Solder joints on rustless alloys

74 • T H E J O U R N A L O F T H E A M E R I C A N D E N T A L A S S O C I A T I O N

technique: its a dvan tage s and d isad van tage s. J. Pros. Den. 1:727 Nov. 1951. 4. Bignell, K. A . H y d ro co llo id technic. Illin o is D. J. 18:163 M a y 1949. 5. Pfeiffer, K. R.. and Jeffreys, F. E. C o m p le te b rid ge technic utilizing tne a lgin a te hydrocolloids. J .A .D .A . 40:66 Jan. 1950. 6. Thompson, M . J. Standardized indirect technic for reversible hydrocolloid. J .A .D .A . 46:1 Jan. 1953. 7. Skinner, E. W . Science of dental materials, ed. 4. Philadelphia, W . B. Saunders Com pan y, 1954. 8. Bailey, L. R. Rubber base impression techniques. Dental C linics of North A m erica. Philadelphia, W . B. Saunders Co m pan y, M arch 1957, p. 157. 9. Phillips, R. W .; Price,# R. R., and Reinking, R. H. Use o f a lgin a te for indirect restorations. J.A .D .A . 46:393 A p ril 1953. 10. Paffenbarger, G . C . H y d ro co llo id a l impression m aterials: physical properties and a specification. J .A .D .A . 27:373 M arch 1940.

11. Burkman, N. W .; Schmidt, H. S., and Crowley, M . C . Prelim inary report of an investigation to study the effectiveness of certain d rugs for sterilizing carious dentin. O ra l Surg., O ra l M e d . & O ra! Path. 7:647 June 1954. 12. Patterson, S. S., and Van Huysen, G . Treatment of pulp exposures. O ra l Surg., O ra l M ed . & O ra l Path. 7:194 Feb. 1954. 13. Klein, Arthur. Personal comm unication. 14. Som m er, R. F.; Ostrander, F. D., a n d Crowley, M . C. C lin ic al endodontics. Philadelphia, W . B. Saunders C om pany, 1956, p. 148-151. 15. Bender, I. B. Penicillin in root canal therapy: report of 53 cases. J .A .D .A . 34:99 Jan. 1947. 16. Grossm an, L. I. Root canal therapy, 4th ed., Philadelphia, Lea & Febiger, 1955, p. 276-279. 17. Stewart, G . G . Preliminary report on penicillin in root therapy. J .A .D .A . 33:1281 O c t. 1946. 18. Healey, H. J. Restoration of the effectively treated pulpless tooth. J. Pros. Den. 4:842 Nov. 1954.

S o ld e r jo in ts on ru s tle ss a llo y s

Saul M . Bien, D.D.S., Lynbrook, N. Y., and Herbert D. Ayers, Jr.,* D.D.S., New York

Passivation is a method of restoring the continuity of the oxide film on metal sur­ faces in order to reduce corrosion. Nickelchromium alloys should be chemically or electrically passivated before these alloys are soldered. Wherever possible, soldered nickel-chromium alloys should be scrubbed mechanically, dried and set aside to be passivated atmospherically.

The increased use of nonprecious rust­ less alloys for dental structures (particu­ larly in prosthetics, oral surgery and orthodontics) coupled with the instabil­ ity of the solder bonds, demands that tests and evaluation be made of the physical properties. Although tissues tolerate nickel-chro-

mium alloys in contact with them, the alloys are subject to corrosion.1 The cor­ rosion resistance o f rustless alloys is be­ lieved to result from a thin, continuous surface film of oxide formed by exposure to atmospheric or to certain oxidizing conditions. The film o f oxide is removed or its continuity is disturbed during spot welding or soldering, or when the alloy is abraded while being fabricated into a dental appliance. When the dental appli­ ance will be exposed immediately to saliva and decomposing food, it is nec­ essary to restore the continuity of the oxide film by a process called passivation (exposure to certain acids or to acid mixtures containing oxidizing agents). Passivation in certain acids or mixtures containing oxidizing agents, or by elec­ trical anodization, reduces the magnitude of corrosion.

B IE N — A Y E R S . . . V O L U M E 58, M A Y 1959 • 75

• Relative pit susceptibility of 18-8 type 302 austenitic steel* T a b le

Physical finishf

Weight loss (mg.)

No. of large pits

No. 1 No. 1-passivated No. 4 No. 6 No. 8

30.2 2.7 25.1 12.6

20 2 13 5 2

5.5

*Based on tests by H. A. Smith.2 |The finishes of the test pieces used by Smith are standard in regular steel production: "Finish no. I, hot rolled, annealed, pickled. Finish no. 4, standard polish. . Finish no. 6, . . . soft and satiny and has low reflectivity. Finish no. 8, mirror finish. . .

SURFACE CH ANG ES R E S U LTIN G FROM PA S S IV A T IO N OF RU STLE SS ALLO YS

H. A. Smith,2 in a study of relative susceptibility to pitting, showed that in 18-8 type 302 austenitic steel the surface finish was related to the amount of cor­ rosion. The corroding solution was 10.8 per cent Fe C 1.3' 6H 20 in 0.05 normal hydrochloric acid with immersion for four hours. The passivation used by Smith was accomplished by a 15 minute immersion in 10 per cent chromic acid at 170°F. T he test area was 10 square centimeters (see table). An analysis of the results shows that the greatest loss o f weight is coupled with the greatest number of pits in an ascending scale relative to the smooth­ ness of the physical finish, with the sole exception of the passivated specimen. The roughest finish shows the greatest weight loss coupled with the greatest number of pits but a magnitude of loss of only 1.51 mg. per pit. The mirror finish shows a weight loss of only 5.5 mg. with only two pits, but each pit indicates a loss o f substance of approximately 2.75 mg. The passivated surface of the roughest finish, on the other hand, shows the lowest weight loss coupled with the lowest weight loss per pit. The usual procedure for passivation is by immersion o f the article to be passi­ vated in a solution containing 10 to 20

per cent nitric acid by volume at 130° to 140° F. for 15 to 20 minutes. An alter­ nate method is by immersion in a solu­ tion of 10 to 20 per cent nitric acid containing 2 per cent sodium dichromate at 110° to 130° F. for 30 minutes, fol­ lowed by immersion in a 5 per cent solution o f sodium dichromate at 140° to 160° F. for one hour. Neither of these methods is o f practical use in passivating dental appliances because of the lack of time as well as the usual absence of a hood. Immersion in a solution of 8 Gm. ferric sulfate (anhydrous), 4 Gm. so­ dium fluoride, 10 ml. sulfuric acid (con c.), and 90 ml. water at room tem­ perature for ten minutes is more suitable in the dental office. A polyethylene con­ tainer with a tightly fitting polyethylene cover should be used to hold the passi­ vating solution. The solution loosens scale by converting the oxides of iron, chro­ mium and nickel to higher oxides more readily soluble in the acid, and also pas­ sivates the surface of the rustless alloy by oxidation. The ferric sulfate in the solution combines with excess hydro­ fluoric acid to form ferric fluoride which is nonvolatile at room temperature. This

F ig . I • E le c tric a lly p a ssiv a te d 18-8 a u ste n itic steel ty p e 3 0 4 w ire ( C ) , s o ld e re d with silver s o ld e r ( A ) . T he p ic r ic a c id etch revea ls w ro u g h t a u ste n itic stru c tu re ( I ) in the center, w h erea s at the p e r ip h e r y the w ire resists the r e a g e n t (2 ). This c h a ra c te ristic is p re se n t t h r o u g h o u t entire le n g th o f the wire, w h e th e r a d ja c e n t to so ld e r o r at extrem ities w h ich h a ve n ot b ee n su b je c te d to th e flux o r the te m p e ra tu re o f s o ld e rin g

76 • T H E J O U R N A L O F T H E A M E R I C A N D E N T A L A S S O C I A T I O N

solution will loosen gross scale but will not polish an appliance. “ Electrical passivation” of rustless al­ loys is a more rapid and more controlled procedure than “ chemical passivation” (Fig. 1 ). In addition to polishing, passivation, and removal of scale, gross reduction in wire diameter can be made rapidly. The appliance is used as anode in a solution of 90 ml. phosphoric acid (con c.), 30 ml. glycerine and 30 ml. water, contained in a stainless steel dish acting as cathode, with a current sup­ ply of 10 amperes at 12 volts. T h e time is 30 seconds except where gross reduc­ tion in size is desired. SO LD ER J O IN T S I N G O L D -P L A T IN U M A N D N IC K E L -C H R O M IU M A L L O Y S

In this study o f solder bonds, the mate­ rials joined were 18-8 type 304 austenitic steel wrought wire and tubes supplied by two different manufacturers; 80 nickel20 chromium wire, and gold-platinum wire. The solders used were 450 fine 14 carat yellow gold solder, melting point of 1415° F.; 400 fine 10 carat white gold solder, melting point of 1390° F., and a silver solder containing silver, copper, cadmium, zinc and nickel, melting point o f 1170° F. A fluoride flux, gas-air ortho­ dontic blow torch and a wire jig served as accessory materials. The specimens were mounted in acrylic resin, polished, and examined microscopically before and after etching. The etching reagent for the nickel-chromium alloys was saturated picric acid in ethyl alcohol plus 5 per cent hydrochloric acid, and for the goldplatinum alloy it was 10 per cent potas­ sium cyanide, plus 10 per cent ammo­ nium persulfate. The term soldering applies to the join­ ing o f alloys by the process o f melting a metal or alloy with a lower fusing point, flowing it over the alloy to be joined at the point for junction, and allowing it to cool below the melting range o f the solder, to solidify and form the joint.

Generally, at temperatures over 800° F., this is classed as brazing. Microscopic examination indicates dis­ tinct differences between the solder bonds involving nickel-chromium alloys and those involving gold-platinum alloys. This is in accord with the reported dif­ ferences in the type o f failure during function. At temperatures suitable for soldering without reducing valuable phys­ ical properties, there is a solution of the gold-platinum alloys in the solder. At similar temperatures, any alloying o f the nickel-chromium that may occur appears to be superficial. The nickel-chromium alloys studied appear to have a microstructural framework that remains intact as a physical boundary or interface dur­ ing soldering (Fig. 1). “ Soldering by solution” applies to the joining o f alloys where there is a dis­ persion of one or more substances in another so as to form a homogeneous mixture. Coleman3 and Crawford4 have shown evidence o f solution of wrought gold alloys in gold solder. With an excess o f a constituent with a higher freezing point present, a continu­ ous series of alloys soluble in the liquid phase will occur. When a bit of solder is melted on a gold alloy wire, the wire starts to go into solution. In a solution, the particles of the solute are in a state of extremely minute subdivision and they are distributed throughout the solvent. The limit o f solubility can, in general, be extended by a rise in temperature. In the gold solder to gold alloy joints, these conditions are met. The melting range o f the solder rises as long as any wire remains and the heat is increased to maintain the solder in molten condi­ tion. These specimens (Fig. 2-4) were prepared as rapidly as possible. The al­ loying did not go as far as on the longer heated specimen shown by Coleman,3 who indicated that the gold-to-gold sol­ dered bonds fail most often adjacent to the bond, because o f structural failure o f the wire.

B IE N — A Y E R S . . . V O L U M E 58. M A Y 1959 • 77

perature used. The junction is easy to identify. It invariably appears as a struc­ ture continuous with the original surface. When the bond of a rustless alloy fails, investigation shows that it is the bond itself, not the alloy or solder, that has failed. The solder separates from the alloy as an integral unit. The junction of solder with the goldplatinum alloy is difficult to identify in the unetched specimens. Etching clearly reveals the invasion of the wrought struc­ ture by the solder. This has not been demonstrated with nickel-chromium al­ loys. Since the wrought nickel-chromium and gold-platinum alloy wire structures shown in Figures 2, 3 and 4 were in the same mass of molten solder, they must have been in about the same temperature range. Fig. 2 • J u n c tio n o f 18-8 a u ste n itic steel ty p e 3 0 4 tu b e ( C ), so ld e re d with silve r so ld e r ( A ) to q o ld -p la tin u m w ire ( G ) . O u t lin e o f g o ld p la tin u m allo y ( I ) is v a g u e or a lm o st im p o ssib le to see

“ Soldering by wetting” applies to a superficial attachment of the lower fusing metal or alloy to the other. Wetting may be accompanied by adsorption, diffusion, or leaching. In all specimens of the nickel-chromium alloys, the boundary between the wrought structure and the solder is clear when polished but not etched. It makes no difference how the wire was cleaned prior to soldering, or which solder was used. The surface irregularities o f the wrought structure are evident against the solder. One might assume that the orig­ inal surface irregularities, cleaned and corroded by flux, provide means of re­ tention for the solder when it is applied to rustless alloys. Microscopic examination of the inter­ face and structure of joints between solder and the nickel-chromium alloys studied indicates that the wrought struc­ ture remains intact. This is true in every instance, at the magnification and tem­

CORROSION OF SOLDER JO INTS ON N IC K E L -C H R O M IU M A LLO Y S

Nickel-chromium alloys used for ortho­ dontic, surgical, prosthetic and orthopedic purposes, have as their most limiting factor characteristic types of corrosion. For the most part these types o f corrosion have been mentioned in the literature, but not fully explained. This study was made of two types of rustless alloys which are

F ig . 3 • S a m e a rea as in F ig u re 2, w ith p ic r ic a c id etch w h ich m akes c le a r the ¡unction ( I ) o f the silv e r so ld e r ( A ) with the g o ld -p la t in u m allo y ( G )

78 • T H E J O U R N A L O F T H E A M E R I C A N D E N T A L A S S O C I A T I O N

Fig. 4 • S a m e area as in F ig u re s 2 a n d 3. P o ta s­ sium c y a n id e etch show s th a t ch e m ic a l c o m p o s i­ tion o f s o ld e r ( A ) is n ot co n sta n t

resistant to body fluids and are commonly used for orthodontic bands and appli­ ances: 18-8 type 304 austenitic steel and 80 nickel-20 chromium. These two al­ loys are commonly subject to several types o f corrosion: “ crevice corrosion,” “ stress corrosion cracking” and “ fire cracking.” Stress corrosion cracking is character­ ized by a “ spontaneous failure of metals by cracking under combined action of corrosion and stress, residual or applied.” 8 Fire cracking results where “ residual stresses . . . cause metals to crack quickly . . . when heated.” 6 Crevice corrosion is characterized by a failure of the bond, with little or no macrocorrosion of the soldered joints or the solder. In the appliances where the bond failed either clinically or experi­ mentally, there was no obvious corrosion or pitting of any metal.

Based on the research7 into the nature of crevice corrosion in type 430 martensitic steel which contains no nickel, it was suggested in the Handy and Harman technical bulletin T9 that a nickel-con­ taining brazing alloy might eliminate crevice corrosion in 18-8 austenitic steel as well. Silver-copper alloys o f this type are available commercially. A silvercopper-cadmium-zinc-nickel alloy (Fig. 5) was used in this series in addition to the 400 fine white gold solder (Fig. 6) and 450 fine yellow gold solder (Fig. 7 ). Samples of soldered joints were washed in detergent to remove flux. They were placed in the ferric fluoride passivating solution for ten minutes. The samples were then mounted in acrylic resin and polished with abrasive paper, and with aluminum oxide in water. In the m icro­ scopic examination at 365 magnifica­ tion, crevice corrosion in varying degrees was readily observable. Other samples o f soldered joints were immersed in a ferric fluoride solution for 12 hours. Total separation o f the parts on certain of the samples ensued (Fig. 8). Immersion of joints for relatively short periods in a ferric fluoride solution in vitro would appear to simulate similar corrosion which takes place in vivo over

i I____

—

I--------— ------------ —— i— i— i------—

Fig. 5 • A b o v e : 80 n ick e l-2 0 c h ro m iu m a llo y ( N ) , s o ld e re d w ith s ilv e r-c o p p e r-c a d m iu m -z in c nickel a llo y s o ld e r ( A ) , c h e m ic a lly p a ssiv a te d . Below : 18-8 a u ste n itic steel t y p e 3 0 4 a llo y ( C J , so ld e re d with silv e r-c o p p e r-c a d m iu m -z in c -n ic k e l allo y s o ld e r ( A ) , c h e m ic a lly p a ss iv a te d

B IE N — A Y E R S . . . V O L U M E 58, M A Y 1959 • 79

Fig. 6 • A b o v e : 80 nickel-20 ch ro m iu m alloy ( N ) , s o ld e re d with 4 0 0 fine w h ite g o ld so ld e r ( W ) , c h e m ic a lly p a ssiv a te d . Below : 18-8 austenitic steel t y p e 3 0 4 a llo y ( C ) , s o ld e re d with 4 0 0 fine w hite g o ld so ld e r ( W ) , c h e m ic a lly p a ssi­ va te d

a period o f months or, in some instances, years. Crevice corrosion can be clearly dem­ onstrated in 18-8 type 304 austenitic steel and in 80 nickel-20 chromium alloy, with all three types o f solder used, despite the nickel content of the silver solder. H. L. Logan,8 in his investigation of stress corrosion, found a difference in electric potential between areas of 18-8 type 302 austenitic steel in which the atmospherically - formed protective film was ruptured by abrasion, and areas in which the film was intact. The sur­ faces in which the film was ruptured

Fig. 8 • 80 nickel-2 0 c h ro m iu m a llo y ( N ) , so ld e re d t o 18-8 a u ste n itic steel ty p e 3 0 4 ( C ) , with 4 5 0 fine ye llo w g o ld ( Y ) , 4 0 0 fine w hite g o ld ( W ) , sllv e r-c o p p e r-c a d m iu m -z in c -n ic k e l a lloy ( A ) , in fe rro u s flu o rid e solu tion fo r 12 h ours

Fig. 7 • A b o v e : 80 n ick e l-2 0 c h ro m iu m a llo y ( N ) , so ld e re d with 4 5 0 fine g o ld s o ld e r (Y ) , c h e m i­ ca lly p a ssiv a te d . B elow : 18-8 a u ste n itic steel typ e 3 0 4 a lloy ( C ), s o ld e re d with 4 5 0 fine g o ld so ld e r (Y ), ch e m ic a lly p a ssiv a te d

were 0.28 to 0.78 volts more negative with respect to a calomel electrode of the saturated KC1 type, than the intact sur­ faces. With this evidence of differences of potential between passivated and ac­ tive surfaces, the mechanism of crevice corrosion is attributed to the formation of a cell of the active-passive type (Fig. 9 ). It would need but the slightest active surface on the junction o f the bond to provide a foothold for the formation of such a cell. Frank LaQue of the International Nickel Company, in an address on the “ Nature of Corrosion of Metals” before the New York Chapter of the American Society for Metals on October 10, 1955, indicated that an oxygen concentration cell of differential aeration type (Fig. 9) might be a responsible factor in crev­ ice corrosion. A situation favorable to the formation of such cells might well exist in the mouth. Curiously enough, although crevice corrosion will take place consistently in the solution used for the in vitro ex­ periments, it does not take place consist­ ently with gross separation in the mouth. Not all of the bonds will disintegrate in the same mouth, nor necessarily disin­ tegrate in the mouths o f all patients.

80 • T H E J O U R N A L O F T H E A M E R I C A N D E N T A L A S S O C I A T I O N

SALIVA

Fig. 9 • C e lls of a c tiv e -p a ssiv e ty p e a n d d iffe re n ­ tial a e ra tio n typ e

Flq. 10 • A b o v e : 18-8 a u ste n itic steel t y p e 3 0 4 ( C ), s o ld e re d with s ilv e r-c o p p e r-c a d m iu m -z in c nickel a llo y s o ld e r ( A ) , e le c tric a lly p a ssiv a te d . Below : 80 nickel-2 0 ch ro m iu m a llo y ( N ) , so ld e re d with silv e r-c o p p e r-c a d m iu m -z in c -n ic k e l a llo y so ld e r ( A ) , e le c tric a lly p a ssiv a te d

SOLDERED JO IN TS AFTER E LECTRICAL P ASSIV ATIO N

Samples of the 80 nickel-20 chromium alloy and 18-8 type 304 austenitic steel wire were soldered with a silver-coppercadmium-zinc-nickel alloy. They were washed with detergent to remove the flux. The samples were immersed as anode in a phosphoric acid-glycerinewater solution with a current supply of 10 amperes at 12 volts for 30 seconds. In Figure 10, above, there appears the formation of a shallow depression in the rustless alloy directly adjoining the solderrustless alloy junction. In Figure 10, be­ low, there appears to be crevice corrosion. C O N C LU S IO N S

Since crevice corrosion was demonstrated after both chemical and electrical passi­ vation, it would appear that chemical or electrical passivation of the nickel-chro-

mium alloys might best precede solder­ ing. Wherever possible, soldered nickelchromium alloys should be mechanically scrubbed, cleaned, dried and then set aside to be passivated atmospherically. 76 Union Avenue

*Associate clinical professor of dentistry, School of Dental and Oral Surgery, Columbia University. 1. Bien, S. M._ Disintegration of solder Joints on base metal orthodontic appliance. Unpublished. 2. Smith, H. A. Pit corrosion of stainless steel. Metal Progress, 33:596 June 1938. 3. Coleman, R. L. Some effects of soldering and other heat treatments on orthodontic alloys. Internat. J. Orthodont. & Den. Children 19:1238 Dec. 1933. 4. Crawford, W. Personal communication to H. D. Ayers, Jr. 5. The American Society for Metals. Metals hand­ book. Cleveland, The American Society for Metals, 1948, p. 14. 6. The American Society for Metals. Supplement to metals handbook. Cleveland, The American Society for Metals, 1955, p.89. 7. Halbig, J.J.; Grenell, L H., and Sistare, G. H. New alloys stop corrosion in silver-brazed type 430 joints. Iron Age 172:159 Dec. 10, 1953. 8. Logan, H. L. Fijm-rupture mechanism of stress corrosion. J. Res. National Bureau of Standards 48:99 Feb. 1952.

Health and W ater • So long as men inhabit the earth, their success, happiness and health will, in a large measure, depend upon how wisely and how well they control and use water. Thom as King.