Citation
Spectro-electrochemical studies of the inhibition effect of 4, 7-diphenyl-1,10-phenanthroline on the corrosion of 304 stainless steel

Material Information

Title:
Spectro-electrochemical studies of the inhibition effect of 4, 7-diphenyl-1,10-phenanthroline on the corrosion of 304 stainless steel
Creator:
Huang, Hsuan-Jung, 1944-
Publication Date:
Language:
English
Physical Description:
xii, 115 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Absorption spectra ( jstor )
Adsorption ( jstor )
Corrosion ( jstor )
Electrodes ( jstor )
Ions ( jstor )
Isotherms ( jstor )
Molecules ( jstor )
Stainless steels ( jstor )
Steels ( jstor )
Surface areas ( jstor )
Chemistry thesis Ph. D
Diphenylphenanthroline ( lcsh )
Dissertations, Academic -- Chemistry -- UF
Stainless steel -- Corrosion ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 109-114.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Hsuan-Jung Huang.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022577227 ( ALEPH )
04673990 ( OCLC )

Downloads

This item has the following downloads:


Full Text













SPECTRO-ELECTROCHEMICAL STUDIES OF THE INHIBITION
EFFECT OF 4,7-DIPHENYL-I,10-PHENANTHROLINE
ON THE CORROSION OF 304 STAINLESS STEEL











By

HSUAN-JUNG HUANG


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY





UNIVERSITY OF FLORIDA


1978























The author dedicates this dissertation to his wife,

Shang-Cheng C. Huang














ACKNOWLEDGEMENTS


The author would like to express his gratitude to his research director, Dr. G. M. Schmid, for the interest and assistance given to him in the course of this investigation and in the preparation of this manuscript. Thanks are also due to the members of his committee, Dr. E. D. Verink, Jr., Dr. J. D. Winefordner, Dr. R. G. Bates and Dr. R. C. Stoufer.

He would also like to thank Dr. K. P. Li for the generosity of sharing his laboratory facilities and helpful discussions, and Mr. R. Strohschein and Mr. W. Axson for their help in the technical aspects of this work.


iii














TABLE OF CONTENTS


Page
ACKNOWLEDGEMENTS iii LIST OF TABLES vi LIST OF FIGURES vii ABSTRACT x CHAPTER

I. INTRODUCTION 1 II. EXPERIMENTAL 12 Experimental Design 12 Experimental Technique 17

Potentiostatic Polarization 17 Chemicals and Equipment 18 Spectrophotometric Measurement 20 III. DATA AND RESULTS 23

Spectrophotometric Measurements by the
Modified Dual-Wavelength Method 23
Adsorption of DPP on Stainless Steel
Electrodes 31 Potentiostatic Polarization 39

Polarization Behavior of the Steel
Electrode at Various Constant Fractional
Surface Coverages 0 of DPP 43 IV. DISCUSSION 55 Selection of Metal Adsorbent and
Adsorbate 55








CHAPTER


Precision of the Spectophotometric
Measurement and Surface Coverage
Determination 57

The Dependence of Surface Coverage
on Inhibitor Concentration and Applied
Polarization Potential 62 Determination of the Adsorption Isotherm 63

Adsorption and the Structure of the
Electrical Double Layer 84

Proposed Inhibition Mechanism of DPP
on the Corrosion of Stainless Steel in
0.1 M HCI Solution 97 V. SUMMARY 105 REFERENCES CITED 109 BIOGRAPHICAL SKETCH 115


Page













LIST OF TABLES


Table Page

I. Open Circuit Potential, E, and Rest
Potential, E n in SolutioR of Various
DPP Concentrations 44


II. Tafel Slope (Cathodic) and Corrosion Current of the Steel Electrode in Solutions with
Various DPP Concentrations (Determined
from Figure 16) 47



III. Particle - Particle Interaction Parameters
b; p, q and p', q Used for Fitting Data into
Various Isotherms and their Calculated Adsorption Constant K and Standard Free Energy
of Adsorption Gu 82


IV. Comparison of Inhibition Efficiencies (Determined from Potentiostatic Polarization
Data) with the Fractional Surface Coverage (Obtained from Spectrophotometric Measurements) of 4,7-Diphenyl-l,10-phenanthroline
on the Stainless Steel Electrode 96













LIST OF FIGURES


Figure Page
1. Stainless Steel Electrode 13

2. Kel-F Electrode Holder 14

3. Electrochemical Cell 16

4. Block Diagram of the Potentiostatic Polarization Circuit 19
5. UV Spectrum of 4,7-Diphenyl-1,10-phenanthroline 24
6. Calibration Curve (Absorbance vs. Concentration) of 4,7-Diphenyl-l,10-phenanthroline at 286.5 nm and 319.0 nm 25

7. Calibration Curve, Absorbance Difference vs. Concentration (AA is the Absorbance
Difference at 286.5 nm and 319.0 nm) 27

8. Demonstration of the Modified DualWavelength Method 30

9. Absorbance vs. Concentration of 4,7-Diphenyl1,10-phenanthroline in the Presence of
2.10 x 10-3 Fe++ 32

10. Absorbance vs. Concentration of Fe++ in
the Presence of 6.58 x 10-0 M 4,7-Diphenyl-l,10-phenanthroline, at 286.5 nm
and 319.0 nm 33

11. Dependence of Surface Coverage on 4,7-Diphenyl-l,10-phenanthroline Concentration
at Various Fixed Potentials 36

12. Three Possible Configurations of 4,7-Diphenyl-l,10-phenanthroline 37


vii








Figure


13. Dependence of Fractional Surface Coverage 0
on 4,7-Diphenyl-l,10-phenanthroline
Concentration at Various Fixed Potentials 41

14. Fractional Surface Coverage 0 vs. Applied
Potential at Various Constant 4,7-Diphenyl1,10-phenanthroline Concentrations 42

15. Potentiostatic Polarization Current vs.
Potential Curves (in the Absence and Presence of Various Constant 4,7-Diphenyl-l,10phenanthroline Concentrations) 45

16. Plot of Cathodic Tafel Slopes from the
Potentiostatic Polarization Curves 46

17. Fractional Surface Coverage 0 vs. 4,7Diphenyl-l,10-phenanthroline Concentration
at Various Fixed Potentials 50

18. Polarization Current vs. 4,7-Diphenyl-l,10phenanthroline Concentration at Various
Fixed Potentials 52

19. Constructed Potentiostatic Polarization
Curves at Various Constant Fractional
Surface Coverage 0 54

20. Dependence of the Surface Coverage 0 on
the Relative Concentration a/a0.., Calculated from Equation (6) for kih, erent
Values of b 71

21. Data Plotted According to the Langmuir
Isotherm 72

22. Logarithmic Plot of the Freundlich
Isotherm 74

23. Data Plotted According to the Frumkin
Isotherm 75

24. Data Plotted According to the Virial
Coefficient Isotherm 76

25. Data Plotted According to the Hill-de Boer
Isotherm 77

26. Data Plotted According to the Parsons
Isotherm 78


viii


Page








Figure


Page


27. Data Plotted According to the BlomgrenBockris Isotherm 79

28. Data Plotted According to the ConwayBarradas Isotherm 80

29. Schematric Distribution of Charges and
Potential (Due to Ions) at the Metal
Electrolyte Interface at Positive (a) and Zero (b) Electrode Charge with no Specific
Adsorption of Ions 87

30. Schematic Representation of a Metal
Solution Interface with Specifically Adsorbed
Anions at Positive (a) and Zero (b) Electrode Charge 88







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


SPECTRO-ELECTROCHEMICAL STUDIES OF THE INHIBITION
EFFECT OF 4,7-DIPHENYL-1,10-PHENANTHROLINE
ON THE CORROSION OF 304 STAINLESS STEEL By

Hsuan-Jung Huang

August 1978

Chairman: Gerhard 11. Schmid
Major Department: Chemistry

The compound 4,7-diphenyl-l,10-phenanthroline, which

has a strong absorption band in the UV absorption region, is slightly soluble in acidic aqueous solution (to about 10-4 M) and is potentially a good inhibitor, was selected to study the relationship between adsorption and inhibition on the corrosion of 304 stainless steel in 0.1 M IICI solution. A 304 stainless steel cylinder with an approximate surface
2
area of 5 cm was used as the test electrode and solid adsorbent. The current vs. potential behavior of the system was determined potentiostatically and the amount of the inhibitor adsorbed on the electrode surface was measured by the concentration depletion method using the modified dual-wavelength spectrophotometric method.

The surface coverage of DPP on the steel electrode was found to depend on the DPP concentration in solution and on the applied polarization potential. At constant applied potential the surface coverage of DPP increases with







increasing DPP concentration in bulk solution (except at the applied potential of -0.200 V and 0.000 V), but levels off at high bulk concentration, c > 10-5 M, for potentials more negative than -0.400 V. At constant concentration, generally, the adsorption of DPP decreases with the shift of applied potential in the anodic direction. The adsorption of the inhibitor was plotted against the applied potential (from

-0.700 V to 0.000 V vs. S.C.E.) in the inhibitor concentration range of 5.0 x 10-7 to 1.0 x 10-5 M. Fifteen different adsorption isotherms were tested for their fit to the experimental data. Five of them (Frumkin, Virial Coefficients, Hill-de Boer, Blomgren-Bockris and Conway-Barradas isotherm) fit the experimental data with a correlation coefficient >0.95 in the surface coverage range 0 = 0 to 0 = 0.5. At higher surface coverages deviations from the isotherms are found. The particle - particle interaction parameters b, p, q or p', q of the isotherms above were obtained with curve fitting processes. Judging from the interaction parameters obtained, an attractive interaction exists between the adsorbed species. The adsorption constant K calculated according to the different isotherms agrees reasonably well at the same applied potential, but decreases slightly as the applied potential shifts anodically. The calculated mean adsorption constant (from different isotherms) changes from (2.2 � 0.3) x 105 at -0.600 V to (1.2 � 0.1) x 105 at -0.200 V. The standard free energy of adsorption -,-o calculated behaves in the same manner. It has a mean value (from different







isotherms) of 7.3 � 0.1 kcal mole- at -0.600 V and 6.9 � 0.1
-l
kcal mole at -0.200 V.

From the data, a series of polarization curves at various constant surface coverage with DPP, 0 = 0.1 to 0 = 0.5 were obtained by extra- and interpolation. The curves all have Tafel slopes (cathodic) identical to that in pure 0.1 M HCI, the current decreasing with increasing 0 at constant potential.

According to the data, the inhibition effect of DPP is mainly due to a surface blocking effect. The adsorbed DPPH+ ion may form a thin layer of chelate film with the Fe++ ion on the steel electrode surface which then retards both the cathodic hydrogen evolution reaction and the anodic metal ion dissolution reaction. Mechanisms for both interactions are proposed.

The inhibition efficiencies of DPP (obtained from potentiostatic polarization data) were compared with the fractional surface coverages of DPP on the electrode surface (measured with the UV spectrophotometric method). Good agreement was obtained only in the most cathodic potential region and at low DPP concentration. The inhibition efficiency ranges from about 15% at 1.02 x 10-6 M DPP concentration to about 65% at 1.28 x 10-5 M DPP concentration in the potential region of -0.500 V to -0.700 V. It decreases slightly as the applied potential is shifted anodically.


xii













CHAPTER I
INTRODUCTION


There are three prerequisites for corrosion to take place; a potential difference, a conduction path and the availability of electrode reactions for transferring charges across the metal-solution interface.1'2 In order to control corrosion, it is then necessary to control one of the prerequisites. This is often most easily done by the use of inhibitors. An inhibitor which is adsorbed on the metal surface may function by (1) increasing the true ohmic resistance of the interface and thus limiting the charge transfer processes by a blocking effect or (2) interfering with the anodic, the cathodic, or both the anodic and the cathodic electrochemical processes. Examples of case (1) are inhibition by the formation of an oxide film or by precipitation of a nonconducting reaction product onto the metal.3'4 Inhibition caused by an increase in the hydrogen activation overpotential and/or a decrease of the potential difference on the metal surface are examples of case (2).5

Adsorption of organic inhibitors on the electrode surface (metal surface) is not only due to physical forces but is often accompanied by chemical interactions between the metal surface and the adsorbate. Van der Waals forces and electrostatic interactions are two general types of forces that








cause the physical adsorption of an organic substance. Van der Waals forces are generally weak and operate over the entire surface. The electrostatic interaction arises when a charged organic species (cation or anion) is close to a polarized electrode surface (negatively or positively charged). Inhibitors such as organic onium compounds, e.g., quaternary phosphonium and arsonium ions which are reported as highly effective inhibitors in acidic solution, are adsorbed electrostatically in the electrode-solution interface. 6,7 Compounds which are able to form the onium structure in proton containing solvents, act in a similar manner. Formation of a dative link between the metal and the organic molecule results in chemisorption. The bond is formed through the sharing of a pair of electrons between the organic adsorbate and the metal. Chemisorption might cause inhibition either by stabilizing the metal ions in the surface lattice to decrease the dissolution tendency of the metal at anodic areas or by increasing the activation operpotential for hydrogen evolution at cathodic areas. A special type of chemisorption is the complex formation between inhibitors and metal ions on the electrode surface leading to the establishment of nonconducting chelate films which usually exhibit very high inhibition efficiency. '9 Among the factors that critically influence the extent of chemisorption are the nature of the electronic configuration of the adsorbed species. The strength of bonding is a function of the metal, since it relates to the residual valence orbitals








existing at the metal surface; in addition, unfilled atomic orbitals are also required and these are principally of d character. An elementary approach to the influence of the electronic configuration of the adsorbed group is possible in terms of the availability of electrons for bond formation. This can be considered to be a function of the electron density and the polarizability of the functional group, or, for simple molecules, of the Vb or VIb elements that the 10
compound contains.

Inhibitors are classified as anodic type, cathodic

type or mixed type inhibitors according to their interactions with electrochemical processes. Anodic inhibitors function by stifling the acidic dissolution reaction by, generally, increasing the overpotential of the anodic process. An example of this class are the sulfide series inhibitors.11 Cathodic inhibitors increase the overpotential of the cathodic process. The increase of the overpotential of the cathodic process shifts the corrosion potential (open circuit potential) in the negative direction and thus retards corrosion on the cathodic area. Quaternary amine and onium salts belong to this class.6'7'12 Inhibitors of the mixed type influence both the cathodic and anodic processes, raising the overpotential of both. Substances whose molecules consist of an organic base cation and an acid anion having oxidizing properties are widely used as inhibitors of this type. Amines, thiols, derivatives of thiourea and quinolines are examples.1315







In order to understand the role of organic inhibitors in the inhibition mechanism of corrosion, the adsorption behavior of the organic adsorbates on the electrode surface must be known. The earliest investigation of the adsorption process on solid electrodes was carried out by Frumkin and coworkers who studied the adsorption of hydrogen, oxygen and halide ions on metals of the platinum group and on gold.16'17 Quantitative measurements of the adsorption of organic substances on solid metals has started only during the last 15 - 20 years, having become possible after the introduction of the radioactive tracer method, the pulse potentiodynamic and galvanostatic methods.1820 Recently, studies of adsorption processes at a metal adsorbent have also attracted special attention in investigations of electrochemical redox reactions, the electrochemical synthesis of organic compounds and the effect of organic additives in the electrodeposition of metals.2123

Due to the ideal polarizability of the liquid mercury electrode and some of its liquid amalgams to which the Gibbs adsorption isotherm can be applied without any loss or rigor and with much simplification, precise and detailed thermodynamic information has been obtained which has led to the understanding of the structure of the double layer, the adsorption behavior and the mechanisms of electrode processes.24 The use of capillary electrometry and capacitance methods for obtaining the surface charge q, and the surface excess ri of component i at a given potential E







and solution composition Pk in terms of the respective chemical potentials pPi' ." .k have been reviewed thoroughly in the literature.25'26

The adsorption of organic compounds on liquid electrodes can be studied by measuring the interfacial tension y, the electrode charge q, and the differential capacity of the double layer C. Measurements of the interfacial tension y are generally done with the Lippmann capillary electrometer. The charge on the electrode (per unit area) is then obtained from the Lippmann equation


q = - 'Y-)
6E k


i.e., from the slope of the surface tension vs. potential curve keeping all the chemical potentials Vk constant.27 Alternatively, the differential capacity of the double layer is related to the electrode charge q by


c = Ek
6E k


In contrast to the case of liquid metal electrodes, there are as yet no methods available for determining the absolute values of interfacial tension between a solid electrode and the surrounding solution. However, the dependence of the interfacial tension on the electrode potential and the solution composition can be investigated by a number of methods based on the study of mechanical properties such







as hardness, creep, and coefficient of friction.28,29 From the variations of the mechanical properties with the change of potential, the adsorption of organic species on a solid electrode may thus be determined. However, on solid electrodes the work done in increasing the surface area is accompanied by irreversible changes and is therefore greater than the interfacial tension. This makes the application of a thermodynamic treatment difficult.

Methods for measurement of the double layer capacity with alternating current and of potential decay curves are widely used for the investigation of adsorption phenomena. 30-32 Various bridges and procedures for capacity measurement have been described employing regular RC (resistance-capacitance) networks or transformer ratio arms.33,34 Recently, a method for automatic recording of the ohmic and capacitive component of the impedance at electrodes was developed by Breiter for use with a linear potential sweep.35 The method employs phase-sensitive amplifiers and the C and R components of the impedance are directly plotted out together with the potentiodynamic current profile. The adsorption of an organic substance on an electrode can be calculated from capacity data using the following equation36


C -C
o = 0
C - C'
0


Here, 0 is the fraction of surface covered by the adsorbate, CO is the capacity of the double layer measured in the pure







electrolyte solution, C is the capacity of the double layer in the solution with organic substance added, C' is the capacity of the double layer at maximum (saturation) coverage of organic substance on the electrode surface.

Capacity measurements at solid electrodes suffer from

the disadvantage of frequency dependence, which some workers attribute to surface roughness, while others attribute it to the penetration of the solution between the electrode and the insulating coating.37,38 Also, the adsorption of hydrogen or oxygen on electrode surfaces (e.g., on metals of the platinum group) interferes with the capacity measurement and thus makes the interpretation of data difficult.

Steady state adsorption at solid electrodes can be

studied with radioactive tracers. This can be done either by measuring the change of activity in solution due to adsorption at the electrode or by direct monitoring of the surface of the electrode, e.g., with a proportional counter whose window itself is also the electrode under study.

The first method is applicable to studies of adsorption of ions and molecules that are strongly adsorbed so that the adsorption measurements can be carried out in very dilute solutions where the amount adsorbed is comparable in magnitude to the amount in the solution with which the electrode is in equilibrium.39 In the procedure developed by Balashova and her coworkers a normal electrolytic cell is used in conjunction with a counting monitor through which the solution can be circulated and its activity monitored.40 Alternatively,







samples of the solution can be withdrawn and counted. The advantage of this method is that the adsorption can be followed continuously in situ with control of potential at the electrode being maintained. However, the initial concentration of adsorbate in the solution must generally be below about 10-4 M.

The direct counting through a thin window electrode was developed for electrochemical adsorption processes by Blomgren and Bockris from the procedure of Aniansson and of Joliot.41-43 The adsorption of thiourea- 35S and ethylene14C has been studied using gold foil and platinized gold foil electrodes respectively.44'45 When using this direct monitoring method, it is necessary to ensure that the measured radiation comes only from adsorbed molecules, and not from radioactive molecules in the solution in the vicinity of the electrode. To achieve this, cells are used whose cross-section is very narrow near the electrode, with most of the solution being present in reservoirs on either side of the electrode. 46 Radiation is thus counted only from the electrode and a very thin layer of solution immediately adjacent to it. Green, Swinkels and Bockris developed another technique for the determination of adsorption of radioactive species at solid electrodes which cannot conveniently be formed into thin films.47 The method consists of using an endless metal tape electrode which is run past two proportional counters to monitor both surfaces of the tape after the tape has passed through an electrolytic cell in which adsorption of the tagged material takes place.








Radiotracer methods measure directly the total number of labeled species adsorbed on the electrode, but they are insensitive to the structure of the adsorbed species and to any changes in the nature of the adsorbed substance.

If the specific surface area of the electrode is large relative to the solution volume, adsorption of a surface active substance will be accompanied by an appreciable change of the solution concentration. In such cases, the amount of substance adsorbed on the electrode surface can be calculated directly from the change in solution concentration. Depending on the nature of the adsorbed substance and the magnitude of the solution concentration, various methods have been used for the determination of such concentration changes. Conway, Barradas and Zawidzki used a spectrophotometric method to study the adsorption of acridine and quinoline on copper, nickel and silver.48 Balezin and coworkers used the chromatographic method for determining the adsorption of sodium benzoate and dicyclohexylamine on iron and
49
magnetite.
The measurement of adsorption by determination of

changes in solute concentration permits the study of adsorption effects at steady state, but it can be applied only to solutions with very small initial concentrations (less than 10- M) for reasonable accuracy, and to systems which have large electrode surface area to solution volume ratios.
From this brief review of the techniques used in studies of the adsorption of organic substances on solid electrodes,








it is apparent that none of the above methods are entirely satisfactory for the purpose of thorough understanding of the adsorption behavior. Each technique has its own advantage and also its own unavoidable inherent defects. However, among the techniques, the concentration depletion method is straightforward and simple and it supplies adsorption information under steady state conditions. It also allows the study of the adsorption isotherm and of the relationship between the corrosion inhibition effect and the degree of adsorption of organic adsorbate on solid metals. Conway, Barradas and Zawidzki were the first group to develop the UV spectrophotometric technique for the determination of the steady state adsorption of organic 48
adsorbates at a solid electrode. Later, Newmiller and Pontius applied the same technique to studies of the adsorption of photographic developers on silver.50 After these, few reports concerning the spectrophotometric measurement of adsorption appeared in the literature. 51,52 This is due to several serious difficulties inherent in this technique. First, the UV absorption spectra of the organic adsorbate are often distorted in the presence of traces of metal ions (e.g., copper, nickel, chromium, iron) and/or complexes are formed between the metallic ions and organic adsorbates. Thus the dissolution of the metal adsorbent, e.g.-, in acidic solution, may give rise to interferences in the analytical determination of the organic adsorbate. Second, the applicable potential range is







severely restricted because of either the possible accelerated dissolution rate of the metal adsorbent at certain anodic potentials or the possible electrocatalytic redox reactions of the organic adsorbate. Thus, adsorption studies could only be made successfully with noble metal electrodes, such as platinum, gold or iridium and at moderately anodic potentials. Also, these adsorption measurements were performed either on powder or on gauze electrodes and studies on ordinary cylindrical or disc electrodes have been lacking. In order to demonstrate the feasibility of spectrophotometric concentration depletion studies of organic adsorbates on metal surfaces, such adsorption work was undertaken in conjunction with a corrosion inhibition study to examine the adsorption behavior of a specifically selected organic inhibitor on the semi-noble stainless steel electrode and to determine the relationship between adsorption and inhibition.













CHAPTER II
EXPERIMENTAL


Experimental Design


The sample used as the test electrode was stainless steel, AISI 304, provided by the United States Steel Corporation. Its composition was given as 0.03 C, 0.027 P, 1.10 Mn, 0.022 S, 0.43 Si, 9.26 Ni, 18.6 Cr, 0.39 Mo and

0.04 N (weight percent). Bar stock was machined into cylinders with a diameter of 6 mm and a height of 10 mm (Figure 1). The cylinders were annealed at 11500 C for 30 minutes and cooled rapidly by quenching in water. They were then tapped, threaded and mechanically polished at 266 rpm with 400 followed by 600 grit emery paper. Before use, the cylinders were tapped, degreased with benzene in an ultrasonic cleaner, soaked in acetone and rinsed with triply distilled water.

The test electrode holder (Figure 2) was made from a

Kel-F rod machined to approximately 1 cm in diameter in which a 3 mm center hole was drilled. The rod was heated and a threaded stainless steel rod was inserted so that thread protruded on both ends. The Kel-F rod was then shrunk in an ice bath to facilitate its insertion into a 1 cm ID glass stirrer bearing sleeve which had a 24/40 standard taper joint




13













~6 mm

3mm






5 mm I mm


Figure 1 - Stainless Steel Electrode





















Kel-F Rod


24/40 Standard Taper Joint









Stainless Steel Rod


Finger Nut Teflon Washer



































Teflon Washer


Figure 2 - Kel-F Electrode Holder








attached. The steel sample was affixed to the protruding steel rod. The sample-to-sample holder seal could be tightened by turning a finger nut at the other end of the protruding steel rod. Teflon washers were placed between both the sample and holder, and the nut and holder, and a thin layer of Kel-F polymer wax was coated on the top surface of the test electrode to give a water proof seal. The shielded top surface of the test electrode was checked after each experiment to make sure no leakage and corrosion occurred there. If corrosion was detected, the data of that run were discarded. The electrode area (geometric) exposed
2
to solution was approximately 5 cm

The design of the electrochemical cell is shown in

Figure 3. The cell was made of Pyrex and was modified from the conventional three-electrode electrolytic cell. The test electrode compartment which basically consists of a 14 mm ID Pyrex tube and a 24/40 female standard taper joint was separated from the auxiliary electrode compartment by a coarse frit and Teflon stopcock. A Luggin capillary and a stopcock were used to connect the reference electrode compartment to the test electrode compartment. A gas inlet was connected through another coarse frit to the bottom of the test electrode compartment. The frit breaks up the helium gas stream into bubbles which enhances the deaeration and stirring effect. A side arm made of a 14/35 standard taper joint connects to the middle part of the test electrode compartment. It allows the withdrawal and return

























Gas Inlet * "-


Gas Outlet








FTest


/etCompartment

Auxiliary
Electrode Reference Compartment Electrode 'p-Compartment


Stopcock


Luggin Capillary


(Coarse) Frit


Figure 3 - Electrochemical Cell







of electrolyte solution. Another small side arm with a 5/20 standard taper joint which also facilitates the replacement of electrolyte solution was built into the auxiliary electrode compartment. A platinum electrode and a saturated calomel electrode were used as the auxiliary and reference electrodes respectively in this work.


Experimental Technique


Potentiostatic Polarization

Solutions were deaerated with helium (99.99 percent) for a minimum of three hours prior to use. Immediately before each experiment, the cell was washed with chromosulfuric acid cleaning solution, tap water, and then rinsed thoroughly with triply distilled water. The stainless steel sample was secured on the Kel-F holder, rinsed with distilled water and the solution to be studied, and immersed in solution. All samples were pretreated at -0.900 V in a separate electrochemical cell for twenty minutes to reduce possible air-formed surface films. Solution stirring was accomplished with a magnetic stirring disc and the helium flow which was continued throughout the experiment. All experiments were conducted at room temperature.

After the pretreatment at -0.900 V the test electrode was moved quickly to another identical electrochemical cell containing the fresh deaerated solution to be studied. The applied potential was shifted anodically in step-wise







increments. The magnitude of the imposed step depended on the potential range under investigation and the electrochemical reaction(s) associated with it. In regions where changes in potential caused significant changes in current density, steps of 20 to 30 mV were generally employed. In regions of approximately constant current, 50 mV steps were used. The current flowing in the auxiliary - test electrode circuit was determined after an arbitrary time interval of 10 minutes.



Chemicals and Equipment

All chemicals used in solution preparation were reagent grade. The water employed for solution preparation was triply distilled, first from alkaline potassium permanganate and then from a two stage Heraeus quartz still, and collected in a two liter Pyrex volumetric flask. Solutions used were

0.1 M HC1. They were normalized with standardized sodium hydroxide and were always 0.1000 � 0.0005 M. The organic compound, 4,7-diphenyl-l,10-phenanthroline (bathophenanthroline) (DPP) was obtained from Matheson, Coleman and Bell Co. The organic chemical was used directly without further purification.

The block diagram of the system employed in potentiostatic experiments is shown in Figure 4. Polarization was accomplished with a slightly modified Harrar potentiostat.53 (Two each of obsolete transistors 2N333A and HA7534 were replaced with the equivalent circuit elements 2N3568, 2N5869, and 2N3644, 2N5867). A Keithley model 660 differential





Electrometer





Pot





A BRcd
C Recorder


entiostat


Test


Reference Auxiliary


Figure 4 - Block Diagram of the Potentiostatic Polarization Circuit
A: Test electrode; B: Reference electrode; C: Auxiliary
electrode.







voltmeter was used to monitor the applied potential. The current flowing in the auxiliary - test electrode circuit was determined from the potential drop across a standard precision resistor (� 1%) using a model SRG Sargent potentiometric recorder. All potentials are reported relative to the saturated calomel electrode. Current densities were calculated using the geometric electrode area.

Spectrophotometric measurements were performed with an AMINCO DW-2 UV-VIS spectrophotometer.


Spectrophotometric Measurement

The amount of the inhibitor adsorbed on the electrode

surface was determined by the concentration depletion method using UV spectrophotometry. The inhibitor concentration in the bulk solution of electrolyte was measured before and after the introduction of a polarized steel electrode into the solution. The number of inhibitor molecules corresponding to the solution concentration change of these two measurements is considered as the number of molecules adsorbed on the electrode surface. Due to the continuous dissolution
-H- ++ 4+
of metal ions (Fe , Ni+, Cr from stainless steel) in electrolytic solution during polarization, the inhibitor concentration was determined from a modified dual wavelength method based on the scanned UV spectrum. The general procedures used for the determination of the adsorption of inhibitor on the electrode surface are as follows:

1. After the pretreated electrode was moved to the cell








containing fresh deaerated 0.1 M HCl solution, the open

circuit potential of the electrode was recorded after

10 minutes.

2. A fixed potential (-0.600 V, -0.400 V, -0.300 V,

-0.200 V or 0.000 V, respectively) was applied to the

steel sample for at least 40 minutes, using the potentiostat.

3. After polarization, the sample electrode was moved

quickly back to the pretreatment cell and was polarized there continuously at the same set potential (-0.600 V

to 0.000 V, respectively).

4. A portion (about 2.5 ml) of the electrolytic solution

in the test electrode compartment was drawn into a

cuvette for UV spectrum measurement and a spectrum was

obtained by scanning the range of 200 to 350 nm at 2

rnm s-1

5. The base line correction control of the spectrophotometer

was used to flatten any absorption spectrum shown on

the scan.

6. After the solution was returned to the electrochemical

cell, a proper amount of concentrated inhibitor solution (e.g., 200 pl of the 1.03 x 10-4 M DPP) was added

to the electrolytic solution in the test electrode

compartment (with an approximate total volume of 12 ml).

Ten minutes were allowed for the added inhibitor solution

to be well mixed with the electrolytic solution.

7. The original electrolytic solution in the auxiliary








electrode compartment was withdrawn with a Teflon tip syringe and the solution containing inhibitor in the

test electrode was allowed to flow freely through the coarse frit and the open stopcock into the auxiliary

electrode compartment. Another 10 minutes was allowed

to have the solution in both the test electrode and auxiliary electrode compartments reach steady state.

8. The solution remaining in the test electrode compartment

was finally adjusted to a known solution volume (6.5 ml)

and the stopcock between the test and auxiliary electrode compartments closed.

9. The UV spectrum of the electrolytic solution in the test

electrode compartment was measured again, as in step 4. 10. The steel electrode was returned to the cell and was

polarized again at the same set potential.

11. The spectra of the electrolytic solution were taken

twice at 10 minute intervals during polarization.

The inhibitor concentration changes in solution before and after the introduction of a polarized steel electrode were then determined from the spectra measured in stels 9 and 11.













CHAPTER III
DATA AND RESULTS


Spectrophotometric Measurements
by the Modified Dual-Wavelength Method



The UV spectrum of 4,7-diphenyl-l,10-phenanthroline (DPP) is shown in Figure 5. The adsorption peak used for quantitative determination is located at 286.5 nm. The molar extinction coefficient at this wavelength was determined from an absorbance vs. concentration plot (Figure 6). The slope of the plot was calculated by the least squares method and was found to be (4.68 � 0.01) x 104 cm- M-. In the presence
++ ++ 4
of metal ions (Fe+, Cr , Ni in this experiment) the absorption spectrum of metal ions superimposed onto that of DPP thus distorting the spectrum of DPP. The degree of distortion is dependent on the concentration of the metal ions present. This matrix ion interference had to be removed or corrected for before the inhibitor concentration in solution could be determined. Correction was achieved as follows: First, the electrode was polarized at the desired potential at least 40 minutes to introduce an appreciable amount of matrix ions into the bulk solution so that any further short term polarization of the electrode would not cause an appreciable concentration change of the matrix ions. Second,

















0.3




0
Co


0.2





0.1


250 nm 300 run
Wavelength


Figure 5 - UV Spectrum of 4,7-Diphenyl-1,10-phenanthroline

















































1.0 2.0 3.0 4.0 5.0 6.0


Concentration, N x


106


Figure 6 -


Calibration Curve (Absorbance vs. Concentration) of 4,7-Diphenyl-l,10-phenanthroline at 286.5 nm (o) and 319.0 nm (0)


0.3 0.2


0.1








the UV absorption due to the matrix ions in solution was then removed electronically from the spectrum by using the base line correction control of the spectrophotometer. After this adjustment, the inhibitor added to the electrolytic solution shows its original undistorted spectrum. However, further polarization of the electrode introduces additional matrix ions into the solution and although there was already an appreciable amount of matrix ions in the bulk electrolyte, this caused changes in the matrix ion concentrations sufficient to lead to slight distortions of the inhibitor spectrum.

In order to eliminate this residual error, a modified dual-wavelength measurement was used. Instead of using the absorbance at the wavelength of maximum absorption (286.5 rm in DPP), the difference of the absorbance at the wavelength of maximum absorption and a reference wavelength (319.0 nm, arbitrarily chosen) was used. The absorbance difference, AA, between 286.5 nm and 319 nm, see Figure 6, was plotted vs. concentration. Figure 7 shows such a plot. The slope of this plot was calculated to be (3.37 � 0.01) x 104 cm- M-I. For an undistorted spectrum, the concentration corresponding to the measured absorbance difference was taken directly from the graph. If the spectrum was distorted by the change of the matrix ion concentrations, a correction procedure was used to find the possible shift of the matrix line. The correction procedure is as follows: On the undistorted inhibitor spectrum (Figure 5) an arbitrary horizontal line was drawn across the spectrum, resulting




















0.3


0.2 0.1


1.0 2.0 3.0 4.0 5.0


Concentration,


M x 106


Figure 7 -


Calibration Curve, Absorbance Difference vs. Concentration (AA is the Absorbance Difference at 286.5 nm and 319.0 nm)


6.0








in three intersection points which have the same absorbance (also the same molar extinction coefficient). The absorbance at these three wavelengths will stay at the same value as long as no spectral interference is present, but will become different if matrix ion interference appears. Thus, for a distorted spectrum, a smooth curve drawn through these three wavelength points will show the relative interference of the matrix ions to the spectrum. The curve is parallel to the true matrix line and allows to determine absorbance changes due to the matrix ion interference at any two proximate wavelengths (< 50 nim). The procedure is illustrated in Figure 8. Here, curves a and b represent the interference (matrix line). An arbitrary horizontal line d is drawn which intercepts curve a at wavelengths 247.0 nm, 257.8 nm and 319.0 rm. The three wavelength points on curve b are then connected by a smooth curve (curve e). The absorbance difference at wavelength 286.5 nm and 319.0 nm determined from curve e is the residual error arising from the matrix ion interference. After correcting this value for the measured absorbance difference on curve b, the true absorbance difference for the undistorted spectrum is then obtained. For practical purposes, the extraction of the absorbance difference AA for the pure organic spectrum can be achieved by measuring the absorbance difference between curves b and e at the wavelength of 286.5 nm and 319.0 nm, see Figure 8.















Figure 8 - Demonstration of the Modified Dual-Wavelength Method. Curve
a, 4,7-Diphenyl-l,10-phenanthroline Spectrum; Curve b, Distorted Spectrum; Curve c, Matrix Line (Interference); Curve d,
Arbitrary Straight Line; Curve e, Rebuilt Simulated Matrix Line












0.4 0.3


0.2


0.1


250 nm 300 nrm
Wavelength








The inhibitor used in this work, DPP, is known to be a complexing agent for Fe-. It forms a very stable red complex with iron(II), Fe(DPP)3, in solutions having a pH of

2.0 to 9.0.54 Since in the present work the base electrolyte was 0.1 M HCI, the pH of the electrolytic solution,

1.0, is lower than the complex formation range and no Fe(DPP)3 complex formation between Fe++ and DPP should be expected. In order to test if there is any interaction between Fe++ and DPP in the solution studied two calibration titrations were run. First 40 Pi and then successively 10 pl of

1.03 x 10-4 M DPP solution were added to 2.5 ml of a solution 2.10 x 10- M in Fe+. The plot of the absorbance at 286.5 nm vs. DPP concentration (Figure 9) showed a straight line with a slope of (4.76 � 0.01) x 104 cm-1 M-I1 which is very close to that of the calibration curve in Figure 6. Similarly, successively 20 il of 1.05 x 10-2 M Fe++ solution were added to 2.5 ml of a 5.579 x 10-6 M DPP solution. Two horizontal lines were obtained for the absorbances at 286.5 nm and 319.0 nm vs. Fe++ concentration (Figure 10). This confirms that there is no chemical interaction between Fe++ and DPP in 0.1 M HCI solution in the UV spectral range studied.


Adsorption of DPP on Stainless Steel Electrodes


The amount of inhibitor adsorbed on the electrode surface was calculated from the concentration change resulting












0.20 0.16


0.12 0.08


0.04


0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 Concentration, 1 x 106


Figure 9 - Absorbance vs.
Presence of 2.


Concent ation of 4,7-Diphenyl-1,10-phenanthroline in the 10 x 10-5 Fe++













0.26


0.22f





0 8
I

0.14






I.0 20 3.0 4.0 Concentration, M x 104 Figure 10 - Absorbance vs. Concentration of Fe"+ in the Presence of 6.58 x 10- 6M 4,7Diphenyl-1,10-phenanthroline, at 286.5 nm (o) and 319.0 nm (e).








from introduction of the polarized electrode into the solution. The total number of molecules adsorbed on the electrode surface was obtained by multiplying the change in concentration of the bulk solution with the volume of the bulk solution and Avogadro's number. The calculated number was then divided by the total surface area of the electrode (a roughness factor of 4.0 was used to convert the geometric area into an estimated true area) to obtain the number of molecules adsorbed per unit surface area. Figure 11 shows plots of number of molecules adsorbed per unit surface area vs. inhibitor concentration at various fixed applied potentials.
To estimate the effective area covered by the inhibitor molecule, Stuart and Briegleb atom models were used to construct the DPP molecule. The molecular model was arranged in a position corresponding to a possible configuration of the adsorbed molecule and then photographed. In arranging the models, the nitrogen atom was placed in a position corresponding to the formation of a metal-nitrogen bond. Two possible configurations were assumed, an extended configuration with the molecule lying flat on the surface, covering maximum area, and a compacted configuration with the molecule standing perpendicular to the metal surface, covering minimum area. Figure 12 shows three possible configurations. Pictures a and b show the extended configuration, the molecules lie flat on the surface, whereas picture c shows the compacted configuration, the molecule stands perpendicular to the




















Figure 11 - Dependence of Surface Coverage on 4,7-Diphenyl-1,10phenthroline Concentration at Various Fixed Potentials









10.0
C14
I


8.0
C
~-0. 400 V



S6.0


4-J

00 10 4.0 Q)
-e
o 0 -0. 200 V


O0. 000 V r 2.0

Q)
0
z

0


1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
Concentration in Solution, M x 106





















(a)












(b)












(c)




Figure 12 - Three Possible Configurations of 4,7-Diphenyl1,10-phenanthroline.
(Molecule built from Stuart and Briegleb Atom
Models)







metal surface. Due to the bulky steric effect of the two phenyl groups at the 4,7 position on DPP, these phenyl groups are not coplanar with the other parts of the molecule (the planar phenanthroline molecule) and thus DPP cannot properly lie flat on the metal surface. In picture a, the two phenyl rings stay in a position which is perpendicular to the phenanthroline plane, have less repulsion between atoms and the molecule probably will be more stable, but the contact with the metal surface seems quite weak. In picture b, the two phenyl groups are rotated to a position that is neither coplanar with nor perpendicular to the phenanthroline group. The intramolecular repulsion may be larger but the molecule has better contact with the metal surface. In picture c, the two nitrogen atoms rest on the metal surface and a strong nitrogen - metal bond is expected to be formed. The two phenyl groups are located on top of the DPP molecule and do not take part directly in the adsorption process. The effective areas covered by the three molecular configurations were calculated from the projected areas in Figure 12. A carbon - carbon double bond length of 1.395 � 0.003 . in aromatic rings was used as a scale factor to estimate the effective area.55 The effective area calculated for configurations in picture a, b and c are 132.4 g2, 153.0 R2 and 87.5 2, respectively. Because of the complexing character of DPP, the configuration of picture c seems to be the most probable arrangement for the inhibitor molecule adsorbed on the metal surface. Thus the estimated effective






area of 87.5 g2 was used for the surface coverage calculation. The plot of fractional surface coverage vs. inhibitor concentration at fixed applied potentials (-0.600 V, -0.400 V, -0.300 V, -0.200 V and 0.000 V) is shown in Figure 13. From this figure it is possible to plot the fractional surface coverage vs. applied potential at constant inhibitor concentration in the bulk solution. Figure 14 shows such a plot for seven different inhibitor concentrations, 1.0 x 10-5 M to 5.0 x 10-7 M.


Potentiostatic Polarization


In order to maintain the inhibitor concentration at a

constant value during an experiment, a different electrochemical cell with a larger test electrode compartment was used in the polarization experiments. The volume of solution used in this study was about 100 ml, so that the total number of moles of inhibitor adsorbed on the electrode surface (< 10-9 mole, depleted from bulk solution) becomes negligible (< 1%) when compared with the total number of moles of inhibitor in the bulk solution (> 1.0 x 10-7 mole). Solutions with 1.02 x 10-6 M, 3.68 x 10-6 M, 1.28 x 10-5 M and 1.15 x 10-4 M DPP in 0.1 M HC supporting electrolyte were used in the polarization experiments.

After the twenty minutes prepolarization period at

-0.900 V, net currents were cathodic with values of (8.0 �

0.5) x 10-3 A cm-2. The open circuit potential, Eoc, of the prepolarized electrode in fresh deaerated bulk solutions


















Figure 13 - Dependence of Fractional Surface Coverage 0 on 4,7-Diphenyl1,10-phenanthroline Concentration at Various Fixed Potentials







































8.0


1.0 2.0 3.0 4.0 5.0 6.0 7.0
Concentration in Solution, M x 106


1.01 0. 81


0.61 o 4( 0.2(



















































0.0


-0.5 -0.4 -0.3 -0.2 -0.1
Applied Potential, V (vs. S.C.E.)


Figure 14


- Fractional Surface Coverage 0 vs. Applied Potential at Various Constant 4,7-Diphenyl1,10-phenanthroline Concentrations


0.9 0.8 0.7


0.6 0.5 0.4 0.3 0.2


0.1


-0.7


-0.6




43

was measured after 10 minutes. The recorded Eo 's are shown in Table I. Figure 15 shows the polarization behavior of the pretreated steel electrode in 0.1 M HCI electrolyte and in 0.1 M HC electrolyte plus various DPP concentrations. The polarization started from -0.700 V and the potentials were shifted anodically to a potential of +0.200 V. As the potential was shifted anodically, the measured current decreased and became zero at the rest potential, Er* The rest potential is defined as that potential at which the absolute values of the internal anodic and cathodic currents become equal, resulting in a net external current flow of zero. Values of the rest potential determined here are -0.372 to 0.345 V. Table I also shows the Er's of the steel electrode in different DPP concentrations. The E., and Er measured agree very well, their discrepancies are generally within 10 mV which may be attributed to the change of surface state of the electrode after polarization has started. The cathodic Tafel slopes, i.e. a plot of 6E/61oglil, of the polarization curves are plotted in Figure 16. The corrosion current of the steel electrode at different DPP concentrations was determined from the extrapolated cathodic Tafel line at the rest (corrosion) potential. Table II shows the Tafel slopes and corrosion currents in solutions of various DPP concentrations.



Polarization Behavior of the Steel Electrode at Various
Constant Fractional Surface Coverages 0 of DPP


From Figure 14, it is clear that the fractional surface

coverage 0 of DPP on the steel electrode is not only a function









Table I. Open Circuit Potential, Eoc, and Rest Potential, Er in Solutions of Various
DPP Concentrations


DPP Concentration 0 1.02 x 10-6 3.68 x 10-6 1.28 x 10-5 1.15 x 10-4 in 0.1 M HCI Solution M M M M M E V -0.382 -0.378 -0.373 -0.365 -0.349 E r V -0.372 -0.367 -0.363 -0.355 -0.345











3.00 - . I- U A% ..J '
1.15 x 1o-4 H (o)





2.001 , //


e,,% , / . ~1.00- 11,'






0.00-I
II r


I I - p.r~




-0.600 -0.400 -0.200 0.000 Applied Potential, V (vs. S.C.E.) Figure 15 - Potentiostatic Polarization Current vs. Potential Curves (in the Absence
and Presence of Various Constant 4,7-Diphenyl-l,10-phenanthroline Concentrations).















0.10 M HC1 (o) 1.02 x 10-6 M (") 3.68 x 10-6 1 (1 ) 1.28 x 10-5 M (A) 1.15 x 10-4 M (0)


-0.600


-0.500


Applied Potential, V (vs.


Figure 16 - Plot of Cathodic Tafel Slopes
static Polarization Curves.


-0.400 S.C.E. )


from the Potentio-


h~ %


3.00


2.00


1.00


0.00 -







Table II. Tafel Slope (Cathodic) and Corrosion Current of the Steel Electrode in
Solutions with Various DPP Concentrations (Determined from Figure 16)


DPP Concentration 0 1.02 x 10-6 3.68 x 10-6 1.28 x 10-5 1.15 x 10-4

in 0.1 M HCI Solution M M M M M


Tafel Slope
1 116 122 130 141 199 (mV decade )16

Corrosion Current
12.6 10.2 7.4 6.5 9.8icr (pA cm-2)









of the inhibitor concentration but also a function of applied potential. Thus, the polarization behavior of the steel electrode presented in Figure 15 only shows the currentpotential relationship at constant bulk concentration. The actual coverage varies from point to point (potential to potential) in the polarization curve.

Since the fractional surface coverages of DPP on the steel electrode at various concentrations and at various applied potentials are known (Figure 13), it is possible to construct the polarization curves at constant surface coverage. A more detailed surface coverage vs. concentration plot is shown in Figure 17, where the curves representing the surface coverage vs. concentration relationship at -0.700 V,

-0.500 V, -0.350 V and -0.320 V were extrapolated or interpolated from Figure 14. Figure 18 shows a plot of the logarithm of the current density as a function of DPP concentration at various applied potentials. From Figure 17, five straight lines representing a constant fractional surface coverage of 0.1, 0.2, 0.3, 0.4 and 0.5, respectively, were drawn across the curves. The intercept obtained shows the concentration of the bulk solution needed for the specified fractional surface coverage at that (specified) potential. After interpolating the needed DPP concentrations (at the specified applied potential) from Figure 18, the polarization current density at the specified fractional surface coverage and applied potential was obtained. Figure 19















Figure 17 - Fractional Surface Coverage 0 vs. 4,7-Diphenyl-1,10-phenanthroline Concentration at Various Fixed Potentials. (Data for -0.700 V, -0.500 V, -0.350 V and -0.320 V are Extrapolated or Interpolated from Figure 14)







































10.0


4.0 6.0 8.0
Concentration in Solution, M x 106


0.80


0.60 0.40 0.20


2.0

















Figure 18 - Polarization Current vs. 4,7-Diphenyl-1,10-phenanthroline
Concentration at Various Fixed Potentials











~-0. 700 V 3.00 -0.650 V ~-0. 600 V


~-0. 550 V 2.00 -0. 500 V



0
-- ~ 0. 320 V


1.00t0.0V




-0.350 V
0 2.0 4.0 6.0 8.0 10.0
Concentration, M x 10O6








shows such a plot for constant fractional surface coverages of 0.0, 0.1, 0.2, 0.3, 0.4 and 0.5, respectively.

The major difference between Figure 19 and Figure 15 is that the cathodic Tafel slopes of all the polarization curves in Figure 19 become identical even though their surface coverages are different. The polarization curves shift downward regularly with increase of fractional surface coverage. The inhibitor effect on both the cathodic and the anodic part of the curves is clearly evident. The plots of anodic polarization curves in Figure 19 were stopped at

-0.300 V because after the steel electrode has become passivated, the reproducibility of the measured polarization current is not sufficient to allow for the correction of the surface coverage effect on the polarization curve.


















3.0 ..,'-,
3. 0.40 (A) 0.50 (-)







2.0




o , ,,' "jx,


"- 1.0 I ' ~ C14



1-4 Ill ,|







� IN ,II
1.01








0- -
-0.7 -0.6 -0,5 -0.4 -0.3
Applied Potential, V (vs. S.C.E.)

Figure 19 - Constructed Potentiostatic Polarization Curves at
Various Constant Fractional Surface Coverage 0














CHAPTER IV
DISCUSSION


Selection of Metal Adsorbent and Adsorbate



The selection of stainless steel as a solid adsorbent was based on its versatility as a construction material and mainly on its corrosion resistance properties. It has a low dissolution rate in acidic solution and is therefore expected to give less interference in the spectrophotometric measurements.

More critical for the success of the contemplated experiments is the proper choice of an organic adsorbate as inhibitor. Basically, the organic adsorbate should fit the following requirements in order to assure that acceptable results could be obtained. First, the organic material used should be adsorbed on the solid adsorbent and function as an acceptable corrosion inhibitor. Second, the organic adsorbate used as an inhibitor should have a very strong absorption band in the UV absorption range in order to facilitate measurements at very low concentrations (< 10-5 M ). For this purpose, the extinction coefficient of the absorption peak should be larger than 1.0 x 104 cm- M-. This requirement is predicated by four factors, the total surface area of the solid electrode, the volume of the solution, the








sensitivity of the spectrophotometer used, and the space requirements of the adsorbed molecule on the electrode surface. The larger the surface area of an electrode, the more organic adsorbate will be adsorbed and thus the greater the change in solution concentration. Similarly, the smaller the solution volume the larger the change in solution concentration for a finite amount of molecules adsorbed on the electrode surface (amount of organic inhibitor depleted from the bulk solution). The relationships between these factors can be demonstrated by looking at the following calculations. Assume a system, similar to the one designed for the present
2
work, where the electrode has a surface area of 5.0 cm , the volume of the solution used is 6.5 ml and the organic adsorbate molecule has a projected area of 50 g2 on the electrode surface when it is adsorbed. Then it takes 1.0 x 1015 molecules to form a monolayer coverage on the solid electrode. If about 10% of the electrode surface is covered (0 = 0.1), the total number of molecules depleted from the bulk solution is 1.0 x 1014. For a reasonably precise spectrophotometric measurement, a 1% variation in the original concentration (proportional to the optical signal) should be detectable. Thus, a bulk solution with an original total number of 1.0 x 1016 organic adsorbate molecules should be used, i.e., a concentration of about 2.5 x 10-6 M. From Beer's Law, A = abc, where A is absorbance, a is the molar extinction coefficient, b is the light path and c is the concentration of
4 -I -i
the solution. With a = 1.0 x 10 cm 1 if b = 1 cm, c








2.5 x 10-6 M, A is equal to 2.5 x 10-2. Thus, in order to detect the desired 1% concentration variation, a rather sensitive spectrophotometer is needed. The Aminco DW-2 spectrophotometer which has a scale (full scale) down to 0.005 absorbance units fits this requirement. From the above example, it is clear that the larger the molar extinction coefficient of the organic adsorbate, the higher the solution concentration allowable and the smaller the measurement errors. The third factor which needs to be considered in the selection of the organic adsorbate is the solubility of the organic molecule in acidic aqueous solution. There is a large number of organic molecules which are potentially good corrosion inhibitors and also have a strong absorption band in the UV absorption region. Most of them, however, are insoluble in aqueous solutions. Thus, the selection is narrow and restricted. The organic compound chosen here as a corrosion inhibitor for this work, 4,7-diphenyl-l,10-phenanthroline fits the requirements mentioned above reasonably well.


Precision of the Spectrophotometric Measurement
and Surface Coverage Determination


As mentioned previously, the dissolution of metal ions in the solutions studied causes serious interference to the spectrum of the organic inhibitor and thus affects the precision of the inhibitor concentration determination. The ++C +-+ N++
different metal ions (Fe , Cr , Ni ) have different








contributions (or interference) to the overall spectrum.56,57 During the polarization of the electrode, more and more ions are dissolving in solution and the interference to the measured spectrum becomes larger and larger. However, at a constant polarization potential, the concentration of metal ions should be proportional to the polarization time (assuming the surface area of the electrode is invariant during polarization). The interference is then a systematic error, and may be eliminated on the basis of a calibration curve. In fact, the proportionality of the spectrum change with the concentration of metal ions dissolved in solution (during polarization) has been used as a method for the quantitative determination of the corrosion rate of metals or alloys. For example, Foley and Alexander used the UV and Visible spectra to monitor the rate of iron corrosion.58 They also contributed to the knowledge of the iron dissolution process through a spectrophotometric study of the various iron species and the corrosion products that appeared in different solutions after the corrosion reaction has proceeded for a finite time. Because of the existence of multiple elements in the steel sample and because the dissolution rate of each metal component may vary with the applied polarization potential, the spectral interference caused by the presence of matrix ions will vary from potential to potential and the calibrations become very tedious work. The proposed modified dual-wavelength method supplies an alternative solution to this problem. From this method, it is possible to build a








simulated matrix line (interference spectrum due to the presence of Fe++ Cru, Ni+ and Fe+) which differs from the true matrix line only at the horizontal location (absolute absorbance). The accuracy of the inhibitor concentration measurement depends upon how exactly the matrix line is reinstalled. Here, only three wavelength points were used for the installation of the matrix line. The simulated matrix line built from these three points should agree reasonably well with the true matrix line if no special absorption peak on the matrix line overlaps with the absorption peak of the organic inhibitor (used for quantitative measurement). That is the case here where the matrix line of the metal ions generally appears as a decay curve from lower wavelength to higher wavelength and no absorption peak was found in the scanned UV range of 250 nm to 350 nm.

The prerequisite for the correction of systematic errors and the usage of the spectra for the determination of the corrosion rate is that the solution must be completely free of oxygen. If there is any trace amount of oxygen present, part of the ferrous ion will be oxidized to ferric ion. Ferrous ion and ferric ion have completely different UV spectra. According to Effrenfruend and Leibengoth, Fe+++ has an absorption maximum at 304 nm with a molar extinction coefficient of a = 3300, whereas, at the same wavelength, Fe++ has a molar extinction coefficient of about a = 10.60 The drastic difference between these two species in their UV spectra would cause a potential additional error in the inhibitor








concentration determination. To avoid the possibility of solutions used here to be contaminated by ambient air during transfer to and from the cuvette, a special cap was designed for the cuvette and a stream of helium was applied to the cuvette before and after the transfer of the solution.

Even if the solution was contaminated slightly with

ambient air and the matrix line was not a smooth curve, it would still be possible to reinstall the true matrix line with reasonable success. The reinstallation procedure would become more tedious, i.e., instead of just one horizontal line on the spectrum in Figure 5, two or more horizontal lines would be drawn. Those interception wavelength points on the same line represent the group of wavelength points which have the same absorbance and the same molar extinction coefficient. Checking the wavelength points of the same group on the distorted spectrum, the relative difference of absorbance at those wavelengths allows to deduce a more exact true matrix line and thus to determine a more accurate inhibitor concentration.

An optimal estimation of the accuracy for the concentration determination using the modified dual-wavelength method is about � 5%. The accuracy of the surface coverage measurement may be very poor, however, because it involves not only the errors that may come from solution transfer and from determination of the solution volume, but also from the estimation of the true surface area of the electrode which is difficult to determine. Thus the accuracy obtained is







probably not better than � 20%. Even though a constant roughness factor has been used for the conversion of the true surface area from the geometric surface area, deviations of the true surface area still exist from experiment to experiment and this error passes into the surface coverage measurement. The deviation essentially comes from two sources. First, it is impossible to produce an exactly reproducible surface state on the electrode. Even though the same surface treatment procedures were followed, it is still not to be expected that the surface states obtained are the same. Secondly, the surface of the treated electrode roughens during polarization. The standard reduction - oxidation potentials of Fe, Cr and Ni (the main components in the 304 stainless steel sample) are -0.707, -1.007 and -0.517 V, respectively, in aqueous solution at 250C. In the surface coverage measurement experiments, the steel electrodes had been polarized at -0.600, -0.400, -0.300, -0.200 and 0.000 V, respectively, for 50 minutes before the spectrophotometric measurements were taken. The applied polarization potentials are all positive to the standard redox potentials of Fe, Cr and Ni listed above (except at -0.600 V which is negative to the standard redox potential of Ni). Thus an appreciable amount of iron, chromium and nickel should have dissolved into the solution and the electrode surface might have become quite rough. The degree of roughness might depend upon the applied potential and the length of the polarization period. The more anodic the applied potential is, the larger the








metal dissolution rate will be and the sooner the surface will become rough. As long as the polarization is still going on, more and more metal ions will be peeled off from the electrode surface as time elapses and thus the rougher the surface may become. But after a certain period of polarization the surface may have reached a steady state and the surface area of the electrode will then just fluctuate within a certain range. In this experiment, 50 minutes were allowed for the electrode to be polarized and a steady state was assumed to have been reached.

Since the peeling of metal ions from the electrode surface may produce various surface irregularities (e.g., kinks, terraces, cracks and projections) the true area may be considerably greater than the geometric area and the roughness factor may be quite large. A roughness factor of 4.0 was used in this experiment and seems reasonable and justifiable.


The Dependence of Surface Coverage on Inhibitor Concentration and Applied Polarization Potential


The dependence of fractional surface coverage on inhibitor concentration and applied potential is shown in Figures 13 and 14. At constant applied potential the surface coverage of DPP increases with increasing DPP concentration in bulk solution (except at the applied potential of -0.200 V and 0.000 V), but levels off at high bulk concentrations, c > 10-5 M for potentials more negative than -0.400 V. At constant concentration, generally, the adsorption of DPP decreases with the








shift of the applied potential in the anodic direction. The drastic decrease of the adsorption found at -0.300 V in Figure 14 can be attributed to the change of the surface state from active to passive. Stainless steel starts to become passivated at around -0.300 V. The adsorption levels off in the potential range -0.200 V to 0.000 V, where complete passivation exists. In the passive state, oxides are present on the metal surface, which intensify the hydrophilic properties of the surface.61 Thus the adsorbability of organic adsorbates decreases sharply and becomes constant in the passive potential range.


Determination of the Adsorption Isotherm



When there is equilibrium between a species in the adsorbed state and in the bulk of the solution at an ideally polarized electrode, the corresponding electrochemical potentials are equal, i.e.,


p* + RT ln{f(r)} - 11* - RT ln a = 0 (1)


where the P and P* are the standard electrochemical potentials in the adsorbed state and in solution, respectively, the quantity f(P), a function of the surface concentration r, represents the activity of the adsorbate and a is the corresponding activity in the bulk of the solution. 62 The standard free energy of adsorption, -Go, is equal to - *. Thus, equation (1) can be rewritten as








AU = RT ln a - RT ln {f(r)}


f(r) = a exp (-AG�/RT)


r = r (Ka) (2)

where K represents the equilibrium constant of adsorption and is equal to exp (-T�/RT). Equation (2) shows that the isotherm is some function r(Ka) of the product of the bulk activity a and the quantity K which depends on the electrical parameters such as charge and potential characterizing con62
ditions at the ideally polarized electrode. The explicit form of equation (2) depends on particle - particle and particle - electrode interactions. Equations (1) and (2) have been used as the starting point to derive different isotherms. Several common adsorption isotherms which have been used to represent adsorption on electrodes are listed below:

1. The Henry isotherm

0 = Ka (3) where K and a are as defined above and 0 is the fractional surface coverage of the electrode.63

2. The Freundlich isotherm


0 = Kan (4)

where 0 < n < 1.64

3. The Langmuir isotherm65


o Ka (5)







4. The Frumkin isotherm
0
exp (-2bO) = Ka (6) where b is a quantity characterizing interactions between the adsorbed particles.66

5. The Virial Coefficients isotherm


0 exp (-2b@) = Ka (7)

where b is always less than zero and represents repulsive interaction between adsorbed molecules.67

6. The Volmer isotherm68


exp =Ka (8)


7. The Amagat isotherm67


1- x (-u) Kan (9)

8. The Helfand-Frisch-Lebowitz isotherm69


0 exp ( 2 2 = Ka (10)
1 - 0(1 0)2

9. The Hill-de Boer isotherm70'71
0 0
1 - exp ( - ) exp (-2bO) = Ka (11)


10. The Parsons isotherm72


0 exp ( 2 o ) exp (-2bO) = Ka (12)
1i - 0)2








11. The Temkin isotherm73


exp (bo) - I
1 - exp {-b(l - 0)}

and

12. The Blomgren-Bockris isotherm


1 - exp (p02 - qO3) = Ka (14)


where p and q are constants, expressed in terms of the dipole moment, the area occupied by the adsorbed molecule and the dielectric constant of the surface layer.74

The simplest isotherm is Henry's isotherm which follows directly from the equation of state which treats the adsorbate as a two-dimensional ideal gas and assumes that the heat of adsorption is directly proportional to the activity a, in the bulk of the solution. The Freundlich isotherm was derived from the assumption that the heat of adsorption falls logarithmically with coverage.

The Langmuir isotherm is based on the equation of state for noninteracting particles adsorbed on fixed sites. It can be derived simply by the kinetic approach or thermodynamically by applying the law of mass action to an equilibrium process. Because the Langmuir isotherm neglects the effect of mutual repulsions between the adsorbed species, it is rarely ever applicable to real systems.

The Frumkin isotherm was derived from the Langmuir isotherm by assuming that the apparent standard free energy of adsorption is a linear function of coverage, i.e.,








K = K0=0 exp (2bo) (15)

By substituting equation (15) into equation (5) the Frumkin isotherm was then obtained. The empirical term exp (-2b0) which was first introduced by Frumkin in the equation of state corresponding to the Langmuir isotherm accounts for the particle - particle interactions between adsorbed
75
species. The parameter b is positive if attraction exists between adsorbed particles and is negative if there is repulsion between adsorbed particles.

The same modification can be applied to derive the

Virial isotherm which takes the form of a Henry isotherm with the factor K defined by equation (15). A similar derivation can also be made for the Volmer, Amagat and Helfand-FrischLebowitz isotherms, but the dependence of K on coverage is more involved than that expressed by equation (15). The Volmer equation treats adsorption as a two-dimensional fluid of rigid particles and so does the Helfand-Frisch-Lebowitz isotherm. It follows from equation (15). that analysis of isotherms as a deviation from Henry's Law or a Langmuir isotherm should lead to a linear variation of the standard free energy of adsorption with coverage provided that either the Virial or Frumkin isotherm can be applied. Some results given by Blomgren, Bockris and Jesch show that this correlation is fairly well obeyed, at least when the coverage is not too high.76 Parsons and Hill-de Boer further modified the Helfand-Frisch-Lebowitz and Volmer isotherms respectively by introducing the empirical term exp (-2b0) to account for








particle - particle interaction in the same way as in the Frumkin and Virial Isotherms.

Based on the assumption that the variation of the standard free energy of adsorption is a linear function of surface coverage, the Temkin isotherm was derived by further assuming that the surface is made up of a great number of small patches, ds, on each of which the Langmuir isotherm is applicable, but with the standard free energy of adsorption increasing in small steps. Equation (13) can be derived by integrating the equation


do K(s) a ds
d + K(s) a

where K(s) is the equilibrium constant for adsorption over the whole area, for s going from zero to unity.

Expressions for the particle - particle interaction

other than the term exp (-2b0) in the isotherm can be derived by transposition of the corresponding treatment for metal gas adsorption. Blomgren and Bockris started with the Langmuir isotherm and corrected the standard free energy of adsorption, in the case of ion adsorption, for coulombic interaction and dispersive forces, introducing the following K factor into the Langmuir isotherm74 K = K0=0 exp {-(p02 - q03)} They obtained the equation for the adsorption of organic ions as









0-exp (po qO3) Ka (14)


Similarly, Conway and Barradas wrote the Langmuir isotherm with the following factor for dipole - dipole interaction in the adsorption of an uncharged substance77

K = K_=0 exp {-(p'O3/2 - q03)}


The isotherm obtained differs from that of Blomgren and Bockris by having the pO term replaced by p'O3/2
S,3/2 3
exp (p- qO = Ka (16)



In equations (14) and (16) the term 02 and 03/2 will be predominant over the 03 term if 0 is not too close to unity.

It is known from experimental data that the adsorption isotherms of organic substances may be either sigmoid (Sshaped) or logarithmic, dependent on whether attractive or repulsive interaction predominates between the adsorbed particles.78'79 If attractive interaction predominates and the isotherm is S-shaped then the isotherm is characterized by the following feature. The slope in the middle region, when 0 = 0*, must be greater than the initial slope at 0 = 0. In other words, the following condition must be satisfied in this case:

f'(E*) > 1 (17)

f'(0)

where f'(O*) and f'(0) represent the value of 6016a at 0 = 0*







0 = 0. Analysis of the isotherms shows that only equations
(6), (11), (12) and (14) satisfy the condition (17). The others are therefore not applicable to experimental S-shaped adsorption isotherms. In isotherms (7) to (13), one arbitrary parameter b is used to represent two-dimensional interaction between the adsorbed particles while in the Blomgren-Bockris isotherm (isotherm (14)) the two-dimensional interaction is represented with the aid of two arbitrary parameters, p and q.

The change of the shape of isotherm with parameter b in equations (6), (11) and (12) is shown in Figure 20, where the dependence of 0 on the relative concentration a/a0=0, is calculated from equation (6) for different values of b.80 In Figure 20, the isotherm changes from a logarithmic form for b < 0 (curve 1, repulsive interaction exists between adsorbed particles) to a sigmoid shape (curves 3 and 4, characterizing attractive interaction) when b > 0.

To express the adsorption behavior of DPP on a stainless steel electrode at different polarization potentials with the proper isotherm curve fitting procedures are required in order to find the parameters for the best fit of the experimental data.
It is obvious from Figure 13, a plot of surface coverage

0 vs. concentration, that the adsorption behavior of DPP on a stainless steel electrode cannot be expressed by Henry's Law. Figure 21 shows the plot of the Langmuir isotherm for four different polarization potentials, -0.600 V, -0.400 V,

-0.300 V and -0.200 V, respectively. The figure indicates









1.0









S 0.5
0

C)







0 1 2 3 a/a� = 0.5 Figure 20 - Dependence of the Surface Coverage 0 on the Relative Concentration
a/aO = 0*, Calculated from Equation (6) for Different Values of b.
1) b =-1; 2) b = 0; 3) b = 1; 4) b = 1.5. See Reference 79














1.0


0.9 -0.600 V -0.400 V


0.8


0.7 0.300 V


0.6






0.4
-0.200 V

0.3 0.2


0.1 -/



0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

Concentration, a x 106


- Data Plotted According to


Figure 21


the Langmuir Isotherm








that the adsorption does not follow the Langmuir type. Due to the similar character of the Langmuir, the Volmer and the Helfand-Frisch-Lebowitz isotherms, fitting of the experimental data to these isotherms is also ruled out. Figure 22 shows the logarithmic plot of 0 vs. concentration. A set of straight lines will be obtained and the slopes will characterize the Freundlich parameter n, if the adsorption behavior fits the Freundlich isotherm. The definite curvature of the plots indicates that the adsorption behavior does not obey the Freundlich equation.
By adjusting the parameter b in the empirical term

exp (-2b0) in equations (6), (7), (11) and (12), it is possible to fit the experimental data to the represented Frumkin, Virial Coefficient, Hill-de Boer and Parsons isotherms to a certain degree. The determination of the parameter b for the different isotherms was achieved by trial and error. Figures 23, 24, 25 and 26 show these plots. For the more complicated isotherms, such as equations (13), (14) and (16), the curve fitting process was achieved with the help of computer calculations. Figures 27 and 28 show the plot of equations (14) and (16), respectively. It is clear from these figures that the isotherms fit the experimental data very well only in the low surface coverage range (0 < 0.4), while in the higher surface coverage range (0 > 0.5) curvatures are inevitably obtained. Table III lists the fitted particle - particle interaction parameter b and the calculated equilibrium constant of adsorption, K, of different isotherms. The adsorption













-0.600 V
-.400 V





-0. 300 V


0 -0.4

0


-0.7



-1.0
I 1 1

-0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 log (a x 106) Figure 22 - Logarithmic Plot of the Freundlich Isotherm
















0.9


0.8
C


0.7 -0.600 V


0.6 '
Q
,0
CJ -0. 400 V
0.5
x

o 0.4
li



0.3
-0.200 V

0.2


0.1



0.5 1.0 1.5 2.0 2.5 3.0 a x 106


Figure 23 - Data Plotted According to the Frumkin Isotherm
















0.8 0.7 0.6 0.5 0.4 0.3


0.2


0.1



0.5 1.0 1.5 2.0 2.5 3.0 a x 10 6 Figure 24 - Data Plotted According to the Virial Coefficient
Isotherm


















0.8 0.7


( 0.6

-0.600 V ) 0.5

o -0.400 V S0.4



0.3 -V


0.2 0.200 V


0.1



0.5 1.0 1.5 2.0 2.5 3.0

a x 106


Figure 25 - Data Plotted According to the Hill-de Boer Isotherm

















































1.0 1.5 2.0 2.5 3.0

a x 106


Figure 26 - Data Plotted According to the Parsons Isotherm


4.0 3.0


(D
0q ,0









0li


2.0 1.0


0.5

















0.8 0.7


0.6 -0.600 V
0

0.5 -0.400 V


0.4


10. 0.

0 .3 --0 . 200 V

0.2 0.1



0.5 1.0 1.5 2.0 2.5 3.0 a x 106 Figure 27 - Data Plotted According to the Blomgren-Bockris
Isotherm






80











0.8 o


0.7 -0.600 V



0.6
o -0.400 V
0.5

CnJ
0 0.4


0.3



CD 1 0.2 - . 0


0.1



0.5 1.0 1.5 2.0 2.5 3.0 a x 106 Figure 28 - Data Plotted According to the Conway-Barradas
Isotherm








constants, K, were obtained from the slopes of the plots. The least squares method was used to find the values of the slopes. From Table III it is seen that the equilibrium constants of adsorption obtained from the different isotherms agree quite well at the same applied polarization potential while the interaction parameters deviate from each other.

There are two thoughts concerning the dependence of the interaction parameter b on potential.81 Parsons, in agreement with Frumkin's original suggestion, considers that b is independent of the potential or of the charge on the electrode, and that the effect of the electrical field at the interface appears solely in term K. Damaskin retains the dependence of K on potential but also assumes that b is a function of potential.81 Table III shows that the parameter b is the same at -0.600 V as at -0.400 V but differs at -0.300 V and -0.200 V. Since the stainless steel electrode is partially passivated at -0.300 V and totally passivated at -0.200 V, the change in surface state may well account for the change in the b parameter at those potentials. A general decrease of parameter b from -0.400 V to -0.300 V and -0.200 V agrees also well with the hydrophilic property of the oxides developed on the electrode surface. Considering that the differences in interaction parameter b at -0.300 V and -0.200 V are due to the change of the state of the surface and that the parameter b is the same at -0.600 V and -0.400 V, it is concluded that b is independent of the applied potential in the system stainless steel/DPP. This would support Parsons arguments.





Table III. Particle - Particle Interaction Parameters b; p, q and p', q Used for Fitting
Data into Various Isotherms and their Calculated Adsorption Constant K and
Standard Free Energy of Adsorption AGO


Applied Potential
Isotherms -0.600 V -0.400 V -0.300 V -0.200 V


b 0.50 0.50 0.00 -0.050 Frumkin K x 10-5 2.36 � 0.01 1.76 � 0.01 1.45 � 0.01 1.24 � 0.01

-AG -1 7.31 � 0.01 7.15 � 0.01 7.04 � 0.01 6.94 � 0.01 in kcal mole


b -0.50 -0.50 -0.60 -1.00
Virial K x 10-5 2.61 � 0.01 1.94 � 0.01 1.39 � 0.01 1.22 � 0.01 Coefficient
-G a 7.38 � 0.01 7.21 � 0.01 7.01 � 0.01 6.93 � 0.01 in kcal mole-" - ".....



b 1.40 1.40 1.00 0.030
Hill-de Boer K x 10-5 1.76 � 0.01 1.31 � 0.01 1.17 � 0.01 1.16 � 0.01

inG c 7.15 � 0.01 6.69 � 0.01 6.91 � 0.01 6.90 � 0.01
in kcal mole-







Table 1ll--Extended


Blomgren and Bockris


q

K x 10-5
-o
- G-1
in kcal mole


-0.50

0.50

2.02 � 0.01 7.23 � 0.01


-0.50

0.50

1.50 � 0.01 7.06 � 0.01


-0.20

-0.20

1.28 � 0.01 6.96 � 0.01


0.20

-0.60

1.17 � 0.01 6.91 � 0.01


p' 1.00 1.00 0.50 -3.0

q 1.00 1.00 0.50 40 Conway and
Barradas K x 10-5 2.38 � 0.01 1.77 � 0.01 1.30 � 0.01 1.07 � 0.01


no a 7.33 � 0.01 7.16 � 0.01 6.97 � 0.01 6.86 � 0.01
in kcal mole-








With the exception of that of the Virial Coefficient isotherm, the interaction parameters obtained here are all positive. This indicates that an attraction exists between the adsorbed species. Generally, an attractive interaction exists in the case of competitive adsorption. Due to the fact that specific adsorption of CI- may be present in the system studied, and that the DPP will first adsorb mainly in the form of DPPH+, the coulombic interaction between the adsorbed Cl- and DPPH+ may enhance the adsorption of DPP on the polarized stainless steel electrode surface.

The standard free energy of adsorption To which relates to the adsorption constant K by -RT ln K, was also calculated for the different isotherms and is listed in Table III. The standard free energy of adsorption calculated at the same potential agrees very well between the different isotherms. The values obtained here are comparable to those of other organic adsorbated reported in the literature.82



Adsorption and the Structure
of the Electrical Double Layer


The "charge" on an electrode signifies the presence near the surface of unequal amounts of mobile electrons and their accompanying positive ions; the counter charge consists of an excess of ions of one type of charge in the adjacent solution. The analogy of such a system to a parallel-plate condenser, as utilized by Helmholtz, forms much of the basis of present double layer theory.83 The potential of zero charge








(p.z.c.) of an electrode with respect to some reference electrode thus serves as a natural reference potential for the double layer. But unlike the parallel-plate condenser, the potential difference across a double layer is not zero at the p.z.c., owing to the dipole contribution to the surface
M soln.
potential X and X in each separate phase, electrode and electrolyte. Therefore, a simple potential measurement does not indicate the charge on an electrode. Changes in the electrolyte may result in variations in surface potential XM and Xsoln and thus in the p.z.c. To correlate the changes in p.z.c. with solution variation and also to understand its effect on adsorption, it is useful to consider the structure of the electrical double layer. Since reviews are available, only the structural components of the interface and their contributions to the metal - solution potential difference A will be mentioned here.62'84

(i) No specific adsorption of solute:

Consider a mercury electrode in an aqueous NaF solution. Over a wide potential range (including the p.z.c.) both ions remain fully hydrated and hence separated from the electrode by at least one water molecule. The most significant interaction between ions in the solution and the electrode surface follows from the theory proposed by Gouy and Chapman, the distance of the closest approach of ions being controlled by the size of the hydrated ions. 85,86 Figure 29(a) qualitatively depicts the distribution of net charge and the potential due to this charge at a positively charged electrode.87








At the p.z.c. (Figure 29(b)) there is no net electrode charge and hence only random distribution of ions. Therefore, there 87
is no metal - solution potential difference due to ions. However, an orientation of solvent dipoles undoubtedly still exists, and this X solvent probably varies with electrode potential. 88

(ii) Specific adsorption of solute:

Many ions have the tendency to shed their primary hydration shell on the side facing the electrode and to be adsorbed in closer proximity to the electrode than hydrated ions are. Overcoming the forces of hydration in solution and causing such "specific" adsorption are electrode - ion interactions due to image, dispersion and covalent forces. The nature of specific adsorption thus allows for an amount of adsorbed charge different than the amount of opposite charge on the electrode. The double layer near a polarizable positively charged electrode in the presence of specifically adsorbed anions is depicted in Figure 30(a) and the same system at the p.z.c. is shown in Figure 30(b).87 An example of such a system is mercury in aqueous KI solution.

From the large number of adsorption measurements made

at Hg electrodes, definite trends for the electrosorption of organic molecules in general can be suggested. Electrocapillary measurements show that the decrease in interfacial tension in the presence of organic adsorbates reaches its maximum in the vicinity of the potential of zero charge. Similarly, a "bell-shaped" curve for the variation of coverage with potential







Diffuse layer


Bulk
solution
,�(
'0



,0


Potential profile
(due to ions)


plane


Helmholtz plane


Potential profile
(due to ions)


(b)


Figure 29 - Schematic Distribution of Charges and Potential
(Due to Ions) at the Metal - Electrolyte Interface at Positive (a) and Zero (b) Electrode
Charge with no Specific Adsorption of lons (See
Reference 87)












Inner Helmholtz plane Bulk solution -)


Potential profile
(due to ions)


Diffuse layer
(a)


- Inner Helmholtz plane Outer Helmholtz plane






Bulk solution :--*

Potential profile


Diffuse
layer
(b)

Figure 30 - Schematic Representation of a Metal - Solution
Interface with Specifically Adsorbed Anions at
Positive (a) and Zero (b) Electrode Charge (See
Reference 87)




Full Text

PAGE 1

SPECTROELECTROCHEMICAL STUDIES OF THE INHIBITION EFFECT OF 4, 7-DIPHENYL-l , 10-PHENANTHROLINE ON THE CORROSION OF 304 STAINLESS STEEL By HSUAN-JUNG HUANG A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1978

PAGE 2

The author dedicates this dissertation to his wife, Shang-Cheng C. Huang

PAGE 3

ACKNOWLEDGEMENTS The author would like to express his gratitude to his research director, Dr. G. M. Schmid, for the interest and assistance given to him in the course of this investigation and in the preparation of this manuscript. Thanks are also due to the members of his committee. Dr. E. D. Verink, Jr., Dr. J. D. Winefordner, Dr. R. G. Bates and Dr. R. C. Stoufer He would also like to thank Dr. K. P. Li for the generosity of sharing his laboratory facilities and helpful discussions, and Mr. R. Strohschein and Mr. W. Axson for their help in the technical aspects of this work. Xll

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii LIST OF TABLES vi LIST OF FIGURES vii ABSTRACT x CHAPTER I. INTRODUCTION 1 II. EXPERIMENTAL 12 Experimental Design 12 Experimental Technique 17 Potentiostatic Polarization 17 Chemicals and Equipment 18 Spectrophotometric Measurement 20 III. DATA AND RESULTS 23 Spectrophotometric Measurements by the Modified Dual -Wave length Method 23 Adsorption of DPP on Stainless Steel Electrodes 31 Potentiostatic Polarization 39 Polarization Behavior of the Steel Electrode at Various Constant Fractional Surface Coverages 0 of DPP 43 IV. DISCUSSION 55 Selection of Metal Adsorbent and Adsorbate 55 XV

PAGE 5

CHAPTER Page Precision of the Spectophotometric Measurement and Surface Coverage Determination 57 The Dependence of Surface Coverage on Inhibitor Concentration and Applied Polarization Potential 62 Determination of the Adsorption Isotherm 63 Adsorption and the Structure of the Electrical Double Layer 84 Proposed Inhibition Mechanism of DPP on the Corrosion of Stainless Steel in 0.1 M HCl Solution 97 V. SUMMARY 105 REFERENCES CITED 109 BIOGRAPHICAL SKETCH 115 V

PAGE 6

LIST OF TABLES Table Page I. Open Circuit Potential, E , and Rest Potential, E , in Solutions of Various DPP Concentrations 44 II. Tafel Slope (Cathodic) and Corrosion Current of the Steel Electrode in Solutions with Various DPP Concentrations (Determined from Figure 16) 47 III. Particle Particle Interaction Parameters b; p, q and p', q Used for Fitting Data into Various Isotherms and their Calculated Adsorption Constant K and Standard Free Energy of Adsorption 82 IV. Comparison of Inhibition Efficiencies (Determined from Potentiostatic Polarization Data) with the Fractional Surface Coverage (Obtained from Spectrophotometric Measurements) of 4, 7-Diphenyl-l , 10-phenanthroline on the Stainless Steel Electrode 96 vi

PAGE 7

LIST OF FIGURES Figure Page 1. Stainless Steel Electrode 13 2. Kel-F Electrode Holder 14 3. Electrochemical Cell 16 4. Block Diagram of the Potentiostatic Polarization Circuit 19 5. UV Spectrum of 4 , 7-Diphenyl-l , 10-phenanthroline 24 6. Calibration Curve (Absorbance vs. Concentration) of 4, 7-Diphenyl-l , 10-phenanthroline at 286.5 nm and 319.0 nm 25 7. Calibration Curve, Absorbance Difference vs. Concentration (AA is the Absorbance Difference at 286.5 nm and 319.0 nm) 27 8. Demonstration of the Modified DualWavelength Method 30 9. Absorbance vs. Concentration of 4, 7-Diphenyl1 , 10-phenanthroline in the Presence of 2.10 X 10-3 Fe’'"'’ 32 10. Absorbance vs. Concentration of Fe in the Presence of 6.58 x 10“° M 4, 7-Diphenyl-l , 10-phenanthroline , at 286.5 nm and 319.0 nm 33 11. Dependence of Surface Coverage on 4, 7 -Diphenyl-1 , 10-phenanthroline Concentration at Various Fixed Potentials 36 12. Three Possible Configurations of 4, 7-Diphenyl-l, 10-phenanthroline 37 vii

PAGE 8

Figure Page 13. Dependence of Fractional Surface Coverage 0 on 4, 7-Diphenyl-l , 10-phenaiithroline Concentration at Various Fixed Potentials 41 14. Fractional Surface Coverage 0 vs. Applied Potential at Various Constant 4 , 7-Diphenyl1, 10-phenanthroline Concentrations 42 15. Potentiostatic Polarization Current vs. Potential Curves (in the Absence and Presence of Various Constant 4, 7-Diphenyl-l , 10phenanthroline Concentrations) 45 16. Plot of Cathodic Tafel Slopes from the Potentiostatic Polarization Curves 46 17. Fractional Surface Coverage 0 vs. 4,7Diphenyl-1, 10-phenanthroline Concentration at Various Fixed Potentials 50 18. Polarization Current vs. 4, 7-Diphenyl-l, 10phenanthroline Concentration at Various Fixed Potentials 52 19. Constructed Potentiostatic Polarization Curves at Various Constant Fractional Surface Coverage 0 54 20. Dependence of the Surface Coverage 0 on the Relative Concentration a/a 0 _Q.u, Calculated from Equation ( 6 ) for Different Values of b 71 21. Data Plotted According to the Langmuir Isotherm 72 22. Logarithmic Plot of the Freundlich Isotherm 74 23. Data Plotted According to the Frumkin Isotherm 75 24. Data Plotted According to the Virial Coefficient Isotherm 76 25. Data Plotted According to the Hill-de Boer Isotherm 77 26. Data Plotted According to the Parsons Isotherm 78 viii

PAGE 9

Figure Page 27. Data Plotted According to the BlomgrenBockris Isotherm 79 28. Data Plotted According to the ConwayBarradas Isotherm 80 29. Schematric Distribution of Charges and Potential (Due to Ions) at the Metal Electrolyte Interface at Positive (a) and Zero (b) Electrode Charge with no Specific Adsorption of Ions 87 30. Schematic Representation of a Metal Solution Interface with Specifically Adsorbed Anions at Positive (a) and Zero (b) Electrode Charge 88 IX

PAGE 10

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 SPECTROELECTROCHEMICAL STUDIES OF THE INHIBITION EFFECT OF 4, 7-DIPHENYL1 , lO-PHENANTHROLINE ON THE CORROSION OF 304 STAINLESS STEEL By Hsuan-Jung Huang August 1978 Chairman: Gerhard M. Schmid Major Department: Chemistry The compound 4, 7-diphenyl-l , 10-phenanthroline , which has a strong absorption band in the UV absorption region, is slightly soluble in acidic aqueous solution (to about 10”^ M) and is potentially a good inhibitor, was selected to study the relationship between adsorption and inhibition on the corrosion of 304 stainless steel in 0.1 M HCl solution. A 304 stainless steel cylinder with an approximate surface 2 area of 5 cm was used as the test electrode and solid adsorbent. The current vs. potential behavior of the system was determined potentiostatically and the amount of the inhibitor adsorbed on the electrode surface was measured by the concentration depletion method using the modified dual-wavelength spectrophotometric method. The surface coverage of DPP on the steel electrode was found to depend on the DPP concentration in solution and on the applied polarization potential. At constant applied potential the surface coverage of DPP increases with X

PAGE 11

increasing DPP concentration in bulk solution (except at the applied potential of -0.200 V and 0.000 V) , but levels off at high bulk concentration, c > 10"-^ M, for potentials more negative than -0.400 V. At constant concentration, generally, the adsorption of DPP decreases with the shift of applied potential in the anodic direction. The adsorption of the inhibitor was plotted against the applied potential (from -0.700 V to 0.000 V vs. S.C.E.) in the inhibitor concentration range of 5.0 x 10 ^ to 1.0 x 10 ^ M. Fifteen different adsorption isotherms were tested for their fit to the experimental data. Five of them (Frumkin, Virial Coefficients, Hill-de Boer, Blomgren-Bockris and ConwayBarradas isotherm) fit the experimental data with a correlation coefficient >0.95 in the surface coverage range 0=0 to 0=0. 5. At higher surface coverages deviations from the isotherms are found. The particle particle interaction parameters b, p, q or p', q of the isotherms above were obtained with curve fitting processes . Judging from the interaction parameters obtained, an attractive interaction exists between the adsorbed species. The adsorption constant K calculated according to the different isotherms agrees reasonably well at the same applied potential, but decreases slightly as the applied potential shifts anodically. The calculated mean adsorption constant (from different isotherms) changes from (2.2 ± 0.3) X 10^ at -0.600 V to (1.2 ± 0.1) x 10^ at -0.200 V. The standard free energy of adsorption -AG° calculated behaves in the same manner. It has a mean value (from different XI

PAGE 12

isotherms) of 7.3 + 0.1 kcal mole“^ at -0.600 V and 6.9 ± 0.1 kcal mole"^ at -0.200 V. From the data, a series of polarization curves at various constant surface coverage with DPP, 0=0.1 to 0=0. 5 were obtained by extraand interpolation. The curves all have Tafel slopes (cathodic) identical to that in pure 0.1 M HCl, the current decreasing with increasing 0 at constant potential. According to the data, the inhibition effect of DPP is mainly due to a surface blocking effect. The adsorbed DPTH”^ ion may form a thin layer of chelate film with the Fe"'~*’ ion on the steel electrode surface which then retards both the cathodic hydrogen evolution reaction and the anodic metal ion dissolution reaction. Mechanisms for both interactions are proposed. The inhibition efficiencies of DPP (obtained from potentiostatic polarization data) were compared with the fractional surface coverages of DPP on the electrode surface (measured with the UV spectrophotometric method) . Good agreement was obtained only in the most cathodic potential region and at low DPP concentration. The inhibition efficiency ranges from about 157o at 1.02 x 10 ^ M DPP concentration to about 657o at 1.28 X 10 ^ M DPP concentration in the potential region of -0.500 V to -0.700 V. It decreases slightly as the applied potential is shifted anodically. xii

PAGE 13

CHAPTER I INTRODUCTION There are three prerequisites for corrosion to take place a potential difference, a conduction path and the availability of electrode reactions for transferring charges across the metal-solution interface. Â’ In order to control corrosion, it is then necessary to control one of the prerequisites. This is often most easily done by the use of inhibitors. An inhibitor which is adsorbed on the metal surface may function by (1) increasing the true ohmic resistance of the interface and thus limiting the charge transfer processes by a blocking effect or (2) interfering with the anodic, the cathodic, or both the anodic and the cathodic electrochemical processes. Examples of case (1) are inhibition by the formation of an oxide film or by precipitation of a nonconducting reaction O / product onto the metal. Â’ Inhibition caused by an increase in the hydrogen activation overpotential and/or a decrease of the potential difference on the metal surface are examples of case (2) . ^ Adsorption of organic inhibitors on the electrode surface (metal surface) is not only due to physical forces but is often accompanied by chemical interactions betv/een the metal surface and the adsorbate. Van der Waals forces and electrostatic interactions are two general types of forces that 1

PAGE 14

2 cause the physical adsorption of an organic substance. Van der Waals forces are generally weak and operate over the entire surface. The electrostatic interaction arises when a charged organic species (cation or anion) is close to a polarized electrode surface (negatively or positively charged). Inhibitors such as organic onium compounds, e.g., quaternary phosphonium and arsonium ions which are reported as highly effective inhibitors in acidic solution, are adsorbed electrostatically in the electrode-solution interface.^Â’^ Compounds which are able to form the onixm structure in proton containing solvents, act in a similar manner. Formation of a dative link between the metal and the organic molecule results in chemisorption. The bond is formed through the sharing of a pair of electrons between the organic adsorbate and the metal. Chemisorption might cause inhibition either by stabilizing the metal ions in the surface lattice to decrease the dissolution tendency of the metal at anodic areas or by increasing the activation operpotential for hydrogen evolution at cathodic areas . A special type of chemisorption is the complex formation between inhibitors and metal ions on the electrode surface leading to the establishment of nonconducting chelate films which usually exhibit very high inhibition efficiency. Â’ Among the factors that critically influence the extent of chemisorption are the nature of the electronic configuration of the adsorbed species. The strength of bonding is a function of the metal, since it relates to the residual valence orbitals

PAGE 15

3 existing at the metal surface; in addition, unfilled atomic orbitals are also required and these are principally of d character. An elementary approach to the influence of the electronic configuration of the adsorbed group is possible in terms of the availability of electrons for bond formation. This can be considered to be a function of the electron density and the polarizability of the functional group, or, for simple molecules, of the Vb or VIb elements that the compound contains . Inhibitors are classified as anodic type, cathodic type or mixed t)q5e inhibitors according to their interactions with electrochemical processes. Anodic inhibitors function by stifling the acidic dissolution reaction by, generally, increasing the overpotential of the anodic process. An example of this class are the sulfide series inhibitors . Cathodic inhibitors increase the overpotential of the cathodic process. The increase of the overpotential of the cathodic process shifts the corrosion potential (open circuit potential) in the negative direction and thus retards corrosion on the cathodic area. Quaternary amine and onium A 7 1 9 salts belong to this class . Â’ Â’ Inhibitors of the mixed type influence both the cathodic and anodic processes, raising the overpotential of both. Substances whose molecules consist of an organic base cation and an acid anion having oxidizing properties are widely used as inhibitors of this type. Amines, thiols, derivatives of thiourea and 13-1 3 quinolines are examples.

PAGE 16

4 In order to understand the role of organic inhibitors in the inhibition mechanism of corrosion, the adsorption behavior of the organic adsorbates on the electrode surface must be known. The earliest investigation of the adsorption process on solid electrodes was carried out by Frumkin and coworkers who studied the adsorption of hydrogen, oxygen and halide ions on metals of the platinxim group and on gold. Â’ Quantitative measurements of the adsorption of organic substances on solid metals has started only during the last 15 20 years, having become possible after the introduction of the radioactive tracer method, the pulse 18 " 20 potentiodynamic and galvanostatic methods. Recently, studies of adsorption processes at a metal adsorbent have also attracted special attention in investigations of electrochemical redox reactions, the electrochemical synthesis of organic compounds and the effect of organic additives 21-23 in the electrodeposition of metals. Due to the ideal polarizability of the liquid mercury electrode and some of its liquid amalgams to which the Gibbs adsorption isotherm can be applied without any loss or rigor and with much simplification, precise and detailed thermodynamic information has been obtained which has led to the understanding of the structure of the double layer, the adsorption behavior and the mechanisms of electrode o / processes. The use of capillary electrometry and capacitance methods for obtaining the surface charge q, and the surface excess of component i at a given potential E

PAGE 17

5 and solution composition in terms of the respective chemical potentials y2‘ • • have been reviewed thoroughly in the literature . The adsorption of organic compounds on liquid electrodes can be studied by measuring the interfacial tension y, the electrode charge q, and the differential capacity of the double layer C. Measurements of the interfacial tension y are generally done with the Lippmann capillary electrometer. The charge on the electrode (per unit area) is then obtained from the Lippmann equation q (§X] k i.e., from the slope of the surface tension vs. potential 27 curve keeping all the chemical potentials p^ constant. Alternatively, the differential capacity of the double layer is related to the electrode charge q by In contrast to the case of liquid metal electrodes, there are as yet no methods available for determining the absolute values of interfacial tension between a solid electrode and the surrounding solution. However, the dependence of the interfacial tension on the electrode potential and the solution composition can be investigated by a number of methods based on the study of mechanical properties such

PAGE 18

6 28 29 as hardness, creep, and coefficient of friction. ' From the variations of the mechanical properties with the change of potential, the adsorption of organic species on a solid electrode may thus be determined. However, on solid electrodes the work done in increasing the surface area is accompanied by irreversible changes and is therefore greater than the interfacial tension. This makes the application of a thermodynamic treatment difficult. Methods for measurement of the double layer capacity with alternating current and of potential decay curves are widely used for the investigation of adsorption phenom30-32 ena. Various bridges and procedures for capacity measurement have been described employing regular RC (resistance-capacitance) networks or transformer ratio arms. Â’ Recently, a method for automatic recording of the ohmic and capacitive component of the impedance at electrodes was 35 developed by Breiter for use with a linear potential sweep. The method employs phase-sensitive amplifiers and the C and R components of the impedance are directly plotted out together with the potentiodynamic current profile. The adsorption of an organic substance on an electrode can be O ^ calculated from capacity data using the following equation C C ^ _ o Here, 0 is the fraction of surface covered by the adsorbate, is the capacity of the double layer measured in the pure

PAGE 19

7 electrolyte solution, C is the capacity of the double layer in the solution with organic substance added, C is the capacity of the double layer at maximum (saturation) coverage of organic substance on the electrode surface. Capacity measurements at solid electrodes suffer from the disadvantage of frequency dependence, which some workers attribute to surface roughness, while others attribute it to the penetration of the solution between the electrode 07 00 and the insulating coating. Â’ Also, the adsorption of hydrogen or oxygen on electrode surfaces (e.g., on metals of the platinum group) interferes with the capacity measurement and thus makes the interpretation of data difficult. Steady state adsorption at solid electrodes can be studied with radioactive tracers. This can be done either by measuring the change of activity in solution due to adsorption at the electrode or by direct monitoring of the surface of the electrode, e.g., with a proportional counter whose window itself is also the electrode under study. The first method is applicable to studies of adsorption of ions and molecules that are strongly adsorbed so that the adsorption measurements can be carried out in very dilute solutions where the amount adsorbed is comparable in magnitude to the amount in the solution with which the electrode 39 is in equilibrium. In the procedure developed by Balashova and her coworkers a normal electrolytic cell is used in conjunction with a counting monitor through which the solution can be circulated and its activity monitored. Alternatively,

PAGE 20

8 samples of the solution can be withdrawn and counted. The advantage of this method is that the adsorption can be followed continuously in situ with control of potential at the electrode being maintained. However, the initial concentration of adsorbate in the solution must generally be below about 10 ^ M. The direct counting through a thin window electrode was developed for electrochemical adsorption processes by Blomgren and Bockris from the procedure of Aniansson and of Joliot. The adsorption of thioureaand ethylene14 C has been studied using gold foil and platinized gold foil electrodes respectively. When using this direct monitoring method, it is necessary to ensure that the measured radiation comes only from adsorbed molecules , and not from radioactive molecules in the solution in the vicinity of the electrode. To achieve this, cells are used whose cross-section is very narrow near the electrode, with most of the solution being present in reservoirs on either 46 side of the electrode. Radiation is thus counted only from the electrode and a very thin layer of solution immediately adjacent to it. Green, Swinkels and Bockris developed another technique for the determination of adsorption of radioactive species at solid electrodes which cannot A "7 conveniently be formed into thin films. ' The method consists of using an endless metal tape electrode which is run past two proportional counters to monitor both surfaces of the tape after the tape has passed through an electrolytic cell in which adsorption of the tagged material takes place .

PAGE 21

9 Radiotracer methods measure directly the total number of labeled species adsorbed on the electrode, but they are insensitive to the structure of the adsorbed species and to any changes in the nature of the adsorbed substance. If the specific surface area of the electrode is large relative to the solution volume, adsorption of a surface active substance will be accompanied by an appreciable change of the solution concentration. In such cases, the amount of substance adsorbed on the electrode surface can be calculated directly from the change in solution concentration. Depending on the nature of the adsorbed substance and the magnitude of the solution concentration, various methods have been used for the determination of such concentration changes. Conway, Barradas and Zawidzki used a spec tropho tome trie method to study the adsorption of acridine and quinoline on copper, nickel and silver. Balezin and coworkers used the chromatographic method for determining the adsorption of sodiiam benzoate and dicyclohexylamine on iron and 49 magnetite. The measurement of adsorption by determination of changes in solute concentration permits the study of adsorption effects at steady state, but it can be applied only to solutions with very small initial concentrations (less than 10 ^ M) for reasonable accuracy, and to systems which have large electrode surface area to solution volume ratios . From this brief review of the techniques used in studies of the adsorption of organic substances on solid electrodes.

PAGE 22

10 it is apparent that none of the above methods are entirely satisfactory for the purpose of thorough understanding of the adsorption behavior. Each technique has its own advantage and also its own unavoidable inherent defects. However, among the techniques, the concentration depletion method is straightforward and simple and it supplies adsorption information under steady state conditions. It also allows the study of the adsorption isotherm and of the relationship between the corrosion inhibition effect and the degree of adsorption of organic adsorbate on solid metals. Conway, Barradas and Zawidzki were the first group to develop the UV spectrophotometric technique for the determination of the steady state adsorption of organic adsorbates at a solid electrode. Later, Newmiller and Pontius applied the same technique to studies of the adsorption of photographic developers on silver. After these, few reports concerning the spectrophotometric measurement of adsorption appeared in the literature . This is due to several serious difficulties inherent in this technique. First, the UV absorption spectra of the organic adsorbate are often distorted in the presence of traces of metal ions (e.g., copper, nickel, chromium, iron) and/or complexes are formed between the metallic ions and organic adsorbates. Thus the dissolution of the metal adsorbent, e.g., in acidic solution, may give rise to interferences in the analytical determination of the organic adsorbate. Second, the applicable potential range is

PAGE 23

11 severely restricted because of either the possible accelerated dissolution rate of the metal adsorbent at certain anodic potentials or the possible electrocatalytic redox reactions of the organic adsorbate. Thus, adsorption studies could only be made successfully with noble metal electrodes, such as platinum, gold or iridium and at moderately anodic potentials. Also, these adsorption measurements were performed either on powder or on gauze electrodes and studies on ordinary cylindrical or disc electrodes have been lacking. In order to demonstrate the feasibility of spectrophotometric concentration depletion studies of organic adsorbates on metal surfaces , such adsorption work was undertaken in conjunction with a corrosion inhibition study to examine the adsorption behavior of a specifically selected organic inhibitor on the semi-noble stainless steel electrode and to determine the relationship between adsorption and inhibition.

PAGE 24

CHAPTER II EXPERIMENTAL Experimental Design The sample used as the test electrode was stainless steel, AISI 304, provided by the United States Steel Corporation. Its composition was given as 0.03 C, 0.027 P, 1.10 Mn, 0.022 S, 0.43 Si, 9.26 Ni, 18.6 Cr, 0.39 Mo and 0.04 N (weight percent). Bar stock was machined into cylinders with a diameter of 6 ram and a height of 10 mm (Figure 1) . The cylinders were annealed at 1150° C for 30 minutes and cooled rapidly by quenching in water. They were then tapped, threaded and mechanically polished at 266 rpm with 400 followed by 600 grit emery paper. Before use, the cylinders were tapped, degreased with benzene in an ultrasonic cleaner, soaked in acetone and rinsed with triply distilled water. The test electrode holder (Figure 2) was made from a Kel-F rod machined to approximately 1 cm in diameter in which a 3 mm center hole was drilled. The rod was heated and a threaded stainless steel rod was inserted so that thread protruded on both ends. The Kel-F rod was then shrunk in an ice bath to facilitate its insertion into a 1 cm ID glass stirrer bearing sleeve which had a 24/40 standard taper joint 12

PAGE 25

Figure 1 Stainless Steel Electrode

PAGE 26

14 Kel-F Rod 24/40 Standard Taper Joint Stainless Steel Rod Finger Nut Teflon Washer Teflon Washer Figure 2 Kel-F Electrode Holder

PAGE 27

15 attached. The steel sample was affixed to the protruding steel rod. The sampleto-sample holder seal could be tightened by turning a finger nut at the other end of the protruding steel rod. Teflon washers were placed between both the sample and holder, and the nut and holder, and a thin layer of Kel-F polymer wax was coated on the top surface of the test electrode to give a water proof seal. The shielded top surface of the test electrode was checked after each experiment to make sure no leakage and corrosion occurred there. If corrosion was detected, the data of that run were discarded. The electrode area (geometric) exposed 2 to solution was approximately 5 cm . The design of the electrochemical cell is shown in Figure 3. The cell was made of Pyrex and was modified from the conventional three-electrode electrolytic cell. The test electrode compartment which basically consists of a 14 mm ID Pyrex tube and a 24/40 female standard taper joint was separated from the auxiliary electrode compartment by a coarse frit and Teflon stopcock. A Luggin capillary and a stopcock were used to connect the reference electrode compartment to the test electrode compartment. A gas inlet was connected through another coarse frit to the bottom of the test electrode compartment. The frit breaks up the helium gas stream into bubbles which enhances the deaeration and stirring effect. A side arm made of a 14/35 standard taper joint connects to the middle part of the test electrode compartment. It allows the withdrawal and return

PAGE 28

16 Figure 3 Electrochemical Cell

PAGE 29

17 of electrolyte solution. Another small side arm with a 5/20 standard taper joint which also facilitates the replacement of electrolyte solution was built into the auxelectrode compartment. A platinum electrode and a saturated calomel electrode were used as the auxiliary and reference electrodes respectively in this work. Experimental Technique Potentiostatic Polarization Solutions were deaerated with helium (99.99 percent) for a minimum of three hours prior to use. Immediately before each experiment, the cell was washed with chromosulfuric acid cleaning solution, tap water, and then rinsed thoroughly with triply distilled water. The stainless steel sample was secured on the Kel-F holder, rinsed with distilled water and the solution to be studied, and immersed in solution. All samples were pretreated at -0.900 V in a separate electrochemical cell for twenty minutes to reduce possible air-formed surface films. Solution stirring was accomplished with a magnetic stirring disc and the heliimi flow which was continued throughout the experiment. All experiments were conducted at room temperature. After the pretreatment at -0.900 V the test electrode was moved quickly to another identical electrochemical cell containing the fresh deaerated solution to be studied. The ^Pplf^d potential was shifted anodically in step-wise

PAGE 30

18 increments. The magnitude of the imposed step depended on the potential range under investigation and the electrochemical reaction(s) associated with it. In regions where changes in potential caused significant changes in current density, steps of 20 to 30 mV were generally employed. In regions of approximately constant current, 50 mV steps were used. The current flowing in the auxiliary test electrode circuit was determined after an arbitrary time interval of 10 minutes . Chemicals and Equipment All chemicals used in solution preparation were reagent grade. The water employed for solution preparation was triply distilled, first from alkaline potassium permanganate and then from a two stage Heraeus quartz still, and collected in a two liter Pyrex volumetric flask. Solutions used were 0.1 M HCl. They were normalized with standardized sodium hydroxide and were always 0.1000 ± 0.0005 M. The organic compound, 4 , 7-diphenyl1 , 10-phenanthroline (bathophenanthroline) (DPP) was obtained from Matheson, Coleman and Bell Co. The organic chemical was used directly without further purification . The block diagram of the system employed in potentiostatic experiments is shown in Figure 4. Polarization was accomplished with a slightly modified Harrar potentiostat . (Two each of obsolete transistors 2N333A and HA7534 were replaced with the equivalent circuit elements 2N3568, 2N5869, and 2N3644, 2N5867) . A Keithley model 660 differential

PAGE 31

Electrometer 19 4J ct) •H *H 0 r-^ O -H cj <: c o •• •H O 4-1 03 N QJ •H T) ^4 O 03 ^4 r-l 4J O O PM cu OJ QJ O •r4 4-1 03 O 4-J C W 03 O P4 •r4 0) U 14-4 C 03 03 Prf 4-4 O PM •• PQ 03 rC -~ 4-4 03 X) 4-4 O O M 4J E O 03 03 p4 I— 4 M) 03 • 03 03 •iH -U X) Q W O 03 P4 PiJ H 4J O CJ O 03 PC) <: 03 I vJ' 03 d bO •H

PAGE 32

20 voltmeter was used to monitor the applied potential. The current flowing in the auxiliary test electrode circuit was determined from the potential drop across a standard precision resistor (± 1%) using a model SRG Sargent potentiometric recorder. All potentials are reported relative to the saturated calomel electrode. Current densities were calculated using the geometric electrode area. Spectropho tometric measurements were performed with an AMINCO DW-2 UV-VIS spectrophotometer. Spectrophotometric Measurement The amount of the inhibitor adsorbed on the electrode surface was determined by the concentration depletion method using UV spectrophotometry. The inhibitor concentration in the bulk solution of electrolyte was measured before and after the introduction of a polarized steel electrode into the solution. The number of inhibitor molecules corresponding to the solution concentration change of these two measurements is considered as the number of molecules adsorbed on the electrode surface. Due to the continuous dissolution of metal ions (Fe''~^, Ni''”'’, Cr^ ' ' from stainless steel) in electrolytic solution during polarization, the inhibitor concentration was determined from a modified dual wavelength method based on the scanned UV spectrum. The general procedures used for the determination of the adsorption of inhibitor on the electrode surface are as follows : 1. After the pretreated electrode was moved to the cell

PAGE 33

21 containing fresh deaerated 0.1 M HCl solution, the open circuit potential of the electrode was recorded after 10 minutes . 2. A fixed potential (-0.600 V, -0.400 V, -0.300 V, -0.200 V or 0.000 V, respectively) was applied to the steel sample for at least 40 minutes, using the potentiostat . ^ftier polarization, the sample electrode was moved quickly back to the pretreatment cell and was polarized there continuously at the same set potential (-0.600 V to 0.000 V, respectively). A portion (about 2.5 ml) of the electrolytic solution in the test electrode compartment was drawn into a cuvette for UV spectrum measurement and a spectrum was obtained by scanning the range of 200 to 350 nm at 2 nm s~^ . The base line correction control of the spectrophotometer was used to flatten any absorption spectrxom shown on the scan. After the solution was returned to the electrochemical cell, a proper amount of concentrated inhibitor solution (e.g., 200 111 of the 1.03 x 10"^ M DPP) was added to the electrolytic solution in the test electrode compartment (with an approximate total volume of 12 ml). Ten minutes were allowed for the added inhibitor solution to be well mixed with the electrolytic solution. The original electrolytic solution in the auxiliary 7.

PAGE 34

22 electrode compartment was withdrawn with a Teflon tip syringe and the solution containing inhibitor in the test electrode was allowed to flow freely through the coarse frit and the open stopcock into the auxiliary electrode compartment. Another 10 minutes was allowed to have the solution in both the test electrode and auxiliary electrode compartments reach steady state. 8. The solution remaining in the test electrode compartment was finally adjusted to a known solution volume (6.5 ml) the stopcock between the test and auxiliary electrode compartments closed. 9. The UV spectrum of the electrolytic solution in the test electrode compartment was measured again, as in step 4. 10. The steel electrode was returned to the cell and was polarized again at the same set potential. The spectra of the electrolytic solution were taken twice at 10 minute intervals during polarization. The inhibitor concentration changes in solution before and after the introduction of a polarized steel electrode were then determined from the spectra measured in step 9 and 11 .

PAGE 35

CHAPTER III DATA AND RESULTS Spectrophotometric Measurements by the Modified Dual-Wavelength Method The UV spectrum of 4, 7-diphenyl-l , 10-phenanthroline (DPP) is shown in Figure 5. The adsorption peak used for quantitative determination is located at 286.5 nm. The molar extinction coefficient at this wavelength was determined from an absorbance vs. concentration plot (Figure 6). The slope of the plot was calculated by the least squares method and was found to be (4.68 ± 0.01) x 10^ cm ^ M In the presence of metal ions (Fe , Cr , Ni in this experiment) the absorption spectrum of metal ions superimposed onto that of DPP thus distorting the spectrum of DPP. The degree of distortion is dependent on the concentration of the metal ions present. This matrix ion interference had to be removed or corrected for before the inhibitor concentration in solution could be determined. Correction was achieved as follows: First, the electrode was polarized at the desired potential at least 40 minutes to introduce an appreciable amount of matrix ions into the bulk solution so that any further short term polarization of the electrode would not cause an appreciable concentration change of the matrix ions. Second, 23

PAGE 36

286 . 5 nm 24 aouBqaosqv 250 nm 300 nm Wavelength Figure 5 UV Spectrum of 4 , 7-Diphenyl1 , 10-phenanthroline

PAGE 37

Absorbance 25 0.3 0.2 0.1 Figure 6 Calibration Curve (Absorbance vs. Concentration) of 4, 7-Diphenyl-l, 10-phenanthroline at 286.5 nm (o) and 319.0 nm (•)

PAGE 38

26 the UV absorption due to the matrix ions in solution was then removed electronically from the spectrum by using the base line correction control of the spectrophotometer. After this adjustment, the inhibitor added to the electrolytic solution shows its original undistorted spectrum. However, further polarization of the electrode introduces additional matrix ions into the solution and although there was already an appreciable amount of matrix ions in the bulk electrolyte, this caused changes in the matrix ion concentrations sufficient to lead to slight distortions of the inhibitor spectrum. In order to eliminate this residual error, a modified dual -wave length measurement was used. Instead of using the absorbance at the wavelength of maximijm absorption (286.5 nm in DPP) , the difference of the absorbance at the wavelength of maximum absorption and a reference wavelength (319.0 nm, arbitrarily chosen) was used. The absorbance *^iffs^6nce, AA, between 286.5 nm and 319 nm, see Figure 6, was plotted vs. concentration. Figure 7 shows such a plot. The slope of this plot was calculated to be (3.37 ± 0.01) x 10^ cm ^ M For an undistorted spectrum, the concentration corresponding to the measured absorbance difference was taken directly from the graph. If the spectrum was distorted by the change of the matrix ion concentrations , a correction procedure was used to find the possible shift of the matrix line. The correction procedure is as follows: On the undistorted inhibitor spectrum (Figure 5) an arbitrary horizontal line was drawn across the spectrum, resulting

PAGE 39

Absorbance Difference AA 27 0.3 0.2 0.1 Figure 7 Calibration Curve, Absorbance Difference vs. Concentration (AA is the Absorbance Difference at 286.5 nm and 319.0 nm)

PAGE 40

28 in three intersection points which have the same absorbance (also the same molar extinction coefficient) . The absorbance at these three wavelengths will stay at the same value as long as no spectral interference is present, but will become different if matrix ion interference appears. Thus, for a distorted spectrum, a smooth curve drawn through these three wavelength points will show the relative interference of the matrix ions to the spectrum. The curve is parallel to the true matrix line and allows to determine absorbance changes due to the matrix ion interference at any two proximate wavelengths (< 50 nm) . The procedure is illustrated in Figure 8. Here, curves a and b represent the interference (matrix line) . An arbitrary horizontal line d is drawn which intercepts curve a at wavelengths 247.0 nm, 257.8 nm and 319.0 nm. The three wavelength points on curve b are then connected by a smooth curve (curve e) . The absorbance difference at wavelength 286.5 nm and 319.0 nm determined from curve e is the residual error arising from the matrix ion interference. After correcting this value for the measured absorbance difference on curve b, the true absorbance difference for the undistorted spectr\am is then obtained. For practical purposes, the extraction of the absorbance difference AA for the pure organic spectrum can be achieved by measuring the absorbance difference between curves b and e at the wavelength of 286.5 nm and 319.0 nm, see Figure 8.

PAGE 41

cu w [> *H M Q 0 CJ 01 C X) "H h 4 0) > X U -H 0 J-l -U 4J ,Q cd T3 0)^^ 0 > oj n3 ^ H O 0 ) U 0 C 4-1 QJ U 0) td S M r— I •' ( 1 ) ;d ^ B 4 -< g •u 3 bO V-i CO 0 cd 0) 43 ^3 cu q cu 1 q pc; 1-4 *H tJ q I — I •' q o X cu Q q -H 4 :; q cu qj +J 4J > cu q q q •H qs q 44 q o •q q 'q 4d o •o pL, q S I q q o > -q q rq q hj ,0 q P rH C_) 4J I rC cq iH •50 o S ’q q q q q q q q O 4H P P •q q, a CO p -q q q Q a >. q I CO q p r-q w -T) q q
PAGE 42

30 s c o CM 3ouBqj;osqv Wavelength

PAGE 43

3i The inhibitor used in this work, DPP, is known to be a complexing agent for Fe"^. It forms a very stable red complex with iron(II), FeCDPP)^, in solutions having a pH of 2.0 to 9.0.^^ Since in the present work the base electrolyte was 0.1 M HCl, the pH of the electrolytic solution, 1.0, is lower than the complex formation range and no FeCDPP)^ complex formation between Fe"^ and DPP should be expected. In order to test if there is any interaction between Fe”^ and DPP in the solution studied two calibration titrations were run. First 40 yl and then successively 10 yl of 1.03 X 10 ^ M DPP solution were added to 2.5 ml of a solu-3 -Htion 2.10 X 10 M in Fe . The plot of the absorbance at 286.5 nm vs. DPP concentration (Figure 9) showed a straight line with a slope of (4.76 ± 0.01) x lo'^ cm"^ which is very close to that of the calibration curve in Figure 6. Similarly, successively 20 yl of 1.05 x 10"^ M Fe"'”'’ solution were added to 2.5 ml of a 5.579 x 10“^ M DPP solution. Two horizontal lines were obtained for the absorbances at 286.5 nm and 319.0 nm vs. Fe"^ concentration (Figure 10). This confirms that there is no chemical interaction between I I Fe and DPP in 0.1 M HCl solution in the UV spectral range studied . Adsorption of DPP on Stainless Steel Electrodes The amount of inhibitor adsorbed on the electrode surface was calculated from the concentration change resulting

PAGE 44

.20 32 o t— I C o •H 4 -) Ctl 4J c OJ a C o u aouBqjLOsqv Figure 9 Absorbance vs. Concentration of 4 , 7-Diphenyl-l , 10-phenanthroline in the Presence of 2.10 x 10“-^ Fe"*~^

PAGE 45

33 o X 2 a o •H J-J cd U 4-1 C 0) o c o o I X. aouBqjiosqv Figure 10 Absorbance vs. Concentration of Fe in the Presence of 6.58 Diphenyl-1 , 10-phenanthroline , at 286.5 nm (o) and 319.0 nm (

PAGE 46

from introduction of the polarized electrode into the solution. The total mmiber of molecules adsorbed on the electrode surface was obtained by multiplying the change in concentration of the bulk solution with the volume of the bulk solution and Avogadro's number. The calculated number was then divided by the total surface area of the electrode (a roughness factor of 4.0 was used to convert the geometric area into an estimated true area) to obtain the number of molecules adsorbed per unit surface area. Figure 11 shows plots of number of molecules adsorbed per unit surface area vs. inhibitor concentration at various fixed applied potentials . To estimate the effective area covered by the inhibitor molecule, Stuart and Briegleb atom models were used to construct the DPP molecule. The molecular model was arranged in a position corresponding to a possible configuration of the adsorbed molecule and then photographed. In arranging the models, the nitrogen atom was placed in a position corresponding to the formation of a metal-nitrogen bond. Two possible configurations were assumed, an extended configuration with the molecule lying flat on the surface, covering maximum area, and a compacted configuration with the molecule standing perpendicular to the metal surface, covering minimum area. Figure 12 shows three possible configurations. Pictures a and b show the extended configuration, the molecules lie flat on the surface, whereas picture c shows the compacted configuration, the molecule stands perpendicular to the

PAGE 47

m I tTJ O -H 1 — I 4-) c 1 — I (U I JJ r-l O >.P-I c P <} o •r4 C J-I O c\5 > 0) M-U cfl cd U o O -H U 4-J cd 0) >-l o -u cd p m Q> cn o u O 0) Pi QJ -H O r-4 P O CU P T3 Xi P -U 0) C PU
PAGE 48

36 X aoBjjins uo paqaospv sapnoapoi,i jo // Concentration in Solution,

PAGE 49

(b) (c) Figure 12 Three Possible Configurations of 4 , 7-Diphenyl1 , 10-phenanthroline . (Molecule built from Stuart and Briegleb Atom Models)

PAGE 50

33 metal surface. Due to the bulky steric effect of the two phenyl groups at the 4,7 position on DPP, these phenyl groups are not coplanar with the other parts of the molecule (the planar phenanthroline molecule) and thus DPP cannot properly lie flat on the metal surface. In picture a, the two phenyl rings stay in a position which is perpendicular to the phenanthroline plane, have less repulsion between atoms and the molecule probably will be more stable, but the contact with the metal surface seems quite weak. In picture b, the two phenyl groups are rotated to a position that is neither coplanar with nor perpendicular to the phenanthroline group. The intramolecular repulsion may be larger but the molecule has better contact with the metal surface. In picture c, the two nitrogen atoms rest on the metal surface and a strong nitrogen metal bond is expected to be formed. The two phenyl groups are located on top of the DPP molecule and do not take part directly in the adsorption process. The effective areas covered by the three molecular configurations were calculated from the projected areas in Figure 12. A carbon carbon double bond length of 1.395 ± 0.003 R in aromatic rings was used as a scale factor to estimate the effective area.^^ The effective area calculated for configurations in picture a, b and c are 132.4 153.0 and 87.5 respectively. Because of the complexing character of DPP, the configuration of picture c seems to be the most probable arrangement for the inhibitor molecule adsorbed on the metal surface. Thus the estimated effective

PAGE 51

J9 2 area of 87.5 8 was used for the surface coverage calculation. The plot of fractional surface coverage vs. inhibitor concentration at fixed applied potentials (-0.600 V, -0.400 V, -0.300 V, -0.200 V and 0.000 V) is shown in Figure 13. From this figure it is possible to plot the fractional surface coverage vs. applied potential at constant inhibitor concentration in the bulk solution. Figure 14 shows such a plot for seven different inhibitor concentrations, 1.0 x lO”'^ M to 5.0 x 10”^ M. Potentiostatic Polarization In order to maintain the inhibitor concentration at a constant value during an experiment, a different electrochemical cell with a larger test electrode compartment was used in the polarization experiments. The volume of solution used in this study was about 100 ml, so that the total number of moles of inhibitor adsorbed on the electrode surface (< 10~^ mole, depleted from bulk solution) becomes negligible (< 1%) when compared with the total number of moles of inhibitor in the bulk solution (> 1.0 x 10"^ mole). Solutions with 1.02 x 10"^ M, 3.68 X 10 ^ M, 1.28 X 10“^ M and 1 . 15 x 10'^ M DPP in 0.1 M HCl supporting electrolyte were used in the polarization experiments . After the twenty minutes prepolarization period at -0.900 V, net currents were cathodic with values of (8.0 ± _ 3 _ 2 0.5) X 10 A cm . The open circuit potential. E of oc ’ the prepolarized electrode in fresh deaerated bulk solutions

PAGE 52

CO I I-H 1 — I cc3 kN'H C o d x: a) (X4-J •H o p p T) -> 0 ) •H C h o CO d o •rH cu bO d ct) (\3 M > ( 1 ) > O u cu CJ cx) 4 ^ }-l d cn cd d o u u cd P o •H cd d 4 J d cu o d o u o cd d fit O d M-i x: o -u d cu cd o d d
PAGE 53

.0C441 0 a§BaaA 03 •[BuoT:joBj:j Concentration in Solution,

PAGE 54

Fractional Surface Coverage 42 Figure 14 Fractional Surface Coverage 0 vs. Applied Potential at Various Constant 4, 7-Diphenyl1 , 10-phenanthroline Concentrations

PAGE 55

43 was measured after 10 minutes. The recorded E^^'s are shown in Table I. Figure 15 shows the polarization behavior of the pretreated steel electrode in 0.1 M HCl electrolyte and in 0.1 M HCl electrolyte plus various DPP concentrations. The polarization started from -0.700 V and the potentials were shifted anodically to a potential of +0.200 V. As the potential was shifted anodically, the measured current decreased and became zero at the rest potential, E^. The rest potential is defined as that potential at which the absolute values of the internal anodic and cathodic currents become equal, resulting in a net external current flow of zero. Values of the rest potential determined here are -0.372 to 0.345 V. Table I also shows the E^ ' s of the steel electrode in different DPP con centrations . The and E^ measured agree very well, their discrepancies are generally within 10 mV which may be attributed to the change of surface state of the electrode after polarization has started. The cathodic Tafel slopes, i.e. a plot of 6E/61og|i|, of the polarization curves are plotted in Figure 16. The corrosion current of the steel electrode at different DPP concentrations was determined from the extrapolated cathodic Tafel line at the rest (corrosion) potential. Table II shows the Tafel slopes and corrosion currents in solutions of various DPP concentrations. Eq l^^lz3-t ion Behavior of the Steel Electrode at Various Constant Fractional Surface Coverages 0 of DPP From Figure 14, it is clear that the fractional surface coverage 0 of DPP on the steel electrode is not only a function

PAGE 56

Table I. Open Circuit Potential, and Rest Potential, E , in Solutions of Various DPP Concentrations ^ 44 -4O X M crv LO m CO CO • . • o o i—i 1 1 m o rH X M m LO VO in 00 CO CO U > O PM o PS PM C o U P -H w w

PAGE 57

45 0 )

PAGE 58

uio 46 Applied Potential, V (vs. S.C.E,) Figure 16 Plot of Cathodic Tafel Slopes from the Potentiostatic Polarization Curves.

PAGE 59

Table II. Tafel Slope (Cathodic) and Corrosion Current of the Steel Electrode in Solutions with Various DPP Concentrations (Determined from Figure 16) 47 o I-) X M o\ oa • CTv . rH I — 1 CO m o r~l X M 00 CM Ip LO • . I — 1 VO vD 1 O ! — 1 X M oo >X) o <100 . CO IP o rH X M (N CM O CN CM O i—i rp 1 — 1 VD 0 M VjD . 1 — 1 CM 1 — 1 1 — 1 a o •H 4-J C P P O 1 — 1 a •H O cu ^ •p ca }-4 CM cfl Ml 1 V-i 1 — 1 1 — 1 P 0 P CJ (U 1 U O a K PM Ml O PM a cC . B O O P3 -H H O -H

PAGE 60

48 of the inhibitor concentration but also a function of applied potential. Thus, the polarization behavior of the steel electrode presented in Figure 15 only shows the currentpotential relationship at constant bulk concentration. The actual coverage varies from point to point (potential to potential) in the polarization curve. Since the fractional surface coverages of DPP on the steel electrode at various concentrations and at various applied potentials are known (Figure 13), it is possible to construct the polarization curves at constant surface coverage. A more detailed surface coverage vs. concentration plot is shown in Figure 17, where the curves representing the surface coverage vs. concentration relationship at -0.700 V, -0.500 V, -0.350 V and -0.320 V were extrapolated or interpolated from Figure 14. Figure 18 shows a plot of the logarithm of the current density as a function of DPP conccntration at various applied potentials. From Figure 17, five straight lines representing a constant fractional surface coverage of 0.1, 0.2, 0.3, 0.4 and 0.5, respectively, were drawn across the curves. The intercept obtained shows the concentration of the bulk solution needed for the specified fractional surface coverage at that (specified) potential. After interpolating the needed DPP concentrations (at the specified applied potential) from Figure 18, the polarization current density at the specified fractional surface coverage and applied potential was obtained. Figure 19

PAGE 61

c O > <1U rH o QJ O CU C m k •H • d H O bO O 1 •H rC p-l 4J > 0 C o CO O >4 C O QJ 42 • 'd OhO CU 1 1 4-1 o CO r-H }-l > o o tH 4-1 Ph 1 >4 i—t (0 CU >^4J J-1 C CO p CU Q M 42 a u •H o Q • 1 CO 'a i-H 4-J cO O p O Ph p ><{ CU 'd w bO CU CO CU }4 -H p CU Px^ CO > o cn > o ;3 O O CU •H CM u P cn CO CO • CJ4 > O U f P p CO p d P 1 — 1 P P CO o P •H > 0 P •r4 P O P P m o p cn p P • p p o fin O 1 I OJ bO •H

PAGE 62

50 o o O o o v£) O O r-i OO X o CN 0 aSBaaAOQ aoBjang Concentration in Solution,

PAGE 63

cu C •H o c cij C cu Ph rH CO I — I rH CCJ I *H iH 4J P cu CU -u ^ o P,PL| •H O T) I CU X *• *H < 1 cn > u P cn P O •H p p cu > p p p p CD p o •H P P O *H V •U P P p N 4-1 •P P P P P O —I P o o CP CD CO p p p M •H

PAGE 64

.700 52 CO c-j" (^_uio yrl) T Sox Concentration ,

PAGE 65

53 shows such a plot for constant fractional surface coverages of 0.0, 0.1, 0.2, 0.3, 0.4 and 0.5, respectively. The major difference between Figure 19 and Figure 15 is that the cathodic Tafel slopes of all the polarization curves in Figure 19 become identical even though their surface coverages are different. The polarization curves shift downward regularly with increase of fractional surface coverage. The inhibitor effect on both the cathodic and the anodic part of the curves is clearly evident. The plots of anodic polarization curves in Figure 19 were stopped at -0.300 V because after the steel electrode has become passivated, the reproducibility of the measured polarization current is not sufficient to allow for the correction of the surface coverage effect on the polarization curve.

PAGE 66

uio yrt) 54 3.0 2.0 CN I 1.0 M O 0.0 •0.7 -0.6 -0,5 -0.4 Applied Potential, V (vs. S.C.E.) -0.3 Figure 19 Constructed Potentiostatic Polarization Curves at Various Constant Fractional Surface Coverage 0

PAGE 67

CHAPTER IV DISCUSSION Selection of Metal Adsorbent and Adsorbate The selection of stainless steel as a solid adsorbent was based on its versatility as a construction material and mainly on its corrosion resistance properties. It has a low dissolution rate in acidic solution and is therefore expected to give less interference in the spectrophotometric measurements . More critical for the success of the contemplated experiments is the proper choice of an organic adsorbate as inhibitor. Basically, the organic adsorbate should fit the following requirements in order to assure that acceptable results could be obtained. First, the organic material used should be adsorbed on the solid adsorbent and function as an acceptable corrosion inhibitor. Second, the organic adsorbate used as an inhibitor should have a very strong absorption band in the UV absorption range in order to facilitate measurements at very low concentrations (< 10"^ M ) . For this purpose, the extinction coefficient of the absorption peak should be larger than 1.0 x lO'^ cm"^ M"^. This requirement is predicated by four factors, the total surface area of the solid electrode, the volume of the solution, the 55

PAGE 68

56 sensitivity of the spectrophotometer used, and the space requirements of the adsorbed molecule on the electrode surface. The larger the surface area of an electrode, the more organic adsorbate will be adsorbed and thus the greater the change in solution concentration. Similarly, the smaller the solution volume the larger the change in solution concsntration for a finite amount of molecules adsorbed on the electrode surface (amount of organic inhibitor depleted from the bulk solution) . The relationships between these factors can be demonstrated by looking at the following calculations. Assxime a system, similar to the one designed for the present work, where the electrode has a surface area of 5.0 cm^, the volume of the solution used is 6.5 ml and the organic adsorbate molecule has a projected area of 50 on the electrode surface when it is adsorbed. Then it takes 1.0 x 10^^ molecules to form a monolayer coverage on the solid electrode. If about 107o of the electrode surface is covered (0 = 0.1), the total number of molecules depleted from the bulk solution is 1.0 X 10^^. For a reasonably precise spectrophotometric measurement, a l^ variation in the original concentration (proportional to the optical signal) should be detectable. Thus, a bulk solution with an original total number of 1.0 X 10^^ organic adsorbate molecules should be used, i.e., a concentration of about 2.5 x 10“^ M. From Beer's Law, A = abc, where A is absorbance, a is the molar extinction coefficient, b is the light path and c is the concentration of the solution. With a = 1.0 x 10^ cm"^ M"^, b = 1 cm, c =

PAGE 69

57 _ 6 _ 9 2 . 5 X 10 M, A is equal to 2 . 5 x 10 . Thus, in order to detect the desired 17o concentration variation, a rather sensitive spectrophotometer is needed. The Aminco DW-2 spectrophotometer which has a scale (full scale) down to 0.005 absorbance units fits this requirement. From the above example, it is clear that the larger the molar extinction coefficient of the organic adsorbate, the higher the solution concentration allowable and the smaller the measurement errors . The third factor which needs to be considered in the selection of the organic adsorbate is the solubility of the organic molecule in acidic aqueous solution. There is a large number of organic molecules which are potentially good corrosion inhibitors and also have a strong absorption band in the UV absorption region. Most of them, however, are insoluble in aqueous solutions. Thus, the selection is narrow and restricted. The organic compound chosen here as a corrosion inhibitor for this work, 4, 7diphenyl1, 10-phenanthroline fits the requirements mentioned above reasonably well. Precision of the Spec trophotome trie Measurement and Surface Coverage Determination As mentioned previously, the dissolution of metal ions in the solutions studied causes serious interference to the spectrum of the organic inhibitor and thus affects the precision of the inhibitor concentration determination. The different metal ions (Fe^, Cr * ' ' , Ni"*~^) have different

PAGE 70

58 contributions (or interference) to the overall spectrum. During the polarization of the electrode, more and more ions are dissolving in solution and the interference to the measured spectrum becomes larger and larger. However, at a constant polarization potential, the concentration of metal ions should be proportional to the polarization time (assuming the surface area of the electrode is invariant during polarization). The interference is then a systematic error, and may be eliminated on the basis of a calibration curve. In fact, the proportionality of the spectrum change with the concentration of metal ions dissolved in solution (during polarization) has been used as a method for the quantitative determination of the corrosion rate of metals or alloys. For example, Foley and Alexander used the UV and Visible spectra C Q to monitor the rate of iron corrosion. They also contributed to the knowledge of the iron dissolution process through a spectrophotometric study of the various iron species and the corrosion products that appeared in different solutions after the corrosion reaction has proceeded for a 59 finite time. Because of the existence of multiple elements in the steel sample and because the dissolution rate of each metal component may vary with the applied polarization potential, the spectral interference caused by the presence of matrix ions will vary from potential to potential and the calibrations become very tedious work. The proposed modified dual -wavelength method supplies an alternative solution to this problem. From this method, it is possible to build a

PAGE 71

59 simulated matrix line (interference spectrum due to the presence of Fe"^, Cr"^, Ni"^ and Fe"^) which differs from the true matrix line only at the horizontal location (absolute absorbance) . The accuracy of the inhibitor concentration measurement depends upon how exactly the matrix line is reinstalled. Here, only three wavelength points were used for the installation of the matrix line. The simulated matrix line built from these three points should agree reasonably well with the true matrix line if no special absorption peak on the matrix line overlaps with the absorption peak of the organic inhibitor (used for quantitative measurement) . That is the case here where the matrix line of the metal ions generally appears as a decay curve from lower wavelength to higher wavelength and no absorption peak was found in the scanned UV range of 250 nm to 350 nm. The prerequisite for the correction of systematic errors and the usage of the spectra for the determination of the corrosion rate is that the solution must be completely free of oxygen. If there is any trace amount of oxygen present, part of the ferrous ion will be oxidized to ferric ion. Ferrous ion and ferric ion have completely different UV spectra. According to Effrenfruend and Leibengoth, Fe* ' ' has an absorption maximum at 304 nm with a molar extinction coefficient of a = 3300, whereas, at the same wavelength, Fe”^ has a molar extinction coefficient of about a = 10.^^ The drastic difference between these two species in their UV spectra would cause a potential additional error in the inhibitor

PAGE 72

60 concentration determination. To avoid the possibility of solutions used here to be contaminated by ambient air during transfer to and from the cuvette, a special cap was designed for the cuvette and a stream of helium was applied to the cuvette before and after the transfer of the solution. Even if the solution was contaminated slightly with ambient air and the matrix line was not a smooth curve, it would still be possible to reinstall the true matrix line with reasonable success. The reinstallation procedure would become more tedious, i.e., instead of just one horizontal line on the spectrum in Figure 5, two or more horizontal lines would be drawn. Those interception wavelength points on the same line represent the group of wavelength points which have the same absorbance and the same molar extinction coefficient. Checking the wavelength points of the same group on the distorted spectrum, the relative difference of absorbance at those wavelengths allows to deduce a more exact true matrix line and thus to determine a more accurate inhibitor concentration. An optimal estimation of the accuracy for the concentration determination using the modified dual -wave length method is about ± 57o. The accuracy of the surface coverage measurement may be very poor, however, because it involves not only the errors that may come from solution transfer and from determination of the solution volume, but also from the estimation of the true surface area of the electrode which is difficult to determine. Thus the accuracy obtained is

PAGE 73

61 probably not better than ± 20%. Even though a constant roughness factor has been used for the conversion of the true surface area from the geometric surface area, deviations of the true surface area still exist from experiment to experiment and this error passes into the surface coverage measurement. The deviation essentially comes from two sources. First, it is impossible to produce an exactly reproducible surface state on the electrode. Even though the same surface treatment procedures were followed, it is still not to be expected that the surface states obtained are the same. Secondly, the surface of the treated electrode roughens during polarization. The standard reduction oxidation potentials of Fe, Cr and Ni (the main components in the 304 stainless steel sample) are -0.707, -1.007 and -0.517 V, respectively, in aqueous solution at 25°C. In the surface coverage measurement experiments, the steel electrodes had been polarized at -0.600, -0.400, -0.300, -0.200 and 0.000 V, respectively, for 50 minutes before the spectrophotometric measurements were taken. The applied polarization potentials are all positive to the standard redox potentials of Fe, Cr and Ni listed above (except at -0.600 V which is negative to the standard redox potential of Ni) . Thus an appreciable amount of iron, chromium and nickel should have dissolved into the solution and the electrode surface might have become quite rough. The degree of roughness might depend upon the applied potential and the length of the polarization period. The more anodic the applied potential is, the larger the

PAGE 74

62 metal dissolution rate will be and the sooner the surface will become rough. As long as the polarization is still going on, more and more metal ions will be peeled off from the electrode surface as time elapses and thus the rougher the surface may become. But after a certain period of polarization the surface may have reached a steady state and the surface area of the electrode will then just fluctuate within a certain range. In this experiment, 50 minutes were allowed for the electrode to be polarized and a steady state was assumed to have been reached. Since the peeling of metal ions from the electrode surface may produce various surface irregularities (e.g., kinks, terraces, cracks and projections) the true area may be considerably greater than the geometric area and the roughness factor may be quite large. A roughness factor of 4.0 was used in this experiment and seems reasonable and justifiable. The Dependence of Surface Coverage on Inhibitor Concentration and Applied Polarization Potential The dependence of fractional surface coverage on inhibitor concentration and applied potential is shown in Figures 13 and 14. At constant applied potential the surface coverage of DPP increases with increasing DPP concentration in bulk solution (except at the applied potential of -0.200 V and 0.000 V) , but levels off at high bulk concentrations, c > 10”^ M for potentials more negative than -0.400 V. At constant concentration, generally, the adsorption of DPP decreases with the

PAGE 75

shift of the applied potential in the anodic direction. The drastic decrease of the adsorption found at -0.300 V in Figure 14 can be attributed to the change of the surface state from active to passive. Stainless steel starts to become passivated at around -0.300 V. The adsorption levels off in the potential range -0.200 V to 0.000 V, where complete passivation exists. In the passive state, oxides are present on the metal surface, which intensify the hydrophilic properties of the surface' Thus the adsorbability of organic adsorbates decreases sharply and becomes constant in the passive potential range. Determination of the Adsorption Isotherm When there is equilibrium between a species in the adsorbed state and in the bulk of the solution at an ideally polarized electrode, the corresponding electrochemical potentials are equal, i.e., y* + RT ln{f(r)} y* RT In a = 0 (1) where the y^ and y* are the standard electrochemical potentials in the adsorbed state and in solution, respectively, the quantity f(F), a function of the surface concentration F, represents the activity of the adsorbate and a is the corresponding activity in the bulk of the solution. The standard free energy of adsorption, ~KG° , is equal to y^ y*. Thus, equation (1) can be rewritten as

PAGE 76

64 ^ = RT In a RT In {f(r)} f(r) = a exp (-AG®/RT) r = r (Ka) (2) where K represents the equilibrium constant of adsorption and is equal to exp (-A^/RT). Equation (2) shows that the isotherm is some function r(Ka) of the product of the bulk activity a and the quantity K which depends on the electrical parameters such as charge and potential characterizing conditions at the ideally polarized electrode. The explicit form of equation (2) depends on particle particle and particle electrode interactions. Equations (1) and (2) have been used as the starting point to derive different isotherms. Several common adsorption isotherms which have been used to represent adsorption on electrodes are listed below: 1. The Henry isotherm 0 = Ka (3) where K and a are as defined above and 0 is the fractional surface coverage of the electrode. 2. The Freundlich isotherm 0 = Ka’^ 3 . where 0 < n < 1.^^ The Langmuir isotherm 0 1 0 (4) Ka (5)

PAGE 77

65 4. The Frumkin isotherm Q exp (-2b0) = Ka (6) where b is a quantity characterizing interactions befi ft tween the adsorbed particles . 5. The Virial Coefficients isotherm 0 exp (-2b0) = Ka (7) where b is always less than zero and represents repulfi 7 sive interaction between adsorbed molecules. ' 6. The Volmer isotherm^^ r0 exp = Ka (8) f\ 7 7. The Amagat isotherm r~-0 exp Cr~ "0 ^ " 8. The Helfand-Frisch-Lebowitz isotherm^^ exp Q ) = Ka (10) ^ ^ (1 0)^ 9. The Hill-de Boer isotherm^^ ’ ]_ ? Q exp (]— t-q) exp (-2b0) = Ka (11) 72 10. The Parsons isotherm Y ~ J o exp (— ^ ^) exp (-2b0) = ^ ^ (1 0)^ Ka ( 12 )

PAGE 78

66 11. The Temkin isotherm^^ (13) and 12. The Blomgren-Bockris isotherm 1 ^ Q exp (p0^ q 0 ^) Ka (14) where p and q are constants, expressed in terms of the dipole moment, the area occupied by the adsorbed molecule and the dielectric constant of the surface layer. The simplest isotherm is Henry's isotherm which follows directly from the equation of state which treats the adsorbate as a two-dimensional ideal gas and assumes that the heat of adsorption is directly proportional to the activity a, in the bulk of the solution. The Freundlich isotherm was derived from the assumption that the heat of adsorption falls logarithmically with coverage. The Langmuir isotherm is based on the equation of state for noninteracting particles adsorbed on fixed sites. It can be derived simply by the kinetic approach or thermodynamically by applying the law of mass action to an equilibrium process. Because the Langmuir isotherm neglects the effect of mutual repulsions between the adsorbed species, it is rarely ever applicable to real systems. The Frumkin isotherm was derived from the Langmuir isotherm by assuming that the apparent standard free energy of adsorption is a linear function of coverage, i.e..

PAGE 79

67 ^ = ^ 0=0 ( 15 ) By substituting equation (15) into equation (5) the Fr\imkin isotherm was then obtained. The empirical term exp (-2b0) which was first introduced by Frumkin in the equation of state corresponding to the Langmuir isotherm accounts for the particle particle interactions between adsorbed species. The parameter b is positive if attraction exists between adsorbed particles and is negative if there is repulsion between adsorbed particles. The same modification can be applied to derive the Virial isotherm which takes the form of a Henry isotherm with the factor K defined by equation (15) . A similar derivation can also be made for the Volmer, Amagat and Helf and-FrischLebowitz isotherms, but the dependence of K on coverage is more involved than that expressed by equation (15) . The Volmer equation treats adsorption as a two-dimensional fluid of rigid particles and so does the Helfand-Frisch-Lebowitz isotherm. It follows from equation (15). that analysis of isotherms as a deviation from Henry's Law or a Langmuir isotherm should lead to a linear variation of the standard free energy of adsorption with coverage provided that either the Virial or Frumkin isotherm can be applied. Some results given by Blomgren, Bockris and Jesch show that this correlation is fairly well obeyed, at least when the coverage is not 7 6 too high. Parsons and Hill-de Boer further modified the Helfand-Frisch-Lebowitz and Volmer isotherms respectively by introducing the empirical term exp (-2bG) to account for

PAGE 80

68 particle particle interaction in the same way as in the Frumkin and Virial Isotherms. Based on the assumption that the variation of the standard free energy of adsorption is a linear function of surface coverage, the Temkin isotherm was derived by further assuming that the surface is made up of a great number of small patches, ds, on each of which the Langmuir isotherm is applicable, but with the standard free energy of adsorption increasing in small steps. Equation (13) can be derived by integrating the equation d0 K(s) a 1 + K(s) a ds where K(s) is the equilibrium constant for adsorption over the whole area, for s going from zero to unity. Expressions for the particle particle interaction other than the term exp (-2b0) in the isotherm can be derived by transposition of the corresponding treatment for metal gas adsorption. Blomgren and Bockris started with the Langmuir isotherm and corrected the standard free energy of adsorption, in the case of ion adsorption, for coulombic interaction and dispersive forces, introducing the following K factor into the Langmuir isotherm^^ ^ " ^0=0 q0^)> They obtained the equation for the adsorption of organic ions as

PAGE 81

69 X ' l Q exp (p0^ q0^) = Ka (14) Similarly, Conway and Barradas wrote the Langmuir isotherm with the following factor for dipole dipole interaction in the adsorption of an uncharged substance^^ K = Kq^q exp {-(p'0^/^ q0^)} The isotherm obtained differs from that of Blomgren and Bockris by having the p0^ term replaced by p'0^^^t Y~—Q exp (p'0-'/^ q0-^) = Ka (16) In equations (14) and (16) the term 0^ and 0^^^ will be predominant over the 0 term if 0 is not too close to unity. It is known from experimental data that the adsorption isotherms of organic substances may be either sigmoid (Sshaped) or logarithmic, dependent on whether attractive or repulsive interaction predominates between the adsorbed particles, If attractive interaction predominates and the isotherm is Sshaped then the isotherm is characterized by the following feature. The slope in the middle region, when 0 = 0*, must be greater than the initial slope at 0 = 0 . In other words, the following condition must be satisfied in this case : f ’ (0*) ~FW > 1 (17) where f ' (0*) and f'(0) represent the value of 60/5a at 0 = 0*

PAGE 82

7U 0=0. Analysis of the isotherms shows that only equations (6) , (11) , (12) and (14) satisfy the condition (17) . The others are therefore not applicable to experimental S-shaped adsorption isotherms. In isotherms (7) to (13), one arbitrary parameter b is used to represent two-dimensional interaction between the adsorbed particles while in the Blomgren-Bockris isotherm (isotherm (14)) the two-dimensional interaction is represented with the aid of two arbitrary parameters, p and q. The change of the shape of isotherm with parameter b in equations (6), (11) and (12) is shown in Figure 20, where the dependence of 0 on the relative concentration a/aQ_Q^ is calculated from equation (6) for different values of b.^^ In Figure 20, the isotherm changes from a logarithmic form for b < 0 (curve 1, repulsive interaction exists between adsorbed particles) to a sigmoid shape (curves 3 and 4, characterizing attractive interaction) when b > 0. To express the adsorption behavior of DPP on a stainless steel electrode at different polarization potentials with the proper isotherm curve fitting procedures are required in order to find the parameters for the best fit of the experimental data. It is obvious from Figure 13, a plot of surface coverage 0 vs. concentration, that the adsorption behavior of DPP on a stainless steel electrode cannot be expressed by Henry's Law. Figure 21 shows the plot of the Langmuir isotherm for four different polarization potentials, -0.600 V, -0.400 V, -0.300 V and -0.200 V, respectively. The figure indicates

PAGE 83

71 C • o ^ •H -l (U 4-1 CO O C (U C QJ ID 0 ) O .-I }-i C rt < 1 ) 0 > 4-1 U QJ 4-1 pc: 0 ) p > 0) OJ •H p CU 4-1 QJ cn nj 4-4 rH 4-1 cr 0 w ^ u B 01 o ^ o P n P 4-4 4-1 P X) O P 01 W 4J II P P .-H ^ rP P 4-1 O 4-1 P Cs| O U P ~r 4 O -!< I P CD P II Xl II P rO P CD P^ P Q) O c\ 3 1 — I I O
PAGE 84

0/(1 72 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 Concentration, a x 10 Figure 21 Data Plotted According to the Langmuir Isotherm

PAGE 85

7J that the adsorption does not follow the Langmuir type. Due to the similar character of the Langmuir, the Volmer and the Helfand-Frisch-Lebowitz isotherms, fitting of the experimental data to these isotherms is also ruled out. Figure 22 shows the logarithmic plot of 0 vs . concentration. A set of straight lines will be obtained and the slopes will characterize the Freundlich parameter n, if the adsorption behavior fits the Freundlich isotherm. The definite curvature of the plots indicates that the adsorption behavior does not obey the Freundlich equation. By adjusting the parameter b in the empirical term exp (-2b0) in equations (6), (7), (11) and (12), it is possible to fit the experimental data to the represented Frumkin, Virial Coefficient, Hill-de Boer and Parsons isotherms to a certain degree. The determination of the parameter b for the different isotherms was achieved by trial and error. Figures 23, 24, 25 and 26 show these plots. For the more complicated isotherms, such as equations (13), (14) and (16), the curve fitting process was achieved with the help of computer calculations. Figures 27 and 28 show the plot of equations (14) and (16), respectively. It is clear from these figures that the isotherms fit the experimental data very well only in the low surface coverage range (0 < 0.4), while in the higher surface coverage range (0 > 0.5) curvatures are inevitably obtained. Table III lists the fitted particle particle interaction parameter b and the calculated equilibrium constant of adsorption, K, of different isotherms. The adsorption

PAGE 86

.600 V 74 0 So-[ Figure 22 Logarithmic Plot of the Freundlich Isotherm

PAGE 87

exp (-2b0) 75 Figure 23 Data Plotted According to the Frumkin Isotherm

PAGE 88

exp (-2b0) 76 Figure 24 Data Plotted According to the Virial Coefficient Isotherm

PAGE 89

( r _ q ) exp (-2b0) 77 25 ~ Data Plotted According to the Hill-de Boer Isotherm

PAGE 90

exp { y 2b0} 78 Figure 26 Data Plotted According to the Parsons Isotherm

PAGE 91

exp (p0 79 Figure 27 Data Plotted According to the Blomgren-Bockris Isotherm

PAGE 92

exp (p ' 0 80 CO o cr I CN OO CD I Figure 28 Data Plotted According to the ConwayBarradas Isotherm

PAGE 93

81 constants, K, were obtained from the slopes of the plots. The least squares method was used to find the values of the slopes. From Table III it is seen that the equilibrium constants of adsorption obtained from the different isotherms agree quite well at the same applied polarization potential while the interaction parameters deviate from each other. There are two thoughts concerning the dependence of the interaction parameter b on potential. Parsons, in agreement with Fr^Imkin's original suggestion, considers that b is independent of the potential or of the charge on the electrode, and that the effect of the electrical field at the interface appears solely in term K. Damaskin retains the dependence of K on potential but also assumes that b is a function of potential. Table III shows that the parameter b is the same at -0.600 V as at -0.400 V but differs at -0.300 V and -0.200 V. Since the stainless steel electrode is partially passivated at -0.300 V and totally passivated at -0.200 V, the change in surface state may well account for the change in the b parameter at those potentials. A general decrease of parameter b from -0.400 V to -0.300 V and -0.200 V agrees also well with the hydrophilic property of the oxides developed on the electrode surface. Considering that the differences in interaction parameter b at -0.300 V and -0.200 V are due to the change of the state of the surface and that the parameter b is the same at -0.600 V and -0.400 V, it is concluded that b is independent of the applied potential in the system stainless steel/DPP. This would support Parsons^ arguments .

PAGE 94

82 bO C •iH 4-1 T) -t-* C •rH Cti Pi-( ^4 O 4J M-l c cd T) 4-) (1) m w C D O u cr C > o » *rH CX4J a 'O M C o tC CO X) cr< -X) eu Pi 1 |i| o
PAGE 95

Table III--Extended 83 t— 1 t-g rH i-g o o o o o o o o +1 +1 +1 +1 o o VO eg VO i-g C3V o o CXD o o rH vO ro o 1 — 1 VO ' 1 g 0 o O rH o pq pq cr* X X) C nJ nJ b C o I
PAGE 96

84 With the exception of that of the Virial Coefficient isotherm, the interaction parameters obtained here are all positive. This indicates that an attraction exists between the adsorbed species. Generally, an attractive interaction exists in the case of competitive adsorption. Due to the fact that specific adsorption of Cl~ may be present in the system studied, and that the DPP will first adsorb mainly in the form of DPPH"*", the coulombic interaction between the adsorbed Cl and DPPH"*" may enhance the adsorption of DPP on the polarized stainless steel electrode surface. The standard free energy of adsorption AG° which relates to the adsorption constant K by -RT In K, was also calculated for the different isotherms and is listed in Table III. The standard free energy of adsorption calculated at the same potential agrees very well between the different isotherms. The values obtained here are comparable to those of other 82 organic adsorbated reported in the literature. Adsorption and the Structure of the Electrical Double Layer The "charge" on an electrode signifies the presence near the surface of unequal amounts of mobile electrons and their accompanying positive ions; the counter charge consists of an excess of ions of one type of charge in the adjacent solution. The analogy of such a system to a parallel-plate condenser, as utilized by Helmholtz, forms much of the basis of OO present double layer theory. The potential of zero charge

PAGE 97

85 (p.z.c.) of an electrode with respect to some reference electrode thus serves as a natural reference potential for the double layer. But unlike the parallel-plate condenser, the potential difference across a double layer is not zero at the p.z.c., owing to the dipole contribution to the surface potential x X in each separate phase, electrode and electrolyte. Therefore, a simple potential measurement does not indicate the charge on an electrode. Changes in the electrolyte may result in variations in surface potential x^ and x^°^^ and thus in the p.z.c. To correlate the changes in p.z.c. with solution variation and also to understand its effect on adsorption, it is useful to consider the structure of the electrical double layer. Since reviews are available, only the structural components of the interface and their contributions to the metal solution potential difference ft 9 ^h. A(J) will be mentioned here. ’ (i) No specific adsorption of solute: Consider a mercury electrode in an aqueous NaF solution. Over a wide potential range (including the p.z.c.) both ions remain fully hydrated and hence separated from the electrode by at least one water molecule. The most significant interaction between ions in the solution and the electrode surface follows from the theory proposed by Gouy and Chapman, the distance of the closest approach of ions being controlled O C QC by the size of the hydrated ions. ’ Figure 29(a) qualitatively depicts the distribution of net charge and the poo~t tential due to this charge at a positively charged electrode.

PAGE 98

86 At the p.z.c. (Figure 29(b)) there is no net electrode charge and hence only random distribution of ions. Therefore, there is no metal solution potential difference due to ions.^^ However, an orientation of solvent dipoles undoubtedly still exists, and this varies with electrode potential . (ii) Specific adsorption of solute: Many ions have the tendency to shed their primary hydration shell on the side facing the electrode and to be adsorbed in closer proximity to the electrode than hydrated ions are. Overcoming the forces of hydration in solution and causing such ''specific'' adsorption are electrode ion interactions due to image, dispersion and covalent forces. The nature of specific adsorption thus allov7s for an amount of adsorbed charge different than the amount of opposite charge on the electrode. The double layer near a polarizable positively charged electrode in the presence of specifically adsorbed anions is depicted in Figure 30(a) and the same system at the p.z.c. is shown in Figure 30(b). An example of such a system is mercury in aqueous KI solution. From the large number of adsorption measurements made at Hg electrodes, definite trends for the electrosorption of organic molecules in general can be suggested. Electrocapillary measurements show that the decrease in interfacial tension in the presence of organic adsorbates reaches its maximum in the vicinity of the potential of zero charge. Similarly, a bell-shaped" curve for the variation of coverage with potential

PAGE 99

Metal Metal 87 Diffuse (b) Figure 29 Schematic Distribution of Charges and Potential (Due to Ions) at the Metal Electrolyte Interface at Positive (a) and Zero (b) Electrode Charge with no Specific Adsorption of Ions (See Reference 87)

PAGE 100

Metal Metal 88 layer (b) Figure 30 Schematic Representation of a Metal Solution Interface with Specifically Adsorbed Anions at Positive (a) and Zero (b) Electrode Charge (See Reference 87)

PAGE 101

89 was observed for adsorption at solid electrodes, particularly at lower solute concentrations. A "water competition" model has been proposed by Bockris, Devanathan and Muller to explain the potential dependence of the organic adsorption, while Frumkin thought the potential dependence of organic adsorption on Pt to be related to competition with hydrogen at low potentials and oxygen at high potentials and not primaOQ QQ rily with water molecules. ' The adsorption maximum is in general slightly negative to the p.z.c., which can be interpreted, at least in the water competition model, as due to the slightly preferred orientation of water molecules with oxygen toward the metal surface at the p.z.c. The p.z.c. was found to shift to positive values by the adsorption of ali7 f\ phatic compounds and to negative values by aromatics . ° According to Bockris et al. the shift of p.z.c. is due to the substitution of organic molecules for water dipoles. Compounds such as benzene, naphthalene, anthracene, phenanthrene and chrysene, although nonpolar, shift the p.z.c. in the 1 r negative direction. The adsorbability of these compounds at positively charged electrodes increases with increasing number of benzene rings in the organic molecule. The shift of the p.z.c. in the negative direction and the adsorption of aromatic compounds on positively charged electrode surfaces are attributed to ir-electron interaction which is facilitated by the benzene ring lying flat on the surface. The TT-electron interaction between adsorbed molecules and the electrode surface usually enhances the adsorption of

PAGE 102

90 organic substances. The replacement of hydrogen atoms in aromatic compounds by fluorine atoms leads to ir-electron exhaustion of the aromatic nucleus on account of the higher electron affinity of fluorine atoms. For this reason 7 r-electron interaction will virtually have no influence on adsorption of pentaf luoroaniline , pentaf luorobenzoic acid and pentafluorophenol molecules on mercury. Blomgren and Bockris used the electrocapillary curve method for studying the adsorption of aromatic amines aniline, o-toluidine, 2,3and 2 , 6-dimethylanilines , pyridine and quinoline from 0.1 M HCl solutions on mercury. They concluded that these substances are adsorbed predominantly in the form of RNH^ ions which lie flat on the electrode surface. The amount adsorbed varies little with variation of potential over a range of 1 V; the adsorption is determined predominantly by ir-electron interaction along the positive branch of the curve and by coulombic forces along the negative branch. The adsorption of the organic cations on positively charged surfaces diminishes but does not fall to zero. This result shows that forces of interaction between positive surface charges and rr-electrons of the aromatic nucleus predominate over electrostatic repulsions between the organic cations and the surface. Introduction of nucleophilic (electron-donor) substituents leads to an increase in adsorbability both owing to the increase in ir-electron density in the aromatic nucleus and to the increase of unshared electron density at the heteroatom

PAGE 103

91 (0 or N) . Introduction of electrophilic substituents has the opposite effect. Foreign surface active anions also play an important role in the studies of organic adsorption. Adsorption of organic cations and onium compounds which exist in cationic form in acid solutions, increases sharply on introduction of halide ions into the sulfuric or perchloric acid solutions. The increased adsorption of organic cations can be attributed to the changes of the surface charge of the metal in the presence of halide ions . Chemisorption of halide ions makes the metal surface appear to be negatively charged, and the cations are attracted by electrostatic forces to the metal surface. XÂ’/hen chemisorption of anions occurs and stable metal halogen covalent bonds are formed, the adsorbed ions enter the metal lattice on the surface and the charge of the ions forms a part of the charge of the metal surface.^^ The resultant dipoles shift the p.z.c. toward more positive values. As a result, the p.z.c. in the presence of chemisorbed anions is at values more positive than the p.z.c. in a solution not containing surface active anions. The surface of a metal with adsorbed halide ions can therefore be regarded as the surface of an electrode having a more positive potential of zero charge than the pure metal. If chloride and other halide ions enter only the ionic part of the double layer (e.g., in the case of the Hg electrode), adsorption shifts the p.z.c. to more negative potentials,. The adsorption and inhibiting action of organic cations and of organic compounds of the onium

PAGE 104

92 (ammonium, oxonium) type on metals of the iron group increase in the sequence from Cl~ to Br“ and especially l" ions; this is due to the more stable chemisorption of I~ ion which has the highest deformability of its electron shells. The p.z.c. of Fe, Cr and Ni has been determined to be -0.64, -0.52 and -0.69 V, respectively, in acidic solution. Thus the p.z.c. of stainless steel used in this study is as93 sumed to be -0.69 V. In the presence of the surface active chloride ion, the p.z.c. of the steel electrode may have shifted negatively to about -0.70 V. Thus a "bell-shaped" adsorption curve would be expected with an adsorption maximum in the vicinity of -0.700 V. The positive branch of this "bell-shaped" adsorption curve was obtained in the studies of the adsorption of DPP on a stainless steel electrode as shown in Figure 14. The inhibitor used in this study, DPP, is known as a chelating agent. It forms a very strong and stable complex with Fe(II) and has been used for Fe(II) deteirminations in solutions containing ferrous and ferric ions. Due to the presence of two electron-rich nitrogen atoms with two unshared electrons each which are available for bond formation, the adsorption behavior of DPP on a stainless steel electrode should be quite different from those adsorption behaviors that are only governed by coulombic forces and ir-electron interaction. Using chelating agents as corrosion inhibitors is not new in the literature, but a thorough study of the adsorption behavior of a chelating agent on a metal adsorbent

PAGE 105

93 8 9 is not available. Â’ The adsorption of benzotriazole on copper is the more extensively studied topic reported in the literature. Benzotriazole forms a waterinsoluble precipitate with both cuprous and cupric ions and has been used as an efficient corrosion inhibitor for copper in a wide variety of environments. Â’ Â’ Cotton concluded that benzotriazole inhibits the corrosion of copper by forming a Cu(I)benzotriazole compound on the copper surface.^ Poling showed from infrared reflectance spectra that the protective film was highly inert Cu(I)-benzotriazole complex polymer.^ Wall and Davis suggested that benzotriazole reacts to form an invisible insoluble chelate film on the copper surface. On the other hand, Mansfeld et al. suggested, on the basis of ellipsometric studies, that benzotriazole is chemisorbed on the copper surface. Ushenina and Klyuchnikov studied the influence of various chelating agents on the corrosion rate of iron, cobalt, zinc and copper, and found that when the chelate is formed directly on the surface as a compact film having good adhesion to the metal it retards corrosion. When the chelate is formed in the bulk of the solution and is then deposited, the film has only a slight retarding effect on corrosion. Niki, Hackerman et al. studied the effects of salicylaldoxime (SAG) and of 8-hydroxyquinoline (HQ) on the anodic dissolution of copper at various pH values. They found that at pH 1.5, where cupric ions do not precipitate as the chelate of SAG and HQ, the anodic dissolution of copper was retarded slightly by these additives. At pH 2 . 8 and 5.0

PAGE 106

94 where SAO precipitates cupric ions quantitatively, the anodic dissolution of copper was markedly retarded by SAO. The anodic polarization curves of copper in the SAO solutions were similar to those obtained in the passivation of iron. A thin protective chelate film was formed on the copper electrode. At very positive potentials, the chelate film ruptured and the anodic dissolution of copper started. On the other hand, HQ chemisorbed strongly on the copper electrode and retarded markedly the dissolution of copper, but the Cu HQ chelate did not act as a barrier for the copper dissolution in this pH range. By referring to the adsorption and inhibition behaviors of ben23triazole , salicylaldoxime and 8-hydroxyquinoline on a copper surface, a similar conclusion can be reached for DPP on stainless steel. A thin protective chelate film may form on the stainless steel electrode surface. Although no data are available to prove the existence of a chelate film, a red precipitate on the polarized steel electrode surface which may indicate the existence of a DPP chelate film was found when a concentrated DPP solution (about 1.0 x 10~^ M) was used. By adding a 4.1 x 10 ^ M Fe ^ ^ solution directly to a solution of 1.15 x 10 ^ M DPP no precipitate was found, thus the red precipitates on the steel electrode surface could be from the Fe DPP chelate film formed on the electrode surface. The red precipitate on the electrode surface could also arise from the accumulation of Fe DPP complex formed in the diffuse layer region. The pH value in the diffuse layer region may be substantially higher than that in the bulk solution

PAGE 107

95 because the proceeding hydrogen evolution reaction may have consumed the hydronium ions in the electrode electrolyte interface. The high pH (> 2.0) environment thus created would favor the formation of Fe DPP complex. The inhibitor effect of DPP on the corrosion (polarization) of 304 stainless steel can be found from Figure 15, the polarization current vs. potential relationship at different DPP concentrations. The inhibitor efficiency of DPP at various polarization potentials was calculated according to % Inh = i? . . . X 100 ^o where i^^^ and i^ represent the polarization current density measured with and without DPP in 0.1 M HCl solution, respectively. Table IV shows the calculated inhibitor efficiency of DPP at DPP concentrations between 1.02 x 10“^ M and 1.28 X 10 ^ M. For the purpose of comparison, the corresponding fractional surface coverages with DPP expressed in \ are also shown. The agreement is only good at the most cathodic potential (-0.700 V) at low DPP concentration. The measured /o surface coverage is in general larger than the corresponding inhibition efficiency. This tendency becomes more pronounced as the DPP concentration becomes larger. In the vicinity of the rest potential (corrosion potential) , the complicated competition between cathodic and anodic reactions makes the true inhibition efficiency difficult to evaluate and the difference between inhibition efficiency and % surface

PAGE 108

96 •H nJ r~l o PM C I -H c I -H C O cfl 0) VJ 4-> > U W O (U Q) CM^ (U 4J •H bO U tr) CO CO g o o •rH •u Mm oj C (1) t3 bO ct! !m QJ > O C a ‘rH c o U (1) >-H ,-1 o a o rt PM ‘iH >M P COrJH O S 4-J Ml -H O ^ Pd 4-3 Ml O tfl o 4-1 C . (U Ml X! a) pl, ~ c •rH 6 Ml M P iH P H > O o cn 00 VO cr> o iH 1 — 1 o VO CN CO m 1 — 1 CTv 1 tH 1 — 1 CO CO o LO oo VO iH VO o 1 — 1 vO o o iH P o 'r4 o 4-1 VO VO cov VO CO Pd • . . . . p o C3V lO o m o CO 4-1 1 1 I — I 1 — 1 CN 00 o PM > XJ P o •H o iH to <1VO 00 LO <} CN PM • • . . . . CM o VO 00 pI — 1 VO 1 — 1 < 1 iH tn tn OV > o o vO 00 00 p'' CO tn o VO o oo tn tn CO ' 1 — 1 CN o o CJv VO to VO iH CN o vjCN CN CD> VO tn 1 iH CN tn vO ov 1 — 1 pd kh 1 — 1 C >v 1 — 1 o o p o o P O O p •H pd pd p •H p pd p •H C pd P Ml p O bO 4-3 P O bO 4-1 P O bO *H *r4 •H P •H *r4 •H P •rH •rH •H p rQ P 4-1 Ml rO o 4-1 Ml rQ O 4-1 Ml •H *H o p •r4 'r4 o p •H *H O P 4d 44 p > -d 44 p > 4d 44 p > a 44 Ml O pd 44 Ml O Pd 44 Ml O M W Pm CJ M W Pm CJ M pa Pm CJ Pd o S S s •H VO VO in 4-1 1 1 1 P o o o Ml rH T-H 4-1 Pd X X X p o CnJ 00 oo Pi o VO CM o • u 1— 1 CO

PAGE 109

97 coverage becomes large. After the electrode surface becomes partially passivated, the agreement becomes better again. Proposed Inhibition Mechanism of DPP on the Corrosion of Stainless Steel in 0.1 M HCl Solution Three different types of inhibition mechanisms have been proposed for organic inhibitors; the indifferent coverage (blocking effect), the deactivating coverage and the reactive 99 coverage mechanism. In the case of indifferent coverage the basis electrode reaction does not take place at the covered area, which is usually relatively large, the degree of coverage 0^1. In the case of deactivating coverage the irihibitor particles are preferentially blocking the active sites of the electrode surface and therefore the degree of coverage 0 << 1. Reactive coverage arises when the primary adsorbed inhibitor species undergoes itself a chemical or electrochemical reaction or when the basis electrode reaction takes place at the covered electrode surface. The degree of coverage is usually large in this case, i.e., 0 -> 1. An experimental distinction between these three types of inhibition mechanisms is possible by combination of different kinds of electrochemical measurements, such as current vs. potential measurements under steady state as well as nonsteady state conditions, capacity vs. potential measurements and the rotating disc electrode technique. By referring to the experimental results obtained here, it is possible to interpret the inhibition effect of DPP on

PAGE 110

98 a stainless steel electrode in terms of the inhibition mechanisms described above. Since the fractional coverages measured at different applied polarization potentials are all quite large, the deactivating coverage (coverage of active sites) mechanism is ruled out and thus the inhibition effect may be interpreted by either the indifferent or reactive coverage mechanism. If the inhibition mechanism is the indifferent coverage one, 0 must be strongly dependent on the electrode potential and a complicated function 0 = f(E) may result from this dependence. Also a decrease of the current density in the normal mass transport controlled potential range can be expected. On the other hand, if the inhibition mechanism is the reactive mechanism one and the basis reaction can take place at the covered area, there will be no noticable decrease of the limiting diffusion current density in the mass transport controlled potential range. In this case, the increase in Tafel slope with increasing inhibitor concentration can be explained assuming that an additional energy barrier exists as a result of inhibitor adsorption. Thus the electrons of the basis reaction have to surmount two energy barriers, the first corresponding to the adsorption layer and the second to the ionic double layer. A mathematical treatment for such a dual energy barrier has been reported by Mayer and applied by Conway and Vijh to explain abnormally high Tafel slopes . A modern interpretation of this classical double barrier model in terms of the quantum mechanical view of electron tunneling is also available . Here , the

PAGE 111

99 probability of electron tunneling decreases at the covered area and therefore the Tafel slope has to increase. In the present system, it is difficult to tell which mechanism is followed based on the obtained experimental data alone. Since the Tafel slopes found are independent of the variation in DPP concentration (Figure 19) , the surface coverage 0 varies with applied potential (Figure 14) and the current density decreases with increasing DPP concentration (Figure 19) , the indifferent coverage mechanism seems to apply. However, from Table IV, the surface coverages measured are in general larger than the inhibition efficiencies calculated and that would favor the reactive coverage mechanism. Also, if the reaction rate of the reaction occurring on the covered area is much less than that on the uncovered area, the distinction between the two mechanisms will become even more difficult. Yamaoka and Fischer measured the steady state and transient polarization curves and potential vs. capacitance curves to study the electrochemical kinetics of iron corrosion in acid chloride solution in the presence of 1 , 10 -phenan thro line (o-ph) . According to their results the adsorbed o-phH"'' can inhibit both the cathodic and anodic partial reactions on the iron electrode surface. Moreover, they found that a parallel reaction takes place on the covered area. But the reaction rate on the covered area is substantially less than that on the uncovered area. Desorption of adsorbate from the electrode surface was also found when the electrode was anodically polarized to about -0.250 V.

PAGE 112

100 Because of the similar chemical properties of DPP and o-ph, the inhibition mechanism of DPPH"^ on the corrosion of a stainless steel electrode can be proposed to be similar to that of o-phH"** on an iron electrode. The interactions of DPPH on the cathodic and anodic reactions of the polarized stainless steel electrode are proposed as follows: (i) Interaction with cathodic hydrogen evolution: (18) + e^ (DPPH) ^^3 (19) (DPPH)^,3 I (DPP)^^^ + (20) ads + J . (21) + e" % H 2 (22) In this mechanism, hydronium ions diffuse from the bulk of the solution to the electrode surface either directly (step 21) or through the transport of PPPH*^ (step 18) . Since the adsorbed DPPH"*” ions undergo a charge transfer reaction (step 19) , the postulated inhibition may in fact act as an accelerator if the reaction rate of equation (19) is larger than the rate of hydrogen evolution on the uncovered surface. (ii) Interaction with the anodic dissolution reactions: Due to the incomplete understanding of the dissolution mechanism of stainless steel in acidic solution, the interaction of DPP on the anodic dissolution reaction of stainless steel seems complicated and ambiguous. According to Mueller,

PAGE 113

101 most of the dissolution current observed on stainless steels containing nickel should represent the oxidation of iron,^^^ Chromium and nickel should passivate as soon as they are exposed to solution by removal of iron from the lattice. Thus it is quite possible that the kinetics and the mechanism of stainless steel dissolution are similar to that of pure iron. Two different reaction mechanisms have been proposed for the anodic dissolution of iron in acidic sulfates and perchlorates (in the absence of halides) . The more generally accepted Bockris-Kelly mechanism (in acidic solution) Fe + HjO J (23) Fe-H^O j 2 ads H+ (24) Fe-OH" , ads F"-0»ads + e~ (25) Fe-OH , ads ^ Fe • OH"^ (r .d.s) + e (26) Fe • OH'*' + H"^ Z Fe'^ + H 2 O (27) Here reaction (26) is the rate determining step. A Tafel slop e of 0.040 V decade ^ and a reaction order with respect to hydroniijm ion of -1 are characteristic of this mechanism. Heusler proposed that the dissolution mechanism involves the conversion of into a surface catalyst Fe(FeOH)^j^ which is then oxidized to release the FeOH"*" snecies.^^^ ^^0\ds + Fe J Fe(FeOH)^^^ (26a) Fe(FeOH)^^g + OH" ^ FeOH"^ + FeOH^^^ + 2 e" (26b)

PAGE 114

102 FeOH'*' + t Fe"^ + H 2 O (27) Reaction (26b) is the rate determing step. The value of the kinetic parameters characterizing this reaction sequence depends on the manner in which they are measured. Steady state polarization results in a Tafel slope of 0.030 V decade ^ and a hydronium ion reaction order of -2. Nonsteady state measurements yield values of 0.060 V decade”^ J -1 X. • T 108 and -1, respectively. Recent work has suggested that the Bockris-Kelly mechanism predominates at low energy surface sites while the Heusler mechanism is favored by a high imperfection density and a large grain boundary to grain area ratio. In the presence of chloride ion which is more surface active than sulfate and perchlorate ion, there is competitive adsorption between chloride and hydroxide ion.^^^ Chloride adsorption will tend to prevail at high chloride ion activity and low pH, and accordingly the dissolution mechanism will be different from the Bockris-Kelly or Heusler case. Lorenz, Yamaoka, Fischer et al. first proposed the following sequence : Fe + H 2 O : Fe-H20^^^ ^^•Vads + i Fe.x;^^ + H2O (23) (28) Fe-X ads Fe-OH .+ ads + OH + X“ + 2 e" (29)

PAGE 115

103 The chemisorbed X ions displace adsorbed water molecules and then interact with adjacent adsorbed hydroxyls. Since an appreciable amount of Cl~ ion is present in the studied DPP adsorption system, the chemisorption of Cl" might have participated in the inhibition action of DPP on the stainless steel electrode according to the Lorenz, Yamaoka and Fischer mechanism. The interaction of DPP on the dissolution mechanism of stainless steel may thus be proposed as : + H+ + Cl+ 2 e(30) On the other hand, if the chemisorption of Cl” does not prevail on the surface, the adsorbed DPPH”^ may interact with FeOH^^g according to the Bockris-Kelly mechanism, + H^O + 2 (30a) A complete proposed mechanism for stainless steel dissolution in the presence of DPP can thus be written as : Fe + H 2 O J Fe-H20^^^ and .+ + 0’’P
PAGE 116

104 and + =1' i + HjO FeCl‘, + ads °™tds i (FeDPP)^^^ + -> (FaDPP)^+3 + e
PAGE 117

CHAPTER V SUMMARY Concentration depletion measurements are a direct method for measuring the amount of substance adsorbed on an electrode surface. To determine the concentration change in bulk solution, the UV spectrophotoraetric , radioactive tracer and chromatographic methods are generally used. Among these methods, UV spectrophotometry is the simplest, most convenient and precise method for concentration determination. An inherent difficulty is that the spectrum of an organic species is very easily interfered with by the presence of other UV absorbing species, e.g., metal ions dissolved in solution. The proposed modified dual -wave length method removes part of this difficulty and allows to determine the concentration of an organic substance in solution in the presence of other interfering ions (matrix ions) . The method thus makes possible an extension of Foley and Alexander's work, i.e., to monitor spectrophotometrically the corrosion rate of iron in acidic solution in the presence of an organic inhibitor having strong absorption in the UV, VIS range. In this experiment, the spectrum measurement was done with an ordinary cuvette, An alternative way would be the application of a flow cell with the whole system sealed 105

PAGE 118

106 and the solution pumped through a raicrocuvette where the spectrum of the solution can be taken continuously, Spectrophotometric and electrochemical data could thus be obtained simultaneously allowing more information about the corrosion behavior to be gathered. The polarization curves obtained in 0.1 M HCl and in various DPP concentrations were corrected according to their fractional surface coverage. The interpolated polarization curves at various constant fractional surface coverage (0 = 0.1 to 0.5) are parallel to the polarization curve obtained in 0.1 M HCl solution (without DPP) with identical Tafel slopes (cathodic) (Figure 19), indicating a simple blocking effect. The proposed inhibition mechanisms of DPP on stainless steel electrodes was derived from Yamaoka and Fischer's inhibition mechanism for o-phenanthroline on the corrosion of iron in acidic solution. Although no direct data to support the proposed mechanism are presented, clues are available to discuss their possibility. Table I shows that the open circuit potential shifts anodically with the increase of DPP concentration in solution. According to Kaesche, the shift of the in the presence of an inhibitor can be used as an indication as to which electrochemical process the inhibitor influences predominantly. Thus it is concluded that the inhibitor DPP has a more pronounced effect on the anodic process. This is in agreement with the proposed mechanisms which require that the inhibition effect of DPP on stainless steel electrodes is mainly due to the blocking effect

PAGE 119

107 (i^diffeiT6nt coveragG rnGchanisin) on ttiG GlGctrodG surfacG but that a parallel reaction (hydrogen evolution) may also occur on the covered area. This parallel reaction thus eliminates part of the inhibition effect of DPP on the cathodic process and the shifts slightly in the anodic direction. Fifteen different adsorption isotherms were tested against the experimental data; five of them were found to fit the spectrophotometric adsorption data very well (with correlation coefficients larger than 0.95) in the low surface coverage region (from 0 = 0 . 1 to 0 = 0 . 5) . The Frumkin isotherm, Virial Coefficients isotherm and Hill-de Boer isotherm are basically the same type, all having the same predominant empirical term exp (-2b0) . Thus they show the same results (Table III) . The Blomgren-Bockris and the ConwayBarradas isotherms are also very similar, although the former is dedicated to the adsorption of organic ions and the latter describes the adsorption of neutral organic species. Both are equally applicable to the experimental data. From the interaction parameters obtained, an attraction force is existing between the adsorbed species . The inhibition efficiencies range from about 157o at 1.02 X 10 M DPP concentration to about 65% at 1.28 x 10"^ M DPP concentration in the potential region of -0.500 V to -0.700 V, while the fractional surface coverages range from about 207o to 93% at the same corresponding DPP concentrations and the same applied potential region. Good agreement between inhibition efficiency and fractional surface coverage

PAGE 120

108 was obtained only in the most cathodic potential region and at low DPP concentration range (Table IV) .

PAGE 121

REFERENCES CITED 1. U. R. Evans, "Metallic Corrosion, Passivity and Protection," p. 5, London, Edward Arnold Co., 1948. 2. N. Hackerman, Corrosion, 8, 143 (1952). 3. W. Mehl , J. M. Hale, and F. Lohmann, J. Electrochem. Soc . , 113 , 1166 (1966). 4. I. Dugdale and J. B. Cotton, Corrosion Science, 3, 69 (1963). 5. J. O'M. Bockris and B. E. Conway, J. Phys . and Colloid Chem., 527 (1949). 6. L. Horner, Werkstoffe Korros . , 466 (1972). 7. H. Fuchs and L. Horner, Chem. Ber. , %, 3141 (1963). 8. J. B. Cotton, Trans. Inst. Marine Eng., 11 _, 165 (1965). 9. G. W. Poling, Corrosion Sci., 1^, 359 (1970). 10. N. Hackerman and A. C. Makrides, Ind. Eng. Chem., 46, 523 (1954). — 11. E. I. Mikhailova and Z. A. lofa, Elektrokhimiya , 1, 107 (1965). 12. Z. A. lofa, V. V. Batrakov and Cho Ngok Ba, Electrochim. Acta, 9, 1645 (1964). 13. Z. A. lofa. Proceedings of the 2nd European Symposium on Corrosion Inhibitors, Ferrara, Ann. Univ. Ferrara, Ser. V, 151 (1965). 14. K. S. Rajagopalan, European S}nnposium on Corrosion Inhibitors, Ferrara, I960; Compt. Rend. Univ. Ferrara, 685 (1961). 15. B. Damaskin, 0. Petrii and V. Batrakov, "Adsorption of Organic Compounds onElectrodes , " Plenum Press, New York 1971, p. 308. 109

PAGE 122

110 16. A. N. Frumkin and A. I. Shlygin, Acta Physicochim. URSS, 5, 819 (1936). 17. G. A. Deborin and B. V. Ershler, Zh. Fiz . Khim. , 14, 708 (1940). ~ 18. N. A. Balashova and V. E. Kazarinov, Usp. Khim., 34, 1721 (1965) . — 19. M. W. Breiter and S. Gilman, J. Electrochem. Soc., 109, 622 (1962). 20. G. M. Schmid and N. Hackerman, J. Electrochem. Soc., 110, 440 (1963). 21. N. S. Hush, "Reactions of Molecules at Electrodes," John Wiley and Sons, Ltd., London, 1971, Chapter 5. 22. R. N. Adams, "Electrochemistry at Solid Electrodes," Marcel Dekker, Inc., New York, 1969, Chapter 10. 23. E. Yeager and A. J. Salkind, "Techniques of Electrochemistry," Vol 3, Wiley-Interscience Publication, New York, 1978, Chapter 6 24. D. C. Grahame and R. W. Whitney, J. Am. Chem. Soc., 64, 1548 (1942). ~ 25. D. M. Mohilner, in A. J. Bard, Electroanalytical Chemistry, Vol. 1, Marcel Dekker, New York, 1966, p. 242. I 26. D. C. Grahame, Ann. Rev. Phys . Chem., 337 (1955). 27. G. Lippmann, Ann. Chim. Phys., 5, 494 (1875). 28. J O'M. Bockris and R. Parry-Jones, Nature, 1^, 930 (1953). 29. F. P. Bowden and D. Tabor, "Properties of Metallic Surfaces," Inst. Metals, 1953. 30. T. Murakawa and N. Hackerman, Cossorion Sci., 4, 387 (1964). 31. B. E. Conway and M. Dzieciuch, Can. J. Chem., 41, 21, 38, 55 (1963). ~ 32. 0. Volk and H. Fischer, Electrochim. Acta, 112 (1961). 33. G. H. Hills and R. Payne, Trans. Faraday Soc., 61, 316 (1965). ~ 34. G. H. Nancollas and P. Vincent, J. Sci. Inst., 40, 306 (1963). ~

PAGE 123

Ill 35. M. W. Breiter, J. Electrochem. Soc., 112 , 845 (1965). 36. P. J. Hilson, Trans. Faraday Soc., 462 (1952). 37. U . V . Pain, V. E. Past, and R. Ya. Pullerits, Elektrokhimiya, 2, 604 (1966). 38. R. de Levie, Electrochim. Acta, 9, 1231 (1964), 10, 113 39. N. A. Balashova, Z. Physik. Chem. , 2^, 340 (1957). 40. N. A. Balashova and N. S. Merkulova, Radiokhimiya , 2, 699, 704 (1960). 41. E. Blomgren and J. O'M. Bockris, Nature, 1 ^, 305 (1960). 42. G. Aniansson, J. Phys . Chem., 1286 (1951). 43. F. Joliot, J. Chim. Phys., 119 (1930). 44. H. Wroblowa and M. Green, Electrochim. Acta, 8, 679 (1963). 45. E. Gileadi, B. T. Rubin, and J. O'M. Bockris, J. Phys Chem., M, 3335 (1965). 46. K. Schwabe and W. Schwenke, Electrochim. Acta, 9, 1003 (1964). 47. M. Green, D. A. J. Swinkels, and J. O'M. Bockris, Rev. Sci. Inst., 33, 18 (1962). 48. B. E. Conway, R. G. Barradas and T. Zawidzki, J. Phys. Chem., § 1 , 676 (1958). 49. S. A. Balezin, P. G. Kuznetsov, and I. A. Podolnyi, "Corrosion Inhibitors," Sudostroenie , Moscow-Leningrad , 1965. 50. R. J. Newmiller and R. B. Pontius, J. Phys. Chem., 64, 584 (1960). — 51. R. G. Barradas and B. E. Conway, J. Electroanal. Chem., 6, 314 (1963). 52. R. R. Annand, R. M. Hurd, and N. Hackerman, J. Electrochem. Soc., m, 138 (1965). 53. J. E. Harrar, F. B. Stephens, and R. E. Pechacek, Anal. Chem., 1036 (1962). G. F. Smith, W. H. McCurdy and H. Diehl, Analyst, 77, 418 (1952). 54 .

PAGE 124

112 55. Handbook of Chemistry and Physics, 53rd Edition, The Chemical Rubber Co., Cleveland, Ohio, 1972, p. F-179. 56. lUPAC, Commission on Spectrochemical and Other Optical Procedures for Analysis, "Spectrophotometric Data for Colorimetric Analysis," Butterworths , 1963. 57. R. S. Drago, "Physical Methods of Inorganic Chemistry," Reinhold Publishing Corp . , 1965. 58. B. J. Alexander and R. T. Foley, Corrosion, M, 417 59. B. J. Alexander and R. T. Foley, Corrosion, (1976). 32, 297 60. M. Effrenfruend and J. L. Leibengoth, Bulletin de la Societe Chimique de France, 1_, 2498 (1970). 61. R. A. Erb, J. Phys. Chem. , M, 1306 (1965). 62. P. Delahay , "Double Layer and Electrode Kinetics," Interscience Publishers, New York, 1965, p. 82. 63. R. Parsons, Reports of the Fourth Soviet Conference on Electrochemistry, English Translation, Consultants Bureau, New York, 1958, p. 18. 64. H. Freundlich, "Colloid and Capillary Chemistry," Methuen, London, 1926. 65. I. Langmuir, J. Am. Chem. Soc., 4D, 1369 (1918). 66. A. N. Frumkin, Z. Phys. Chem., 116 , 466 (1925). 67. R. Parsons, Trans. Faraday Soc., 51, 1518 (1955): 55. 999 (1959). ~ — 68. M. Volmer, Z. Phys. Chem., 115 , 253 (1925). 69. E. Helfand, H. L. Frisch, and J. L. Lebowitz, J. Chem. Phys., 1037 (1961). 70. T. L. Hill, J. Chem. Phys., W, 141 (1952). 71. J. H. De Boer, "The Dynamical Character of Adsorption," Oxford Univ. Press, 1953. 72. R. Parsons, J. Electroanal. Chem., 7, 136 (1964); 8, 93 (1964) . 73. M. I. Terakin, Zh. Fiz . Khim. , 1^, 296 (1941).

PAGE 125

113 74. E. Blomgren and J. O’M. Bockris, J. Phys . Chem. 63 1475 (1959). — ’ 75. A. N. Frumkin, Z. Physik, 792 (1926). 76. E. Blomgren, J. O'M. Bockris, and C. Jesch, J. Phys Chem., 2000 (1961). 77. B. E. Conway and R. G. Barradas, Electrochim. Acta 5 319 (1961), 78. B. B. Damaskin, A. A. Survila and L. E. Rybalka, Elektrokhimiya, 3, 146, 927, 1138 (1967). 79. W. Lorenz, F. Mockel, and W. Muller, Z. Phys. Chem. (N F ) 145 (1960). V 80. B. B. Damaskin, Electrochim. Acta, 9, 231 (1964). 81. P. Delahay, "Double Layer and Electrode Kinetics," Interscience Publishing Co., New York, 1965, p. 84. 82. B. J. Piersma, "Organic Adsorption at Electrodes," in Gileadi's Electrosorption, Plenum Press, New York, 1967. 83. H. L. F. von Helmholtz, Ann. Physik, (2), 211 (1853). 84. B. E. Conway, "Theory and Principles of Electrode Processes," The Ronald Press Co., New York, 1965. 85. G. Gouy, J. Phys. Radium, 9^, 457 (1910). 86. D. L. Chapman, Phil. Mag., 475 (1913). 87. J. O'M. Bockris and B. E. Conway, "Modern Aspects of Electrochemistry," No. 5, Plenum Press, New York. 1969 p. 213. 88. J. O'M. Bockris, M. A. V. Devanathan, and K. Muller, Proc. Roy. Soc., 274A , 55 (1963). 89. A. N. Friimkin, Elektrokhimiya , ]^, 394 (1965). 90. B. B. Damaskin, M. M. Andrusev, V. M. Gerovich, and R. I. Kaganovich, Elektrokhimiya, 3, 667 (1967). 91. N. Hackerman, E. S. Snavely, Jr., and J. S. Payne, Jr. J. Electrochem. Soc., U3, 677 (1966). B. B. Damaskin, 0. A. Petrii, and V. V. Batrakov, "Adsorption of Organic Compounds on Electrodes," Plenum Press, New York, 1971.

PAGE 126

114 93. 94. 95. 96. 97. 98. 99. 100 . 101 . 102 103. 104. 105. 106. 107. 108. 109. 110 . 111 . 112 . M. E. Curley-Fiorino, Doctoral Dissertation, Univ. of Florida, 1975. J. B. Cotton and I. R. Scholes , Brit. Corrosion J 2, 1 (1967). F. Mansfeld, T. Smith, and E. P. Parry, Corrosion 27 289 (1971). ’ — ’ K. H. Wall and I. Davies, J. Appl. Chem. , 389 (1965) U. F. Ushenina and N. G. Klyuchnikov, Zhur. Prikl Khira 191 (1971). K. Niki, F. M. Delnick, and N. Hackerman, J. Electrochem. Soc., m, 855 (1975). M. S. Abdelaal, A. A. Miligy, G. Reiners, and W. G Lorenz, Electrochim. Acta, 507 (1975). H. Fischer, Werkstoffe Korros., 1 ^, 445, 453 (1972). R. E. Mayer, J. Electrochem. Soc., ]^, 847 (1960). B. E. Conway and A. K. Vijh, J. Phys . Chem., 71, 3637 (1967) . — R. R. Dogonadze, J. Ulstrup, and Yu. I. Kharkats J Electroanal. Chem., 161 (1973); 47 (1972). H. Yamaoka and H. Fischer, Electrochim. Acta, 10, 679 (1965) . — ’ W. A. Mueller, Corrosion, 1^, 73t (1962). J. O'M. Bockris, D. Drazic, and A. R. Despic, Electrochim. Acta, 4, 325 (1961). E. J. Kelly, J. Electrochem. Soc., 112 , 124 (1965). K. E. Heusler, Z. Elektrochem. , 582 (1958). F. Hilbert, Y. Miyoshi, G. Eichkorn, and W. J. Lorenz J. Electrochem. Soc., 118 , 1919 (1971). 108^’ 732^(1961)’^^ Electrochem. Soc., W. J. Lorenz, H. Yamaoka, and H. Fischer, Ber. Bunsenses Phys. Chem., 932 (1963). H. Kaesche and N. Hackerman, J. Electrochem. Soc 105 191 (1958) . ’ — ’

PAGE 127

BIOGRAPHICAL SKETCH The author was born in Taipei, Taiwan, Republic of China, on January 5, 1944. He attended high school in Taiwan and graduated with the Bachelor of Science degree from the National Taiwan Normal University in June 1966. He taught as a high school chemistry teacher for one year. In June 1969, he received the Master of Science degree from the National Tsing Hua University. He began his graduate program at the University of Florida in September 1972. He married the former Shang-Cheng Chiu in July, 1972, and is now the father of Poyin W. Huang. 115

PAGE 128

I certify that I have read this study and that in niy 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! Gj^rhard M. Schmid , Chairman Associate Professor of Chemistry 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! Roger /G . Bates Profefe'sor of Chemistry 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! R. Carl Stoufe'f Associate Professor of Chemistry

PAGE 129

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! 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! This dissertation was submitted to the Graduate Faculty of the Department of Chemistry in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1978 Professor of Materials Science and Engineering J ame s D . Winef qrdner Graduate Research Professor of Chemistry Dean, Graduate School