ANALYTICAL CHEMISTRY, VOL. 50, NO. 8, JULY 1978
I
jot
I
11°
1-
TIME MIN
Figure 6. High permeability 3 YO silicon-iron containing aluminum nitride
The surface condition, particularly the presence of a thin silica layer, can markedly alter the rate of nitrogen egress. In 3% Si-Fe, it can also influence the identity of the species observed. A MnS inhibited steel in the decarburized condition was examined with and without surface preparation (mechanical abrasion to remove approximately 20 pm of surface). Before abrasion, a ‘‘soluble’’nitrogen and large a-Si3N4peak are observed (Figure 5). By abrading, the “soluble” signal is relatively unchanged but the a-Si3N4signal is removed. Apparently the surface region contains a significant concentration of a-Si3N4. This observation has been made for 3% Si-Fe containing MnS and high permeability 3% Si-Fe containing boron. The removal of the a-Si3N4allows better resolution of the AlN signal. With the current interest in Si-Fe containing AlN and acknowledging the stability of AlN, a detectable signal should
1049
be apparent from steels “inhibited” by A1N. Using the composition of AlN “inhibited” material shown in Table 11, a trace as illustrated in Figure 6 is observed for a decarburized alloy. A1N is the predominant species with lesser quantities of a-Si3N4. The soluble fraction is very low. The shift to higher temperature for the dissolution of A1N in this sample (-80 “C higher) is believed a result of the higher A1N dissolution temperature a t this composition as illustrated by Leslie (5) for A1 and N in austenite.
CONCLUSIONS A technique has been developed to continously monitor the nitrogen egress from bulk samples of 3% Si-Fe annealed in hydrogen. The apparatus and technique are applicable to variable heating rates, isothermal conditioning, and other metal systems. By comparison with published data on nitride solubilities and by selecting materials of different chemistry, peak assignments have been made for “soluble” nitrogen, a-Si3N4,AlN, and BN which help define the nature and role of nitrogen bearing species in the secondary recrystallization of silicon-iron. LITERATURE CITED H. E. Grenoble and H. C. Fiedler, J . Appl. Phys., 40, 1575 (1969). R. Fisher, British Steel Corporation, PB234305 (June 1974). W. Oelsen and K. H Saurer, Arch. Eisenbuetfenwes.,38, 141 (1967). J. 8 . Headridge and G. D. Long, Analyst. (London), 101, 103 (1976). (5) W. C Leslle, Am. Iron Steel Inst. Monogr., New York, N.Y., 1959. (6) W. Roberts, Jernkontarifs Ann., 155, 286 (1971). (7) T. N. Yanovaskaya et al., Izv. Akad, Nauk SSSR, 39, 1543 (1975). (8) N. S . Corney and E. T. Turkdogan, J . Iron SteelInst., 180, 344 (1955). (9) G. White, G. D. Hall, and R . Fisher, BISRA Open Rep., MG/D/695/70. (IO) K. Kawamura, T. Otsubo, and T. Mori, Trans. Iron SteelInst Jpn., 14, 347 (1974). (1) (2) (3) (4)
RECEIVED for review December 23,1977. Accepted April 10, 1978.
Determination of the Standard Electrode Potential of Potassium Amalgam in Ethylenediamine Suraj P. Makhija‘
Department of Chemistry, Indiana University, Bloomington, Indiana 4740 1
To determine potentiometrically the standard electrode potential of potassium amalgam in ethylenediamine, two different salts, KBr and K I , with widely different ion-pair association constants were used. Liquid junction potentials corrections seemed very important in this study. An average of -0.1656 V for Eo of potassium amalgam vs. Zn(Hg)lZnCI,INaCI references in ethylenediamine was obtained. Relative standard deviation was 1.647%.
The “exact” expression for the activity of a n ion of uniunivalent salt in a solvent of low dielectric constant, as shown Present address, D e p a r t m e n t of Chemistry, A l a b a m a S t a t e U n i v e r s i t y , M o n t g o m e r y , Ala. 36101.
0003-2700/78/0350-1049$01 .OO/O
by Bruckenstein and Mukerjee ( 1 ) and Schaap (2) and coworkers, is given by
(M’) = (X-) =
(1+- C ~ K ~ & 2 )- ”1 2 2 KIMXfi
where (M’) is activity of cation, (X-) is the activity of anion, CMx is the stoichiometric concentration of the 1:l salt (MX), fi is the mean ionic activity coefficient and Km is the ion-pair formation constant of MX. Nernst’s equation for the potential of metal-metal ion electrode, vs. a reference electrode, is
RT
E = EoM,w+ Ej + - In (M)’ nF where EoM,M+ is the standard electrode potential and Ej is the 0 1978 American Chemical Society
1050
ANALYTICAL CHEMISTRY, VOL. 50, NO. 8, JULY 1978 ~~~~
Zn(Hg)-ZnClz-NaC1 (Sat’d)
NaCl (Sat’d)
KX
K(Hg)-KX
Bridge
Electrode Compartment
Figure 1. The
four-compartment
H shaped cell
liquid junction potential, if any. Substituting the expression for M+ into Equation 2, we get
E = EoM,M+
Table I. Electrode Potential of Potassium Amalgam in Contact with Ethylenediamine Solution Containing Potassium Bromidea
RT + Ej + In nF
Eobsd
(3) in which E is the potential measured vs. the same reference electrode. To evaluate E o M , M + values in a solvent of low dielectric constant, it is necessary to measure the potential ( E ) of the indicating electrode as a function of the stoichiometric concentration of the salt (CMx), vs. a suitable reference electrode, and t o know, or t o be able to calculate, values of K M X , fi, and Ej. These data are becoming available for solutions in anhydrous ethylenediamine so that calculation of Eo values in this solvent can now be attempted.
EXPERIMENTAL Two-phase potassium amalgam was prepared by electrolyzing a 5% solution of KOH in methanol using a mercury pool as cathode and a platinum wire gauze as anode. The amalgam was thoroughly washed with methanol to remove excess KOH and then stored under ether. Ethylenediamine (98-100%) was doubly distilled from calcium hydride under reduced pressure in an atmosphere of nitrogen. Reagent grade KI and KBr were used. A vacuum oven was used for drying these salts. The four-compartment H-shaped cell was used for these experiments, as shown in Figure 1. Potassium amalgam was placed at the bottom of the indicating electrode compartment and contact to the potentiometer was made by means of a platinum wire sealed through the glass. The indicating electrode and reference electrode (2) were separated by two salt bridges, one of which contained ethylenediamine saturated with sodium chloride and the other of which contained the potassium salt dissolved in ethylenediamine. The concentration of the potassium salt in the bridge compartment and in the indicating electrode compartment was changed simultaneously and equally by adding appropriate amounts of concentrated solution of KX to both compartments with a buret. The bridge containing KX was introduced in order to avoid any direct mixing of saturated sodium chloride solution with the KX solution in the indicator cell. The indicator cell was continuously flushed with dry nitrogen between potentiometric readings. All potentials were measured with a null-type Rubicon potentiometer capable of being read to f O . O 1 mV. A Honeywell high-gain electronic null-indicator, Model No, 104 WIG, was used with the potentiometer. RESULTS AND DISCUSSION Equation 3 can be written
0.05916 log
(1+ 4 C,X,fi’)”’ 2 KKxh
-
1
CKBrr
mol/L
1.5094 X 1.8120 X 2.5000 X 3.0770 X 4.5801 X 6.0606 X 8.9552X 1.1765 X 1.7143 X 2.9333 X 4.9411 X 6.5263 X 7.8095 X 8.8650 X
(volts) vs. ref, electrodeb
fi
lom48.50 X l o - ’ 8.41 X lo-’ 8.25 X lo-’ 8.14 X 7.92 X 7.67 X 7.54X 7.37 X 7.13 X 6.77 X 6.42 X 6.22 X 6.09 X 6.00 X
lo-’ lo-’ lo-’
lo-’ lo-’ lo-’ lo-’ lo-’ lo-’ lo-’ lo-’
-0.3958 -0.3946 -0.3907 -0.3881 -0.3825 -0.3794 -0.3738 -0.3692 -0.3637 -0.3565 -0.3480 -0.3434 -0.3406 -0.3390
Eo
(volts) vs. ref. elecEj (mV) trodeb -3.69 -3.64 -3.54 -3.48 -3.37 -3.29 -3.18 -3.11 -3.02 -2.88 -2.75 -2.69 -2.64 -2.61
-0.1503 -0.1526 -0.1543 -0.1553 -0.1564 -0.1578 -0.1585 -0.1582 -0.1584 -0.1592 -0.1582 -0.1576 -0.1573 -0.1565
The ion- air association constant used to calculate E o was 8134. The reference electrode used was Zn(Hg)ZnC1, (sat’d.), NaCl (sat’d.) in ethylenediamine. Relative standard deviation is of the order of 1.647%. coefficient equation
fi,
were obtained by successively solving the
(5) until a constant result was obtained. In this equation CM+ is the concentration of the dissociated ion, CMx is the stoichiometric concentration of potassium salt, and KMxis the ion-pair association constant. A computer program was set u p t o change the value of fi until the values of CM+ and fi changed insignificantly (one part per thousand). Ej is the liquid junction potential, the calculations of which were published earlier (4). Nernst’s equation is applied to a junction involving the same electrolyte a t two different activities and is written
in which t+ and t- are the transport numbers of cation and anion, respectively. The Lewis and Sargent equation is applied t o a junction involving two different electrolytes (like KBr, NaBr or NaC1, NaBr) having one common ion a t constant activity. In approximate form, it may be expressed as
(4)
where E°K(Hg)K+ is the standard potential of potassium amalgam in ethylenediamine, E is the experimentally measured potential of potassium amalgam in contact with a solution containing KX, vs. the reference electrode, and CKX is the stoichiometric concentration. K K Xis the ion-pair formation constant of KX and was determined independently by conductance measurements (3);values of the mean activity
where AIo and Azo are the limiting equivalent conductances of the two salts. Application of Equations 4-1 to the data obtained using saturated two-phase potassium amalgam as an indicator electrode gave an average value of -0.1565 V with RSD = 1.647% for the Eo of potassium amalgam vs. the Zn(Hg)ZnClz-NaC1 reference electrode in ethylenediamine. Two different salts were used, KBr and KI, which have widely
ANALYTICAL CHEMISTRY, VOL. 50, NO.
different ion-pair association constants. The data for KBr are given in Table I. Once a value of Eo is well established for a particular metal/metal-ion electrode, it is possible to determine potentiometrically the ion-pair formation constants of all the soluble 1:1salts of that metal by application of the equation:
where (M') = antilog ( E - E0M,M+)/0.05916 and the experimental values of E are corrected for liquid junction potentials.
8,JULY 1978
1051
ACKNOWLEDGMENT I am grateful to Ward B. Schaap of Indiana University for his suggestions and help in this work.
LITERATURE CITED (1) S. Bruckenstein and L. Mukerjee, J . Phys. Chem., 84, 1601 (1960). (2) W. 8. Schaap, R. E. Bayer, J. R. Seifker, J. Y. Kim, P. W. Brewster, and F. C. Schmidt, Rec. Chem. Prop., 22, 197 (1961). (3) F. Puspanaden, Ph.D dissertation, Indiana University, Bloomington, Ind., 1966. (4) S.P. MakhiJa, Can. J . Chem., 55, 2962 (1977).
RECEIVED for review January 30, 1978. Accepted April 12, 1978.
Fabrication and Characterization of a Kel-F-Graphite Composition Electrode for General Voltammetric Applications Jeffrey E. Anderson, Dennis E. Tallman," David J. Chesney, and James L. Anderson* Department of Chemistry, North Dakota State University, Fargo, North Dakota 58 102
An electrode composed of compressed powdered graphite and Kel-F (the Kelgraf electrode) Is descrlbed, whlch shows promise for application In a wide range of solvents. Fabricatlon of the electrode and its application to aqueous and nonaqueous cyclic voltammetry and aqueous stripping analysis are discussed. The electrode offers significant advantages over many other carbon formulations in nonaqueous applications and in strlpplng analysls wlth deposition at thin mercury films. Relative selectivity of the Kelgraf electrode for polar and nonpolar reactants can be partlally modlfled as a functlon of surface pretreatment. Roughening of the surface enhances current response for polar solutes (e.g., Fe(CN):in water) while little affectlng current response for relatively nonpolar solutes wlth similar formal reduction potentials (e.g., ferrocene in methanol).
Numerous carbon electrodes have been described in the literature (1--11)for use in voltammetry and stripping analysis. The carbon used in the construction of these electrodes has ranged from glassy carbon to graphite powder. Glassy carbon is very hard and brittle and its electrochemical behavior is dependent on surface preparation ( I ) as well as its source (2). With the exception of glassy carbon and basal-plane pyrolytic graphite, an electrochemically inert binder or impregnator is required to attain a nonporous surface. Often, however, it is desirable to use an impregnator which allows easier electrode fabrication, cleaning, and polishing than is possible with glassy carbon. Carbon paste electrodes which use graphite powder mixed with various organic oils offer an easily renewable surface. These electrodes usually are restricted to aqueous systems to prevent dissolution of the binder ( 3 ) ,although a nonaqueous formulation has been described ( 4 ) . The binder may restrict the potential range of the electrode (5). Furthermore, the type of paste used appears to affect the electrochemical activity of various substances in different ways ( 5 ) . Wax impregnated graphite (WIG) electrodes have been used in anodic stripping voltammetry (ASV). These electrodes appear to have a finite shelf life which is attributed to 0003-2700/78/0350-1051$01.00/0
crystallization of the wax, resulting in increased porosity which, in turn, leads to a broadening of the stripping peaks (6). Memory effects may also accrue with WIG electrodes ( 2 ) . Various polymers have also been used as impregnators or as binders. Chemical polymerization was used in the curing of an epoxy binder in the graphoxy electrode (7) which works well for differential pulse ASV (DPASV) when plated with a thin mercury layer. Large background currents, however, limit the use of this electrode in other forms of voltammetry. Both Clem and Sciamanna (8) and McLaren and Batley ( 2 ) have used radiation curing of various polymer impregnators to fabricate electrodes with lower residual currents than was obtained with the epoxy binder. Klatt et al. (9) described the construction of a Teflongraphite electrode and demonstrated its use in cyclic voltammetry (CV). Teflon is particularly attractive due to its inertness in both aqueous and most organic solvents. T o achieve good mechanical strengths and low background currents, low ratios of carbon to Teflon were required yielding relatively high resistances (>20 R). We have found that sintering of these electrodes at 360 "C substantially improves their mechanical strengths at higher carbon concentrations (10). Kel-F has long been used by chemists because of its inertness in a wide range of solvents including concentrated alkalies and strong acids. I t is more rigid than Teflon and easier to machine, with a working temperature range of -200 to +200 OC. Morcos and Yeager (11) have described the use of this polymer for encapsulating edge oriented pyrolytic graphite. We describe in this paper the use of Kel-F as a binder for powdered graphite to form electrodes with a wide range of desirable properties including ease of fabrication in practically any desired shape, low residual currents, and wide potential range. These "Kelgraf' electrodes can be fabricated with typical resistances of less than 2 R. Applications in aqueous and nonaqueous solvents are described.
EXPERIMENTAL Instrumentation. A Princeton Applied Research (PAR) model 174A Polarographic Analyzer or model 173 Potentiostat equipped with a model 179 Digital Coulometer was used in 0 1978 American Chemical Society