>v> *^ V

m- ^/ #fe v* "% \/ .•$&••\^ .•:

a <v

"° ^ ^ * is

*o . » » A

o V •^d*

%̂ ^ A^ '^

A^

V^ ^^^; ^\. c

^

aV V '•

* v- , -^^^¥ * A

if

x<*

* ^

•JO. •» *(> »

O^ * o H o ° -0

*> <>. *^M^J<" "Pa A

V 'O, 'o .A& ^, "° • *

:^°^

c, ... *^ ..„.

tv « l ' • * o A r a o . ^

IC 8985

Bureau of Mines Information Circular/1984

Pickling of Stainless Steels—A Review

By Bernard S. Covino, Jr., John V. Scalera,

and Philip M. Fabis

UNITED STATES DEPARTMENT OF THE INTERIOR

" a' 11 * '.afaw^iiwrewsawragpfflp

Information Circular 8985

Pickling of Stainless Steels—A Review

By Bernard S. Covino, Jr., John V. Scalera,

and Philip M. Fabis

' " •:•

UNITED STATES DEPARTMENT OF THE INTERIOR

William P. Clark, Secretary

BUREAU OF MINESRobert C. Norton, Director

Library of Congress Cataloging in Publication Data:

V<&

<*

Covino, B. S. (Bernard S.)

Pickling of stainless steels—a review.

(Information circular / United States Department of the Interior, Bu-reau of Mines ; 8985)

Bibliography: p. 13-15.

Supt. of Docs, no.: I 28.27:8985.

1. Steel, Stainless—Pickling. I. Scalera, John V. II. Fabis, PhilipM. III. Title. IV. Series: Information circular (United States. Bu-reau of Mines) ; 8985.

TN295.U4 [TS654] 622s [669'. 142] 84-600164

k

4£ CONTENTSPage

i

V Abstract 1

j Introduction 2

/ Effect of hot and cold working on pickling 3

Hot working 3

Cold working 4

Effect of annealing on pickling 4

•XEf fect of conditioning on pickling 6

Degreas ing 6

Abrasive blasting 6

^ Chemical conditioning 6

N. Reducing acids 7

Qv/N Oxidizing acids 7

Electrolytic acid conditioning 7

rAnodic conditioning « ... .

rf7

v~/ Cathodic conditioning '.'. .'..*.. 8

Alternating current conditioning ' 8

Electrolytic neutral conditioning 8

Salt bath conditioning 8

Reducing bath 9

Oxidizing bath 9

Electrolytic bath 9

Effect of pickling bath variables on pickling 10

Pickling operation 10

Bulk alloy dissolution 11

Solution composition 11

Acid concentration 11

Mixed acids 11

Temperature 12

Chromium-depleted zone dissolution 12

Research needs 12

References 13

ILLUSTRATION

/^l. Schematic of steps involved in processing stainless steels 2

\

9\

UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT

A/ft 2 ampere per square foot ym/yr micrometer per year

°C degree Celsius rpm revolution per minute

min minute vol pet volume percent

Um micrometer wt pet weight percent

PICKLING OF STAINLESS STEELS-A REVIEW

By Bernard S. Covino, Jr., John V. Scalera, and Philip M. Fabis

ABSTRACT

The Bureau of Mines is conducting a study of methods to improve theefficiency of the process used to pickle stainless steels. A review ofthe literature on pickling of stainless steels showed that the chemistryof several process operations involved in the pickling of stainlesssteels is not fully understood, and that further research could improvethe pickling efficiency. The benefits of this research would be a re-duction in the annual loss of several thousands of tons of criticalminerals such as nickel and chromium, and a reduction in the amount of

solids and spent acid solution that are currently discarded. The con-clusion from this review is that further research is needed in fouroperations that either directly or indirectly influence the picklingprocess: hot working, annealing, conditioning, and the actual operationof pickling. Laboratory studies of the pickling operation are presentlyin progress.

'Research chemist.^Materials engineer.Avondale Research Center, Bureau of Mines, Avondale, MD.

INTRODUCTION

The various operations involved in pro-cessing stainless steels are given infigure 1. After being worked by hot orcold rolling, the steel is softened byannealing. Oxide forms on the stainlesssteels during this annealing process,which, as shown in figure 1, occurs sev-eral times. Conditioning is used to

facilitate the pickling process. Mixed-acid pickling, or pickling by a solutionof two or more acids , is then used for

cleaning the oxide-covered stainlesssteels. In addition to removing the an-nealing scale, pickling also removes avery thin (1- to 5-pm) region depleted inchromium between the oxide and the bulkstainless steel. Loss of chromium andnickel from this region and the oxide isinherent in this part of the pickling op-eration. In practice, the stainlesssteel may be left in the pickling solu-tion longer than necessary, causing ex-cessive dissolution of the bulk steel,resulting in losses of several thousandtons of chromium and nickel annually.The combined dissolution products canbuild up to a point where the action atthe pickling bath stops, resulting in asizable disposal problem when the bath is

replaced. The dissolved metals also in-crease the use of acids in the picklingbath by complexing or precipitating acidsalts. The need to study the picklingprocess was formulated during discussionsbetween the Bureau of Mines and the Amer-ican Iron and Steel Institute (AISI).Both groups concluded that the problemsof loss of critical minerals, excess useof acids , and disposal of spent solutionscould be lessened by a better understand-ing of the entire pickling process.

The literature pertinent to the pick-ling of austenitic stainless steels wasreviewed. Data bases such as ChemicalAbstracts, Metadex, Compendex, and NTIS(National Technical Information Service)were searched from 1900 to 1983 whereapplicable. Although articles in alllanguages were accepted in the search,the review was done mainly on articles inEnglish. The general purpose of the re-view was to assess the technology of

stainless steel pickling and to present a

critical examination of the mechanism ofmixed-acid pickling of stainless steelsin terms of all the important processoperations and operating parameters.

All such operations and parameters areconsidered in light of how much knowledgeis available and what further knowledgeis necessary for a better understandingand control of the pickling process.This review begins by considering thosefactors in the metal-forming operationthat can affect subsequent pickling be-havior. The effect of annealing on pick-ling is then addressed. Conditioningtreatments prior to pickling have a very

Ingot Hot work Anneal

j

Conditioning:

mechanical

or

chemical

1

j

Pickling

"

Cold work

'

Anneal

1

Conditioning:

chemical

or

electrolytic

or

salt bath

•

PicklingAs necessary

'

Finished

material

FIGURE 1. - Schematic of steps involved in process-

ing stainless steels. Indicated steps are repeated as

often as necessary to thin the material to the desired

thickness.

significant effect on pickling and are

addressed next. Finally, the operationof the pickling bath is considered. The

present understanding of the mechanism of

pickling of stainless steel is developedin detail in this last section.

EFFECT OF HOT AND COLD WORKING ON PICKLING

In order to be inclusive, the effectsof hot and cold working on the picklingof stainless steels are briefly consid-ered. Working of the stainless steelusually has no direct effect on the pick-ling operation because an annealing op-eration is interposed between the workingand the pickling steps. However, someeffects of hot and cold working can in-fluence the annealing operation or remainunchanged after the annealing operation.Of the two types of working, hot workingpresents the greatest potential problembecause of its ability to change thechemistry and grain structure of thestainless steel, resulting in subsequentchanges in the scale formed duringannealing.

HOT WORKING

Hot working, the process of mechanicaldeformation of a material at temperaturesabove its recrystallization temperature,can significantly alter the grain size of

metals. The degree to which this altera-tion occurs for alloys such as stainlesssteels depends on the degree of deforma-tion, the number and frequency of defor-mation steps, and the initial and finalworking temperatures. The degree of

deformation determines the stored energyin a material that is the driving forcefor recrystallization to occur. The fre-quency of deformation steps and initialand final temperatures determines therate of crystallization and the occur-rence and rate of grain growth. Both alower deformation temperature and agreater amount of deformation produce a

smaller ultimate grain size. It isusually the temperature at which hotworking is completed, the finishing tem-perature, that determines the averagegrain size (1_).

3

^Underlined numbers in parentheses re-fer to items in the list of references atthe end of this report.

When hot working operations are fol-

lowed by an anneal, the major effect of

this altered grain size is on the result-ing annealing scale. If the mill scaleconforms to the morphology of the metalsurface, then a fine-grained scale formson a fine-grained metal and a largergrained scale forms on a large-grainedmetal. Recent research (2) has shownthat the scale formed on smaller-grain-size stainless steels contains morechromium than that on larger-grain-sizesteels. Oxide films containing morechromium could significantly affect the

rate of pickling and possibly cause com-positional changes, such as chromium de-pletion, in the metal adjacent to the

oxide.

Severe problems can occur after hotworking if there is significant chemicalinhomogeneity or large quantities of in-

clusions in the initial ingot. Regionsof the ingot depleted of an alloy con-stituent could be smeared in the rollingdirection, resulting in bands of differ-ent composition (3). Another phenomenon,similar in appearance to banding, is

fiber. This elongated structure consistsof nonmetallic inclusions which are elon-gated as the steel is worked (4^) . Whilethe spontaneous recrystallization that

occurs during hot working is usually una-ble to affect these localized compositionchanges , annealing usually eliminatesthem. If these variations in metal com-position are still present after the an-

nealing step, however, preferential at-tack of the banded regions or pittingnear the fibers could result in an ir-

regular surface morphology. Since stain-less steel chemistry is closely con-trolled, these structures are rarely seen

in commercial stainless steels; if pre-sent, they would be expected to signifi-cantly affect the pickling process.

Sensitization to intergranular corro-sion may occur in austenitic stainless

steels when they are subjected to aworking temperature range of 450° to900° C or when the steel is slow-cooledfrom 1,050° C (_5_) . A chromium-depletiontheory proposes that the precipitation of

M23C6 carbides along a grain boundaryresults in a region depleted of chromiumadjacent to the carbides. In some casesof very high purity (low-carbon) alloys,although there is no detectable carbideprecipitation, a solute segregationtheory proposes that a chromium-depletedregion still exists (6) . The resistanceto the pickling solution is greatly re-duced in these regions. Essentially, twodissimilar metals are in contact and anunfavorable anode-cathode area ratio is

present (7^, pp. 35-36). In this case,the depleted zone sets up active-passivecells with large-area alloy grains actingas cathodes in contact with material inthe grain boundaries of limited area act-ing as anodes. In addition, the grainboundary carbides can be susceptible tothe mixed acid solution, and their dis-solution would produce further surface

degradation. As with the banding andfiber phenomena, however, any sensitiza-tion should be removed in a proper an-nealing step, making hot-work-inducedsensitization a minor problem area.

COLD WORKING

Cold working increases the storedenergy of a material and, when coupledwith an annealing process , can causegrain size changes similar to those dis-cussed for hot working. But, unless a

deformation-induced segregation or trans-formation occurs that brings dissimilarconcentrations of atoms into contact (8)

,

cold working has no other effect on thepickling behavior of properly annealedstainless steels. Simultaneous deforma-tion and pickling never occur in the in-dustrial processing scheme of stainlesssteels; therefore, cold working is im-

portant only if new phases or the segre-gation of alloyed components result fromthe plastic deformation.

EFFECT OF ANNEALING ON PICKLING

Compared to hot and cold working, an-nealing probably has a more significanteffect on the pickling rate of stainlesssteels than any other process precedingpickling. Annealing can be thought of as

a relatively uncontrolled oxidation re-action; that is, the temperature can becontrolled fairly accurately, but theatmosphere within the annealing furnaceis usually not controlled. Also, frombatch to batch of stainless steels, the

heat-up time, time at temperature, andcool-down time are different for variousmetallurgical reasons. This variabilityin the annealing process can have a sig-nificant effect on the mill scale formedand thus on the ease of pickling of the

stainless steel. Numerous investigationsconcerning the high-temperature oxidationof stainless steels in various atmos-pheres are available throughout the met-allurgical literature. Although there is

considerable disagreement about thestructure, thickness, and growth mechan-isms of the films, some agreement con-cerning film characteristics exists.

Because of the short annealing timesinvolved in stainless steel processing,the gaseous annealing environment reactspreferentially with the more reactivechromium component. This produces a re-gion of the base metal near the oxidethat is significantly depleted in chromi-um. Consequently, the thin oxide films

that form on these alloys during the ini-tial oxidation stages are composed mainlyof Cr

2 3 with small amounts of Fe 2 3 or a

spinel phase FeFe(

2

_ x) Cr x 4 where 0<x<2

(9-12) . These films are fairly adherentand protective, and since their ionicconductivity is low (13) , they preventdiffusive penetration of other ions and

atoms through the scale. Although the

Cr 2 3 films are adequate barriers, the

iron oxide films are quite permeable,especially to carbon (14) . Under condi-

tions of severe scaling, a stratifiedstructure occurs that contains Cr 2 3 ,

FeFe (2 _ x )Cr x 4 , FeO, Fe 3 4 , and Fe 2 3 (8_,

15). In certain cases, as in extendedheating periods, the initially protectivescales then become nonprotective.

The scaling behavior of stainlesssteels under conditions similar to thosein annealing furnaces has not been ade-quately studied. Brief exposures (1- to5-min time at temperature) in complex(02 , H20, CO, C0 2 ) high-temperature (871°

to 1,038° C) environments should be stud-ied. Complete characterization of thescales resulting from these exposuresmust be done in order to understand thechemistry and structure of the complexscales formed on stainless steels. It

would then be necessary to correlate the

ease of pickling of the steels with thechemistry and structure of the annealingscales, an area of research that couldhave significant impact on the picklingprocess as currently used.

Because of the elevated-temperatureanneal, three other metal-related phenom-ena that could have an impact on subse-quent pickling operations are embrittle-ment, sensitization, and a transformationto a dual-phase structure.

Embrittlement (_1_) may occur upon heat-ing or slowly cooling certain ferriticand high-nickel austenitic steels throughthe 425° to 760° C temperature range. Aprecipitation reaction occurs in thistemperature range producing a complexstructure FeCr phase, the sigma (a)phase. Because this phase is rich inchromium, the presence of a chromium con-centration gradient between the a-phaseand adjoining phases may affect the pick-ling behavior and corrosion resistance ofstainless alloys (8) . With proper tem-perature control, however, this phasewill not form.

Elemental concentration gradients arisein the alloy owing to diffusional pro-cesses at elevated temperatures. For in-stance, if an 18Cr-8Ni stainless steel(i.e., 304) is heated to 1,150° C (a

common heat-treatment temperature) andwater-quenched, the single-phase austen-itic alloy is formed. The quenched-inaustenitic structure is thermodynamicallyunstable at room temperature, but thetransformation kinetics are extremelyslow. If the same alloy, assuming alow carbon concentration, is heated to1,200° C and allowed to soak at that

temperature for several hours , the FCCaustenite partially transforms to fer-rite, a BCC structure. Therefore, a two-phase (austenite-ferrite) microstructurecan be present. A problem arises in that

the austenite phase contains more nickeland less chromium relative to the ferritephase (8) . Thus , an elemental concentra-tion gradient is created between adjacentgrains. The susceptibility of the alloyin this condition to localized corrosionin a mixed-acid pickling bath may be sig-nificantly enhanced. This enhancement is

attributable to a galvanic effect arisingbetween grains of different compositionin close proximity.

The final problem produced by elevated-temperature exposure of stainless steelsis sensitization. Depending upon thecarbon content of the alloy, precautionsmay have to be taken to avoid sensitizingconditions during annealing. As men-tioned previously, sensitization resultsin intergranular precipitation of M

2 3C 6carbides of high chromium content andchromium depletion of regions adjacent tothese grain boundaries to below thenecessary 12 wt pet required for stablepassivity (_7, pp. 35-36). While thiswould cause an increased localized at-tack, sensitization is rarely a problemin the production of austenitic stainlesssteel.

In summary, hot working and, to a smalldegree, cold working can have potentiallydeleterious effects on the pickling of

stainless steels. Grain-size variationscan cause variations in scale chemistry,and ingot inhomogeneity can cause bandingand fibering. Sensitization can occur inboth hot working and annealing, but em-brittlement usually occurs only duringannealing. Typical annealing processesare poorly controlled, resulting in vari-ations in scale chemistry. All of theseproblems do not necessarily occur duringpresent-day stainless steel processing,but there is a very real possibility fortheir occurrence. Therefore, in order tofully understand the pickling process andeverything that affects it , the more im-portant of these problems (grain-sizevariations and annealing variations)should be studied further.

EFFECT OF CONDITIONING ON PICKLING

Conditioning of stainless steels is aprocess used to prepare annealed stain-less steel for the pickling process. Itspurpose is to alter the annealing scalein order to reduce the time, temperature,and acid concentration used in the pick-ling process. The scale is eithercracked by thermal or mechanical means

,

or constituents in the scale are chem-ically or electrolytically altered to In-crease their solubility in order to beable to reduce either the amount of HNO3and HF in the final pickling solution orthe times and temperatures in the pick-ling process.

In contrast to the well-understoodmechanical conditioning process, thechemical and electrolytic conditioningtechniques need to be studied further inorder to understand exactly what chemicaland structural changes result from therespective technique.

DEGREASING

Although degreasing is usually done be-fore annealing stainless steels, it alsomay be necessary, prior to a conditioningstep, to remove surface contaminants suchas oil, grit, graphite, metal chips, orother foreign matter that may have beentransferred to the steel surface. De-pending on the type of foreign mattersuspected on the surface of the steel,various techniques are employed. Thesetechniques include the use of vapor de-greasing with organic solvents, water-soluble emulsifiers, chelates, and water-soluble alkaline cleaners (16) . Contam-inants on the surface of the heat-treatedsteel can often result in an oxidizedgrime during scale conditioning, whichcan severely inhibit the effectiveness ofthe subsequent pickling processes. Sur-face contamination also pollutes thepickling solutions, which reduces theireffectiveness.

Once the surface has been effectivelycleaned, conditioning techniques help to

oxidize, crack, and loosen scale to en-hance the effectiveness of final pickling

operations. Depending upon the nature ofthe oxide film, various techniques areemployed. Both mechanical and chemicaltechniques have been used successfully.The basic mechanical technique for scaleconditioning is abrasive blasting.

ABRASIVE BLASTING

Abrasive blasting is one of the fastestof all conditioning techniques and hasbeen successfully used in both batch and

continuous on-line processing of stain-less steel strip (16-18) , although it is

used almost exclusively on the latter.Abrasive blasting uses steel shot or sil-ica in sizes ranging from 100 to 2,500mesh depending on the material being con-ditioned. The abrasive is either droppedor directed by air pressure toward the

surface of the steel at angles and ve-locities that allow the cracking, loosen-ing, and partial removal of heavy oxidescales without cold-working the steel'ssurface. The technique has limited ap-plication for batch-processing complexshapes because some oxidized areas not

exposed to the shot blast are not totallyconditioned. The use of carbon steelshot as an abrasive has been a topic of

controversy (19) . Opponents feel thatits use can embed iron particles in the

surface of the stainless steel, resultingin future corrosion problems. Thiswould, of course, depend on what treat-ments followed the carbon steel blasting.

The abrasive silica used in blasting mustbe low in iron in order to avoid surfacecontamination. As with most mechanicaldescaling techniques, caution must bepracticed to avoid work-hardening the

steel surface. Heat-treated surfaceswith thin oxide films , such as cold-rolled products, are usually not condi-tioned using mechanical techniques.

CHEMICAL CONDITIONING

The choice of chemical conditioningparameters varies significantly betweenstainless steel processing plants. Some

of the parameters that must be taken into

consideration are the nature of the base

metal and oxide being removed, strip or

batch processing, and chemical costs.For example, when working with an inter-mediate pickled 304 stainless steel, somecompanies use a salt bath conditioningprocess followed by an electrolyticnitric acid bath before the final HN03-HFacid pickle; other processors go directlyfrom the salt bath into the final pick-ling bath. The following paragraphsdescribe some of the more common chemicalconditioning processes in use.

Reducing Acids

Reducing acid baths descale a metal byreducing the oxides in the scale and alsoliberate hydrogen at the oxide-metal in-terface. The most common reducing acidsused in conditioning processes for an-nealed stainless steels are H

2S0 4 andHC1. The specific acid solution used de-pends on the nature of the base materialand oxide being treated; chemical costs,as well as the time permitted for condi-tioning, are usually determined bywhether the metal is prepared by eithercontinuous strip or batch processing.Solutions of 10 to 15 vol pet H2 S04 at60° to 71° C are used for conditioningheavy oxides (20) . H 2 S04 solutions arerelatively slow in comparison to solu-tions combining H2S0 4 (6 to 10 vol pet)with HC1 (6 to 10 vol pet) at 54° to 60°

C (_21, pp. 684-689). Although HC1 is notas aggressive as H 2S0 4 in attacking thebase metal at elevated temperatures (77°

to 93° C) , it is much more aggressive inattacking iron oxides than is H2 S04 . Theuse of conditioning solutions containingHC1 requires critical control becauseferric chlorides formed during the condi-tioning process can result in pitting ofthe base material ( 19 ) . Another condi-tioning solution that can cause base met-al pitting is H 2 S0 4 (8 to 11 wt pet) com-bined with NaCl (5 to 6 wt pet) at 60° to65° C U8).

Oxidizing Acids

In contrast to reducing acids , oxidiz-ing acids descale by oxidizing the scaleto a higher oxidation state, thus in-creasing the solubility of the scale.

HNO3 (8 to 10 vol pet at 38° to 54° C) is

the most commonly used oxidizing acid

(19) . Because of the greater cost of

HNO3 compared to H 2 S0 4 , however, its usehas been limited in the conditioningprocesses.

ELECTROLYTIC ACID CONDITIONING

Experimental studies on electrolyticpickling in acid solutions were describedby Tamba, Azzerri, Bombara, and others(22-25). Industrial electrolytic condi-tioning using an H2 S0 4 bath can be tracedback to the 1920's and 1930' s where therewas a need for faster pickling techniquesto maximize output from on-line stainlesssteel strip processing (26)

.

The most common acids associated withelectrolytic conditioning are H2S0 4 andHNO3. Electrolytic acid conditioningtechniques are used presently on bothmartensitic and ferritic steels which,

during the annealing process , have devel-oped tightly adhering thin oxide films

(21 , pp. 689-690). Austenitic chromium-nickel steels can also be electrolytical-ly conditioned, although this condition-ing requires caution because of the

greater potential for surface pitting.The probability of pitting is increasedfor thicker oxide scales because breaksin the oxide scale react more rapidlythan scale-covered base metal. Thishighly localized activity can result in

pitting before the bulk of the scale is

removed

.

Anodic Conditioning

There are three basic types of electro-lytic conditioning: anodic, cathodic,and alternating current (27) . Anodicconditioning either electrically oxidizes(dissolves) the surface or forces thesurface into a passive state by an ap-plied anodic potential. When the surfaceis being electrically oxidized, the basemetal is being attacked and the scaledislodges. When the surface is passive,then the base metal undergoes far lessattack than the oxide scale, releasingoxygen gas at its surface. The oxygenmechanically agitates the solution,

uplifts the surface scale, and can oxi-dize contaminants such as organic impuri-ties. An example of an anodic electro-lytic solution is 2 pet HN0 3 used at acurrent density of 75 to 200 A/ft 2 (21,pp. 689-690; 27_)

.

Cathodic Conditioning

During cathodic electrolysis, the work-piece is charged to act as a cathode.The base metal is electrochemically pro-tected while the oxide scale is being re-duced. Cathodic electrolysis is fasterthan anodic; hydrogen gas generated onthe metal surface helps to agitate thesolution and lift off the oxides. Inmartensitic steels, however, this reac-tion can result in hydrogen embrittlementof the surface (27-28).

solution is 10 vol pet H2 S0 4 used at 88°

C with a current density of 100 to 150A/ft 2 (27).

Both ferritic and martensitic steelsundergo annealing processes which formthin, tightly "skinned" oxides. Removalof these oxides by mechanical means couldcold-work the surface. However, the useof reducing acids, such as H2SO4, whichrelease hydrogen during the alternatingcurrent descaling process when the work-piece is the cathode, can be detrimentalto martensitic steels. As in cathodicconditioning, these steels are subject to

hydrogen embrittlement. Alternative con-ditioning techniques for martensiticsteels include those in which the libera-tion of hydrogen on the steel surface is

eliminated.

Alternating Current Conditioning

Very low frequency (one cycle per sev-eral minutes) alternating current elec-trolysis is used in conjunction withstainless strip processing where directelectrical contact to the workpiece isdifficult. In fact, the current is usu-ally reversed only from tank to tank in atwo-tank method, or, in a one-tank meth-od, it is reversed once in the tank. Thebasic reactions, however, are the same asdescribed in anodic and cathodic condi-tioning. Operationally, the electrodesare placed above and below the strip,forming an anodic potential on one sideof the strip and a cathodic potential onthe other side (27) . During each alter-nating cycle, the polarity is reversed onthe electrodes and, in turn, on the facesof the stainless steel strip. A varia-tion is the use of two tanks where thestrip in the first tank is made cathodicwith respect to the anodic electrode inthe tank (26) . Here the base metal is

electrochemically protected and hydrogengas is formed to lift off the oxides.The strip then enters a second tank whereit becomes anodic with respect to the ca-thodic electrode placed in the secondtank. Here the surface and scale are ox-idized and release oxygen at the surface,which uplifts the scale. An example ofan alternating current electrolysis

ELECTROLYTIC NEUTRAL CONDITIONING

Electrolytic neutral (the Ruthner"Neolyte" process) conditioning is simi-lar in mechanical design to electrolyticacid conditioning in that alternatingcathodic and anodic electrodes are usedto polarize the workpiece and induce oxi-dation and reduction of the surface scale

(29) . The electrolytic neutral picklinginvolves a Na 2 S04 solution at tempera-tures of 65° to 85 C, resulting in saferoperation conditions and reduced energycosts (29) . The final stages in the pro-cess result in a regeneration of the

Na 2 S04 .

SALT BATH CONDITIONING

The use of both aqueous and molten salt

baths as a pretreatment for acid pickling

has been successful in increasing the

ease of scale removal of all stainlesssteel types, permitting the use of re-

duced acid concentration and decreasingthe time of pickling needed to remove the

scale Q6, JLjU 30-31). This reduction in

acid concentration and pickling time re-

duces the likelihood of hydrogen embrit-tlement in susceptible alloys. Otherbenefits of salt bath descaling are its

fast process time and uniform surfacefinish (19).

Salt bath treatments can be either re-ducing, oxidizing, or electrolytic (16,30 ) . Reducing and electrolytic bathshave proven to be more effective in at-tacking heavy, tightly adhering oxidescales.

Reducing Bath

Reducing salt baths consist of a moltenbath of NaOH (371° ±11° C) , containing1.5 to 2.0 wt pet sodium hydride (16,30 ) . The sodium hydride is formed ingenerators along the side of the descal-ing tank. Sodium and hydrogen gas reactto produce the hydride, which must beconstantly generated since the level of

sodium hydride in solution is depletedduring descaling. There are safety haz-ards involved in the use of sodium andhydrogen. Descaling takes place as theoxide films are reduced to a lower oxida-tion state:

M2 3 + xNaH = xNaOH + M2 3 . x .

After the workpiece is removed from thesalt bath, it is water-quenched. Thisquenching thermally shocks the remainingoxide layer, fracturing and loosening itand making it more susceptible to theacid pickling.

Oxidizing Bath

As with the previously described caus-tic reducing salt bath, oxidizing saltbaths also use molten NaOH (60 to 90 wtpet) at elevated temperatures (482° C)

(16, 30) . Other constituents includesodium nitrate (7 to 32 wt pet) and so-dium chloride (1.5 to 6 wt pet). Some ofthe oxidizing reactions that take placeare (31)--

2Fe0 + NaN03 = Fe 2 3 + NaN0 2

Cr 2 3 + 3NaN0 3 = Cr 2 6 + 3NaN0 2

2Ni0 + NaN03

= Ni 2 3+ NaN0 2

C + 2NaN03

= C0 2 + 2NaN0 2

As long as the bath is in contact withair, the NaN0

3 can be regenerated by theoxidation of the NaN0

2 formed. Trivalent

chromium is oxidized to the hexavalentstate, which readily dissolves into acid

pickling solutions.

Upon leaving the salt bath, the work-piece is quenched with water. The ther-mal shock involved in quenching the work-piece disrupts the oxidized scale. Afterthe water quench, an acid pickling bathis needed to dissolve any remainingoxide.

An aqueous oxidizing salt bath consistsof NaOH (20 wt pet) and KMn0 4 (5 wt pet)

(32) . This strong oxidizing solution maybe used on all grades of stainless steel,

especially where a light, tightly adher-ing scale has formed, such as is the casein cold rolling or in some controlledannealing atmospheres. The KMn0 4-NaOHsolution may also be used on heavier ox-ide scales. As with the other previouslymentioned molten salt baths , the treat-ment is followed by acid pickling.

Electrolytic Bath

The use of an electrolytic technique incombination with molten salt baths has

been applied successfully to continuousstainless steel strip processing lines.A typical bath consists of 75 pet NaOH,10 pet NaCl-NaF, 14 pet Na 2C03 , and 1 petother carbonates and is operated at 482°

C. The salt baths are neither oxidizingnor reducing in nature until electricallypolarized. Polarization of the workpieceis achieved by the use of a series of

cathodes followed by anodes in the neu-tral salt bath (_16, 30). Opposite the

cathode, the strip acts as an anode withoxidation taking place. As some of thescale is dissolved into solution, itreacts with the oxygen released duringthe oxidation process to form insolublehydroxides. These hydroxides will notinterfere with the surface of the steelstrip as it becomes cathodic and under-goes reduction. The strip becomes ca-

thodic as it passes below the anode.As reduction takes place at the scale-surface interface, hydrogen evolves,lifting the scale off the surface. Afterthe sample passes through the salt bath,

it is water-rinsed and acid-pickled for

final scale removal and surface finish.

10

EFFECT OF PICKLING BATH VARIABLES ON PICKLING

The pickling of stainless steels re-quires three distinct processes. Thefirst process is the removal of thethermally grown oxide scale for appear-ance purposes and to facilitate furthercold working of the steel. The secondprocess maximizes the corrosion resist-ance of the final steel product by com-pletely dissolving the chromium-depletedzone that is generally formed duringshort high-temperature anneals in oxidiz-ing environments. The third process dis-solves the minimum amount of bulk steelnecessary to give the desired whiteningeffect. These three processes occur, to

some degree, simultaneously during thepickling operation and most probably areinterdependent. Therefore, to understandthe pickling operation, it is necessaryto understand the effect of pickling bathvariables, solution composition, and tem-perature on the pickling operation as awhole and on the dissolution of both thechromium-depleted region and the bulksteel.

PICKLING OPERATION

Before considering the effect of bathvariables on the pickling of stainlesssteels, it is necessary to assess thestate of understanding of the mechanismof pickling. It is generally agreed inthe literature (24) that the pickling of

stainless steels is accomplished by un-dercutting the oxide and that the highchemical reactivity of the chromium-depleted zone essentially controls therate of the pickling operation. Therewere, however, no in-depth studies of themechanism or even any proof found in thebody of literature reviewed here. Whileit was difficult to determine how thismechanism was discovered, it is clearthat little is known empirically aboutthe mixed acid pickling of stainlesssteels. Other research indicates thatdissolution of the oxide is inconsequen-tial in the actual removal of the oxide(23) . What appears to be important is

that the oxide be sufficiently disrupted

to permit the penetration of the picklingsolution. Most changes in oxide porosityoccur during the conditioning process.

The ideal pickling solution is one thatwill easily penetrate the thermally grownoxide, rapidly dissolve the chromium-depleted zone, and dissolve only a smalllayer of the bulk steel. This solutionwould optimize the rate of pickling whileminimizing the unnecessary and excessiveloss of bulk alloy. The most commonlyused pickling solution for austeniticstainless steels is a mixed acid consist-ing of HN03 and HF in various propor-tions. By varying the ratio of HN0 3 toHF, it is possible to pickle differenttypes of austenitic steels. This solu-tion accomplishes all three processeslisted above and provides a very whitesurface. There are, however, potentialproblems of overpickling the steel, re-sulting in excessive grain boundary at-tack. This could affect the reflectivityof the surface and decrease the corrosionresistance. This mixed acid is also ex-pensive and extremely dangerous. A pos-sible alternative to HNO3-HF pickling wasreported in a series of research articles(22-25) . This research resulted in a

technique (24) that used an applied po-tential to dissolve the chromium-depletedzone while minimizing the dissolution of

the base material in an H2 S04 bath. Theresearchers claim that a proper choice of

potential can optimize the pickling of

stainless steels while using the lesshazardous H2 S0 4 solution. However, this

technique does not produce as reflective

a surface as produced in free (not poten-tial controlled) pickling in HNO3-HF mix-tures. The researchers also did not in-

vestigate the effects of iron, chromium,and nickel buildup in the bath on the po-

tential controlled pickling or the quali-

ty of the pickled surface. As in HNO3-HFsolutions , these impurities could affect

the electrochemical reactions and the

rate of pickling, and could even lead

to pitting or increased intergranularattack.

11

Other pickling solutions have been ex-amined in previous research efforts. Inone study (33) the solutions were select-ed for their ability to dissolve the bulksteel. This research reported on the re-

lationship of the composition of thepickling bath to the dissolution rate of

304 stainless steel. Using various com-binations of H 2S0 4 , HC1, HN03 , and NaN03 ,

it was shown that HC1 and HNO3 have aboutan equal effect on the dissolution rateof scale-free 304 SS, while H2S0 4 has a

much lower effect and NaN03 has a muchgreater effect (NaN0 3>HCl«HN03 >H2 S04 )

.

The results of this study assumed thateach chemical in the pickling bath canoperate individually with no synergisticeffect. However, the same group of chem-icals produced different results on an-nealed and oxide-covered 304 stainlesssteel. The NaN0 3 solution had the slow-est pickling rate, while HC1 had thefastest pickling rate. Both H

2S04 andHNO3 had intermediate rates. This sug-gests that while dissolution may be im-portant in the mechanism of pickling,dissolution of the bulk alloy may not becritical. When solutions containing HFwere tested, results showed that it hadas strong an effect on pickling rate as

HC1. HC1 is usually not used because theFeCl

3 formed generally promotes pittingof the stainless steel. No studies of

the effect of temperature on the picklingof stainless steels were found in thissearch of the literature.

BULK ALLOY DISSOLUTION

Solution Composition

The aforementioned report (33) suggest-ed that dissolution of the bulk steel maynot be critical in determining the rateof pickling of stainless steels. Thedissolution rate of the bulk steel does,however, control the amount of bulk steellost to the pickling solution and to someextent the amount of dissolved metalspecies in solution. The factors thatcould affect this dissolution rate aretemperature, acid concentration, dis-solved metal concentration, and agita-tion. Since many dissolution reactionsexhibit some dependence on convection and

diffusion, agitation of the solutionshould be important. A study (34) wasdone that showed that agitation, as simu-

lated by a rotating disk electrode, sig-nificantly affects the dissolution of a

304-type stainless steel (Khl8N10T) in

HNO3 + NaCl solutions. Dissolution ratesfor samples rotated at to 300 rpm were

not affected, whereas rates increasedsteadily with increasing rotation speedfrom 300 to 10,000 rpm. The investigatorassumed that the agitation affected main-ly the cathodic reduction of nitric to

nitrous acid, but offers no evidence.Another investigator has shown (35) , to

the contrary, that the anodic reaction of

nitric acid on platinum is diffusion de-pendent while the cathodic reaction is

only reaction dependent.

Acid Concentration

The effect of HN03 concentration on the

dissolution of austenitic stainlesssteels is minimal over a range of concen-trations and temperatures that would beused in pickling solutions. Isocorrosiondiagrams (7_, p. 243) for 18-8S4 steelsshow that the corrosion rate ranges from

to 125 um/yr for HN03 concentrations upto 50 wt pet from 30° C to the boilingpoint. The dissolution behavior of sev-eral of the more popular austeniticstainless steels and of some iron-chromium-nickel alloys has been thorough-ly reviewed ( 36 ) elsewhere. The resultssimply show that HNO3 solutions (<50 pet)do not rapidly dissolve austenitic stain-less steels.

Mixed Acids

To dissolve the austenitic stainlesssteels , a mixed acid is usually consid-ered. Mixtures of HNO3 and H2 S0 4 in-crease the dissolution rate of the steelsby about a factor of four over HNO3 alone

(7^, p. 243); however, this combination is

rarely used in pickling operations. Themost commonly used solutions for pick-ling austenitic stainless steels containHNO3 and HF. The effect of this mixed

^"S" means lower carbon content to pre-vent carbide precipitation.

12

acid on various austenitic steels hasbeen extensively studied and adequatelyreviewed (37) . HF significantly in-creases the rate of dissolution of aus-tenitic steels in HN0 3 solutions. Forexample, as little as 0.5 wt pet HF in18 wt pet HN0 3 can increase the dissolu-tion rate of 304 stainless steel by twoorders of magnitude at 60° C and by al-most three orders of magnitude at 80° C

(37) . This corrosion rate is approxi-mately 15,000 um/yr at 80° C.

Temperature

The above results suggest that tempera-ture has a significant effect on the dis-solution of austenitic stainless steels.A study ( 37 ) conducted on 309SCb 5 stain-less steel indicated that a temperaturechange from 20° to 100° C increased thedissolution rate in HNO3-HF solutions byover two orders of magnitude. This re-sponse to increasing temperature wasshown to be true regardless of solutioncompositions from 4.5 wt pet HNO3 + 0.2wt pet HF to 27 wt pet HNO3 + 2 wt petHF. These results imply that the activa-tion energy for dissolution is not af-fected by solution composition, althoughthe absolute magnitude of that dissolu-tion rate was found to be affected by so-lution composition.

CHROMIUM-DEPLETED ZONE DISSOLUTION

The dissolution behavior of thechromium-depleted zone should be affected

by the same factors as those affectingthe bulk steel: temperature, acid con-centration, and agitation. Dissolutionrates should be higher than those for thebulk steel because of the reduced chromi-um and nickel contents. Direct studiesof the dissolution behavior of this de-pleted zone have not been conducted be-cause of the extreme thinness of this

zone. Indirect studies, however, havebeen done by fabricating alloys that sim-ulate different regions of the depletedzone. A study (38) of a series of iron-chromium-nickel alloys (2 to 18 wt pet

Cr) showed that the dissolution rate insulfuric acid was high for very low con-centrations of chromium. Surprisingly,however, the dissolution rate passedthrough a minimum at 12 wt pet Cr wherethe dissolution rate was only 40 pet of

that for a 19 wt pet Cr alloy. Studiesof similar alloys in HNO3-HF would be

very important for developing an under-standing of the pickling of stainlesssteels. Equally important would be stud-ies of the effect of acid concentration,temperature, and agitation on the disso-lution behavior of these chromium-depleted alloys.

RESEARCH NEEDS

Based on this review of the literatureon the pickling of stainless steels, thefollowing areas have been identified asneeding further study:

3. The effect of the various condi-tioning techniques on the structure and

composition of the scale related to its

effect on pickling rate.

1. The effect of hot working on thegrain size of metal and subsequent oxidescale composition.

2. The effect of annealing parameterssuch as furnace environment and time attemperature on the annealing scale andultimately on the pickling rate.

S" means lower carbon to prevent car-bide precipitation, and "Cb" means nio-bium (columbium) added to prevent carbideprecipitation

.

4. The dissolution behavior of bulksteels and the chromium-depleted zone of

bulk steels, and the effects of the pick-le bath variables, acid concentration,temperature, and agitation, on the disso-lution rate.

In conclusion, additional studies in

any one of the above four areas will con-

tribute to the understanding of the pick-ling process. However, knowledge fromall four areas is necessary to develop

the relationships needed to quantify the

13

pickling process. This quantificationshould lead to an improvement in the ef-ficiency of pickling, a reduced cost toprocess stainless steels, and minimizingthe loss of critical metals such as nick-el and chromium. Laboratory studies are

presently being done to understand the

effect of acid concentration, tempera-ture, and dissolved metal concentrationon the pickling of 304 and 430 stainlesssteels.

REFERENCES

1. Camp, J. M. , and B. C. Francis.The Making, Shaping, and Treating of

Steel. U.S. Steel Co., 6th ed. , 1951,

pp. 580-586.

10. Yearian, H. J., E. C. Randell, andT. A. Longo. The Structure of OxideScales on Chromium Steels. Corrosion,v. 12, 1956, pp. 515t-525t.

2. Baer, D. R. , and M. D. Merz. Dif-ferences in Oxides on Large and Small-grained 304 Stainless Steel. Metall.Trans., v. 11A, 1980, pp 1973-1980.

3. Bramfitt, B. L. Effect of HotRolling on the Dendrite Texture of Di-rectionally Solidified Steel Ingots.Metall. Trans., v. 1A, 1970, pp 2495-2505.

4. Keissling, R. Non-Metallic Inclu-sions in Steel. The Iron and SteelInst., London, I.S.I. Publ. 115, pt. 3,

1948, pp. 51-73.

5. Binder, W. , C. Brown, and R.

Franks. Resistance to Sensitization ofAustenitic Chromium-Nickel Steels of0.03% Maximum Carbon Content. Trans. Am.Soc. Met., v. 41, 1949, pp. 1301-1302.

6. Hanninen, H. , and E. Minni. OnGrain Boundary Segregation in AusteniticStainless Steels. Metall. Trans, v. 13A,1982, pp. 2281-2285.

7. Fontana, M. G. , and N. D. Greene.Corrosion Engineering. McGraw-Hill,1978, pp. 35-36, 243.

8. Uhlig, H. H. Effect of Metal Com-position and Structure on Corrosion andOxidation. Corrosion, v. 19, 1963, pp.231t-237t.

9. Seybolt, A. V. Observations on theFe-Cr-0 System. J. Electrochem. Soc. , v.

107, 1960, pp. 147-156.

11. Edstrom, J. 0. Scaling of 18-8

Stainless Steel on Reheating Furnace At-mospheres. J. Iron and Steel Inst., v.

185, 1957, pp. 450-466.

12. McCullough, H. M. , M. G. Fontana,and F. H. Beck. Formation of Oxides onSome Stainless Steels at High Tempera-tures. Trans. Am. Soc. Met., v. 43,

1951, pp. 404-425.

13. Hoar, T. P. Anodic Behavior ofMetals. Sec. in Modern Aspects of Elec-trochemistry, ed. by J. 0. M. Bockris.Butterworths , v. 2, 1958, pp. 290-291.

14. Meier, G. H. , W. C. Coon,and R. A. Perkins. Corrosion of Iron-,Nickel-, and Cobalt-Base Alloys in At-mospheres Containing Carbon and Oxy-gen. Oxid. Met., v. 17, No. 3/4, 1982,pp. 235-262.

15. Fabis, P. M. , R. H. Heidersbach,C. W. Brown, and T. Rockett. Oxide ScaleFormation on Iron-Chromium Alloys in Ele-vated Temperature Air Environments. Cor-rosion, v. 37, 1981, pp. 700-715.

16. Wood, W. G. The Cleaning ofStainless Steels. Sec. in Handbook of

Stainless Steels, ed. by D. Peckner andI. M. Berstein. McGraw-Hill, 1977, pp.35-1 to 35-16.

17. Spencer, L. G. Descaling theStainless Steels. Met. Finishing, v. 52,

1954, pp. 54-59.

14

18. McFee, W. E. Pickling StainlessTo Remove Scale. Steel, v. 141, 1977,

pp. 103-107.

19. AISI Committee of Stainless SteelProducers. Cleaning and Descaling Stain-less Steel. AISI, Washington, DC, De-signers' Handbook Series, 1982, 27 pp.

20. Gurklis, J. A., and L. D. McGraw.Pickling and Descaling Stainless Steelsand High Temperature Alloys. Met. Prog.,v. 83, No. 6, 1963, pp. 95-96.

21. Mark, H. (ed.). Pickling of

Steel. Ch. in Kirk-Othmer Encyclopediaof Chemical Technology. Wiley, 2d ed.

—

suppl. vol., 1971, pp. 684-689, 689-690.

22. Vincentini, B., and G. Bombara.On the Mechanism of Scale Removal in theAcid Pickling of Austenitic StainlessSteels. Electrochim. Met., v. 3, 1968,

pp. 313-322.

28. American Society For Metals Com-mittee on the Pickling of Iron and Steel.Pickling Iron and Steel. Sec. in MetalsHandbook, ed. by T. Lyman. ASM, v. 2,

8th ed., 1964, pp. 346-355.

29. Braun, E. How To Improve Picklingof Stainless Steel Strip. Iron and SteelEng., v. 57, No. 4, 1980, pp. 79-81.

30. American Society for Metals Com-mittee on Salt Bath Descaling. Salt BathDescaling. Sec. in Metals Handbook, ed.

by T. Lyman. ASM, v. 2, 8th ed. , 1964,

pp. 356-363.

31. Shoemaker, R. H. No Acid- or LowAcid-Pickling of Stainless Steel WithoutPollution Problems. Paper in Proceedingsof the 17th Mechanical Working and SteelProcessing Conference (Pittsburgh, PA,

Jan. 22-23, 1975). Mechanical Workingand Steel Processing Division of ISS-AIME, 1975, pp. 42-49.

23. Bombara, G. , A. Tamba, and N. Az-zerri. Potentiostatic Anodic Pickling ofStainless Steels. J. Electrochem. Soc.

:

Electrochem. Technol. , v. 118, 1971, pp.676-681.

24. Tamba, A., and N. Azzerri. AnodicPickling of Stainless Steels in SulphuricAcid. J. Appl. Electrochem., v. 2, 1972,

pp. 175-181.

25. Azzerri, N. , and A. Tamba. Poten-tiostatic Pickling: A New Technique ForImproving Stainless Steel Processing. J.

Appl. Electrochem., v. 6, 1976, pp. 347-352.

26. Geiser, J. D. Electrolytic Pick-ling of Stainless Steel Strip. BlastFurnace and Steel Plant, v. 27, 1939, pt.

1, pp. 258-261; pt. 2, pp. 258-261; pt.

3, pp. 352-361; pt. 4, pp. 487-520.

27. Shrier, L. L. (ed.).Acid. Sec. in Corrosion.1963, pp. 12.16-12.23.

Pickling inWiley, v. 2,

32. Spencer, L. F. Scale RemovalTechnique With the Stainless Steels.Steel Processing, v. 36, 1950, pp. 623-

628.

33. Frantiu, I. (Corrosion Rate andPickling Time of Austenitic StainlessSteel in Acid Solutions). Metallurgia,v. 2, 1975, pp. 79-80 (British Industrialand Scientific International TranslationService, Transl. B151 14195).

34. Zhursolev, V. K. , M. Kurtepov, andV. I. Oreshkin. (Effect of Agitation onthe Corrosion and Electrochemical Beha-vior of Khl8N10T Stainless Steel inNitric Acid Solutions With a ChlorineIon Addition). Protection of Metals,

v. 8, 1972, pp. 161-164 (translated fromZashchita Metallov, v. 8, No. 2, 1972,

pp. 183-187).

35. Vetter, K. J. ElectrochemicalKinetics. Academic, 1967, pp. 490-493.

15

36. Wilding, M. W. , and B. E. Paige.Survey on Corrosion of Metals and Alloysin Solutions Containing Nitric Acid. Al-lied Chemical Corp., ICP-1107, 1976, 56

pp.

37. Cole, H. S. Corrosion of Aus-tenitic Stainless Steel Alloys Due to

HNO3-HF Mixtures. Allied Chemical Corp.,

ICP-1036, 1974, 42 pp.

38. Kaneko, S., Y. Inoue, M. Komori,and H. Sunaga. (Electrolytic Descalingof Austenitic Stainless Steels). Tetsuto Hagane', v. 59, 1973, p. 5588 (Brutch-er Translations, Transl. 9394).

INT.-BU.OF MINES, PGH., PA. 27632

HA9- 85

\3 'a.,* A,^ OSt* •'Ussw.tr + (P .*

L

4* ^ 0° *W?'of

^•>^.'> >*,-fc\ 6°.i^> X.-^b-^ c°*.-A^ °"

. , , • ftV o * „ „ o o "V * " •> * ^

\ /&&>> /VjS&X /*$&>* A5S&-S

V A

"I Pa.

Pj. *CH0' .0-

» ^ \* „4P*

F % •:a)"^<- J

£. * Jew's «.r cv

• AV^- :

<P*$fcS ,/\.^X £*iak?°» /'smkS **.*

4 Pa.

*bV°

C,vP

» ^ V

V** •£ K:V -Mfe. %<** .*«*•- \/ -'M£', %S .-Mfe*.vv v>vf * AV"^A : '

a0V t' lAf* O, aV c V»» '«*a

A-^i.% ^^mk°<> /^w'\ c°**c^:- °o ^.^^^>^6*«*<?* . ?^ A°*

irV v>V

^

>v^

V Si

^ °^.

-» o

» * * °- > v %

<

/ %^^r^\^ V'*^%o0 V^^\/ \-..7.«

**»<?