US2784124A - Method of coating ferrous metal surface and composition therefor - Google Patents

Description

United States. Patent METHOD OF COATING FERROUS METAL SUR- FACE AND COMPOSITION THEREFOR Howard R. Moore, Hatboro, Pa.

No Drawing. Application July 1, 1953, Serial No. 365,565

17 Claims. (Cl. 1486.15) (Granted under Title 35, U. S. Code (1952), see. 266) The invention described herein may be manufactured and used by or for the Government of the United States of America for governmental purposes without the payment of any royalties thereon or therefor.

The invention relates to a method of treating ferrous metal surfaces at normal temperatures to resist corrosion, a composition of matter for use in treating the ferrous metal surfaces at prevailing outdoor temperatures, and a coating formed on the ferrous metal surface by the treatment.

The invention relates specifically to a physical method of simultaneously cleaning, pre-conditioning and reacting ferrous metal surfaces under prevailing weather conditions with alkaline inhibitor solutions capable of forming continuous inorganic coatings derived from a ferrousion-containing film coverng the cleaned surface; an alternate chemical method of cleaning and preconditioning ferrous metal surfaces to produce an identical ferrousion-containing film capable of reacting with the same alkaline inhibitor solutions under varying weather conditions to produce identical coatings held to the ferrous metal or steel substratum by secondary valence forces; the functional role of the two separate alkaline constituents in the inhibitor bath and methods of ascertaining their optimum weight percent ratios, total concentrations, solution temperatures and reaction times; the composition, weight and thickness of the coatings formed under optimum conditions of reaction; the performance characteristics of the coatings in prolonging the time of the exposure of treated, but unpainted, surfaces before corrosion occurs; and lastly, the beneficial effect of these integral coatings in prolonging the durability of subsequently applied primer paints and complete paint systems.

The most outstanding feature of this invention is the unusually high degree of corrosion resistance obtained by reacting the preferred combinations of alkaline inhibitors with properly cleaned and preconditioned ferrous metal surfaces. It must be clearly understood, however, that the desired results are obtained only when the physical and/or chemical cleaning and preconditioning procedures are carried out in such a way as to yield iron or steel surfaces of prescribed electrical potentials. Thus it has been found by several laboratory and practical service trials that adequate preconditioning is obtained only when previously descaied mild steel specimens attain surface potentials of 680 to -700 millivolts, when measured according to the method of La Que and Cox. (F. L. La Que and G. L. Cox Some Observations of the Potentials of Metals and Alloys in Sea Water. Proc. Am. Soc. Testing Materials, vol. 40 p. 670687 (1940)).

' 2 This method records surface potentials developed between metal surfaces and a saturated calomel half cell on instant zero time periods of immersion in a 3% sodium chloride solution.

The foregoing values of 680 to -700 millivolts obtained for satisfactorily preconditioned steel surfaces are in fair agreement with the theoretical value of 681.5 millivolts, representing the sum of the standard electrode potentials for iron and the saturated calomel electrode, -440 and 24l.5 millivolts respectively. (H. H. Uhlig. Corrosion Handbook, pp. 1134 and 1137, New York: John Wiley and Sons, 1948). Furthermore, since molal electrode potentials are obtained only when reversible reactions corresponding to the reaction Fe:Fe++|-2e are established, the existence of an alkaline surface film of ferrous hydroxide having the formula Fe(OH)z, in contact with properly cleaned and preconditioned steel becomes a practical certainty. Several years ago, Whitman, Russell, and Altieri demonstrated the existence of such an alkaline saturated solution of hydrous ferrous oxide in in aqueous media containing the normal quantity of dissolved oxygen for the temperature in question, e. g. 31 ml. per liter at 20 C.

Before my discovery of the preferred blends of alkaline phosphatizing and oxidizing agents capable of reacting completely with the hydrated ferrous oxide film covering pro. erly cleaned iron or steel surfaces to produce a well int grated, homogeneous, coating of hydrated iron phosphates and oxides in different stages of valence, it was customary to attempt to prevent corrosion by adding a wide variety of different alkaline or acidic chemicals to the sand-water slurry used in the physical cleaning process known as wet sandblasting. To expedite this process, the Pangborn Corporation, Hagerstown, Maryland,

manufactures a special type of mixer for the delivery of V Hitherto, prior to the development of the improved inhibitor combinations of this invention, a typical inhibitor formula used by the Department of the Navy in blasting the hulls of ships 'in drydock comprised a 4:1 blend of trisodium phosphate with potassium dichromate, added w. Whitman, R. Russell and v. Altieri, Industrial and Engineering Chemistry, vol. 16, p. 665 (1924).

Patented Mar. 5, 1957' in suflicient quantity to correspond to a 1.6% by Weight solution in the water of the sand-water slurry mixed in the Pangborn machine. This composition is identical with the Protecta-Tin formulation described by British Patent No. 535,670 assigned to the Tin Research Institute. Alkaline chromate combinations of this type were gradually abandoned in favor of 2% by volume solutions of phosphoric acid because of the toxicity of the atomized spray containing hexavalent chromium and the generally unreliable and nonuniform resistance to corrosion obtained by their use.

None of the foregoing combinations was found satisfactory in producing a substantial degree of inhibition in the time interval usually prevailing between the cleaning process and application of organic protective finishes, because of the failure to recognize the necessity of selecting chemicals capable of reacting with the ferrous hy-' droxide film initially formed in the preconditioning process. Such chemicals should be capable of reacting with this film to form insoluble surface conversion coatings integral with the underlying ferrous metal. Such coatings, to be serviceable, must possess a satisfactory degree of resistance to oxidation by ambient air with a wide range of relative humidities, and also resistance to steady and/ or intermittent rainfall over a 24 hour period at least. Furthermore, the inhibitor solutions must be sufficiently responsive to Wide variations in outside temperatures and varying temperatures of the steel substrate to give an acceptable degree of inhibition under adverse conditions, for example, ambient temperatures and relative humidities in excess of 80 degrees F. and 80%, respectively, and metal surface temperatures in excess of 100 degrees F. due to the sun beating on the structures to be processed.

The reasons for failure of the -4:1 blend of trisodium phosphate with potassium dichromate and the 2% by volume solutions of phosphoric acid previously used to meet the above requirements provide valuable guides in the development of improved formulations. The main coating material formed by both of these reagents was shown to be vivianite, a hydrated ferrous phosphate which has limited stability under the prevailing environmental conditions because of its rapid conversion, in the presence of moist air, to ferric phosphate, ferrous hydroxide, and ferric hydroxide by the reactions:

There is also the possibility that the potassium dichromate ingredient of the T SP--'K2Cr207 inhibitor oxidizes the vivianite initially formed to form a brownish partially bydroxylated ferric phosphate coating similar to beraunite having the formula Fe3(OH)3(PO4)z.2 /2H2O (I. W. Mellor, A Comprehensive Treatise on Inorganic and Theoretical Chemistry, volume XIV, p. 408). Later, in the final step of hosing down the treated surface to remove weathered paint sludge and wet sand deposits, both inhibitors show the same disadvantage of developing a lightly adherent combined deposit of ferric hydroxide (rust) and ferricphosphate dust which must be removed by brooms, brushes, or other mechanical means. Unfortunately, this hosing operation provides practically ideal conditions for simultaneously hydrolyzing and oxidizing surface coatingfilms of poor integrity.

Because of the aforementioned drastic conditions of simultaneously removing weathered paint and inorganic corrosion deposits and finally removing the soil, as well as adherent sand, by hosing with a high ,pressure water stream, the wet sandblasting process had limited application before the preferred blends of dibasic alkali .phosphates with alkali nitrites, as stipulated Iby'this invention, became available. Excessive materials costs and labor charges were involved in the use of either of "the foregoing alkaline chromate or phosphoric acid inhibitors.

In addition to the toxicity hazards accompanying their use, a supplemental quantity of phosphoric acid must be added to the wash water to remove the surface film of rust formed during the initial blasting. This final dosage of dilute phosphoric acid has the inherent disadvantage, as mentioned above, of producing a second thinner coating of hydrated ferrous phosphate subject to rapid hydrolysis and oxidation reactions on exposure to rain before application of the primer paint coating.

As contrasted to the unsatisfactory performance of alkaline and acid inhibitors hitherto used, the mild alkaline compositions prescribed by this invention are so effective in reacting instantly with freshly cleaned ferrous metal surfaces to produce an integrally bound hydrolysis and oxidation resistant coating that the blasted areas can be hosed freely with fresh water, having a pH preferably in excess of 5.0, without any trace of rust formation. It is unnecessary, of course, to add acidic removers or other chemicals to the water used in hosing down partly decomposed paint film residues or other corrosion products. The degree of inhibition is so effective that cleaned, unpainted areas can be exposed overnight to intermittent or steady rain without incidence of corrosion the following day. The usually invisible coatings remain intact after 72 hours exposure to ambient air temperatures of -90 F. and relative humidities of 80 to 97% or after 144 hours exposure to air having temperatures and relative humidities of F. and 65%, respectively. Exposure times of 1 to 2 Weeks have been recorded for steel cleaned in comparatively cool and dry weather (temperature and relative humidities below 70 F. and 50%, respectively) Because of the long times of exposure tolerated by the inorganic surface coatings of this invention before breakdown occurs, brush and/or spray paint application schedules are readily met without interruptions caused by the necessity of removing newly formed rust. For this reason, the wet-sandblasting process, hitherto considered impracticable for large structures, is now being used successfully for cleaning and reconditioning large steel structures, such as ships, bridges, locks and dams for rivers and canals, and sections of prefabricated housing at the site of assembly, As a consequence, the older method of dry sandblasting is being abandoned in favor of wetsandblasting for these and other applications because of the health hazards involved in the inhalation of siliceous dusts, partially decomposed paint residues, toxic pigments, and other refuse widely dispersed in operating dry sandblasting equipment. Hosing is frequently necessary, after dry sandblasting is completed, to remove old coatings and corrosion products redeposited to a large extent on cleaned areas. This causes fresh rusting and the need of corrosion'removers and supplementary chemical treatments to prepare surfaces suitable for painting. Unfortunately, the dry-sandblasting process does not lend itself to useof the integral surface reacting chemicals of this invention, because the surface potential of sandblasted steel instantly attains values of 390 to 4l0 millivolts, characteristic of ferric oxide (Fe203) and ferroso-ferric oxide (1 8304) coatings which form immediately onaccess of air to the cleaned areas. Obviously such surfaces are not adequately preconditioned for re action withthe preferred inhibitors. Furthermore, unless dry sandblasting is conducted in a dry atmosphere, rerusting will occur before the primer paint can be applied.

The only limitation to successful use of myinvention is that mill scale or other continuous FezOs or F6304 coatings originally present on hot or cold rolled steel be removed by dry sandblasting or by chemical descaling before preconditioning is effected by wet sandblasting or iby uninhibited non-oxidizing acids to secure the required surface potential of 680 to 700 millivolts for reaction with the preferred inhibitors. Mill scale removal should 'be accomplished by chemical descaling with the usual inhibitors before fabrication to avoid the necessity of dry sandblasting to secure the required surface'for effective preconditioning at some later time after the organic protective finish has failed. Unfortunately, the sand-water slurry of the wet-sandblasting process does not have sufiicient cutting action to remove mill scale from prefabricated structures, nor is uninhibited acid suificiently strong or hot to remove mill scale completely by spray application.

An object of this invention is to provide physical and chemical methods for treating ferrous metal surfaces at prevailing indoor and outdoor temperatures to resist corrosion, to provide compositions of matter useful in treating ferrous metal surfaces at normal temperatures, and to provide surface conversion coatings on ferrous metal surfaces under varied ambient Weather conditions.

A second object is to provide methods of treating ferrous metal surfaces by the steps of physically or chemically removing iron oxide, rust, and severely weathered organic coatings, simultaneously preconditioning the surface to surface potentials of 680 to -700 millivolts, and contacting the surface with solutions of secondary alkali phosphate and alkali nitrite to produce composite, integrally bound, coatings of vivianite (Fe3(PO4)2.8H2O), hydrous ferrous oxide (FeO.H2O) and hydrous ferric oxides (Fe2O3.XH2O and Fe3O4.XH2O).

A third object is to provide surface conversion coatings capable of improving the durability of primer and top coat organic finishes systems by virtue of their exceptional resistance to hydrolysis and oxidation reactions.

The only immediately applicable physical method of producing integrally bound surface conversion coatings is the Wet-sandblasting process, which provides ideal conditions for reaction since the water soluble inhibitors are able to react instantly with the preconditioned surface before it has a chance to dry.

Unfortunately, cleaning and preconditioning by the physical method of wet sandblasting is not always applicable for the treatment of small parts. In this case, a chemical method of preconditioning is usually more feasible to insure the maintenance of dimensional tolerances.

Chemical preconditioning is based on a process comprising:

(1) immersion or spray application of chemically pure, non-oxidizing uninhibited acids on previously descaled metal surfaces from which paint has been previously removed, (2) thorough rinsing of the work in cold circulating water to remove all traces of cleaning acids used in step 1, and (3) contact of the work while still wet with the same improved inhibitor solution used for Wet sandblasting. It has been found that the composition and performance characteristics of the coatings produced by chemical preconditioning are identical with those obtained by the wet-sandblasting process.

In respect to the physical method of wet sandblasting (a single step process if rinsing to remove sand, weathered paint, and miscellaneous corrosion products can be considered a supplementary operation), it is clear that inhibitors capable of surface reactions on outside structures must function under variable weather conditions, solution temperatures, and reaction times. For example, the temperature of the metal surface contacted may vary from 40 to 120 F. depending upon the time of year, the geographic location, and the degree of exposure to the sun. Also, the reaction temperature of the water slurry carrying the inhibitors may vary from 50 to 85 F, depending on the source of the Water and local weather conditions. The time of contact also Will'vary from /2 to 3 minutes per square foot of surface cleaned, depending on the amount of paint and corrosion products to be removed. Fortunately, the new combinations of inhibitive chemicals of this invention give integral coatings of high degree of corrosion resistance, independent of the surthree-step face temperatures of the steel base and sand-water slurry, and reaction times within the stated limits.

In determining the eligibility of various classes of inorganic chemicals for integral reactions with previously descaled and preconditioned steel or iron surfaces, the three step chemical preconditioning method was found much more adaptable for investigational purposes than the wet-sandblasting procedure. size and cumbersome nature of the wet-sandblasting equip ment preclude the possibility of direct evaluation of the hundreds of tests required to establish the optimum composition of inhibitor for the prescribed conditions of use. For this reason, the inventive steps were accomplished by the three step chemical preconditioning procedure, which fortunately can be controlled to reproduce the varied outdoor conditions under which the wet-sandblasting equipment is operated.

Before making the transition to chemically preconditioned surfaces for convenience in development Work, it was ascertained that the surface potentials of both wet sandblasted and chemically preconditioned steel specimens were identical, within experimental error. In making these tests, inhibitors of all types were omitted from both the sand-water slurry and acid cleaning baths, respectively. To prevent corrosion of the small steel specimens subjected to wet sandblasting in the absence of inhibitors, they were instantly placed, after blasting, in distilled water carrying the normal amount of dissolved oxygen at 20 degrees C., 31 ml. per liter.

Chemical preconditioning to produce the ferrous hydroxide film is preferably carried out with non oxidizing, uninhibited inorganic acids, such as hydrochloric acid, and water soluble organic acids of high dissociation constants, such as mono-, di-, and trichloroacetic acids. Cold dilute sulfuric acid also may be used to remove loose rust deposits prior to establishing a preconditioned surface, providing any mill scale present was previously eliminated by pickling. Hot sulfuric acid solutions cannot be used reliably to produce suitably preconditioned metal surfaces, because of the accelerated action of nascent hydrogen in converting the hexavalent sulfur to reduction products which promote the establishment of a hydrated ferric oxide film, as contrasted to the desired ferrous hydroxide film on rinsing with cold water.

In producing the inhibitive coatings of this invention by the prior step of chemically preconditioning ferrous metal surfaces carrying mill scale and loose surface rust, hot dilute hydrochloric acid is generally used because of its rapid action in dissolving inorganic contaminants and corrosion products with minimum secondary effects due to the reducing action of nascent hydrogen on impurities in the steel such as sulfur. in using this acid, however, great care must be taken in removing the last traces of ferrous and ferric chloride by thorough rinsing with water having a temperature less than degrees F. The pH of the rinse water preferably should be in the range of 5.0 to 8.5 to insure the formation of a highly adherent ferrous hydroxide film as prescribed by Whitman, Russell, and Altieri, op. cit.

The required concentrations of preconditioning acids are less if chemical descaling and/ or mechanical removal of scale have been previously accomplished. In any event, however, the acid concentrations and temperatures must be sufiiciently high, and the time of treatment sufiiciently prolonged to secure complete removal of corrosion products and organic pickling inhibitors adsorbed in the first chemical descaling operation.

Example 1 (parts A, B, and C) gives the range in concentration, temperature and time of treatment for three representative acids that may be optionally used to precondition ferrous metal surfaces for the new and improved alkaline inhibitors of this invention. Acid concentrations are expressed in terms of their anhydrous contents in the various solutions.' Technical grades are satisfactory.

Actually, the physical' Example 1.-Precnditioning acids and application PART Err-RANGE IN TEMPERATURE OF ACID, DEGREES F Hydrochloric Acid Sulfuric Acid a.

120-160 not recommended.

PART C.RANGE IN TIME OF TREATMENT, MINUTES 7-15 not recommended Hydrochloric Acid Sulfuric Acid When scale removal is required, all preconditioning must be accomplished by immersion; otherwise, spray application of the various acids accomplishes the desired results. In preconditioning ferrous metal surfaces coated with protective oils or slushing compounds, vapor phase degreasing should precede acid treatment. Steel or iron surfaces carrying only traces of oily substances are satisfactorily preconditioned with acid solutions of the same concentrations, but with a 70-1-30 blend of isopropanol and diethylene dioxide (1,4-dioxane) replacing to 30% of the water in the solutions.

After chemical treatment to remove adherent oxide and/or adsorbed organic films, the steel specimens are rinsed or sprayed thoroughly with cold water before making surface potential determinations. Alternately, they are immersed or sprayed with experimental modifications of the preferred composition of alkaline inhibitors established by this invention. Because of its convenience, the immersion technique is preferred in establishing the optimum proportions and concentrations of inhibitor ingredients, as well as the most effective temperatures and times of treatment. Various accelerated and outdoor exposure tests of the longevity of the different types of inorganic surface conversion coatings, to be described presently, served as the sole criteria of the preferred inhibitor formulas.

it is essential, of course, thatall traces of surface active agents .(anionic, cationic, nonionic), commercial pickling inhibitors, and other water soluble or insoluble organic and/ or inorganic films be completely desorbed from the ferrous metal surfaces in both physical and chemical preconditioning processcs. Obviously, the complete or partial retention of these barrier films will modify the metal surface potential and hence the uniformity of the ferrous hydroxide film. Although the barrier function of these extraneous coatings makes the surface more noble from the standpoint of passivity, they effectively prevent the desired surface conversion reactions normally obtained by the preferred blends of secondaryalkali phosphates with alkali nitrites of this invention. In chemical preconditioning, the importance of using cold water to remove iron salts cannot be overstressed. The slow evaporation rate of cold water retards drying which would cause rapid conversion of the thin ferrous hydroxide film to ferric hydroxide.

A weighed analysis of the results of outdoor exposure tests and accelerated laboratory tests was the sole criterion for the selection of inhibitor formulas best adapted for producing corrosion resistant coatings of exceptional durability on ferrous metal surfaces preconditioned by chemical techniques or the physical method of .wet sandblasting. In respect to wet sandblasting,various factors such as the widely varying temperatures of the surfaces treated, temperature of the sand-water slurry carrying the inhibitive chemicals, and the high degree of surface roughness obtained in blasting do not permit the attainment of as high a degree of rust inhibition by this method as by chemical preconditioning. Ncltt rtheless, it is possible to realize at least of the maximum protection given by chemical treatment, which is high in consideration of the reduced concentration of chemicals generally used to reduce costs, and the lack of controlled processing conditions.

Evaluation of qualifying inhibitors is accomplished by averaging the results of the following performance tests carried out on small steel specimens preconditioned by physical (wet sandblasting) or by chemical techniques:

1. Time of outdoor exposure required to secure the first incidence of breakdown, and later stages corresponding to 20, 40 and 60% of the panel areas covered with rust.

2. Number of rinses of tap water required to secure the first incidence of breakdown in an inclined-panel wash test, operated by inserting the panels in the slot of a Bakelite rod running the length of a stainless steel tank.

3. Time of exposure in a condensation-type humidity chamber in which the panels are suspended vertically on copper coils through which cold water circulates. This chamber is provided with a thermostatically controlled electric heater, that maintains a temperature of to degrees F. within the box under a state of complete moisture saturation. The weighted averages of the results given by these three tests are utilized in computing the accelerated test protective efiiciency ratings of various inhibitor solutions. An independent rating of protective efficiency is also made by outdoor exposure of treated panels.

Using the foregoing criteria, it was concluded that acidic formulations having "a pH less than 4.0 should be abandoned in normal temperature reactions of inhibitor solutions on preconditioned surfaces carrying the ferrous hydroxide film. In thisrespect, it was found that primary alkali phosphates similar to NaHzPOi, have definite value, probably because the pH of a 0.5% aqueous solution of the salt is 5.0, well above the 4.0 minimum required for maintenance of the film. The degree of inhibition is considerably less, however, than that obtained with the secondary phosphate, NazHPOi, which has a pH of 8.5 at the same concentration. When various blends of the two phosphates are used, the inhibition decreases measurably when the pH drops below 8.0, corresponding to a 30/70 blend of the primary and secondary salts.

It was further ascertained that the best inhibitive properties are obtained with solutions of a fairly narrow pH range, 8.0 to 9.0, and thatalkaline hydroxides, trisodium phosphate andother salts having a pH in excess of 9.0 give inferior results. Other tests indicated the desirability of limiting the selection of phosphates to the dibasic monovalent alkali salts, the hydrogen phosphate anions of which (HPOr) are capable of reacting with the ferrous hydroxide film to produce a coatingof hydrated ferrous phosphate (vivianite) by metathesis, in accordance with the equation:

furnished by thesaturated alkalinefilrn of preconditioned t el sur aces torrnduce insoluble c t n Since the degree of inhibition furnished by surface reactions of dilute solutions of secondary alkali phosphates on preconditioned steel at ordinary temperatures (65- 80 F.) is unexpectedly higher than can be attributed to the hydrated ferrous phosphate (vivianite) component of the coating alone, which undergoes fairly rapid decom position by hydrolysis and oxidation (vide supra), qualitative chemical tests were performed to detect other coating ingredients that might be responsible.

In conducting these tests, preconditioned steel panels were immersed in a 5% solution of secondary alkali phosphate (dibasic sodium, potassium or ammonium phosphates), a concentration which falls considerably short of saturation at temperatures of 65 to 80 F., but which gives the greatest thickness of coating and maximum protection. Treated specimens are allowed to dry after thoroughly immersing, and washed carefully to remove unreacted water soluble residues of the phosphate salt.

Chemical analysis of the coating is based on the solubility of vivianite (I. W. Mellor, op. cit., page 393) and hydrated ferrous and ferric oxides in a 30% solution of pure acetic acid. A preliminary qualitative analysis of these acetic acid extracts revealed the presence of phosphate ions and substantial quantities of iron in both the ferrous and ferric states. The ferrous form predominates, however, as shown by the heavier precipitates of Prussian Blue obtained when potassium ferricyanide was used as the precipitating reagent as contrasted to those obtained when potassium ferrocyanide was used.

Quantitative analyses of the phosphorus and the ferrous and ferric iron contents of 120 ml. of 30% acetic acid solution used in removing the coating from both sides of 12 steel panels, with an aggregate area of 2.73 sq. ft., were the basis for determining the composition of the coating obtained by immersing chemically preconditioned steel in a 5% solution of disodium phosphate. Example 2 gives the weights of the separate components of the coating thus obtained and their relative proportions.

Example 2.Cmp0sition of coating obtained with a 5% solution of disodium phosphate The formation of appreciable amounts of hydrous ferrous and ferric oxides integrally bound with vivianite by this procedure is reasonable from published data on the accelerated oxidation capacity of dissolved oxygen in alkaline solutions, as pointed out by Bauer, Kronhnke, and Masing in their book entitled Die Korrosion Metallischer Werkstoffe, vol. I, page 151: S. Hirzel, Leipzig, 1936. These investigators found that the passivity of iron in 1 N sodium hydroxide solution could be attributed to dissolved oxygen. Likewise, solutions of dibasic alkali phosphates, because of their alkalinity due to hydrolysis, are capable of absorbing much larger quantities of oxygen from the atmosphere than neutral or acidic phosphate salts. This accounts for the action of secondary alkali phosphates in promoting the formation of insoluble iron oxides as well as \u'vianite in accordance with the equations Bivalent and trivalent oxides of iron in the hydrated form comprise 61.3% of the total quantity of the surface 10 conversion coating, and vivianite the remaining 38.7%. Nearly four times as much iron is present in the ferrous as compared to the ferric state. It is further ascertained that the mechanical scrubbing technique, required to accelerate the chemical action of acetic acid on the coating, does not remove the underlying steel by blank tests of the effect of the same quantity, 120 ml. of 30% acetic acid, on 12 panels that were chemically preconditioned only and not immersed in the inhibitor solution before scrubbing.

The usual qualitative and quantitative tests for iron and phosphorous in. acetic acid washings ofruntreated panels are negative, as might be expected since 30% acetic acid is too weak to attack clean iron or steel surfaces at ordinary temperatures in the time required to rub both sides of the panels with a glass rod ",policeman covered with a glass fiber mat held in position with glass fibers. After oxidation with sodium pyrosulfate, however, minute traces of ferric ion are detected in the acetic acid solution by adding ammonium thiocyanate, as might be expected from the solvent action of acetic acid on the very thin hydrous ferrous oxide film covering preconditioned steel.

The coating is further proved to be in the amorphous state, since prolonged exposure of panel scrapings to -ray radiation from a FeKa source of X-ray radiation failed to reveal diffracted lines due to the presence of crystalline iron oxides or ferrous phosphate. Likewise, electron diffraction photographs of an electron beam of 300 microamperes intensity at 50,000 volts incident at 21 1 angle on small steel specimens gave negative results, in further proof of the non-crystalline nature of the coating. Hence, analytical chemistry techniques offer the best means, not only for identifying the various coating ingredients present, but for measuring the quantities present.

In exploring methods of improving the inhibitive properties of secondary alkali phosphate solutions by increasing their oxidizing effects on preconditioned ferrous metal surfaces, I have found that various neutral and alkaline oxidizing salts vary widely in their capacity for producing increased quantities of hydrated ferrous and ferric oxides in approximately 1:4 proportions as contrasted to 4:1 ratios of the same oxides obtained with solutions of secondary alkali phosphates alone. Of the various salts tested, sodium nitrite and other alkali nitrites are alone capable of increasing the proportions of hydrous ferrous and ferric oxides in approximately the reverse ratio as was found for a 5% solution of anhydrous disodium phosphate alone (vide supra).

By varying the relative proportions and total quantities of diammonium phosphate and sodium nitrite over a wide range, I have found that an.80/20 blend of the two chemicals at 5% total solids concentration gave the greatest protection. In terms of the times of exposure in various accelerated test devices and outdor weather to secure the first evidence of breakdown of the coating, the degree of protection was at least three times that obtained with 5% solutions of secondary alkali phosphate alone. Furthermore, the 80/20 blend was more effective than other blends in giving best results over a total concentration range of 0.25 to 10 Concentrations of inhibitor higher than 5% were generally ineffective in improving corrosion resistance. Related studies showed that the invention may be practiced with equal effectivenes's, as expected, with substantially the same concentrations of diba sic sodium and potassium phosphates substituted for diammonium phosphate, and ammonium nitrite and potassium nitrite for sodium nitrite in the proportions set forth by Example '3. Larger quantities of 7 dibasic sodium phosphates are always required, making allowance for 2, 7 or 12 water molecules of crystallization present in commercial products. The term anhydrousj hereinafter used, refers to the calculated quantity -11 of secondary sodium phosphate present in various crystalline modifications.

' 12 f mmon m it i i h he se on a y k l pho ph cannot be stored for the same reason. Solid ammonium Example 3.--Quantities of materials for optimum protective efiiciency Ingredient A Ingredient B Ingredient Percent Formula No. Water olids Dibasic Alkali lbs. Alkali Nitrite lbs.

Phosphate Pounds gollons* (*NH4) HPO4.... 80 NHiNOa soln. 93.0 1, 901. 4 221 4. 93 (NHfigHP 04-... 80 N aNOa 20. 0 l, 900. 0 221 5. 00 (N114 zHP04 80 KNO: 24. 7 1, 895. 3 220 5. 23 NA-iHP 04"- 86. 1 20% NHANO: soln... 93. 0 1, B95. 3 220 5. 23 NmHPOF-fi 2 NO 20. 0 1.804. 0 220 -5. 31 NazHPOr 24. 7 l, 889. 2 220 5. 54 K HPOL 03. 0 1, 875. S 218 6. 21 KgHPOi. 20. 0 1, 874. 4 '21 8 6. 28 K 111 O4. 24. 7 l, 869. 7 217 6. 52

Based on an average specific gravity of 1.03 of Formulas 1-9 calculated to the nearest gallon. Based on the anhydrous N azHPO; content of various hydrated crystalline modifications.

In compounding Formulas Nos. 1 and 2-9, inclusive, sufficient quantities of the various dibasic alkali phosphate and monovnlent alkali nitrites are dissolved to provide the same concentrations of hydrogen phosphate and nitrite anions (HPOF and NO2 58.16 and 13.34 pounds per 2000 pounds of solution, as given by the standard diammonium phosphate sodium nitrite composition, Formula No. 2. The varying quantities of secondary alkali phosphates and alkali nitrites in the other formulas are thus due to molecular weight differences. The necessity of using equal quantities of hydrogen phosphate and nitrite anions in the different formulas indicates that these anions, rather than the cations, are responsible for the ferrous metal surface conversion coatings produced in these alkaline media containing dissolved oxygen.

, Formula No. 2 is most frequently used in the wet sandblasting of the hulls of Naval vessels, because of the low cost and high solubility of the mixed crystals which can be added directly to the Pangborn mixer; alternately Formula No. S is sometimes preferred for this application when anhydrous disodium phosphate can be procured at a price advantage and when steam heated vats are available for preparing the required solution concentrates. Formula No. 8, on the other hand, is frequently selected for industrial applications based on both wet sandblasting and the chemical method of preconditioning the steel before spray or immersion application of the inhibitor. This blend can be prepared as a highly concentrated solution, in the neighborhood of for economical transportation to the point of use. Formula No. 2 blend, although equally soluble in high concentrations, unfortunately cannot be stored in steel drums because of the development of increasingly higher gas pressures due to the evolution of nitrogen gas in the reaction which causes the storage drums to burst. Formulas Nos. 1, 4, and 7 based on 80/20 blends of a 20% solution mgr./sq. ft. with a 5% nitrite should not be used in formulating this inhibitor because of its explosive properties.

The inhibitive properties of Formulas Nos. 1-9 can best be understood with reference to the composition of the coating obtained with a typical dibasic alkali phosphate alone, as explained in the accompanying discussion. The oxide ingredients of the coating are formed as a result of the accelerated oxidation caused by the presence of the alkali nitrite component of the inhibitor. The vivianite ingredient is formed, as in Example 2, by an exchange reaction of the hydrogen phosphate anions with the ferrous hydroxide film formed by preconditioning. The chemical analysis of acetic acid extracts has likewise shown that the coatings formed by the preferred secondary alkali phosphate and alkali nitrite blends of Example 3 contain the same ingredients but in different proportions by weight and in different absolute quantities.

The exceptional rust inhibitivc properties of the surface conversion coatings produced by the formulas of Example 3 are believed due to the more than three fold increase in weight of coating per unit area of surface (total weight of 192 mgL/sq; ft. as contrasted to 57.5 solution of disodium phosphate alone), and the presence of a sufiicient, though smaller, amount of vivianite to bond the hydrous ferrous and ferric oxide components of the coating. In this connection, it is of interest to note that the replacement of roughly 20% of the secondary alkali phosphate by an alkali nitrite, a mild oxidizing agent, has caused about a twenty-fold increase in the amount of hydrou ferric oxides formed as compared to only a 30% increase in the hydrous ferrous oxide content. Example 4 gives further details on the compositions and absolute quantities of coating ingredients formed as a function of the total concentration of dissolved salts over the range of 05-10%, which includes the 5% optimum concentration.

Example 4.--Chemicul analyses of surface conversion coatings produced with Formulas 1-9 inhibitors Formula Concentration, Weight Percent Coating Ingredients I p Inga] Wt. mgrJ Wt. mgr.l Wt. rngn/ Wt. mgnl Wt. mgrJ Wt.

so. it. persq. ft. persq. ft. persq. ft. persq. ft. persq. ft. percent cent cent cent cent cent Vivianite. 4s 7 5. 0 V 9. 3 -7. 2 1i. 1 7.7 13. 0 8. 0 15. 8 8. 2 17. 6 8. 4 Fe0.XHaO 16. 7 17.3 24.2 18.3 27.6 18,6 31. 6 18.8 36. 2 18.8 39. 0 18.0 FeiOnxHiot 73.4 77.7 99.7 74. 5 106 73. 7 123 V 73.2 14; 73.0 151 72.7

Total- 95. 0 1 00 133 144 100 167 100 193 100 208 100 i of watercognbinfisl' Wit fih oxideslsilndetermlnate.

It will be noted that as the concentration drops tenfold, from to 0.5%, the total coating weight per unit area decreases 50%; further, as the concentration drops the oxidation reaction leading to the formation of higher percentages of hydrous ferric oxide increases at the expense of lower percentages of hydrous ferrous oxide and vivianite. In respect to the minimum concentration of inhibitor giving maximum performance, the 5% solution, the total weight per unit area is approximately of the average weight of 185.8 mgr./ sq. ft. reported for anodized aluminum surfaces (H. H. Uhlig, op. cit., p. 861). On the basis of a calculated average density of 4.9 for the three coating ingredients, the maximum thickness of the surface conversion coatings produced with the 4.93-6.52 percent phosphate-nitrite blends of Example 4 is 0.000042 cm., or 0.017 mil.

In establishing the optimum 4:1 phosphate/nitrite blend, it was found that higher proportions of sodium nitrite increased the production of iron oxides at the expense of vivianite in the coating. Although the total thickness of the coating was greater, its durability was less. In the limiting case, when all the phosphate was replaced by nitrite, the inhibitive power was reduced considerably below the level obtained with the dibasic alkali phosphate salt alone. This undesirable effect is undoubtedly due to the occurrence of either one, or both, of the following reactions of sodium nitrite with the preconditioned film:

However, the protection obtained with sodium nitrite and other alkali nitrites is considerably better than given by other alkaline oxidizers of the type sodium chlorate, sodium peroxide, and sodium molybdate tested alone and in various blends with a dibasic alkali phosphate. The greater oxidation tendencies of these materials evidently suppresses the phosphating reaction and the mild oxidizing action of the dibasic alkali phosphate salts, leading to the formation of hydrous ferrous oxide. These experiments were valuable in demonstrating the importance of limited oxidizing action favoring the formation of hydrous ferrous oxide and suflicient vivianite to bond the iron oxide ingredients.

In sizeaole outdoor applications of my invention, such as the wet sandblasting of ships hulls and the protective maintenance of locks, dams, and bridges, it is unnecessary to use the 4.93 to 6.52% optimum concentrations of Example 3 to obtain coatings with a density of 185.8 mgr./ sq. ft. or higher. For applications of this type, I have found that 1.6% solutions of inhibitor, less than one third the optimum concentration, are strong enough to insure rust-proof surfaces of adequate longevity before application of paint.

In particular reference to the wet sandblasting of ships hulls in naval shipyards, I prepare a 1.6% solution of inhibitor at the time of use by adding 2 pounds of the required 4:1 blend as mixed crystals from quart can dispensers directly to the Pangborn sand-water mixers. Each charge of one of these mixers requires 300 pounds of sand, gallons of water, and 2 pounds of solid inhibitor (corresponding to a. 1.6% solution), usually sufiicient for blasting 100 square feet of surface. Alternately, larger quantities of inhibitor solution concentrates can be used if a very high degree of rust inhibition is required. In any case, to derive the full benefit of the inhibitor, the crystals or solution concentrates must be added after the full charge of sand and water has been fed into the mixer. This addition should be made prior to turning on the air pressure, since premature addition causes a reduced concentration of inhibitor at the gun before the end of the cleaning period, with consequent ineffective utilization of the inhibitor. To expedite this process, several cans of inhibitor crystals or solution concentrates are stationed at each Pangborn unit. These units, with auxiliary air lines and blasting equipment, are disposed at 20 to 30 foot intervals alongside the areas to be cleaned. The speed of the cleaning operation depends on the extent of corrosion, the amount of old paint to be removed, and the air pressure, which should not be allowed to drop below 90 p. s. i. for maximum efliciency. The process requires a final washing down of sand from the blasted areas by hosing with a stream of water.

Investigations of the protective efliciency of wet sandblasted panels as a function of the concentration of inhibitor solutions applied at satisfactory temperatures and a suitable time of contact have amply justified the use of lower than optimum quantities to effect necessary economies in the treatment of large structures. In these studies, protective efiiciency is defined as the percentage corrosion inhibition at lower concentrations relative to the maximum obtained for concentrations of 5% and higher. Protective efficiency ratios are evaluated by comparative measurements, visually and by chemical means, of the relative amounts of surface rust formed during the progress of outdoor exposure and various accelerated tests of panels, wet sandblasted or chemically preconditioned prior to treatment with inhibitor solutions. Example 5 is a summary of these measurements of protective efficiency as a function of inhibitor concentration, temperature, and contact time.

Example 5.Pr0tective efiicz'ency as a function of inhibitor concentration, temperature, and comact time PART A.EFFEOT 0F CONCENTRATION AT PREFERRED TEMPERATURES OF 70 DEGREES I AND CONTACT TIMES OF 2 MINUTES OR IVIORE Concentration (Weight percent) 0.25 Protection (percent). 60

PART B.EFFECT OF TEMPERATURE OF 5% INHIBITOR SOLUTION APPLIED FOR A MINIMUM CONTACT TIME OF 2 MINUTES Temperature, deg. F

80 Protection (percent) 90 PART C.--EFFECT OF CONTACT TIME OF 1.65% INHIB ITOR SOLUTIONS ADJUSTED TO TEMPERATURES 0F 70 DEGREES F.

Contact Time, Minutes 1 2 5 10 20 Protection, percent 80 90 ,100 100 100 100 A comparison of the above results with the data on coating composition of Example 4 indicates that there is a direct relation between protective efficiencies at lower than optimum concentration and coating thickness. Thus the protective efficiency of 75% obtained with the 1.6% solution used in wet sandblasting appears to be directly proportional to coating thickness, which in this case is 70% of the thickness obtained with a 5% solution.

The desirability of carrying out the wet sandblasting process at low, rather than high, ambient temperatures is clearly shown by Part B of the foregoing example. This precaution is especially important for lower than optimum concentrations where the effect of high surface temperatures of the treated surface, as well as ambient temperatures, is apt to be more pronounced. This is due to the fact that temperatures in excess of 100 degrees F. speed up the oxidizing reactions of the alkali nitrite component of the inhibitor to a prohibitive degree, which accounts for the marked reduction in protective efficiency. Higher temperatures also cause a noticeable darkening of the coating, which at lower temperatures has a bluish gray to mediiun gray tint due to light interference effects;

Consistent with its ability to inhibit rusting of unpainted steel under different conditions of exposure, steel 1.5 panels reacted with the preferred 4:1 blend of a secondary alkali phosphate with an alkali nitrite after physical and/or chemical preconditioning give a marked improvement in the durability of either a single primer paint coat, or complete paint system comprising two coats of primer and two finish coats.

In the case of chemically preconditioned panels treated with inhibitor, a three-fold improvement in paint durability is obtained in terms of the time required to obtain the first evidence of breakdown of the paint on control panels that are preconditioned only before painting. Physically preconditioned (wet sandblasted) panels give slightly less improvement as might be expected with rough surfaces due to blasting, lower inhibitor concentrations, and limited control of processing variables.

Obviously many modifications and variations of the present invention are possible in the light of the above teachings. It is therefore to be understood that within thescope of the appended claims the invention may be practiced otherwise than as specifically described.

.I claim:

1. A composition of matter consisting essentially of:

Secondary alkali phosphate, anhydrous, about 25.6

pounds to about 105.6 pounds Alkali nitrite, about 6.4 pounds to about 24.7 pounds, said nitrite combined with said phosphate in a ratio of 1 to 4.

2. A composition of matter consisting essentially of:

Secondary alkali phosphate, anhydrous, about 80 pounds Alkali nitrite, about 20 pounds.

3. A composition of matter consisting essentially of:

Secondary alkali phosphate, anhydrous, from about 25.6

pounds to about 105.6 pounds Alkali nitrite, from about 6.4 pounds to about 24.7 pounds, said nitrite combined With said phosphate in a ratio of l to 4 Water, from about 217 gallons to about 229 gallons.

4. A composition of matter for use in coating ferrous metal surfaces to resist corrosion, said composition consisting essentially of:

Secondary sodium phosphate, anhydrous, about 80 pounds Sodium nitrite, about 20 pounds Water, about 228 gallons.

5. A composition of matter for use in coating ferrous metal surfaces to resist corrosion, said composition consisting essentially of:

Diammonium phosphate, anhydrous, about 80 pounds Sodium nitrite, about 20 pounds Water, about 228 gallons.

6. In combination a ferrous member having a surface and a coating on the surface to resist corrosion, said coating comprising:

Vivianite (Fea(PO4)2.8HzO), about 8.2% Hydrous ferrous oxide (FeO.I-Iz), about 18.8% Hydrous ferric oxide (FezOaI-IzO), about 73.0%.

7. In combination a ferrous member having a surface and a coating on the surface to resist corrosion, said coating comprising:

Vivianite (Fes(PO4)2.8I-I2O), from about 7.7% to about Hydrous ferrous oxide (FeO.H2O), from about 18.6%

to about 18.8%

Hydrous ferric oxide (FezQaHzO) from about 73.0%

to about 73.7%.

8. A method for treating ferrous metal surfaces to resist corrosion, said method comprising the steps of removing paint and scale, applying a non-oxidizing acid to said descaled surface, rinsing the acid from said surface, whereby a surface to surface potential of -680 to 700 millivoltsis obtained, and contacting the surface with a composition consisting of secondary alkali phosphate, anhydrous from about 25.6 pounds to about 105.6 poundsalkali nitrite from about 6.4 pounds to about 24.7 pounds, said nitrite combined with said phosphate in a ratio of 1 to 4water from about 217 gallons to about 229 gallons to provide the coating of claim 7.

9. A method for treating ferrous metal surfaces to resist corrosion, said method comprising the steps of removing paint and scale, applying a non-oxidizing acid to said descaled surface, rinsing the acid from said surface, whereby a surface to surface potential of 680 to -700 millivolts is obtained, and contacting the surface with a composition consisting of secondary alkali phosphate, anhydrous from about 25 .6 pounds to about 105.6 pounds-alkali nitrite from about 6.4 pounds to about 24.7 pounds, said nitrite combined with said phosphate in a ratio of 1 to 4-water from about 217 gallons to about 229 gallons to provide the coating of claim 7, said composition having a concentration of about 1.0% by weight to about 5.0% by Weight, a temperature of about 50 F. to about F., and the time of contact being from about /6 minute to about 20 minutes.

10. A method for treating ferrous metal surfaces to resist corrosion, said method comprising the steps of removing paint and scale, applying a non-oxidizing acid to said descaled surface, rinsing the acid from said surface, whereby a surface to surface potential of 680 to 700 millivolts is obtained, and contacting the surface with a composition consisting of secondary alkali phosphate, anhydrous from about 25.6 pounds to about 105.6 pounds-alkali nitrite from about 6.4 pounds to about 24.7 pounds, said nitrite combined with said phosphate in a ratio of l to 4Water from about 217 gallons to about 229 gallons to provide the coating of claim 7, said composition having a concentration of about 5.0% by weight, a temperature of about 70 degrees F., and the time of contact being 2 minutes.

11. A method of treating ferrous metal surfaces to resist corrosion, said method comprising the steps of removing paint and scale, applying a non-oxidizing acid to said descaled surface, rinsing the acid from said sur face, whereby a surface to surface potential of -680 to -700 millivolts is obtained, and blasting the surface with a slurry of abrasive and a composition consisting of secondary alkali phosphate, anhydrous from about 25.6 pounds to about 105.6 pounds--a.lkali nitrite from about 6.4 pounds to about 24.7 pounds, said nitrite combined with said phosphate in a ratio of 1 to 4-water from about 217 gallons to about 229 gallons, said composition having a concentration about 1.6% by weight, a temperature of about 50 degrees F., and the time of contact being from about /2 minute to about 3 minutes.

12. In combination a ferrous member having a surface and a coating on the surface to resist corrosion, said coating comprising:

Vivianite (Fe3(PO4)2.8H2O), about 8.2%

Hydrous ferrous oxide (FeOl-IzO), about 18.8%

Hydrous ferric oxide (FezOaHzO), about 73.0%, the aggregate Weight of said coating ranging from 144 to 193 mgr./ft.

13. In combination a ferrous member having a surface and a coating to resist corrosion, said coating comprising: Vivianite (Fes(PO4)z.8I-I2O), from about 7.7% to about Hydrous ferrous oxide (FeO.XHzO), from about 18.6%

to about 18.8%

Hydrous ferric oxide (Fe20s.XH20), from about 73.7% to about 73.0% the aggregate Weight of said coating ranging from about 144 to 193 mgr/ft? 14. A blend of inorganic chemicals consisting essentially of secondary alkali phosphate and alkali nitrite mixed in proportions corresponding to 58.16 pounds hydrogen phosphate anions (HPOr) and 13.34 pounds nitrite anions (NOT).

15. A blend of inorganic chemicals consisting of secondary alkali phosphate and alkali nitrite dissolved in water in proportions corresponding to from about 5.82 to 58.2 pounds of hydrogen phosphate anions (HPOr) and from about 1.33 to 13.3 pounds of nitrite anions (NO2) per 2000 pounds of solution.

16. A chemical method for treating ferrous metal surface to resist corrosion, said method comprising the sequential steps of removing rust and scale of said surface with non-oxidizing acid solutions, rinsing the acid from said surface with cold Water having a pH of 4.5 to 8.5, whereby a surface to surface potential of 680 to 700 millivolts is obtained, and contacting the surface with the composition consisting of secondary alkali phosphate, anhydrous from about 25.6 pounds to 105.6 pounds-- alkali nitrite from about 6.4 pounds to about 24.7 pounds, said nitrite combined with said phosphate in a ratio of 1 to 4-water from about 217 gallons to about 229 gallons to provide the coating of claim 7.

17. A physical method for treating ferrous metal surfaces to resist corrosion, said method comprising the steps of simultaneously wet-sandblasting the surface with a slurry of abrasive admixed with the composition consisting of secondary alkali phosphate, anhydrous from about 25.6 pounds to about 105.6 pounds-alkali nitrite from about 6.4 pounds to about 24.7 pounds, said nitrite combined with said phosphate in a ratio of 1 to 4 water from about 217 to about 229 gallons to provide the coating of claim 7, said composition having solids concentrations of about 1.6 to 5% by weight, temperatures of F. to F., and times of contact ranging from about /2 minute to about 3 minutes.

References Cited in the file of this patent UNITED STATES PATENTS 1,215,463 Allen Feb. 13, 1917 1,549,409 Gravell Aug. 11, 1925 1,781,507 Gravell Nov. 11, 1930 1,850,726 Pfalzgrafi Mar. 22, 1932 2,217,586 Zapf Oct. 8, 1940 2,403,426 Douty et al. July 2, 1946 2,479,423 Snyder Aug. 16, 1949 2,609,308 Gibson Sept. 2, 1952 FOREIGN PATENTS 517,049 Great Britain Jan. 18, 1940

Claims (1)

1. A COMPOSITION OF MATTER CONSISTING ESSENTIALLY OF: SECONDARY ALKALI PHOSPHATE, ANHYDROUS, ABOUT 25.6 POUNDS TO ABOUT 105.6 POUND ALKALI NITRITE, ABOUT 6.4 POUNDS TO ABOUT 24.7 POUNDS, SAID NITRITE COMBINED WITH SAID PHOSPHATE IN A RATIO OF 1 TO 4