P
c
REVIOUS studies (3) have confirmed that metallic copper is a t least as effective as is cuprous oxide when employed as the toxic ingredient in short-oil antifouling paints. There is, however, considerable reluctance on the part of many users of marine bottom paints to approve metallic copper in antifouling compositions intended for application to steel hulls. This prejudice arises from the apparent belief that metallic copper paints will accelerate the corrosion of steel by simple “couple” action. Such accelerating action, if proved to take place, would indeed introduce serious limitations on the use of metallic copper in bottom paints. Accordingly, a controlled study was undertaken of a variety of commercial and experimental antifouling paints based on metallic copper, cuprous oxide, and mixtures of these with diverse other pigments. The investigation involved sea water immersion of numerous coated medium-carbon steel panels with and without the usual intermediate primer and anticorrosive coats, and with both intact paint films and films which were deliberately broken through to bare metal by scribing with a sharpened scalpel. These experiments prove conclusively that there can be accelerated attack on the basis metal, when coated directly with antifouling paints carrying either metallic copper, or the usual cuprous oxide toxics. It is important to note that the attack was actually as severe under the cuprous oxide paints as it was under the metallic copper paints in these tests. On the other hand, no marked accelerating action was found where the antifouling paints were applied over even a single primer coat, unless the toxic content was inordinately high.
ANTIFOULING PAINTS G. H . Y o u n g , G . W . Seagren, W . IC. Schneider, and J . C . Zehner MELLON INSTITUTE, PITTSBURGH, PA.
Quantitative measurement of the comparative rates at which pre-scribed painted steel panels corrode when immersed in sea water show that there can be accelerated attack on the exposed steel. The intensity of attack appears to vary directly with the concentration of copper pigment in the antifouling coat, and may be as severe with cuprous oxide paints as with metallic copper paints. The rate of such accelerated attack decreases materially if even a single barrier coat is interposed between the steel surface and the antifouling paint, and is substantially eliminated if multiple barrier coats are employed. The studies are being continued.
Accelerated Corrosion of Steel Coated with Heavy Metal Antifouling Compositions This appeared to be true whether the primer carried “inhibitive” pigment, such as zinc yellow, or a noninhibitive pigment, such as iron oxide. Even pre-scribed panels with bare metal thus initially exposed showed no deeper scribe etching than did the controls carrying no antifouling paint. Whether the “protection” thus afforded by barrier undercoats is operative when there are finite areas of metal exposed, remains to be demonstrated. Quantitative data will be reported when the experimental study of this point is completed. CORROSION OF STEEL COUPLED TO COPPER
By accepted definition, a “couple” consists of two or more
i
metals having dissimilar solution potentials, in metal-to-metal contact with one another. When such an assembly is exposed to the action of a corrosive medium, there is usually accelerated solution of the more anodic member of the series, and “cathodic protection” (more or less complete) of the member or members having the lower solution potential. The phenomenon has been investigated widely, and is one of the strongest pieces of evidence validating the electrochemical theory of corrosion reactions ( 1 ) . Evans and others ( I ) have pointed out that the corrosion of iron coupled to copper is, in most media, cathodically controlled; that is, the intensity of the attack is primarily a function of the cathode (copper) area. That this is true for sea water corrosion of this couple was conclusively demonstrated at Kure Beach, N. C., by LaQue (9), who obtained a constant acceleration factorbf 2 when iron was coupled to copper in area ratios varying from 1/6 to 6/1. Furthermore, under these experimental conditions the attack was not localized; there was equally accelerated sacrifice of the steel 72 inches removed from the dissimilar-metal junction. Certain portions of these experiments were repeated a t the Kure Beach exposure site, and it was again conclusively demonstrated that medium-carbon steel coupled with metallic copper is corroded a t a rate of 2 to 2.6 times as fast as the same steel uncoupled. I n this case equal areas of the two metals were involved.
These experiments are important, for they show that the couple acceleration factor is a t least 2, over a thirty-six-fold range of relative anode-cathode areas. If, then, an antifouling paint film carrying metallic copper powder as pigment behaves as though it were a continuous sheath of copper coupled td the underlying steel, we should expect to find severe, uniform sacrifice of the steel a t any breaks in the sheath, a t a rate approximately twice that a t which the steel would normally corrode. If, on the other hand, the paint film does not behave as a continuous conductive sheath, any acceleration a t breaks or bare spots must be attributed to purely local action a t the break, and the rate of attack may vary unpredictably. CORROSION OF STEEL COATED WITH METALLIC COPPER PAINTS
I n a preliminary series of experiments, 6 X 12 X inch medium-carbon steel panels were freshly sandblasted before painting. One set of panels was given a single primin@;coat of a 25gallon tung-phenolic varnish pigmented a t a 2/1 pigment/binder ratio with 90% zinc tetroxy chromate (ZTO) and 10% asbestine. The second set was unprimed. Both were then painted with a single coat of an experimental copper-containing antifouling paint, AF-2, previously described (3). For control purposes a top coat, comprising an inert pigment (barytes) in the same vehicle as that carrying the copper, was used over both bare and primed steel duplicates. The panels were cross-scribed along the diag9nals on both front and back faces, and immersed for 8 months a t the North Florida Test Service marine site a t Daytona Beach, Fla. They were then returned, the paint was carefully removed, and accurate measurements were made of the depth of etching of the scribes, as well as of the average depth of underfilm pits. For this purpose a No. 77 Ames dial was used, which is accurate to 0.5 mil and fitted with a needle point probe. Readings were taken a t every inch of scribe length, for a total of 48-52 readings per panel. The 341
*
I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
342
Vol. 36, No. 4
resulting data ( Table I) show that there was more severe attack on the panels carrying this copper paint. The acceleration factor does not, however, even closely approach the value of 2 to be anticipated if simple couple action is all that is involved. I n a following series of experiments the effect of varying the concentration of toxic ingredient in the same short-oil cumar vehicle was explored. The experimental formulas involved are listed in Table 11, together with data on scribe and pit depths after 4-month immersion at Daytona Beach. As in the preliminary study, freshly sandblasted 6 X 12 X '/* inch steel panels were employed. The panels were prepared in duplicate, with and without an undercoat of a phenolic-tung oil primer pigmented with a commercial zinc yellow, and were scribed along the diagonals before exposure in the manner previously described. The appearance of the panels carrying the highest toxic loading for each typeof formulation, with and without the undercoat, at the end of 4-month exposure, is shown in Figure 1. After removal of the residual paint films, the panels were again photographed and are shown in Figure 2. The deep etching of the scribe lines shows clearly, as does the fact that the attack (even on the scribes) is nonuniform. Although the data from these two series of experiments cannot be compared directly, inasmuch &s the test formulations and the exposure times were different, it is apparent that there are appreciable differences among the formulations of this second series. I n general, the higher the toxic concentration in the antifouling paint, the more severe is the attack at exposed areas of underlying steel, With only one exception, however (panel 269), the distribution and severity of pitting away from the scribe marks appear to be substantially Figure 1. Panels 233, 265, and 256 over Primed Steel (above); Panels 259, 268, and 262 over Bare Steel (below) independent of the toxic concentration and even of its type. The data show that, over unprimed steel, the severity of scribe etching is approxitually significant. This arises from the very high magnitude of mately proportional to the weight per cent of copper, calculated the deviations in the scribe depth measurements, resulting from as such, in the overlying paint film (Figure 3). the highly nonuniform attack on the scribe line; many of the This relation is obscured when the antifouling paint is applied scribes are deeply etched for only short distances along the line over a primer. The data in Table I1 were tested by the usual and are substantially unattacked at other places. statistical methods; from the results it is extremely doubtful Statistical analysis shows the difference between panels 253 whether the apparent trend shown by the primed panels is acand 254 to be questionable, that between 253 and 255 to be real. Similarly, the difference between 256 and 258 is real; the difference between 265 and 267 is not, however, significant. TABLE I ACCELERATED CORROSION UNDER COPPERPAINT Thus it cannot be stated positively that the concentration effect Panel Av. Scribe Av. Pit Depth, Mils Depth, Mils Paint Pigment Primed observed on unprimed panels is actually inoperative over one 20.5 * 2 . 5 7.5 * 1 . 7 Y es 60/40 Cu/harytes coat of primer, a t least. 14.4 * 3 . 6 6.7 A 1.0 Yes Barytes 31.5 * 3 3 8.6 * 2 7 NO 60/40 Cu/barytes 26.8 * 3 . 5 5.0 * 1.6 NO Barytes
TABLE 11. EFFECTOF TOXIC CONCENTRATION O N ACCELERATED CORROSION
+
-
4 gallons linseed oil 65%; mineral spirita, 35%) (Vehicle: 100 pounds cumar V-3 -I-4 gallons tung oil, 7 Panels Primed -, Panels Unprimed-Wt. % Pit depth, mils cu on panel Scribe depth, mils panel Scribe depth, mils Pit depth, m i l s Paint Pigmentation Amount Solids KO. Av. Max. Min. Av. ApTearanoe No. Av. Max. Min. Av. Appearance C.P. Cu powder 12 lb./gal. 68.0 253 5.2 22 0 0 Vs. etchingb 259 18.0 35 0 10.0 Deep 3.3 14 0 2.0 254 Shallow 260 5.2 15 0 2.5 Shallow 6 Ih./gal. 49.0 2.1 14 0 0 14.5 255 Vs. etching 261 9 0 5.0 Medium 1 lb./gal. 0.7 P . D . Cu pasten 12lb./gal. 55.0 265 4.4 21 0 4.0 Shallow 268 14.6 25 0 8.6 Deep 5.3 24 0 3.7 266 Shallow 269 9.4 18 1 16.8 V.deep 6 lh./gal. 38.5 2.7 12 0 4.0 Shallow 3.6 10 0 4.5 Medium 12.0 267 270 1 lb./gal. 4.4 21 0 0 Deep 38.0 256 Vs. etching 262 25 0 10.2 7.2 (50-50 CuaO6/1 11 0 0 Vs.etching 263 7.5 Deep 1.4 5.2 16 0 33.5 257 ZnO)/P.D. 1.7 8.0 Deep 8 0 3.0 Shallow 264 4.5 11 0 22.0 258 a P . D . Cu paste is a proprietary metallic copper pigment manufactured by Phelps Dodge Corporation. b Vs. = very slight.
3;
April, 1944
INDUSTRIAL AND ENGINEERING CHEMISTRY
343
TABLE 111. ACCELERATED CORROSION AND COMPARATIVE ANTIFOULING EFFICIENCIES UNDER VARIOUSCOMMERCIAL BOTTOMPAINTS Panel No.
233 240 237
236 234 235 239 238
*
Pigmentation Control (barytes only) Cue0 inert Cu &der Cu flake Cu powder inert Cu owder: CUZO,HgO, inert c u r 8 inert c~*o: Z n o , inert
Corrosion of Unprimed Panels Scribe depth, mils Pit depth, milsa Av. Max. Min. Av. Max. Min. 1 0 0.0 0 0 2.6 5.2
7.1
8.5
0
0
20
12.4 13.6 24.8
Measured on 10 of the most obvious pits only.
14 16
28 24
28
32 b
0
.
0 0 0 16
P. * primed; U.
5.1
10.3
12.8 4.5 9.2
-
CORROSION O F STEEL COATED WITH CUPROUS OXIDE PAINTS
Table TI shows that accelerated corrosion of unprimed steel is possible under paints which do not contain metallic copper. Thus, panels 256-8, and 262 -4 employed red cuprous oxide as toxicant. The weight per cent of copper (calculated as such), even for the highest pigment loading, is sufficiently below that of the highest copper-containing paints to account for the observed lower average scribe depth on panel 262 in comparison with 259 and 268. Thus, Figure 3 shows rather satisfactory agreement of the data on cuprous oxide with those for metallic copper when the oxide is plotted in term8 of copper content. Accordingly, the corrosion-accelerating effect of several representative commercial antifouling paints containing cuprous oxide as toxicant were investigated. There were included a number of commercial bottom paints based on metallic copper and Figure 2.
8.0 3.0
12 20
37
9 15
28 5
Over-all Fouling Rating (10 = Perfect)b After 1 mo. After 3 mo. After 4 mo. P. u. P. u. P. u. 7 4 5 4 9 8 5 5 10 7 9 8 10 10 8 9 10 10 10 10 10 10 10 9 10 10 8 6 10 10 8 7
unprimed.
on a mixture of copper, cuprous oxide, and mercuric oxide. No information is available on t h e actual pigment and vehicle compositions; t h e y probably vary widely, however, because their antifouling characteristics are known to vary over quite a broad range,
701
0
X
0
0
0
0 0
Ob
2
4
6 8 IO I2 14 4%Scribe Dsprtl (mils)
16
Figure 3. Scribe Depth after 4-Month Immersion vs. Copper Content of Overlying Paint
Panels of Figure 1 after Removal of Residual Paint
The paints were applied direotly to freshly sandblasted steel panels, diagonally scribed in the usual manner and immersed 4 months at Daytona Beach. At the same time a duplicate set of panels was immersed in which the steel was first coated with a standard primer and an anticorrosive paint before antifouling paints were applied. The duplicate set was used for evaluating the comparative antifouling efficiencies of the paints. Table I11 lists scribe and pit depths, arranged in order of increasing severity of attack, and compares antifouling efficiencies over both primed and unprimed steel during the same exposure period. Fouling ratings are on a 0-10 scale, with 10 being perfect (3). The data in Table I11 confirm that there is a t least as great a likelihood of accelerated corrosion under cuprous oxide bottom paints as under metallic copper paints when applied to bare steel. Actually, within this group of commercial formulations, cuprous oxide paints show a more deleterious action than do copper paints. The antifouling efficiency of these formulations is rapidly lost if they are applied over bare steel. Within 3 months they were all as badly fouled as the control panel which carried no antifouling paint. Since a t least five of the paints have demonstrated satisfactory antifouling action for 6-7 months when applied in the conventional manner over the usual primer and anticorrosive coats, it seems established that their immediate proximity to corroding steel is responsible for inactivation of the toxicants. It is important to note that this inactivation did not manifest itself until after the first month of exposure. This fact throws doubt on simple “coupling” with the paint film as the causative factor, because all fouling exposures involving unpainted couples of copper and steel show immediate fouling of the ‘‘cathodically protected” copper (& 4 ) .
INDUSTRIAL AND ENGINEERING CHEMISTRY
344
Vol. 36, No. 4
cuprous oxide-zinc oxide paint, and a control pigmented only with barytes were used. The formulations are described in FormulaCompn.* a‘t. % Table IV. The panel preparation schedule is shown in Table V, tion No. Pigmentation Vehicle Pigment Vehicle together with the inspection results after 4-month immersion a t a Zn dust, ZnO 25- a1 1/1 tung/linseed 71.5 28.5 ptenolio varnish Daytona Beach. b Znasbestine yellow, TiOz, Same 67.0 33.0 The presence of tIyo barrier coats separating the antifouling c Iron oxide Same 40.0 60.0 paiht from the metal surface appears to have substantially elimd Blue lead, ZnO, 33- a1 dehvdratedcastor 6 0 . 0 40..0 inated any accelerated corrosion a t the scribe marks, although asbestine, celite pf~enolic~varnish e Red lead 25-ga1. tung phenolic 7 0 . 0 30.0 the steel was initially laid bare a t these marks as in the previous varnish I C u powder, barytes 8-gal. 1/1 t m d 1 i n - d 38.5 Cu content experiments. Although the relation is not clear-cut, it seems oumar varnish likely that, as the thickness of the barrier layer is increased, there CUZO,ZnO Same 38.0 cu content Barytes Same 75.0 23.0 is a rapid decrease in the tendency t o accelerate corrosion localls a t exDosed areas of steel substrate. These findings confirm best shipyard practiceTABLE Jr. FOULIXG US. C O R R O S I O X AFTER MONTH IMI\fERSION namely, the use of multiple barrier coats, or of Antivery heavy undercoats, beneath antifouling paints. 2nd fouling Over-all Panel 1st Corrosion Coat Coat Fouling NO. Coat Scribe marksa P i t ratingb They further suggest that thick, relatively impera C 241 Slight 6 s-m meable barrier coats, topped by medium coats of a C Slight; deep a t one spot 6+ m 242 antifouling paints, should give better corrosion prob C 243 4 f m Slight but varying b C Occasional deep sections 7f s 244 tection coupled with efficient antifouling action C C Slight b u t varying 4 m 245 C c Slight 4 m 246 than can be had with light priming coats and a C C 247 Slight 5+ heavy antifouling layer. This point is currently C Slight 248 C IO 249 Vs., few small pits 9+ OS f under investigation. d C 9 250 Slight 9 us
TABLE ITT. PAINT COMPOSITIONS
a
---
:
251 252
.
e
e
C C
r”
8 9
Fairly uniformly etched Fairly uniformly etched
3 5
s m-d
LITERATURE CITED
(1) Evans, U. R., “Metallic Corrosion, Passivity and 0 Average depths were less than 1 mil, with spotty variations from 0 t o 5 mils. There v,as Protection”, pp. 513-35, London, Edward Arnold no significant relation between composition and depth of attack. b There were too few pits to allow valid measurements: ratings are on a 0-10 scale, with & Co., 1937; LaQue and Cox, Proc. Am. SOC. depth indicated as s (shallow), m (medium), and d (deep). Testing Materials, 40, 670-89 (1940) ; Wesley, Z b i d . 40, 690-io4 (1940). (2) LaQue, F. L., Gibson Island Symposium on Corrosion, 1912; private communications. INFLUEYCE O F INTERMEDIATE BARRIER COATS (3) Young, G . H., and eo-workers, IND.ENG.CHEM., 35, 432. 436 (1943). A4notherseries of saribed test panels was exposed which carried (4) young, G , H., and co-workers, unpublidlled data. ,
a variety of undercoat permutations involving five different barrier paint,s For the antifouling compositions a copper paint, a
CoxTRIBvTIox from the Multiple Fellowship on protective coatings of Stoner-Mudge, I ~ c . ,a t ~ \ i e i i o nInstitute.
Carotenoids in Corn By- products W. Baumgarten, J. C. Bauernfeind C. S. Boruff H I R A M WALKER & SONS, INC., PEORIA, ILL.
N ANIMAL nutrition certain carosenoids serve as precursors of vitamin -4;others not possessing vitamin activity are mainly of interest because of their pigmenting characteristics. Steenbock (16) noted that vitamin potency was associated with the yellow color of plants, possibly carotene, and demonstrated that yellow corn was a better source of provitamin A than white corn. At present (14) nine or more pigments are reported to possess vitamin A activity. Previously i t was reported (3, 10,11) that the carotenoid pigments present in egg yolk and in the body fat of the fowl consist almost entirely of xanthophylls with small amounts of cryptoxanthol and carotene, whereas cattle deposit carotene primarily in their body fat and in the fat of their milk. Corn distillers’ by-products are used primarily in dairy rations (7) as desirable ingredients, high in total digestible nutrients, although during the past few years they have been recognized as significant sources of the water-soluble vitamins in the rations of poultry and swine (6, 16). These corn distillers’
by-products are made from yeast-fermented grain mixtures in which yellow corn predominates, a grain regarded as an important source of the carotenoids in the feeding of farm animals. I n view of these facts an investigation was initiated to determine the carotenoid content of distillers’ by-products. By-products used in this investigation have been defined by the Association of American Feed Control Officials (1). Briefly i t may be stated that corn distillers’ dried grains with solubles is the total dried stillage containing both the spent grain residues and the thin stillage; corn distillers’ dried grains is the spent grain residues or screenings; corn distillers’ dried solubles, as the name suggests, is the dried thin stillage or solubles. Usually six samples of each product were analyzed, each representing a t least 20 tons or more of the product. Hybrid dent corn (1-2 years old) was analyzed for comparison. EXTRACTION AND CHROMATOGRAPHIC TECHNIQUE
Samples of distillers’ by-products or of yellow corn were continuously extracted in a Soxhlet apparatus with Skellysolve B (65.5-70.5’ C.) in the dark until the effluent solvent was colorless by visual inspection. It was soon discovered that two or three times the amount of carotenoid pigment could be