmeter is the reading of the voltmeter immediately after breaking the circuit. The results in Table V are plotted in Plate V. The readings at the end of each period being the ones used. All three curves make a very sudden bend at 2.81 to 3.75 amperes per sq. dcm. This is the point a t which the anode quits corroding quantitatively. The change in the anode discharge potential a t this point is very marked, from +0.285 to -1.317, a change of 1.602 volts. The totalpolarization increases from -0.146 to -2.464, an increase of 2.318 volts. Curve C of Plate 111 shows that up to 2.68 amperes per square decimeter the corrosion is IOO per cent, but when this is increased to 4 amperes the corrosion drops to less than 50
A n o d e Corrosiovl of Lead, Etc.
377
TABLE V Discharge potentials and total polarization at different current densities. Temperature-75 O C. Electrolyte-550 cc of a solution containing 152 grams of sodium hydroxide )er liter. Minutes after start a t which measurement was made
Current per s q dcm in amperes
Blectrode tension in volts
Discharge potentials in volts Anode
1
t
Total polarization
Voltmeter
Cathode
1
70 85 86 IO0
IO1 1I5 I 16'
1.56 1.56
0.5
2.10 2.10 2.81 2.81
0.6
'49
3-75 3.75
146
4.22
0.42
0.58 0.68 0.75 0.95 3.6 4.0
+O.292 t0.292 +0.285 +0.292 +0.285 +0.285
+0.277 .317 -I .317
-I
.
+0.471
-0.179
+0.3g2
--o.IOO
0.09
+0.457 +0.400
-0.165 -0.107
0.12
+0.442 +0.428 4-1.002
-0.150
-0.146
0.12 0.12
-0.788
0.14
-2.464 -2.484
2.0
+I.
104
+I.
167
AMP€R€S PER SQ.dcm.
Fig. 5 1
Oxides of lead started t o appear a t the anode.
" Anode was gassing vigorously.
0.1
0.1
1.3
378
0. W , Brown, C. 0. Henke and L. T . Smith
percent. Comparing this with Plate V it will be seen that when the anode is corroding properly the discharge potential is low, about 0.2 volt, but, when it stops corroding quantitatively its discharge potential immediately rises to about -1.3 I7 Volts. The voltmeter reading immediately after breaking the circuit is in each case a little lower than the total polarization due to the fact that it cannot be read quick enough. Yet the change in voltmeter reading is over one volt, being from 0 . 1 2 to 1.3, a change of 1-18 volt. Hence the corrosion of the anode can be carefully watched by merely noting the voltmeter reading immediately after interrupting the current. Laboratory of Physical Chemistry Indiana University B loontingtoiz
.
