volts, 20 seconds, aluminum cathode

TABLE VII Reactions of Minerals Conducting minerals Elements sought Electrolyte Volts T, sec. Specific reagent Color of print Bismuth (Native) Bi HC1 (1:20) 4 30 ΚI—Cinchonine Orange Galena Pb S Acetic acid 5% NaOH (cathodic reduction) 4 4 30 15 KI—Sn CI, SbCl3 + HC1 Yellow-orange Orange Pyrites, \ Marcasite > Pyrrhotine j

Fe S HC1 or HN03, (1:20) 5% NaOH (cathodic reduction) 4 4 20 15 K4Fe(CN)e SbCl3 + HC1 Blue Orange Millerite Ni S NH4OH 5% NaOH (cathodic reduction) 4 4 10 15 Dimethylglyoxime SbCl3 + HC1 Red Orange Pentlandite Fe Ni S HC1 or HNO3, (1:20) NH4OH 5% NaOH (cathodic reduction)

4 4 4

30 15 14

K4Fe(CN)6 Dimethylglyoxime SbCl3 + HC1

Blue Red Orange Linnaeites Co Ni Fe Cu S

5% KCN NH4OH HC1(1:20) NH4OH 5% NaOH (cathodic reduction)

8 8 8 8 "8

30 15 30 30 15

(Direct print) Dimethylglyoxime Chromotropic acid a-Benzoinoxime SbCl3 + HC1

Yellow-orange Red Green Green Orange Chalcosine Covellite Cu s NH4OH (dilute) 5% NaOH (cathodic reduction) 4 4 5 15 Rubeanic acid SbCl3 + HC1 Dark green Orange Bornite Chalcopyrite Gray Copper

Cu Fe S Cu

NH4OH HC1 (1:20) 5% NaOH (cathodic reduction) NH4OH

4 4-8 4 4-8

15 30 30 15

Rubeanic acid Chromotropic acid SbCl3 + HC1 Rubeanic acid Dark green Green Orange Dark green 216 Η. ΛΥ. HERMANCE AND Η. V. WADLOW

Ag HN03 (1:4) 8-12 60-180 Reducing agent Black Sb 10% Tartaric acid + H3P04 8-12 60 Methyltrioxyfluorone Red As HC1 (1:1) 8-12 60 SnCl2 + HC1 Brown S 5% NaOH (cathodic reduction) 8-12 30 SbCl3 + HC1 Orange Mispickel Fe HC1 or HN03 (1:20) 4-8 30 K4Fe(CN)6 Blue Danaite Co NH4OH 8-12 60 Rubeanic acid Yellow-brown As NH4OH + H202 (5:1) 4-8 30 AgN03 Brown S — — — NaN3 + I2 — Cobaltite Co NH4OH 4-8 30 a-Nitroso-j8-Naphthol Brown As NH4OH + H202 (5:1) 8 30 AgN03 Brown S — — — NaN3 + I2 — Gersdorffite Ni NH4OH 4 15 Dimethylglyoxime Red Ullmannite As Η CI (1:1) 4-8 30 SnCl2 + HC1 Brown Sb Tartaric acid + Η NO 3 8 60 Methyltrioxyfluorone Red s 5% NaOH (cathodic reduction) 8 30 SbCl3 + HC1 Orange Lollingite Fe Η CI or HNO3 (1:20) 4 30 K4Fe(CN)6 Blue As HC1 (1:1) 4 30 SnCl3 + HC1 Brown Smaltite-Co NH4OH (dilute) 8 30 a-Nitroso-j8-Naphthol Brown Chloanthite Ni NH4OH 8 15 Dimethylglyoxime Red Safflorite-Fe HC1 or HNO3 (1:20) 8 60 K4Fe(CN)6 Blue Rammelsbergite As NH4OH + H202 (5:1) 4 30 AgN03 Brown Nickeline Ni NH4OH 4 10 Dimethylglyoxime Red Breithauptite As HC1 (1:1) 4 30 SnCl2 + HC1 Brown Sb Tartaric acid + H3P04 4 30 Methyltrioxyfluorone Red Magnetite Fe HC1 or HNO3 (1:10) 4-8 30 K4Fe(CN)6 Blue Ilmenite Fe HC1 or HNO3 (1:10) 12-16 60 K4Fe(CN)6 Blue Ti 25% H2S04 + H3P04 16 180 Chromotropic acid Red-brown

ELECTROGRAPHY AND ELECTRO-SPOT TESTING 217

218 Η. W. HERMANCE AND Η. V. WADLOW

the print. I n general, Hiller has used low voltages, short printing periods and sensitive reagents. T h e printing was done largely on gelatin coated

papers.

T h e mineral specimen is provided with a ground face, from which t h e print is made. W h e n there is sufficient electrical continuity t h r o u g h the mineral, contact m a y be m a d e b y mounting, face u p , in a bed of crumpled aluminum foil or in a low melting alloy such as Wood's metal (see p . 173). Hiller discusses several manipulative techniques which m a y be used when t h e conducting mineral is more or less isolated in a nonconducting mass. T h e mineral m a y exist as a vein which emerges a t an u n k n o w n point on t h e unground surface of t h e specimen. I n such case t h e electrographic assembly is completed, b u t t h e surface is explored with a probing needle contact until t h e milliammeter indicates establishment of a circuit, when t h e print is completed.

I n the case of mineral grains which are completely isolated, t h e printing operation m u s t be localized so as t o permit electrical connection to t h e grain t o be m a d e with a small probe. Hiller uses a small spatula-shaped electrode with which he holds t h e printing medium against a portion of the grain with one hand, while the probing contact is manip­

ulated with the other. Where t h e mineral inclusions are very small, Hiller has devised a micro electro-spot testing technique in which a tiny square of #575 paper, held under a needle electrode, is brought in contact with t h e grain under a binocular microscope.

W h e n t h e specimen contains b o t h highly and poorly conducting minerals, it m a y be difficult t o m a i n t a i n sufficient current density through t h e poorly conducting mineral t o produce a print. I n such cases Hiller covers t h e good conductor with a film of lacquer.

4-8.5. Quantitative Applications. T h e q u a n t i t y of metal dissolved in t h e electro-transfer can be brought under fairly close regulation. I t is only natural, therefore, t h a t t h e colorimetric possibilities of t h e m e t h o d should h a v e received some exploration. T h e rotating d r u m m e t h o d of Fritz (8), while applied primarily t o t h e s t u d y of color-reaction sen­

sitivities, involved t h e application of F a r a d a y ' s law t o t h e solution a n d transfer of metal ions a n d t h e relation of t h e quantities t h u s estimated t o the intensity of t h e line traced on t h e reagent paper. T h e conventional electro-transfer m e t h o d was given q u a n t i t a t i v e study b y Glazunov a n d Krivohlavy in a special application t o iron-nickel alloys (25). T h e original article contains a r a t h e r thorough discussion of t h e theoretical basis for t h e m e t h o d b u t leaves m u c h unsaid concerning experimental details, particularly t h e technique of comparing print densities a n d t h e precision obtained. I t would appear t o t h e present authors t h a t m u c h work remains before a generalized q u a n t i t a t i v e technique can be realized.

ELECTROGRAPHY AND ELECTRO-SPOT TESTING 2 1 9

It is hoped t h a t t h e equipment, circuits, a n d techniques described in this chapter m a y facilitate t h e controlled operations necessary t o systematic q u a n t i t a t i v e investigations. T h e usefulness of semiquantitative methods for routine control inspections should easily justify such effort.

The q u a n t i t a t i v e approach involves: (a) estimation, from t h e color density of t h e print, of t h e q u a n t i t y of an alloy component dissolved, and (b) estimation, t h r o u g h application of F a r a d a y ' s laws of electrolytic solution, of t h e q u a n t i t y of t h e alloy dissolved.

Approximate determination of certain metals b y comparison of spot densities on reagent-impregnated papers h a s been described b y Yagoda (45) and b y Clarke a n d H e r m a n c e (7), using independent techniques.

In each case, it was concluded t h a t differences of a b o u t 1 0 % could be distinguished within t h e favorable range of concentration of t h e metal ion. Unless more sensitive methods of spot comparison are developed, therefore, it would appear doubtful t h a t a n y t h i n g beyond a semiquantita-tive order of accuracy is possible from direct examination of electro-transfer prints. Greater accuracy m a y be obtainable b y ashing t h e transfer and determining t h e metal b y conventional micromethods (see ref. 7), using t h e electro-transfer solely as a means of obtaining a

" w e i g h e d " microsample. However, m u c h of t h e rapidity of t h e method would be sacrificed in such a procedure. On t h e whole, it would appear t h a t the chief value of semiquantitative electro-transfer methods would be in their use for classifying alloys a n d for estimating minor components rather t h a n for obtaining precise information.

Assuming t h a t a given metal ion m a y be estimated satisfactorily from t h e print density, its q u a n t i t y in t h e print can be related t o t h e alloy composition only if t h e components go into solution in t h e propor-tions in which t h e y occur in t h e alloy. Glazunov points out t h a t this condition is rigorously met only when t h e alloy is a solid solution, presenting an equipotential surface t o t h e transfer medium. If more t h a n one phase is present, each phase will h a v e its own solution potential a n d t h e yield of t h e ion under test will be governed b y the individual compositions a n d relative rates of solution of t h e separate phases. As t h e potential impressed on a polyphase specimen is raised, however, t h e phase potential differences exert diminishing control over t h e solution rates of t h e components. B y employing experimentally established conditions for t h e electro-transfer, it is often possible t o approximate t h e composition of such alloys on a n empirical basis.

Glazunov and Drescher (23) have reported such a semiquantitative method for lead in lead-tin alloys t h a t are mechanical mixtures. They use an electrolyte consisting of 1.85 g. K N 0³ a n d 1.35 g. K I in 100 ml., applying 6 volts (aluminum cathode) for 20 seconds. An empirical lead

220 Η. W. HERMANCE AND Η. V. WADLOW

iodide color scale was prepared from known specimens which, they report, permits the estimation of lead contents ranging from 0.25 to 1 0 % .

Estimation of the dissolved alloy: Assume a binary alloy to consist of a solid solution of y in ζ and t h e proportion of y is to be determined. A print is obtained with a suitable color reagent for y, t h e density of which will depend on t h e q u a n t i t y of alloy dissolved and t h e proportion of y contained in it.

n xWa

where Z>, the print density is expressed in terms of the color produced per microgram of y/sq. cm., ζ is the fractional content of y in the alloy, W^a is t h e weight, in micrograms, of t h e alloy dissolved, and A is the print area in square centimeters. Then, applying the F a r a d a y law,

n xltEa D =

where / is t h e current, in milliamperes, t, t h e time, a n d E^a is the electro­

chemical equivalent of the alloy in micrograms per millicoulomb. Since in this t y p e of alloy, the metals y and ζ dissolve proportionately, E^a is the sum of the electrochemical equivalents of y and z, each multiplied by its fractional content in t h e alloy, hence,

D - U x { x E" Y ~ X ) E' \ or = z*E, - x>E> + xEz (3)

E q u a t i o n (3) is general for binary alloys. T h e quantities E^v a n d E² are constants. D^} A, I, and t are measurable experimentally, x, T h e fractional content of component y, then is obtained by solving the result­

ing quadratic equation. T h e equation applies to single phase alloys, dissolved a t current densities low enough t o permit all of t h e electrical energy to be utilized in the solution of the metals. With two phase binary alloys, the results will depart more or less from the theoretical, depending on the magnitude of the phase potential differences a n d the loss of electrolytic efficiency attending the increased current density necessary to minimize such differences.

Measurement of J · t: T h e current is never entirely constant during the transfer, the electrolysis being a t t e n d e d by changes in t h e concentration a n d composition of the electrolyte with precipitation of products in t h e paper t h a t m a y block the electrolytic p a t h s and increase t h e resistance.

Changes in the ρ Η at anode and cathode also m a y exert considerable influence on the current. However, by careful choice of electrolyte a n d the use of sensitive reagents requiring only a light a t t a c k of t h e alloy, these changes m a y be held to a minimum. Passivation of the specimen

ELECTROGRAPHY AND ELECTRO-SPOT TESTING 221 surface or its coating by precipitated films m u s t be avoided. Unsized papers containing " f i x e d " reagents (see Table I I I ) , combined with a thick backing p a d are helpful in holding t h e current constant. P r e ­ cipitated products t h e n show less tendency t o clog t h e paper a n d t h e augmented reservoir of electrolyte is less affected b y concentration changes. Buffered electrolytes hold a constant p H at t h e specimen surface a n d facilitate a uniform rate of solution. W i t h m a n y organic reagents this is necessary, for excessive anodic acidity would prevent complete reaction. If t h e printing p a d extends beyond t h e specimen edges, some of t h e current will flow outward t h r o u g h t h e surrounding uncompressed pad. This produces an increased density at t h e periphery of t h e print a n d a corresponding increase in t h e measured current. Appre­

ciable errors m a y result. T h e simplest remedy is t o employ a p a d cut t o conform t o t h e specimen.

Fig. 17 shows a current-time curve u n d e r conditions suited t o q u a n t i ­ tative printing. After t h e first 5 seconds, t h e current remains prac­

tically unchanged for t h e 60 second observation period. M e a s u r e m e n t οϊ I - t for very short periods would involve some error due to t h e very rapid initial drop. However, it is probable t h a t even then t h e error would not exceed 1 0 % . For printing periods of 15 seconds and over, t h e error would be insignificant for semiquantitative work. U n d e r less favorable conditions, t h e current m a y continue t o fall, b u t after t h e initial sharp drop, the rate is usually almost linear and it is possible to effect a compensation by manipulation of the circuit rheostat or poten­

tiometer. In general it is advisable t o employ a current density as low as is consistent with satisfactory solution of t h e specimen surface. In this way the extremely short printing periods m a y be avoided. Because of t h e difficulty of following simultaneously b o t h stop watch a n d milliam-meter, it will be found convenient t o time t h e printing b y means of a metronome adjusted t o a 1 second beat. T h e eye is t h e n free t o give full attention to the milliammeter a n d readings can be m a d e at regular intervals.

Measurement of D: T h e density of t h e u n k n o w n print is estimated by comparison with a series of standards, similarly prepared from t h e pure metal or from known alloys. Q u a n t i t a t i v e printing is best done on a relatively thick, fine-textured paper, such as C.S. & S. #598 and the reagent impregnation should be fairly light. T h e n the colored reaction product will tend t o coat t h e fibers in a uniform, thin layer which will extend progressively deeper into t h e paper as more metal is dissolved.

In this way, t h e ratio of t h e surface t o t h e weight of colored product is held more constant a n d t h e paper interstices remain open. Under these conditions, t h e print density is most nearly proportional t o t h e q u a n t i t y

222 Η. W. HERMANCE AND Η. V. WADLOW

of metal dissolved. For each reaction, there will exist a favorable density-range, within which differentiation can be m a d e most accurately. For example, copper on cadmium or zinc sulfide-impregnated #598 paper is best estimated in t h e range from 1 to 50 jug./sq. cm. W h e n t h e print

c/) Id

rr u 4 0 Q_

<

- I

I-z UJ

2 0

3 . 0 V O L T S

ELECTROLYTE - 1 0 % HAc PAPER - S b²S³ + B A C K I N G S P E C I M E N - P U R E C u , 6 . 3 C M² CATHODE - C A R B O N

10 15 6 0

FIG. 17.

3 0

T I M E IN S E C O N D S

Current-time curve for typical quantitative print.

Electrolyte: 1 0 % acetic acid Paper: S b2S3 + backing Specimen: pure copper, 6.3 cm.2

Cathode: carbon

density is light and t h e precipitation of colored product is largely con­

fined t o t h e upper layers of t h e paper, comparison is m a d e best in incident light, with t h e print illuminated from above a n d viewed at an angle of 30-45 degrees. Dense prints, in which t h e colored product has pene­

t r a t e d deeply, are best compared in t r a n s m i t t e d light. Immersion in

ELECTROGRAPHY AND ELECTRO-SPOT TESTING 223 clarifying agents such as cyclohexanol improves t h e light transmission a n d facilitates gomparison. A special illuminator for viewing spots in b o t h incident a n d t r a n s m i t t e d light h a s been described b y Clarke a n d Hermance (7).

Area of print: For comparison purposes, color a n d current intensities must be expressed in t e r m s of print area. W h e n this is of irregular outline, its measurement m a y become troublesome. If, regardless of the size of t h e specimen, t h e printing is confined t o a definite, reproducible area, one variable is eliminated a n d t h e operations, in general, are m a d e more convenient. T h e writers h a v e used circular cathodes, 2 - 3 cm. in diameter, with t h e printing a n d backing papers cut t o discs of exactly the same size. I n this w a y t h e area is precisely defined, b o t h for printing and for t h e current flow. T h e discs are laid on t h e specimen without blotting. T h e cathode is t h e n carefully aligned over t h e m , pressure applied, a n d t h e excess electrolyte is absorbed from t h e edges of t h e sand-wich with blotting paper. T h e q u a n t i t y of electrolyte remaining in t h e paper, a n d hence t h e conductivity, will depend on t h e pressure applied.

Reproducibility of printing conditions therefore requires control of t h e pressure. This can be accomplished with t h e calibrated spring press described earlier in t h e chapter.

Comparison techniques: Experimentally, t h e relation between t h e q u a n t i t y of t h e dissolved alloy a n d t h e print density m a y be established in various ways. I n their analysis of nickel-iron alloys,* Glazunov a n d Krivohlavy (25) prepare dimethylglyoxime prints on "filter p a p e r "

with an acetic acid electrolyte, t T h e iron interference is eliminated b y washing t h e print with dilute acetic acid, leaving t h e pure nickel color.

Their procedure involves essentially t h e following steps:

1. A series of nickel prints are m a d e from a known alloy, with condi-tions held identical except for t h e time, which is varied in regular steps.

2. P r i n t s of t h e u n k n o w n alloy are likewise prepared with varying time intervals b u t other conditions t h e same as for t h e k n o w n alloy.

3. T h e prints of t h e k n o w n a n d t h e u n k n o w n alloys are compared to obtain t h e times for which identical color intensity is obtained.

4. New prints are now m a d e for b o t h alloys, using t h e times which give equal color intensity. Now, however, t h e current is read at regular intervals t o obtain t h e product I · t for each alloy.

* Alloys having nickel contents from 0 to 6.5%, and from 51 to 100% nickel are solid solutions, to which eq. (3) may be applied accurately,

t The electrolyte used was:

Acetic acid, 5%, 2 parts Dimethylglyoxime, 1% alcoholic, 1 part.

224 Η. W. HERMANCE AND Η. V. WADLOW

5. Steps 1 to 4 give the time and current conditions necessary to produce prints of equal density from t h e two alloys. Since D = D ' , applying eq. (3),

^ (x2Ey - x*Ez + xEz) = ^ (a*Ey - a*Ez + aEz) or x*Ey - x*Ez + xEz = (a*Ey - a}Et + aEz)

where t', and A' are current, time, and area respectively for t h e known alloy and a is t h e fractional content of component y in it.

T h e electrochemical equivalents for nickel a n d iron are so nearly alike t h a t for approximate analyses, t h e y m a y be con­

sidered equal. If E^y = E^z, t h e above equation reduces to

Ft'A

χ = 7 ΰ Γ ^α

I n a somewhat different technique employed by t h e writers, t h e values for I, t, a n d A are held constant for t h e two alloys t o be compared, and t h e common component, y, is estimated from t h e relative densities of t h e prints obtained. T h e density ratio is arrived at b y comparing the prints with a series of permanent s t a n d a r d prints, suitably m o u n t e d and protected.

The method involves placing the printing cells in series in t h e same press and printing t h e two specimens simultaneously. Figure 18 shows the arrangement. I n this way, the same current flows t h r o u g h each cell for the same period and the pressure m u s t be t h e same on both. If cathodes and printing papers of uniform area are used, A, 7, and t will be identical for the two prints. F r o m eq. (3),

A = x2Ev - x*Ez + xEi I t D

Since t h e left h a n d member of the above equation is identical for b o t h alloys, the relation between t h e print density and the composition in binary alloys m a y be expressed by the following equation:

x*Ey - x*Ez + xEz = ^ (a*Ey - a*E2 + aEk)

where χ and a are the fraction contents of y in t h e two alloys a n d D a n d D' are the print densities, respectively. As pointed out in the dis­

cussion of t h e m e t h o d of Glazunov a n d Krivohlavy, when E^z and E^y are nearly the same, t h e equation m a y be simplified for approximate work T h e above equation then becomes:

D

x = D'^a

ELECTROGRAPHY AND ELECTRO-SPOT TESTING 225 W h e n there is available a series of known alloy specimens in which t h e y content increases regularly, t h e simultaneous printing m e t h o d is particularly helpful, for t h e n t h e u n k n o w n m a y be compared directly until a color m a t c h is obtained in t h e prints. I t s composition will then approximate t h a t of t h e matching known, without further calculation.

T h e estimation of m i n u t e quantities · of bisrjiuth in copper b y t h e electro-transfer m e t h o d was reported b y Garino and C a t t o (18). T h e electrolyte reagent was cinchonine iodide, and t h e print was m a d e on

FIG. 1 8 . Series electro-transfer.

cotton fabric, with a veiling sheet containing 1 0 % HNO3 between it and t h e specimen. For quantities of b i s m u t h of t h e order of . 0 1 % , differences of . 0 0 2 % could be distinguished.

An advantageous application of t h e q u a n t i t a t i v e electro-transfer might be found in t h e controlled stripping and analysis of surface layers to determine t h e concentration gradient when one metal penetrates another b y diffusion. For example, protective metal coatings m a y alloy with t h e base metal t h r o u g h m u t u a l solution. This process is accelerated when h e a t is involved, as in hot-dip applications. Some-times heat t r e a t m e n t s are purposely applied t o bring about such alloy

226 Η. W. HERMANCE A N D Η. V. WADLOW

formation. Successive prints m a y be made, t h e q u a n t i t y of surface dissolved in each case being determined b y application of F a r a d a y ' s law. F r o m this a n d t h e print density, t h e concentration of t h e metal in question m a y be estimated.

Related t o this application is t h e estimation of t h e weight of thin metal coatings. Successive prints for t h e base metal are m a d e with measurement of t h e q u a n t i t y of coating stripped each time. W h e n t h e print indicates exposure of t h e base metal, t h e individual strippings are totaled, giving t h e weight of coating over t h e print area. I n addition to the q u a n t i t a t i v e information, this method would also reveal any varia­

tions in t h e coating thickness in the area printed.

4.8.6. Electrography of Salt and Structural Patterns. There is one special application of electrography which, while not directly related to metallurgical problems, deserves brief mention. This is t h e reproduc­

tion of p a t t e r n s derived from conductivity differences in a material.

Such conductivity p a t t e r n s m a y result from localized salt concentrations or from specific electrolytic p a t h s afforded b y t h e structure of t h e material.

Yagoda (46) has obtained remarkably detailed electrographs of fresh sections of plant and animal tissues, using t h e potential gradient between anode and cathode platens to drive t h e naturally occurring chloride ions into a silver chromate-impregnated gelatin paper. T h e printing cell is built u p as follows: On t h e anode plate is placed a thick absorbent pad, t h e n the printing paper, b o t h soaked in s a t u r a t e d cal­

cium sulfate. On the printing paper is laid t h e specimen section, 1-3 m m . in thickness, then t h e cathode plate is lowered t o form t h e sandwich.

45 Volts are applied through a 500 ohm variable resistance until 200 millicoulombs/sq. cm. have passed. T h e yield of t h e chloride ion is controlled both b y t h e local salt concentration and b y t h e electrolytic p a t h s in t h e specimen. Since t h e speed of ionic migration differs in the various tissue elements, prints are often obtained which show con­

siderable a m o u n t of cellular structure.

Electrolytic p a t h s through wax or resin impregnated fibrous products m a y reduce their value as electrical insulators. Yagoda's work suggests a method of testing for such p a t h s electrographically. A material such as phenol fiber would be dried first, then soaked in a suitable electrolyte for several hours. At the end of this time, it would be placed on a gelatin printing paper containing the same electrolyte and a copper sen­

sitive reagent. T h e printing paper, with t h e usual backing sheet, rests on an aluminum cathode platen. A copper anode platen is placed over the specimen, and t h e sandwich is placed under considerable pressure.

A fairly high voltage is applied for sufficient time t o drive t h e dissolved copper ions through the p a t h s in the specimen to t h e printing paper,

ELECTROGRAPHY A N D ELECTRO-SPOT TESTING 227 where it is registered as the colored product, revealing t h e n u m b e r and location of t h e p a t h s .

REFERENCES 1. Ammerman, E., Stahl u. Eisen 51, 207 (1931).

2. Arnold, E., Chem. Listy 27, 73 (1933) (Brit. Chem. Abstracts B393, 1933).

In document Electrography and Electro-Spot Testing BY (Pldal 61-73)