Citation
The dezincification of alpha and beta brasses

Material Information

Title:
The dezincification of alpha and beta brasses
Creator:
Heidersbach, Robert Henry, 1940-
Publication Date:
Language:
English
Physical Description:
xi, 145 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Alloys ( jstor )
Brasses ( jstor )
Chlorides ( jstor )
Corrosion ( jstor )
Electrolytes ( jstor )
Electrons ( jstor )
Kinetics ( jstor )
Photomicrographs ( jstor )
Teeth ( jstor )
Zinc ( jstor )
Brass -- Corrosion ( lcsh )
Dissertations, Academic -- Metallurgical and Materials Engineering -- UF
Metallurgical and Materials Engineering thesis Ph. D
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis - University of Florida.
Bibliography:
Bibliography: leaves 137-144.
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025062112 ( ALEPH )
20143492 ( OCLC )

Downloads

This item has the following downloads:


Full Text















T HE 1) E ZIN" .' I.C"T T ON 0F ALPHIA AIND ET1A B RAS SEF





By



r Henry He idersb ach, J-fr.













A Dissvao Presernted to the Gradutate Courcil
of the University of Flcridla
i-n Partial FulfilJ mont of the Requirei~ents for the
Degree of Doctor of Philosophy



UNIVEPRSITY 0,11 FLORIDA\ ELC E'"BER, 1971.
































Dedicated to my wife, Dianne Katherine.














ACKNOWLEDGEMENTS



I would like to express my thanks to Dr. Ellis D. Verink, Jr., for his guLdance and inspiration. Thanks are also extended to the members of my committee, Dr. A. D. Wrallace, Dr. R. T. DeHoff, Dr. R. W. Gould and Dr. J. J. Hren.

Dr. M. Pourbaix provided encouragement during the early stages of this investigation. Dr. S. R. Bates provided

guidance on the use of and interpretation of results from the scanning electron microscope and the electron microprobe.

Mr. W. C. Fort III, provided invaluable assistance in obtaining the experimental results. Technical assistance was provided by Mr. W. A. Acree, Mr. E. J. Jenkins, Mr. E. C. Logsdon, Mr. P. D. Kalb, Mr. C. J. Minier and Mr. C. Simmons.

I particularly wish to acknowledge the financial support supplied by the National Association of Corrosion Engineers and by the International Nickel Company. In addition, certain of the equipment used in the electrochemical studies was purchased with funds from the Office of Saline Water.

The Chase rass and Copper Company donated the alpha brass used in this investigation, and this contribution is grateful ly acknowlcded.




iii









The author is grateful to his parents, who taught him the dignity of hard work, and to his wife, who made life bearable when things were not going according to plans.














































iv















TABLE OF CONTENTS


Page

ACKNOWLEDGEMENTS . . . . . . . . iii

LIST OF TABLES . . . . . . . . vii

LIST OF FIGURES . . . . . . . . viii

ABSTRACT . . . . . . . . . . xi

INTRODUCTION . . . . . . . . . 1

EXPERIMENTAL PROCEDURE . . . . . . 5
Immersion Test Apparatus . . . . . 6
Electrochemical Test Apparatus . . . 9
Sample Preparation . . . . . . is

X-RAY DIFFRACTION . . . . . . . . 18
Present Work . . . . . . . . 24

ELECTRON MICROPROBE . . . . . . . 30
Present Work . . . . . . . . 32

OPTICAL METHODS . . . . . . . . 41
Present Work . . . . . . . . 45

SOLUTION ANALYSIS . . . . . . . ... 54
Present Work . . . . . . . . 58

ELECTROCHEMICAL INVESTIGATIONS . . . . 83
Present Work . . . . . . . . 88

DISCUSSION . . . . . . . . . 105

CONCLUSIONS . . . . . . . . . 116

RECOMMENDATIONS FOR FURTHER RESEARCH . . . 119





v










TABLE OF CONTENTS (Continued)


Page

APPENDIX

1 EQUIPMENT USED IN ELECTROCHEMICAL
TESTS...................122

2 ELECTROLYTES ...............123

3 CHEMICAL ANALYSES OF ALPHA BRASS ....125 4 BETA BRASS INGOT PREPARATION. .......127

5 PROCEDURE FOR THE ELECTROGRAVIMETRIC
CHEMICAL ANALYSIS OF BINARY COPPERZINC ALLOYS...............129
6 ANALYSIS OF BETA BRASS INGOTS
USED IN THIS INVESTIGATION. ........131

7 ELECTROCHEMICAL SAMPLE PREPARATION .. 132

8 EQUATIONS USED IN CONSTRUCTION OF
POTENTIAL VERSUS pH DIAGRAMS FOR
THE Cu-C1-H20 SYSTEM AND THE
ZnO-H20 SYSTEM ...............133

BIBLIOGRAPHY....................137

BIOGRAPHICAL SKETCH ................145




















vi















LIST OF TABLES


Table Page

1 Sample 6-19, Weight Information. ......53

2 Atomic Absorption Data for Alpha
Brass in SN HC1 at 90.350C .. ........60

3 Atomic Absorption Data for Alpha
Brass in SN HC1 at 98.500C .. ........61

4 Atomic Absorption Data for Beta
Brass in SN HCl at 59.60'C .. ........62

5 Atomic Absorption Data for Beta
Brass in SN HCl at 67.500C .. ........63

6 Atomic Absorption Data for Beta
Brass in SN HCl at 89.350C .. ........64

7 Atomic Absorption Data for Beta
Brass in SN HCl at 99.850C .. ........66

8 Alpha Brass Potentiostatic
Test Data ...................96

9 Beta Brass Potentiostatic
Test Data..................98

10 Alpha Brass Free Corrosion Potential Tests...............102

11 Beta Brass Free Corrosion Potential Tests...............103

12 Approximate Copper and Zinc Concentrations in Solutions
Where Copper Deposits Were
Observed on Alpha Brass...........109







vii














LIST OF FIGURES


Figure Page

1 Immersion test cell ..... ........... 7

2 Circuit diagram of equipment used
to polarize specimen and automatically record corrosion current density as a
function of potential ..... .......... 10

3 Corrosion cell used in electrochemical
investigations .... ............. 11
4 Assembled and exploded views of
sample holder .... .............. .. 13

5 Copper-zinc phase diagram .. ........ .. 19

6 Copper-gold phase diagram .. ........ .. 20

7 The lattice parameter of alpha brass
as a function of the atomic percent
zinc ...... .................. 22

8 (111) peaks of a mixture of 70-30
brass filings and copper filings . 25

9 (111) peaks from sample of 70-30
brass dezincified for 20 days in
SN HCI at 750C ...... ............. 26

10 (111) peaks from sample of 70-30 brass
dezincified for 30 days in SN HCI
at 750C ........ ................. 27

11 Diffraction pattern from sample of
beta brass dezincified for two days
in SN HCI at 75C ... ............ 29

12 Zinc intensity profile from a
sample of alpha brass dezincified
for 79 days in 1N NaCl .. ......... 34

13 Zinc intensity profile from a
sample of alpha brass dezincified
for 10 days in SN HCl at 75C ....... .. 35

viii









LIST OF FIGURES (Continued)


Figure Page

14 Zinc intensity profile from a
sample of beta brass dezincified
for two days in SN HCI at 75C ..... 36

15 Photomicrograph of sample shown
in Figure 12 .... .............. 38

16 Photomicrograph of sample shown
in Figure 13 .... .............. 39

17 Photomicrograph of sample shown
in Figure 14 .... .............. 40

18 Dezincification plug in 70-30
alpha brass exposed for 79 days in
1N NaCl at room temperature ....... ... 42

19 Copper deposits on the surface of
dezincified alpha brass sample .... .. 47

20 Deposit on surface of dezincified
alpha brass sample .. ........... 48

21 Scanning electron micrograph of
copper slab protruding from the surface of a dezincified alpha
brass sample .... .............. 49

22 Nondispersive x-ray analyzer pattern
of deposit shown in Figure 21 ....... .. 51

23 Dezincified cross-section of sample
shown in Figure 21 .. ........... 52

24 Atomic-absorption calibration curve
for copper .... ............... 68

25 Copper and zinc dissolution from
alpha brass in SN HCl at 90.35'C . . 71

26 Copper and zinc dissolution from
alpha brass in SN HCI at 98.50C . . 72

27 Dezincification factors, Z, for
alpha brass in SN HCl at 98.50'C 73

28 Copper dissolution from beta brass
in SN HCI at 67.50.C ............ 75

ix










LIST OF FIGURES (Continued)


Figure Page

29 Zinc dissolution from beta brass
in 5N HCI at 67.50'C .. .......... 76

30 Dezincification factors, Z, for
beta brass in SN HCI at 67.500C .. ..... 77

31 Zinc dissolution from beta brass
in SN HCl at various temperatures . . 78

32 Arrhenius plot of zinc dissolution
rate versus 1000/T ... .......... 80

33 Experimental potential versus pH
diagram for 70 Cu 30 Zn in
0.1M C1- at 250C ...... ............ 89

34 Simplified Cu-Cl-H20 diagram at
25C for solution containing 0.1M
chloride ions .... .............. .. 90

35 Simplified Zn-H20 diagram for
concentrations of ionic
species = 10-6M .... ............ 91

36 70 Cu 30 Zn alloy in 0.1M
chloride solution ... .......... 92

37 Beta brass held at +0.050VSHE for
2-3/4 hours ..... ............... 94

38 Typical potentiokinetic scan in
acid solutions .. ....... ..... 100

39 Cross-section of alpha brass
showing where holidays in the
stop-off lacquer caused dezincification in the stagnant
regions beneath the holidays .I..... 110

40 Theoretical domains for dealloying
in a given solution based upon the
Nernst equation .... ............. ..114






x









Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



THE DEZINCIFICATION OF ALPHA AND BETA BRASSES



By


Robert Henry Heidersbach, Jr.

December, 1971



Chairman: Dr. Ellis D. Verink, Jr. Major Department: Metallurgical and Materials Engineering



The mechanisms of dezincification of single-phase

alpha and beta brasses were studied using x-ray diffraction, electron microprobe, metallographic, atomic absorption, and electrochemical techniques.

The two mechanisms of dezincification which had been

previously reported, (1) dissolution of both alloy constituents followed by redeposition of the more noble species, and (2) the selective removal of the less noble constituent, were found to be operative under certain conditions of potential and pH for both alpha and beta brasses.

An electrochemical explanation of the circumstances

under which dealloying can be expected to occur was developed based on the use of Pourbaix diagrams.




xi















INTRODUCTION



Dealloying is a corrosion process whereby one constituent of an alloy is preferentially removed from the alloy, leaving an altered residual structure.(1)

While dezincification, the loss of zinc from brasses, is the most commonly experienced form of dealloying,(2-12) other examples have been reported in practice. These include loss of nickel, (2,13-15) aluminum (9,16-22) and tin (23,24) from copper alloys; iron from cast iron; 25) nickel from alloy steels (2)and cobalt from Stellite. (7

Since the phenomenon was first reported in 1866, the

literature has been filled with reports of research efforts aimed at clarifying the mechanisms of dealloying. Nonetheless, there still is no general agreement as to the detailed mechanism involved. One group contends that the entire alloy is dissolved and that one of its constituents then is replated from slto.(23,27-40) Another contends that

one species is selectively dissolved from the alloy, leaving a porous residue of the more noble spce.(15,41-47) Still others believe that both mechanisms take place. (16,4854)

A number of literature surveys have appeared,and these summarize the situation up to the time they were written. (1,49,55-58)





2



The study of dealloying phenomena is fraught with a number of complications. Generally, such reactions are relatively slow and a lengthy exposure period is required to cause a sufficiently great amount of dealloying to facilitate evaluation. Consequently, there is considerable interest in accelerated tests for evaluation of tendencies of alloys to dealloy. Many techniques have been employed. For example, electrolyte compositions have been adjusted by using more concentrated solutions or solutions having variations in oxidizing power. (52) Specific ions have been added to stimulate dealloying, e.g., saturated cuprous chloride

solutions have been used to accelerate the dezincification of copper-base alloys. (40) Electrochemical stimulation also has been used. Unfortunately, all too often the test methods emploved have been vulnerable to criticism as having biased the experimental result and, although specific techniques are now available which can cause dealloying to occur in the laboratory, nonetheless, there still is no firm basis

for predicting the likelihood of dealloying in service based on these tests.

Single-phase alpha brasses and duplex alpha-plus-beta brasses are the only forms of copper-zinc alloys having enineer i, g significance at present. Bengough and May reported, in 1922, that small additions of arsenic would prevent dezincification of alpha brasses but would not protect the beta phases cf duplex alloy-,. (59) The reasons for





3



arsenic protecting alpha brass but not the beta phase of duplex alloys have remained controversial.(1,55,56) However, the addition of small amounts of arsenic, or of antimony or phosphorus, which have similar effects, has become a standard means of inhibiting dezincification in alpha brasses.(1) No inhibitor is presently available for duplex alloys, although thie addition of tin retards most forms of brass corrosion to include dezincification.(1)

The purpose of this investigation has been to elucidate the mechanism of dezincification of alpha and of beta brasses and to develop a basis for predicting the conditions under which dezincification of these alloys might be expected to occur .

Particular emphasis was placed on selecting exposure

conditions and test methods which would not bias the experimental results. For example, it was felt that accelerated

tests using copper-chloride solutions could not give unbiased evidence for a dissolution-redeposition mechanism, although this method of accelerated testing has been reported frequently in the literature. (29,40,60-62) Electrochemi.cal stimulation (in which the specimen was a driven anode) has also been used as a means of producing dezincification.(62-65) This method of producing accelerated attack also can bias the experimental results by masking the presence of diffusioin-related selective-removal processes, becaMuse diffusion, a: it is commonly understood, is a quite






4


slow process at the low temperatures encountered in aqueous environments.

Mechanism studies were performed on brass immersion

samples placed in NaCl and HCl solutions. No copper salts were added to the test electrolyte except as the result of corrosion of the test specimens. Test durations of up to 90 days were employed at temperatures ranging from room temperature (approximately 25'C) to 100'C.

Potentiostatic and potentiokinetic methods were used to define conditions under which dealloying might be expected to occur.














EXPERIMENTAL PROCEDURE



The possibility of bias of results due to the testing method used was discussed in the introduction to this dissertation. Immersion testing of brass samples in environments known to produce dezincification was chosen as the laboratory test method most suited for a study of the mechanism of dealloying.

Electrochemical tests were employed in later stages of this investigation, after the mechanism studies were completed. These tests were intended to define the conditions of potential and pH under which particular modes of dealloying of copper-zinc alloys could be expected to occur.























5





6



Iunersion Test Apparatus


Figure 1 is a diagram of the immersion test cell used in this investigation. It is similar in many respects to that recommended by National Association of Corrosion Engineers Standard TM-01-69, "Laboratory Corrosion Testing of Metals for the Process Industries."(65)

The cell is constructed of a Pyrex* glass resin reaction kettle held together by an external metal clamp. All fixtures inserted into the cell also are made of Pyrex, with the exception of the Teflon**- coated stirring bar.

Samples are suspended from a sample holder as shown in the diagram. In tests requiring an oxygen-free environment, liquid in the cell is sparged with dried and purified argon which enters the cell through a fritted-glass diffuser immersed in the liquid. Gas is passed out of the system by wav of a Liebig-type reflux condenser which is backed up by a liquid trap to prevent solution loss or contamination. The volume of solution within the cell was measured at the start and finish of each test. Losses did not exceed 10 ml from an initial volume of 750 ml in tests of up to 30 days. This corresponds to a maximum change in solution volume of 1.3 percent. The normality of acids being used was also checked


*Registered trademark, Corning Glass Company.

**Registerd trademark, E. I. duPont de Nemours and Company, Inc.






7







G


-F


































Figure 1. Immersion test cell. A -sample holder,
B sample, C -immersion heater, D -stirring bar, E J-tube thermostat, F -gas
inlet diffuser, and G gas outlet through
reflux condenser.






8



before and after each exposure by titration with a 1N NaOH standard. It was found to vary by no more than 0.05 normality units from start to finish of any test.

The liquid temperature was controlled to within O.3S0C by means of an electric thermoregulator coupled to a 100 watt immersion heater.

Immersion tests were conducted in 1N NaCi and SN HCl. These solutions were chosen because they are nonoxidizing and do not form soluble metal complexes with the exception of CuCl 2 Because of the historic significance of dezincification failures in salt water and other chloride environments, it was felt that chlorides should be used despite the existence of this one complex metal ion. The high acid concentration was chosen to minimize acidity changes within the solution. This high acid concentration was also very close to the maximum concentration that can be maintained at 100*C without boiling. (67)

Previous immersion tests in HC1 and NaCi have appeared in the lieaue,(94,4,26)2%8%9 and comparisons

with previous results are contained in the discussion section of this work.





9



Electrochemical Test Apparatus


The electrochemical hysteresis technique was used for experimental determination of the corrosion behavior of alloys as a function of potential and pH. The equipment, techniques and theories upon which investigations of this type are based have been discussed at some length in other reports. (70- 73)

The equipment is described schematically in Figure 2

and listed in Appendix 1. The scanning potentiostat applies a continuously varying potential to the sample over a certain range. The differential amplifier isolates the corrosion cell from the recording equipment and eliminates ground loops. The log converter allows the corrosion current to be plotted as a logarithm on the x axis of the x-y recorder, while the sample potential is plotted linearly on the y axis. An automatic step switching apparatus extendss the range of the log converter and allows the continuous recording of corrosion current densities from 5 x 108 to 10 amps/cm 2 The low-pass RC filter, consisting of a 100 microfarad capacitor in parallel with a 25 K-ohm potentiometer adjusted to 10 K-ohms, reduces electrical noise in the system to negligible values.

A schematic diagram of the corrosion cell is seen in Figure 3. The buffered electrolyte is vacuum deaerated prior to transference to the corrosion cell. The solution is continuously purged with hydrogen during the run via the






10
















L SPECIMEN
ASZIL IA T
ELECTAODE

POTENTIOSTAT CELL
-IEftE I N CE 0ELECTIRODE
-o0 )




RESISTANCE 0_ DIFFERENTIAL
SELECTOR c AMPLIFIER
SWITCH BOX



LOG CONVERTER





R. C.
-zX-Y RECORDER FILTER

-0
I-AIIS I NP T CURRNT DEMSITT

Figure 2. Circuit diagram of equipment used to polarize
specimen and automatically record corrosion current density as a function of potential.





11







CORROSION CELL SCHEMATIC



SAMPLE PLATINUM HYDROGEN
HOLDER PROBE ELECTRODE DIFFUSER
THERMOMETER

























MAGNETIC
STIRRER

Figure 3. Corrosion cell used in electrochemical
investigations.






12



gas diffuser as suggested by ASTM committee G.( )A bright platinum screen serves as the auxiliary electrode. The current from the potentiostat to the auxiliary electrode is measured as a potential across a precision resistor selected to provide the required logarithmic converter input voltage. A Luggin-Haber probe connected to a standard calomel electrode is used to measure the sample potential. The thermometer is used to indicate the temperature of the electrolyte. The solution is stirred using a magnetic, waterpowered stirrer. The cell itself is made of Pyrex glass with a Teflon lid bolted to a polycarbonate Van-Stone backing ring which makes the system airtight.

The sampleholder is shown in Figure 4. The main body of the holder is constructed of Teflon so that it will not react with the test solution. Copper parts encased in the Teflon allow electrical contact with the sample. A polycarbonate nut fas tens the sample in place and allows 1 cm2 to be exposed to the electrolyte. The Teflon gasket avoids leakage and minimizes crevice effects.

Care must be exercised when choosing buffered electrolytes to insure that solution ions will not form complexes with the metal ions from sample dissolution. The electrolytes used in these studies are listed in Appendix 2. All solutions used in this portion of the investigation had a

0.1 M chloride content. The effects of copper-chloride complexes were discussed in the section on immersion testing.





13







BUS BAR
















TEFLON HOLDER






SPACER SAMPLE TEFLON GASKET


POLYCARBONATE NUT DISK SPECIMEN HOLDER Pc



Figure 4. Assembled and exploded views of sample holder.





14



The effects of scan rate were investigated. Reinoehl and coworkers(76) show that passivation current density is a function of scan rate, but that critical passivation potentials and rupture potentials are independent of scan rate, at least for iron in lN H2SO4. A more detailed discussion of the importance of scan rates is contained in the discussion section. A scan rate of 33 mv/min was chosen for this investigation after examining scan rates from 5.5 mv/min to 667 mv/min.(77) The selected rate was slow enough to avoid losing any configuration of the polarization curve, yet fast enough to minimize masking effects due to dezincification of the sample.

After experimental Pourbaix diagrams were established by potentiokinetic polarization techniques, potentiostatic tests were run to verify the potential regions where dezincification occurs. The six-cell potentiostat described by Cusamano(72) was used for these potentiostatic tests.





15



Sample Preparation


The alpha brass samples used in this investigation were donated by the Chase Brass and Copper Company. The brass was obtained in two 25-pound lots of as-extruded cartridge brass wire having diameters of 0.850 in and 0.687 in, respectively. The chemical analyses of these alpha brass samples are shown in Appendix 3.

Beta brass ingots having a nominal composition of 52 w/o Cu and 48 w/o Zn were prepared from 99.99) pure copper and zinc stock purchased from the 'Americn" Smelting and Refining Company. The ingot melting procedures are de:cribed in detail in Appendix 4. TIlie tops and bottoms of each i;ngot were analyzed according to the procedu-re described in Appendix 5. Inrgots which varied by less than 1.0 w/o Cu from top to bottom: were then used in the as-cast condition for these experiments. Each ingot weighed approximately 150 grams and was 15 mr in diameter. The average comlpositions of the beta brass ingots used are listed in Appendix 6.

Disc-shaped cross-sections were cut from the alpha

brass wire and from the as-cast beta brass ingots. Sample surfaces we!-e prepared according to the technique described in Appendix 7.

Five types of samples were used in this study. Samples intended for x-ray analysis were simple discs of brass with a hole drilled through one side so they could be suspended from the sample holder as sho:n in Figure 1. These samples





16



were also used for the electron microprobe a:d optical investigations.

For the atomic-absorption experiment where metal-ion dissolution was to be monitored, cylindrical specimens were mounted in dental mount so as to leave exposed a circular cross-section of known geometric area. The specimens were then polished according to the procedure described in Appendix 7 and suspended in the reaction ketitles by means of a hole drilled in the denial mount. This procedure is similar to that described by Fisher and Halperin for their calorimetry studies on the dealloying of copper-7old alloys.(78)

Disc specimens prepared according to the procedure described in Appendix 7 were used for the potentiokinetic experiments using a sample holder shown in Figure 4. For testing details see the section on electrochemical test apparatus.

Early high-temperature tests in this investigation used samples in which all but a 1/2 cm x 1/2 cm square section of the sample was "stopped off" using Miccroshield* stop-off lacquer. Examination of the samples after exposure revealed that leaks through the lacquer were present and this type of sample was abandoned. The significance of these leaks is discussed in the discussion section of this report.

Potentiostatic tests at room temperature were conducted using discs with an insulated copper wire soldered to the


*Trade name, Michigan Chrome and Chemical Company.






17



back. Miccrostop stop-off lacquer was used to coat the back and sides of these samples leaving a circular exposure surface. Examination of the samples after exposure revealed no apparent leaks in the lacquer.















X-RAY DIFFRACTION



X-ray diffraction and electron diffraction have been used by several researchers in an attempt to ascertain whether dealloying is a process involving the selective removal of one constituent or a dissolution-redeposition process (3,5263,6,7983)If a redeposition process occurs, the diffraction pattern should show diffracted radiation due to remnants of the original alloy and also due to redeposited metal. Diffracted radiation between the two peaks would supposedly indicate an alloy having a composition between that of the original alloy and that of the final, dealloyed, relatively pure metal. This intermediate composition, if detected, would indicate that a selective-removal process is operative.

Most x-ray diffraction and electron diffraction studies of dealloying have involved single-phase, binary alloys of either copper-zinc or copper-gold. The phase diagrams for these two binary systems are shown in Figures 5 and 6.

One of the principal limitations of the copper-zinc system is that alpha brass is stable up to only about 37 w/o zinc at room temperature. This limits the possible changes in lattice parameter, and thus the changes in the resulting x-ray or electron diffraction angles, to a small


18





19
















00
1000
900 .8)'+


800
0
= 700 2 30
re a p
14 60030 +



200 L
10 20 30 40 50 60 70 80 90 WEIGHT PERCENTAGE ZINC



Figure 5. Copper-zinc phase diagram (Metals Handbook,
1948 edition, p. 1206).






20












1100

1000L




0
w700
cc
600
L
Ea

400

/ I
300 -I \
I a I a
200-II\
I I
100 1 f I I I I I I I
10 20 30 40 50 60 70 80 90 100
WEIGHT PERCENTAGE GOLD




Figure 6. Copper-gold phase diagram (Metals Handbook,
1948 edition, p. 1171).





21



range. (83 The lattice parameter of alpha brass as a function of zinc content is shown in Figure 7.

Pickering points out that by choosing an alloy system, such as copper-gold, with elements having substantial differences in atomic diameter, the lattice parameter, and thus the separation between peaks of the original alloy and those of the corroded residue, will be increased. (79)

An alternative to this would be to start with an alloy which, upon dealloying, would undergo a phase change as well as a change in lattice parameter.

Of the large number of reports concerning diffraction studies of dealloying, only two previous researchers have concluded that their data supported a selective-removal mechanism. (52,63,64,79)

Pickering subjected copper-gold alloys to electrolytic dissolution and showed the appearance of an intermediate peak which, as additional current was passed, increased in intensity and moved closer to the position to be expected for pure gold. The alloy peak decreased appropriately in intensity.

A later investigation by the same author reported the formation of new, intermediate, phases during the anodic dissolution of gamma brass and of epsilon brass.(13 ) This report substantiated the results of Stillwell and Turnipseed who subjected epsilon brass to various corrosive media, and whose x-ray diffraction results indicated the presence of intermediate phases in some of their experiments. (52)





22










3.68



0 cc
S3.66





3.64




3.62




3"600 10 20 30 4G

ATOMIC PERCENTAGE ZINC





Figure 7. The lattice parameter of alpha brass as a
function of the atomic percent zinc. (84)






23



The lack of similar results in other studies of dealloying may be due to the use of conventional film methods for recording diffracted intensity. (8s)





24



Present Work


All x-ray diffraction mechanism studies performed during this investigation were performed on powder samples obtained from discs of alpha brass and of beta brass exposed to hydrochloric acid in the immersion apparatus shown in Figure 1 and described in the section on experimental procedure.

There are several problems associated with the use of a diffractometer for investigating dezincification of alpha brasses. If the peaks due to the original alloy occur too close to the copper peaks, then "tails" of diffraction peaks may overlap and be misinterpreted as being due to an alloy of intermediate composition.

This is true of the diffraction patterns obtained using Cu Ka radiation. The separation is less than one degree 2e between the (111) peaks of pure copper and of 70-30 alpha brass. At higher angles the separation between peaks becomes greater, but intermediate intensity, if present, would be spread out also, and thus be harder to detect.

Figure 8 shows the (111) peaks from a sample obtained by mixing filings of alpha brass with annealed filings of copper. A slight increase of intensity due to the overlap of the "tails" of the peaks is apparent.

Figures 9 and 10 show the (111) peaks obtained from

samples of alpha brass dezincified for 20 and 30 days, respectively, in 5N HC1 at 75'C. The intensity between the








25














04 P4
0
u


Cc


bo





M
M Cd
$.4 .
Q 4J
4-1



CD 0
r- 4
4
4-4 00


09 C.0 41
to
-H CIS

P4

t4-4 M 0 M
4
tA
-X
C13










oj








26




































Cd










La 0 u
0
(D Ln La r= 4J Cd Cd
th


0

4-4 Ln


cd H
0
sa tn




r-4 C)






(D








27

















Lq
0










tq





Cd



C)


C:)
t-I C4
0 u 71- col) 0
LIJ (1) Ln
LAj r--q r4-) LAJ CIS CIS
i= V)








cd -H V) >1 CIS 10






C>
r--q

(D


W





28



two peaks on the corroded samples is higher than would be obtained due to overlap of the "tails" of the copper and alpha brass peaks. However, this intensity would probably be undetected on a standard powder pattern obtained by film methods.

If dezincification of beta brass occurs by a volumediffusion mechanism, the diffraction pattern obtained from a powder of dezincified beta brass could be expected to show scattered intensity characteristic of FCC alpha brass having a lattice parameter different from that of FCC copper. Figure 11 shows a diffraction pattern obtained from a sample of beta brass dezincified for two days in SN HCl at 75*C. The peak at 43.3 degrees is due to a superposition of beta brass (110) scattering and copper (111) scattering. The broad peak at 42.4 degrees is characteristic of alpha brass having approximately 36 w/o zinc.

The above-mentioned figures thus give x-ray diffraction evidence which supports a selective-removal mechanism for the loss of zinc from both alpha and beta brasses when exposed to SN HC1 at 75'C.

No unidentified peaks were detected in the powder patterns of any of the samples shown above; thus, the possibility of this scattering being due to an extraneous source is eliminated.





29






BETA BRASS 110) PLUS
COPPER (111)

















ALPHA BRASS (110)








41 42 43 44 45
DEGREES 2e



Figure 11. Diffraction pattern from sample of beta
brass dezincified for two days in SN HCl
at 750C.















ELECTRON MICROPROBE



The electron microprobe, with its unique capability

for chemical analysis of very small regions, has been used by other researchers in an attempt to show the "diffusion gradients" to be expected from a selective-removal mechanism for dealloying. (64,79,86)

Birks describes the electron microprobe as having the capability of giving an x-ray spectrochemical analysis of areas between 0.1 and 3p' in diameter. (87 ) The principal factors affecting the size of the region contributing x-rays to the spectrum are the electron beam size and the density of the sample under investigation.

Any quantitative method for x-ray spectrochemical analysis requires comparison with standards having, as nearly as possible, the same physical characteristics as the sample under investigation.i) If a selective removal of zinc atoms occurs during dezincification, the atomic sites formerly occupied by these atoms will become vacancies. (79) As there is no way of reproducing these vacancies with their subsequent effects on density and surface roughness, in a calibration standard, it is misleading to attempt to assign quantitative chemical analysis values to electron microprobe data from dealloyed specimens.


30






31



The use of chart recorders to show diffusion gradients can also be misleading. Both the response time of the chart recorder and the scan rate at which the sample is swept under the electron beam can affect the apparent fall-off distance that is recorded on the chart.

Sugawara and Ebk 8)and Pikrn(64'"7) have published microprobe data obtained with chart recorders to show relative zinc concentrations as a function of distance. Sugawara and Ebiko concluded that their data gave no indication of a concentration gradient along the interface between corroded and uncorroded brass. Pickering reports a region approximately 9p~ thick where the copper concentration fell off in a copper-gold alloy which had been subjected to electrochemical anodic dissolution.

Other researchers have used x-ray images to demonstrate the occurrence of dealloying. (88,89) None of these authors used microprobe data to support arguments regarding the mechanisms involved.





32



Present Work


As the above discussion shows, it would be extremely difficult) if not impossible, to obtain meaningful quantitative analyses through electron microprobe analysis of a region having a large, and changing, number of vacancies. This problem has been discussed by Anusavice (90) who, because of his interest in eliminating vacancy and pore formation in high-temperature diffusion couples, was able to apply pressure and significantly reduce porosity formation.

The technique that was decided upon in this study was to use the University of Florida's Acton Electron Microprobe, operated under conditions which would produce a minimum beam size, to obtain point counts across the region where zinc concentration was found to change. These regions were identified either by using a chart recorder or, as experience was obtained in identifying precise regions of interest, by using the light-optics microscope attachment on the microprobe.

The published data of Pickering (64,79) and of Forty and Humble(91) indicated that a diffusion zone, if it were to be found., would be rather narrow, of the order of 10p or even less. Results in these laboratories confirmed these observations and perhaps explain why Sugawara and Ebiko, (86) who used a chart recorder and a relatively fast scan rate, were unable to find a diffusion zone.





33



In the present study point counts were taken at intervals as small as one micron across the region of interest. It is recognized that making point counts at narrow intervals might involve some slight overlap of the irradiated volumes between two adjacent points. However, the results obtained by varying operating conditions sc that the beam

size would be reduced to a bare minimum did not give a change in the results obtained, and it is felt that the counts obtained in this manner are of phYsical significance.

Results were tabulated as intensity ratios by taking the average of three successive counts on the same point, subtracting the average background intensity on that sample, and dividing by the average point counts on an elemental standard minus the average background intensity on the standard. This can be expressed as: IA(Sample) IA(Sample BG)
(Star' -9-TG
A 00A Il 0a l 0 a r G


where A is the element of interest. (87)

Intensity ratios of this type are the raw data for

computer programs such as the MAGIC program written by J. W. Colby and in use at the University of Florida for quanti(93)
tative electron microprobe analysis.

Figures 12, 13 and 14 show plots of intensity ratios

versus distance for samples of alpha and beta brass exposed in 5N HCI or in lN NaCl. The accompanying photomicrographs,





34

















3=





15

10 "
250







0 [ I
0 4 8 12 16 20
MICRONS





Figure 12. Zinc intensity profile from a sample of
alpha brass dezincified for 79 days in
iN NaCI.








35















-wr
CN



tA
M




t CN cd



Cd
0
4-4 Lo 0 r

co 0 4J
C13


as u tn :=

Cd z
LO

Cd 0
1.4 ,


(D
I crj .,.1 10
4-4
0 C)
$-4
PL, cc > 0 4J 4-4 .rq tn


4J 4-4 r .,.q
u

U -r-4 0 N







&ZLI C= ull cm
Cd

oz 001 ",1








36






















CM tr)



as
41
Q)o 'm c)
Ln 4-4 rC=) 0
qw
Q (t r--l
04 r=

CO) tA

CD 4= C-) r 9
0
4
tA
>1



4-4 0 0 : 4J


0
4-) 44 .,.4
V) I:$


C:o

U
r
U H r t-4 .H (D





CD CD C2
Ln qw C-2 Cd r-A

UZ COL I/ UZI





37



Figures 15, 16 and 17, show'the carbon traces left on the sample by the interaction of small amounts of diffusion pump oil with the high-energy electron beam. The diffusion zone occurs in each case at the very edge of the obvious change in color as shown on the photomicrographs.

Figures 12, 13 and 14 are thus, to the author's knowledge, the first experimental plots of alloy composition versus distarce to be obtained from samples subjected to dealloying under free corrosion conditions and which indicate the existence of a diffusion zone.

The diffusion distances shown in Figures 12, 13 and 14 are typical of those found on the samples examiied. Traces were made cross the edges of polished pieces of brass and copper to determine the "distance" which would be recorded due to beam overlap between one-micron settings of the microprobe stage. Results indicated one intermediate point between the intensity ratio of the metal and the background reading on the nonconductiive Lmounting material. A beam at the intermediate-reading setting which was only partially on the sample could explain the intermediate point. The intermediate points of Figures 12, 13 and 14 cover an interval of at least six microns and are thus not due to overlap of the beam at adjacent position settings.





38










































Figure 15. Photomicrograph of sample shown in Figure 12.
Probe trace is indicated by the arrow. SOOX






39













































Figure 16. Photomicrograph of sample shown in Figure 13.
Probe trace is indicated by the arrow. Note
surface deposit of copper. 250X






40












































Figure 17. Photomicrograph of sample shown in Figure 14.
Probe trace is indicated by the arrow. 250X















OPTICAL METHODS



Early work on the dezincification mechanism relied

heavily on optical observations. Most of the research tools used in this study were not developed at that time, but the microscope and metallograph were available and were used. The results obtained from microscopic investigation were, however, subject to "opinion-type" interpretations, and researchers did not have the benefit of present-day knowledge of crystal structure, grain growth, epitaxial electrolytic deposition, the concept of an occluded cell, and other ideas which are part of the present-day researcher's background.

In 1922 Abrams (29) suggested that dezincification

occurred when a membrane of some type was available to hold dissolved copper in contact with the brass surface or when a large excess of copper was present in the solution. His experiments with copper chloride solutions led to further wo rk by Bengough and May, (69) and solutions of this type soon became an accepted method of accelerated testing for the susceptibility of alloys to dezincification. (60)

Early researchers noted that dezincification could be classified as occurring either in "layers" or in "plugs" such as the one shown in Figure 18. These plugs were



41






42












































Figure 18. Dezincification plug in 70-30 alpha brass
exposed for 79 days in IN NaCI at room
temperature. 200X





43



attributed to breaks in a protective scale on the metal surface which allowed localized attack in certain specific locations. It seemed obvious that breaks of this type would cause flow restrictions which would lead to a high concentration of copper in a region adjacent to the corroding surface. Bengough and May pointed out that precipitation of copper due to concentration effects seemed likely,(69) and for a long time this seemed to be the logical mechanism for dezincification. Simmons later pointed out that dezincification plugs occurred under circumstances which, if slightly altered, would result in pits. (8)

Most early investigators were concerned with dezincification of condenser tubes under conditions where scale buildup, and subsequent cracks in the scale,were likely to occur. The ideas put forth by Abrams and by Bengough and May gained wide acceptance. However, this explanation did not seem to cover dealloying in ship propellers and in other high-flowrate situations.

Interest in the field remained high, and reports of

metallographic investigations into the dezincification mechanism continued. Bassett (45) and Polushkin and Shuldener(7) decided that dezincification was a selective removal process after examining large numbers of samples which had failed in service. This contrasted with the opinion of Horton(39) and served to emphasize that conclusions based on metallographic observation are often dependent on the opinions and prior experience of the observer.





44



The difficulties of defining and comparing the service conditions reported by various authors are additional.drawbacks to arriving at a general mechanism based on optical observations of in-service failures.

In 1965, V. F. Lucey reported on laboratory experiments conducted in closed containers containing saturated copper chloride solutions and containing excess undissolved cuprous chloride. (40) It is hardly surprising that he observed copper deposition from such a solution.(92) However, Lucey neglected to point out that his experiments did not show that a selective removal process could not occur under other

circumstances or, indeed, under the conditions described in his papers.

The above discussion is not intended to indicate that optical observations have no value in the determination of a mechanism for dealloying but merely to point out that experim!,nts should be carefully controlled and that limitations of optical methods should be recognized. A concurrent observation is that information which lends support to one theory should not be misinterpreted as proof of the invalidity of a contrasting theory. Some, but by no means all, of the authors discussed above have recognized this,while others, unfortunately, have not.






45



Present Work


Exposure samples intended for x-ray diffraction analysis were prepared according to the procedures described in the section on sample preparation. NAGE Standard TM-Ol69 (66) suggests that closed exposure cells, such as were used in this investigation, should contain several samples to be pulled after certain periods of time. Each time a sample is removed from the cell a replacement sample is added. The reason which is given for this is to check for possible changes in the corrosivity of the environment.(6 6) As an example, in one series of tests run during this investigation five samples were prepared for each cell. Three. were immersed in the original solution. After ten days one sample was removed and replaced with a fresh sample. This

was repeated at the end of twenty days, and the test was terminated at the end of thirty days. Thus five samples were obtained from each test and, if no changes in the corrosivity of the environment occurred, the amount of dezincification experienced by the sample exposed for the first ten days of the test would correspond to that of the sample exposed for the last ten. The same should hold for samples exposed for the first twenty days and the last twenty days.

As was mentioned above, these tests were intended primarily as a source of dezincified brass to be used for x-ray powder samples. Because of this each sample was a simple disc with a hole drilled near one edge so it could be





46



supported in the exposure cell. No effort was made to mask off all but a certain area for exposure and the exposed area was about 19-1/2 cm2 in 750 ml of solution (~2.6 x 10-2 cm2/ml) which is quite high. The tests were conducted in 5N HC1 at varying temperatures. During these tests a number of the samples appeared to have surface deposits which had a metallic luster and gre, la-rgeor with the passage of time. These "deposits" were apparent on samples which had beer, exposed from the beginning of the test as well as cn samples which were added when other samples were removed.

Figure 19 shows a photomicrograph of a cross-section

of one of these samples. The normal deziracified texture at the bottom of the sample contrasts sharply with the appearance of the shiny surface deposits.

That these formations are, in fact, deposits is shown in Figure 20 where a portion of the original brass surface

is visible in the photomicrograph.

Each sample in these tests was weighed before and after the test. The weight-loss data did not follow any recognizable pattern. One sample in these tests actually gained weight.

This particular sample, which was exposed for the last

ten days of a twenty-day test, had numerous surface deposits which were copper colored and had a metallic luster.

Figure 21 is a scanning electron micrograph of one of

these deposits which fits the "ridge depoAs tioni" description





47










































Figure 19. Copper deposits on the surface of dezincified
alpha brass sample. Sample was exposed to
5N HC1 for 20 days at 1000C. 50X





48











































Figure 20. Deposit on surface of dezincified alpha brass
sample. Sample was exposed in 5N HCl at
1000C for 10 days. 10OX





49












































Figure 21. Scanning electron micrograph of copper slab
protruding from the surface of a dezincified
alpha brass sample. SOOX





so



of copper deposits provided by Bockris and Damjamovic. (3 other deposits had the appearance of "long, curled-up wood shavings." Figure 22 is a nondispersive x-ray pattern of the formation shown in Figure 21, and this pattern clearly shows the deposit to be metallic copper. The sample was exposed to hydrochloric acid in a Pyrex reaction kettle, and the silicon and chlorine peaks-'are probably due to contamination on the surface of the deposit.

One of the deposits was pulled from the sample surface

with a pair of tweezers and made into an x-ray powder sample. The diffraction pattern obtained from this deposit confirmed that the deposits were metallic copper.

Table 1 summarizes the weight change data from the

sample discussed above. Figure 23 is a photomicrograph of this sample and shows that the sample did dezincify as well as provide sites for the deposition of copper. No deposits appear in photomicrographs of this sample because they were removed for the x-ray diffraction and weight-change analyses

described above.

The above results clearly show that large surface-areato-cell-volume ratios and the addition of samples to solutions already containing dissolved copper are situations to be avoided. All further tests were conducted on a one-sampleper-cell basis to avoid the effects described above.





















16 CuKa

14

S12




CC
L.


=s4 S C
= 1 CuKp
2

2 3 4 5 6 7 8 9 10
EMISSION ENERGY- KeV





Figure 22. Nondispersive x-ray analyzer pattern of
deposit shown in Figure 21.





52










































Figure 23. Dezincified cross-section of sample shown
in Figure 21. 500X





53



Table 1

Sample 6-19, Weight Information



Original weight 5.2121 gm


Final weight 5.3901 gm.


Weight gained 0.1780 gm


Weight of sample with deposits removed by tweezers and ultrasonic cleaning 4.505S gm


Weight of deposits (by difference) 0.8846 gm














SOLUTION ANALYSIS



Chemical changes in the electrolyte surrounding samples undergoing dealioying can provide information regarding the mechanisms inolved. Analytical methods which have been used to monitor metal-ion pickup in solution include electrochemical methods [split-ring electrodes(63,88) and polarog(44,596) (7897- 00)
raphy(44,47,95',96)], colorimetry,(78 97-100) and radioactive tracer techniques. (99) Alloy systems investigated include the copper- inc (44,63,S8,95,100) copper-gold,(78,97) and copper-nickel(47) systems.

Of particular interest is whether or not the more noble species of a binary alloy dissolves during dealloying. (44,63, 78,88,97-99)

Colorimetry and polarography are the only two techniques which have been reported heretofore which can detect the presence of trace elements in solution. (47'101) Atomic absorption is a technique having detection levels as good as or better than either colorimetry or polarography,(100) and this is the method employed in the present study.

Fisher and Halperin(78) report that no gold was detected during thoir colorimetry experiments with copper-gold alloys. This is in agreement with the colorimetric data of Pickering and Byrne on the same system. (98)By contrast, i a r54n
S4





5S



copper dissolves from copper-zinc alloys under certain dealloying conditions. (44,96-97,99)

Marshakov and coworkers(44) introduced the concept of

a "dezincification factor" which can be defined by the equation

Z = (Zn/Cu) solution
(Zn/Cu) alloy



The (Zn/Cu) ratio in solution is determined by chemical analysis of the solution, and (Zn/Cu) alloy is the ratio of weight percents of zinc to copper in the alloy. Marshakov and coworkers studied dezincification under a variety of conditions in both NaCl and HCl solutions. They state that alpha brasses in acid media have a dezincification factor slightly in excess of unity. This would mean that the ratio of zinc to copper in solution is slightly greater than it is in alpha brass. No copper was detected in acid solutions which had been in contact with beta and gamma brass (Z = -). No other reports on solution analysis have appeared for beta brass.

The dissolution of copper from alpha brass has been

interpreted as evidence for a redeposition mechanism,(44,99) or as an indication that the dezincified copper layer was undergoing dissolution.(47,88',95) Recent radiotracer experiments indicate that exchange of copper between a solution and a copper-containing metal surface can occur.(99)






56



This would seem to indicate that the presence of copper in solution from a dealloyed metal could also be interpreted as being the result of varying dissolution rates or, in other words, a preferential removal process in which one metal dissolves faster than the other.

Solution analysis has been used to measure the rate of dealloying as a function of time and/or temperature. (47,78, 95,96 ) Rubin (47) reported the "partial apparent heats of activation as a function of composition" for the dissolution of a series of copper-nickel alloys, some of which denickelified, in 1N HCl. The total apparent activation energies, the sums of the two partial values, varied from 5 to 10 Kcal/mole of alloy. The lower figure, corresponding to pure copper, compares quite well with the 5.44 Kcal/mole obtained by Halperin (102) for the dissolution of copper in ammonia solutions. The values reported are also in the general range described by Vetter for the dissolution of metals in electrolytes. (103)

Fisher and Halperin reportedd that copper-gold alloys showed a decrease in dissolution rate with an increase in temperature for all of their alloys. They also observed parabolic corrosion rates. Parabolic corrosion rates are characteristic of systems in which the rate is determined by-transfer of reactants through an adherent surface film which thickens as the reaction progresses. (104 ) Thus they concluded that the rate was limited by a surface film,






57



probably Cu 2 0 and/or CuO, which formed under stagnant conditions within the residual gold sponge. Higher temperatures would favor precipitation of copper oxides and the formation of denser film structures, thus accounting for the observed inverse temperature dependence of the reaction kinetics.

The concentration of metal ions in solution has also been used to calculate metal-dissolution currents. (97,98)

The results of these investigations are discussed in the section on electrochemical investigations.






58



Present Work


Atomic absorption spectrophotometry was used to analyze SN HCl solutions which had been in contact with freely corroding alpha and beta brasses. Information was obtained on dissolution rates, on the presence or absence of dissolved copper from beta brass, and on the effect of temperature on the reaction kinetics.

While valuable information can be gained from analysis of the electrolyte, this procedure cannot be used as the sole investigative method in the study of dealloying phenomena because of certain severe limitations imposed by the character and structure of the dealloyed metal surface. The morphology of the dezincified copper sponge can have an important effect on dissolution kinetics. The possibility exists that precipitated reaction products can form a surface film which retards dissolution. The surface area of the sponge is substantially greater than that of the alloy at the corrosion interface. This means that sites for the electrodeposition of copper are increased as well as the surface area for dissolution of copper from the sponge. Blocked or narrow passageways can lead to stagnant conditions and the precipitation of salts which would be soluble in the bulk solution. All these effects can alter the metal concentration in the bulk solution, and their possible influence must be considered in the analysis of data on metal dissolution.





59



The exposure method used for these tests was similar

to that reported by Fisher and Halperin. (78) Reaction kettles were used to expose samples, cut from the same ingot and having the same surface area (2.38 cm2 for alpha brass, 1.53 cm2 for beta brass), in 750 m! of argon-sparged 5N HCl at various temperatures. Tw:enty-five milliliter aliquots of acid were removed from the cells at 12-hour intervals. Each aliquot of sample was replaced with an aliquot of the stock solution used to fill the cell. This caused a dilution of 3.3 % (25 ml out of 750 ml total) each time a sample was removed. Corrections were made for these dilutions in the data which follow.

The aliquots were analyzed using a Heath Model EU703-D Atomic Absorption Spectrophotometer (see Appendix 1). Both copper and zinc concentrations were determined for each sample.

Data obtained from atomic absorption determinations are tabulated in Tables 2 through 7. Atomic absorption data are reported to only two significant figures consistent with the accuracy of the method.

Figure 24 shows an atomic-absorption calibration curve for copper. This figure shows that copper can be determined to at least the nearest 1/4 ppm with a reasonable degree of certainty. However, the precision, or the relative accuracy of the measurement when compared to the amount of metal







60












C
r--4 r--l r--4 r-A r-A r-4 r-I r-4 q f--4




0 00 r-I tn %0 m C) to Ln
.,q r--4 r-4 r-i r-q r--l cq cla cq
4-J
0 0 41 z
Ln LH "-I ::r 00 C) Ln rl tn r-I tn 1--t 11:t
41) 0 u r--4 r-4 cq to tn t1o t-I t1o
r-4 tn

Ch

+-) tn
cis 4-) Cd 0 r. tn LO 00 cq -ct r- C:) C)
r-I I-- =- -H -r-4 t-4 r--q r- r-A r-I to
r-i cz tx 4-)
u 41 -H 41 0 :j
0 0 Cd 0 r--4 z m r-I r- r- cn (14 :r
E- t Q) 0 U tn tn V) :t
.0 tn
Ln

00
V) r--4 Ln r- r- r-4 C)
r-i 4J 4)
tn CIS > C r4 r4 cl; C4 t L 1
tn 4-) W 0
cis 0 -r-4 E
4 E- (1) (D Lr) r-q 00 r- %0 r-I 00 m "-I tn

u c; 1 r4 1 t 1; L 0 C;
Cd


,a r--4 t*l 11* I t Itt %0 %D 1- 00 00 m
Cd
E 1 C; C C C
>
0 to 0
4-4 00 cl 00 -:1,
%0 ko 00 C) r--i C) r--4 cq cq
cd
41 C C C; 14 1-4 r4 1 1
C13



0 0 cis 0) r- vd, %0 C) r--4 I:dl 1 0 r-I
.rj -H r-4-3 4-) 4J

(M r-I %-0 00 I t rq Lt") % o C:)
0 0 r--l r-I cq K) tol te) tn tn r-I
tn V)
o 9
cq Ln 00 -o 00 cq lzt ID r-4
r-q r-4 r-I cli C14 tn t-I V)
u tn
4J

0
m Ln t-- Ln 00 LO Ln 00 -4
5 r--q to Rzr


r-I


eq It -c 00 C:) 00 C) C13
tn 1-d' \ o r- 00 C) CD
r-I 0 r-I








61


:t
41 1 1 1 o4 14 14 14 4






4J 0 m r--q r, tn tn r- cn C>
.,y -,y tn Ln c r- 00 (n ry
4-j oy
0 4J :3
14-4 ol
0 0 Ln C:) oy *tztl C) C) C) C> r-y
oy En oy tn %D C) oy tn Ln t- 00
oW MA oy my my


Q
0 M 0) r-I Ln C) I'D Lf) oy V)
Cl 4-j M 0 Lf) %.0 r OD 0) C)
Ln r-4 -C C ol ry oy
cis to 4-J
00 +j _ry 41 r ::5
m 0 C) Z (1) r--i Itt C) 00 rl- C) C) CD CD CD
4 0 0 :z oy tn U) 00 CD tn q* Ln t4-) 41 c tn u oy o-4 oq oy oy o-I
0

r-y
cu o-q
V) CD rq M 00 Wy

z oy 41 0) t-4 oy eq tn Ln 00 oy CD
LO 4 > r-4 oy r--4 rlj
to 0
0 ry I:j
(D Ln Ln Ln Ln
4 z
u C) mt 00 VO 0) 11:4*
M oy oy oy tn
cis


tn r-I
cis to r- C14 t- m cq %0 C l r-I \0
4
P4 > 14 14 1; 1 C4 t 1 t
'o oy bt 0
Cd < H E
00 Ln Ln

0 u oy tr) 11; 4 L4 10
4-4

ccs
4-J Ln
cli r0) ?--q C:) 00 00 C) 00 Itt \0
0 ow C*4 tn Ln Ln \0 r- cc m C)
.H H r-A
4-) +J 4-J
:3 : :
r-q Ln CD C) C=) C) CD C) C) C) C)
0 0 oy tn \0 0) C=) tn Ln 10 00 to
tA V) u oy oy r--( ry r-I r-I
C)
r-I
.of Ln Ln
u 9
.'y 4-) t-j cq 00 00 t- r- C) C> CD CD C> ry
r Z "y r- C) CD r1l C14 11*
0 : oy r-4 r-4 oy M
4-J 0 P4

C) C:) CD C) C:) C) C) C) C) CD to
r-j lmr 00 mt \0 00 C) oy lt:t ry
cu m4 my my my CY 64 cq Q)
1:
r-4
oy

CD 4 C14 Itt %0 00 C) cq t %D 00 CD Cd
=$ oy to -:t \.O t 00 0) C) cq
0 o-I r-q








62


Ln %0 C:) \0 Irr Irr
tn I zt rl- \0 \10



C) CD CD CD CD
D CD C) C) CD
rl- %0 \.D C) C)
+j 0
.-I r--i
4J
0 +j z
4.4 r-I
0 0 Itt CD CD C) C:)
r-4 tn Ln Itil 00 co
r-i r--4



C) C> CD C) C) C)
0 C) C) CD C) C) CD
C:) V 00 00 C) C:) C) C)
C r 0
4-- -H -H :t C>
cz w -j r-q r-I r-4
Lr) 4J -r-4 +J
0 (1) Cd
41 a) 0 :3 C) C) C) C) CD
Cd 4-) C V) Ln C) CD eq
tn tn
r--4


C:) C C) C> C) C
00 \.D C) C)
Ln co r--4
r--l 4J (1)
CTJ 4-- > r--4
+j bb 0
0
tn C) \.D \0
M
cz 4 U) C




r-I 4-) C> C) CD CD C:) C)
.0 0) 0 \0 \D %D 00 \0 m
Cd +j a) T-q tn Lr) \.D

W 0
0 m 00 C14 Ln
4-4
tn Ln \.D C)


Cd

C) C) C) C:) C) C:)
CD C C) C CD C)
0 00 r- 0) CD C) C)
.,4 0 cd
4J F-4 r-- -J, r- r--i r- C:
4J 4-) r-i r-4 C14
4 :j :
0 q %.
V) 0 r- C) CD 00 CD C) LO
V) Ln r-I Ln 00 00 C I
r-I V-4 r--4 t-*) C)

u
.H t1r) C) CD C> C)
rz 4-) eq \.D C) Lr) C14 OD
0 r. r--i r4
:j
4-)
LO 19
r- Lr) Ln 00

cl; 14


r-q
r-q

00 Cl Cd
tn "I*, \0 r0







63


r-4 r--q \.c CD tn
r-4 Ln 1 0 r- 00



Ln C:) C) C) CD CD
C:) C) CD CD C:) CD

+j z 0 t1i : C Ci Ci Ci
-,4 -H tn to C:)
r--4 r--4 V)
0 41
LH
a 0 CD C) C:) C) C:) C:)
r-4 tn 00 cn Ln C)
tn tn



C) C) C:) C) CD
0 C) C) CD C CD C)
C m tn 00 C) C:) C=) C)
4-) CZ 0
r-4 H tn tn
M 4-) r-i r-4 C14 V)
4-J -H 4-) ;:: :
0 cz r--4
4-) -r 0 C) Cl C:) C) C) C)
CIS 4-) 0 V) tn C) 1-:31 m Itt

r-4 L)

Lr) C) C:) C) CD CD
r- t- CD C) C>
Ln Ln r--4 m C:)
r--l 4J
cd x > r--l V-4 t1o
4-)'tC 0
0 -1 rg
M

cz L) Itt C) 00 00 C) Rt
Lr)
Lf)

0 Cd
r-4 4J Ln C) C C: C) C)
c 0 C) CD
r-4 "tt Ln 00 C)
4-) (1)
4 >
0 bAD 0
4-4
LO Ln
C13
4-) U qtll 110 r C) r4
cis



0 C) C) C) C) C) C)
.F.4 C) C) C) C> C C)
4-) 9 I:t 00 C> C:) C) C)
0
-r-i Cd r--l tn r-) r- Ln r1l
0 4j e-- ?--I rn
tA : 4J
r-4 : Ln
0 C:) C) C:) C) CD CD I
tn 0) C) \10 C) C
u tn


0 00 C) CD LO C:) C)
4-) 4-) Lf) %D to
r. tn RT M
:3 4J
0 P4 C:
rl- Lr) 00 IZI" to
Z
1; LA




a) $-4 cq Itil \.o 00 C:) C14 Cd
rz: ::s
.rq 0
F-4






64



Table 6

Atomic Absorption Data for Beta Brass
in 5N HCI at 89.35'C





Time Amount in solution Weight
(hours) (ppm) (wt)a removed
Cu Zn Cu Zn Cu Zn

12 10 420 780 32,000 26 1,100

24 32 1,200 2,400 94,000 80 3,100

36 60 2,000 4,500 150,000 150 5,000

48 100 2,800 7,500 210,000 250 7,100

aAll weights in 10-5 gm.






6S



Table 6 (Extended)





Total
we ight
Total that has Amount
weight been in left in
removed solution solution Z
Cu Zn Cu Zn Cu Zn

26 1,100 780 32,000 750 31,000 44.6

110 4,200 2,400 95,000 2,300 91,000 42.6 260 9,200 4,600 150,$000 4,440 140,000 36.5 510 16,000 7,800 220,000 31.3






66


Table 7

Atomic Absorption Data for Beta Brass
in 5N HC1 at 99.850C




Time Amount in solution Weight
(hours) (ppm) (wt)a removed
Cu Zn Cu Zn Cu Zn

12 34 1,200 2,500 94,000 84 3,100

24 150 3,100 11,000 230,000 380 7,800

36 380 4,600 28,000 350,000 940 11,500

48 820 5,500 62,000 410,000 2,100 14,000

aAll weights in 10-5 gm.






67



Table 7 (Extended)





Total
weight
Total that has Amount
weight been in left in
removed solution solution Z
Cu Zn Cu Zn Cu Zn

84 3,100 2,500 94,000 2,400 91,000 40.70 460 11,000 11,000 240,000 11,000 230,000 22.85 1,400 22,000 29,000 360,000 27,000 340,000 13.64 3,500 36,000 63,000 430,000 .-- 7.50








68


























cc 0
CL. u
0.

B. 0
C.


CW) z








Cl.0






060



4w


-4-
B. 1-5
0
4-)





69



available for measurement, is lower for more dilute samples. As an example, a 1/4 ppm deviation in a sample having 2 ppm copper yields a 12-1/2 percent error, whereas in a sample having 8 ppm copper the same 1/4 ppm deviation would produce an error of only 3-1/8 percent. This type of precision deviation is common to all atomic absorption methods. The

confidence limits shown on the figures which follow were calculated by assuming a possible -1/4 ppm deviation for the original instrument reading obtained on the sample in que s t i on.

The Heath spectrophotometer is a single-beam instrument. This means t-at no provision is made for automatically adjusting for instrumental drift due to power deviations or lamp fluctuations. (05) It was found in practice that the

instrument would drift significantly in a short period of time, thus negating the value of a calibration curve. For this reason, a procedure was developed whereby each sample was introduced into the absorption flame once to determine

the general range of the concentration of the element being anal.yzed. Then standards were introduced into the flame to confirm the concentration range of the unknown. The values tabulated in Tables 2 through 7 were then determined by introducing one standard, then the unknown, and then another standard, so that each unknown reading was bracketed by two standard readings.





70



The limits of detection for the unit under the operating conditions used in this investigation were approximately

0.15 ppm for zinc and approximately 0.25 ppm for copper. Theoretical limits of detection for these elements by atomic absorption aae somewhat lower than this, but the solutions analyzed had copper and zinc contents well above these liinits, and most of them required dilution to be brought within the operating range of the instrument.

Figure 25 shows the dilution-corrected values for copper and zinc dissolved from alpha brass at various times (see also Column 4 of Table 3). The zinc values obey linear kinetics, but the coppel values fluctuate.

Copper and zinc dissolution from alpha brass at 98.50'C is shown in Figure 26. Once again, linear kinetics are observed for zinc,and the copper values fluctuate somewhat before they too become linear.

The deziicification factor, Z, that was introduced by Marshakov and coworkers(44) is plotted for this cell in Figure 27. When the confidence limits are taken into consideration, Z for alpha brass is seen to be between 1 and 1.5 for this investigation. This is in general agreement with Marsh,kov and coworkers. The usefulness of determining Z is limited since the absolute magnitudes of the values of Z are strongly influenced by the precision of chemical analysis methods in processes (such as dealloying) where the quanttitis 1heine1 an.i'vzeJd are small.








71










c
CL.
Im. 2c
cz c :1




4J
Ln LH
0


tn 0 (A H
cl

'm u







0
$-4 0 4-4

CD
0 tA .r-4
co 4J

0 M
tA $ 4




u u




0
Cd Ln r--l V) C13




0 4J u CTS 0 cm C=p
w

_PL X W2 IV13w a3AIOSSia jo IND13M Ln









72






















Ln

r- 0


co (n 0
V) -H
tA
H
'a u













0 tn .r-j
co +j
cr :z

0 CA
tn
tn
-rq 4j 0 '1:$ 4J

u u $-4

C-4 tn



0
cd c) --I Ln CIS 4 9
la) 00 -H P. C) tz C=) CD CD C H
co 0 4-) 4
u Cd 0 x W2 IV13W G3AlOSSla 30 1HO13M








73
















0
C=p


u

Ln -4


-r-i 4.4
0
W
tn tn
Cd 4J
.,i
r

CKS r--4



CN cis 41


0 0
4-44 (n M 4J r



co (n Cd -4
0 (n
4J -4 CIS u a) a) cri +j Ei LH 4J cfj --I
u
0 tn

4J bo
cis r-I u u ;A
-H 0 0 LH C)
-H Ln U .4
CO 4J

$-4
0 4J 0 CR Lq CIS 4-4
C=





74



Figures 28 and 29 show the corrected amounts of copper and zinc which have dissolved from beta brass at 67.500C as a function of time. They are not plotted on the same scale, as was done in Figures 25 and 26 for alpha brass, because the dissolution rate for zinc (Figure 29) is much greater than the dissolution rate for copper (Figure 28). The reaction kinetics for both copper and zinc appear to be

linear after the first 24 hours.

Slight fluctuations in the copper or zinc dissolution

kinetics make disproportionately large changes in the magnitudes of the dezincification factor Z. This is seen in the positions of the calculated points in Figue 30. The values shown in Figure 30 for beta brass are up to fifty times the values plotted in Figure 27 for alpha brass. It is emphasized here that the values for beta brass are finite in contrast to the results of Marshakov and coworkers who reported dezincification factors which were infinite for beta brass because no dissolved copper was detected by ther. This again illustrates that Z has no fundamental significance with regard to the mechanism of dealloying.

The effect of temperature on the dissolution kinetics of zinc from beta brass is shown in Figure 31 where the amount of zinc dissolved for beta brass is plotted versus time for four different temperatures. There is an obvious increase of zinc dissolution rate with temperature.








7S
















0 C)
Ln

r_: CD
LO








co Ln




V)
M CIS


co C-7 CD C





0
-4
44

CN
0
i 4-1


0

V)
CN H
10

4








CN co


x W2 HUM G3AIOSSIO 30 1HO13M









76





















0
C)
Ln CD co











Ln







Cd



Cd
-P
(D



0

C*Q


0
.H
4J
















CD C CD C=
CD CD



Cpi x W2 3NIZ 03AIOSSIa 10 1HO13M









77
















0
.H


1= u

z $-4 Ln rL

LH
0

tA
M 4J







,c 41


0 0 co tn

V) 0


tn cz 4 $4
0
+j $-4 Cd u a) C) CN 0 +j r
t4-i 4J CIS r-i u Cd
0 CIO


cts
u u
-HO 0
CN 4-4 C)
-r-4 LO Q) U .4
9 t- 4-)


4-) 0 Cd



CD CZ) CD CD C:)
nr C14

N






78








5000 [




4000

99.85 0C

E
3000




2000

8&35 0C
C.

B 100067.50 0C 59.60 0C

0
0 12 24 36 48
HOURS



Figure 31. Zinc dissolution from beta brass in SN HC1
at various temperatures. Scatter bands show
the limits of precision for the original
measurements.






79



Many chemical reactions, including metal dissolution, are found to obey an empirical equation first proposed by Arrhenius

K = K e-Q/RT


where the pre-exponential factor, K 0, is usually found to be temperature-independent, at least within the experimental accuracy of the observations, (104,106) and Q is the so-called activationn energy." K and K 0in the above equation are rate-related measurements such as weight loss, weight gain, depth of penetration, and metal dissolution for corrosion experiments. The amount of zinc dissolved in 48 hours was used in the discussion which follows. This is a reasonable choice since Figure 31 shows the kinetics to be linear. R in the Arrhenius equation is the gas constant, 1.986 calories gm-mole 1 degree K-1,1 (04) and T is the absolute temperature in degrees Kelvin.

If reaction kinetics can be represented by the above Arrhenius formula, then a plot of log K as a function of l/T will give a straight-line having a slope of -Q/2.303R. (104, 106)

Figure 32 is an Arrhenius plot of the amount of zinc

dissolved from beta brass in 48 hours. The activation energy for zinc dissolution over the temperature interval 59.6099.8S0C is found from the slope of this graph to be 18 Kcal gm-mole 1. This value is somewhat higher than those for






80



5






2




E




0

M 0.5C


0.2






0.1
I lI -I
98.85 89.35 67.50 59.60
TOC
I I I I I
2.6 2.8 3.0
1000/T K"



Figure 32. Arrhenius plot of zinc dissolution rate versus
1000/T. Scatter bands show the limits of precision for the original measurements.





81

-I(103)

electrochemical metal dissolution (10 Kcal gia-mole 1) the dissolution of copper-nickel alloys as reported by Rubin (5-10 Kcal gm-mole-), or the dissolution of copper reported by Halperin. (102)

Kofstad points out the difficulty of ascribing physical

interpretations to experimentally determined activation
(101)
energies.

Several conclusions can be drawn from the atomic-absorption data. There can be no doubt that at least some copper enters solution from both alpha and beta brass when freely exposed under conditions used in these tests. This contradicts the results of Marshakov and coworkers(44a who reported that no dissolved copper was detected in solution for their experiments with beta brass in O.5M NaCl and in

0.SM HCI.

There is evidence for a linear dissolution rate for

zinc from both alpha and beta brasses (see Figures 25, 26, 29 and 31). This would support the metallographic observations of Langenegger and Robinson. (62)

The irregular dissolution kinetics for copper seen in Figures 25, 26, 27 and 30 are believed to be due at least in part to the effects of the morphology of the "spongy" dezincified surface on the corrosion processes.

A comparison of the data obtained from alpha and beta

brasses confirms that beta brass corrodes much more rapidly than does alpha brass in hydrochloric acid.






82



The dezincification factor introduced by Marshakov

and coworkers provides an indication of whether or not dezincification has occurred.

However, the factor Z has no fundamental significance and the absolute magnitude of the dezincification factor may vary considerably for a given exposure cell and cannot be compared with the values from other cells. The sensitivity of the dezincification factor to the irregularities of the dissolution kinetics of copper during dezincification is well illustrated.













ELECTROCHEMICAL INVESTIGATIONS



Early), electrochemical tests measured the potential of brasses in solutions known to produce dezincification.(69) At that time it was not generally recognized that duplex alloys, such 3s alpha-plus-beta brasses, would exhibit microstructure-dependent electrochemical behavior.

In 1967, Joseph and Arce(107) showed that the corrosion behavior of a 63 w/o Cu 37 w/o Zn brass was strongly dependent on structure. This particular alloy can have either a single-phase alpha or a duplex alpha-plus-beta structure, depending on heat treatment, as revealed in the copper-zinc phase diagram, Figure 5. Several recent electrochemoical studies of dealloying have examined the behavior of single-phase copper alloys and of more complex copper alloys and have taken microstructure into account.(44,97,108-110)

Some researchers have used driven anodes as a rapid

means of producing specimens to be examined by non-electrochemical means.(62-65) Others have used electrochemical observations to arrive at conclusions as to the mechanism of
dezin ifiction (43,88,95,96,1_12)
dezincification. (43,S,9961) Sti]l others have tried
(4,98,112) (44,86,97,98, 110,113)
to define current, 9 12 potential, (4,869798,110,113) or potential -pH(58'86,1jI,15) conditions where dealloying




83





84



might be expected to occur, or to determine potentials where cathodic, protection may be possible.(3,116-119)

Electrochemical tests have a number of drawbacks and

may be subject to misinterpretation. For example, constantpotential tests conducted for short periods, e.g., two hours, at room temperature which fail to produce observable dezincification(0) may produce measurable dezincification after longer periods of time. (3,44,86)

Electrochemical tests on alloys subject to dealloying have an added difficulty in that the surface of the sample will quickly change to that of a dealloyed metallic "sponge" if the electrode is subjected to dealloying conditions.(44'86' 98,113) Sugawa-ra and Ebiko(86) point out that thick anodic films, such as those caused by dezincification, may produce additional resistance, and thus a potential drop, between the surface and the uncorroded portion of the sample. This means that the true potential of the uncorroded sample cannot be determined. However, this potential change, which is expressed by Ohm's Law, E = IR, is very small, because the currents being measured in a typical electrochemical cell are of the order of 108 to 10-2 amperes. (71'98)

Wilde and Teterin(113) reported that their anodic

polarization cirve: for alpha brasses and duplex alpha-plusbeta brasses were almost identical with the curve for pure copper. The only measurable differences were in current

density. They attributed the similarities to a copper layer









which formed on the sample surfaces. Similar results were
(86)
reported by Sugawara and Ebiko 86 for alloys in the same composition range. Alloys having 47 w/o zinc (beta brass) or more were reported as having anodic polarization curves which became more like that of zinc as the zinc composition increased.
( 44)
Marshakov and cov.'orkers' stated that "The anodic

behavior of alloys is determined by the rate of dissolution of the noble component as the sloiest stage." They investigated single-phase alpha and beta brasses as well as duplex alpha-plus-beta brasses.

Pickering and Byrne(98) investigated the effects of a dealloyed sponge on the surface of a sample by holding copper-gold alloys at dealloying potentials for varying periods of time and then "jumping" the potential to a more noble level. They showed that below a certain "critical potential" the copper-dissolution current from copper-gold alloys was dependent on the rate of copper solid-state diffusion from the alloy. Above the critical potential the copper-dissolution current became strongly potential-dependent. This critical potential was shown to be compositiondependent for the copper-gold alloy system, which is a homogeneous alloy system over the composition range investigated, as well as for copper-zinc alloys,(97) where phase changes

could be interpreted as causing shifts in the critical potential.





86



Some of the potentials that Pickering and Byrne investigated were below the hydrogen evolution potentials for the electrolytes they used. Because there was no accurate way to subtract the hydrogen-evolution current from the measured electrode current, they used chemical analyses of the metals in solution to determine dissolution currents. This also allowed them to plot the dissolution currents for each dissolving species separately at potentials where the more noble metal in the alloy was also dissolving.

Potential-versus-pH plots (Pourbaix diagrams) of regions of chemical stability of elements, their ions, and their salts in aqueous environments, present one possible way of providing a basis for predicting the tendency for dealloying. Latanision and Staehle,(114) Verink and Parrish,(115) and Verink and Heidersbachl58) have suggested that the superposition of Pourbaix diagrams(120) for the constituent elements of the alloy may reveal a region of potential and pH where one element would tend to corrode while the other(s) would be immune. Latanision and Staehle later concluded that their hypothetical denickelification of Fe-Ni-Cr alloys was not substantiated by their experimental observations.(121) However, the copper-nickel system suggested by Verink and Parrish and the copper-zinc system discussed by Verink and Heidersbach both are widely reported to deallov in service.(2 ,13-15 ,89 ,122 ,123)






87



Another advantage of the Pourbaix diagram, or potential pH, approach to dealloying is that information obtained by various authors in different solutions can be plotted in a manner which provides meaningful correlations.






88



Present Work


The experimental potential-pli (Pourbaix) diagram for

70 w/o Cu 30 w/o Zn alpha brass in 0.1M chloride solutions is shown in Figure 33 and was determined by W. C. Fort,

1,,1 accordingg to the method of Pourbaix. (120) Figure 34 is a simplified version of the equilibrium diagram calculated by van Muylder, Zoubov and Pourbaix (123) assuming a chloride ion concentration of O.1M and all other ionic species to be present in 10- concentrations. Figure 35 is a similar simplified diagram for the zinc-H 20 systems. (120) Figures in the diagrams correspond to calculations which are listed in Appendix 8. Superposition of diagrams, Figures 33, 34 and 35, gives Figure 36. It is evident that the experimentally constructed diagram shows a number of features in common with the equilibrium (potential versus pH) diagram for copper. Verink and Heidersbach have discussed these similarities at some length. (58)

For pure copper in deaerated solutions, the zero current potential on the upward potential sweep of the electrochemical hysteresis circuit indicates the position of the so-called "immunity" line at a given pH. The potential at which this zero current is observed often is fairly close to the calculated position of the metal/metal ion coexistence potential for a metal ion concentration of 10- molar. Thus the arbitrary choice of 10-6 molar as a definition of






89




1.8


1.6 o ZERO CURRENT POTENTIAL
\ PASSIVATION POTENTIAL
A + PROTECTION POTENTIAL
\ A FIRST MAXIMUM
A SECONDARY MAXIMA
1.2


1.0


.8
CORROSION ... A

A NA

.4A
PASSIVATION

.- .2 -4 .-44-#--+++4 -+ +- --0 o o

-.2

s IMMUNITY
*.4 ...


-.6 ... .3 4 5 6 7 8 9 10 11 12 13
pH



Figure 33. Experimental potential versus pH diagram for
70 Cu 30 Zn in 0.1M Cl- at 25oC.(77)




Full Text
55
copper dissolves from copper-zinc alloys under certain de-
(44,96-97,99)
alloying conditions.
Marshakov and coworkers introduced the concept of
a "dezincification factor" which can be defined by the equa
tion
7 (Zn/Cu) solution
(Zn/Cu) alloy
The (Zn/Cu) ratio in solution is determined by chemical
analysis of the solution, and (Zn/Cu) alloy is the ratio of
weight percents of zinc to copper in the alloy. Marshakov
and coworkers studied dezincification under a variety of
conditions in both NaCl and HC1 solutions. They state that
alpha brasses in acid media have a dezincification factor
slightly in excess of unity. This would mean that the
ratio of zinc to copper in solution is slightly greater than
it is in alpha brass. No copper was detected in acid solu
tions which had been in contact with beta and gamma brass
(Z = o). No other reports on solution analysis have ap
peared for beta brass.
The dissolution of copper from alpha brass has been
(44 991
interpreted as evidence for a redeposition mechanism,v J
or as an indication that the dezincified copper layer was
undergoing dissolution.^^^^^ Recent radiotracer ex
periments indicate that exchange of copper between a solu-
(991
tion and a copper-containing metal surface can occur. J


144
119.A. H. Taylor, Journal of the Electrochemical Society,
118, 854 (1971).
120. M. Pourbaix, Atlas of Electrochemical Equilibria in
Aqueous Solutions, Pergamon Press, New York, 1966.
121. R. W. Staehle, Private communication.
122. K. D. Efrid, MS Thesis, University of Florida, 1970.
123. J. Van Muylder, N. deZoubov and M. Pourbaix, Report
No. 101, CELBECOR, July, 1962, Brussels, Belgium.
124. J. D. Harrison and C. Wagner, Acta Metallurgica, 7,
722 (1959).
125. F. L. LaQue, Private communication.
126. E. D. Verink, Jr., Quarterly Report for the Period
Ending June 30, 1971, submitted to the Office of
Saline Water, U.S. Department of the Interior, by the
Engineering Industrial and Experiment Station, Uni
versity of Florida, Gainesville, Florida.
127. L. S. Darken and R. W. Gurry, Physical Chemistry of
Metals, McGraw-Hill, New York, 1953, p. 249.
128. N. Ohtani, Journal of the Japanese Institute of Metals,
30, 729 (1366).
129. R. Covert, Private communication.
130. C. L. Bulow, Private communication.
131. P. R. Swann, 'Corrosion, 25 147 (1969).


X-RAY DIFFRACTION
X-ray diffraction and electron diffraction have been
used by several researchers in an attempt to ascertain
whether dealloying is a process involving the selective re
moval of one constituent or a dissolution-redeposition pro
cess .,52,63,6479'83) If a redeposition process occurs,
the diffraction pattern should show diffracted radiation
due to remnants of the original alloy and also due to re
deposited metal. Diffracted radiation between the two peaks
would supposedly indicate an alloy having a composition be
tween that of the original alloy and that of the final, de-
alloyed, relatively pure metal. This intermediate composi
tion, if detected, would indicate that a selective-removal
process is operative.
Most x-ray diffraction and electron diffraction studies
of dealloying have involved single-phase, binary alloys of
either copper-zinc or copper-gold. The phase diagrams for
these two binary systems are shown in Figures 5 and 6.
One of the principal limitations of the copper-zinc
system is that alpha brass is stable up to only about 37
w/o zinc at room temperature. This limits the possible
changes in lattice parameter, and thus the changes in the
resulting x-ray or electron diffraction angles, to a small
18


DISCUSSION
The two main theories which have been proposed for
dezincification have often been considered to be mutually
exclusive. Authors claiming that they have found supporting
evidence for one or the other mechanism have claimed without
further justification that the other conclusion was unwar
ranted.
Evidence has been presented herein which indicates that
each (and sometimes both) mechanism can occur. X-ray analy
sis and electron microprobe data which support a selective-
removal explanation of the dealloving process were obtained
from some of the same samples which show copper deposits.
Figure 15 shows a diffusion zone in alpha brass which had
been exposed in hydrochloric acid. Figure 16 is a photo
micrograph of this sample and shows the probe trace in the
center of the picture. Figure 16 also shows a copper depos
it at the top of the picture, thus illustrating that both
mechanisms can occur on the same piece of metal. The two
samples which produced the x-ray patterns in Figures 9 and
10 also had deposits on the surface and produced electron-
microprobe data indicating the existence of diffusion zones.
105


23
The lack of similar results in other studies of de
alloying may be due to the use of conventional film methods
f 8 51
for recording diffracted intensity.^ J


14 2
85. R. Heidersbach, Corrosion, 2_6, 445 (1970).
86. H. Sugawara and H. Ebiko, Corrosion Science, 7_, 513
(1967) .
87. L. S. Birks, Electron Probe Microanalysis, Inter
science Publishers, New York, 1963.
83. H. G. Feller, Corrosion Science, 7, 359 (1968);
Zeitschrift fur Metallkunde, 5_8 875 (1968).
89. G. M. Ugiansky and G. A. Ellinger, Corrosion, 24,
134 (1968).
90. K. J. Anusavice, PhD Dissertation, University of
Florida, 1970.
91. A. J. Forty and P. H. Humble, Proceedings of the
Second International Congress on Metallic Corrosion,
1963 p. 80.
92. H. W. Pickering, Private communication.
93. J. W.*Colby, "MAGIC--A Computer Program for Quanti
tative Electron Microprobe Analysis," Bell Telephone
Laboratories, Allentown, Pennsylvania, 1967.
94. J. OM. Bockris and A. Damjamovic, The Mechanism of
the Electrodeposition of Metals, Modern Aspects of
Electrochemistry Series, No. 3, Butterworths,
Washington, 1964.
95. J. Bumbulis and W. F. Graydon, Journal of the Elec
trochemical Society, 109, 1130 (1962).
96. T. J. Kagetsu and W. F. Graydon, Journal of the
Electrochemical Society, 110, 709 (1963).
97. H. W. Pickering and P. J. Byrne, Journal of the
Electrochemical Society, 116, 1492 (1969).
98.Ibid. 118 209 (1971).
99.R. Potzl and K. H. Leiser, Zeitschrift fur Metallkunde,
61, 525-527 (1970).
100. L. H. Jenkin and R. B. Durham, Journal of the Electro
chemical Society, 1M7 768 (1970).
101. L. L. Leais, "A Comparison of Atomic Absorption with
Some Other Techniques of Chemical Analysis," in Atomic
Absorption Spectroscopy, ASTM STP 443, ASTM, Baltimore,
Maryland, 1969. "


m rf
Table 8
Alpha Brass Potentiostatic Test Data
Copper
'otential
esiie
Dezincified
Surface
Appearance ?
Was
Solution
Stirred?
Duration
of
Exposure
Zinc
in
Solution?
Copper
in
Solution?
on
Platinum
Electrode?
Remarks
+0.150
No
Yes
2 2 hours
Yes
Yes
Yes
+0.150
Yes
No
22 hours
Yes
Yes
Yes
+0.050
No
Yes
5 days
Yes
No
Yes
+0.050
Yes
No
5 days
Yes
No
Yes
-
-0.050
Yes
Yes
10 days
Yes
No
No
-0.050
Yes
No
10 days
Yes
No
No
-0.750
?
Yes
8 days
Yes
No
No
Surface has
black tarnish
at end of
test
-0.850
?
Yes
12 days
Yes
No
No
Surface has
black tarnish
at end of
test
-1.050
?
Yes
13 days
Yes
No
No
Surface has
gray tarnish
which turns
black when dry


2
The study of dealloying phenomena is fraught with a
number of complications. Generally, such reactions are
relatively slow and a lengthy exposure period is required
to cause a sufficiently great amount of dealloying to facil
itate evaluation. Consequently, there is considerable inter
est in accelerated tests for evaluation of tendencies of
alloys to dealloy. Many techniques have been employed. For
example, electrolyte compositions have been adjusted by us
ing more concentrated solutions or solutions having varia-
tions in oxidizing power. JJ Specific ions have been added
to stimulate dealloying, e.g., saturated cuprous chloride
solutions have been used to accelerate the dezincification
of copper-base alloys. Electrochemical stimulation also
has been used. Unfortunately, all too often the test meth
ods employed have been vulnerable to criticism as having
biased the experimental result and, although specific tech
niques are now available which can cause dealloying to occur
in the laboratory, nonetheless, there still is no firm basis
for predicting the likelihood of dealloying in service based
on these tests.
Single-phase alpha brasses and duplex alpha-plus-beta
brasses are the only forms of copper-zinc alloys having
engineering significance at present. Bengough and May re
ported, in 1922, that small additions of arsenic would pre
vent dezincification of alpha brasses but would not protect
f 591
the beta phases of duplex alloys.'- J
The reasons for


LIST OF FIGURES (Continued)
Figure Page
29 Zinc dissolution from beta brass
in 5N HC1 at 67.50C 76
30 Dezincification factors, Z, for
beta brass in 5N HC1 at 67.50C 77
31 Zinc dissolution from beta brass
in 5N HC1 at various temperatures .... 78
32 Arrhenius plot of zinc dissolution
rate versus 1000/T 80
33 Experimental potential versus pH
diagram for 70 Cu 30 Zn in
0.1M Cl" at 25C 89
34 Simplified CU-CI-H2O diagram at
25C for solution containing 0.1M
chloride ions 90
35 Simplified Zn-H20 diagram for
concentrations of ionic
species = 10" 6m 91
36 70 Cu 30 Zn alloy in 0.1M
chloride solution ; 92
37 Beta brass held at +0.050Vshe fr
2-3/4 hours 94
38 Typical potentiokinetic scan in
acid solutions 100
39 Cross-section of alpha brass
showing where holidays in the
stop-off lacquer caused de
zincif ication in the stagnant
regions beneath the holidays 110
40 Theoretical domains for dealloying
in a given solution based upon the
Nernst equation 114
x


110
Figure 39. Cross-section of alpha brass showing where
holidays in the stop-off lacquer caused
dezincification in the stagnant regions
beneath the holidays. The unmasked region
had no flow restrictions and suffered gen
eral dissolution. 25X


15
Sample Preparation
The alpha brass samples used in this investigation were
donated by the Chase Brass arid Copper Company. The brass
was obtained in two 25-pound lots of as-extruded cartridge
brass wire having diameters of 0.850 in and 0.687 in, re
spectively. The chemical analyses of these alpha brass sam
ples are shown in Appendix 3.
Beta brass ingots having a nominal composition of 52
w/o Cu and 48 w/o Zn were prepared from 99.99 pure copper
and zinc stock purchased from the American Smelting and Re
fining Company. The ingot melting procedures are described
in detail in Appendix 4. The tops and bottoms of each ingot
were analyzed according to the procedure described in Appen
dix 5. Ingots which varied by less than 1,0 w/o Cu from top
to bottom were then used in the as-cast condition for these
experiments. Each ingot weighed approximately 150 grams and
was 15 mm in diameter. The average compositions of the beta
brass ingots used are listed in Appendix 6.
Disc-shaped cross-sections were cut from the alpha
brass wire and from the as-cast beta brass ingots. Sample
surfaces we re prepared according to the technique described
in Appendix 7.
Five types of samples were used in this study. Samples
intended for x-ray analysis were simple discs of brass with
a hole drilled through one side so they could be suspended
from the sample holder as shown in Figure 1. These samples


Table 9
Beta Brass Potentiostatic Test Data
Copper
Potential
12 SI IE
Dezincified
Surface
Appearance?
Was
Solution
Stirred?
Duration
of
Exposure
Z inc
jn
Solution?
Copncr
in
Solution?
on
Platinum
Electrode?
Remarks
+0.050
Yes
Yes
1
day
Yes
ND
Yes
+0.050
Yes
No
1
day
Yes
ND
No
-0.050
Yes
Yes
2
days
Yes
ND
No a
-C 050
Yes
No
o
L
days
Yes
ND
No
-0.100
Yes
Yes
3
days
Yes
ND
Yes
.
-0.150
Yes
Yes
3
days
Yes
ND
Yes
-0.250
Yes
Yes
6
days
Yes
ND
No
-0.250
Yes
No
6
days
Yes
ND
No
-0.850
?
Yes
13
days
Yes
ND
No
Red tarnish
after one day
is black by
end of test
-1.050
?
Yes
7
days
Yes
ND
No
Silver-gray
tarnish visi
ble after one
day. Remains
silver after
drying.
^Comparison with the test at -O.lOOVspp suggests that copper might have been
found on the platinum electrodes after longer periods of time.


Beta Brass Free
Solution
Final Potential
66 g/l HC1
+13.4 g/l CuCl2
+0.050VgHE
66 g/l HC1
+15.8 g/l CuCl
+o.ioovSHE
1M H2S04
+ 0.05 0VgHE
30 g/l NaCl
+32 g/l NiCl2-6H20
+o.ioovSHe
Table 11
Corrosion Potential Tests
Dezincification? Reference Authors
Yes
Langenegger
and Robinson^^)
Yes
Langenegger
and Robinson^^
Yes
Stillwell and Turnipseed^^
Yes
Falleiro and Pieske
(61)
103


106
The diffusion distance's in Figures 12 and 13 are small
but are still wide enough to represent several tens of thou
sands of interatomic distances, and thus they can represent
significant diffusion distances. Pickering points out that
the diffusion interface is not flat but becomes rough, or
rippled.Although this irregular interface usually
is too small to be seen metallographically in the case of
#'92')
dealloying in aqueous environments, } the same effect has
been noted on samples subjected to liquid-metal corrosion
where the perturbations are large enough to be seen metal
lographically. This roughening results in a shorter
average diffusion path length than would be available with
a flat interface. The distances which are detected with the
microprobe are thus probably larger than the actual diffu
sion path and represent the width of this roughened area
rather than the length of the diffusion path itself. None
theless, the net diffusion zone still is large enough to
be observed with confidence using the electron microprobe.
The electrodeposition of copper from solutions having
a significant copper ion concentration can be explained by
the Nernst equation
E
+ 2.3 RT
n
log
aCuCl2
aCu(alloy)
(Eq. 1)
which at room temperature becomes
*The sign convention used herein conforms
in calculation of Pourbaix diagrams. See Eq.
to that used
76, Appendix 8.


COPPER
0
HOURS
Figure 25. Copper and zinc dissolution from alpha brass in 5N HC1 .
at 90.35C. Scatter bands show the precision of the
original measurements.


RECOMMENDATIONS FOR FURTHER RESEARCH
The results of this investigation reveal a number of
areas of research which should be explored, both for their
scientific possibilities and because of their possible
economic benefits.
The development of experimental Pourbaix diagrams should
be extended to cover all commercial, and potentially commer
cial, alloy systems. Pourbaix diagrams at elevated temper
atures are also needed. Preliminary investigations with
single-phase beta brasses reveal that the electrochemical
hysteresis technique must be augmented by potentiostatic
testing in conjunction with solution analysis investigations.
Identification of in situ reaction products, such as
the tarnishes found on potentiostatic samples at low poten
tials in this study, is an area of importance. Identifica
tion of thin films of passive species could lead to impor-
tant advances in alloy development.^ A The possibilities
of reflectance spectroscopy in the ultraviolet, visible, and
infrared spectra should be investigated.
Recent dealuminization failures in aluminum bronzes
indicate a need for further understanding of this system of
alloys.
119


DEGREES 26
hO
^3
Figure 10. (Ill) peaks from sample of 70-30 brass dezincified for
. 30 days in 5N MCI at 75C.


APPENDICES


TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS iii
LIST OF TABLES vii
LIST OF FIGURES viii
ABSTRACT xi
INTRODUCTION 1
EXPERIMENTAL PROCEDURE 5
Immersion Test Apparatus 6
Electrochemical Test Apparatus 9
Sample Preparation 15
X-RAY DIFFRACTION 18
Present Work 24
ELECTRON MICROPROBE 30
Present Work 32
OPTICAL METHODS 41
Present Work 45
SOLUTION ANALYSIS 54
Present Work 58
ELECTROCHEMICAL INVESTIGATIONS 83
Present Work 88
DISCUSSION 105
CONCLUSIONS 116
RECOMMENDATIONS FOR FURTHER RESEARCH 119
v


78
Figure 31. Zinc dissolution from beta brass in 5N HC1
at various temperatures. Scatter bands show
the limits of precision for the original
measurements.


hour. Then the furnace controllei was turned down to 850C
and held for four hours at this temperature.
The capsules were then removed from the furnace and
broken, and the ingots were quenched under a water tap.
Top and bottom sections were removed from each ingot
and analyzed according to the procedure in Appendix 5.
Grain size was quite large, and a typical cross-section
would have four or five grains visible after polishing.


LIST OF FIGURES (Continued)
Figure Page
14 Zinc intensity profile from a
sample of beta brass dezincified
for two days in 5N HC1 at 75C 36
15 Photomicrograph of sample shown
in Figure 12 38
16 Photomicrograph of sample shown
in Figure 13 39
17 Photomicrograph of sample shown
in Figure 14 40
18 Dezincification plug in 70-30
alpha brass exposed for 79 days in
IN NaCl at room temperature 42
19 Copper deposits on the surface of
dezincified alpha brass sample 47
20 Deposit on surface of dezincified
alpha brass sample 48
21 Scanning electron micrograph of
copper slab protruding from the
surface of a dezincified alpha
brass sample 49
22 Nondispersive x-ray analyzer pattern
of deposit shown in Figure 21 51
23 Dezincified cross-section of sample
shown in Figure 21 52
24 Atomic-absorption calibration curve
for copper 68
25 Copper and zinc dissolution from
alpha brass in 5N HC1 at 90.35C .... 71
26 Copper and zinc dissolution from
alpha brass in 5N HC1 at 98.50C .... 72
27 Dezincification factors, Z, for
alpha brass in 5N HC1 at 98.50C .... 73
28 Copper dissolution from beta brass
in 5N HC1 at 67.50C 75
ix


135
Cu-Cl-I^O System (continued)
Eg. #
59 4CuCl + 6H20 = 3Cu(OH)2-CuC12 + 2Cl' + 6H+ + 4e
a) E = 0.785 0.0886 pH + 0.0295 log(Cl')
62 2Cu20 + 6H20 + 2C1" = 3Cu(OH)2-CuCl2 6H+ + 8e
a) E = 0.498 0.0148 log(Cl") 0.0443 pH
67 2CuC1~2 + H20 = Cu20 + 4Cl" + 2H+
a) log(CuCl"2) = 4.45 pH + 2 log(Cl')
76 Cu + 2C1~ = CuCl"2 + e
a) E = 0.208 + 0.0591 log(CuCl"2)
- 0.1182 log(Cl")
80 CuCl"2 + H20 = CuO + 2C1" + 2H+ + e
a) E = 0.932 0.1182 pH 0.059 log(CuCl"2)
+ 0.1182 log(Cl~)
82 CuCl = Cu++ + Cl" + e
a) E = 0.537 + 0.0591 log(Cu++) + 0.0591 log(Cl")


133
18. S. C. Hamstead, Industrial and Engineering Chemistry,
50 [10] 87A (1958) .
19. A. H. Hesse, E. T. Myskocoski and B. M. Loring, Trans
actions o£ the American Foundrymen's Association, 51,
821 (1944).
20. B. F. Peters, J. A. H. Carson and R. D. Barer, Mate
rials Protection, 4_ [5], 24 (1 965).
21-. M. Schussler and D. S. Napolitan, Corrosion, 1_2, 107t
v '(1956).
22. 'J. Serre and J. Lawreys, Corrosion Science, 2i, 135
(1965) .
23. W. D. Clark, Journal of the Institute of Metals, 73,
263 (1947).
24. P. Trautzel and W. D. Treadwell, Helvitia Chemica Acta
3_4 1723 (1951).
25. L. M. Leedon, Journal of the American Water Works Asso
ciation, 3^8, 1392 (1946).
26. A. Dvorak, Strojirentsvi, 1_2 [1], 39 (1962), refer
enced by Corrosion Abstracts, 2_, 11 (1963).
27. M. G. Fontana, Industrial and Engineering Chemistry,
39 [5], 87A (1947).
28. W. Lynes, Proceedings ASTM, 41_, 859 (1941).
29. R. B. Abrams, Transactions of the American Electro
chemical Society, 4_2 39 (1922).
30. C. F. Nixon, Transactions of the American Electro
chemical Society, 4_5 297 (1924) .
31. J. H. Hollomon and J. Wulff, Transactions AIME, 147,
297 (1942).
32. D. B. Thompson, Australasian Engineering, p. 48
(October, 1954).
33. A. R. Zender and C. L. Bulow, Heating, Piping and Air
Conditioning, 16_, 273 (1944).
34. E. S. Dixon, ASTM Bulletin No. 102, p. 21 (1940).
35. F. H. Rhodes and J. T. Carty, Industrial and Engineer
ing Chemistry, 17, 909 (1925).


139
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
J. C. Scully, The Fundamentals of Corrosion, Pergamon
Press, New York, 1966.
H. H. Uhlig, The Corrosion Handbook, John Wiley and
Sons, New York, 1948.
K. Hashimoto, W. Ogawa and S. Shimodaira, Journal of
the Japanese Institute of Metals, £ [1], 42 (1963).
R. M. Horton, Corrosion, 2_6 [7], 160 (1970).
V. F. Lucey, British Corrosion Journal, 1, 7 (1965);
ibid., 2, 53 (1965).
L. E. Tabor, Transactions of the Water Works Associ
ation, 48 [3], 239 (1963).
S. S. Gastev, Izvest. Akad. Nauk. SSR. Metally, _3
(1965), referenced in Corrosion Abstracts, 6^, 446
(1966) .
J. M. Bialosky, Corrosion and Metal Protection, 4_,
15 (1947).
K. K. Marshakov, V. P. Bogdanov and S. M. Aleikina,
Russian Journal of Physical Chemistry, 3_8 [7], 960
(1964); ibid., 38 [8]', 104 (1964); ibid., 39 (6),
804 (1965).
W. H. Bassett, Chemical and Metallurgical Engineering,
27_, 340 (1922).
J. O'M. Bockris, Private communication.
B. T. Rubin, PhD Dissertation, University of Pennsyl
vania, 1969.
L.Piatti and R. Grauer, Werkstoffe und Korrosion,
14 [7] 551 (1963).
G. T. Colegate, Metal Industry, 7_3 [507], 531 (1948).
L. Kenworthy and W. G. O'Driscoll, Corrosion Tech
nology 2_, 247 (1955).
U. R. Evans, The Corrosion and Oxidation of Metals,
Edward Arnold and Company, London, 1946i
C. W. Stillwell and E. S. Turnipseed, Industrial and
Engineering Chemistry, 2_6 740 (3 934).
E. D. Verink, Jr., and P. A. Parrish, Corrosion, 26,
5 (1970).


50
40
30
O
CD
20
10
L
Tr
i
10
20
30
MICRONS
40
50
_J
60
Figure 14.
Zinc intensity profile from a sample of beta brass
dezincified for two days in 5N MCI at 750C.
O'


Table 5
Atomic Absorption
Data for
Beta
Brass
in 5N HC1
at
67.50C
Time
(hours)
Amount in solution
(ppm) (wt)a
Weight
removed
Total
weight
removed
Total
weight
that has
been in
solution
Amount
left in
solution
z
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
12
1.7
18
130
1,400
4.2
45
4.2
45
130
1,300
120
1,305
11
24
2.5
50
190
3,800
6.2
120
10
170
190
3,800
180
3,000
21
36
3
160
220
12,000
7.5
400
18
570
240
12,000
220
12,000
56
48
4
225
300
17,000
10
520
28
1,100
320
17,000
290
16,000
60
60
4.8
330
360
25,000
12
820
40
1,900
390
26,000
350
23,000
73
72
5.4
420
400
32,000
13.5 1
,000
54
3,000
440
33,000
390
30,000
82
aAll
weight
s in
10"5
gm.


4
slow process at the low temperatures encountered in aqueous
environments.
Mechanism studies were performed on brass immersion
samples placed in NaCl and HC1 solutions. No copper salts
were added to the test electrolyte except as the result of
corrosion of the test specimens. Test durations of up to
90 days were employed at temperatures ranging from room tem
perature (approximately 25C) to 100C.
Potentiostatic and potentiokinetic methods were used
to define conditions under which dealloying might be ex
pected to occur.




29
DEGREES 29
Figure 11.
Diffraction pattern from sample
brass dezincified for two days
at 75C.
of beta
in 5N HC1


43
attributed to breaks in a protective scale on the metal sur
face which allowed localized attack in certain specific loca
tions. It seemed obvious that breaks of this type would
cause flow restrictions which would lead to a high concen
tration of copper in a region adjacent to the corroding sur
face. Bengough and May pointed out that precipitation of
copper due to concentration effects seemed likely, and
for a long time this seemed to be the logical mechanism for
dezincification. Simmons later pointed out that dezincifi-
cation plugs occurred under circumstances which, if slightly
altered, would result in pits. J
Most early investigators were concerned with dezincifi
cation of condenser tubes under conditions where scale build
up, and subsequent cracks in the scale, were likely to occur.
The ideas put forth by Abrams and by Bengough and May gained
wide acceptance. However, this explanation did not seem to
cover dealloying in ship propellers and in other high-flow
rate situations.
Interest in the field remained high, and reports of
metallographic investigations into the dezincification mecha
nism continued. Bassett and Polushkin and Shuldener^)
decided that dezincification was a selective removal process
after examining large numbers of samples which had failed in
service. This contrasted with the opinion of Horton and
served to emphasize that conclusions based on metallographic
observation are often dependent on the opinions and prior
experience of the observer.


39
Figure 16. Photomicrograph of sample shown in Figure 13.
Probe trace is indicated by the arrow. Note
surface deposit of copper. 250X


TEMPERATURE C
19
Figure 5. Copper-zinc phase diagram (Metals Handbook,
1948 edition, p. 1206).


10
Figure 2. Circuit diagram of equipment used to polarize
specimen and automatically record corrosion
current density as a function of potential.


67
Table 7 (Extended)
Total
weight
removed
Total
weight
that has
been in
solution
Amount
left in
solution
Z
Cu
Zn
Cu Zn
Cu
Zn
84
3,100
2,500 94,000
2,400
91,000
40.70
460
11,000
11,000 240,000
11,000
230,000
22.85
1,400
22,000
29,000 360,000
27,000
340,000
13.64
3,500
36,000
63,000 430,000
7.50


13
Figure 4.
Assembled and exploded views of sample
holder.


and
113
B B++ + 2e~
of a hypothetical binary alloy shown in Figure 40. The com
position regions in which the activities of the metals in
the alloy become significant have been exaggerated so that
they can be shown on the graph. There are two regions of
interest--the dotted region in the upper left where metal A
will not dissolve and element B will, and the larger region
where element B will dissolve and element A will not. Both
are regions where dealloying is theoretically possible.
This argument would lead to the conclusion that the
line on a Pourbaix diagram which separates a region of non
corrosion from a corrosion region would not be displaced
radically by alloying. This is contrary to experimental
observations made during this study and to the work of
others.
The experimental observations of Pickering and Byrne
indicate that copper dissolves from copper-gold alloys at
a different potential for copper 13 percent gold than it
does for copper 18 percent gold. This confirms earlier
('97')
resultsv J by the same authors on the dissolution of zinc
from epsilon, gamma and alpha brasses.
Clearly the Nernst equation cannot explain this aspect
of dealloying. The equilibrium potential-pH diagrams for
copper and zinc, Figures


108
The approximate copper and zinc concentrations which
were attained in the three cells in which copper deposits
were observed on alpha brass are listed in Table 12. The
dezincification factors, Z, vary from 3.8 for 50C to 25
for 75C. All these values are substantially higher than
the values (slightly above unity) reported by Marshakov and
coworkers and in this report for the dezincification of
alpha brasses.
The fact that stagnant conditions can lead to dezincifi
cation is illustrated in Figure 39 where "holidays" in the
stop-off lacquer used to coat all but a certain portion of
a sample caused dezincification in areas under the holidays
while the uncoated area suffered general dissolution. Three
explanations are possible:
1. The copper ion concentration may satisfy
conditions necessary for deposition accord
ing to the Nernst equation.
2. The concentration of copper in localized
areas of restricted flow could exceed
solubility limits and cause precipitation
of copper.
3. The local electrochemical potential of
the brass surface at the corrosion inter
face could have been altered due to local
ized differences in the electrolyte within
the "flow-restricted" area.


16
were also used for the electron microprobe and optical in
vest gat ions.
For the atomic absorption experiments where metal-ion
dissolution was to be monitored, cylindrical specimens were
mounted in dental mount so as to leave exposed a circular
cross-section of known geometric area. The specimens were
then polished according to the procedure described in Appen
dix 7 and suspended in the reaction kettles by means of a
hole drilled in the dental mount. This procedure is similar
to that described by Fisher and Halperin for their calorim-
7 81
etry studies on the dealloying of copper-gold alloys. -
Disc specimens prepared according to the procedure de
scribed in Appendix 7 were used for the potentiokinetic ex
periments using a sample holder shown in Figure 4. For
testing details see the section on electrochemical test
apparatus.
Early high-temperature tests in this investigation used
samples in which all but a 1/2 cm x 1/2 cm square section of
the sample was "stopped off" using Miccroshield* stop-off
lacquer. Examination of the samples after exposure revealed
that leaks through the lacquer were present and this type of
sample was abandoned. The significance of these leaks is
discussed in the discussion section of this report.
Potentiostatic tests at room temperature were conducted
using discs with an insulated copper wire soldered to the
*Trade name, Michigan Chrome and Chemical Company.


86
Some of the potentials that Pickering and Byrne inves
tigated were below the hydrogen evolution potentials for
the electrolytes they used. Because there was no accurate
way to subtract the hydrogen-evolution current from the
measured electrode current, they used chemical analyses of
the metals in solution to determine dissolution currents.
This also allowed them to plot the dissolution currents for
each dissolving species separately at potentials where the
more noble metal in the alloy was also dissolving.
Potential-versus-pH plots (Pourbaix diagrams) of re
gions of chemical stability of elements, their ions, and
their salts in aqueous environments, present one possible
way of providing a basis for predicting the tendency for
dealloying. Latanision and Staehle, ^Verink and Par
rish, and Verink and Heidersbach^^ have suggested
that the superposition of Pourbaix diagrams^*^ for the
constituent elements of the alloy may reveal a region of
potential and pH where one element would tend to corrode
while the other(s) wrould be immune. Latanision and Staehle
later concluded that their hypothetical denickelification
of Fe-Ni-Cr alloys was not substantiated by their experi
mental observations.^ However, the copper-nickel system
suggested by Verink and Parrish and the copper-zinc system
discussed by Verink and Heidersbach both are widely reported
to deallov in service.^15^12z,lz3)


14
The effects of scan rate were investigated. Reinoehl
and coworkers^ ) show that passivation current density is
a function of scan rate, but that critical passivation poten
tials and rupture potentials are independent of scan rate,
at least for iron in IN l^SO^. A more detailed discussion
of the importance of scan rates is contained in the discus
sion section. A scan rate of 33 mv/min was chosen for this
investigation after examining scan rates from 5.5 mv/min to
667 mv/min.The selected rate was slow enough to avoid
losing any configuration of the polarization curve, yet fast
enough to minimize masking effects due to dezincification
of the sample.
After experimental Pourbaix diagrams were established
by potentiokinetic polarization techniques, potentiostatic
tests were run to verify the potential regions where de-
zincification occurs. The six-cell potentiostat described
by Cusamano^^^ was used for these potentiostatic tests.


APPENDIX 3
CHEMICAL ANALYSES OF ALPHA BRASS
Sample Number 6
Element
Weight Percent
Copper
70.47
Lead
0.006
Iron
0.004
Tin
<0.001
Nickel
<0.001
Manganese
r> o ^
Silicon
<0.002
Aluminum
<0.001
Tellurium
plus Selenium
None detected
Phosphorus
None detected
Bismuth
<0.0005
Zinc
Remainder
12 5


112
exposure tests of spinning discs of copper alloys, that de
alloying (when it occurred) was always observed near the
center of the discs w'here effective water velocity was the
lowest. The outer portions of the same discs were ob
served to have undergone general corrosion. Although the
electrochemical potential was not reported, it is likely
that it was in the region indicated by cross-hatching on
Figure 36.
Components exposed to conditions involving high rela
tive motion between the part and the electrolyte, such as
ship propellers and pump impellers,have also been observed
to undergo dealloying. Under these circumstances, it was
likely that the potential was in the region indicated by
small dots in Figure 36.
Examination of the Nernst equation as written for the
dissolution of a divalent metal ion of a hypothetical metal,
A,
c -o 2.3 RT ,
H = E + log aA+.
2.3 RT
log aA (Eq. 4)
reveals that the term relating
per in the alloy is almost neg
from 10-100 percent A (activit
to concentration). This is il
equilibrium potentials for the
to the
acti
vity
of
the
cop
ligible
for
concent:
rati
ons of
ies ass
umed
to be
Pi
-oportional
lustrat
ed by
a plot
of
the
equati
ons
= A
+ +
A


58
Present Work
Atomic absorption spectrophotometry was used to analyze
5N HC1 solutions which had been in contact with freely cor
roding alpha and beta brasses. Information was obtained on
dissolution rates, on the presence or absence of dissolved
copper from beta brass, and on the effect of temperature
on the reaction kinetics.
While valuable information can be gained from analysis
of the electrolyte, this procedure cannot be used as the
sole investigative method in the study of dealloying phenome
na because of certain severe limitations imposed by the char
acter and structure of the dealloyed metal surface. The
morphology of the dezincified copper sponge can have an im
portant effect on dissolution kinetics. The possibility
exists that precipitated reaction products can form a sur
face film which retards dissolution. The surface area of
the sponge is substantially greater than that of the alloy
at the corrosion interface. This means that sites for the
electrodeposition of copper are increased as well as the
surface area for dissolution of copper from the sponge.
Blocked or narrow passageways can lead to stagnant condi
tions and the precipitation of salts which would be soluble
in the bulk solution. All these effects can alter the metal
concentration in the bulk solution, and their possible in
fluence must be considered in the analysis of data on metal
dissolution.


Table 3
Atomic Absorption Data for Alpha Brass in 5N HC1 at 98.50C
Total
weight
T ime
(hours)
Amount in
(ppm)
solution
(t)a
Weight
removed
Total
weight
removed
that has
been in
solution
Amount
left in
solution
Z
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
12
20
12.5
15
9.375
. 5
. 31
. 5
. 31
14
9
15
9.4
1.4
24
40
28
30
21
1
. 7
1.5
1.0
30
21
30
21
1.6
36
80
48
60
36
2
1.2
3.5
2.2
58
35
61
37
1.4
48
120
67.5
90
50
3
1.7
6.5
3.9
87
49
94
53
1.3
60
140
77
100
58
3.5
1.9
10
5.8
100
56
110
62
1.3
72
160
90
120
68
4
2.2
14
8.1
120
65
130
73
1.3
84
180
100
130
79
4.4
2.6
18
11
130
76
150
87
1.4
96
200
120
150
88
4.9
3.0
23
14
140
86
170
99
1.4
108
210
120
160
94
5.2
3.1
29
17
150
91
181
110
1.4
120
240
140
180 106
6
3.6
34
20
174
103
1.4
aAll weights in 10 3 gm.



PAGE 1

7+( '(=,1&,),&$7,21 2) $/3+$ $1' %(7$ %5$66(6 %\ 5REHUW +HQU\ +HLGHUVEDFK -U $ 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH &RXQFLO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WK 'HJUHH RI 'RFWRU RI 3KLORVRSK\ 81,9(56,7< 2) )/25,'$ '(&(0%(5

PAGE 2

'HGLFDWHG WR P\ ZLIH 'LDQQH .DWKHULQH

PAGE 3

$&.12:/('*(0(176 ZRXOG OLNH WR H[SUHVV P\ WKDQNV WR 'U (OOLV 9HULQN -U IRU KLV JXLGDQFH DQG LQVSLUDWLRQ 7KDQNV DUH DOVR H[n WHQGHG WR WKH PHPEHUV RI P\ FRPPLWWHH 'U $ :DOODFH 'U 5 7 'HOORII 'U 5 : *RXOG DQG 'U +UHQ 'U 0 3RXUEDL[ SURYLGHG HQFRXUDJHPHQW GXULQJ WKH HDUO\ VWDJHV RI WKLV LQYHVWLJDWLRQ 'U 6 5 %DWHV SURYLGHG JXLGDQFH RQ WKH XVH RI DQG LQWHUSUHWDWLRQ RI UHVXOWV IURP WKH VFDQQLQJ HOHFWURQ PLFURVFRSH DQG WKH HOHFWURQ PLFURSUREH 0U : & )RUW ,,, SURYLGHG LQYDOXDEOH DVVLVWDQFH LQ REn WDLQLQJ WKH H[SHULPHQWDO UHVXOWV 7HFKQLFDO DVVLVWDQFH ZDV SURYLGHG E\ 0U : $ $HUHH 0U ( -HQNLQV 0U ( & /RJVGRQ 0U 3 .DOE 0U & 0LQLHU DQG 0U & 6LPn PRQV SDUWLFXODUO\ ZLVK WR DFNQRZOHGJH WKH ILQDQFLDO VXSn SRUW VXSSOLHG E\ WKH 1DWLRQDO $VVRFLDWLRQ RI &RUURVLRQ (QJLn QHHUV DQG E\ WKH ,QWHUQDWLRQDO 1LFNHO &RPSDQ\ ,Q DGGLWLRQ FHUWDLQ RI WKH HTXLSPHQW XVHG LQ WKH HOHFWURFKHPLFDO VWXGLHV ZDV SXUFKDVHG ZLWK IXQGV IURP WKH 2IILFH RI 6DOLQH :DWHU 7KH &KDVH %UDVV DQG &RSSHU &RPSDQ\ GRQDWHG WKH DOSKD EUDVV XVHG LQ WKLV LQYHVWLJDWLRQ DQG WKLV FRQWULEXWLRQ LV JUDWHIXOO\ DFNQRZOHGJHG LLL

PAGE 4

7KH DXWKRU LV JUDWHIXO WR KLV SDUHQWV ZKR WDXJKW KLP WKH GLJQLW\ RI KDUG ZRUN DQG WR KLV ZLIH ZKR PDGH OLIH EHDUDEOH ZKHQ WKLQJV ZHUH QRW JRLQJ DFFRUGLQJ WR SODQV ,9

PAGE 5

7$%/( 2) &217(176 3DJH $&.12:/('*(0(176 LLL /,67 2) 7$%/(6 YLL /,67 2) ),*85(6 YLLL $%675$&7 [L ,1752'8&7,21 (;3(5,0(17$/ 352&('85( ,PPHUVLRQ 7HVW $SSDUDWXV (OHFWURFKHPLFDO 7HVW $SSDUDWXV 6DPSOH 3UHSDUDWLRQ ;5$< ',))5$&7,21 3UHVHQW :RUN (/(&7521 0,&52352%( 3UHVHQW :RUN 237,&$/ 0(7+2'6 3UHVHQW :RUN 62/87,21 $1$/<6,6 3UHVHQW :RUN (/(&752&+(0,&$/ ,19(67,*$7,216 3UHVHQW :RUN ',6&866,21 &21&/86,216 5(&200(1'$7,216 )25 )857+(5 5(6($5&+ Y

PAGE 6

7$%/( 2) &217(176 &RQWLQXHGf 3DJH $33(1',; (48,30(17 86(' ,1 (/(&752&+(0,&$/ 7(676 (/(&752/<7(6 &+(0,&$/ $1$/<6(6 2) $/3+$ %5$66 %(7$ %5$66 ,1*27 35(3$5$7,21 352&('85( )25 7+( (/(&752*5$9,0(75,& &+(0,&$/ $1$/<6,6 2) %,1$5< &233(5 =,1& $//2<6 $1$/<6,6 2) %(7$ %5$66 ,1*276 86(' ,1 7+,6 ,19(67,*$7,21 (/(&752&+(0,&$/ 6$03/( 35(3$5$7,21 (48$7,216 86(' ,1 &216758&7,21 2) 327(17,$/ 9(5686 S+ ',$*5$06 )25 7+( &X&+ 6<67(0 $1' 7+( =Q2+ 6<67(0 %,%/,2*5$3+< %,2*5$3+,&$/ 6.(7&+ YL

PAGE 7

DJH /,67 2) 7$%/(6 6DPSOH :HLJKW ,QIRUPDWLRQ $WRPLF $EVRUSWLRQ 'DWD IRU $OSKD %UDVV LQ 1 +& DW r& $WRPLF $EVRUSWLRQ 'DWD IRU $OSKD %UDVV LQ 1 +& DW r& $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& $OSKD %UDVV 3RWHQWLRVWDWLF 7HVW 'DWD %HWD %UDVV 3RWHQWLRVWDWLF 7HVW 'DWD $OSKD %UDVV )UHH &RUURVLRQ 3RWHQWLDO 7HVWV %HWD %UDVV )UHH &RUURVLRQ 3RWHQWLDO 7HVWV $SSUR[LPDWH &RSSHU DQG =LQF &RQFHQWUDWLRQV LQ 6ROXWLRQV :KHUH &RSSHU 'HSRVLWV :HUH 2EVHUYHG RQ $OSKD %UDVV 9OO

PAGE 8

/,67 2) ),*85(6 )LJXUH 3DJH ,PPHUVLRQ WHVW FHOO &LUFXLW GLDJUDP RI HTXLSPHQW XVHG WR SRODUL]H VSHFLPHQ DQG DXWRPDWLFDOO\ UHFRUG FRUURVLRQ FXUUHQW GHQVLW\ DV D IXQFWLRQ RI SRWHQWLDO &RUURVLRQ FHOO XVHG LQ HOHFWURFKHPLFDO LQYHVWLJDWLRQV $VVHPEOHG DQG H[SORGHG YLHZV RI VDPSOH KROGHU &RSSHU]LQF SKDVH GLDJUDP &RSSHUJROG SKDVH GLDJUDP 7KH ODWWLFH SDUDPHWHU RI DOSKD EUDVV DV D IXQFWLRQ RI WKH DWRPLF SHUFHQW ]LQF +,f SHDNV RI D PL[WXUH RI EUDVV ILOLQJV DQG FRSSHU ILOLQJV +,f SHDNV IURP VDPSOH RI EUDVV GH]LQFLILHG IRU GD\V LQ 1 +& DW r& +,f SHDNV IURP VDPSOH RI EUDVV GH]LQFLILHG IRU GD\V LQ 1 +& DW r& 'LIIUDFWLRQ SDWWHUQ IURP VDPSOH RI EHWD EUDVV GH]LQFLILHG IRU WZR GD\V LQ 1 +& DW r& =LQF LQWHQVLW\ SURILOH IURP D VDPSOH RI DOSKD EUDVV GH]LQFLILHG IRU GD\V LQ ,1 1D&O =LQF LQWHQVLW\ SURILOH IURP D VDPSOH RI DOSKD EUDVV GH]LQFLILHG IRU GD\V LQ 1 +& DW r& YLLL

PAGE 9

/,67 2) ),*85(6 &RQWLQXHGf )LJXUH 3DJH =LQF LQWHQVLW\ SURILOH IURP D VDPSOH RI EHWD EUDVV GH]LQFLILHG IRU WZR GD\V LQ 1 +& DW r& 3KRWRPLFURJUDSK RI VDPSOH VKRZQ LQ )LJXUH 3KRWRPLFURJUDSK RI VDPSOH VKRZQ LQ )LJXUH 3KRWRPLFURJUDSK RI VDPSOH VKRZQ LQ )LJXUH 'H]LQFLILFDWLRQ SOXJ LQ DOSKD EUDVV H[SRVHG IRU GD\V LQ ,1 1D&O DW URRP WHPSHUDWXUH &RSSHU GHSRVLWV RQ WKH VXUIDFH RI GH]LQFLILHG DOSKD EUDVV VDPSOH 'HSRVLW RQ VXUIDFH RI GH]LQFLILHG DOSKD EUDVV VDPSOH 6FDQQLQJ HOHFWURQ PLFURJUDSK RI FRSSHU VODE SURWUXGLQJ IURP WKH VXUIDFH RI D GH]LQFLILHG DOSKD EUDVV VDPSOH 1RQGLVSHUVLYH [UD\ DQDO\]HU SDWWHUQ RI GHSRVLW VKRZQ LQ )LJXUH 'H]LQFLILHG FURVVVHFWLRQ RI VDPSOH VKRZQ LQ )LJXUH $WRPLFDEVRUSWLRQ FDOLEUDWLRQ FXUYH IRU FRSSHU &RSSHU DQG ]LQF GLVVROXWLRQ IURP DOSKD EUDVV LQ 1 +& DW r& &RSSHU DQG ]LQF GLVVROXWLRQ IURP DOSKD EUDVV LQ 1 +& DW r& 'H]LQFLILFDWLRQ IDFWRUV = IRU DOSKD EUDVV LQ 1 +& DW r& &RSSHU GLVVROXWLRQ IURP EHWD EUDVV LQ 1 +& DW r& L[

PAGE 10

/,67 2) ),*85(6 &RQWLQXHGf )LJXUH 3DJH =LQF GLVVROXWLRQ IURP EHWD EUDVV LQ 1 +& DW r& 'H]LQFLILFDWLRQ IDFWRUV = IRU EHWD EUDVV LQ 1 +& DW r& =LQF GLVVROXWLRQ IURP EHWD EUDVV LQ 1 +& DW YDULRXV WHPSHUDWXUHV $UUKHQLXV SORW RI ]LQF GLVVROXWLRQ UDWH YHUVXV 7 ([SHULPHQWDO SRWHQWLDO YHUVXV S+ GLDJUDP IRU &X =Q LQ 0 &O DW r& 6LPSOLILHG &8&,+2 GLDJUDP DW r& IRU VROXWLRQ FRQWDLQLQJ 0 FKORULGH LRQV 6LPSOLILHG =Q+ GLDJUDP IRU FRQFHQWUDWLRQV RI LRQLF VSHFLHV P &X =Q DOOR\ LQ 0 FKORULGH VROXWLRQ %HWD EUDVV KHOG DW 9VKH IrU KRXUV 7\SLFDO SRWHQWLRNLQHWLF VFDQ LQ DFLG VROXWLRQV &URVVVHFWLRQ RI DOSKD EUDVV VKRZLQJ ZKHUH KROLGD\V LQ WKH VWRSRII ODFTXHU FDXVHG GHn ]LQFLI LFDWLRQ LQ WKH VWDJQDQW UHJLRQV EHQHDWK WKH KROLGD\V 7KHRUHWLFDO GRPDLQV IRU GHDOOR\LQJ LQ D JLYHQ VROXWLRQ EDVHG XSRQ WKH 1HUQVW HTXDWLRQ [

PAGE 11

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f GLVVROXWLRQ RI ERWK DOOR\ FRQVWLWn XHQWV IROORZHG E\ UHGHSRVLWLRQ RI WKH PRUH QREOH VSHFLHV DQG f WKH VHOHFWLYH UHPRYDO RI WKH OHVV QREOH FRQVWLWXHQW ZHUH IRXQG WR EH RSHUDWLYH XQGHU FHUWDLQ FRQGLWLRQV RI SRn WHQWLDO DQG S+ IRU ERWK DOSKD DQG EHWD EUDVVHV $Q HOHFWURFKHPLFDO H[SODQDWLRQ RI WKH FLUFXPVWDQFHV XQGHU ZKLFK GHDOOR\LQJ FDQ EH H[SHFWHG WR RFFXU ZDV GHYHOn RSHG EDVHG RQ WKH XVH RI 3RXUEDL[ GLDJUDPV [L

PAGE 12

,1752'8&7,21 'HDOOR\LQJ LV D FRUURVLRQ SURFHVV ZKHUHE\ RQH FRQVWLWXn HQW RI DQ DOOR\ LV SUHIHUHQWLDOO\ UHPRYHG IURP WKH DOOR\ OHDYLQJ DQ DOWHUHG UHVLGXDO VWUXFWXUHA :KLOH GH]LQFLILFDWLRQ WKH ORVV RI ]LQF IURP EUDVVHV nnf LV WKH PRVW FRPPRQO\ H[SHULHQFHG IRUP RI GHDOOR\LQJ RWKHU H[DPSOHV KDYH EHHQ UHSRUWHG LQ SUDFWLFH 7KHVH LQn FOXGH ORVV RI QLFNHO DOXPLQXP A A DQG WLQAA!f eURP &2SSHU DOOR\V LURQ IURP FDVW LURQ QLFNHO IURP DOOR\ VWHHOVrAf DQ FREDOW IURP 6WHOOLWH 6LQFH WKH SKHQRPHQRQ ZDV ILUVW UHSRUWHG LQ WKH OLWHUDWXUH KDV EHHQ ILOOHG ZLWK UHSRUWV RI UHVHDUFK HIIRUWV DLPHG DW FODULI\LQJ WKH PHFKDQLVPV RI GHDOOR\LQJ 1RQHWKHn OHVV WKHUH VWLOO LV QR JHQHUDO DJUHHPHQW DV WR WKH GHWDLOHG PHFKDQLVP LQYROYHG 2QH JURXS FRQWHQGV WKDW WKH HQWLUH DOn OR\ LV GLVVROYHG DQG WKDW RQH RI LWV FRQVWLWXHQWV WKHQ LV U UHSODWHG IURP VROXWLRQY f $QRWKHU FRQWHQGV WKDW RQH VSHFLHV LV VHOHFWLYHO\ GLVVROYHG IURP WKH DOOR\ OHDY U LQJ D SRURXV UHVLGXH RI WKH PRUH QREOH VSHFLHVY f 6WLOO RWKHUV EHOLHYH WKDW ERWK PHFKDQLVPV WDNH SODFHA f $ QXPEHU RI OLWHUDWXUH VXUYH\V KDYH DSSHDUHGDQG WKHVH VXPPDUL]H WKH VLWXDWLRQ XS WR WKH WLPH WKH\ ZHUH ZULWn WHQ &!!f

PAGE 13

7KH VWXG\ RI GHDOOR\LQJ SKHQRPHQD LV IUDXJKW ZLWK D QXPEHU RI FRPSOLFDWLRQV *HQHUDOO\ VXFK UHDFWLRQV DUH UHODWLYHO\ VORZ DQG D OHQJWK\ H[SRVXUH SHULRG LV UHTXLUHG WR FDXVH D VXIILFLHQWO\ JUHDW DPRXQW RI GHDOOR\LQJ WR IDFLOn LWDWH HYDOXDWLRQ &RQVHTXHQWO\ WKHUH LV FRQVLGHUDEOH LQWHUn HVW LQ DFFHOHUDWHG WHVWV IRU HYDOXDWLRQ RI WHQGHQFLHV RI DOOR\V WR GHDOOR\ 0DQ\ WHFKQLTXHV KDYH EHHQ HPSOR\HG )RU H[DPSOH HOHFWURO\WH FRPSRVLWLRQV KDYH EHHQ DGMXVWHG E\ XVn LQJ PRUH FRQFHQWUDWHG VROXWLRQV RU VROXWLRQV KDYLQJ YDULD WLRQV LQ R[LGL]LQJ SRZHU -6SHFLILF LRQV KDYH EHHQ DGGHG WR VWLPXODWH GHDOOR\LQJ HJ VDWXUDWHG FXSURXV FKORULGH VROXWLRQV KDYH EHHQ XVHG WR DFFHOHUDWH WKH GH]LQFLILFDWLRQ RI FRSSHUEDVH DOOR\V (OHFWURFKHPLFDO VWLPXODWLRQ DOVR KDV EHHQ XVHG 8QIRUWXQDWHO\ DOO WRR RIWHQ WKH WHVW PHWKn RGV HPSOR\HG KDYH EHHQ YXOQHUDEOH WR FULWLFLVP DV KDYLQJ ELDVHG WKH H[SHULPHQWDO UHVXOW DQG DOWKRXJK VSHFLILF WHFKn QLTXHV DUH QRZ DYDLODEOH ZKLFK FDQ FDXVH GHDOOR\LQJ WR RFFXU LQ WKH ODERUDWRU\ QRQHWKHOHVV WKHUH VWLOO LV QR ILUP EDVLV IRU SUHGLFWLQJ WKH OLNHOLKRRG RI GHDOOR\LQJ LQ VHUYLFH EDVHG RQ WKHVH WHVWV 6LQJOHSKDVH DOSKD EUDVVHV DQG GXSOH[ DOSKDSOXVEHWD EUDVVHV DUH WKH RQO\ IRUPV RI FRSSHU]LQF DOOR\V KDYLQJ HQJLQHHULQJ VLJQLILFDQFH DW SUHVHQW %HQJRXJK DQG 0D\ UHn SRUWHG LQ WKDW VPDOO DGGLWLRQV RI DUVHQLF ZRXOG SUHn YHQW GH]LQFLILFDWLRQ RI DOSKD EUDVVHV EXW ZRXOG QRW SURWHFW I WKH EHWD SKDVHV RI GXSOH[ DOOR\Vn 7KH UHDVRQV IRU

PAGE 14

DUVHQLF SURWHFWLQJ DOSKD EUDVV EXW QRW WKH EHWD SKDVH RI GXSOH[ DOOR\V KDYH UHPDLQHG FRQWURYHUVLDOAA+RZHYHU WKH DGGLWLRQ RI VPDOO DPRXQWV RI DUVHQLF RU RI DQWLPRQ\ RU SKRVSKRUXV ZKLFK KDYH VLPLODU HIIHFWV KDV EHFRPH D VWDQ I GDUG PHDQV RI LQKLELWLQJ GH]LQFLILFDWLRQ LQ DOSKD EUDVVHVA 1R LQKLELWRU LV SUHVHQWO\ DYDLODEOH IRU GXSOH[ DOOR\V DOn WKRXJK WKH DGGLWLRQ RI WLQ UHWDUGV PRVW IRUPV RI EUDVV FRUn URVLRQ WR LQFOXGH GH]LQFLILFDWLRQA 7KH SXUSRVH RI WKLV LQYHVWLJDWLRQ KDV EHHQ WR HOXFLGDWH WKH PHFKDQLVP RI GH]LQFLILFDWLRQ RI DOSKD DQG RI EHWD EUDVVHV DQG WR GHYHORS D EDVLV IRU SUHGLFWLQJ WKH FRQGLWLRQV XQGHU ZKLFK GH]LQFLILFDWLRQ RI WKHVH DOOR\V PLJKW EH H[SHFWHG WR RFFXU 3DUWLFXODU HPSKDVLV ZDV SODFHG RQ VHOHFWLQJ H[SRVXUH FRQGLWLRQV DQG WHVW PHWKRGV ZKLFK ZRXOG QRW ELDV WKH H[SHULn PHQWDO UHVXOWV )RU H[DPSOH LW ZDV IHOW WKDW DFFHOHUDWHG WHVWV XVLQJ FRSSHUFKORULGH VROXWLRQV FRXOG QRW JLYH XQn ELDVHG HYLGHQFH IRU D GLVVROXWLRQUHGHSRVLWLRQ PHFKDQLVP DOWKRXJK WKLV PHWKRG RI DFFHOHUDWHG WHVWLQJ KDV EHHQ UHn SRUWHG IUHTXHQWO\ LQ WKH OLWHUDWXUH Af (OHFWURn FKHPLFDO VWLPXODWLRQ LQ ZKLFK WKH VSHFLPHQ ZDV D GULYHQ DQRGHf KDV DOVR EHHQ XVHG DV D PHDQV RI SURGXFLQJ GH]LQFLILn FDWLRQ 7KLV PHWKRG RI SURGXFLQJ DFFHOHUDWHG DWWDFN DOVR FDQ ELDV WKH H[SHULPHQWDO UHVXOWV E\ PDVNLQJ WKH SUHVn HQFH RI GLIIXVLRQ UHODWHG VHOHFWLYHUHPRYDO SURFHVVHV EHn FDXVH GLIIXVLRQ DV LW LV FRPPRQO\ XQGHUVWRRG LV D TXLWH

PAGE 15

VORZ SURFHVV DW WKH ORZ WHPSHUDWXUHV HQFRXQWHUHG LQ DTXHRXV HQYLURQPHQWV 0HFKDQLVP VWXGLHV ZHUH SHUIRUPHG RQ EUDVV LPPHUVLRQ VDPSOHV SODFHG LQ 1D&O DQG +& VROXWLRQV 1R FRSSHU VDOWV ZHUH DGGHG WR WKH WHVW HOHFWURO\WH H[FHSW DV WKH UHVXOW RI FRUURVLRQ RI WKH WHVW VSHFLPHQV 7HVW GXUDWLRQV RI XS WR GD\V ZHUH HPSOR\HG DW WHPSHUDWXUHV UDQJLQJ IURP URRP WHPn SHUDWXUH DSSUR[LPDWHO\ r&f WR r& 3RWHQWLRVWDWLF DQG SRWHQWLRNLQHWLF PHWKRGV ZHUH XVHG WR GHILQH FRQGLWLRQV XQGHU ZKLFK GHDOOR\LQJ PLJKW EH H[n SHFWHG WR RFFXU

PAGE 16

(;3(5,0(17$/ 352&('85( 7KH SRVVLELOLW\ RI ELDV RI UHVXOWV GXH WR WKH WHVWLQJ PHWKRG XVHG ZDV GLVFXVVHG LQ WKH LQWURGXFWLRQ WR WKLV GLVn VHUWDWLRQ ,PPHUVLRQ WHVWLQJ RI EUDVV VDPSOHV LQ HQYLURQn PHQWV NQRZQ WR SURGXFH GH]LQFLILFDWLRQ ZDV FKRVHQ DV WKH ODERUDWRU\ WHVW PHWKRG PRVW VXLWHG IRU D VWXG\ RI WKH PHFKD QLVP RI GHDOOR\LQJ (OHFWURFKHPLFDO WHVWV ZHUH HPSOR\HG LQ ODWHU VWDJHV RI WKLV LQYHVWLJDWLRQ DIWHU WKH PHFKDQLVP VWXGLHV ZHUH FRPn SOHWHG 7KHVH WHVWV ZHUH LQWHQGHG WR GHILQH WKH FRQGLWLRQV RI SRWHQWLDO DQG S+ XQGHU ZKLFK SDUWLFXODU PRGHV RI GHDOOR\ LQJ RI FRSSHU]LQF DOOR\V FRXOG EH H[SHFWHG WR RFFXU

PAGE 17

OLPQHUVLRQ 7HVW $SSDUDWXV )LJXUH LV D GLDJUDP RI WKH LPPHUVLRQ WHVW FHOO XVHG LQ WKLV LQYHVWLJDWLRQ ,W LV VLPLODU LQ PDQ\ UHVSHFWV WR WKDW UHFRPPHQGHG E\ 1DWLRQDO $VVRFLDWLRQ RI &RUURVLRQ (QJLn QHHUV 6WDQGDUG 70 /DERUDWRU\ &RUURVLRQ 7HVWLQJ RI 0HWDOV IRU WKH 3URFHVV ,QGXVWULHVAf 7KH FHOO LV FRQVWUXFWHG RI D 3\UH[r JODVV UHVLQ UHDFn WLRQ NHWWOH KHOG WRJHWKHU E\ DQ H[WHUQDO PHWDO FODPS $OO IL[WXUHV LQVHUWHG LQWR WKH FHOO DOVR DUH PDGH RI 3\UH[ ZLWK WKH H[FHSWLRQ RI WKH 7HIORQrrFRDWHG VWLUULQJ EDU 6DPSOHV DUH VXVSHQGHG IURP D VDPSOH KROGHU DV VKRZQ LQ WKH GLDJUDP ,Q WHVWV UHTXLULQJ DQ R[\JHQIUHH HQYLURQPHQW OLTXLG LQ WKH FHOO LV VSDUJHG ZLWK GULHG DQG SXULILHG DUJRQ ZKLFK HQWHUV WKH FHOO WKURXJK D IULWWHGJODVV GLIIXVHU LPn PHUVHG LQ WKH OLTXLG *DV LV SDVVHG RXW RI WKH V\VWHP E\ ZD\ RI D /LHELJW\SH UHIOX[ FRQGHQVHU ZKLFK LV EDFNHG XS E\ D OLTXLG WUDS WR SUHYHQW VROXWLRQ ORVV RU FRQWDPLQDWLRQ 7KH YROXPH RI VROXWLRQ ZLWKLQ WKH FHOO ZDV PHDVXUHG DW WKH VWDUW DQG ILQLVK RI HDFK WHVW /RVVHV GLG QRW H[FHHG PO IURP DQ LQLWLDO YROXPH RI PO LQ WHVWV RI XS WR GD\V 7KLV FRUUHVSRQGV WR D PD[LPXP FKDQJH LQ VROXWLRQ YROXPH RI SHUFHQW 7KH QRUPDOLW\ RI DFLGV EHLQJ XVHG ZDV DOVR FKHFNHG r5HJLVWHUHG WUDGHPDUN &RUQLQJ *ODVV &RPSDQ\ rr5HJLVWHUHG WUDGHPDUN ( GX3RQW GH 1HPRXUV DQG &RPSDQ\ ,QF

PAGE 18

&4 2 $ )LJXUH ,PPHUVLRQ WHVW FHOO $ VDPSOH KROGHU % VDPSOH & LPPHUVLRQ KHDWHU VWLUn ULQJ EDU ( -WXEH WKHUPRVWDW ) JDV LQOHW GLIIXVHU DQG JDV RXWOHW WKURXJK UHIOX[ FRQGHQVHU

PAGE 19

EHIRUH DQG DIWHU HDFK H[SRVXUH E\ WLWUDWLRQ ZLWK D ,1 1D2+ VWDQGDUG ,W ZDV IRXQG WR YDU\ E\ QR PRUH WKDQ s QRUn PDOLW\ XQLWV IURP VWDUW WR ILQLVK RI DQ\ WHVW 7KH OLTXLG WHPSHUDWXUH ZDV FRQWUROOHG WR ZLWKLQ sr& E\ PHDQV RI DQ HOHFWULF WKHUPRUHJXODWRU FRXSOHG WR D ZDWW LPPHUVLRQ KHDWHU ,PPHUVLRQ WHVWV ZHUH FRQGXFWHG LQ ,1 1D&O DQG 1 +& 7KHVH VROXWLRQV ZHUH FKRVHQ EHFDXVH WKH\ DUH QRQR[LGL]LQJ DQG GR QRW IRUP VROXEOH PHWDO FRPSOH[HV ZLWK WKH H[FHSWLRQ RI &X&A %HFDXVH RI WKH KLVWRULF VLJQLILFDQFH RI GH]LQF LILFDWLRQ IDLOXUHV LQ VDOW ZDWHU DQG RWKHU FKORULGH HQYLURQn PHQWV LW ZDV IHOW WKDW FKORULGHV VKRXOG EH XVHG GHVSLWH WKH H[LVWHQFH RI WKLV RQH FRPSOH[ PHWDO LRQ 7KH KLJK DFLG FRQFHQWUDWLRQ ZDV FKRVHQ WR PLQLPL]H DFLGLW\ FKDQJHV ZLWKLQ WKH VROXWLRQ 7KLV KLJK DFLG FRQFHQWUDWLRQ ZDV DOVR YHU\ FORVH WR WKH PD[LPXP FRQFHQWUDWLRQ WKDW FDQ EH PDLQWDLQHG DW r& ZLWKRXW ERLOLQJAf 3UHYLRXV LPPHUVLRQ WHVWV LQ +& DQG 1D&O KDYH DSSHDUHG f f P WKH OLWHUDWXUH }!}!}}} DQG FRPSDULVRQV ZLWK SUHYLRXV UHVXOWV DUH FRQWDLQHG LQ WKH GLVFXVVLRQ VHFWLRQ RI WKLV ZRUN

PAGE 20

(OHFWURFKHPLFDO 7HVW $SSDUDWXV 7KH HOHFWURFKHPLFDO K\VWHUHVLV WHFKQLTXH ZDV XVHG IRU H[SHULPHQWDO GHWHUPLQDWLRQ RI WKH FRUURVLRQ EHKDYLRU RI DOn OR\V DV D IXQFWLRQ RI SRWHQWLDO DQG S+ 7KH HTXLSPHQW WHFKQLTXHV DQG WKHRULHV XSRQ ZKLFK LQYHVWLJDWLRQV RI WKLV W\SH DUH EDVHG KDYH EHHQ GLVFXVVHG DW VRPH OHQJWK LQ RWKHU f UHSRUWVY 7KH HTXLSPHQW LV GHVFULEHG VFKHPDWLFDOO\ LQ )LJXUH DQG OLVWHG LQ $SSHQGL[ 7KH VFDQQLQJ SRWHQWLRVWDW DSSOLHV D FRQWLQXRXVO\ YDU\LQJ SRWHQWLDO WR WKH VDPSOH RYHU D FHUn WDLQ UDQJH 7KH GLIIHUHQWLDO DPSOLILHU LVRODWHV WKH FRUURn VLRQ FHOO IURP WKH UHFRUGLQJ HTXLSPHQW DQG HOLPLQDWHV JURXQG ORRSV 7KH ORJ FRQYHUWHU DOORZV WKH FRUURVLRQ FXUUHQW WR EH SORWWHG DV D ORJDULWKP RQ WKH [ D[LV RI WKH [\ UHFRUGHU ZKLOH WKH VDPSOH SRWHQWLDO LV SORWWHG OLQHDUO\ RQ WKH \ D[LV $Q DXWRPDWLF VWHS VZLWFKLQJ DSSDUDWXVn H[WHQGV WKH UDQJH RI WKH ORJ FRQYHUWHU DQG DOORZV WKH FRQWLQXRXV UHFRUGLQJ RI FRUURVLRQ FXUUHQW GHQVLWLHV IURP [ WR DPSVFP 7KH ORZSDVV 5& ILOWHU FRQVLVWLQJ RI D PLFURIDUDG FDSDFLWRU LQ SDUDOOHO ZLWK D .RKP SRWHQWLRPHWHU DGMXVWHG WR .RKPV UHGXFHV HOHFWULFDO QRLVH LQ WKH V\VWHP WR QHJn OLJLEOH YDOXHV $ VFKHPDWLF GLDJUDP RI WKH FRUURVLRQ FHOO LV VHHQ LQ )LJXUH 7KH EXIIHUHG HOHFWURO\WH LV YDFXXP GHDHUDWHG SULRU WR WUDQVIHUHQFH WR WKH FRUURVLRQ FHOO 7KH VROXWLRQ LV FRQWLQXRXVO\ SXUJHG ZLWK K\GURJHQ GXULQJ WKH UXQ YLD WKH

PAGE 21

)LJXUH &LUFXLW GLDJUDP RI HTXLSPHQW XVHG WR SRODUL]H VSHFLPHQ DQG DXWRPDWLFDOO\ UHFRUG FRUURVLRQ FXUUHQW GHQVLW\ DV D IXQFWLRQ RI SRWHQWLDO

PAGE 22

&25526,21 &(// 6&+(0$7,& 0$*1(7,& 67,55(5 )LJXUH &RUURVLRQ FHOO XVHG LQ HOHFWURFKHPLFDO LQYHVWLJDWLRQV

PAGE 23

JDV GLIIXVHU DV VXJJHVWHG E\ $670 FRPPLWWHH *O $ EULJKW SODWLQXP VFUHHQ VHUYHV DV WKH DX[LOLDU\ HOHFWURGH 7KH FXUUHQW IURP WKH SRWHQWLRVWDW WR WKH DX[LOLDU\ HOHFWURGH LV PHDVXUHG DV D SRWHQWLDO DFURVV D SUHFLVLRQ UHVLVWRU VHn OHFWHG WR SURYLGH WKH UHTXLUHG ORJDULWKPLF FRQYHUWHU LQSXW YROWDJH $ /XJJLQ+DEHU SUREH FRQQHFWHG WR D VWDQGDUG FDORn PHO HOHFWURGH LV XVHG WR PHDVXUH WKH VDPSOH SRWHQWLDO 7KH WKHUPRPHWHU LV XVHG WR LQGLFDWH WKH WHPSHUDWXUH RI WKH HOHFn WURO\WH 7KH VROXWLRQ LV VWLUUHG XVLQJ D PDJQHWLF ZDWHU SRZHUHG VWLUUHU 7KH FHOO LWVHOI LV PDGH RI 3\UH[ JODVV ZLWK D 7HIORQ OLG EROWHG WR D SRO\FDUERQDWH 9DQ6WRQH EDFNn LQJ ULQJ ZKLFK PDNHV WKH V\VWHP DLUWLJKW 7KH VDPSOHKROGHU LV VKRZQ LQ )LJXUH 7KH PDLQ ERG\ RI WKH KROGHU LV FRQVWUXFWHG RI 7HIORQ VR WKDW LW ZLOO QRW UHDFW ZLWK WKH WHVW VROXWLRQ &RSSHU SDUWV HQFDVHG LQ WKH 7HIORQ DOORZ HOHFWULFDO FRQWDFW ZLWK WKH VDPSOH $ SRO\ FDUERQDWH QXW IDVWHQV WKH VDPSOH LQ SODFH DQG DOORZV FP WR EH H[SRVHG WR WKH HOHFWURO\WH 7KH 7HIORQ JDVNHW DYRLGV OHDNDJH DQG PLQLPL]HV FUHYLFH HIIHFWV &DUH PXVW EH H[HUFLVHG ZKHQ FKRRVLQJ EXIIHUHG HOHFWURn O\WHV WR LQVXUH WKDW VROXWLRQ LRQV ZLOO QRW IRUP FRPSOH[HV ZLWK WKH PHWDO LRQV IURP VDPSOH GLVVROXWLRQ 7KH HOHFWURn O\WHV XVHG LQ WKHVH VWXGLHV DUH OLVWHG LQ $SSHQGL[ $OO VROXWLRQV XVHG LQ WKLV SRUWLRQ RI WKH LQYHVWLJDWLRQ KDG D 0 FKORULGH FRQWHQW 7KH HIIHFWV RI FRSSHUFKORULGH FRPn SOH[HV ZHUH GLVFXVVHG LQ WKH VHFWLRQ RQ LPPHUVLRQ WHVWLQJ

PAGE 24

)LJXUH $VVHPEOHG DQG H[SORGHG YLHZV RI VDPSOH KROGHU

PAGE 25

7KH HIIHFWV RI VFDQ UDWH ZHUH LQYHVWLJDWHG 5HLQRHKO DQG FRZRUNHUVA f VKRZ WKDW SDVVLYDWLRQ FXUUHQW GHQVLW\ LV D IXQFWLRQ RI VFDQ UDWH EXW WKDW FULWLFDO SDVVLYDWLRQ SRWHQn WLDOV DQG UXSWXUH SRWHQWLDOV DUH LQGHSHQGHQW RI VFDQ UDWH DW OHDVW IRU LURQ LQ ,1 OA62A $ PRUH GHWDLOHG GLVFXVVLRQ RI WKH LPSRUWDQFH RI VFDQ UDWHV LV FRQWDLQHG LQ WKH GLVFXVn VLRQ VHFWLRQ $ VFDQ UDWH RI PYPLQ ZDV FKRVHQ IRU WKLV LQYHVWLJDWLRQ DIWHU H[DPLQLQJ VFDQ UDWHV IURP PYPLQ WR PYPLQ7KH VHOHFWHG UDWH ZDV VORZ HQRXJK WR DYRLG ORVLQJ DQ\ FRQILJXUDWLRQ RI WKH SRODUL]DWLRQ FXUYH \HW IDVW HQRXJK WR PLQLPL]H PDVNLQJ HIIHFWV GXH WR GH]LQFLILFDWLRQ RI WKH VDPSOH $IWHU H[SHULPHQWDO 3RXUEDL[ GLDJUDPV ZHUH HVWDEOLVKHG E\ SRWHQWLRNLQHWLF SRODUL]DWLRQ WHFKQLTXHV SRWHQWLRVWDWLF WHVWV ZHUH UXQ WR YHULI\ WKH SRWHQWLDO UHJLRQV ZKHUH GH ]LQFLILFDWLRQ RFFXUV 7KH VL[FHOO SRWHQWLRVWDW GHVFULEHG E\ &XVDPDQRAAA ZDV XVHG IRU WKHVH SRWHQWLRVWDWLF WHVWV

PAGE 26

6DPSOH 3UHSDUDWLRQ 7KH DOSKD EUDVV VDPSOHV XVHG LQ WKLV LQYHVWLJDWLRQ ZHUH GRQDWHG E\ WKH &KDVH %UDVV DULG &RSSHU &RPSDQ\ 7KH EUDVV ZDV REWDLQHG LQ WZR SRXQG ORWV RI DVH[WUXGHG FDUWULGJH EUDVV ZLUH KDYLQJ GLDPHWHUV RI LQ DQG LQ UHn VSHFWLYHO\ 7KH FKHPLFDO DQDO\VHV RI WKHVH DOSKD EUDVV VDPn SOHV DUH VKRZQ LQ $SSHQGL[ %HWD EUDVV LQJRWV KDYLQJ D QRPLQDO FRPSRVLWLRQ RI ZR &X DQG ZR =Q ZHUH SUHSDUHG IURP r SXUH FRSSHU DQG ]LQF VWRFN SXUFKDVHG IURP WKH $PHULFDQ 6PHOWLQJ DQG 5Hn ILQLQJ &RPSDQ\ 7KH LQJRW PHOWLQJ SURFHGXUHV DUH GHVFULEHG LQ GHWDLO LQ $SSHQGL[ 7KH WRSV DQG ERWWRPV RI HDFK LQJRW ZHUH DQDO\]HG DFFRUGLQJ WR WKH SURFHGXUH GHVFULEHG LQ $SSHQn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

PAGE 27

ZHUH DOVR XVHG IRU WKH HOHFWURQ PLFURSUREH DQG RSWLFDO LQn YHVW JDW LRQV )RU WKH DWRPLF DEVRUSWLRQ H[SHULPHQWV ZKHUH PHWDOLRQ GLVVROXWLRQ ZDV WR EH PRQLWRUHG F\OLQGULFDO VSHFLPHQV ZHUH PRXQWHG LQ GHQWDO PRXQW VR DV WR OHDYH H[SRVHG D FLUFXODU FURVVVHFWLRQ RI NQRZQ JHRPHWULF DUHD 7KH VSHFLPHQV ZHUH WKHQ SROLVKHG DFFRUGLQJ WR WKH SURFHGXUH GHVFULEHG LQ $SSHQn GL[ DQG VXVSHQGHG LQ WKH UHDFWLRQ NHWWOHV E\ PHDQV RI D KROH GULOOHG LQ WKH GHQWDO PRXQW 7KLV SURFHGXUH LV VLPLODU WR WKDW GHVFULEHG E\ )LVKHU DQG +DOSHULQ IRU WKHLU FDORULP  HWU\ VWXGLHV RQ WKH GHDOOR\LQJ RI FRSSHUJROG DOOR\V 'LVF VSHFLPHQV SUHSDUHG DFFRUGLQJ WR WKH SURFHGXUH GHn VFULEHG LQ $SSHQGL[ ZHUH XVHG IRU WKH SRWHQWLRNLQHWLF H[n SHULPHQWV XVLQJ D VDPSOH KROGHU VKRZQ LQ )LJXUH )RU WHVWLQJ GHWDLOV VHH WKH VHFWLRQ RQ HOHFWURFKHPLFDO WHVW DSSDUDWXV (DUO\ KLJKWHPSHUDWXUH WHVWV LQ WKLV LQYHVWLJDWLRQ XVHG VDPSOHV LQ ZKLFK DOO EXW D FP [ FP VTXDUH VHFWLRQ RI WKH VDPSOH ZDV VWRSSHG RII XVLQJ 0LFFURVKLHOGr VWRSRII ODFTXHU ([DPLQDWLRQ RI WKH VDPSOHV DIWHU H[SRVXUH UHYHDOHG WKDW OHDNV WKURXJK WKH ODFTXHU ZHUH SUHVHQW DQG WKLV W\SH RI VDPSOH ZDV DEDQGRQHG 7KH VLJQLILFDQFH RI WKHVH OHDNV LV GLVFXVVHG LQ WKH GLVFXVVLRQ VHFWLRQ RI WKLV UHSRUW 3RWHQWLRVWDWLF WHVWV DW URRP WHPSHUDWXUH ZHUH FRQGXFWHG XVLQJ GLVFV ZLWK DQ LQVXODWHG FRSSHU ZLUH VROGHUHG WR WKH r7UDGH QDPH 0LFKLJDQ &KURPH DQG &KHPLFDO &RPSDQ\

PAGE 28

EDFN 0LFFURVWRS VWRSRII ODFTXHU ZDV XVHG WR FRDW WKH EDFN DQG VLGHV RI WKHVH VDPSOHV OHDYLQJ D FLUFXODU H[SRVXUH VXUIDFH ([DPLQDWLRQ RI WKH VDPSOHV DIWHU H[SRVXUH UHYHDOHG QR DSSDUHQW OHDNV LQ WKH ODFTXHU

PAGE 29

;5$< ',))5$&7,21 ;UD\ GLIIUDFWLRQ DQG HOHFWURQ GLIIUDFWLRQ KDYH EHHQ XVHG E\ VHYHUDO UHVHDUFKHUV LQ DQ DWWHPSW WR DVFHUWDLQ ZKHWKHU GHDOOR\LQJ LV D SURFHVV LQYROYLQJ WKH VHOHFWLYH UHn PRYDO RI RQH FRQVWLWXHQW RU D GLVVROXWLRQUHGHSRVLWLRQ SURn FHVV fnf ,I D UHGHSRVLWLRQ SURFHVV RFFXUV WKH GLIIUDFWLRQ SDWWHUQ VKRXOG VKRZ GLIIUDFWHG UDGLDWLRQ GXH WR UHPQDQWV RI WKH RULJLQDO DOOR\ DQG DOVR GXH WR UHn GHSRVLWHG PHWDO 'LIIUDFWHG UDGLDWLRQ EHWZHHQ WKH WZR SHDNV ZRXOG VXSSRVHGO\ LQGLFDWH DQ DOOR\ KDYLQJ D FRPSRVLWLRQ EHn WZHHQ WKDW RI WKH RULJLQDO DOOR\ DQG WKDW RI WKH ILQDO GH DOOR\HG UHODWLYHO\ SXUH PHWDO 7KLV LQWHUPHGLDWH FRPSRVLn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

PAGE 30

7(03(5$785( r& )LJXUH &RSSHU]LQF SKDVH GLDJUDP 0HWDOV +DQGERRN HGLWLRQ S f

PAGE 31

)LJXUH &RSSHUJROG SKDVH GLDJUDP 0HWDOV +DQGERRN HGLWLRQ S f

PAGE 32

UDQJH 7KH ODWWLFH SDUDPHWHU RI DOSKD EUDVV DV D IXQFn WLRQ RI ]LQF FRQWHQW LV VKRZQ LQ )LJXUH 3LFNHULQJ SRLQWV RXW WKDW E\ FKRRVLQJ DQ DOOR\ V\VWHP VXFK DV FRSSHUJROG ZLWK HOHPHQWV KDYLQJ VXEVWDQWLDO GLIn IHUHQFHV LQ DWRPLF GLDPHWHU WKH ODWWLFH SDUDPHWHU DQG WKXV WKH VHSDUDWLRQ EHWZHHQ SHDNV RI WKH RULJLQDO DOOR\ DQG WKRVH I RI WKH FRUURGHG UHVLGXH ZLOO EH LQFUHDVHG $Q DOWHUQDWLYH WR WKLV ZRXOG EH WR VWDUW ZLWK DQ DOOR\ ZKLFK XSRQ GHDOOR\LQJ ZRXOG XQGHUJR D SKDVH FKDQJH DV ZHOO DV D FKDQJH LQ ODWWLFH SDUDPHWHU 2I WKH ODUJH QXPEHU RI UHSRUWV FRQFHUQLQJ GLIIUDFWLRQ VWXGLHV RI GHDOOR\LQJ RQO\ WZR SUHYLRXV UHVHDUFKHUV KDYH FRQFOXGHG WKDW WKHLU GDWD VXSSRUWHG D VHOHFWLYHUHPRYDO f PHFKDQLVP 3LFNHULQJ VXEMHFWHG FRSSHUJROG DOOR\V WR HOHFWURO\WLF GLVVROXWLRQ DQG VKRZHG WKH DSSHDUDQFH RI DQ LQWHUPHGLDWH SHDN ZKLFK DV DGGLWLRQDO FXUUHQW ZDV SDVVHG LQFUHDVHG LQ LQWHQVLW\ DQG PRYHG FORVHU WR WKH SRVLWLRQ WR EH H[SHFWHG IRU SXUH JROG 7KH DOOR\ SHDN GHFUHDVHG DSSURSULDWHO\ LQ LQWHQVLW\ $ ODWHU LQYHVWLJDWLRQ E\ WKH VDPH DXWKRU UHSRUWHG WKH IRUPDWLRQ RI QHZ LQWHUPHGLDWH SKDVHV GXULQJ WKH DQRGLF GLVVROXWLRQ RI JDPPD EUDVV DQG RI HSVLORQ EUDVVAf 7KLV UHSRUW VXEVWDQWLDWHG WKH UHVXOWV RI 6WLOOZHOO DQG 7XUQLSVHHG ZKR VXEMHFWHG HSVLORQ EUDVV WR YDULRXV FRUURVLYH PHGLD DQG ZKRVH [UD\ GLIIUDFWLRQ UHVXOWV LQGLFDWHG WKH SUHVHQFH RI I nf LQWHUPHGLDWH SKDVHV LQ VRPH RI WKHLU H[SHULPHQWVA `

PAGE 33

/$77,&( 3$5$0(7(5 $ )LJXUH 7KH ODWWLFH SDUDPHWHU RI DOSKD EUDVV DV D IXQFWLRQ RI WKH DWRPLF SHUFHQW ]LQF f

PAGE 34

7KH ODFN RI VLPLODU UHVXOWV LQ RWKHU VWXGLHV RI GHn DOOR\LQJ PD\ EH GXH WR WKH XVH RI FRQYHQWLRQDO ILOP PHWKRGV I IRU UHFRUGLQJ GLIIUDFWHG LQWHQVLW\A -

PAGE 35

3UHVHQW :RUN $OO [UD\ GLIIUDFWLRQ PHFKDQLVP VWXGLHV SHUIRUPHG GXUn LQJ WKLV LQYHVWLJDWLRQ ZHUH SHUIRUPHG RQ SRZGHU VDPSOHV REn WDLQHG IURP GLVFV RI DOSKD EUDVV DQG RI EHWD EUDVV H[SRVHG WR K\GURFKORULF DFLG LQ WKH LPPHUVLRQ DSSDUDWXV VKRZQ LQ )LJXUH DQG GHVFULEHG LQ WKH VHFWLRQ RQ H[SHULPHQWDO SURn FHGXUH 7KHUH DUH VHYHUDO SUREOHPV DVVRFLDWHG ZLWK WKH XVH RI D GLIIUDFWRPHWHU IRU LQYHVWLJDWLQJ GH]LQFLILFDWLRQ RI DOSKD EUDVVHV ,I WKH SHDNV GXH WR WKH RULJLQDO DOOR\ RFFXU WRR FORVH WR WKH FRSSHU SHDNV WKHQ WDLOV RI GLIIUDFWLRQ SHDNV PD\ RYHUODS DQG EH PLVLQWHUSUHWHG DV EHLQJ GXH WR DQ DOOR\ RI LQWHUPHGLDWH FRPSRVLWLRQ 7KLV LV WUXH RI WKH GLIIUDFWLRQ SDWWHUQV REWDLQHG XVLQJ &X .D UDGLDWLRQ 7KH VHSDUDWLRQ LV OHVV WKDQ RQH GHJUHH EHWZHHQ WKH f SHDNV RI SXUH FRSSHU DQG RI DOSKD EUDVV $W KLJKHU DQJOHV WKH VHSDUDWLRQ EHWZHHQ SHDNV EHn FRPHV JUHDWHU EXW LQWHUPHGLDWH LQWHQVLW\ LI SUHVHQW ZRXOG EH VSUHDG RXW DOVR DQG WKXV EH KDUGHU WR GHWHFW )LJXUH VKRZV WKH f SHDNV IURP D VDPSOH REWDLQHG E\ PL[LQJ ILOLQJV RI DOSKD EUDVV ZLWK DQQHDOHG ILOLQJV RI FRSSHU $ VOLJKW LQFUHDVH RI LQWHQVLW\ GXH WR WKH RYHUODS RI WKH WDLOV RI WKH SHDNV LV DSSDUHQW )LJXUHV DQG VKRZ WKH f SHDNV REWDLQHG IURP VDPSOHV RI DOSKD EUDVV GH]LQFLILHG IRU DQG GD\V UHn VSHFWLYHO\ LQ 1 +& DW r& 7KH LQWHQVLW\ EHWZHHQ WKH

PAGE 36

1L /Q )LJXUH ,OOf SHDNV RI D PL[WXUH RI EUDVV ILOLQJV DQG FRSSHU ILOLQJV %UDVV SHDN LV WR WKH OHIW ,

PAGE 37

'(*5((6 WVf 2n )LJXUH ,OOf SHDNV IURP VDPSOH RI EUDVV GH]LQFLILHG IRU GD\V LQ 1 +& DW r&

PAGE 38

'(*5((6 K2 A )LJXUH ,OOf SHDNV IURP VDPSOH RI EUDVV GH]LQFLILHG IRU GD\V LQ 1 0&, DW r&

PAGE 39

WZR SHDNV RQ WKH FRUURGHG VDPSOHV LV KLJKHU WKDQ ZRXOG EH REWDLQHG GXH WR RYHUODS RI WKH WDLOV RI WKH FRSSHU DQG DOSKD EUDVV SHDNV +RZHYHU WKLV LQWHQVLW\ ZRXOG SUREDEO\ EH XQGHWHFWHG RQ D VWDQGDUG SRZGHU SDWWHUQ REWDLQHG E\ ILOP PHWKRGV ,I GH]LQFLILFDWLRQ RI EHWD EUDVV RFFXUV E\ D YROXPH GLIIXVLRQ PHFKDQLVP WKH GLIIUDFWLRQ SDWWHUQ REWDLQHG IURP D SRZGHU RI GH]LQFLILHG EHWD EUDVV FRXOG EH H[SHFWHG WR VKRZ VFDWWHUHG LQWHQVLW\ FKDUDFWHULVWLF RI )&& DOSKD EUDVV KDYn LQJ D ODWWLFH SDUDPHWHU GLIIHUHQW IURP WKDW RI )&& FRSSHU )LJXUH VKRZV D GLIIUDFWLRQ SDWWHUQ REWDLQHG IURP D VDPSOH RI EHWD EUDVV GH]LQFLILHG IRU WZR GD\V LQ 1 +& DW r& 7KH SHDN DW GHJUHHV LV GXH WR D VXSHUSRVLWLRQ RI EHWD EUDVV f VFDWWHULQJ DQG FRSSHU f VFDWWHULQJ 7KH EURDG SHDN DW GHJUHHV LV FKDUDFWHULVWLF RI DOSKD EUDVV KDYLQJ DSSUR[LPDWHO\ ZR ]LQF 7KH DERYHPHQWLRQHG ILJXUHV WKXV JLYH [UD\ GLIIUDFWLRQ HYLGHQFH ZKLFK VXSSRUWV D VHOHFWLYHUHPRYDO PHFKDQLVP IRU WKH ORVV RI ]LQF IURP ERWK DOSKD DQG EHWD EUDVVHV ZKHQ H[n SRVHG WR 1 +& DW r& 1R XQLGHQWLILHG SHDNV ZHUH GHWHFWHG LQ WKH SRZGHU SDWn WHUQV RI DQ\ RI WKH VDPSOHV VKRZQ DERYH WKXV WKH SRVVLn ELOLW\ RI WKLV VFDWWHULQJ EHLQJ GXH WR DQ H[WUDQHRXV VRXUFH LV HOLPLQDWHG

PAGE 40

'(*5((6 )LJXUH 'LIIUDFWLRQ SDWWHUQ IURP VDPSOH EUDVV GH]LQFLILHG IRU WZR GD\V DW r& RI EHWD LQ 1 +&

PAGE 41

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

PAGE 42

7KH XVH RI FKDUW UHFRUGHUV WR VKRZ GLIIXVLRQ JUDGLHQWV FDQ DOVR EH PLVOHDGLQJ %RWK WKH UHVSRQVH WLPH RI WKH FKDUW UHFRUGHU DQG WKH VFDQ UDWH DW ZKLFK WKH VDPSOH LV VZHSW XQGHU WKH HOHFWURQ EHDP FDQ DIIHFW WKH DSSDUHQW IDOORII GLVWDQFH WKDW LV UHFRUGHG RQ WKH FKDUW 6XJDZDUD DQG (ELNRAAf DQG 3LFNHULQJ KDYH SXEn OLVKHG PLFURSUREH GDWD REWDLQHG ZLWK FKDUW UHFRUGHUV WR VKRZ UHODWLYH ]LQF FRQFHQWUDWLRQV DV D IXQFWLRQ RI GLVWDQFH 6XJDZDUD DQG (ELNR FRQFOXGHG WKDW WKHLU GDWD JDYH QR LQGLn FDWLRQ RI D FRQFHQWUDWLRQ JUDGLHQW DORQJ WKH LQWHUIDFH EHn WZHHQ FRUURGHG DQG XQFRUURGHG EUDVV 3LFNHULQJ UHSRUWV D UHJLRQ DSSUR[LPDWHO\ \ WKLFN ZKHUH WKH FRSSHU FRQFHQWUDn WLRQ IHOO RII LQ D FRSSHUJROG DOOR\ ZKLFK KDG EHHQ VXEMHFWHG WR HOHFWURFKHPLFDO DQRGLF GLVVROXWLRQ 2WKHU UHVHDUFKHUV KDYH XVHG [UD\ LPDJHV WR GHPRQVWUDWH WKH RFFXUUHQFH RI GHDOOR\LQJ1RQH RI WKHVH DXWKRUV XVHG PLFURSUREH GDWD WR VXSSRUW DUJXPHQWV UHJDUGLQJ WKH PHFKDQLVPV LQYROYHG

PAGE 43

3UHVHQW :RUN $V WKH DERYH GLVFXVVLRQ VKRZV LW ZRXOG EH H[WUHPHO\ GLIILFXOW LI QRW LPSRVVLEOH WR REWDLQ PHDQLQJIXO TXDQWLn WDWLYH DQDO\VHV WKURXJK HOHFWURQ PLFURSUREH DQDO\VLV RI D UHJLRQ KDYLQJ D ODUJH DQG FKDQJLQJ QXPEHU RI YDFDQFLHV 7KLV SUREOHP KDV EHHQ GLVFXVVHG E\ $QXVDYLFHAA ZKR EHn FDXVH RI KLV LQWHUHVW LQ HOLPLQDWLQJ YDFDQF\ DQG SRUH IRUPDn WLRQ LQ KLJKWHPSHUDWXUH GLIIXVLRQ FRXSOHV ZDV DEOH WR DSSO\ SUHVVXUH DQG VLJQLILFDQWO\ UHGXFH SRURVLW\ IRUPDWLRQ 7KH WHFKQLTXH WKDW ZDV GHFLGHG XSRQ LQ WKLV VWXG\ ZDV WR XVH WKH 8QLYHUVLW\ RI )ORULGDnV $FWRQ (OHFWURQ 0LFURn SUREH RSHUDWHG XQGHU FRQGLWLRQV ZKLFK ZRXOG SURGXFH D PLQLn PXP EHDP VL]H WR REWDLQ SRLQW FRXQWV DFURVV WKH UHJLRQ ZKHUH ]LQF FRQFHQWUDWLRQ ZDV IRXQG WR FKDQJH 7KHVH UHJLRQV ZHUH LGHQWLILHG HLWKHU E\ XVLQJ D FKDUW UHFRUGHU RU DV H[n SHULHQFH ZDV REWDLQHG LQ LGHQWLI\LQJ SUHFLVH UHJLRQV RI LQn WHUHVW E\ XVLQJ WKH OLJKWRSWLFV PLFURVFRSH DWWDFKPHQW RQ WKH PLFURSUREH I 7KH SXEOLVKHG GDWD RI 3LFNHULQJA r DQG RI )RUW\ DQG +XPEOH LQGLFDWHG WKDW D GLIIXVLRQ ]RQH LI LW ZHUH WR EH IRXQG ZRXOG EH UDWKHU QDUURZ RI WKH RUGHU RI O2\ RU HYHQ OHVV 5HVXOWV LQ WKHVH ODERUDWRULHV FRQILUPHG WKHVH REVHUYDn WLRQV DQG SHUKDSV H[SODLQ ZK\ 6XJDZDUD DQG (ELNRAAf ZKR XVHG D FKDUW UHFRUGHU DQG D UHODWLYHO\ IDVW VFDQ UDWH ZHUH XQDEOH WR ILQG D GLIIXVLRQ ]RQH

PAGE 44

LQ WKH SUHVHQW VWXG\ SRLQW FRXQWV ZHUH WDNHQ DW LQWHUn YDOV DV VPDOO DV RQH PLFURQ DFURVV WKH UHJLRQ RI LQWHUHVW ,W LV UHFRJQL]HG WKDW PDNLQJ SRLQW FRXQWV DW QDUURZ LQWHUn YDOV PLJKW LQYROYH VRPH VOLJKW RYHUODS RI WKH LUUDGLDWHG YROXPHV EHWZHHQ WZR DGMDFHQW SRLQWV +RZHYHU WKH UHVXOWV REWDLQHG E\ YDU\LQJ RSHUDWLQJ FRQGLWLRQV VF WKDW WKH EHDP VL]H ZRXOG EH UHGXFHG WR D EDUH PLQLPXP GLG QRW JLYH D FKDQJH LQ WKH UHVXOWV REWDLQHG DQG LW LV IHOW WKDW WKH FRXQWV REn WDLQHG LQ WKLV PDQQHU DUH RI SK\VLFDO VLJQLILFDQFH 5HVXOWV ZHUH WDEXODWHG DV LQWHQVLW\ UDWLRV E\ WDNLQJ WKH DYHUDJH RI WKUHH VXFFHVVLYH FRXQWV RQ WKH VDPH SRLQW VXEWUDFWLQJ WKH DYHUDJH EDFNJURXQG LQWHQVLW\ RQ WKDW VDPSOH DQG GLYLGLQJ E\ WKH DYHUDJH SRLQW FRXQWV RQ DQ HOHPHQWDO VWDQGDUG PLQXV WKH DYHUDJH EDFNJURXQG LQWHQVLW\ RQ WKH VWDQn GDUG 7KLV FDQ EH H[SUHVVHG DV ,A6DPSOHf ,A6DPSOH %*f rSr $ a 7aLRT$6WDQGDUGn@ f ALTR ? 6WDQGDUG %*U V ZKHUH $ LV WKH HOHPHQW RI LQWHUHVW ,QWHQVLW\ UDWLRV RI WKLV W\SH DUH WKH UDZ GDWD IRU FRPSXWHU SURJUDPV VXFK DV WKH 0$*,& SURJUDP ZULWWHQ E\ : &ROE\ DQG LQ XVH DW WKH 8QLYHUVLW\ RI )ORULGD IRU TXDQWL 4 WDWLYH HOHFWURQ PLFURSUREH DQDO\VLV )LJXUHV DQG VKRZ SORWV RI LQWHQVLW\ UDWLRV YHUVXV GLVWDQFH IRU VDPSOHV RI DOSKD DQG EHWD EUDVV H[SRVHG LQ 1 +& RU LQ ,1 1D&O 7KH DFFRPSDQ\LQJ SKRWRPLFURJUDSKV

PAGE 45

)LJXUH =LQF LQWHQVLW\ SURILOH IURP D VDPSOH RI DOSKD EUDVV GH]LQFLILHG IRU GD\V LQ ,1 1D&O

PAGE 46

0,&5216 )LJXUH =LQF LQWHQVLW\ SURILOH IURP D VDPSOH RI DOSKD EUDVV GH]LQFLILHG IRU GD\V LQ 1 +& DW r&

PAGE 47

2 &' / 7U L 0,&5216 B)LJXUH =LQF LQWHQVLW\ SURILOH IURP D VDPSOH RI EHWD EUDVV GH]LQFLILHG IRU WZR GD\V LQ 1 0&, DW r& 2n

PAGE 48

)LJXUHV DQG VKRZn WKH FDUERQ WUDFHV OHIW RQ WKH VDPSOH E\ WKH LQWHUDFWLRQ RI VPDOO DPRXQWV RI GLIIXVLRQ SXPS RLO ZLWK WKH KLJKHQHUJ\ HOHFWURQ EHDP 7KH GLIIXVLRQ ]RQH RFFXUV LQ HDFK FDVH DW WKH YHU\ HGJH RI WKH REYLRXV FKDQJH LQ FRORU DV VKRZQ RQ WKH SKRWRPLFURJUDSKV )LJXUHV DQG DUH WKXV WR WKH DXWKRUnV NQRZOn HGJH WKH ILUVW H[SHULPHQWDO SORWV RI DOOR\ FRPSRVLWLRQ YHUn VXV GLVWDQFH WR EH REWDLQHG IURP VDPSOHV VXEMHFWHG WR GHn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n YDO RI DW OHDVW VL[ PLFURQV DQG DUH WKXV QRW GXH WR RYHUODS RI WKH EHDP DW DGMDFHQW SRVLWLRQ VHWWLQJV

PAGE 49

)LJXUH 3KRWRPLFURJUDSK RI VDPSOH VKRZQ LQ )LJXUH 3UREH WUDFH LV LQGLFDWHG E\ WKH DUURZ ;

PAGE 50

)LJXUH 3KRWRPLFURJUDSK RI VDPSOH VKRZQ LQ )LJXUH 3UREH WUDFH LV LQGLFDWHG E\ WKH DUURZ 1RWH VXUIDFH GHSRVLW RI FRSSHU ;

PAGE 51

, )LJXUH 3KRWRPLFURJUDSK RI VDPSOH VKRZQ LQ )LJXUH 3UREH WUDFH LV LQGLFDWHG E\ WKH DUURZ ;

PAGE 52

237,&$/ 0(7+2'6 (DUO\ ZRUN RQ WKH GH]LQFLILFDWLRQ PHFKDQLVP UHOLHG KHDYLO\ RQ RSWLFDO REVHUYDWLRQV 0RVW RI WKH UHVHDUFK WRROV XVHG LQ WKLV VWXG\ ZHUH QRW GHYHORSHG DW WKDW WLPH EXW WKH PLFURVFRSH DQG PHWDOORJUDSK ZHUH DYDLODEOH DQG ZHUH XVHG 7KH UHVXOWV REWDLQHG IURP PLFURVFRSLF LQYHVWLJDWLRQ ZHUH KRZHYHU VXEMHFW WR RSLQLRQW\SH LQWHUSUHWDWLRQV DQG UHn VHDUFKHUV GLG QRW KDYH WKH EHQHILW RI SUHVHQWGD\ NQRZOHGJH RI FU\VWDO VWUXFWXUH JUDLQ JURZWK HSLWD[LDO HOHFWURO\WLF GHSRVLWLRQ WKH FRQFHSW RI DQ RFFOXGHG FHOO DQG RWKHU LGHDV ZKLFK DUH SDUW RI WKH SUHVHQWGD\ UHVHDUFKHUnV EDFNJURXQG ,Q $EUDPVA ` VXJJHVWHG WKDW GH]LQFLILFDWLRQ RFFXUUHG ZKHQ D PHPEUDQH RI VRPH W\SH ZDV DYDLODEOH WR KROG GLVVROYHG FRSSHU LQ FRQWDFW ZLWK WKH EUDVV VXUIDFH RU ZKHQ D ODUJH H[FHVV RI FRSSHU ZDV SUHVHQW LQ WKH VROXWLRQ +LV H[SHULPHQWV ZLWK FRSSHU FKORULGH VROXWLRQV OHG WR IXUWKHU ZRUN E\ %HQJRXJK DQG 0D\AAf DQ VROXWLRQV RI WKLV W\SH VRRQ EHFDPH DQ DFFHSWHG PHWKRG RI DFFHOHUDWHG WHVWLQJ IRU WKH VXVFHSWLELOLW\ RI DOOR\V WR GH]LQFLILFDWLRQ (DUO\ UHVHDUFKHUV QRWHG WKDW GH]LQFLILFDWLRQ FRXOG EH FODVVLILHG DV RFFXUULQJ HLWKHU LQ OD\HUV RU LQ SOXJV VXFK DV WKH RQH VKRZQ LQ )LJXUH 7KHVH SOXJV ZHUH

PAGE 53

)LJXUH 'H]LQFLILFDWLRQ SOXJ LQ DOSKD EUDVV H[SRVHG IRU GD\V LQ ,1 1D&O DW URRP WHPSHUDWXUH ;

PAGE 54

DWWULEXWHG WR EUHDNV LQ D SURWHFWLYH VFDOH RQ WKH PHWDO VXUn IDFH ZKLFK DOORZHG ORFDOL]HG DWWDFN LQ FHUWDLQ VSHFLILF ORFD WLRQV ,W VHHPHG REYLRXV WKDW EUHDNV RI WKLV W\SH ZRXOG FDXVH IORZ UHVWULFWLRQV ZKLFK ZRXOG OHDG WR D KLJK FRQFHQn WUDWLRQ RI FRSSHU LQ D UHJLRQ DGMDFHQW WR WKH FRUURGLQJ VXUn IDFH %HQJRXJK DQG 0D\ SRLQWHG RXW WKDW SUHFLSLWDWLRQ RI FRSSHU GXH WR FRQFHQWUDWLRQ HIIHFWV VHHPHG OLNHO\ DQG IRU D ORQJ WLPH WKLV VHHPHG WR EH WKH ORJLFDO PHFKDQLVP IRU GH]LQFLILFDWLRQ 6LPPRQV ODWHU SRLQWHG RXW WKDW GH]LQFLIL FDWLRQ SOXJV RFFXUUHG XQGHU FLUFXPVWDQFHV ZKLFK LI VOLJKWO\ DOWHUHG ZRXOG UHVXOW LQ SLWV 0RVW HDUO\ LQYHVWLJDWRUV ZHUH FRQFHUQHG ZLWK GH]LQFLILn FDWLRQ RI FRQGHQVHU WXEHV XQGHU FRQGLWLRQV ZKHUH VFDOH EXLOGn XS DQG VXEVHTXHQW FUDFNV LQ WKH VFDOH ZHUH OLNHO\ WR RFFXU 7KH LGHDV SXW IRUWK E\ $EUDPV DQG E\ %HQJRXJK DQG 0D\ JDLQHG ZLGH DFFHSWDQFH +RZHYHU WKLV H[SODQDWLRQ GLG QRW VHHP WR FRYHU GHDOOR\LQJ LQ VKLS SURSHOOHUV DQG LQ RWKHU KLJKIORZn UDWH VLWXDWLRQV ,QWHUHVW LQ WKH ILHOG UHPDLQHG KLJK DQG UHSRUWV RI PHWDOORJUDSKLF LQYHVWLJDWLRQV LQWR WKH GH]LQFLILFDWLRQ PHFKD QLVP FRQWLQXHG %DVVHWW DQG 3ROXVKNLQ DQG 6KXOGHQHUAf GHFLGHG WKDW GH]LQFLILFDWLRQ ZDV D VHOHFWLYH UHPRYDO SURFHVV DIWHU H[DPLQLQJ ODUJH QXPEHUV RI VDPSOHV ZKLFK KDG IDLOHG LQ VHUYLFH 7KLV FRQWUDVWHG ZLWK WKH RSLQLRQ RI +RUWRQ DQG VHUYHG WR HPSKDVL]H WKDW FRQFOXVLRQV EDVHG RQ PHWDOORJUDSKLF REVHUYDWLRQ DUH RIWHQ GHSHQGHQW RQ WKH RSLQLRQV DQG SULRU H[SHULHQFH RI WKH REVHUYHU

PAGE 55

7KH GLIILFXOWLHV RI GHILQLQJ DQG FRPSDULQJ WKH VHUYLFH FRQGLWLRQV UHSRUWHG E\ YDULRXV DXWKRUV DUH DGGLWLRQDO GUDZn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n SHULPHQWV VKRXOG EH FDUHIXOO\ FRQWUROOHG DQG WKDW OLPLWDWLRQV RI RSWLFDO PHWKRGV VKRXOG EH UHFRJQL]HG $ FRQFXUUHQW REn VHUYDWLRQ LV WKDW LQIRUPDWLRQ ZKLFK OHQGV VXSSRUW WR RQH WKHRU\ VKRXOG QRW EH PLVLQWHUSUHWHG DV SURRI RI WKH LQYDOLGn LW\ RI D FRQWUDVWLQJ WKHRU\ 6RPH EXW E\ QR PHDQV DOO RI WKH DXWKRUV GLVFXVVHG DERYH KDYH UHFRJQL]HG WKLVZKLOH RWKHUV XQIRUWXQDWHO\ KDYH QRW

PAGE 56

3UHVHQW :RUN ([SRVXUH VDPSOHV LQWHQGHG IRU [UD\ GLIIUDFWLRQ DQDO\n VLV ZHUH SUHSDUHG DFFRUGLQJ WR WKH SURFHGXUHV GHVFULEHG LQ WKH VHFWLRQ RQ VDPSOH SUHSDUDWLRQ 1$&( 6WDQGDUG 70 f 68JJHVWV WKDW FORVHG H[SRVXUH FHOOV VXFK DV ZHUH XVHG LQ WKLV LQYHVWLJDWLRQ VKRXOG FRQWDLQ VHYHUDO VDPSOHV WR EH SXOOHG DIWHU FHUWDLQ SHULRGV RI WLPH (DFK WLPH D VDPSOH LV UHPRYHG IURP WKH FHOO D UHSODFHPHQW VDPSOH LV DGGHG 7KH UHDVRQ ZKLFK LV JLYHQ IRU WKLV LV WR FKHFN IRU SRVVLEOH FKDQJHV LQ WKH FRUURVLYLW\ RI WKH HQYLURQPHQW $V DQ H[DPSOH LQ RQH VHULHV RI WHVWV UXQ GXULQJ WKLV LQn YHVWLJDWLRQ ILYH VDPSOHV ZHUH SUHSDUHG IRU HDFK FHOO 7KUHH ZHUH LPPHUVHG LQ WKH RULJLQDO VROXWLRQ $IWHU WHQ GD\V RQH VDPSOH ZDV UHPRYHG DQG UHSODFHG ZLWK D IUHVK VDPSOH 7KLV ZDV UHSHDWHG DW WKH HQG RI WZHQW\ GD\V DQG WKH WHVW ZDV WHUPLQDWHG DW WKH HQG RI WKLUW\ GD\V 7KXV ILYH VDPSOHV ZHUH REWDLQHG IURP HDFK WHVW DQG LI QR FKDQJHV LQ WKH FRUn URVLYLW\ RI WKH HQYLURQPHQW RFFXUUHG WKH DPRXQW RI GH]LQFLIL FDWLRQ H[SHULHQFHG E\ WKH VDPSOH H[SRVHG IRU WKH ILUVW WHQ GD\V RI WKH WHVW ZRXOG FRUUHVSRQG WR WKDW RI WKH VDPSOH H[n SRVHG IRU WKH ODVW WHQ 7KH VDPH VKRXOG KROG IRU VDPSOHV H[SRVHG IRU WKH ILUVW WZHQW\ GD\V DQG WKH ODVW WZHQW\ GD\V $V ZDV PHQWLRQHG DERYH WKHVH WHVWV ZHUH LQWHQGHG SULn PDULO\ DV D VRXUFH RI GH]LQFLILHG EUDVV WR EH XVHG IRU [UD\ SRZGHU VDPSOHV %HFDXVH RI WKLV HDFK VDPSOH ZDV D VLPSOH GLVF ZLWK D KROH GULOOHG QHDU RQH HGJH VR LW FRXOG EH

PAGE 57

VXSSRUWHG LQ WKH H[SRVXUH FHOO 1R HIIRUW ZDV PDGH WR PDVN RII DOO EXW D FHUWDLQ DUHD IRU H[SRVXUH DQG WKH H[SRVHG DUHD ZDV DERXW 6 FPA LQ PO RI VROXWLRQ a [ A FP POf ZKLFK LV TXLWH KLJK 7KH WHVWV ZHUH FRQGXFWHG LQ 1 +& DW YDU\LQJ WHPSHUDWXUHV 'XULQJ WKHVH WHVWV D QXPEHU RI WKH VDPSOHV DSSHDUHG WR KDYH VXUIDFH GHSRVLWV ZKLFK KDG D PHWDOOLF OXVWHU DQG JUHZ ODUJHU ZLWK WKH SDVVDJH RI WLPH 7KHVH GHSRVLWV ZHUH DSSDUHQW RQ VDPSOHV ZKLFK OLDG EHHQ H[n SRVHG IURP WKH EHJLQQLQJ RI WKH WHVW DV ZHOO DV RQ VDPSOHV ZKLFK ZHUH DGGHG ZKHQ RWKHU VDPSOHV ZHUH UHPRYHG )LJXUH VKRZV D SKRWRPLFURJUDSK RI D FURVV VHFW LRQ RI RQH RI WKHVH VDPSOHV 7KH QRUPDO GH]LQFLILHG WH[WXUH DW WKH ERWWRP RI WKH VDPSOH FRQWUDVWV VKDUSO\ ZLWK WKH DSSHDUn DQFH RI WKH VKLQ\ VXUIDFH GHSRVLWV 7KDW WKHVH IRUPDWLRQV DUH LQ IDFW GHSRVLWV LV VKRZQ LQ )LJXUH ZKHUH D SRUWLRQ RI WKH RULJLQDO EUDVV VXUIDFH LV YLVLEOH LQ WKH SKRWRPLFURJUDSK (DFK VDPSOH LQ WKHVH WHVWV ZDV ZHLJKHG EHIRUH DQG DIWHU WKH WHVW 7KH ZHLJKWORVV GDWD GLG QRW IROORZ DQ\ UHFRJQL]n DEOH SDWWHUQ 2QH VDPSOH LQ WKHVH WHVWV DFWXDOO\ JDLQHG ZHLJKW 7KLV SDUWLFXODU VDPSOH ZKLFK ZDV H[SRVHG IRU WKH ODVW WHQ GD\V RI D WZHQW\GD\ WHVW KDG QXPHURXV VXUIDFH GHSRVLWV ZKLFK ZHUH FRSSHU FRORUHG DQG KDG D PHWDOOLF OXVWHU )LJXUH LV D VFDQQLQJ HOHFWURQ PLFURJUDSK RI RQH RI WKHVH GHSRVLWV ZKLFK ILWV WKH ULGJH GHSRVLWLRQ GHVFULSWLRQ

PAGE 58

)LJXUH &RSSHU GHSRVLWV RQ WKH VXUIDFH RI GH]LQFLILHG DOSKD EUDVV VDPSOH 6DPSOH ZDV H[SRVHG WR 1 +& IRU GD\V DW r& ;

PAGE 59

)LJXUH 'HSRVLW RQ VXUIDFH RI GH]LQFLILHG VDPSOH 6DPSOH ZDV H[SRVHG LQ 1 r& IRU GD\V ; DOSKD EUDVV +& DW

PAGE 60

)LJXUH 6FDQQLQJ HOHFWURQ PLFURJUDSK RI FRSSHU VODE SURWUXGLQJ IURP WKH VXUIDFH RI D GH]LQFLILHG DOSKD EUDVV VDPSOH ;

PAGE 61

RI FRSSHU GHSRVLWV SURYLGHG E\ %RFNULV DQG 'DPMDPRYLF 2WKHU GHSRVLWV KDG WKH DSSHDUDQFH RI ORQJ FXUOHGXS ZRRG VKDYLQJV )LJXUH LV D QRQGLVSHUVLYH [UD\ SDWWHUQ RI WKH IRUPDWLRQ VKRZQ LQ )LJXUH DQG WKLV SDWWHUQ FOHDUO\ VKRZV WKH GHSRVLW WR EH PHWDOOLF FRSSHU 7KH VDPSOH ZDV H[n SRVHG WR K\GURFKORULF DFLG LQ D 3\UH[ UHDFWLRQ NHWWOH DQG WKH VLOLFRQ DQG FKORULQH SHDNV DUH SUREDEO\ GXH WR FRQWDPLn QDWLRQ RQ WKH VXUIDFH RI WKH GHSRVLW 2QH RI WKH GHSRVLWV ZDV SXOOHG IURP WKH VDPSOH VXUIDFH ZLWK D SDLU RI WZHH]HUV DQG PDGH LQWR DQ [UD\ SRZGHU VDPSOH 7KH GLIIUDFWLRQ SDWWHUQ REWDLQHG IURP WKLV GHSRVLW FRQILUPHG WKDW WKH GHSRVLWV ZHUH PHWDOOLF FRSSHU 7DEOH VXPPDUL]HV WKH ZHLJKW FKDQJH GDWD IURP WKH VDPSOH GLVFXVVHG DERYH )LJXUH LV D SKRWRPLFURJUDSK RI WKLV VDPSOH DQG VKRZV WKDW WKH VDPSOH GLG GH]LQFLI\ DV ZHOO DV SURYLGH VLWHV IRU WKH GHSRVLWLRQ RI FRSSHU 1R GHSRVLWV DSSHDU LQ SKRWRPLFURJUDSKV RI WKLV VDPSOH EHFDXVH WKH\ ZHUH UHPRYHG IRU WKH [UD\ GLIIUDFWLRQ DQG ZHLJKWFKDQJH DQDO\VHV GHVFULEHG DERYH 7KH DERYH UHVXOWV FOHDUO\ VKRZ WKDW ODUJH VXUIDFHDUHD WRFHOOYROXPH UDWLRV DQG WKH DGGLWLRQ RI VDPSOHV WR VROXWLRQV DOUHDG\ FRQWDLQLQJ GLVVROYHG FRSSHU DUH VLWXDWLRQV WR EH DYRLGHG $OO IXUWKHU WHVWV ZHUH FRQGXFWHG RQ D RQHVDPSOH SHUFHOO EDVLV WR DYRLG WKH HIIHFWV GHVFULEHG DERYH

PAGE 62

)LJXUH 1RQGLVSHUVLYH [UD\ DQDO\]HU SDWWHUQ RI GHSRVLW VKRZQ LQ )LJXUH

PAGE 63

)LJXUH 'H]LQFLILHG FURVVVHFWLRQ RI VDPSOH VKRZQ LQ )LJXUH ;

PAGE 64

7DEOH 6DPSOH :HLJKW ,QIRUPDWLRQ 2ULJLQDO ZHLJKW )LQDO ZHLJKW :HLJKW JDLQHG JP JP JP :HLJKW RI VDPSOH ZLWK GHSRVLWV UHPRYHG E\ WZHH]HUV DQG XOWUDn VRQLF FOHDQLQJ JP :HLJKW RI GHSRVLWV E\ GLIIHUHQFHf JP

PAGE 65

62/87,21 $1$/<6,6 &KHPLFDO FKDQJHV LQ WKH HOHFWURO\WH VXUURXQGLQJ VDPSOHV XQGHUJRLQJ GHDOOR\LQJ FDQ SURYLGH LQIRUPDWLRQ UHJDUGLQJ WKH PHFKDQLVPV LQYROYHG $QDO\WLFDO PHWKRGV ZKLFK KDYH EHHQ XVHG WR PRQLWRU PHWDOLRQ SLFNXS LQ VROXWLRQ LQFOXGH HOHFWUR  FKHPLFDO PHWKRGV >VSOLWULQJ HOHFWURGHV f/ DQG SRODURJ f f UDSK\Y f f FRORULPHWU\ f DQG UDGLRDFWLYH f WUDFHU WHFKQLTXHV n $OOR\ V\VWHPV LQYHVWLJDWHG LQFOXGH f f WKH FRSSHU]LQF f f f f FRSSHUJROG f DQG I FRSSHUQLFNHOY V\VWHPV 2I SDUWLFXODU LQWHUHVW LV ZKHWKHU RU QRW WKH PRUH QREOH VSHFLHV RI D ELQDU\ DOOR\ GLVVROYHV GXULQJ GHDOOR\LQJf f &RORULPHWU\ DQG SRODURJUDSK\ DUH WKH RQO\ WZR WHFKQLTXHV ZKLFK KDYH EHHQ UHSRUWHG KHUHWRIRUH ZKLFK FDQ GHWHFW WKH SUHVHQFH RI WUDFH HOHPHQWV LQ VROXWLRQ $WRPLF DEn VRUSWLRQ LV D WHFKQLTXH KDYLQJ GHWHFWLRQ OHYHOV DV JRRG DV RU EHWWHU WKDQ HLWKHU FRORULPHWU\ RU SRODURJUDSK\ApA DQG WKLV LV WKH PHWKRG HPSOR\HG LQ WKH SUHVHQW VWXG\ U )LVKHU DQG +DOSHULQA UHSRUW WKDW QR JROG ZDV GHWHFWn HG GXULQJ WKHLU FRORULPHWU\ H[SHULPHQWV ZLWK FRSSHUJROG DOOR\V 7KLV LV LQ DJUHHPHQW ZLWK WKH FRORULPHWULF GDWD RI 3LFNHULQJ DQG %\UQH RQ WKH VDPH V\VWHP %\ FRQWUDVW

PAGE 66

FRSSHU GLVVROYHV IURP FRSSHU]LQF DOOR\V XQGHU FHUWDLQ GH f DOOR\LQJ FRQGLWLRQV f f 0DUVKDNRY DQG FRZRUNHUV LQWURGXFHG WKH FRQFHSW RI D GH]LQFLILFDWLRQ IDFWRU ZKLFK FDQ EH GHILQHG E\ WKH HTXDn WLRQ B =Q&Xf VROXWLRQ =Q&Xf DOOR\ 7KH =Q&Xf UDWLR LQ VROXWLRQ LV GHWHUPLQHG E\ FKHPLFDO DQDO\VLV RI WKH VROXWLRQ DQG =Q&Xf DOOR\ LV WKH UDWLR RI ZHLJKW SHUFHQWV RI ]LQF WR FRSSHU LQ WKH DOOR\ 0DUVKDNRY DQG FRZRUNHUV VWXGLHG GH]LQFLILFDWLRQ XQGHU D YDULHW\ RI FRQGLWLRQV LQ ERWK 1D&O DQG +& VROXWLRQV 7KH\ VWDWH WKDW DOSKD EUDVVHV LQ DFLG PHGLD KDYH D GH]LQFLILFDWLRQ IDFWRU VOLJKWO\ LQ H[FHVV RI XQLW\ 7KLV ZRXOG PHDQ WKDW WKH UDWLR RI ]LQF WR FRSSHU LQ VROXWLRQ LV VOLJKWO\ JUHDWHU WKDQ LW LV LQ DOSKD EUDVV 1R FRSSHU ZDV GHWHFWHG LQ DFLG VROXn WLRQV ZKLFK KDG EHHQ LQ FRQWDFW ZLWK EHWD DQG JDPPD EUDVV = R}f 1R RWKHU UHSRUWV RQ VROXWLRQ DQDO\VLV KDYH DSn SHDUHG IRU EHWD EUDVV 7KH GLVVROXWLRQ RI FRSSHU IURP DOSKD EUDVV KDV EHHQ LQWHUSUHWHG DV HYLGHQFH IRU D UHGHSRVLWLRQ PHFKDQLVPY RU DV DQ LQGLFDWLRQ WKDW WKH GH]LQFLILHG FRSSHU OD\HU ZDV XQGHUJRLQJ GLVVROXWLRQAAAAA 5HFHQW UDGLRWUDFHU H[n SHULPHQWV LQGLFDWH WKDW H[FKDQJH RI FRSSHU EHWZHHQ D VROX WLRQ DQG D FRSSHUFRQWDLQLQJ PHWDO VXUIDFH FDQ RFFXU -

PAGE 67

7KLV ZRXOG VHHP WR LQGLFDWH WKDW WKH SUHVHQFH RI FRSSHU LQ VROXWLRQ IURP D GHDOOR\HG PHWDO FRXOG DOVR EH LQWHUSUHWHG DV EHLQJ WKH UHVXOW RI YDU\LQJ GLVVROXWLRQ UDWHV RU LQ RWKHU ZRUGV D SUHIHUHQWLDO UHPRYDO SURFHVV LQ ZKLFK RQH PHWDO GLVVROYHV IDVWHU WKDQ WKH RWKHU 6ROXWLRQ DQDO\VLV KDV EHHQ XVHG WR PHDVXUH WKH UDWH RI U GHDOOR\LQJ DV D IXQFWLRQ RI WLPH DQGRU WHPSHUDWXUH f f f 5XE LQ UHSRUWHG WKH SDUWLDO DSSDUHQW KHDWV RI DFWLYDWLRQ DV D IXQFWLRQ RI FRPSRVLWLRQ IRU WKH GLVVROXWLRQ RI D VHULHV RI FRSSHUQLFNHO DOOR\V VRPH RI ZKLFK GHQLFNHO LILHG LQ ,1 +& 7KH WRWDO DSSDUHQW DFWLYDWLRQ HQHUJLHV WKH VXPV RI WKH WZR SDUWLDO YDOXHV YDULHG IURP WR .FDOPROH RI DOOR\ 7KH ORZHU ILJXUH FRUUHVSRQGLQJ WR SXUH FRSSHU FRPSDUHV TXLWH ZHOO ZLWK WKH .FDOPROH REWDLQHG E\ +DOSHULQAr IRU WKH GLVVROXWLRQ RI FRSSHU LQ DPPRQLD VROXWLRQV 7KH YDOXHV UHSRUWHG DUH DOVR LQ WKH JHQHUDO UDQJH GHVFULEHG E\ 9HWWHU IRU WKH GLVVROXWLRQ RI PHWDOV LQ f HOHFWURO\WHVY )LVKHU DQG +DOSHULQAUHSRUWHG WKDW FRSSHUJROG DOn OR\V VKRZHG D GHFUHDVH LQ GLVVROXWLRQ UDWH ZLWK DQ LQFUHDVH LQ WHPSHUDWXUH IRU DOO RI WKHLU DOOR\V 7KH\ DOVR REVHUYHG SDUDEROLF FRUURVLRQ UDWHV 3DUDEROLF FRUURVLRQ UDWHV DUH FKDUDFWHULVWLF RI V\VWHPV LQ ZKLFK WKH UDWH LV GHWHUPLQHG E\ WUDQVIHU RI UHDFWDQWV WKURXJK DQ DGKHUHQW VXUIDFH ILOP ZKLFK WKLFNHQV DV WKH UHDFWLRQ SURJUHVVHV7KXV WKH\ FRQFOXGHG WKDW WKH UDWH ZDV OLPLWHG E\ D VXUIDFH ILOP

PAGE 68

SUREDEO\ &X DQGRU &X2 ZKLFK IRUPHG XQGHU VWDJQDQW FRQn GLWLRQV ZLWKLQ WKH UHVLGXDO JROG VSRQJH +LJKHU WHPSHUDn WXUHV ZRXOG IDYRU SUHFLSLWDWLRQ RI FRSSHU R[LGHV DQG WKH IRUPDWLRQ RI GHQVHU ILOP VWUXFWXUHV WKXV DFFRXQWLQJ IRU WKH REVHUYHG LQYHUVH WHPSHUDWXUH GHSHQGHQFH RI WKH UHDFWLRQ NLQHWLFV 7KH FRQFHQWUDWLRQ RI PHWDO LRQV LQ VROXWLRQ KDV DOVR  EHHQ XVHG WR FDOFXODWH PHWDOGLVVROXWLRQ FXUUHQWV f 7KH UHVXOWV RI WKHVH LQYHVWLJDWLRQV DUH GLVFXVVHG LQ WKH VHFWLRQ RQ HOHFWURFKHPLFDO LQYHVWLJDWLRQV

PAGE 69

3UHVHQW :RUN $WRPLF DEVRUSWLRQ VSHFWURSKRWRPHWU\ ZDV XVHG WR DQDO\]H 1 +& VROXWLRQV ZKLFK KDG EHHQ LQ FRQWDFW ZLWK IUHHO\ FRUn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n SRUWDQW HIIHFW RQ GLVVROXWLRQ NLQHWLFV 7KH SRVVLELOLW\ H[LVWV WKDW SUHFLSLWDWHG UHDFWLRQ SURGXFWV FDQ IRUP D VXUn IDFH ILOP ZKLFK UHWDUGV GLVVROXWLRQ 7KH VXUIDFH DUHD RI WKH VSRQJH LV VXEVWDQWLDOO\ JUHDWHU WKDQ WKDW RI WKH DOOR\ DW WKH FRUURVLRQ LQWHUIDFH 7KLV PHDQV WKDW VLWHV IRU WKH HOHFWURGHSRVLWLRQ RI FRSSHU DUH LQFUHDVHG DV ZHOO DV WKH VXUIDFH DUHD IRU GLVVROXWLRQ RI FRSSHU IURP WKH VSRQJH %ORFNHG RU QDUURZ SDVVDJHZD\V FDQ OHDG WR VWDJQDQW FRQGLn WLRQV DQG WKH SUHFLSLWDWLRQ RI VDOWV ZKLFK ZRXOG EH VROXEOH LQ WKH EXON VROXWLRQ $OO WKHVH HIIHFWV FDQ DOWHU WKH PHWDO FRQFHQWUDWLRQ LQ WKH EXON VROXWLRQ DQG WKHLU SRVVLEOH LQn IOXHQFH PXVW EH FRQVLGHUHG LQ WKH DQDO\VLV RI GDWD RQ PHWDO GLVVROXWLRQ

PAGE 70

7KH H[SRVXUH PHWKRG XVHG IRU WKHVH WHVWV YDV VLPLODU WR WKDW UHSRUWHG E\ )LVKHU DQG +DOSHULQAf 5HDFWLRQ NHWn WOHV ZHUH XVHG WR H[SRVH VDPSOHV FXW IURP WKH VDPH LQJRW DQG KDYLQJ WKH VDPH VXUIDFH DUHD FP IRU DOSKD EUDVV FP IRU EHWD EUDVVf LQ PO RI DUJRQVSDUJHG 1 +& DW YDULRXV WHPSHUDWXUHV 7ZHQW\ILYH PLOOLOLWHU DOLTXRWV RI DFLG ZHUH UHPRYHG IURP WKH FHOOV DW KRXU LQWHUYDOV (DFK DOLTXRW RI VDPSOH ZDV UHSODFHG ZLWK DQ DOLTXRW RI WKH VWRFN VROXWLRQ XVHG WR ILOO WKH FHOO 7KLV FDXVHG D GLOXn WLRQ RI b PO RXW RI PO WRWDOf HDFK WLPH D VDPSOH ZDV UHPRYHG &RUUHFWLRQV ZHUH PDGH IRU WKHVH GLOXWLRQV LQ WKH GDWD ZKLFK IROORZ 7KH DOLTXRWV ZHUH DQDO\]HG XVLQJ D +HDWK 0RGHO (8 $WRPLF $EVRUSWLRQ 6SHFWURSKRWRPHWHU VHH $SSHQGL[ f %RWK FRSSHU DQG ]LQF FRQFHQWUDWLRQV ZHUH GHWHUPLQHG IRU HDFK VDPSOH 'DWD REWDLQHG IURP DWRPLF DEVRUSWLRQ GHWHUPLQDWLRQV DUH WDEXODWHG LQ 7DEOHV WKURXJK $WRPLF DEVRUSWLRQ GDWD DUH UHSRUWHG WR RQO\ WZR VLJQLILFDQW ILJXUHV FRQVLVn WHQW ZLWK WKH DFFXUDF\ RI WKH PHWKRG )LJXUH VKRZV DQ DWRPLFDEVRUSWLRQ FDOLEUDWLRQ FXUYH IRU FRSSHU 7KLV ILJXUH VKRZV WKDW FRSSHU FDQ EH GHWHUPLQHG WR DW OHDVW WKH QHDUHVW SSP ZLWK D UHDVRQDEOH GHJUHH RI FHUWDLQW\ +RZHYHU WKH SUHFLVLRQ RU WKH UHODWLYH DFFXUDF\ RI WKH PHDVXUHPHQW ZKHQ FRPSDUHG WR WKH DPRXQW RI PHWDO

PAGE 71

7DEOH $WRPLF $EVRUSWLRQ 'DWD IRU $OSKD %UDVV LQ 1 +& DW r& 7RWDO ZHLJKW 7LPH KRXUVf $PRXQW LQ SSPf VROXWLRQ ZWf D :HLJKW UHPRYHG 7RWDO ZHLJKW UHPRYHG WKDW KDV EHHQ LQ VROXWLRQ $PRXQW OHIW LQ VROXWLRQ = &X =Q &X =Q &X =Q &X =Q &X =Q &X =Q D$O ZHLJKWV LQ r1 JP

PAGE 72

7DEOH $WRPLF $EVRUSWLRQ 'DWD IRU $OSKD %UDVV LQ 1 +& DW r& 7RWDO ZHLJKW 7 LPH KRXUVf $PRXQW LQ SSPf VROXWLRQ }WfD :HLJKW UHPRYHG 7RWDO ZHLJKW UHPRYHG WKDW KDV EHHQ LQ VROXWLRQ $PRXQW OHIW LQ VROXWLRQ = &X =Q &X =Q &X =Q &X =Q &X =Q &X =Q D$OO ZHLJKWV LQ JP

PAGE 73

7DEOH $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& 7 LPH KRXUVf $PRXQW SSPf LQ VROXWLRQ ZWf D :HLJKW UHPRYHG 7RWDO ZHLJKW UHPRYHG 7RWDO ZHLJKW WKDW KDV EHHQ LQ VROXWLRQ $PRXQW OHIW LQ VROXWLRQ = &X =Q &X =Q &X =Q &X =Q &X =Q &X =Q D$OO ZHLJKWV LQ f JP

PAGE 74

7DEOH $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& 7LPH KRXUVf $PRXQW LQ VROXWLRQ SSPf ZWfD :HLJKW UHPRYHG 7RWDO ZHLJKW UHPRYHG 7RWDO ZHLJKW WKDW KDV EHHQ LQ VROXWLRQ $PRXQW OHIW LQ VROXWLRQ ] &X =Q &X =Q &X =Q &X =Q &X =Q &X =Q D$OO ZHLJKW V LQ JP

PAGE 75

7DEOH $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& 7LPH $PRXQW LQ VROXWLRQ :HLJKW KRXUVf SSPf ZWfD UHPRYHG &X =Q &X =Q &X =Q D$OO ZHLJKWV LQ A JP

PAGE 76

7DEOH ([WHQGHGf 7RWDO ZHLJKW UHPRYHG 7RWDO ZHLJKW WKDW KDV EHHQ LQ VROXWLRQ $PRXQW OHIW LQ VROXWLRQ = &X =Q &X =Q &X =Q

PAGE 77

7DEOH $WRPLF $EVRUSWLRQ 'DWD IRU %HWD %UDVV LQ 1 +& DW r& 7LPH $PRXQW LQ VROXWLRQ :HLJKW KRXUVf SSPf ZWfD UHPRYHG &X =Q &X =Q &X =Q D$OO ZHLJKWV LQ A JP

PAGE 78

7DEOH ([WHQGHGf 7RWDO ZHLJKW UHPRYHG 7RWDO ZHLJKW WKDW KDV EHHQ LQ VROXWLRQ $PRXQW OHIW LQ VROXWLRQ = &X =Q &X =Q &X =Q

PAGE 79

330 )LJXUH $WRPLFDEVRUSWLRQ FDOLEUDWLRQ FXUYH IRU FRSSHU

PAGE 80

DYDLODEOH IRU PHDVXUHPHQW LV ORZHU IRU PRUH GLOXWH VDPSOHV $V DQ H[DPSOH D SSP GHYLDWLRQ LQ D VDPSOH KDYLQJ SSP FRSSHU \LHOGV D SHUFHQW HUURU ZKHUHDV LQ D VDPSOH KDYLQJ SSP FRSSHU WKH VDPH SSP GHYLDWLRQ ZRXOG SURGXFH DQ HUURU RI RQO\ SHUFHQW 7KLV W\SH RI SUHFLVLRQ GHYLDWLRQ LV FRPPRQ WR DOO DWRPLF DEVRUSWLRQ PHWKRGV 7KH FRQILGHQFH OLPLWV VKRZQ RQ WKH ILJXUHV ZKLFK IROORZ ZHUH FDOFXODWHG E\ DVVXPLQJ D SRVVLEOH s SSP GHYLDWLRQ IRU WKH RULJLQDO LQVWUXPHQW UHDGLQJ REWDLQHG RQ WKH VDPSOH LQ TXHVWLRQ 7KH +HDWK VSHFWURSKRWRPHWHU LV D VLQJOHEHDP LQVWUXPHQW 7KLV PHDQV WKDW QR SURYLVLRQ LV PDGH IRU DXWRPDWLFDOO\ DGn MXVWLQJ IRU LQVWUXPHQWD@ GULIW GXH WR SRZHU GHYLDWLRQV RU ODPS IOXFWXDWLRQV LW ZDV IRXQG LQ SUDFWLFH WKDW WKH LQVWUXPHQW ZRXOG GULIW VLJQLILFDQWO\ LQ D VKRUW SHULRG RI WLPH WKXV QHJDWLQJ WKH YDOXH RI D FDOLEUDWLRQ FXUYH )RU WKLV UHDVRQ D SURFHGXUH ZDV GHYHORSHG ZKHUHE\ HDFK VDPSOH ZDV LQWURGXFHG LQWR WKH DEVRUSWLRQ IODPH RQFH WR GHWHUPLQH WKH JHQHUDO UDQJH RI WKH FRQFHQWUDWLRQ RI WKH HOHPHQW EHLQJ DQDO\]HG 7KHQ VWDQGDUGV ZHUH LQWURGXFHG LQWR WKH IODPH WR FRQILUP WKH FRQFHQWUDWLRQ UDQJH RI WKH XQNQRZQ 7KH YDOXHV WDEXODWHG LQ 7DEOHV WKURXJK ZHUH WKHQ GHWHUPLQHG E\ LQWURGXFLQJ RQH VWDQGDUG WKHQ WKH XQNQRZQ DQG WKHQ DQRWKHU VWDQGDUG VR WKDW HDFK XQNQRZQ UHDGLQJ ZDV EUDFNHWHG EY WZR VWDQGDUG UHDGLQJV

PAGE 81

7KH OLPLWV RI GHWHFWLRQ IRU WKH XQLW XQGHU WKH RSHUDWn LQJ FRQGLWLRQV XVHG LQ WKLV LQYHVWLJDWLRQ ZHUH DSSUR[LPDWHO\ SSP IRU ]LQF DQG DSSUR[LPDWHO\ SSP IRU FRSSHU 7KHRUHWLFDO OLPLWV RI GHWHFWLRQ IRU WKHVH HOHPHQWV E\ DWRPLF DEVRUSWLRQ DUH VRPHZKDW ORZHU WKDQ WKLV EXW WKH VROXWLRQV DQDO\]HG KDG FRSSHU DQG ]LQF FRQWHQWV ZHOO DERYH WKHVH OLPn LWV DQG PRVW RI WKHP UHTXLUHG GLOXWLRQ WR EH EURXJKW ZLWKn LQ WKH RSHUDWLQJ UDQJH RI WKH LQVWUXPHQW )LJXUH VKRZV WKH GLOXWLRQFRUUHFWHG YDOXHV IRU FRSn SHU DQG ]LQF GLVVROYHG IURP DOSKD EUDVV DW YDULRXV WLPHV VHH DOVR &ROXPQ RI 7DEOH f 7KH ]LQF YDOXHV REH\ OLQHDU NLQHWLFV EXW WKH FRSSHL YDOXHV IOXFWXDWH &RSSHU DQG ]LQF GLVVROXWLRQ IURP DOSKD EUDVV DW r& LV VKRZQ LQ )LJXUH 2QFH DJDLQ OLQHDU NLQHWLFV DUH REVHUYHG IRU ]LQFDQG WKH FRSSHU YDOXHV IOXFWXDWH VRPHZKDW EHIRUH WKH\ WRR EHFRPH OLQHDU 7KH GH]LQFLILFDWLRQ IDFWRU = WKDW ZDV LQWURGXFHG E\ 0DUVKDNRY DQG FRZRUNHUV LV SORWWHG IRU WKLV FHOO LQ )LJXUH :KHQ WKH FRQILGHQFH OLPLWV DUH WDNHQ LQWR FRQn VLGHUDWLRQ = IRU DOSKD EUDVV LV VHHQ WR EH EHWZHHQ DQG IRU WKLV LQYHVWLJDWLRQ 7KLV LV LQ JHQHUDO DJUHHPHQW ZLWK 0DUVKDNRY DQG FRZRUNHUV 7KH XVHIXOQHVV RI GHWHUPLQLQJ = LV OLPLWHG VLQFH WKH DEVROXWH PDJQLWXGHV RI WKH YDOXHV RI = DUH VWURQJO\ LQIOXHQFHG E\ WKH SUHFLVLRQ RI FKHPLFDO DQDO\n VLV PHWKRGV LQ SURFHVVHV VXFK DV GHDOOR\LQJf ZKHUH WKH TXDQWLWLHV EHLQJ DQDO\]HG DUH VPDOO

PAGE 82

&233(5 +2856 )LJXUH &RSSHU DQG ]LQF GLVVROXWLRQ IURP DOSKD EUDVV LQ 1 +& DW r& 6FDWWHU EDQGV VKRZ WKH SUHFLVLRQ RI WKH RULJLQDO PHDVXUHPHQWV

PAGE 83

&233(5 =,1& +2856 WVf )LJXUH &RSSHU DQG ]LQF GLVVROXWLRQ IURP DOSKD EUDVV LQ 1 +& DW r& 6FDWWHU EDQGV VKRZ WKH SUHFLVLRQ RI WKH RULJLQDO PHDVXUHPHQWV

PAGE 84

+2856 )LJXUH 'H]LQFLILFDWLRQ IDFWRUV = IRU DOSKD EUDVV LQ 1 +& DW r& 6FDWWHU EDQGV VKRZ WKH OLPLWV RI SUHFLVLRQ IRU WKH RULJLQDO PHDVXUHPHQWV

PAGE 85

)LJXUHV DQG VKRZ WKH FRUUHFWHG DPRXQWV RI FRSSHU DQG ]LQF ZKLFK KDYH GLVVROYHG IURP EHWD EUDVV DW r& DV D IXQFWLRQ RI WLPH 7KH\ DUH QRW SORWWHG RQ WKH VDPH VFDOH DV ZDV GRQH LQ )LJXUHV DQG IRU DOSKD EUDVV EHFDXVH WKH GLVVROXWLRQ UDWH IRU ]LQF )LJXUH f LV PXFK JUHDWHU WKDQ WKH GLVVROXWLRQ UDWH IRU FRSSHU )LJXUH f 7KH UHDFWLRQ NLQHWLFV IRU ERWK FRSSHU DQG ]LQF DSSHDU WR EH OLQHDU DIWHU WKH ILUVW KRXUV 6OLJKW IOXFWXDWLRQV LQ WKH FRSSHU RU ]LQF GLVVROXWLRQ NLQHWLFV PDNH GLVSURSRUWLRQDWHO\ ODUJH FKDQJHV LQ WKH PDJQLn WXGHV RI WKH GH]LQFLILFDWLRQ IDFWRU = 7KLV LV VHHQ LQ WKH SRVLWLRQV RI WKH FDOFXODWHG SRLQWV LQ )LJXUH 7KH YDOXHV VKRZQ LQ )LJXUH IRU EHWD EUDVV DUH XS WR ILIW\ WLPHV WKH YDOXHV SORWWHG LQ )LJXUH IRU DOSKD EUDVV ,W LV HPSKDn VL]HG KHUH WKDW WKH YDOXHV IRU EHWD EUDVV DUH ILQLWH LQ FRQn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

PAGE 86

+2856 )LJXUH &RSSHU GLVVROXWLRQ IURP EHWD EUDVV LQ 1 +& DW r&

PAGE 87

+2856 )LJXUH =LQF GLVVROXWLRQ IURP EHWD EUDVV LQ 1 +& DW r&

PAGE 88

= )LJXUH A -/ -/ / +2856 / 'H]LQFLILFDWLRQ IDFWRUV = IRU EHWD EUDVV LQ 1 +& DW r& 6FDWWHU EDQGV VKRZ WKH OLPLWV RI SUHFLVLRQ IRU WKH RULJLQDO PHDVXUHPHQWV

PAGE 89

)LJXUH =LQF GLVVROXWLRQ IURP EHWD EUDVV LQ 1 +& DW YDULRXV WHPSHUDWXUHV 6FDWWHU EDQGV VKRZ WKH OLPLWV RI SUHFLVLRQ IRU WKH RULJLQDO PHDVXUHPHQWV

PAGE 90

0DQ\ FKHPLFDO UHDFWLRQV LQFOXGLQJ PHWDO GLVVROXWLRQ DUH IRXQG WR REH\ DQ HPSLULFDO HTXDWLRQ ILUVW SURSRVHG E\ $UUKHQLXV . H57 ZKHUH WKH SUHH[SRQHQWLDO IDFWRU .T LV XVXDOO\ IRXQG WR EH WHPSHUDWXUHLQGHSHQGHQW DW OHDVW ZLWKLQ WKH H[SHULPHQWDO DFFXUDF\ RI WKH REVHUYDWLRQV f DQM T VRBFDLLH> DFWLYDWLRQ HQHUJ\ DQG .T LQ WKH DERYH HTXDWLRQ DUH UDWHUHODWHG PHDVXUHPHQWV VXFK DV ZHLJKW ORVV ZHLJKW JDLQ GHSWK RI SHQHWUDWLRQ DQG PHWDO GLVVROXWLRQ IRU FRUURVLRQ H[SHULPHQWV 7KH DPRXQW RI ]LQF GLVVROYHG LQ KRXUV ZDV XVHG LQ WKH GLVFXVVLRQ ZKLFK IROORZV 7KLV LV D UHDVRQDEOH FKRLFH VLQFH )LJXUH VKRZV WKH NLQHWLFV WR EH OLQHDU 5 LQ WKH $UUKHQLXV HTXDWLRQ LV WKH JDV FRQVWDQW FDORULHV JPPROHGHJUHH Af DQG 7 LV WKH DEVROXWH WHPSHUDWXUH LQ GHJUHHV .HOYLQ ,I UHDFWLRQ NLQHWLFV FDQ EH UHSUHVHQWHG E\ WKH DERYH $UUKHQLXV IRUPXOD WKHQ D SORW RI ORJ DV D IXQFWLRQ RI 7 ZLOO JLYH D VWUDLJKWOLQH KDYLQJ D VORSH RI 45 f f )LJXUH LV DQ $UUKHQLXV SORW RI WKH DPRXQW RI ]LQF GLVVROYHG IURP EHWD EUDVV LQ KRXUV 7KH DFWLYDWLRQ HQHUJ\ IRU ]LQF GLVVROXWLRQ RYHU WKH WHPSHUDWXUH LQWHUYDO r& LV IRXQG IURP WKH VORSH RI WKLV JUDSK WR EH .FDO JPPROH A 7KLV YDOXH LV VRPHZKDW KLJKHU WKDQ WKRVH IRU

PAGE 91

=,1& ',662/9(' ,1 +2856 JP 7 r.a )LJXUH $UUKHQLXV SORW RI ]LQF GLVVROXWLRQ UDWH 7 6FDWWHU EDQGV VKRZ WKH OLPLWV FLVLQ IRU WKH RULJLQDO PHDVXUHPHQWV YHUVXV RI SUH

PAGE 92

L I HOHFWURFKHPLFDO PHWDO GLVVROXWLRQ s .FDO JLQPROH af A WKH GLVVROXWLRQ RI FRSSHUQLFNHO DOOR\V DV UHSRUWHG E\ 5XELQ .FDO JPPROHff RU WKH GLVVROXWLRQ RI FRSSHU UHSRUWHG f f E\ +DOSHULQY .RIVWDG SRLQWV RXW WKH GLIILFXOW\ RI DVFULELQJ SK\VLFDO LQWHUSUHWDWLRQV WR H[SHULPHQWDOO\ GHWHUPLQHG DFWLYDWLRQ f HQHUJLHVY 6HYHUDO FRQFOXVLRQV FDQ EH GUDZQ IURP WKH DWRPLFDEn VRUSWLRQ GDWD 7KHUH FDQ EH QR GRXEW WKDW DW OHDVW VRPH FRSSHU HQWHUV VROXWLRQ IURP ERWK DOSKD DQG EHWD EUDVV ZKHQ IUHHO\ H[SRVHG XQGHU FRQGLWLRQV XVHG LQ WKHVH WHVWV 7KLV U FRQWUDGLFWV WKH UHVXOWV RI 0DUVKDNRY DQG FRZRUNHUVA n ZKR UHSRUWHG WKDW QR GLVVROYHG FRSSHU ZDV GHWHFWHG LQ VROXWLRQ IRU WKHLU H[SHULPHQWV ZLWK EHWD EUDVV LQ 60 1D&O DQG LQ 0 +& 7KHUH LV HYLGHQFH IRU D OLQHDU GLVVROXWLRQ UDWH IRU ]LQF IURP ERWK DOSKD DQG EHWD EUDVVHV VHH )LJXUHV DQG f 7KLV ZRXOG VXSSRUW WKH PHWDOORJUDSKLF REVHUYDn WLRQV RI /DQJHQHJJHU DQG 5RELQVRQ 7KH LUUHJXODU GLVVROXWLRQ NLQHWLFV IRU FRSSHU VHHQ LQ )LJXUHV DQG DUH EHOLHYHG WR EH GXH DW OHDVW LQ SDUW WR WKH HIIHFWV RI WKH PRUSKRORJ\ RI WKH VSRQJ\ GH]LQFLILHG VXUIDFH RQ WKH FRUURVLRQ SURFHVVHV $ FRPSDULVRQ RI WKH GDWD REWDLQHG IURP DOSKD DQG EHWD EUDVVHV FRQILUPV WKDW EHWD EUDVV FRUURGHV PXFK PRUH UDSLGO\ WKDQ GRHV DOSKD EUDVV LQ K\GURFKORULF DFLG

PAGE 93

7KH GH]LQFLILFDWLRQ IDFWRU LQWURGXFHG E\ 0DUVKDNRY DQG FRZRUNHUV SURYLGHV DQ LQGLFDWLRQ RI ZKHWKHU RU QRW GHn ]LQFLILFDWLRQ KDV RFFXUUHG +RZHYHU WKH IDFWRU = KDV QR IXQGDPHQWDO VLJQLILFDQFH DQG WKH DEVROXWH PDJQLWXGH RI WKH GH]LQFLILFDWLRQ IDFWRU PD\ YDU\ FRQVLGHUDEO\ IRU D JLYHQ H[SRVXUH FHOO DQG FDQQRW EH FRPSDUHG ZLWK WKH YDOXHV IURP RWKHU FHOOV 7KH VHQVLn WLYLW\ RI WKH GH]LQFLILFDWLRQ IDFWRU WR WKH LUUHJXODULWLHV RI WKH GLVVROXWLRQ NLQHWLFV RI FRSSHU GXULQJ GH]LQFLILFDWLRQ LV ZHOO LOOXVWUDWHG

PAGE 94

( /(&75&+(0&$/ ,19(67, *$7,216 (DUO\ HOHFWURFKHPLFDO WHVWV PHDVXUHG WKH SRWHQWLDO RI EUDVVHV LQ VROXWLRQV NQRZQ WR SURGXFH GH]LQFLILFDWLRQ $W WKDW WLPH LW ZDV QRW JHQHUDOO\ UHFRJQL]HG WKDW GXSOH[ DOOR\V VXFK DV DOSKDSOXVEHWD EUDVVHV ZRXOG H[KLELW PLFURVWUXFWXUHGHSHQGHQW HOHFWURFKHPLFDO EHKDYLRU ,Q -RVHSK DQG $UFHAAf VKRZHG WKDW WKH FRUURVLRQ EHKDYLRU RI D ZR &X ZR =Q EUDVV ZDV VWURQJO\ GHn SHQGHQW RQ VWUXFWXUH 7KLV SDUWLFXODU DOOR\ FDQ KDYH HLWKHU D VLQJOHSKDVH DOSKD RU D GXSOH[ DOSKDSL XVEHWD VWUXFWXUH GHSHQGLQJ RQ KHDW WUHDWPHQW DV UHYHDOHG LQ WKH FRSSHU]LQF SKDVH GLDJUDP )LJXUH 6HYHUDO UHFHQW HOHFn WURFKHPLFDO VWXGLHV RI GHDOOR\LQJ KDYH H[DPLQHG WKH EHKDYLRU RI VLQJOHSKDVH FRSSHU DOOR\V DQG RI PRUH FRPSOH[ FRSSHU DOn OR\V DQG KDYH WDNHQ PLFURVWUXFWXUH LQWR DFFRXQW"AAfAAA 6RPH UHVHDUFKHUV KDYH XVHG GULYHQ DQRGHV DV D UDSLG PHDQV RI SURGXFLQJ VSHFLPHQV WR EH H[DPLQHG E\ QRQHOHFWUR FKHPLFDO PHDQV2WKHUV KDYH XVHG HOHFWURFKHPLFDO REn VHUYDWLRQV WR DUULYH DW FRQFOXVLRQV DV WR WKH PHFKDQLVP RI GH]LQFLILFDWLRQAAAAM L f VWLOO RWKHUV KDYH WULHG WR GHILQH FXUUHQWfnf SRWHQWLDO-ffff‘f RU SRWHQWLDOS+ A A A A A rr A FRQGLWLRQV ZKHUH GHDOOR\LQJ

PAGE 95

PLJKW EH H[SHFWHG WR RFFXU RU WR GHWHUPLQH SRWHQWLDOV ZKHUH FDWKRGLF SURWHFWLRQ PD\ EH SRVVLEOH A +f (OHFWURFKHPLFDO WHVWV KDYH D QXPEHU RI GUDZEDFNV DQG PD\ EH VXEMHFW WR PLVLQWHUSUHWDWLRQ )RU H[DPSOH FRQVWDQW SRWHQWLDO WHVWV FRQGXFWHG IRU VKRUW SHULRGV HJ WZR KRXUV DW URRP WHPSHUDWXUH ZKLFK IDLO WR SURGXFH REVHUYDEOH GH ]LQFLILFDWLRQAA PD\ SURGXFH PHDVXUDEOH GH]LQFLILFDWLRQ DIWHU ORQJHU SHULRGV RI WLPH AfA!f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n QRW EH GHWHUPLQHG +RZHYHU WKLV SRWHQWLDO FKDQJH ZKLFK LV H[SUHVVHG E\ 2KPnV /DZ ( ,5 LV YHU\ VPDOO EHFDXVH WKH FXUUHQWV EHLQJ PHDVXUHG LQ D W\SLFDO HOHFWURFKHPLFDO FHOO DUH RI WKH RUGHU RI A WR A DPSHUHVAAA :LOGH DQG 7HWHULQAAA UHSRUWHG WKDW WKHLU DQRGLF SRODUL]DWLRQ FXUYHV IRU DOSKD EUDVVHV DQG GXSOH[ DOSKDSOXV EHWD EUDVVHV ZHUH DOPRVW LGHQWLFDO ZLWK WKH FXUYH IRU SXUH FRSSHU 7KH RQO\ PHDVXUDEOH GLIIHUHQFHV ZHUH LQ FXUUHQW GHQVLW\ 7KH\ DWWULEXWHG WKH VLPLODULWLHV WR D FRSSHU OD\HU

PAGE 96

ZKLFK IRUPHG RQ WKH VDPSOH VXUIDFHV 6LPLODU UHVXOWV ZHUH UHSRUWHG E\ 6XJDZDUD DQG (ELNRAAn IRU DOOR\V LQ WKH VDPH FRPSRVLWLRQ UDQJH $OOR\V KDYLQJ ZR ]LQF EHWD EUDVVf RU PRUH ZHUH UHSRUWHG DV KDYLQJ DQRGLF SRODUL]DWLRQ FXUYHV ZKLFK EHFDPH PRUH OLNH WKDW RI ]LQF DV WKH ]LQF FRPSRVLWLRQ LQFUHDVHG 0DUVKDNRY DQG FRZRUNHUVAA VWDWHG WKDW 7KH DQRGLF EHKDYLRU RI DOOR\V LV GHWHUPLQHG E\ WKH UDWH RI GLVVROXWLRQ RI WKH QREOH FRPSRQHQW DV WKH VORZHVW VWDJH 7KH\ LQYHVWLn JDWHG VLQJOHSKDVH DOSKD DQG EHWD EUDVVHV DV ZHOO DV GXSOH[ DOSKDSOXVEHWD EUDVVHV 3LFNHULQJ DQG %\UQH LQYHVWLJDWHG WKH HIIHFWV RI D GHDOOR\HG VSRQJH RQ WKH VXUIDFH RI D VDPSOH E\ KROGLQJ FRSSHUJROG DOOR\V DW GHDOOR\LQJ SRWHQWLDOV IRU YDU\LQJ SHULRGV RI WLPH DQG WKHQ MXPSLQJ WKH SRWHQWLDO WR D PRUH QREOH OHYHO 7KH\ VKRZHG WKDW EHORZ D FHUWDLQ FULWLFDO SRWHQWLDO WKH FRSSHUGLVVROXWLRQ FXUUHQW IURP FRSSHUJROG DOOR\V ZDV GHSHQGHQW RQ WKH UDWH RI FRSSHU VROLGVWDWH GLIn IXVLRQ IURP WKH DOOR\ $ERYH WKH FULWLFDO SRWHQWLDO WKH FRSSHUGLVVROXWLRQ FXUUHQW EHFDPH VWURQJO\ SRWHQWLDOGHSHQn GHQW 7KLV FULWLFDO SRWHQWLDO ZDV VKRZQ WR EH FRPSRVLWLRQ GHSHQGHQW IRU WKH FRSSHUJROG DOOR\ V\VWHP ZKLFK LV D KRPRn JHQHRXV DOOR\ V\VWHP RYHU WKH FRPSRVLWLRQ UDQJH LQYHVWLJDWHG DV ZHOO DV IRU FRSSHU]LQF DOOR\V ZKHUH SKDVH FKDQJHV FRXOG EH LQWHUSUHWHG DV FDXVLQJ VKLIWV LQ WKH FULWLFDO SRWHQWLDO

PAGE 97

6RPH RI WKH SRWHQWLDOV WKDW 3LFNHULQJ DQG %\UQH LQYHVn WLJDWHG ZHUH EHORZ WKH K\GURJHQ HYROXWLRQ SRWHQWLDOV IRU WKH HOHFWURO\WHV WKH\ XVHG %HFDXVH WKHUH ZDV QR DFFXUDWH ZD\ WR VXEWUDFW WKH K\GURJHQHYROXWLRQ FXUUHQW IURP WKH PHDVXUHG HOHFWURGH FXUUHQW WKH\ XVHG FKHPLFDO DQDO\VHV RI WKH PHWDOV LQ VROXWLRQ WR GHWHUPLQH GLVVROXWLRQ FXUUHQWV 7KLV DOVR DOORZHG WKHP WR SORW WKH GLVVROXWLRQ FXUUHQWV IRU HDFK GLVVROYLQJ VSHFLHV VHSDUDWHO\ DW SRWHQWLDOV ZKHUH WKH PRUH QREOH PHWDO LQ WKH DOOR\ ZDV DOVR GLVVROYLQJ 3RWHQWLDOYHUVXVS+ SORWV 3RXUEDL[ GLDJUDPVf RI UHn JLRQV RI FKHPLFDO VWDELOLW\ RI HOHPHQWV WKHLU LRQV DQG WKHLU VDOWV LQ DTXHRXV HQYLURQPHQWV SUHVHQW RQH SRVVLEOH ZD\ RI SURYLGLQJ D EDVLV IRU SUHGLFWLQJ WKH WHQGHQF\ IRU GHDOOR\LQJ /DWDQLVLRQ DQG 6WDHKOH A9HULQN DQG 3DUn ULVK DQG 9HULQN DQG +HLGHUVEDFKAA KDYH VXJJHVWHG WKDW WKH VXSHUSRVLWLRQ RI 3RXUEDL[ GLDJUDPVAfrA IRU WKH FRQVWLWXHQW HOHPHQWV RI WKH DOOR\ PD\ UHYHDO D UHJLRQ RI SRWHQWLDO DQG S+ ZKHUH RQH HOHPHQW ZRXOG WHQG WR FRUURGH ZKLOH WKH RWKHUVf ZURXOG EH LPPXQH /DWDQLVLRQ DQG 6WDHKOH ODWHU FRQFOXGHG WKDW WKHLU K\SRWKHWLFDO GHQLFNHOLILFDWLRQ RI )H1L&U DOOR\V ZDV QRW VXEVWDQWLDWHG E\ WKHLU H[SHULn PHQWDO REVHUYDWLRQVA +RZHYHU WKH FRSSHUQLFNHO V\VWHP VXJJHVWHG E\ 9HULQN DQG 3DUULVK DQG WKH FRSSHU]LQF V\VWHP GLVFXVVHG E\ 9HULQN DQG +HLGHUVEDFK ERWK DUH ZLGHO\ UHSRUWHG WR GHDOORY LQ VHUYLFHAfA]O]f

PAGE 98

$QRWKHU DGYDQWDJH RI WKH 3RXUEDL[ GLDJUDP RU SRWHQWLDO S+ DSSURDFK WR GHDOOR\LQJ LV WKDW LQIRUPDWLRQ REWDLQHG E\ YDULRXV DXWKRUV LQ GLIIHUHQW VROXWLRQV FDQ EH SORWWHG LQ D PDQQHU ZKLFK SURYLGHV PHDQLQJIXO FRUUHODWLRQV

PAGE 99

3UHVHQW :RUN 7KH H[SHULPHQWDO SRWHQWLDOS+ 3RXUEDL[f GLDJUDP IRU ZR &X ZR =Q DOSKD EUDVV LQ 0 FKORULGH VROXWLRQV LV VKRZQ LQ )LJXUH DQG ZDV GHWHUPLQHG E\ : & )RUW nnf ,,, DFFRUGLQJ WR WKH PHWKRG RI 3RXUEDL[ )LJXUH LV D VLPSOLILHG YHUVLRQ RI WKH HTXLOLEULXP GLDJUDP FDLn UQVnf FXODWHG E\ YDQ 0X\OGHU =RXERY DQG 3RXUEDL[A DVVXPLQJ D FKORULGH LRQ FRQFHQWUDWLRQ RI 0 DQG DOO RWKHU LRQLF VSHn FLHV WR EH SUHVHQW LQ FRQFHQWUDWLRQV )LJXUH LV D VLPLODU VLPSOLILHG GLDJUDP IRU WKH ]LQFA2 V\VWHPV )LJXUHV LQ WKH GLDJUDPV FRUUHVSRQG WR FDOFXODWLRQV ZKLFK DUH OLVWHG LQ $SSHQGL[ 6XSHUSRVLWLRQ RI GLDJUDPV )LJXUHV DQG JLYHV )LJXUH ,W LV HYLGHQW WKDW WKH H[SHULn PHQWDOO\ FRQVWUXFWHG GLDJUDP VKRZV D QXPEHU RI IHDWXUHV LQ FRPPRQ ZLWK WKH HTXLOLEULXP SRWHQWLDO YHUVXV S+f GLDJUDP IRU FRSSHU 9HULQN DQG +HLGHUVEDFK KDYH GLVFXVVHG WKHVH VLPLODULWLHV DW VRPH OHQJWK )RU SXUH FRSSHU LQ GHDHUDWHG VROXWLRQV WKH ]HUR FXUn UHQW SRWHQWLDO RQ WKH XSZDUG SRWHQWLDO VZHHS RI WKH HOHFWURn FKHPLFDO K\VWHUHVLV FLUFXLW LQGLFDWHV WKH SRVLWLRQ RI WKH VRFDOOHG LPPXQLW\ OLQH DW D JLYHQ S+ 7KH SRWHQWLDO DW ZKLFK WKLV ]HUR FXUUHQW LV REVHUYHG RIWHQ LV IDLUO\ FORVH WR WKH FDOFXODWHG SRVLWLRQ RI WKH PHWDOPHWDO LRQ FRH[LVn WHQFH SRWHQWLDO IRU D PHWDO LRQ FRQFHQWUDWLRQ RI A PRODU 7KXV WKH DUELWUDU\ FKRLFH RI A PRODU DV D GHILQLWLRQ RI

PAGE 100

327(17,$/ YV 6+( S+ )LJXUH ([SHULPHQWDO SRWHQWLDO YHUVXV S+ GLDJUDP IRU &X =Q LQ 0 &On DW r&f

PAGE 101

S27W+OnA ILJXUH 6LPSOLILHG &&,+2 GLDJUDP DW r& IRL VROXWLRQ FRQWDLQLQJ 0 FKORULGH LRQV FRQFHQWUDWLRQV RI LRQLF VSHFLHV fP nnHQ 5HI @

PAGE 102

327(17,$/ YV 6+-( )LJXUH 6LPSOLILHG =+&f GLDJUDP IRU FRQFHQWUDWLRQV RI LRQLF VSHFLHV fP

PAGE 103

327(17,$/ YV 6+( )LJXUH &X =Q DOOR\ LQ 0 FKORULGH VROXWLRQ 6XSHUSRVLWLRQ RI WKH H[SHULPHQWDO SRWHQWLDO GLDJUDP )LJXUH RQWR )LJXUHV DQG

PAGE 104

QRQFRUURVLRQ VHHPV DSSURSULDWH $OOR\V ZKLFK DUH ULFK LQ RQH FRPSRQHQW HJ &X=Q RU &X1Lf WHQG WR KDYH PDQ\ IHDWXUHV LQ FRPPRQ ZLWK WKH GLDJUDP IRU WKH PDMRU FRPSRQHQW LQ WKLV FDVH FRSSHUf 7KH LPPXQLW\ OLQH KRZHYHU DSSHDUV WR KDYH D VRPHZKDW GLIIHUHQW VLJQLILn FDQFH IRU DOOR\V WKDQ IRU SXUH PHWDOV :KLOH WKH NLQHWLFV RI DOOR\ GLVVROXWLRQ IRU DOSKD EUDVV DUH VORZ EHORZ WKH DOOR\ LPPXQLW\ OLQH WKLV GRHV QRW UXOH RXW WKH SRVVLELOn LW\ RI GHDOOR\LQJ 5HIHUULQJ WR )LJXUH WKH LPPXQLW\ OLQH IRU &X=Q LQ 0 FKORULGH ZDV DERXW 22229JAJ %HWZHHQ WKLV SRWHQWLDO DQG DSSUR[LPDWHO\ 9JAJ OLQH f WKHUH LV D WKHRUHWLFDO WHQGHQF\ IRU WKH VHOHFWLYH UHn PRYDO RI ]LQF IURP WKH DOOR\ $W SRWHQWLDOV PRUH SRVLWLYH WKDQ DERXW 22229JA ERWK FRQVWLWXHQWV RI WKH &X=Q DOOR\ JR LQWR VROXWLRQ f 3LFNHULQJ DQG %\UQHY UHSRUW WKDW WZR PRGHVRI GLVVROXn WLRQ PD\ RFFXU GHSHQGLQJ RQ WKH SRWHQWLDO SUHIHUHQWLDO DQG VLPXOWDQHRXV 3UHIHUHQWLDO GLVVROXWLRQ FKDQJHV JUDGXn DOO\ ZLWK SRWHQWLDO IURP YLUWXDOO\ QR GLVVROXWLRQ RI WKH PRUH QREOH PHWDO WR VLPXOWDQHRXV GLVVROXWLRQ RI ERWK FRPn SRQHQWV 3LFNHULQJ DQG %\UQH REVHUYHG WKLV FKDQJH LQ EHn KDYLRU EHWZHHQ 9&8& DQG 9&8& LQ 1DR6 VROX WLRQV DW S+ 7KHUH ZHUH QR FKORULGH LRQV LQ WKHLU VROXn WLRQV EXW WKH SRWHQWLDOS+ FRQGLWLRQV WKH\ GHVFULEHG ILW LQWR WKH FURVVKDWFKHG DUHD RQ )LJXUH

PAGE 105

,I WKH DERYHVWDWHG K\SRWKHVHV DUH FRUUHFW WKHQ EUDVVHV VKRXOG EH H[SHFWHG WR XQGHUJR GH]LQFLILFDWLRQ LQ DFLG VROXWLRQV DW SRWHQWLDOV IURP a 9f7UU WR 9FWUF U 6+( 6+( WKH UHJLRQ ZLWK VPDOO GRWV RQ )LJXUH f $W SRWHQWLDOV EHWZHHQ 22229JAJ DQG 9JA( WKH FURVVKDWFKHG UHJLRQ RI )LJXUH f FRSSHU DQG ]LQF VKRXOG GLVVROYH EXW QRW QHFHVVDULO\ LQ WKH VDPH UDWLR DV LQ WKH DOOR\ ,Q RUGHU WR WHVW WKHVH K\SRWKHVHV D VHULHV RI SRWHQ WLRVWDWLF WHVWV RI DOSKD EUDVV ZHUH FRQGXFWHG LQ EXIIHUHG 0 FKORULGH VROXWLRQV DW S+ VHH $SSHQGL[ IRU WKH H[DFW FRPSRVLWLRQ RI WKH VROXWLRQ XVHGf 7KH HTXLSPHQW IRU WKHVH WHVWV DQG WKH VDPSOH SUHSDUDn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n? ZKHUH 3LFNHULQJ DQG %\UQH UHSRUWHG GH]LQFLILFDWLRQ RI

PAGE 106

P UI 7DEOH $OSKD %UDVV 3RWHQWLRVWDWLF 7HVW 'DWD &RSSHU nRWHQWLDO HVLLH 'H]LQFLILHG 6XUIDFH $SSHDUDQFH :DV 6ROXWLRQ 6WLUUHG" 'XUDWLRQ RI ([SRVXUH =LQF LQ 6ROXWLRQ" &RSSHU LQ 6ROXWLRQ" RQ 3ODWLQXP (OHFWURGH" 5HPDUNV 1R
PAGE 107

DOSKD EUDVV E\ WKH VHOHFWLYH UHPRYDO RI ]LQF =LQF LRQV ZHUH SUHVHQW EXW QR FRSSHU ZDV GHWHFWHG DW WKHVH SRWHQWLDOV HLWKHU RQ WKH SODWLQXP DX[LOLDU\ HOHFWURGH RU E\ DWRPLF DEVRUSWLRQ DQDO\VLV RI WKH EXON VROXWLRQV &OHDUO\ VHOHFn WLYH GLVVROXWLRQ RI ]LQF LV WKH SUHGRPLQDQW SURFHVV LQ WKLV SRWHQWLDO UDQJH &HUWDLQ RI WKH VDPSOHV H[KLELWHG D GDUN WDUQLVK 94MMU DQG 9JASf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
PAGE 108

7DEOH %HWD %UDVV 3RWHQWLRVWDWLF 7HVW 'DWD &RSSHU 3RWHQWLDO 6, ,( 'H]LQFLILHG 6XUIDFH $SSHDUDQFH" :DV 6ROXWLRQ 6WLUUHG" 'XUDWLRQ RI ([SRVXUH = LQF MQ 6ROXWLRQ" &RSQFU LQ 6ROXWLRQ" RQ 3ODWLQXP (OHFWURGH" 5HPDUNV
PAGE 109

EUDVV ,Q XQVWLUUHG FHOOV WKH FRSSHU DSSHDUHG WR KDYH UHn PDLQHG DW RU QHDU WKH VDPSOH VXUIDFH VLQFH QR FRSSHU ZDV GHWHFWHG LQ WKH EXON VROXWLRQV DIWHU DQ\ RI WKH SRWHQWLR VWDWLF EHWD EUDVV WHVWV )LJXUH VKRZV WKH VXUIDFH RI D VDPSOH RI EHWD EUDVV ZKLFK ZDV KHOG LQ D VDPSOH KROGHU VXFK DV LV VKRZQ LQ )LJXUH 7KH VDPSOH ZDV H[SRVHG IRU D VKRUWWHUP WHVW RI KRXUV DW 9JAJ 7KH UHGGLVK FRSSHU FUHVFHQW LQ WKH XSSHU OHIW RI WKH SKRWRJUDSK LOOXVWUDWHV WKH HIIHFWV RI VWLUULQJ RQ GH]LQFLILFDWLRQ 7KDW SRUWLRQ RI WKH VDPSOH VXUn IDFH ZKLFK ZDV VKHOWHUHG IURP VWLUULQJ E\ WKH FRQILJXUDWLRQ RI WKH VDPSOH KROGHU LH D UHODWLYHO\ VWDJQDQW DUHDf KDV D GH]LQFLILHG DSSHDUDQFH GXH WR WKH GHSRVLWLRQ RI FRSSHU :KHUH VWLUULQJ ZDV HIIHFWLYH WKH EDODQFH RI WKH VXUIDFHf GHWLQFLILFDWLRQ GLG QRW RFFXU 7KH DERYH SKRWRJUDSK FRXSOHG ZLWK WKH REVHUYDWLRQV RQ DOSKD EUDVV GLVFXVVHG DERYH FDQ EH WDNHQ DV IXUWKHU HYLGHQFH WKDW GH]LQFLILFDWLRQ FDQ RFFXU E\ DQ HOHFWURGHSRVLWLRQ PHFKDn QLVP LQ FHUWDLQ SRWHQWLDO UDQJHV )LJXUH VKRZV WKH FRQILJXUDWLRQ RI D W\SLFDO SRWHQ WLRNLQHWLF VFDQ IRU DOSKD EUDVV 3ORWV RI WKH ]HUR FXUUHQW SRWHQWLDO RQ WKH UHWXUQ GRZQZDUGf VFDQ ( DUH VKRZQ E\ WKH OLQH QXPEHUHG LQ )LJXUH DQG LQ $SSHQGL[ $W WKLV SRWHQWLDO RQ GRZQZDUG SRWHQWLRNLQHWLF VFDQV WKH VDPSOHV ZHUH REVHUYHG WR FKDQJH WR D UHGGLVK FRORU

PAGE 110

)LJXUH %HWD EUDVV KHOG DW 9J+( IRU KRXUV &UHVFHQW VKDSHG UHGGLVK UHJLRQ VKRZV HIIHFW RI VWLUULQJ DFWLRQ RQ GH]LQFLILFDWLRQ

PAGE 111

)LJXUH 7\SLFDO SRWHQWLRNLQHWLF VFDQ LQ DFLG VROXWLRQV (S LV IRXQG WR RFFXU QHDU 9VK( DQG EHWZHHQ WKH ]HURFXUUHQW SRWHQWLDO DQG WKH ILUVW PD[LPXP

PAGE 112

W\SLFDO RI 9VKH FRSSHU GHSRVLWV DQG FRUUHVSRQGV 7KLV WR WKH SRWHQWLDO LV DSSUR[LPDWHO\ SRWHQWLDO IRU WKH UHDFWLRQ &X & &X&O H $V D FRQVHTXHQFH GHSRVL DW SRWHQWLDOV EHORZ XVHG FRSSHU FKORULGH VROX RQ UHDFWLRQV $ UHFHQW FDWHV WKDW QLFNHO FKORULGH WLQJ HIIHFW RQ GH]LQFLIL WKH DFFHOHUDWLRQ RI GH]LQFLIL FRXOG EH H[SODLQHG E\ WKH VHV LPPHUVHG LQ WKHP LRQV ZKLFK KDG EHHQ UHn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f $ QXPEHU RI UHVHDUFKHUV KDYH WLRQV WR DFFHOHUDWH GH]LQFLILFDWL SDSHU E\ )DOOHLUR DQG 3LHVNH LQGL VROXWLRQV KDYH D VLPLODU DFFHOHUD FDWLRQ7KLV VXJJHVWHG WKDW IW FDWLRQ FDXVHG E\ WKHVH VROXWLRQV IUHH FRUURVLRQ SRWHQWLDOV RI EUDV 7HVWV ZHUH UXQ LQ IRXU VROXW

PAGE 113

7DEOH $OSKD %UDVV )UHH &RUURVLRQ 3RWHQWLDO 7HVWV 6ROXWLRQ )LQDO 3RWHQWLDO 'H]LQFLILFDWLRQ" 5HIHUHQFH $XWKRUV J= 0&, J= &X&O 9VKH 1R /DQJHQHJJHU DQG 5RELQVRQAAf J= +& J= &X&O 69V+( 1R /DQJHQHJJHU DQG 5RELQVRQAA 0 +6 9J((
PAGE 114

%HWD %UDVV )UHH 6ROXWLRQ )LQDO 3RWHQWLDO JO +& JO &X&O 9J+( JO +& JO &X&O RLRRY6+( 0 +6 9J+( JO 1D&O JO 1L&O+ RLRRY6+H 7DEOH &RUURVLRQ 3RWHQWLDO 7HVWV 'H]LQFLILFDWLRQ" 5HIHUHQFH $XWKRUV
PAGE 115

DOO LQVWDQFHV WKLV SRWHQWLDO ZDV UHDFKHG ZLWKLQ RQH KRXU RI WKH VWDUW RI WKH WHVW 7KH WZR VDPSOHV ZKLFK GLG QRW GH]LQFLI\ ZHUH ERWK DOSKD EUDVVHV ZKLFK FDPH WR SRWHQWLDOV RI 9JMMJ DQG 9JAJ 7KLV LV LQ WKH SRWHQWLDO UDQJH ZKHUH VWLUULQJ ZDV VKRZQ WR SUHYHQW GHSRVLWLRQ RI FRSSHU RQ DOSKD EUDVV VSHFLPHQV

PAGE 116

',6&866,21 7KH WZR PDLQ WKHRULHV ZKLFK KDYH EHHQ SURSRVHG IRU GH]LQFLILFDWLRQ KDYH RIWHQ EHHQ FRQVLGHUHG WR EH PXWXDOO\ H[FOXVLYH $XWKRUV FODLPLQJ WKDW WKH\ KDYH IRXQG VXSSRUWLQJ HYLGHQFH IRU RQH RU WKH RWKHU PHFKDQLVP KDYH FODLPHG ZLWKRXW IXUWKHU MXVWLILFDWLRQ WKDW WKH RWKHU FRQFOXVLRQ ZDV XQZDUn UDQWHG (YLGHQFH KDV EHHQ SUHVHQWHG KHUHLQ ZKLFK LQGLFDWHV WKDW HDFK DQG VRPHWLPHV ERWKf PHFKDQLVP FDQ RFFXU ;UD\ DQDO\n VLV DQG HOHFWURQ PLFURSUREH GDWD ZKLFK VXSSRUW D VHOHFWLYH UHPRYDO H[SODQDWLRQ RI WKH GHDOORYLQJ SURFHVV ZHUH REWDLQHG IURP VRPH RI WKH VDPH VDPSOHV ZKLFK VKRZ FRSSHU GHSRVLWV )LJXUH VKRZV D GLIIXVLRQ ]RQH LQ DOSKD EUDVV ZKLFK KDG EHHQ H[SRVHG LQ K\GURFKORULF DFLG )LJXUH LV D SKRWRn PLFURJUDSK RI WKLV VDPSOH DQG VKRZV WKH SUREH WUDFH LQ WKH FHQWHU RI WKH SLFWXUH )LJXUH DOVR VKRZV D FRSSHU GHSRVn LW DW WKH WRS RI WKH SLFWXUH WKXV LOOXVWUDWLQJ WKDW ERWK PHFKDQLVPV FDQ RFFXU RQ WKH VDPH SLHFH RI PHWDO 7KH WZR VDPSOHV ZKLFK SURGXFHG WKH [UD\ SDWWHUQV LQ )LJXUHV DQG DOVR KDG GHSRVLWV RQ WKH VXUIDFH DQG SURGXFHG HOHFWURQ PLFURSUREH GDWD LQGLFDWLQJ WKH H[LVWHQFH RI GLIIXVLRQ ]RQHV

PAGE 117

7KH GLIIXVLRQ GLVWDQFHnV LQ )LJXUHV DQG DUH VPDOO EXW DUH VWLOO ZLGH HQRXJK WR UHSUHVHQW VHYHUDO WHQV RI WKRXn VDQGV RI LQWHUDWRPLF GLVWDQFHV DQG WKXV WKH\ FDQ UHSUHVHQW VLJQLILFDQW GLIIXVLRQ GLVWDQFHV 3LFNHULQJ SRLQWV RXW WKDW WKH GLIIXVLRQ LQWHUIDFH LV QRW IODW EXW EHFRPHV URXJK RU ULSSOHG$OWKRXJK WKLV LUUHJXODU LQWHUIDFH XVXDOO\ LV WRR VPDOO WR EH VHHQ PHWDOORJUDSKLFDOO\ LQ WKH FDVH RI nnf GHDOOR\LQJ LQ DTXHRXV HQYLURQPHQWV ` WKH VDPH HIIHFW KDV EHHQ QRWHG RQ VDPSOHV VXEMHFWHG WR OLTXLGPHWDO FRUURVLRQ ZKHUH WKH SHUWXUEDWLRQV DUH ODUJH HQRXJK WR EH VHHQ PHWDOn ORJUDSKLFDOO\ 7KLV URXJKHQLQJ UHVXOWV LQ D VKRUWHU DYHUDJH GLIIXVLRQ SDWK OHQJWK WKDQ ZRXOG EH DYDLODEOH ZLWK D IODW LQWHUIDFH 7KH GLVWDQFHV ZKLFK DUH GHWHFWHG ZLWK WKH PLFURSUREH DUH WKXV SUREDEO\ ODUJHU WKDQ WKH DFWXDO GLIIXn VLRQ SDWK DQG UHSUHVHQW WKH ZLGWK RI WKLV URXJKHQHG DUHD UDWKHU WKDQ WKH OHQJWK RI WKH GLIIXVLRQ SDWK LWVHOI 1RQHn WKHOHVV WKH QHW GLIIXVLRQ ]RQH VWLOO LV ODUJH HQRXJK WR EH REVHUYHG ZLWK FRQILGHQFH XVLQJ WKH HOHFWURQ PLFURSUREH 7KH HOHFWURGHSRVLWLRQ RI FRSSHU IURP VROXWLRQV KDYLQJ D VLJQLILFDQW FRSSHU LRQ FRQFHQWUDWLRQ FDQ EH H[SODLQHG E\ WKH 1HUQVW HTXDWLRQ ( 57 Q ORJ D&X&Of D&XDOOR\f (T f ZKLFK DW URRP WHPSHUDWXUH EHFRPHV r7KH VLJQ FRQYHQWLRQ XVHG KHUHLQ FRQIRUPV LQ FDOFXODWLRQ RI 3RXUEDL[ GLDJUDPV 6HH (T WR WKDW XVHG $SSHQGL[

PAGE 118

F FR & ‘ ( ( f§ c ORJ n&X&O D&XDOOR\fD&OA IRU WKH UHDFWLRQ (T f &X & &X&O H (T f 7KH WHUPV RI WKLV HTXDWLRQ FDQ EH VHSDUDWHG ( ( ORJ D&X& ORJ D&XDOOR\fDFrf 9DULDWLRQV LQ HLWKHU WKH &X&O RU WKH &XDOOR\f DFWLYLW\ ZLOO FDXVH D FKDQJH LQ WKH HTXLOLEULXP SRWHQWLDO IRU WKH UHDFWLRQ ,Q RWKHU ZRUGV LQFUHDVHV LQ WKH DFWLYLW\ RI WKH FRPSOH[ FRSSHU LRQ ZLOO VKLIW WKH HTXLOLEULXP LQ WKH QREOH GLUHFWLRQ 7KH H[LVWHQFH RI D FRSSHU DOOR\ KDYLQJ D KLJKHU FRSSHU DFWLYLW\ WKDQ WKDW RI WKH RULJLQDO DOOR\ ZLOO ORZHU WKH SRWHQWLDO $FFRUGLQJO\ FRSSHU LV H[SHFWHG WR EH HOHF WURGHSRVLWHG IURP VROXWLRQV DW D JLYHQ SRWHQWLDO SURYLGHG WKH FRQGLWLRQV RI (T DUH IXOILOOHG 7KH ORJ DFXD@L\f WHUP YDULHV RQO\ VOLJKWO\ IRU FRPSRVLWLRQV RI IURP SHUFHQW FRSSHU (TXDWLRQ LPSOLHV WKDW FRSSHU ZLOO EH GHSRVLWHG IURP XQVWLUUHG 0 FKORULGH VROXWLRQV DW SRWHQWLDOV EHORZ 9 6+( 6FDQQLQJ LQ WKH DFWLYH GLUHFWLRQ IURP SRWHQn WLDOV PRUH QREOH WKDQ 9&/,7U ZLOO UHVXOW LQ WKH RQVHW RI FRSSHU GHSRVLWLRQ DW WKLV SRWHQWLDO 7KXV FRSSHU GHSRn VLWLRQ LV WR EH DQWLFLSDWHG IURP VROXWLRQV FRQWDLQLQJ FRSn SHU LRQV HLWKHU DV FRUURVLRQ SURGXFWV RU DV DGGHG LRQVf DW SRWHQWLDOV EHORZ 9FAS LQ 0 FKORULGH VROXWLRQV

PAGE 119

7KH DSSUR[LPDWH FRSSHU DQG ]LQF FRQFHQWUDWLRQV ZKLFK ZHUH DWWDLQHG LQ WKH WKUHH FHOOV LQ ZKLFK FRSSHU GHSRVLWV ZHUH REVHUYHG RQ DOSKD EUDVV DUH OLVWHG LQ 7DEOH 7KH GH]LQFLILFDWLRQ IDFWRUV = YDU\ IURP IRU r& WR IRU r& $OO WKHVH YDOXHV DUH VXEVWDQWLDOO\ KLJKHU WKDQ WKH YDOXHV VOLJKWO\ DERYH XQLW\f UHSRUWHG E\ 0DUVKDNRY DQG FRZRUNHUV DQG LQ WKLV UHSRUW IRU WKH GH]LQFLILFDWLRQ RI DOSKD EUDVVHV 7KH IDFW WKDW VWDJQDQW FRQGLWLRQV FDQ OHDG WR GH]LQFLILn FDWLRQ LV LOOXVWUDWHG LQ )LJXUH ZKHUH KROLGD\V LQ WKH VWRSRII ODFTXHU XVHG WR FRDW DOO EXW D FHUWDLQ SRUWLRQ RI D VDPSOH FDXVHG GH]LQFLILFDWLRQ LQ DUHDV XQGHU WKH KROLGD\V ZKLOH WKH XQFRDWHG DUHD VXIIHUHG JHQHUDO GLVVROXWLRQ 7KUHH H[SODQDWLRQV DUH SRVVLEOH 7KH FRSSHU LRQ FRQFHQWUDWLRQ PD\ VDWLVI\ FRQGLWLRQV QHFHVVDU\ IRU GHSRVLWLRQ DFFRUGn LQJ WR WKH 1HUQVW HTXDWLRQ 7KH FRQFHQWUDWLRQ RI FRSSHU LQ ORFDOL]HG DUHDV RI UHVWULFWHG IORZ FRXOG H[FHHG VROXELOLW\ OLPLWV DQG FDXVH SUHFLSLWDWLRQ RI FRSSHU 7KH ORFDO HOHFWURFKHPLFDO SRWHQWLDO RI WKH EUDVV VXUIDFH DW WKH FRUURVLRQ LQWHUn IDFH FRXOG KDYH EHHQ DOWHUHG GXH WR ORFDOn L]HG GLIIHUHQFHV LQ WKH HOHFWURO\WH ZLWKLQ WKH IORZUHVWULFWHG DUHD

PAGE 120

7DEOH $SSUR[LPDWH &RSSHU DQG =LQF &RQFHQWUDWLRQV LQ 6ROXWLRQV :KHUH &RSSHU 'HSRVLWV :HUH 2EVHUYHG RQ $OSKD %UDVV 7HPSHUDWXUH &RSSHU =LQF RI FHOO SSPf SSPf r & r& r&

PAGE 121

)LJXUH &URVVVHFWLRQ RI DOSKD EUDVV VKRZLQJ ZKHUH KROLGD\V LQ WKH VWRSRII ODFTXHU FDXVHG GH]LQFLILFDWLRQ LQ WKH VWDJQDQW UHJLRQV EHQHDWK WKH KROLGD\V 7KH XQPDVNHG UHJLRQ KDG QR IORZ UHVWULFWLRQV DQG VXIIHUHG JHQn HUDO GLVVROXWLRQ ;

PAGE 122

,OO 7KH KLJK GH]LQFLILFDWLRQ IDFWRUV IRU EHWD EUDVV XQGHU IUHHFRUURVLRQ FRQGLWLRQV FDQ EH H[SODLQHG DV GXH WR WKH HOHFWURFKHPLFDO SRWHQWLDO RI EHWD EUDVV UHODWLYH WR WKDW RI DOSKD EUDVV XQGHU FRQGLWLRQV ZKHUH WKHVH EUDVVHV XQGHUJR GLVVROXWLRQ 7DQDEHArnA DQG 6XJDZDUD DQG (ELNRAA UHSRUW WKDW GHDOOR\HG FRSSHU DOOR\V TXLFNO\ EHFRPH FRYHUHG ZLWK D VXUIDFH OD\HU RI FRSSHU ZKLFK LI WKH RKPLF UHVLVWDQFH RI WKH OD\HU LV ORZ SUHVHQWV D FRSSHU VXUIDFH WR WKH EXON VROXn WLRQ ZKLFK LV DW WKH IUHH FRUURVLRQ SRWHQWLDO RI WKH DOOR\ Q f 7KLV SRWHQWLDO DFFRUGLQJ WR WKH GDWD RI :LOGH DQG 7HWHULQnf LV ORZHU IRU EHWD EUDVV WKDQ LW LV IRU DOSKD EUDVV 7KXV WKH SRWHQWLDO RU GULYLQJ IRUFH GLIIHUHQFH EHWZHHQ WKH GHSRVLn WLRQ SRWHQWLDO RI FRSSHU IURP FRSSHU FKORULGH VROXWLRQV DSSUR[LPDWHO\ 9JMA IRU 0 FKORULGH VROXWLRQVf DQG WKH SRWHQWLDO RI WKH FRSSHU VXUIDFH OD\HU RQ WKH EUDVV VXUn IDFH LV JUHDWHU IRU EHWD EUDVV WKDQ LW LV IRU DOSKD EUDVV 6XJDZDUD DQG (ELNRAAf SRLQW RXW WKH GLIILFXOW\ RI GHWHUPLQLQJ WKH SRWHQWLDO RI D GHDOOR\HG PHWDO DW WKH FRUURn VLRQ LQWHUIDFH ZKLFK LV EHKLQG D UHVLVWDQFH EDUULHU GXH WF VWDJQDQW VROXWLRQ FRQGLWLRQV DQG WKH SUHVHQFH RI WKH GHn DOOR\HG OD\HU 1RQHWKHOHVV WKH FRQFHSW RI D 3RXUEDL[ GLDn JUDP IRU EUDVVHV VHHPV WR H[SODLQ H[SHULPHQWDO REVHUYDWLRQV DQG SURYLGHV D EDVLV IRU SUHGLFWLQJ FRQGLWLRQV XQGHU ZKLFK GHDOOR\LQJ LV OLNHO\ WR RFFXU 7KH 3RXUEDL[ GLDJUDP DSSURDFK DOVR H[SODLQV WKH HIIHFWV RI VWLUULQJ RQ GHDOOR\LQJ /D4XH KDV REVHUYHG LQ VDOW ZDWHU

PAGE 123

H[SRVXUH WHVWV RI VSLQQLQJ GLVFV RI FRSSHU DOOR\V WKDW GHn DOOR\LQJ ZKHQ LW RFFXUUHGf ZDV DOZD\V REVHUYHG QHDU WKH FHQWHU RI WKH GLVFV ZnKHUH HIIHFWLYH ZDWHU YHORFLW\ ZDV WKH ORZHVW 7KH RXWHU SRUWLRQV RI WKH VDPH GLVFV ZHUH REn VHUYHG WR KDYH XQGHUJRQH JHQHUDO FRUURVLRQ $OWKRXJK WKH HOHFWURFKHPLFDO SRWHQWLDO ZDV QRW UHSRUWHG LW LV OLNHO\ WKDW LW ZDV LQ WKH UHJLRQ LQGLFDWHG E\ FURVVKDWFKLQJ RQ )LJXUH &RPSRQHQWV H[SRVHG WR FRQGLWLRQV LQYROYLQJ KLJK UHODn WLYH PRWLRQ EHWZHHQ WKH SDUW DQG WKH HOHFWURO\WH VXFK DV VKLS SURSHOOHUV DQG SXPS LPSHOOHUVKDYH DOVR EHHQ REVHUYHG WR XQGHUJR GHDOOR\LQJ 8QGHU WKHVH FLUFXPVWDQFHV LW ZDV OLNHO\ WKDW WKH SRWHQWLDO ZDV LQ WKH UHJLRQ LQGLFDWHG E\ VPDOO GRWV LQ )LJXUH ([DPLQDWLRQ RI WKH 1HUQVW HTXDWLRQ DV ZULWWHQ IRU WKH GLVVROXWLRQ RI D GLYDOHQW PHWDO LRQ RI D K\SRWKHWLFDO PHWDO $ F R 57 + ( f§ ORJ D$ 57 ORJ D$ (T f UHYHDOV WKDW WKH WHUP UHODWLQJ SHU LQ WKH DOOR\ LV DOPRVW QHJ IURP SHUFHQW $ DFWLYLW WR FRQFHQWUDWLRQf 7KLV LV LO HTXLOLEULXP SRWHQWLDOV IRU WKH WR WKH DFWL YLW\ RI WKH FRSn OLJLEOH IRU FRQFHQW UDWL RQV RI LHV DVV XPHG WR EH 3L RSRUWLRQDO OXVWUDW HG E\ D SORW RI WKH HTXDWL RQV $ $

PAGE 124

DQG % % Ha RI D K\SRWKHWLFDO ELQDU\ DOOR\ VKRZQ LQ )LJXUH 7KH FRPn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n FRUURVLRQ IURP D FRUURVLRQ UHJLRQ ZRXOG QRW EH GLVSODFHG UDGLFDOO\ E\ DOOR\LQJ 7KLV LV FRQWUDU\ WR H[SHULPHQWDO REVHUYDWLRQV PDGH GXULQJ WKLV VWXG\ DQG WR WKH ZRUN RI RWKHUV 7KH H[SHULPHQWDO REVHUYDWLRQV RI 3LFNHULQJ DQG %\UQH LQGLFDWH WKDW FRSSHU GLVVROYHV IURP FRSSHUJROG DOOR\V DW D GLIIHUHQW SRWHQWLDO IRU FRSSHU SHUFHQW JROG WKDQ LW GRHV IRU FRSSHU SHUFHQW JROG 7KLV FRQILUPV HDUOLHU nnf UHVXOWVY E\ WKH VDPH DXWKRUV RQ WKH GLVVROXWLRQ RI ]LQF IURP HSVLORQ JDPPD DQG DOSKD EUDVVHV &OHDUO\ WKH 1HUQVW HTXDWLRQ FDQQRW H[SODLQ WKLV DVSHFW RI GHDOOR\LQJ 7KH HTXLOLEULXP SRWHQWLDOS+ GLDJUDPV IRU FRSSHU DQG ]LQF )LJXUHV

PAGE 125

:(,*+7 3(5&(17 % )LJXUH 7KHRUHWLFDO GRPDLQV IRU GHDOOR\LQJ LQ D JLYHQ VROXWLRQ EDVHG XSRQ WKH 1HUQVW HTXDWLRQ

PAGE 126

DQG ZHUH FDOFXODWHG XVLQJ WKH 1HUQVW HTXDWLRQ IRU WKH LPPXQLW\ OLQHV 3UHOLPLQDU\ LQYHVWLJDWLRQV LQWR WKH H[n SHULPHQWDO 3RXUEDL[ GLDJUDP IRU SXUH QLFNHO UHYHDO WKDW WKHUH DUH VLJQLILFDQW GLIIHUHQFHV EHWZHHQ WKH H[SHULPHQWDO DQG WKH QRQ L R  K RUHWLFDO GLDJUDPV IRU QLFNHO 7KH VDPH PD\ KROG WUXH IRU ]LQF DQG JROG DOWKRXJK WKH H[SHULPHQWDO FRSSHU GLDn JUDP LQ FKORULGH VROXWLRQV KDV EHHQ GHWHUPLQHG DQG DJUHHV U L R S FORVHO\ ZLWK PDQ\ IHDWXUHV RI WKH WKHRUHWLFDO GLDJUDP f f 2QH SRVVLELOLW\ IRU WKH IDLOXUH RI WKH 1HUQVW HTXDWLRQ WR H[SODLQ WKH H[SHULPHQWDO REVHUYDWLRQV GLVFXVVHG DERYH LV WKDW WKH FDOFXODWLRQV ZHUH PDGH DVVXPLQJ WKHDFWLYLWLHV RI WKH HOHPHQWV LQ WKH DOOR\ ZHUH GLUHFWO\ SURSRUWLRQDO WR WKH FRPSRVLWLRQ 7KLV DSSUR[LPDWLRQ PD\ QRW EH MXVWLILHG 7KH DFWLYLWLHV RI FRQVWLWXHQWV RI DOOR\V XVXDOO\ DUH GHWHUPLQHG DW KLJK WHPSHUDWXUHV DQG PD\ QRW EH WKH VDPH DW ORZ WHPSHU UJJ A "f DWXUHVnn f M +RZHYHU UDGLFDO GHSDUWXUHV IURP LGHDOn LW\ DUH VHOGRP HQFRXQWHUHG LQ DOOR\ V\VWHPV DQG WKHVH GHn SDUWXUHV ZRXOG KDYH WR DOWHU WKH DFWLYLWLHV E\ VHYHUDO RUGHUV RI PDJQLWXGH EHIRUH WKH\ ZRXOG DOWHU WKH SULQFLSDO SRLQWV RI WKH DERYH DUJXPHQW

PAGE 127

&21&/86,216 2Q WKH EDVLV RI DQ DQDO\VLV RI WKH GH]LQFLILFDWLRQ RI DOSKD DQG EHWD EUDVVHV XQGHU IUHH FRUURVLRQ FRQGLWLRQV DQG RQ WKH EDVLV RI HOHFWURFKHPLFDO EHKDYLRU RI WKHVH DOOR\V LW PD\ EH FRQFOXGHG WKDW ;UD\ GLIIUDFWLRQ DQG HOHFWURQ PLFURSUREH GDWD LQGLFDWH WKDW D VHOHFWLYH UHPRYDO SURFHVV IRU WKH GH]LQFLIL FDWLRQ RI EUDVVHV LV RSHUDWLYH DW OHDVW XQGHU FHUWDLQ FRQn GLWLRQV RI SRWHQWLDO DQG S+ 2SWLFDO HYLGHQFH DQG HOHFWURFKHPLFDO HYLGHQFH LQGLn FDWH WKDW D UHGHSRVLWLRQ PHFKDQLVP IRU GH]LQFLILFDWLRQ FDQ RFFXU XQGHU VXLWDEOH FRQGLWLRQV ZLWKLQ D SDUWLFXODU UDQJH RI SRWHQWLDO DQG S+ %RWK RSHUDWLYH PHFKDQLVPV FDQ EH REVHUYHG WR KDYH WDNHQ SODFH RQ WKH VDPH VSHFLPHQ 7KLV FDQ EH H[SODLQHG LQ VHYHUDO ZD\V Df 7KH VHOHFWLYHUHPRYDO SURFHVV PD\ DFWXDOO\ EH RQH RI UHODWLYH NLQHWLFV LH ERWK PHWDOV XQGHUn JR GLVVROXWLRQ EXW WKH ]LQF GLVVROYHV PXFK IDVWHU LQ D FHUWDLQ ZHOOGHILQHG UDQJH RI SRWHQWLDO DQG S+ WKDQ FRSSHU Ef &RQGLWLRQV DW WKH PHWDOVROXWLRQ LQWHUIDFH DUH DOWHUHG E\ WKH SUHVHQFH RI PHWDO LRQV LQ VROXWLRQ

PAGE 128

nr f I &RSSHU HQWHUV VROXWLRQ GXULQJ WKH GH]LQFLILFDWLRQ RI ERWK DOSKD DQG EHWD EUDVVHV XQGHU IUHHO\ FRUURGLQJ FRQn GLWLRQV QR HOHFWURFKHPLFDO VWLPXODWLRQf LQ 1 +& VROXWLRQ /LQHDU GLVVROXWLRQ NLQHWLFV ZHUH REVHUYHG IRU WKH GLVVROXWLRQ RI ]LQF IURP ERWK DOSKD DQG EHWD EUDVVHV IRU WHVWV UDQJLQJ XS WR KRXUV DW WHPSHUDWXUHV XS WR DSSUR[Ln PDWHO\ rF 7KH GLVVROXWLRQ NLQHWLFV IRU FRSSHU DUH LUUHJXODU EDVHG RQ DWRPLF DEVRUSWLRQ GDWD 7KLV LUUHJXODULW\ LV EHn OLHYHG WR EH FDXVHG DW OHDVW LQ SDUW E\ WKH PRUSKRORJ\ RI WKH UHVLGXDO FRSSHU OD\HU 7KH DFWLYDWLRQ HQHUJ\ PHDVXUHG IRU WKH GLVVROXWLRQ RI ]LQF IURP EHWD EUDVV GXULQJ GH]LQFLILFDWLRQ ZDV DSSUR[L B PDWHO\ .FDO JPPROH 7KLV YDOXH LV VRPHZKDW KLJKHU WKDQ DFWLYDWLRQ HQHUJLHV IRU HOHFWURFKHPLFDO GLVVROXWLRQ RI PHWDOV DSSUR[LPDWHO\ .FDO JPPROH %HFDXVH RI WKH VHHPLQJ LUUHJXODULW\ RI FRSSHU GLVn VROXWLRQ NLQHWLFV LW LV FRQVLGHUHG PHDQLQJOHVV WR FDOFXODWH DQ DFWLYDWLRQ HQHUJ\ IRU FRSSHU IURP DWRPLF DEVRUSWLRQ GDWD ,W LV SRVVLEOH WR SUHGLFW GRPDLQV RI SRWHQWLDO DQG S+ IRU HDFK RI WKH RSHUDWLYH PHFKDQLVPV RI GH]LQFLILFDWLRQ E\ VXSHUSRVLWLRQ RI H[SHULPHQWDO SRWHQWLDO YHUVXV S+ GLDn JUDPV RYHU WKH HTXLOLEULXP 3RXUEDL[ GLDJUDPV IRU WKH FRQn VWLWXHQW PHWDOV RI WKH DOOR\

PAGE 129

7KH UHVXOWLQJ FRPSRVLWH GLDJUDP SURYLGHV D EDVLV IRU XQGHUVWDQGLQJ WKH LQIOXHQFH RI VWLUULQJ DQG KLJK IORZ UDWHV RQ WKH GH]LQFLILFDWLRQ SURFHVV 7KH FRPSRVLWH GLDJUDP DOVR SURYLGHV DQ H[SODQDWLRQ IRU ZK\ WHVWV LQ FRSSHU FKORULGH VROXWLRQV SURGXFH DFFHOHUn DWHG GH]LQFLILFDWLRQ 7KH XVH RI FRSSHU FKORULGH VROXWLRQV WR WHVW FRSSH DOOR\V IRU VXVFHSWLELOLW\ WR GHDOOR\ LV D ZHOOHVWDEOLVKHG SUDFWLFH ,W LV QRW YDOLG WR PDNH FRQFOXVLRQV RQ WKH PHFKD QLVP RI GH]LQFLILFDWLRQ EDVHG VROHO\ RQ WHVWV LQ WKHVH VROX WLRQV

PAGE 130

5(&200(1'$7,216 )25 )857+(5 5(6($5&+ 7KH UHVXOWV RI WKLV LQYHVWLJDWLRQ UHYHDO D QXPEHU RI DUHDV RI UHVHDUFK ZKLFK VKRXOG EH H[SORUHG ERWK IRU WKHLU VFLHQWLILF SRVVLELOLWLHV DQG EHFDXVH RI WKHLU SRVVLEOH HFRQRPLF EHQHILWV 7KH GHYHORSPHQW RI H[SHULPHQWDO 3RXUEDL[ GLDJUDPV VKRXOG EH H[WHQGHG WR FRYHU DOO FRPPHUFLDO DQG SRWHQWLDOO\ FRPPHUn FLDO DOOR\ V\VWHPV 3RXUEDL[ GLDJUDPV DW HOHYDWHG WHPSHUn DWXUHV DUH DOVR QHHGHG 3UHOLPLQDU\ LQYHVWLJDWLRQV ZLWK VLQJOHSKDVH EHWD EUDVVHV UHYHDO WKDW WKH HOHFWURFKHPLFDO K\VWHUHVLV WHFKQLTXH PXVW EH DXJPHQWHG E\ SRWHQWLRVWDWLF WHVWLQJ LQ FRQMXQFWLRQ ZLWK VROXWLRQ DQDO\VLV LQYHVWLJDWLRQV ,GHQWLILFDWLRQ RI LQ VLWX UHDFWLRQ SURGXFWV VXFK DV WKH WDUQLVKHV IRXQG RQ SRWHQWLRVWDWLF VDPSOHV DW ORZ SRWHQn WLDOV LQ WKLV VWXG\ LV DQ DUHD RI LPSRUWDQFH ,GHQWLILFDn WLRQ RI WKLQ ILOPV RI SDVVLYH VSHFLHV FRXOG OHDG WR LPSRU WDQW DGYDQFHV LQ DOOR\ GHYHORSPHQWA $ n 7KH SRVVLELOLWLHV RI UHIOHFWDQFH VSHFWURVFRS\ LQ WKH XOWUDYLROHW YLVLEOH DQG LQIUDUHG VSHFWUD VKRXOG EH LQYHVWLJDWHG 5HFHQW GHDOXPLQL]DWLRQ IDLOXUHV LQ DOXPLQXP EURQ]HV LQGLFDWH D QHHG IRU IXUWKHU XQGHUVWDQGLQJ RI WKLV V\VWHP RI DOOR\V

PAGE 131

& / %XORZ KDV VXJJHVWHG WKDW HOHFWURQ PLFURSUREH LQn YHVWLJDWLRQV RI DUVHQLFDO GXSOH[ DOSKDSOXVEHWD EUDVVHV PLJKW RYHUFRPH RXU SUHVHQW ODFN RI XQGHUVWDQGLQJ DQG OHDG WR WKH GHYHORSPHQW RI DQ HIIHFWLYH GH]LQFLILFDWLRQ LQKLELWRU U f IRU WKHVH DOOR\V 7KH LUUHJXODULWLHV LQ WKH FRSSHU GLVVROXWLRQ GDWD REn WDLQHG E\ DWRPLF DEVRUSWLRQ DQDO\VLV PD\ ZHOO EH H[SODLQHG E\ D V\VWHPDWLF VWXG\ RI WKH HYROXWLRQ RI PRUSKRORJ\ RI GHn DO OR\HG VSRQJHV XVLQJ WKH VFDQQLQJ HOHFWURQ PLFURVFRSH 7KH DSSHDUDQFH RI DQQXODU ULQJV VXFK DV WKRVH VKRZQ LQ )LJXUH FDQQRW EH H[SODLQHG PHUHO\ E\ WHPSHUDWXUH YDULDn WLRQV &RUURVLRQ WXQQHOLQJ VXFK DV LV GLVFXVVHG E\ 3 5 6ZDQQ KDV D EHDULQJ RQ VWUHVV FRUURVLRQ RI VRPH DOOR\V 7LLLV LV FRQVLGHUHG WR RFFXU E\ D GHDOOR\LQJ PHFKDQLVP DQG LQGLFDWHV D IXUWKHU QHHG IRU PRUSKRORJLFDO VWXGLHV 7KH IDLOXUH RI WKH 1HUQVW HTXDWLRQ WR H[SODLQ GHDOOR\n LQJ LQGLFDWHV D QHHG IRU LQYHVWLJDWLRQV LQWR D UHOLDEOH PHWKRG RI ORZWHPSHUDWXUH GHWHUPLQDWLRQ RI WKH DFWLYLWLHV RI DOOR\ FRPSRQHQWV

PAGE 132

$33(1',&(6

PAGE 133

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f ;UD\ (QHUJ\ $QDO\VLV 6\VWHP +HDWK 0DOPVWDGW (QNH 0RGHO (8' ZLWK 7HFKWURQ 7\SH $% 6ORWW\SH %XUQHU

PAGE 134

$33(1',; (/(&752/<7(6 S+ sf &RPSRVLWLRQ PRO DU 1D&O PRODU +& PRODU .+&+r PRODU 1D&O PRODU NKFKR PRODU 1D2+ PRODU 1D&O PRODU 1D2+ PRODU .+&+ PRODU 1D&O PRODU 1D2+ PRODU NLLFKR PRODU 1D&O PRODU 1D2+ PRODU .+&+r PRODU 1D&O PRODU 1D2+ PRODU .+&+r PRODU 1D&O PRODU 1D2+ PRODU NKFKR PRODU 1D&O PRODU 1D+& PRODU NKFKR PRODU 1D&O PRODU 1D+&2M PRODU 1D&O PRODU 1D+&2 PRODU 1D2+

PAGE 135

S+ sf &RPSH L" LW ,RQ PRO DU 1D&O PR D U 1D+&2 PRODU 1D 2+ PRO DU 1D&O PRODU 1D2+ PRODU 1D+&2 PRODU 1D&O PRODU 1D2+ PRODU 1D&O PRODU 1D2+ PRODU 1D&O PRODU 1D2+

PAGE 136

$33(1',; &+(0,&$/ $1$/<6(6 2) $/3+$ %5$66 6DPSOH 1XPEHU (OHPHQW :HLJKW 3HUFHQW &RSSHU /HDG ,URQ 7LQ 1LFNHO 0DQJDQHVH U! R A 6LOLFRQ $OXPLQXP 7HOOXULXP SOXV 6HOHQLXP 1RQH GHWHFWHG 3KRVSKRUXV 1RQH GHWHFWHG %LVPXWK =LQF 5HPDLQGHU

PAGE 137

6DPSOH 1XPEHU (OHPHQW :HLJKW 3HUFHQW &RSSHU /HDG ,URQ 7LQ 1LFNHO 0DQJDQHVH 6LOLFRQ $OXPLQXP $QWLPRQ\ 3KRVSKRUXV 1RQH GHWHFWHG %LVPXWK $UVHQLF =LQF 5HPDLQGHU $QDO\VHV ZHUH VXSSOLHG E\ WKH PDQXIDFWXUHU &KDVH %UDVV DQG &RSSHU &RPSDQ\ ,QFRUSRUDWHG ZKR GRQDWHG WKH PDWHULDO 6DPSOH ZDV XVHG IRU WKH PHWDOORJUDSKLF [UD\ HOHFn WURQ PLFURSUREH DQG SRWHQWLRNLQHWLF VWXGLHV 6DPSOH ZDV XVHG IRU DWRPLF DEVRUSWLRQ DQG SRWHQWLRVWDWLF VWXGLHV

PAGE 138

$33(1',; %(7$ %5$66 ,1*27 35(3$5$7,21 %HWD EUDVV DOOR\V ZHUH SUHSDUHG IURP ZR SXUH FRSSHU DQG SXUH ]LQF VWRFN SXUFKDVHG IURP WKH $PHULFDQ 6PHOWLQJ DQG 5HILQLQJ &RPSDQ\ 7KH SURFHGXUH ZKLFK IROORZV If LV VLPLODU WR WKDW GHVFULEHG E\ 0DUVEDNRY DQG FRZRUNHUV 3LHFHV RI FRSSHU DQG ]LQF ZHUH FOHDQHG LQ FRQFHQWUDWHG +12A IROORZHG E\ ULQVLQJ LQ +1&8,A2 LQ GLVWLOOHG ZDWHU DQG ILQDOO\ LQ DFHWRQH DIWHU ZKLFK WKH\ ZHUH GULHG XQGHU D KRWDLU EORZHU &KDUJHV ZHLJKLQJ DSSUR[LPDWHO\ JUDPV ZHUH SODFHG LQ PP 2' 9\FRUr YLDOV ZKLFK WKHQ ZHUH HYDFXDWHG DQG VHDOHG 7KH HQFDSVXODWHG FKDUJHV ZHUH SODFHG LQ D VWDQGDUG KHDWWUHDWLQJ IXUQDFH DQG WKH WHPSHUDWXUH RI WKH IXUQDFH ZDV EURXJKW XS WR r& ZKLFK LV DSSUR[LPDWHO\ r& DERYH WKH OLTXLGXV IRU WKH DOOR\ VHH )LJXUH f $IWHU WKH DOOR\V KDG EHHQ PROWHQ IRU DW OHDVW IRXU KRXUV WKH FDSVXOHV ZHUH UHPRYHG IURP WKH IXUQDFH DQG WXUQHG HQGRYHUHQG DW OHDVW WZLFH LQ RUGHU WR VWLU WKH OLTXLG 7KH FDSVXOHV ZHUH WKHQ SODFHG EDFN LQ WKH IXUQDFH DQG NHSW DW r& IRU DQRWKHU r7UDGH QDPH &RUQLQJ *ODVV &RPSDQ\

PAGE 139

KRXU 7KHQ WKH IXUQDFH FRQWUROOHL ZDV WXUQHG GRZQ WR r& DQG KHOG IRU IRXU KRXUV DW WKLV WHPSHUDWXUH 7KH FDSVXOHV ZHUH WKHQ UHPRYHG IURP WKH IXUQDFH DQG EURNHQ DQG WKH LQJRWV ZHUH TXHQFKHG XQGHU D ZDWHU WDS 7RS DQG ERWWRP VHFWLRQV ZHUH UHPRYHG IURP HDFK LQJRW DQG DQDO\]HG DFFRUGLQJ WR WKH SURFHGXUH LQ $SSHQGL[ *UDLQ VL]H ZDV TXLWH ODUJH DQG D W\SLFDO FURVVVHFWLRQ ZRXOG KDYH IRXU RU ILYH JUDLQV YLVLEOH DIWHU SROLVKLQJ

PAGE 140

$33(1',; 352&('85( )25 7+( (/(&752*5$9,0(75,& &+(0,&$/ $1$/<6,6 2) %,1$5< &233(5=,1& $//2<6 7KH IROOFPLQJ SURFHGXUH LV DGDSWHG IURP WKH SURFHGXUHV UHFRPPHQGHG E\ WKH $PHULFDQ 6RFLHW\ IRU 7HVWLQJ 0DWHULDOV DQG E\ + $\UHV 4XDQWLWDWLYH &KHPLFDO $QDO\VLV 8VH D EDODQFH DFFXUDWH WR JP IRU DOO ZHLJKLQJV 3ODFH D RQHJUDP VDPSOH RI WKH EUDVV WR EH DQDO\]HG LQ D PO EHDNHU $GG UU GLVWLOOHG /2 PO FRQH +62f DQG PO FRQH +12 WR WKH EHDNHU LQ WKH RUGHU OLVWHG 7KLV VROXWLRQ ZLOO EHFRPH KRW DQG LQLWLDWH GLVVROXWLRQ RI WKH EUDVV 'LOXWH WKH PL[WXUH WR PO DQG VWLU XVLQJ DQ LQn HUW PDJQHWLF VWLUULQJ EDU XQWLO WKH EUDVV LV FRPSOHWHO\ GLVVROYHG 'LOXWH WR PO (OHFWURO\]H DW DPS DSSUR[LPDWHO\ YROWVf IRU KRXUV XVLQJ D SUHZHLJKHG SODWLQXP JDX]H HOHFWURGH DV WKH FDWKRGH DQG D SODWLQXP ZLUH SODFHG DW OHDVW FP IURP WKH FDWKRGH DV DQ DQRGH $GG PO RI ZDWHU WR WKH VROXWLRQ ,I QR DGGLn WLRQDO FRSSHU GHSRVLW LV DSSDUHQW DW WKH QHZ VROXWLRQ OHYHO DIWHU PLQXWHV DVVXPH WKH HOHFWURGHSRVLWLRQ LV FRPSOHWH

PAGE 141

,I FRSSHU DSSHDUV FRQWLQXH HOHFWURO\VLV XQWLO QR QHZ FRSSHU DSSHDUV ZKHQ OLTXLG OHYHO LV UDLVHG 5HPRYH FDWKRGH IURP WKH ZDWHU ZLWK WKH FXUUHQW VWLOO RQ 5LQVH WKH FDWKRGH XVLQJ GLVWLOOHG ZDWHU IURP D ZDVK ERWWOH DV WKH FDWKRGH LV UHPRYHG 5LQVH WKH FDWKRGH ZLWK GLVWLOOHG ZDWHU DQG DFHWRQH XVLQJ ZnDVK ERWWOHV 'U\ WKH FDWKRGH XVLQJ D KRWDLU EORZHU :HLJK WKH FDWKRGH 7KH ZHLJKW SHUFHQWDJH RI FRSSHU LQ WKH WHVW VDPSOH LV JLYHQ E\ ZR &X ZHLJKW JDLQ RI FDWKRGH RU UUMPO 7 R M QU 9} I 2 3 & 2+LQO 3 =LQF FRQWHQW LV FDOFXODWHG WR EH WKH UHPDLQGHU RI WKH VDPSOH

PAGE 142

$33(1',; $1$/<6,6 2) %(7$ %5$66 ,1*276 86(' ,1 7+,6 ,19(67,*$7,21 ,QYHVWLJDWLRQV ,QJRW QXPEHU $QDO\VLVD ZRf LQ ZKLFK LQJRW ZDV XVHG &X =Q (OHFWURQ PLFURSUREH ;UD\ GLIIUDFWLRQ HOHFWURQ PLFURSUREH )UHH FRUURVLRQ SRWHQWLDO $WRPLF DEVRUSWLRQ 6, 26 3RWHQWLRVWDWLF 3RWHQWLRVWDWLF 3RWHQWLRVWDWLF $OO DQDO\VHV GHVFULEHG LQ ZHUH SHUIRUPHG XVLQJ WKH SURFHGXUH $SSHQGL[

PAGE 143

$33(1',; (/(&752&+(0,&$/ 6$03/( 35(3$5$7,21 $OO VDPSOHV ZHUH VOLFHG LQWR GLVFV RQ D FXWRII ZKHHO DQG JURXQG DW JULW RQ D EHOW VDQGHU 6DPSOHV ZHUH WKHQ KDQG SROLVKHG RQ DQG JULW VLOLFRQ FDUELGH SDSHU ZDVKHG LQ GLVWLOOHG ZDWHUDQG GULHG XQGHU D KRWDLU EORZHU

PAGE 144

$33(1',; (48$7,216 86(' ,1 &216758&7,21 2) 327(17,$/ 9(5686 S+ ',$*5$06 )25 7+( &X&+ 6<67(0 $1' 7+( =Q+ 6<67(0 &X&O+" 6\VWHP >H[WUDFWHG IURP 9DQ 0X\OGHU GH=RXERY DQG 3RXUEDL[AOf@ (J IW &X + &X + H Df ( S+ QR K\GUDWHG R[LGHVf Ef ( S+ DVVXPHV K\GUDWHG &Xf FXR + &X2 + H Df ( S+ Ef ( S+ DVVXPHG K\GUDWHG EXW QRQK\GUDWHG &X &X2f &X2 + &X + H Df ( 3+ A(TXDWLRQ QXPEHUV UHIHU WR FLUFOHG QXPEHUV RQ GLDJUDPV )LJXUHV DQG

PAGE 145

&X&+ 6\VWHP FRQWLQXHGf (J &X + &X2 + Df ORJ&Xf S+ &X&O &X & H Df ( ORJ f§!f§ &X&O f ORJ&Of &X&O + &X & +r Df ORJ&O f S+ &X2+f&X&O &X2 &Of + + Df ORJ &Of S+ &X &Of &X&O H Df ( ORJ&Of &X + &O &X2+ff&X&O\ + H Df ( S+ ORJ&Off

PAGE 146

&X&O,A2 6\VWHP FRQWLQXHGf (J &X&O + &X2+f&X& &On + H Df ( S+ ORJ&Onf &X + & &X2+f&X&O ‘} + H Df ( ORJ&Of S+ &X&a + &X &O + Df ORJ&X&Of S+ ORJ&Onf &X &a &X&O H Df ( ORJ&X&Of ORJ&Of &X&O + &X2 & + H Df ( S+ ORJ&X&Of ORJ&Oaf &X&O &X &O H Df ( ORJ&Xf ORJ&Of

PAGE 147

=Q+A2 6\VWHP >H[WUDFWHG IURP =RXERY DQG 3RXUEDL[AfA@ (J LI =Q + =Q2 + H Df ( S+ =Q +" =Q2 + Df ORJ=Qf S+ =Q2 + +=Q2f + Df ORJ+=Q2 f S+ =Q =Q H Df ( ORJ=QIf

PAGE 148

%,%/,2*5$3+< 5 +HLGHUVEDFK &RUURVLRQ B f % %LUG DQG / 0RRUH 0DWHULDOV 3URWHFWLRQ f 7 &UHQHOO DQG / ( 6DZ\HU -RXUQDO RI $SSOLHG &KHPLVWU\ B f $ ) %OXPHU &RUURVLRQ f 0 0DWVXGD -RXUQDO RI WKH -DSDQHVH ,QVWLWXWH RI 0HWDOV B f 5 % 1LHGHUEHUJHU 0RGHUQ &DVWLQJV B f ( 3 3ROXVKNLQ DQG 0 6KXOGHQHU 7UDQVDFWLRQV $,0( f $ / 6LPPRQV 0HWDO 3URJUHVV B f $ : 7UDF\ $670 673 f 6M YDQ GHU %DDQ &RUURVLRQ f : + 9HUQRQ 7UDQVDFWLRQV RI WKH )DUDGDY 6RFLHW\ f 5 $ :LONLQV 0HFKDQLFDO (QJLQHHULQJ >@ f / +HQGHUVRQ DQG & / 5RDGKRXVH -RXUQDO RI 'DLU\ 6FLHQFH B f : & 6WHZDUW DQG ) / /D4XH &RUURVLRQ f ) 7D\ORU DQG : :RRG (QJLQHHULQJ f 6 & %ULWWRQ -RXUQDO RI WKH ,QVWLWXWH RI 0HWDOV B f / : *OHHNPDQ DQG 5 6ZDQGE\ &RUURVLRQ W f n f

PAGE 149

6 & +DPVWHDG ,QGXVWULDO DQG (QJLQHHULQJ &KHPLVWU\ >@ $ f $ + +HVVH ( 7 0\VNRFRVNL DQG % 0 /RULQJ 7UDQVn DFWLRQV Re WKH $PHULFDQ )RXQGU\PHQnV $VVRFLDWLRQ f % ) 3HWHUV $ + &DUVRQ DQG 5 %DUHU 0DWHn ULDOV 3URWHFWLRQ B >@ f 0 6FKXVVOHU DQG 6 1DSROLWDQ &RUURVLRQ B W Y nf n6HUUH DQG /DZUH\V &RUURVLRQ 6FLHQFH L f : &ODUN -RXUQDO RI WKH ,QVWLWXWH RI 0HWDOV f 3 7UDXW]HO DQG : 7UHDGZHOO +HOYLWLD &KHPLFD $FWD B f / 0 /HHGRQ -RXUQDO RI WKH $PHULFDQ :DWHU :RUNV $VVR FLDWLRQ A f $ 'YRUDN 6WURMLUHQWVYL B >@ f UHIHUn HQFHG E\ &RUURVLRQ $EVWUDFWV B f 0 )RQWDQD ,QGXVWULDO DQG (QJLQHHULQJ &KHPLVWU\ >@ $ f : /\QHV 3URFHHGLQJV $670 B f 5 % $EUDPV 7UDQVDFWLRQV RI WKH $PHULFDQ (OHFWURn FKHPLFDO 6RFLHW\ B f & ) 1L[RQ 7UDQVDFWLRQV RI WKH $PHULFDQ (OHFWURn FKHPLFDO 6RFLHW\ B f + +ROORPRQ DQG :XOII 7UDQVDFWLRQV $,0( f % 7KRPSVRQ $XVWUDODVLDQ (QJLQHHULQJ S 2FWREHU f $ 5 =HQGHU DQG & / %XORZ +HDWLQJ 3LSLQJ DQG $LU &RQGLWLRQLQJ B f ( 6 'L[RQ $670 %XOOHWLQ 1R S f ) + 5KRGHV DQG 7 &DUW\ ,QGXVWULDO DQG (QJLQHHUn LQJ &KHPLVWU\ f

PAGE 150

& 6FXOO\ 7KH )XQGDPHQWDOV RI &RUURVLRQ 3HUJDPRQ 3UHVV 1HZ @ f 5 0 +RUWRQ &RUURVLRQ B >@ f 9 ) /XFH\ %ULWLVK &RUURVLRQ -RXUQDO f LELG f / ( 7DERU 7UDQVDFWLRQV RI WKH :DWHU :RUNV $VVRFLn DWLRQ >@ f 6 6 *DVWHY ,]YHVW $NDG 1DXN 665 0HWDOO\ B f UHIHUHQFHG LQ &RUURVLRQ $EVWUDFWV A f 0 %LDORVN\ &RUURVLRQ DQG 0HWDO 3URWHFWLRQ B f . 0DUVKDNRY 9 3 %RJGDQRY DQG 6 0 $OHLNLQD 5XVVLDQ -RXUQDO RI 3K\VLFDO &KHPLVWU\ B >@ f LELG >@n f LELG f f : + %DVVHWW &KHPLFDO DQG 0HWDOOXUJLFDO (QJLQHHULQJ B f 2n0 %RFNULV 3ULYDWH FRPPXQLFDWLRQ % 7 5XELQ 3K' 'LVVHUWDWLRQ 8QLYHUVLW\ RI 3HQQV\On YDQLD /3LDWWL DQG 5 *UDXHU :HUNVWRIIH XQG .RUURVLRQ >@ f 7 &ROHJDWH 0HWDO ,QGXVWU\ B >@ f / .HQZRUWK\ DQG : 2n'ULVFROO &RUURVLRQ 7HFKn QRORJ\ B f 8 5 (YDQV 7KH &RUURVLRQ DQG 2[LGDWLRQ RI 0HWDOV (GZDUG $UQROG DQG &RPSDQ\ /RQGRQ L & : 6WLOOZHOO DQG ( 6 7XUQLSVHHG ,QGXVWULDO DQG (QJLQHHULQJ &KHPLVWU\ B f ( 9HULQN -U DQG 3 $ 3DUULVK &RUURVLRQ f

PAGE 151

) : )LQN 7UDQVDFWLRQV RI WKH (OHFWURFKHPLFDO 6RFLHW\ B f 0 1 'HVDL 7DODWL DQG $ 0 7ULYHGL -RXUQDO RI WKH ,QGLDQ &KHPLFDO 6RFLHW\ > f / 3LDWWL DQG 5 *UDXHU :HUNVWRIIH XQG .RUURVLRQ B f 0 & 6WHHOH $XVWUDODVLDQ &RUURVLRQ (QJLQHHULQJ S -DQXDU\ f ( 9HULQN -U DQG 5 +HLGHUVEDFK -U (YDOn XDWLRQ RI WKH 7HQGHQF\ IRU 'HDOOR\LQJ LQ 0HWDO 6\VWHPV DFFHSWHG IRU SXEOLFDWLRQ LQ DQ $670 6SHFLDO 7HFKQLFDO 3XEOLFDWLRQ RQ /RFDOL]HG &RUURVLRQ 5 % $EUDPV 7UDQVDFWLRQV RI WKH $PHULFDQ (OHFWURn FKHPLFDO 6RFLHW\ B f GLVFXVVLRQ E\ %HQJRXJK DQG 5 0D\ & / %XORZ LQ &RUURVLRQ +DQGERRN + + 8KOLJ HGLWRU -RKQ :LOH\ DQG 6RQV 1HZ @ -DQXDU\ f ( ( /DQJHQHJJHU DQG ) 3 $ 5RELQVRQ &RUURVLRQ f LELG B f + : 3LFNHULQJ DQG & :DJQHU -RXUQDO RI WKH (OHFWURn FKHPLFDO 6RFLHW\ f + : 3LFNHULQJ 3URFHHGLQJV RI WKH &RQIHUHQFH RQ WKH )XQGDPHQWDOV RI 6WUHVV &RUURVLRQ &UDFNLQJ &ROXPEXV 2KLR f 1$&( 5 : 6WDHKOH HGLWRU S ) 3 $ 5RELQVRQ DQG 0 6KDOLW &RUURVLRQ 7HFKQRORJ\ $SULO f 1$&( 6WDQGDUG 70 /DERUDWRU\ 7HVWLQJ RI 0HWDOV IRU WKH 3URFHVV ,QGXVWULHV SXEOLVKHG DV DQ LQVHUW WR 0DWHULDOV 3URWHFWLRQ B >@ 0D\ f f ,QWHUQDWLRQDO &ULWLFDO 7DEOHV RI 1XPHLLFDO 'DWD 1DWLRQDO $FDGHP\ RI 6FLHQFHV 0DSOH 3UHVV
PAGE 152

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n G\QDPLF 3RODUL]DWLRQ 6HFWLRQ 6XEFRPPLWWHH ;, $670 &RPPLWWHH $SULO ( 5HLQRHKO ) + %HFN DQG 0 )RQWDQD &RUURn VLRQ B @ f : & )RUW ,,, +LJK +RQRUV 3URMHFW 8QLYHUVLW\ RI )ORULGD )LVKHU DQG +DOSHULQ -RXUQDO RI WKH (OHFWURn FKHPLFDO 6RFLHW\ f + : 3LFNHULQJ -RXUQDO RI WKH (OHFWURFKHPLFDO 6RFLHW\ f / *UDI 0HWDOOZLUWVFKDIW f =HLWVFKULIW IXU 0HWDOONXQGH e f +DVKLPRWR 7 *RWR : 6XHWDND DQG 6 6KLPRGDLUD 7UDQVDFWLRQV RI WKH -DSDQHVH ,QVWLWXWH RI 0HWDOV f 5 .OHLQEH HU + 2NX]XPL DQG 3 3HULR 0HWDX[ &RUURn VLRQ ,QGXVWULHV 6B f + : 3LFNHULQJ -RXUQDO RI WKH (OHFWURFKHPLFDO 6RFLHW\ f ( $ 2ZHQ DQG ( : 5REHUWV 3KLORVRSKLFDO 0DJD]LQH f

PAGE 153

5 +HLGHUVEDFK &RUURVLRQ B f + 6XJDZDUD DQG + (ELNR &RUURVLRQ 6FLHQFH B f / 6 %LUNV (OHFWURQ 3UREH 0LFURDQDO\VLV ,QWHUn VFLHQFH 3XEOLVKHUV 1HZ
PAGE 154

+DOSHULQ -RXUQDO RI WKH (OHFWURFKHPLFDO 6RFLHW\ f 9HWWHU (OHFWURFKHPLFDO .LQHWLFV $FDGHPLF 3UHVV 1HZ @ f & 9UHHODQG DQG 7 %HGIRUG 0DWHULDOV 3URWHFWLRQ >@ f : % %URRNV &RUURVLRQ f

PAGE 155

$ + 7D\ORU -RXUQDO RI WKH (OHFWURFKHPLFDO 6RFLHW\ f 0 3RXUEDL[ $WODV RI (OHFWURFKHPLFDO (TXLOLEULD LQ $TXHRXV 6ROXWLRQV 3HUJDPRQ 3UHVV 1HZ
PAGE 156

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n FDQ ,QVWLWXWH RI 0LQLQJ 0HWDOOXUJLFDO DQG 3HWUROHXP (QJLn QHHUV WKH $PHULFDQ 6RFLHW\ IRU 0HWDOV WKH 1DWLRQDO $VVRn FLDWLRQ RI &RUURVLRQ (QJLQHHUV 3UHVV &OXE DQG 6LJPD $OSKD (SVLORQ

PAGE 157

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


POTENTIAL vs SHJE.
91
Figure 35.
Simplified Z11-H2C) diagram for concentrations
of ionic species = 106m.


118
10. The resulting composite diagram provides a basis
for understanding the influence of stirring and high flow
rates on the dezincification process.
11. The composite diagram also provides an explanation
for why tests in copper chloride solutions produce acceler
ated dezincification.
12. The use of copper chloride solutions to test coppe
alloys for susceptibility to dealloy is a well-established
practice. It is not valid to make conclusions on the mecha
nism of dezincification based solely on tests in these solu
tions.


ACKNOWLEDGEMENTS
I would like to express my thanks to Dr. Ellis D. Verink,
Jr., for his guidance and inspiration. Thanks are also ex
tended to the members of my committee, Dr. A. D. Wallace, Dr.
R. T. Delloff, Dr. R. W. Gould and Dr. J. J. Hren.
Dr. M. Pourbaix provided encouragement during the early
stages of this investigation. Dr. S. R. Bates provided
guidance on the use of and interpretation of results from
the scanning electron microscope and the electron microprobe.
Mr. W. C. Fort III, provided invaluable assistance in ob
taining the experimental results. Technical assistance was
provided by Mr. W. A. Aeree, Mr. E. J. Jenkins, Mr. E. C.
Logsdon, Mr. P. D. Kalb, Mr. C. J. Minier and Mr. C. Sim
mons .
I particularly wish to acknowledge the financial sup
port supplied by the National Association of Corrosion Engi
neers and by the International Nickel Company. In addition,
certain of the equipment used in the electrochemical studies
was purchased with funds from the Office of Saline Water.
The Chase Brass and Copper Company donated the alpha
brass used in this investigation, and this contribution is
gratefully acknowledged.
iii.


80
60
Z
40
20
0
Figure 30.
^4
JL
12
JL
24
I L
36 48
HOURS
L
60
J
72
Dezincification factors, Z, for beta brass in 5N HC1
at 67.50C. Scatter bands show the limits of precision
for the original measurements.


115
54 and 35, were calculated using the Nernst equation for the
"immunity" lines. Preliminary investigations into the ex
perimental Pourbaix diagram for pure nickel reveal that there
are significant differences between the experimental and the-
non i o h
oretical diagrams for nickel. > The same may hold
true for zinc and gold, although the experimental copper dia
gram in chloride solutions has been determined and agrees
r i o p
closely with many features of the theoretical diagram.
126)
One possibility for the failure of the Nernst equation
to explain the experimental observations discussed above is
that the calculations were made assuming the.activities of
the elements in the alloy were directly proportional to the
composition. This approximation may not be justified. The
activities of constituents of alloys usually are determined
at. high temperatures and may not be the same at low temper-
rgg 1^7 1?8)
atures.'' j However, radical departures from ideal
ity are seldom encountered in alloy systems and these de
partures would have to alter the activities by several
orders of magnitude before they would alter the principal
points of the above argument.


56
This would seem to indicate that the presence of copper in
solution from a dealloyed metal could also be interpreted
as being the result of varying dissolution rates or, in
other words, a preferential removal process in which one
metal dissolves faster than the other.
Solution analysis has been used to measure the rate of
r 4 7 78
dealloying as a function of time and/or temperature.1-
95,96) Rub in reported the "partial apparent heats of
activation as a function of composition" for the dissolution
of a series of copper-nickel alloys, some of which denickel-
ified, in IN HC1. The total apparent activation energies,
the sums of the two partial values, varied from 5 to 10
Kcal/mole of alloy. The lower figure, corresponding to pure
copper, compares quite well with the 5.44 Kcal/mole obtained
by Halperin^-102-* for the dissolution of copper in ammonia
solutions. The values reported are also in the general
range described by Vetter for the dissolution of metals in
, (103)
electrolytes.v J
Fisher and Halperin^reported that copper-gold al
loys showed a decrease in dissolution rate with an increase
in temperature for all of their alloys. They also observed
parabolic corrosion rates. Parabolic corrosion rates are
characteristic of systems in which the rate is determined
by transfer of reactants through an adherent surface film
which thickens as the reaction progresses.Thus they
concluded that the rate was limited by a surface film,


120
C. L. Bulow has suggested that electron microprobe in
vestigations of arsenical duplex alpha-plus-beta brasses
might overcome our present lack of understanding and lead
to the development of an effective dezincification inhibitor
r ,, (130)
for these alloys.
The irregularities in the copper dissolution data ob
tained by atomic absorption analysis may well be explained
by a systematic study of the evolution of morphology of de
al loyed sponges using the scanning electron microscope.
The appearance of "annular rings" such as those shown in
Figure 17 cannot be explained merely by temperature varia
tions.
"Corrosion tunneling," such as is discussed by P. R.
(131
Swann, has a bearing on stress corrosion of some alloys.
Tiiis is considered to occur by a dealloying mechanism and
indicates a further need for morphological studies.
The failure of the Nernst equation to explain dealloy
ing indicates a need for investigations into a reliable
method of low-temperature determination of the activities
of alloy components.


BIOGRAPHICAL SKETCH
Robert Henry Heidersbach, Jr., was born December 30,
1940, at El Paso, Texas. He attended public schools in
Illinois and Colorado and was graduated from the Colorado
School of Mines in 1963, with the degree of Metallurgical
Engineer. From 1963 to 1967 he served in the Corps of
Engineers of the United States Army and was stationed in
Germany and Vietnam. In 1967 he entered the University of
Florida; he enrolled in the Graduate School in 1968. He
received the degree Master of Engineering in December, 1969.
He has pursued the degree of Doctor of Philosophy in the
Department of Metallurgical and Materials Engineering since
that date.
Robert Henry Heidersbach, Jr., is married to the former
Dianne Katherine Shrum. He is the father of two children,
Robert Scott and Krista Lynn. He is a member of the Ameri
can Institute of Mining, Metallurgical and Petroleum Engi
neers, the American Society for Metals, the National Asso
ciation of Corrosion Engineers, Press Club, and Sigma Alpha
Epsilon.


79
Many chemical reactions, including metal dissolution,
are found to obey an empirical equation first proposed by
Arrhenius
K K0 e-<5/RT ,
where the pre-exponential factor, Kq is usually found to
be temperature-independent, at least within the experimental
accuracy of the observations, > 106) an(j q so_caiie[
"activation energy." K and Kq in the above equation are
rate-related measurements such as weight loss, weight gain,
depth of penetration, and metal dissolution for corrosion
experiments. The amount of zinc dissolved in 48 hours was
used in the discussion which follows. This is a reasonable
choice since Figure 31 shows the kinetics to be linear. R
in the Arrhenius equation is the gas constant, 1.986 calories
gm-moledegree K 1^104) and T is the absolute temperature
in degrees Kelvin.
If reaction kinetics can be represented by the above
Arrhenius formula, then a plot of log K as a function of
1/T will give a straight-line having a slope of -Q/2.303R.
106)
Figure 32 is an Arrhenius plot of the amount of zinc
dissolved from beta brass in 48 hours. The activation energy
for zinc dissolution over the temperature interval 59.60-
99.85C is found from the slope of this graph to be 18 Kcal
gm-mole ^. This value is somewhat higher than those for


8 PPM
Figure 24. Atomic-absorption calibration curve for copper.


8
before and after each exposure by titration with a IN NaOH
standard. It was found to vary by no more than 0.05 nor
mality units from start to finish of any test.
The liquid temperature was controlled to within 0.35C
by means of an electric thermoregulator coupled to a 100
watt immersion heater.
Immersion tests were conducted in IN NaCl and 5N HC1.
These solutions were chosen because they are nonoxidizing
and do not form soluble metal complexes with the exception
of CuC^". Because of the historic significance of dezinc-
ification failures in salt water and other chloride environ
ments, it was felt that chlorides should be used despite
the existence of this one complex metal ion. The high acid
concentration was chosen to minimize acidity changes within
the solution. This high acid concentration was also very
close to the maximum concentration that can be maintained
at 100C without boiling.^7)
Previous immersion tests in HC1 and NaCl have appeared
+ (29,40,47,52,61,62,68,69) ,
m the literature, >> J and comparisons
with previous results are contained in the discussion section
of this work.


130
If copper appears continue electrolysis until no new copper
appears when liquid level is raised.
8. Remove cathode from the water with the current
still on. Rinse the cathode using distilled water from a
wash bottle as the cathode is removed.
9. Rinse the cathode with distilled water and acetone
using w'ash bottles.
10. Dry the cathode using a hot-air blower.
11. Weigh the cathode.
12. The weight percentage of copper in the test sample
is given by:
w/o Cu =
weight gain of cathode
or rr-jml 1 T o j nr V f" O P C OHinl P
- -
100.
13.Zinc content is calculated to be the remainder of
the sample.


45
Present Work
Exposure samples intended for x-ray diffraction analy
sis were prepared according to the procedures described in
the section on sample preparation. NACE Standard TM-01-
69(66) SUggests that closed exposure cells, such as were
used in this investigation, should contain several samples
to be pulled after certain periods of time. Each time a
sample is removed from the cell a replacement sample is
added. The reason which is given for this is to check for
possible changes in the corrosivity of the environment.
As an example, in one series of tests run during this in
vestigation five samples were prepared for each cell. Three
were immersed in the original solution. After ten days one
sample was removed and replaced with a fresh sample. This
was repeated at the end of twenty days, and the test was
terminated at the end of thirty days. Thus five samples
were obtained from each test and, if no changes in the cor
rosivity of the environment occurred, the amount of dezincifi-
cation experienced by the sample exposed for the first ten
days of the test would correspond to that of the sample ex
posed for the last ten. The same should hold for samples
exposed for the first twenty days and the last twenty days.
As was mentioned above, these tests were intended pri
marily as a source of dezincified brass to be used for x-ray
powder samples. Because of this each sample was a simple
disc with a hole drilled near one edge so it could be


94
If the above-stated hypotheses are correct, then 70-30
brasses should be expected to undergo dezincification in
acid solutions at potentials from ~ 0.940VTrr; to 0. 200Vctrc
r SHE SHE
(the region with small dots on Figure 36). At potentials
between O.OOOVg^g and +0.200Vg^E (the cross-hatched region
of Figure 36), copper and zinc should dissolve, but not
necessarily in the same ratio as in the alloy.
In order to test these hypotheses, a series of poten-
tiostatic tests of alpha brass were conducted in buffered
0.1M chloride solutions at pH 4 (see Appendix 2 for the
exact composition of the solution used).
The equipment for these tests and the sample prepara
tion procedures were discussed in the section on experimental
procedure.
these
tests are tabulated
in
Table
8.
at +0 ,
.150VCIrr, and
SHE
+0.050V
SHE
, in th
e
the cross-hatched
region
of
Figure
36 ,
appeared dezincified in solutions that were not stirred.
Stirred solutions produced samples which appeared to have
undergone general dissolution. Stirring apparently removed
stagnant conditions which would allow redeposition of copper
on the specimen surface.
Those samples which were exposed at potentials from
-0.050Vshe to -0.850Vg^E were observed to dezincify in
stirred solutions. This corresponds to the potential region
f9 7'\
where Pickering and Byrne1 1
reported dezincification of


MICRONS
Figure 13.
Zinc intensity profile from a sample of alpha brass
dezincified for 10 days in 5N HC1 at 75C.


COPPER
ZINC
HOURS
ts)
Figure 26. Copper and zinc dissolution from alpha brass in 5N HC1
at 98.50C. Scatter bands show the precision of the
original measurements.


pOTtHl'^
figure
Simplified C11-CI-H2O diagram at 25C foi
solution containing 0.1M chloride ions;
concentrations of ionic species = 106m
''-en, Ref. 123.]


124
pH (0.2)
Compe i? it Ion
9.5
0.100
mol ar
NaCl
0.100
mo 1 a r
NaHCO
0.020
molar
Na OH
10.0
0.100
mol ar
NaCl
0.001
molar
NaOH
0.003
molar
NaHCO
11.0
0.100
molar
NaCl
0.001
molar
NaOH
12.0
0.100
molar
NaCl
0.010
molar
NaOH
13.0
0.100
molar
NaCl
0.100
molar
NaOH


134
Cu-C1-H20 System (continued)
Eg, #
17 Cu++ + H20 = CuO + 2H+
a) log(Cu++) = 7.89 2pH
42 CuCl"2 = Cu++ + 2C1" + e
a) E = 0.465 = 0.0591 log >
(CuCl 2)
+ 0.1182 log(Cl")
51 2CuCl + H20 = Cu20 + 2C1" + 2H*
a) log(Cl ) = -5.66 + pH
53 3Cu(OH)2-CuCl2 = 4CuO + 2Cl + 2H+ + 2H20
a) log (Cl") = 7.40 + pH
55 Cu + Cl = CuCl + e
a) E = 0.137 0.0591 log(Cl")
57 4Cu + 6H20 + 2Cl" = 3Cu(OH)2CuCl2y + 6H+ + 8e
a) E = 0.461 0.0443 pH 0.0148 log(Cl)


The author is grateful to his parents, who taught him
the dignity of hard work, and to his wife, who made life
bearable when things were not going according to plans.
IV


57
probably Cu20 and/or CuO, which formed under stagnant con
ditions within the residual gold sponge. Higher tempera
tures would favor precipitation of copper oxides and the
formation of denser film structures, thus accounting for
the observed inverse temperature dependence of the reaction
kinetics.
The concentration of metal ions in solution has also
97 981
been used to calculate metal-dissolution currents.
The results of these investigations are discussed in
the section on electrochemical investigations.


Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
THE DEZINCIFI CATION OF ALPHA AND BETA BRASSES
By
Robert Henry Heidersbach, Jr.
December, 1971
Chairman: Dr. Ellis D. Verink, Jr.
Major Department: Metallurgical and Materials Engineering
The mechanisms of dezincification of single-phase
alpha and beta brasses were studied using x-ray diffraction,
electron microprobe, metallographic, atomic absorption, and
electrochemical techniques.
The two mechanisms of dezincification which had been
previously reported, (1) dissolution of both alloy constit
uents followed by redeposition of the more noble species,
and (2) the selective removal of the less noble constituent,
were found to be operative under certain conditions of po
tential and pH for both alpha and beta brasses.
An electrochemical explanation of the circumstances
under which dealloying can be expected to occur was devel
oped based on the use of Pourbaix diagrams.
xi


136
Zn-H^O System [extracted from Zoubov and Pourbaix^^J ]
Eg, if
5 Zn + H20 = ZnO + 2H + 2e
a) E = -0.439 0.0591 pH
6 Zn++ + H?0 = ZnO + 2H+
a) log(Zn++) = 10.96 2 pH
7 ZnO + H20 = HZnO2 + H+
a) log(HZnO 2) = -16.68 + pH
Zn = Zn++ + 2e
a) E = -0.763 + 0.0295 log(Zn+f)
9


98
brass. In unstirred cells the copper appeared to have re
mained at or near the sample surface, since no copper was
detected in the bulk solutions, after any of the potentio-
static beta brass tests.
Figure 37 shows the surface of a sample of beta brass
which was held in a sample holder such as is shown in
Figure 4. The sample was exposed for a short-term test of
2-3/4 hours at +0.050Vg^g. The reddish copper crescent in
the upper left of the photograph illustrates the effects of
stirring on dezincification. That portion of the sample sur
face which was sheltered from stirring by the configuration
of the sample holder (i.e., a relatively stagnant area) has
a dezincified appearance due to the deposition of copper.
Where stirring was effective (the balance of the surface)
detincification did not occur.
The above photograph, coupled with the observations on
alpha brass discussed above, can be taken as further evidence
that dezincification can occur by an electrodeposition mecha
nism in certain potential ranges.
Figure 38 shows the configuration of a typical poten-
tiokinetic scan for alpha brass. Plots of the zero current
potential on the return (downward) scan, E are shown by
the line numbered 55 in Figure 34 and in Appendix 8.
At this potential on downward potentiokinetic scans
the samples were observed to change to a reddish color


11
CORROSION CELL SCHEMATIC
MAGNETIC
STIRRER
Figure 3. Corrosion cell used in electrochemical
investigations.


50
of copper deposits provided by Bockris and Damjamovic. ;
Other deposits had the appearance of "long, curled-up wood
shavings." Figure 22 is a nondispersive x-ray pattern of
the formation shown in Figure 21, and this pattern clearly
shows the deposit to be metallic copper. The sample was ex
posed to hydrochloric acid in a Pyrex reaction kettle, and
the silicon and chlorine peaks are probably due to contami
nation on the surface of the deposit.
One of the deposits was pulled from the sample surface
with a pair of tweezers and made into an x-ray powder sample.
The diffraction pattern obtained from this deposit confirmed
that the deposits were metallic copper.
Table 1 summarizes the weight change data from the
sample discussed above. Figure 23 is a photomicrograph of
this sample and shows that the sample did dezincify as well
as provide sites for the deposition of copper. No deposits
appear in photomicrographs of this sample because they were
removed for the x-ray diffraction and weight-change analyses
described above.
The above results clearly show that large surface-area-
to-cell-volume ratios and the addition of samples to solutions
already containing dissolved copper are situations to be
avoided. All further tests were conducted on a one-sample-
per-cell basis to avoid the effects described above.


42
Figure 18. Dezincification plug in 70-30 alpha brass
exposed for 79 days in IN NaCl at room
temperature. 200X


51
Figure 22.
Nondispersive x-ray analyzer pattern of
deposit shown in Figure 21.


6
limnersion Test Apparatus
Figure 1 is a diagram of the immersion test cell used
in this investigation. It is similar in many respects to
that recommended by National Association of Corrosion Engi
neers Standard TM-01-69, "Laboratory Corrosion Testing of
Metals for the Process Industries."^0)
The cell is constructed of a Pyrex* glass resin reac
tion kettle held together by an external metal clamp. All
fixtures inserted into the cell also are made of Pyrex,
with the exception of the Teflon**-coated stirring bar.
Samples are suspended from a sample holder as shown in
the diagram. In tests requiring an oxygen-free environment,
liquid in the cell is sparged with dried and purified argon
which enters the cell through a fritted-glass diffuser im
mersed in the liquid. Gas is passed out of the system by
way of a Liebig-type reflux condenser which is backed up by
a liquid trap to prevent solution loss or contamination. The
volume of solution within the cell was measured at the start
and finish of each test. Losses did not exceed 10 ml from
an initial volume of 750 ml in tests of up to 30 days. This
corresponds to a maximum change in solution volume of 1.3
percent. The normality of acids being used was also checked
*Registered trademark, Corning Glass Company.
**Registered trademark, E. I. duP-ont de Nemours and
Company, Inc.


ZINC DISSOLVED IN 48 HOURS gm
80
2.6 2.8 10
1000/T K~1
Figure 32. Arrhenius plot of zinc dissolution rate
1000/T. Scatter bands show the limits
cisin for the original measurements.
versus
of pre-


CONCLUSIONS
On the basis of an analysis of the dezincification of
alpha and beta brasses under free corrosion conditions and
on the basis of electrochemical behavior of these alloys
it may be concluded that:
1. X-ray diffraction and electron microprobe data
indicate that a selective removal process for the dezincifi-
cation of brasses is operative, at least under certain con
ditions of potential and pH.
2. Optical evidence and electrochemical evidence indi
cate that a redeposition mechanism for dezincification can
occur under suitable conditions within a particular range of
potential and pH.
3. Both operative mechanisms can be observed to have
taken place on the same specimen. This can be explained in
several ways:
a) The selective-removal process may actually
be one of relative kinetics, i.e., both metals under
go dissolution, but the zinc dissolves much faster,
in a certain well-defined range of potential and pH,
than copper.
b) Conditions at the metal-solution interface
are altered by the presence of metal ions in solution.
116


Ill
The high dezincificati.on factors for beta brass under
free-corrosion conditions can be explained as due to the
electrochemical potential of beta brass relative to that of
alpha brass under conditions where these brasses undergo
dissolution. Tanabe^"*'^ and Sugawara and Ebiko^^1- report
that dealloyed copper alloys quickly become covered with a
surface layer of copper which, if the ohmic resistance of
the layer is low, presents a copper surface to the bulk solu
tion which is at the free corrosion potential of the alloy.
n 13")
This potential, according to the data of Wilde and Teterin,'
is lower for beta brass than it is for alpha brass. Thus the
potential, or "driving force," difference between the deposi
tion potential of copper from copper chloride solutions
(approximately +0.200Vgj,^ for 0.1M chloride solutions) and
the potential of the copper surface layer on the brass sur
face is greater for beta brass than it is for alpha brass.
Sugawara and Ebiko^0^) point out the difficulty of
determining the potential of a dealloyed metal at the corro
sion interface, which is behind a resistance barrier due tc
stagnant solution conditions and the presence of the de
alloyed layer. Nonetheless, the concept of a Pourbaix dia
gram for brasses seems to explain experimental observations
and provides a basis for predicting conditions under which
dealloying is likely to occur.
The Pourbaix diagram approach also explains the effects
of stirring on dealloying. LaQue has observed, in salt water


9
Electrochemical Test Apparatus
The electrochemical hysteresis technique was used for
experimental determination of the corrosion behavior of al
loys as a function of potential and pH. The equipment,
techniques and theories upon which investigations of this
type are based have been discussed at some length in other
. (70-73)
reports.v J
The equipment is described schematically in Figure 2
and listed in Appendix 1. The scanning potentiostat applies
a continuously varying potential to the sample over a cer
tain range. The differential amplifier isolates the corro
sion cell from the recording equipment and eliminates ground
loops. The log converter allows the corrosion current to be
plotted as a logarithm on the x axis of the x-y recorder,
while the sample potential is plotted linearly on the y axis.
An automatic step switching apparatus'- J extends the range
of the log converter and allows the continuous recording of
corrosion current densities from 5 x 10 to 10 amps/cm .
The low-pass RC filter, consisting of a 100 microfarad
capacitor in parallel with a 25 K-ohm potentiometer adjusted
to 10 K-ohms, reduces electrical noise in the system to neg
ligible values.
A schematic diagram of the corrosion cell is seen in
Figure 3. The buffered electrolyte is vacuum deaerated
prior to transference to the corrosion cell. The solution
is continuously purged with hydrogen during the run via the


81
. i f 10 31
electrochemical metal dissolution (10 Kcal gin-mole ~) ^
the dissolution of copper-nickel alloys as reported by Rubin
(5-10 Kcal gm-mole1), or the dissolution of copper reported
, (102)
by Halperin.v
Kofstad points out the difficulty of ascribing physical
interpretations to experimentally determined activation
(104)
energies.v J
Several conclusions can be drawn from the atomic-ab
sorption data. There can be no doubt that at least some
copper enters solution from both alpha and beta brass when
freely exposed under conditions used in these tests. This
r 4 4 >
contradicts the results of Marshakov and coworkers^ who
reported that no dissolved copper was detected in solution
for their experiments with beta brass in 0.SM NaCl and in
0.5M HC1.
There is evidence for a linear dissolution rate for
zinc from both alpha and beta brasses (see Figures 25, 26,
29 and 51). This would support the metallographic observa
tions of Langenegger and Robinson.
The irregular dissolution kinetics for copper seen in
Figures 25, 26, 27 and 30 are believed to be due at least
in part to the effects of the morphology of the "spongy"
dezincified surface on the corrosion processes.
A comparison of the data obtained from alpha and beta
brasses confirms that beta brass corrodes much more rapidly
than does alpha brass in hydrochloric acid.


2.0
0 1 1 1 1 1
0 24 48 72 96 120
HOURS
Figure 27.
Dezincification factors, Z, for alpha brass in 5N HC1
at 98.50C. Scatter bands show the limits of precision
for the original measurements.


APPENDIX 4
BETA BRASS INGOT PREPARATION
Beta brass alloys were prepared from 99.99 w/o pure
copper and pure zinc stock purchased from the American
Smelting and Refining Company. The procedure which follows
f44)
is similar to that described by Marsbakov and coworkers. J
Pieces of copper and zinc were cleaned in concentrated
HNO^ followed by rinsing in 1/1 HNCU-I^O, in distilled
water, and finally in acetone after which they were dried
under a hot-air blower.
Charges weighing approximately 150 grams were placed
in 19 mm O.D. Vycor* vials which then were evacuated and
sealed.
The encapsulated charges were placed in a standard
heat-treating furnace and the temperature of the furnace
was brought up to 1000C, which is approximately 120C above
the liquidus for the alloy (see Figure 5). After the alloys
had been molten for at least four hours, the capsules were
removed from the furnace and turned end-over-end at least
twice in order to stir the liquid. The capsules were then
placed back in the furnace and kept at 1000C for another
*Trade name, Corning Glass Company.
127


Table 2
Atomic Absorption Data for Alpha Brass in 5N HC1 at 90.35C
Total
weight
Time
(hours)
Amount in
(ppm)
solution
(wt) a
Weight
removed
Total
weight
removed
that has
been in
solution
Amount
left in
solution
Z
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
12
19
12
12
9
0.48
0.3
0.5
0.3
14
9
14
8.7
1.5
24
25
15
19
11
0.62
0.4
1.1
0.68
19
12
18
11
1.4
36
27
16
21
12
0.69
0.4
1.8
1.1
22
13
20
12
1.4
48
35
18
26
14
0.88
0.4
2.7
1.5
29
15
25
13
1.2
60
38
22
28
16
0.94
0.6
3.6
2.1
31
18
27
16
1.4
72
45
26
34
20
1.1
0.6
4.7
2.7
37
22
33
19
1.3
84
42
28
32
21
1.01
0.7
5.8
3.4
37
24
31
20
1.5
96
45
32
34
24
1.1
0.8
6.9
4.2
39
27
33
23
1.6
108
46
34
35
26
1.2
0.8
8.1
5.1
42
30
34
25
1.6
120
48
36
36
27
1.2
0.9
9.3
6.0
44
30
34
26
1.6
aAl 1
weights
in
10"
3
gm.


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of
Professor of Metall
Materials Engineering
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Robert T. Dello:
Professor of Metallurgical and
Materials Engineering
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Associate Professor of
Metallurgical and Materials
Engineering
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Jofir/ J. Hren-
Associate Professor of
Metallurgical and Materials
Engineering


LIST OF FIGURES
Figure Page
1 Immersion test cell 7
2 Circuit diagram of equipment used
to polarize specimen and automatically
record corrosion current density as a
function of potential 10
3 Corrosion cell used in electrochemical
investigations 11
4 Assembled and exploded views of
sample holder 13
5 Copper-zinc phase diagram 19
6 Copper-gold phase diagram 20
7 The lattice parameter of alpha brass
as a function of the atomic percent
zinc 2 2
8 (HI) peaks of a mixture of 70-30
brass filings and copper filings .... 25
9 (HI) peaks from sample of 70-30
brass dezincified for 20 days in
5N HC1 at 75C 26
10 (HI) peaks from sample of 70-30 brass
dezincified for 30 days in 5N HC1
at 75C 27
11 Diffraction pattern from sample of
beta brass dezincified for two days
in 5N HC1 at 75C 29
12 Zinc intensity profile from a
sample of alpha brass dezincified
for 79 days in IN NaCl 34
13 Zinc intensity profile from a
sample of alpha brass dezincified
for 10 days in 5N HC1 at 75C 35
viii


APPENDIX 8
EQUATIONS USED IN CONSTRUCTION OF
POTENTIAL VERSUS pH DIAGRAMS FOR THE Cu-C1-H20 SYSTEM
AND THE Zn-H20 SYSTEM
Cu-Cl-H?0 System [extracted from Van Muylder, deZoubov
and Pourbaix^l23)]
Eg. ft
12
2Cu +
H-0 Cu20 + 2H+ +
2e
a)
E = 0.471 0.0591
pH
(no hydrated oxides)
b)
E = 0.572 0.0591
pH
(assumes hydrated
Cu70)
14
cu2o +
H20 = 2CuO + 2H+ +
2e
a)
E = 0.669 0.0591
pH
b)
E = 0.568 0.0591
pH
(assumed hydrated
but non-hydrated
Cu20
CuO)
16
2CuO +
H20 = Cu203 + 2H+
+ 2e
a)
E = 1.648 0.0591
PH
^Equation numbers refer to circled numbers on diagrams,
Figures 34, 35 and 36.
133


3 09
Table 12
Approximate Copper and Zinc Concentrations
in Solutions Where Copper Deposits
Were Observed on Alpha Brass
Temperature
Copper
Zinc
of cell
(ppm)
(ppm)
50 C
1,600
2,600
75C
500
5,500
100C
5,900
12,200


APPENDIX 6
ANALYSIS OF BETA BRASS INGOTS
USED IN THIS INVESTIGATION
Investigations
Ingot
number
Analysisa
(w/o)
in which
ingot was used
Cu
Zn
23
53.10
46.90
Electron microprobe
36
52.15
47.85
X-ray diffraction,
electron microprobe
38
52.56
47.44
Free corrosion
potential
44
52.22
47.68
Atomic absorption
45
51.SI
48. OS
Potentiostatic
56
51.80
48.20
Potentiostatic
58
53.37
46.63
Potentiostatic
All analyses
described in
were performed using the procedure
Appendix 5.
131


INTRODUCTION
Dealloying is a corrosion process whereby one constitu
ent of an alloy is preferentially removed from the alloy,
leaving an altered residual structure.^
While dezincification, the loss of zinc from brasses,
('2-12')
is the most commonly experienced form of dealloying, J
other examples have been reported in practice. These in
clude loss of nickel, aluminum ^ ^ and
tin^^>24) £rom COpper alloys; iron from cast iron;
nickel from alloy steels*-^) an cobalt from Stellite.
Since the phenomenon was first reported in 1866, the
literature has been filled with reports of research efforts
aimed at clarifying the mechanisms of dealloying. Nonethe
less, there still is no general agreement as to the detailed
mechanism involved. One group contends that the entire al
loy is dissolved and that one of its constituents then is
r23 27-401
replated from solution.v J Another contends that
one species is selectively dissolved from the alloy, leav-
r15 41-471
ing a porous residue of the more noble species.v J
Still others believe that both mechanisms take place.(1^,48 54)
A number of literature surveys have appeared,and these
summarize the situation up to the time they were writ
ten. C1>49>55-58)
1


32
Present Work
As the above discussion shows, it would be extremely
difficult, if not impossible, to obtain meaningful quanti
tative analyses through electron microprobe analysis of a
region having a large, and changing, number of vacancies.
This problem has been discussed by Anusavice^^ who, be
cause of his interest in eliminating vacancy and pore forma
tion in high-temperature diffusion couples, was able to
apply pressure and significantly reduce porosity formation.
The technique that was decided upon in this study was
to use the University of Florida's Acton Electron Micro
probe, operated under conditions which would produce a mini
mum beam size, to obtain point counts across the region
where zinc concentration was found to change. These regions
were identified either by using a chart recorder or, as ex
perience was obtained in identifying precise regions of in
terest, by using the light-optics microscope attachment on
the microprobe.
f64 791
The published data of Pickering^ 1 and of Forty and
Humble indicated that a diffusion zone, if it were to be
found, would be rather narrow, of the order of lOy or even
less. Results in these laboratories confirmed these observa
tions and perhaps explain why Sugawara and Ebiko,^^) who
used a chart recorder and a relatively fast scan rate, were
unable to find a diffusion zone.


28
two peaks on the corroded samples is higher than would be
obtained due to overlap of the "tails" of the copper and
alpha brass peaks. However, this intensity would probably
be undetected on a standard powder pattern obtained by film
methods.
If dezincification of beta brass occurs by a volume-
diffusion mechanism, the diffraction pattern obtained from
a powder of dezincified beta brass could be expected to show
scattered intensity characteristic of FCC alpha brass hav
ing a lattice parameter different from that of FCC copper.
Figure 11 shows a diffraction pattern obtained from a sample
of beta brass dezincified for two days in 5N HC1 at 75C.
The peak at 43.3 degrees is due to a superposition of beta
brass (110) scattering and copper (111) scattering. The
broad peak at 42.4 degrees is characteristic of alpha brass
having approximately 36 w/o zinc.
The above-mentioned figures thus give x-ray diffraction
evidence which supports a selective-removal mechanism for
the loss of zinc from both alpha and beta brasses when ex
posed to 5N HC1 at 75C.
No unidentified peaks were detected in the powder pat
terns of any of the samples shown above; thus, the possi
bility of this scattering being due to an extraneous source
is eliminated.


APPENDIX 1
EQUIPMENT USED IN ELECTROCHEMICAL TESTS
1. Magna Anatrol Potentiostat Model 4700M with attached
Linear-Scan Model 4510.
2. Hewlett-Packard Sanborn Differential Amplifier Model
8875A.
3. Hewlett-Packard Logarithmic Converter Model 75.6IA.
4. Hewlett-Packard Moseley X-Y Recorder Model 7000A.
5. Corning Calomel Reference Electrode.
6. Keithley Electrometer Model 602B.
7. Beckman pH Meter Expandomatic Model SS-2.
8. Hewlett-Packard Moseley Model 680 Strip Chart Recorder.
9. Duo Seal Vacuum Pump Model 1405H.
10. Acton Model MS64 Electron Microprobe.
11. Cambridge Stereoscan Scanning Electron Microscope
with Ortec Energy-Dispersive Si(Li) X-ray Energy
Analysis System.
12. Heath Malmstadt Enke Model EU-703-D with Techtron
Type AB51 Slot-type Burner.
122


ELECTRON MICROPROBE
The electron microprobe, with its unique capability
for chemical analysis of very small regions, has been used
by other researchers in an attempt to show the "diffusion
gradients" to be expected from a selective-removal mechanism
for dealloying. (-64,79,86^
Birks describes the electron microprobe as having the
capability of giving an x-ray spectrochemical analysis of
areas between 0.1 and 3y in diameter. phe principal
factors affecting the size of the region contributing x-rays
to the spectrum are the electron beam size and the density
of the sample under investigation.
Any quantitative method for x-ray spectrochemical analy
sis requires comparison with standards having, as nearly as
possible, the same physical characteristics as the sample
under investigation.If a selective removal of zinc
atoms occurs during dezincification, the atomic sites former
f 791
ly occupied by these atoms will become vacancies. J As
there is no way of reproducing these vacancies, with their
subsequent effects on density and surface roughness, in a
calibration standard, it is misleading to attempt to assign
quantitative chemical analysis values to electron microprobe
data from dealloyed specimens.
30


140
54. F. W. Fink, Transactions of the Electrochemical
Society, 7_5, 441 (1939) .
55. M. N. Desai, J. D. Talati and A. M. Trivedi, Journal
of the Indian Chemical Society, 38[, 565 (1961).
56. L. Piatti and R. Grauer, Werkstoffe und Korrosion,
14_, 551 (1963).
57. M. C. Steele, Australasian Corrosion Engineering,
p. 13 (January, 1963).
58. E. D. Verink, Jr., and R. Heidersbach, Jr., "Eval
uation of the Tendency for Dealloying in Metal Systems,"
accepted for publication in an ASTM Special Technical
Publication on Localized Corrosion.
59. R. B. Abrams, Transactions of the American Electro
chemical Society, 42_, 39 (1922); discussion by G. D.
Bengough and R. May.
60. C. L. Bulow, in Corrosion Handbook, H. H. Uhlig,
editor, John Wiley and Sons, New York, 1948.
61. I. G. S. Falleiro and A. Pieske, Metailurgia, 26^ [146],
21 (January, 1970).
62. E. E. Langenegger and F. P. A. Robinson, Corrosion,
24, 411 (1968); ibid. 2_5 59 (1969).
63. H. W. Pickering and C. Wagner, Journal of the Electro
chemical Society, 114 698 (1967) .
64. H. W. Pickering, Proceedings of the Conference on the
Fundamentals of Stress Corrosion Cracking, Columbus,
Ohio (1967), NACE, R. W. Staehle, editor, 1969, p. 159.
65. F. P. A. Robinson and M. Shalit, Corrosion Technology,
11 (April, 1964).
66. NACE Standard TM-01-69, "Laboratory Testing of Metals
for the Process Industries," published as an insert
to Materials Protection, _8 [5] (May, 1969).
67 International Critical Tables of Numeiical Data,
National Academy of Sciences, Maple Press, York,
Pennsylvania, 1929.
68. W. J. Muller, H. Freissler and E. Plettinger, Zeit-
schrift fur Elektrochemie, 42_, 366 (1936).
69. G. D. Bengough and R. May, Journal of the Institute
of Metals, 32, 81 (1924).


85
which formed on the sample surfaces. Similar results were
reported by Sugawara and Ebiko^^-' for alloys in the same
composition range. Alloys having 47 w/o zinc (beta brass)
or more were reported as having anodic polarization curves
which became more like that of zinc as the zinc composition
increased.
Marshakov and coworkers^^ stated that "The anodic
behavior of alloys is determined by the rate of dissolution
of the noble component as the slowest stage." They investi
gated single-phase alpha and beta brasses as well as duplex
alpha-plus-beta brasses.
Pickering and Byrne investigated the effects of a
dealloyed sponge on the surface of a sample by holding
copper-gold alloys at dealloying potentials for varying
periods of time and then "jumping" the potential to a more
noble level. They showed that below a certain "critical
potential" the copper-dissolution current from copper-gold
alloys was dependent on the rate of copper solid-state dif
fusion from the alloy. Above the critical potential the
copper-dissolution current became strongly potential-depen
dent. This critical potential was shown to be composition-
dependent for the copper-gold alloy system, which is a homo
geneous alloy system over the composition range investigated,
as well as for copper-zinc alloys, J where phase changes
could be interpreted as causing shifts in the critical
potential.


93
"noncorrosion" seems appropriate. Alloys which are rich
in one component (e.g., 70-30 Cu-Zn or 90-10 Cu-Ni) tend
to have many features in common with the diagram for the
major component (in this case, copper). The "immunity"
line, however, appears to have a somewhat different signifi
cance for alloys than for pure metals. While the kinetics
of alloy dissolution for alpha brass are slow below the
alloy "immunity" line, this does not rule out the possibil
ity of dealloying. Referring to Figure 37, the "immunity"
line for 70-30 Cu-Zn in 0.1M chloride was about O.OOOVg^g.
Between this potential and approximately -0.940Vg^g (line
#9) there is a theoretical tendency for the selective re
moval of zinc from the alloy.
At potentials more positive than about O.OOOVg^, both
constituents of the 70-30 Cu-Zn alloy go into solution.
(97)
Pickering and Byrnev report that "two modes-of dissolu
tion may occur depending on the potential: preferential
and simultaneous. Preferential dissolution changes gradu
ally with potential from virtually no dissolution of the
more noble metal to simultaneous dissolution of both com
ponents." Pickering and Byrne observed this change in be
havior between +0.050VCUC and +0.200VCUC in NaoS0. solu-
tions at pH 5. There were no chloride ions in their solu
tions, but the potential-pH conditions they described fit
into the cross-hatched area on Figure 36.


EXPERIMENTAL PROCEDURE
The possibility of bias of results due to the testing
method used was discussed in the introduction to this dis
sertation. Immersion testing of brass samples in environ
ments known to produce dezincification was chosen as the
laboratory test method most suited for a study of the mecha
nism of dealloying.
Electrochemical tests were employed in later stages of
this investigation, after the mechanism studies were com
pleted. These tests were intended to define the conditions
of potential and pH under which particular modes of dealloy
ing of copper-zinc alloys could be expected to occur.
5


47
Figure 19. Copper deposits on the surface of dezincified
alpha brass sample. Sample was exposed to
5N HC1 for 20 days at 100C. 50X


65
Table 6 (Extended)
Total
weight
removed
Total
weight
that has
been in
solution
Amount
left in
solution
Z
Cu
Zn
Cu Zn
Cu
Zn
26
1,100
780 32,000
750
31,000
44.6
110
4,200
2,400 95,000
2,300
91,000
42.6
260
9,200
4,600 150,000
4,440
140,000
36.5
510
16,000
7,800 220,000
31.3


104
all instances this potential was reached within one hour
of the start of the test. The two samples which did not
dezincify were both alpha brasses which came to potentials
of +0.090Vgjjg and +0.085Vg^g. This is in the potential
range where stirring was shown to prevent deposition of
copper on alpha brass specimens.


46
supported in the exposure cell. No effort was made to mask
off all but a certain area for exposure and the exposed area
was about 1S-1/2 cm^ in 750 ml of solution (~2.6 x 10 ^
cm /ml) which is quite high. The tests were conducted in
5N HC1 at varying temperatures. During these tests a number
of the samples appeared to have surface deposits which had
a metallic luster and grew larger with the passage of time.
These "deposits" were apparent on samples which liad been ex
posed from the beginning of the test as well as on samples
which were added when other samples were removed.
Figure 19 shows a photomicrograph of a cross sect ion
of one of these samples. The normal dezincified texture at
the bottom of the sample contrasts sharply with the appear
ance of the shiny surface deposits.
That these formations are, in fact, deposits is shown
in Figure 20 where a portion of the original brass surface
is visible in the photomicrograph.
Each sample in these tests was weighed before and after
the test. The weight-loss data did not follow any recogniz
able pattern. One sample in these tests actually gained
weight.
This particular sample, which was exposed for the last
ten days of a twenty-day test, had numerous surface deposits
which were copper colored and had a metallic luster.
Figure 21 is a scanning electron micrograph of one of
these deposits which fits the "ridge deposition" description


107
c co C.0591
E = E + =¡ log
'CuCl.
aCu(alloy)(aCl-^
for the reaction
(Eq. 2)
Cu + 2C1 = CuCl2 + e
(Eq. 3)
The terms of this equation can be separated
E = E + 0.059 log aCuC1 - 0.0591 log (aCu(alloy)(ac1*)
Variations in either the CuCl2 or the Cu(alloy) activity
will cause a change in the equilibrium potential for the
reaction. In other words, increases in the activity of the
complex copper ion will shift the equilibrium in the noble
direction. The existence of a copper alloy having a higher
copper activity than that of the original alloy will lower
the potential. Accordingly, copper is expected to be elec-
trodeposited from solutions at a given potential provided
the conditions of Eq. 1 are fulfilled. The log acu(a]i0y)
term varies only slightly for compositions of from 10-100
percent copper.
Equation 3 implies that copper will be deposited from
unstirred 0.1M chloride solutions at potentials below-
+0.200V
SHE
Scanning in the active direction from poten
tials more noble than +0.200VCLITr will result in the onset
of copper deposition at this potential. Thus copper depo
sition is to be anticipated from solutions containing cop
per ions (either as corrosion products or as added ions) at
potentials below +0.200Vc;^p in 0. 1M chloride solutions.


HOURS
Figure 28. Copper dissolution from beta brass in 5N HC1 at 67.50C.


Dedicated to my wife, Dianne Katherine.


'* f.
117
4. Copper enters solution during the dezincification
of both alpha and beta brasses under freely corroding con
ditions (no electrochemical stimulation) in 5N HC1 solution.
5. Linear dissolution kinetics were observed for the
dissolution of zinc from both alpha and beta brasses for
tests ranging up to 120 hours at temperatures up to approxi
mately 100c.
6. The dissolution kinetics for copper are irregular
based on atomic absorption data. This irregularity is be
lieved to be caused at least in part by the morphology of
the residual copper layer.
7. The activation energy measured for the dissolution
of zinc from beta brass during dezincification was approxi-
_ 1
mately 18 Kcal gm-mole This value is somewhat higher
than activation energies for electrochemical dissolution of
metals (approximately 10 Kcal gm-mole .
8. Because of the seeming irregularity of copper dis
solution kinetics it is considered meaningless to calculate
an activation energy for copper from atomic absorption data.
9. It is possible to predict domains of potential and
pH for each of the operative mechanisms of dezincification
by superposition of experimental potential versus pH dia
grams over the equilibrium Pourbaix diagrams for the con
stituent metals of the alloy.


TABLE OF CONTENTS (Continued)
Page
APPENDIX
1 EQUIPMENT USED IN ELECTROCHEMICAL
TESTS 122
2 ELECTROLYTES 123
3 CHEMICAL ANALYSES OF ALPHA BRASS .... 125
4 BETA BRASS INGOT PREPARATION 127
5 PROCEDURE FOR THE ELECTROGRAVIMETRIC
CHEMICAL ANALYSIS OF BINARY COPPER-
ZINC ALLOYS 129
6 ANALYSIS OF BETA BRASS INGOTS
USED IN THIS INVESTIGATION 131
7 ELECTROCHEMICAL SAMPLE PREPARATION ... 132
8 EQUATIONS USED IN CONSTRUCTION OF
POTENTIAL VERSUS pH DIAGRAMS FOR
THE Cu-C1-H20 SYSTEM AND THE
ZnO-H20 SYSTEM 133
BIBLIOGRAPHY 137
BIOGRAPHICAL SKETCH 145
vi


17
back. Miccrostop stop-off lacquer was used to coat the
back and sides of these samples leaving a circular exposure
surface. Examination of the samples after exposure revealed
no apparent leaks in the lacquer.


POTENTIAL vs S.H.E.
89
pH
Figure 33. Experimental potential versus pH diagram for
70 Cu 30 Zn in 0.1M Cl' at 25C.(77)


38
Figure 15.
Photomicrograph of sample shown in Figure 12.
Probe trace is indicated by the arrow. 500X


20
Figure 6. Copper-gold phase diagram (Metals Handbook,
1948 edition, p. 1171).


59
The exposure method used for these tests vas similar
to that reported by Fisher and Halperin.^8) Reaction ket
tles were used to expose samples, cut from the same ingot
2
and having the same surface area (2.38 cm for alpha brass,
2
1.53 cm for beta brass), in 750 ml of argon-sparged 5N HC1
at various temperatures. Twenty-five milliliter aliquots
of acid were removed from the cells at 12-hour intervals.
Each aliquot of sample was replaced with an aliquot of the
stock solution used to fill the cell. This caused a dilu
tion of 3.3 % (25 ml out of 750 ml total) each time a sample
was removed. Corrections were made for these dilutions in
the data which follow.
The aliquots were analyzed using a Heath Model EU-
703-D Atomic Absorption Spectrophotometer (see Appendix 1).
Both copper and zinc concentrations were determined for
each sample.
Data obtained from atomic absorption determinations
are tabulated in Tables 2 through 7. Atomic absorption
data are reported to only two significant figures consis
tent with the accuracy of the method.
Figure 24 shows an atomic-absorption calibration curve
for copper. This figure shows that copper can be determined
to at least the nearest 1/4 ppm with a reasonable degree of
certainty. However, the precision, or the relative accuracy
of the measurement when compared to the amount of metal


THE DEZINCIFICATION OF ALPHA AND BETA BRASSES
By
Robert Henry Heidersbach, Jr.
A Dissertation Presented to the Graduate Council
of the University of Florida
in Partial Fulfillment of the Requirements for th
Degree of Doctor of Philosophy
UNIVERSITY OF FLORIDA
DECEMBER, 1971


74
Figures 28 and 29 show the corrected amounts of copper
and zinc which have dissolved from beta brass at 67.50C
as a function of time. They are not plotted on the same
scale, as was done in Figures 25 and 26 for alpha brass,
because the dissolution rate for zinc (Figure 29) is much
greater than the dissolution rate for copper (Figure 28).
The reaction kinetics for both copper and zinc appear to be
linear after the first 24 hours.
Slight fluctuations in the copper or zinc dissolution
kinetics make disproportionately large changes in the magni
tudes of the dezincification factor Z. This is seen in the
positions of the calculated points in Figure 30. The values
shown in Figure 30 for beta brass are up to fifty times the
values plotted in Figure 27 for alpha brass. It is empha
sized here that the values for beta brass are finite in con
trast to the results of Marshakov and coworkers who reported
dezincification factors which were infinite for beta brass
because no dissolved copper was detected by them. This
again illustrates that Z has no fundamental significance
with regard to the mechanism of dealloying.
The effect of temperature on the dissolution kinetics
of zinc from beta brass is shown in Figure 31 where the
amount of zinc dissolved for beta brass is plotted versus
time for four different temperatures. There is an obvious
increase of zinc dissolution rate with temperature.


69
available for measurement, is lower for more dilute samples.
As an example, a 1/4 ppm deviation in a sample having 2 ppm
copper yields a 12-1/2 percent error, whereas in a sample
having 8 ppm copper the same 1/4 ppm deviation would produce
an error of only 3-1/8 percent. This type of precision
deviation is common to all atomic absorption methods. The
confidence limits shown on the figures which follow were
calculated by assuming a possible 1/4 ppm deviation for
the original instrument reading obtained on the sample in
question.
The Heath spectrophotometer is a single-beam instrument.
This means that no provision is made for automatically ad
justing for instrumenta] drift due to power deviations or
lamp fluctuations. it was found in practice that the
instrument would drift significantly in a short period of
time, thus negating the value of a calibration curve. For
this reason, a procedure was developed whereby each sample
was introduced into the absorption flame once to determine
the general range of the concentration of the element being
analyzed. Then standards were introduced into the flame to
confirm the concentration range of the unknown. The values
tabulated in Tables 2 through 7 were then determined by
introducing one standard, then the unknown, and then another
standard, so that each unknown reading was bracketed bv two
standard readings.


70
The limits of detection for the unit under the operat
ing conditions used in this investigation were approximately
0.15 ppm for zinc and approximately 0.25 ppm for copper.
Theoretical limits of detection for these elements by atomic
absorption are somewhat lower than this, but the solutions
analyzed had copper and zinc contents well above these lim
its, and most of them required dilution to be brought with
in the operating range of the instrument.
Figure 25 shows the dilution-corrected values for cop
per and zinc dissolved from alpha brass at various times
(see also Column 4 of Table 3). The zinc values obey linear
kinetics, but the coppei values fluctuate.
Copper and zinc dissolution from alpha brass at 98.50C
is shown in Figure 26. Once again, linear kinetics are
observed for zinc,and the copper values fluctuate somewhat
before they too become linear.
The dezincification factor, Z, that was introduced by
Marshakov and coworkers is plotted for this cell in
Figure 27. When the confidence limits are taken into con
sideration, Z for alpha brass is seen to be between 1 and
1.5 for this investigation. This is in general agreement
with Marshakov and coworkers. The usefulness of determining
Z is limited since the absolute magnitudes of the values of
Z are strongly influenced by the precision of chemical analy
sis methods in processes (such as dealloying) where the
quantities being analyzed are small.


37
Figures 15, 16 and 17, show' the carbon traces left on the
sample by the interaction of small amounts of diffusion
pump oil with the high-energy electron beam. The diffusion
zone occurs in each case at the very edge of the obvious
change in color as shown on the photomicrographs.
Figures 12, 13 and 14 are thus, to the author's knowl
edge, the first experimental plots of alloy composition ver
sus distance to be obtained from samples subjected to de
alloying under free corrosion conditions and which indicate
the existence of a diffusion zone.
The diffusion distances shown in Figures 12, 13 and 14
are typical of those found on the samples examined. Traces
were made across the edges of polished pieces of brass and
copper to determine the "distance" which would be recorded
due to beam overlap between one-micron settings of the
microprobe stage. Results indicated one intermediate point
between the intensity ratio of the metal and the background
reading on the nonconductive mounting material. A beam at
the intermediate-reading setting which was only partially
on the sample could explain the intermediate point. The
intermediate points of Figures 12, 13 and 14 cover an inter
val of at least six microns and are thus not due to overlap
of the beam at adjacent position settings.


52
Figure 23. Dezincified cross-section of sample shown
in Figure 21. 500X


34
Figure 12. Zinc intensity profile from a sample of
alpha brass dezincified for 79 days in
IN NaCl.


APPENDIX 7
ELECTROCHEMICAL SAMPLE PREPARATION
All samples were sliced into discs on a cut-off wheel
and ground at 120 grit on a belt sander. Samples were then
hand polished on 240, 320, 480 and 600 grit silicon carbide
paper, washed in distilled water,and dried under a hot-air
blower.
132


44
The difficulties of defining and comparing the service
conditions reported by various authors are additional draw
backs to arriving at a general mechanism based on optical
observations of in-service failures.
In 1965, V. F. Lucey reported on laboratory experiments
conducted in closed containers containing saturated copper
chloride solutions and containing excess undissolved cuprous
chloride.J It is hardly surprising that he observed cop-
f 921
per deposition from such a solution. J However, Lucey
neglected to point out that his experiments did not show
that a selective removal process could not occur under other
circumstances or, indeed, under the conditions described in
his papers.
The above discussion is not intended to indicate that
optical observations have no value in the determination of
a mechanism for dealloying but merely to point out that ex
periments should be carefully controlled and that limitations
of optical methods should be recognized. A concurrent ob
servation is that information which lends support to one
theory should not be misinterpreted as proof of the invalid
ity of a contrasting theory. Some, but by no means all, of
the authors discussed above have recognized this,while
others, unfortunately, have not.


POTENTIAL vs SHE.
92
Figure 36. 70 Cu 30 Zn alloy in 0.1M chloride solution.
Superposition of the experimental potential
diagram, Figure 33, onto Figures 34 and 35.


APPENDIX 2
ELECTROLYTES
pH (0.2)
Composition
3.
0
0.
100
mol ar
NaCl
0.
002
molar
HC1
0 .
005
molar
KHC8H44
4.
,1
0.
,100
molar
NaCl
0.
,088
molar
khc8h4o4
0.
,002
molar
NaOH
4.
,5
0.
,100
molar
NaCl
0.
,016
molar
NaOH
0.
,074
molar
KHC8H404
5.
,0
0.
,100
molar
NaCl
0,
,0 30
molar
NaOH
0,
, 060.
molar
kiic8h4o4
5.
.5
0.
, 100
molar
NaCl
0.
,038
molar
NaOH
0.
.052
molar
KHC8H44
5,
.9
0,
,100
molar
NaCl
0,
.042
molar
NaOH
0,
.048
molar
KHC8H44
7,
.1
0,
.100
molar
NaCl
0,
.046
molar
NaOH
0,
.045
molar
khc8h4o4
8,
. 3
0,
.100
molar
NaCl
0
.100
molar
NaHC03
0,
.001
molar
khc8h4o4
8,
.6
0
.100
molar
NaCl
0
.100
molar
NaHCOj
9
.0
0
.100
molar
NaCl
0
.100
molar
NaHCO,
0
.010
molar
NaOH
123


24
Present Work
All x-ray diffraction mechanism studies performed dur
ing this investigation were performed on powder samples ob
tained from discs of alpha brass and of beta brass exposed
to hydrochloric acid in the immersion apparatus shown in
Figure 1 and described in the section on experimental pro
cedure .
There are several problems associated with the use of
a diffractometer for investigating dezincification of alpha
brasses. If the peaks due to the original alloy occur too
close to the copper peaks, then "tails" of diffraction peaks
may overlap and be misinterpreted as being due to an alloy
of intermediate composition.
This is true of the diffraction patterns obtained using
Cu Ka radiation. The separation is less than one degree 20
between the (111) peaks of pure copper and of 70-30 alpha
brass. At higher angles the separation between peaks be
comes greater, but intermediate intensity, if present, would
be spread out also, and thus be harder to detect.
Figure 8 shows the (111) peaks from a sample obtained
by mixing filings of alpha brass with annealed filings of
copper. A slight increase of intensity due to the overlap
of the "tails" of the peaks is apparent.
Figures 9 and 10 show the (111) peaks obtained from
samples of alpha brass dezincified for 20 and 30 days, re
spectively, in 5N HC1 at 75C. The intensity between the


53
Table 1
Sample 6-19, Weight Information
Original weight
Final weight
Weight gained
5.2121 gm
5.3901 gm
0.1780 gm
Weight of sample with
deposits removed by-
tweezers and ultra
sonic cleaning 4.5055 gm
Weight of deposits
(by difference)
0.8846 gm


LATTICE PARAMETER A
22
Figure 7. The lattice parameter of alpha brass as a
function of the atomic percent zinc. (84)


87
Another advantage of the Pourbaix diagram, or potential
pH, approach to dealloying is that information obtained by
various authors in different solutions can be plotted in a
manner which provides meaningful correlations.


E LECTR0CHEM1CAL INVESTI GATIONS
Early electrochemical tests measured the potential of
brasses in solutions known to produce dezincification.
At that time it was not generally recognized that duplex
alloys, such as alpha-plus-beta brasses, would exhibit
microstructure-dependent electrochemical behavior.
In 1967, Joseph and Arce^^) showed that the corrosion
behavior of a 63 w/o Cu 37 w/o Zn brass was strongly de
pendent on structure. This particular alloy can have
either a single-phase alpha or a duplex alpha-pi us-beta
structure, depending on heat treatment, as revealed in the
copper-zinc phase diagram, Figure 5. Several recent elec
trochemical studies of dealloying have examined the behavior
of single-phase copper alloys and of more complex copper al
loys and have taken microstructure into account.?^^0^^^
Some researchers have used driven anodes as a rapid
means of producing specimens to be examined by non-electro-
chemical means.Others have used electrochemical ob
servations to arrive at conclusions as to the mechanism of
dezincification.^0^^^j,96, i 1 -) still others have tried
to define current/498'112 potentialJ44869798110113
or potential-pH ^ ^ ^ ^ ^ ** ^ conditions where dealloying
83


BIBLIOGRAPHY
1. R. Heidersbach, Corrosion, 2_4, 38 (1968).
2. D. B. Bird and K. L. Moore, Materials Protection, 1,
70 (1962).
3. J. T. Crenell and L. J. E. Sawyer, Journal of Applied
Chemistry, 12_, 1 70 (1962).
4. A. F. Blumer, Corrosion, 5, 144 (1949).
5. M. Matsuda, Journal of the Japanese Institute of
Metals, 2_6, 124 (1962) .
6. R. B. Niederberger, Modern Castings, 4_5, 115 (1964).
7. E. P. Polushkin and M. Shuldener, Transactions AIME,
161, 214 (1945).
8. A. L. Simmons, Metal Progress, 5_7, 496 (1950).
9. A. W. Tracy, ASTM STP, 175, 67 (1956).
10. Sj van der Baan, Corrosion, 6, 14 (1950).
11. W. H. J. Vernon, Transactions of the Faradav Society,
23, 170 (1927).
12. R. A. Wilkins, Mechanical Engineering, 58 [12], 57
(1936).
13. J. L. Henderson and C. L. Roadhouse, Journal of
Dairy Science, 23_, 215 (1940).
14. W. C. Stewart and F. L. LaQue, Corrosion, 8, 259
(1952).
15. F. Taylor and J. W. Wood, Engineering, 149, 58 (1940).
16. S. C. Britton, Journal of the Institute of Metals,
67_, 119 (1941).
17. L. W. Gleekman and R. K. Swandby, Corrosion, 17, 144t
(1961).
137


141
70. E. D. Verink, Jr., "Construction of Pourbaix Diagrams
for Alloy Systems with Special Application to the
Binary Fe-Cr System," Report to the Advanced Research
Projects Agency, June 10, 1970.
71.
P.
A.
Parrish, MS
Thesis ,
University of
Florida, 1970.
72.
R.
L.
Cus amano, MS
Thesis
, University of
Florida, 1971
73.
E.
D.
Verink, Jr.,
and M.
Pourbaix, "Use
of Electro-
chemical Techniques in Developing Alloys for Saline
Exposures," accepted for publication in Corrosion.
74. W. D. France, Jr., and R. W. Liety, "Improved Data
Recording for Automatic Potentiodynamic Polarization
Measurements," Research Publication GMR-762, General
Motors Corp., May 3, 1968.
75. Recommended Practice for a Standard Method for Making
Potentiostatic and Potentiodynamic Polarization Mea-
surements, Task Group 2, Potentiostatic and Potentio
dynamic Polarization, Section I, Subcommittee XI,
ASTM Committee G 1, April 10, 1968.
76. J. E. Reinoehl, F. H. Beck and M. G. Fontana, Corro
sion, _26 ] 41 (1970) .
77. W. C- Fort III, High Honors Project, University of
Florida, 1971.
78. J. I. Fisher and J. Halperin, Journal of the Electro
chemical Society, 103, 282 (1956).
79. H. W. Pickering, Journal of the Electrochemical
Society, 115, 143 (1968).
80. L. Graf, Metallwirtschaft, 11, 77 (1932); Zeitschrift
fur Metallkunde, £6, 275 (1949).
81. K. Hashimoto, T. Goto, W. Suetaka and S. Shimodaira
Transactions of the Japanese Institute of Metals, 6,
107 (1965).
82. R. Kleinbe :;er, H. Okuzumi and P. Perio, Metaux Corro
sion Industries, 3S_, 40 (1960).
83. H. W. Pickering, Journal of the Electrochemical
Society, 117, 8 (1970).
84. E. A. Owen and E. W. Roberts, Philosophical Magazine,
2!7 294 (1939).


96
alpha brass by the selective removal of zinc. Zinc ions
were present but no copper was detected at these potentials
either on the platinum auxiliary electrode or by atomic-
absorption analysis of the bulk solutions. Clearly selec
tive dissolution of zinc is the predominant process in this
potential range.
Certain of the samples exhibited a dark tarnish
(- 0.750VQjjr and -0.850Vg^p). These are presumed to have
undergone dezincification, although the tarnish film was
too thin for unequivocal identification by x-ray analysis.
Superposition of Pourbaix diagrams of copper and zinc
also was used to assess the dealloying behavior of beta
. \
brass. Although a complete experimental Pourbaix diagram
was not available for beta brass, enough data were at hand
to facilitate selection of appropriate potentials for poten
tiostatic studies in 0.1M chloride solutions at pH 4.
The results of these studies are shown in Table 9.
All samples exposed to potentials from +0.050Vg^p to
-0. SSOVc,.,^ were found to dezincify. The effects of stir-
one
ring were not as obvious as for alpha brass, but between
+0.050 and -0.150YgHg in stirred solutions copper was.depos
ited on the platinum auxiliary electrode. The lowest poten
tial at which copper was detected on the platinum auxiliary
electrode in stirred solutions was -0.150Vg^p. Evidently
the high zinc content in beta brass causes this potential
to be more active than was found to be the case for alpha


49
Figure 21.
Scanning electron micrograph of copper slab
protruding from the surface of a dezincified
alpha brass sample. 500X


21
range. The lattice parameter of alpha brass as a func
tion of zinc content is shown in Figure 7.
Pickering points out that by choosing an alloy system,
such as copper-gold, with elements having substantial dif
ferences in atomic diameter, the lattice parameter, and thus
the separation between peaks of the original alloy and those
f 791
of the corroded residue, will be increased. J
An alternative to this would be to start with an alloy
which, upon dealloying, would undergo a phase change as well
as a change in lattice parameter.
Of the large number of reports concerning diffraction
studies of dealloying, only two previous researchers have
concluded that their data supported a selective-removal
, (52,63,64,79)
mechanism.
Pickering subjected copper-gold alloys to electrolytic
dissolution and showed the appearance of an intermediate
peak which, as additional current was passed, increased in
intensity and moved closer to the position to be expected
for pure gold. The alloy peak decreased appropriately in
intensity.
A later investigation by the same author reported the
formation of new, intermediate, phases during the anodic
dissolution of gamma brass and of epsilon brass.^3) This
report substantiated the results of Stillwell and Turnipseed
who subjected epsilon brass to various corrosive media, and
whose x-ray diffraction results indicated the presence of
f 52')
intermediate phases in some of their experiments.^ }


APPENDIX 5
PROCEDURE FOR THE ELECTROGRAVIMETRIC
CHEMICAL ANALYSIS OF BINARY COPPER-ZINC ALLOYS
The follcming procedure is adapted from the procedures
recommended by the American Society for Testing Materials
and by G. H. Ayres, Quantitative Chemical Analysis.
-4
1. Use a balance accurate to 10 gm for all weighings.
2. Place a one-gram sample of the brass to be analyzed
in a 250 ml beaker.
3 A.dd 17 rr distilled 1-LO. 5 ml cone H0SO, and 3 ml
cone HNO., to the beaker in the order listed. This solution
will become hot and initiate dissolution of the brass.
4. Dilute the mixture to 75 ml and stir, using an in
ert magnetic stirring bar, until the brass is completely
dissolved.
5. Dilute to 200 ml.
6. Electrolyze at 1 amp (approximately 2 volts) for
8 hours using a preweighed platinum gauze electrode as the
cathode and a platinum wire placed at least 2 cm from the
cathode as an anode.
7. Add 20 ml of water to the solution. If no addi
tional copper deposit is apparent at the new solution level
after 15 minutes, assume the electrodeposition is complete.
12 9


114
WEIGHT PERCENT B
Figure 40. Theoretical domains for dealloying in a
given solution based upon the Nernst
equation.


31
The use of chart recorders to show diffusion gradients
can also be misleading. Both the response time of the chart
recorder and the scan rate at which the sample is swept
under the electron beam can affect the apparent fall-off
distance that is recorded on the chart.
Sugawara and Ebiko^^) and Pickering have pub
lished microprobe data obtained with chart recorders to show
relative zinc concentrations as a function of distance.
Sugawara and Ebiko concluded that their data gave no indi
cation of a concentration gradient along the interface be
tween corroded and uncorroded brass. Pickering reports a
region approximately 9y thick where the copper concentra
tion fell off in a copper-gold alloy which had been subjected
to electrochemical anodic dissolution.
Other researchers have used x-ray images to demonstrate
the occurrence of dealloying.None of these authors
used microprobe data to support arguments regarding the
mechanisms involved.


Figure 37. Beta brass held at +0.050VgHE for 2-3/4 hours.
Crescent shaped reddish region shows effect
of stirring action on dezincification.


OPTICAL METHODS
Early work on the dezincification mechanism relied
heavily on optical observations. Most of the research tools
used in this study were not developed at that time, but the
microscope and metallograph were available and were used.
The results obtained from microscopic investigation were,
however, subject to "opinion-type" interpretations, and re
searchers did not have the benefit of present-day knowledge
of crystal structure, grain growth, epitaxial electrolytic
deposition, the concept of an occluded cell, and other ideas
which are part of the present-day researcher's background.
In 1922 Abrams^ } suggested that dezincification
occurred when a membrane of some type was available to hold
dissolved copper in contact with the brass surface or when
a large excess of copper was present in the solution. His
experiments with copper chloride solutions led to further
work by Bengough and May,^^) an solutions of this type
soon became an accepted method of accelerated testing for
the susceptibility of alloys to dezincification.
Early researchers noted that dezincification could be
classified as occurring either in "layers" or in "plugs"
such as the one shown in Figure 18. These plugs were
41


Sample Number 55
Element
Weight Percent
Copper
69.82
Lead
<0.006
Iron
<0.004
Tin
<0.001
Nickel
<0.001
Manganese
<0.002
Silicon
<0.002
Aluminum
<0.001
Antimony
<0.01
Phosphorus
None detected
Bismuth
<0.0005
Arsenic
<0.003
Zinc
Remainder
Analyses were supplied by the manufacturer, Chase Brass
and Copper Company, Incorporated, who donated the material.
Sample 6 was used for the metallographic, x-ray, elec
tron microprobe, and potentiokinetic studies. Sample 55 was
used for atomic absorption and potentiostatic studies.


CQ O
7
A
Figure 1. Immersion test cell. A sample holder,
B sample, C immersion heater, D stir
ring bar, E J-tube thermostat, F gas
inlet diffuser, and G gas outlet through
reflux condenser.


143
102.J. Halperin, Journal of the Electrochemical Society,
100, 421 (1953).
103. K. J. Vetter, Electrochemical Kinetics, Academic
Press, New York, 1967.
104. P. Kofstad, High Temperature Oxidation of Metals,
John Wiley and Sons, New York, 1966.
105. H. H. Willard, L. I. Merritt and J. A. Dean? Instru
mental Methods of Analysis, Van Nostrand Reinhold
Company, New York, 1965.
106. F. Daniels and R. A. Alberty, Physical Chemistry,
John Wiley and Sons, New York, 1961
107. G. Joseph and M. T. Arce, Corrosion Science, 7_, 597
(1967).
108. N. Ohtani, Journal of the Jaoanese Institute of
Metals 30, 729 (1965) .
109. S. Sugawara and S. Shimodaira, Journal of the Japanese
Institute of Metals, 3C(, 765 (1966).
110. Z, Tanabe, Corrosion Science, 4_, 413 (1964).
111. II; Gerischer and H. Rickert, Zeitschrift fur Metall-
kunde, 4j6 681 (1955).
112. H. W. Pickering, Journal of the Electrochemical
Society, 115, 693 (1968).
113. B. Wilde and G. A. Teterin, British Corrosion Journal,
2, 125 (1967).
114. R. M. Latanision and R. W. Staehle, Proceedings of
the Conference on the Fundamentals of Stress Corro-
s i on Cracking Columbus, TTio (19 6 7), NACE R. W.
Staehle, editor, 1969, p. 214.
115. E. D. Verink, Jr., and P. A. Parrish, Corrosion, 26,
214 (1970).
116. T. J. Lennox, M. H. Peterson and R. E. Groover, Mate
rials Protection and Performance, 10^ [8], 31 (1971).
117. D. C. Vreeland and G. T. Bedford, Materials Protection,
9, [8] 31 (1970).
118. W. B. Brooks, Corrosion, 24, 171 (1968).


Table 10
Alpha Brass Free Corrosion Potential Tests
Solution
Final Potential
Dezincification?
Reference Authors
66 g/Z MCI
+13.4 g/Z CuCl2
+0.090Vshe
No
Langenegger and Robinson^^)
66 g/Z HC1
+15.8 g/Z CuCl
+0.0S5VsHE
No
Langenegger and Robinson^^
1M H2S04
+0.058VgEE
Yes
Stillwell and Turnipseed
30 g/Z NaCl
+32 g/Z NiCl2
-0.290Vshe
Yes
Falleiro and Pieske^"^
102


Table 4
Atomic Absorption Data for Beta Brass in 5N HC1 at 59.60C
T ime
(hours)
Amount
(ppm)
in solution
(wt) a
Weight
removed
Total
weight
removed
Total
weight
that has
been in
solution
Amount
left in
solution
Z
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
Cu
Zn
12
0.75
24
57
1,800
1.9
60
1.9
60
56
1,800
54
1,700
35
24
1.5
63
110
4,700
3.8
160
5.6
220
110
4,800
110
4,600
46
36
2
100
150
7,900
5
260
10.6
480
160
10,000
140
7,600
70
48
2.5
150
188
11,000
6.2
380
17
860
200
12,000
180
11,000
66
60
3.8
220
280
17,000
9.5
560
26
1,400
300
17,000
280
16,000
64
72
4
280
300
20,000
10
690
36
2,100
320
22,000
74
aAll
weights
in
105
gm.


HOURS
Figure 29.
Zinc dissolution from beta brass in 5N HC1 at 67.50C.


Ni
Ln
Figure 8. (Ill) peaks of a mixture of 70-30 brass filings and copper
filings. Brass- peak is to the left.
.I


66
Table 7
Atomic Absorption Data for Beta Brass
in 5N HC1 at 99.85C
Time Amount in solution Weight
(hours) (ppm) (wt)a removed
Cu
Zn
Cu
Zn
Cu
Zn
12
34
1,200
2,500
94,000
84
3,100
24
150
3,100
11,000
230,000
380
7,800
36
380
4,600
28,000
350,000
940
11,500
48
820
5,500
62 ,000
410,000
2,100
14,000
aAll weights in 10 ^ gm.


40
I
Figure 17. Photomicrograph of sample shown in Figure 14.
Probe trace is indicated by the arrow. 250X


DEGREES 2G
ts)
O'
Figure 9. (Ill) peaks from sample of 70-30 brass dezincified for
20 days in 5N HC1 at 75C.


12
1751
gas diffuser as suggested by ASTM committee Gl. 3 A
bright platinum screen serves as the auxiliary electrode.
The current from the potentiostat to the auxiliary electrode
is measured as a potential across a precision resistor se
lected to provide the required logarithmic converter input
voltage. A Luggin-Haber probe connected to a standard calo
mel electrode is used to measure the sample potential. The
thermometer is used to indicate the temperature of the elec
trolyte. The solution is stirred using a magnetic, water-
powered stirrer. The cell itself is made of Pyrex glass
with a Teflon lid bolted to a polycarbonate Van-Stone back
ing ring which makes the system airtight.
The sampleholder is shown in Figure 4. The main body
of the holder is constructed of Teflon so that it will not
react with the test solution. Copper parts encased in the
Teflon allow electrical contact with the sample. A poly-
2
carbonate nut fastens the sample in place and allows 1 cm
to be exposed to the electrolyte. The Teflon gasket avoids
leakage and minimizes crevice effects.
Care must be exercised when choosing buffered electro
lytes to insure that solution ions will not form complexes
with the metal ions from sample dissolution. The electro
lytes used in these studies are listed in Appendix 2. All
solutions used in this portion of the investigation had a
0.1 M chloride content. The effects of copper-chloride com
plexes were discussed in the section on immersion testing.


3
arsenic protecting alpha brass but not the beta phase of
duplex alloys have remained controversial.^^However,
the addition of small amounts of arsenic, or of antimony or
phosphorus, which have similar effects, has become a stan-
f 11
dard means of inhibiting dezincification in alpha brasses.^ 1
No inhibitor is presently available for duplex alloys, al
though the addition of tin retards most forms of brass cor
rosion to include dezincification.^
The purpose of this investigation has been to elucidate
the mechanism of dezincification of alpha and of beta brasses
and to develop a basis for predicting the conditions under
which dezincification of these alloys might be expected to
occur.
Particular emphasis was placed on selecting exposure
conditions and test methods which would not bias the experi
mental results. For example, it was felt that accelerated
tests using copper-chloride solutions could not give un
biased evidence for a dissolution-redeposition mechanism,
although this method of accelerated testing has been re
ported frequently in the literature ^2) Electro
chemical stimulation (in which the specimen was a driven
anode) has also been used as a means of producing dezincifi
cation. This method of producing accelerated attack
also can bias the experimental results by masking the pres
ence of diffusion related selective-removal processes, be
cause diffusion, as it is commonly understood, is a quite


64
Table 6
Atomic Absorption Data for Beta Brass
in 5N HC1 at 89.35C
Time Amount in solution Weight
(hours) (ppm) (wt)a removed
Cu
Zn
Cu
Zn
Cu
Zn
12
10
420
780
32,000
26
1,100
24
32
1,200
2,400
94,000
80
3,100
36
60
2,000
4,500
150,000
150
5,000
48
100
2,800
7,500
210,000
250
7,100
aAll weights in 10 ^ gm.


82
The dezincification factor introduced by Marshakov
and coworkers provides an indication of whether or not de
zincification has occurred.
However, the factor Z has no fundamental significance
and the absolute magnitude of the dezincification factor
may vary considerably for a given exposure cell and cannot
be compared with the values from other cells. The sensi
tivity of the dezincification factor to the irregularities
of the dissolution kinetics of copper during dezincification
is well illustrated.


84
might be expected to occur, or to determine potentials where
cathodic protection may be possible. ^ H9)
Electrochemical tests have a number of drawbacks and
may be subject to misinterpretation. For example, constant-
potential tests conducted for short periods, e.g., two hours,
at room temperature which fail to produce observable de-
zincification^^ may produce measurable dezincification
after longer periods of time. ^^>86)
Electrochemical tests on alloys subject to dealloying
have an added difficulty in that the surface of the sample
will quickly change to that of a dealloyed metallic "sponge"
if the electrode is subjected to dealloying conditions.^
9b,113j Sugawara and Ebiko^^ point out that thick anodic
films, such as those caused by dezincification, may produce
additional resistance, and thus a potential drop, between
the surface and the uncorroded portion of the sample. This
means that the true potential of the uncorroded sample can
not be determined. However, this potential change, which is
expressed by Ohm's Law, E = IR, is very small, because the
currents being measured in a typical electrochemical cell
are of the order of 10 ^ to 10 ^ amperes.^^^
Wilde and Teterin^^^ reported that their anodic
polarization curves for alpha brasses and duplex alpha-plus-
beta brasses were almost identical with the curve for pure
copper. The only measurable differences were in current
density. They attributed the similarities to a copper layer


101
typical of
+0.200Vshe
copper deposits
and corresponds
This
to the
potential is approximately
potential for the reaction
Cu + 2C1 = CuCl + e
As a consequence, deposi-
at potentials below
used copper chloride solu-
on reactions. A recent
cates that nickel chloride
ting effect on dezincifi-
the acceleration of dezincifi
could be explained by the
ses immersed in them,
ions which had been re
ported in the literature to produce rapid dezincification.
The experimental setup was identical to that used for the
potentiostatic tests except that a Keithley Model 602B
Electrometer attached to a strip-chart recorder was used to
measure the potential of the sample for a period of 24 hours
in the solution of interest. All solutions were stirred
vigorously.
The results of these tests on both alpha and beta
brasses are summarized in Tables 10 and 11. In every case
the observed behavior could be explained in terms of the
final steady-state potential reached by the sample. In
in solutions containing 0.1M Cl .
tion of copper would be expected
+0.200vshE
A number of researchers have
tions to accelerate dezincificati
paper by Falleiro and Pieske indi
solutions have a similar accelera
cation.This suggested that
ft
cation caused by these solutions
free corrosion potentials of bras
Tests were run in four solut


100
Figure 38. Typical potentiokinetic scan in acid solutions.
Ep is found to occur near +0.200VshE and
between the zero-current potential and the
first maximum.


1
2
3
4
5
6
7
8
9
10
11
12
age
53
60
61
62
63
64
66
96
98
102
103
109
LIST OF TABLES
Sample 6-19, Weight Information
Atomic Absorption Data for Alpha
Brass in 5N HC1 at 90.35C . .
Atomic Absorption Data for Alpha
Brass in 5N HC1 at 98.50C . .
Atomic Absorption Data for Beta
Brass in 5N HC1 at 59.60C . .
Atomic Absorption Data for Beta
Brass in 5N HC1 at 67.50C . .
Atomic Absorption Data for Beta
Brass in 5N HC1 at 89.35C . .
Atomic Absorption Data for Beta
Brass in 5N HC1 at 99.85C . .
Alpha Brass Potentiostatic
Test Data
Beta Brass Potentiostatic
Test Data
Alpha Brass Free Corrosion
Potential Tests
Beta Brass Free Corrosion
Potential Tests
Approximate Copper and Zinc
Concentrations in Solutions
Where Copper Deposits Were
Observed on Alpha Brass . .
Vll


33
in the present study point counts were taken at inter
vals as small as one micron across the region of interest.
It is recognized that making point counts at narrow inter
vals might involve some slight overlap of the irradiated
volumes between two adjacent points. However, the results
obtained by varying operating conditions sc that the beam
size would be reduced to a bare minimum did not give a change
in the results obtained, and it is felt that the counts ob
tained in this manner are of physical significance.
Results were tabulated as intensity ratios by taking
the average of three successive counts on the same point,
subtracting the average background intensity on that sample,
and dividing by the average point counts on an elemental
standard minus the average background intensity on the stan
dard. This can be expressed as:
I^(Sample) I^(Sample BG)
*pJ 100A ~ T~i"oqA(Standard'] ^iqo \ (Standard BG"J
r s 71
where A is the element of interest.
Intensity ratios of this type are the raw data for
computer programs such as the MAGIC program written by J.
W. Colby and in use at the University of Florida for quanti-
Q31
tative electron microprobe analysis. J
Figures 12, 13 and 14 show plots of intensity ratios
versus distance for samples of alpha and beta brass exposed
in 5N HC1 or in IN NaCl. The accompanying photomicrographs,


88
Present Work
The experimental potential-pH (Pourbaix) diagram for
70 w/o Cu 30 w/o Zn alpha brass in 0.1M chloride solutions
is shown in Figure 33 and was determined by W. C. Fort,
('77') 11201
III, J according to the method of Pourbaix. J Figure
34 is a simplified version of the equilibrium diagram cai
rns')
culated by van Muylder, Zoubov and Pourbaix^ J assuming a
chloride ion concentration of 0.1M and all other ionic spe
cies to be present in 10 concentrations. Figure 35 is
a similar simplified diagram for the zinc-^O systems.
Figures in the diagrams correspond to calculations which are
listed in Appendix 8. Superposition of diagrams, Figures 33,
34 and 35, gives Figure 36. It is evident that the experi
mentally constructed diagram shows a number of features in
common with the equilibrium (potential versus pH) diagram
for copper. Verink and Heidersbach have discussed these
similarities at some length.
For pure copper in deaerated solutions, the zero cur
rent potential on the upward potential sweep of the electro
chemical hysteresis circuit indicates the position of the
so-called "immunity" line at a given pH. The potential at
which this zero current is observed often is fairly close
to the calculated position of the metal/metal ion coexis
tence potential for a metal ion concentration of 10 ^ molar.
Thus the arbitrary choice of 10 ^ molar as a definition of


SOLUTION ANALYSIS
Chemical changes in the electrolyte surrounding samples
undergoing dealloying can provide information regarding the
mechanisms involved. Analytical methods which have been used
to monitor metal-ion pickup in solution include electro-
6 3 881
chemical methods [split-ring electrodes1 L J and polarog-
(44,47,95,96)-. (78,97-100) ,.
raphyv 5 J, colorimetry, J and radioactive
(99)
tracer techniques. 1 Alloy systems investigated include
+ (44 ,63,88,95,100) , (78,97) ,
the copper-zinc, copper-gold, and
f 4 71
copper-nickelv 1 systems.
Of particular interest is whether or not the more noble
species of a binary alloy dissolves during dealloying.
78,88,97-99)
Colorimetry and polarography are the only two techniques
which have been reported heretofore which can detect the
presence of trace elements in solution. Atomic ab
sorption is a technique having detection levels as good as or
better than either colorimetry or polarography,^^ and
this is the method employed in the present study.
r 7 81
Fisher and Halperin^ report that no gold was detect
ed during their colorimetry experiments with copper-gold
alloys. This is in agreement with the colorimetric data of
Pickering and Byrne on the same system. By contrast,
54


48
Figure 20.
Deposit on surface of dezincified
sample. Sample was exposed in 5N
100C for 10 days. 100X
alpha brass
HC1 at