while the piston advanced, a delay tube was incorporated in the circuit. The tube contacts were normally closed and passed current to hold the external relay closed, thus powering the driving solenoid, despite the fact that the internal timer relay opened. After approximately one minute the delay tube opened, thus opening the external relay and deactivating the driving mechanism solenoid. Simultaneously the timer clutch would be engaged to start another timing cycle. Because of the recycling time and the time needed for the solenoid to advance the piston, 1 minute was the shortest time possible between successive samples; however, with modification this time could be shortened. Polyethylene tubing connected the piston and the reaction system. The exhaust ports were also fitted with lengths of polyethylene tubing which in turn were attached to short
lengths of glass tubing which were then immersed in 125-1111 Erlenmeyer flasks. G a s would now flow, 10-20 cc/min., from the reaction vessel through a mercury trap, (keeping water out of the reaction vessel) through the manifold and piston to exit through the exhaust ports and then be expelled into the water. The sampling device would switch the gas flow to successive exhaust ports after the set time interval and the fractions collected could be subsequently assayed.
REcEivED for review September 8, 1969. Accepted January 22, 1970. We gratefully acknowledge an Undergraduate Research Participation Grant from the National Science Foundation for DCW as well as support from The Research Corporation.
The Tubular Carbon Electrode W. D. Mason' and Carter L. Olson College of Pharmacy, The Ohio State University, Columbus, Ohio 43210
DEVELOPMENT of small tubular electrodes for use in flowing streams was first reported by Blaedel et al. ( I ) . This tubular platinum electrode was applied to the enzymatic determination of glucose and illustrated the convenience of these specially designed electrodes with very low holdup volume for use as analytical monitoring devices in flowing streams (2). Since that time several papers have appeared describing thebehavior andtheory of tubular electrodes (3-6). Oneof the immediate shortcomings of the tubular electrode was that it was constructed from platinum and had all of the problems associated with platinum surfaces. I n order to extend the utility of the tubular electrode, Osterling and Olson prepared and evaluated a mercury-coated tubular platinum electrode (7). A tubular carbon electrode has been described by Sharma and Dutt, but it is a rigid assembly similar in construction to the early tubular platinum electrodes (8). Because a mercury surface has a very limited anodic voltage range and the fact that construction of tubular platinum electrodes presents some fabrication problems, the development and evaluation of a tubular carbon electrode that is easy and rapid to construct was undertaken. EXPERIMENTAL Instrumentation. Voltammetric studies were carried out
using a Heath chopper stabilized polarograph. Recordings were made o n a Varian Model GlOOO strip chart recorder. Solutions were pumped through the electrode by means of a Present address, School of Pharmacy, University of Georgia, Athens, Ga. (1) W. J. Blaedel, Carter L. Olson and L. R. Sharma, ANAL. CHEM., 35, 2100 (1963). (2) W. J. Blaedel and Carter L. Olson, ibid., 36, 393 (1964). (3) W. J. Blaedel and L. N. Klatt, ibid., 38, 879 (1966). (4) L. N. Klatt and W. J. Blaedel, ibid., 39, 1065 (1967). ( 5 ) Zbid., 40, 512 (1968). (6) T. 0. Osterling and Carter L. Olson, ibid., 39, 1546 (1967). (7) Zbid., p. 1543. (8) L. R. Sharrna and J. Dutt, Ztzdian J. Chem., 6 , 593 (1958).
548
0
ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
i(
I"
i SALT BRIDGE
,TYGON
1
(*LASS
SLEEVE
FRIT
*ELECTRODE CONTACT
GLAssg TEFLON
SPACER
N SLEEVE
c
SOLUTION
IN
Figure 1. Electrode assembly
Harvard Apparatus Model 600-1200 peristaltic pump. By placing the pump downstream from the electrode, the solution can be pulled from the sample through a short piece of glass capillary tubing, minimizing the entrance of oxygen into the solution. Electrodes and Cell. The tubular carbon electrode ( W E ) is constructed by press-fitting a tubular piece of wax-impregnated spectroscopic-grade graphite into a specially designed holder (Figure 1). Several factors were considered in arriving at this cell design. The TCE is constructed so that it can be assembled easily and can be quickly replaced if the electrode surface becomes damaged or fouled. In addition it is important to obtain a smooth bore to and through the electrode so that a parabolic velocity profile with laminar flow is established before the solution enters the electrode which would not be disturbed by turbulence as the solution enters the electrode in order that derived equations be valid. This was accomplished conveniently by placing a solid plug in front
Table I. Reduction at TCE.
Depolarizer K3Fe(CN)6 &(I)
Concn 10-4 to I O 10-6M
Bi(II1)
10-3~ 10-6M 10-4~
6 ~
10-4~ 5 x 10-4~
10-3~
Pb(I1) Cd(I1)
5 X 10+M 1x 10-4~ 6.25 X 10F
1 x 10-4 1 a
x 10-3~
Media 1M KCl 1M KNO3 1 M KNOa 1M KNOa 1 M KNOI 1M HCI 1M HCl 1 M HCl 0.5MKC1 0.5MKCl 1 . 2 M NHdOH, 0 . 2 M NHaCl 1 . 2 MM I O H , 0 . 2 M “icl 1.2M NHiOH 0 . 2 M NHaCl
Ell2 us. SCE, mV
+212 185
Epiz =
il/c X lo4 MAIMf 0.2 5‘5
8.0
+245 ++300 305
7.9 7.8
-212 - 210 - 224 - 590 - 575 - 1057
15.7 15.3 16.0 12.4 12.3 10.9
- 1054
11.0
- 1038
10.8
8.0
= 0.24 in. by 0.081 in. i d . Vi = 4 . 0 ml/min Scan = -0.2 V/min
TCE
(upstream) of the carbon rod, a spacer disk behind the carbon rod, and then drilling the assembly after tightening so that a single perfectly-aligned bore is obtained. The Teflon (DuPont) spacer disk, carbon rod, and entry plug are all fabricated in advance as standard stock pieces so that the entire assembly of a new TCE, including drilling the bore, requires only a few minutes. The tight pressure fit obtained by screwing the assembly together prevents any solution leaks t o the outside of the electrode. The Teflon spacer disk should be thin t o minimize solution IR drop between the T C E and the reference electrode. The top Teflon piece of the electrode assembly functions as the electrolysis cell eliminating the need for any auxillary electrolysis cell. Two holes are drilled into the side of the top Teflon piece, one t o permit connection of the electrode to the pump, the other t o permit electrical contact to the outside of the electrode. The electrode is placed in a n upright position so that any gas bubbles which might occur in solution would not collect in the T C E or between the T C E and the reference electrode. A large low resistance saturated calomel electrode is used for the reference electrode. The carbon inserts are fabricated from wax impregnated spectroscopic grade carbon rod, 0.25-inch 0.d. Prior t o wax impregnation the carbon rod was cut into short lengths, about 1 cm long, and drilled with a n undersize hole running through the center t o facilitate degassing the center portion of the carbon rod. These pieces were then immersed in molten ceresin wax and degassed under a vacuum (9, IO). After assembly the final bore was drilled 0.081 inch in diameter. Reagents. All chemicals were reagent grade and were used without further purification. Solutions were prepared using deionized double-distilled water. Stock solutions of Cd2+and Pb2+were analyzed by titration with EDTA (11). RESULTS AND DISCUSSION
Initial measurements made on TCE’s prepared as just described yielded erratic results. Current voltage curves for well defined depolarizers such as ferricyanide in 1M KC1 ~
(9) V. F. Gaylor, A. L. Conrad and J. H. Lander], Indian J. Chem., 29, 224 (1957). (10) J. B. Morris and J. M. Schempf, ibid., 31, 286 (1959). (1 1) G. Schwarzenbach, “Complexometric Titrations,” Interscience Publishers Inc., New York, 1957.
showed varying degrees of irreversibility and yielded erratic diffusion limited currents which were much smaller than expected. What apparently happens is ,that upon drilling the final bore through the electrode, enough heat is generated to melt some of the ceresin wax forming a partially insulating wax film over the electrode surface. This problem is easily and satisfactorily solved by scouring the inside of the electrode with a pipe cleaner soaked in a ceresin soluble solvent such as ethyl acetate. After this treatment, current voltage curves for ferricyanide, for example, appear to be reversible and the experimental diffusion currents are very close to expected values. It was observed on some occasions after the electrode had become old and fouled that ethyl acetate alone did not seem t o recondition the electrode. In this case, a 1 :1 ethyl acetate-acetone mixture worked very well. When the electrode is to be employed in the potential region negative of -200 mV us. a SCE, a n additional pretreatment is necessary. This is the removal of oxygen from the T C E by holding the potential at - 800 mV in deaerated electrolyte until the current levels off. This usually requires 2 to 4 hours. This time can be shortened somewhat if the graphite is wax-impregnated under a nitrogen atmosphere. The voltage range over which the T C E can be used in a variety of supporting electrolytes is similar to that obtained at other carbon electrodes. Four metal ions were studied to determine the suitability of the T C E for the reduction of metal ions. The deposition potentials for Ag(I), Bi(III), Pb(II), and Cd(I1) are dependent upon the history of the electrode and the concentration of the depolarizer. When 0.1 m M Ag(1) in 1M KNOI is deposited on a new TCE, a long drawn out wave is observed with Elizof +0.070 V us. SCE. The deposition potential of subsequent scans is dependent on the length of time the electrode was stripped. If the stripping time is short, the deposition potential moves in a n anodic direction and the slope of the current voltage curve becomes steeper. If the electrode is stripped a t +0.400 V for a time not exceeding 2 minutes (time required for background current to return to initial value), then a reproducible deposition wave (Ellz = 0.245 =t0.005 V us. SCE and Esi4 - E1,4= 0.30 V) was obtained. Longer stripping times resulted in a progressive cathodic shift of the deposition wave. ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
549
Table 11. Oxidation a t TCEQ
a
TCE
Depolarizer
Media
K4Fe(CNh o-Tolidine Uric acid
0.5MKC1 1MHC1 0 . 5M Acetate pH = 6 . 0
=
4.8 8.9 9.0
65 28 35
+210 $605 $330
0.24 in. by 0.081 in. i.d.
Vf = 4.0 ml/min
* Average results for three runs at each of the following concentrations, 1 X
10-4M, 2 X 10-4M, and 4 X 10-4M.
Table 111. Mixture of Pb(I1) and Cd(II1) 40-
AIM x Mixture 0.99 x 1.08 X 0.99 X 2.16 X 1.98 X 1.08 x
601 I
I (ANODIC)
~o-~MP~(II)
10-4MCd(II) 10-4M Pb(I1) 10-4M Cd(I1) 10-4M Pb(I1) 10-4M Cd(I1)
104
Pb(I1)
Cd(I1)
9.30
9.63
9.20
9.65
9.30
9.50
Vf 3.0 ml/min, scan = - 100 mV/rnin Background = 1.2M NH40H, 0.2M NH4Cl XI = 0.44 cm = length calculated from experimental limiting current data Ciu equation by Blaedel and Klatt (3).
Figure 2. Metal ion deposition and stripping Scan rate-500 mV/min V , = 4.0 ml/min TCE = 0.24 in. long and i.d.-O.O81 in. A . Ag(1) in 1 M KN03 E. Bi(I1) in 1 M HCI C. Bi(1II) in 0.5M KCI D . Cd(I1) in 1.2M NHaOH, 0.2M NH4CI
The electrode could be returned to its initial condition by stripping at f1.0 V for 30 minutes followed by cleaning with ethyl acetate. The current voltage curves were scanned at a rate of 0.200 V per minute. The limiting current for Ag(1) reduction in 1M KNO3 is reproducible after the first deposition and stripping and is presented in Table I. Deposition of O.lmM Bi(II1) in 1M HC1 o n a new T C E gave a long drawn out wave with E,,z = -0.260 f 0,010 V cs. SCE. For subsequent scans it was found that if the stripping time at -0.100 V did not exceed 2 minutes, a reproducible wave (Eliz = -0.210 V us. SCE and E J j 4- E,,, = 0.021 V) was obtained. Longer stripping times and more anodic voltages resulted in progressive cathodic shifts of the deposition potential. The electrode was returned to the initial state if stripped a t f0.100 V US. SCE for 2 minutes. The limiting current for Bi(II1) was reproducible, independent of deposition potential, and is presented in Table I. Deposition of O.lmM Pb(I1) in 0.5M KC1 or 1 M K N 0 3 gives reproducible current voltage curves (E,/z = -0.575 =k 0.005 V us. SCE and E3/4-E1j4 = 0.012 V) o n a new TCE. Subsequent scans are identical to those on the new electrode if the electrode is stripped a t -0.400 V us. SCE for 1 minute. This indicates efficient stripping of the deposited lead. Lead depositions show a peak current prior to the diffusion current plateau. The limiting current for Pb(I1) in 0.5M KCl is reproducible and is shown in Table I. 550
ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
Deposition of O.lmM Cd(I1) in 1.2M N H 4 0 H , 0.2M NHICl occurs a t a new TCE with a Ellzof -1.054 i 0.005 V us. SCE. Subsequent deposition waves were identical if the electrode was returned to -0.800 V for 1 minute of stripping. The limiting currents are presented in Table I. Typical cyclic voltammograms showing the deposition wave and subsequent stripping wave for the four ions investigated are illustrated in Figure 2. The suitability of the TCE for electrochemical oxidations was tested and the results for three depolarizers are given in Table 11. Each of the three depolarizers tested were measured at 0.1, 0.2, and 0.4 m M concentrations. Three current voltage curves were run for each of the three depolarizers at each concentration. The shape of the waves was reproducible over the rather narrow concentration range with no shift in run, and the limiting currents were very reproducible. Although one can measure accurately the geometrical area of the electrode during fabrication, it is not necessarily true that the actual effective area o n which electrolysis can take place will be the same. The experimental length of the electrodes is calculated using the limiting current equation derived by Blaedel and Klatt (3) using experimental diffusion currents and appropriate values for the other terms in the equation. In general, the “experimental” length for a used electrode is about 2 5 z shorter than the actual measured length of the electrode. A very brief study using a mixture of Pb(I1) and Cd(I1) ions to demonstrate the possibility of determining mixtures a t the TCE was conducted, the results of which are shown in Table 111. From the brief study conducted, it would appear that in this instance, mixtures can be determined. In conclusion, it was found that the TCE has a relatively long voltage range over which oxidations and reductions can be run. Some metal ion reductions are very sensitive to the
electrode history, while the soluble product oxidations run were essentially independent of electrode history, a t least as long as the electrode had been previously treated in standard manner by ethyl acetate pretreatment and no electrode fouling compound had been electrolyzed The limiting current reproducibil ty is very good as illustrated in Tables I and 11. The electroie has been applied to continuous measurement of monamine oxidase enzyme activity where it has been shown
to yield highly reproducible results over relatively long periods of time (12).
RECEIVED for review September 30, 1970. Accepted January 29, 1970. William D. Mason Was a fellow Of the American Foundation for H-mmaceutical E h ~ a t i o n . (12) W. D. M~~~~ and Carter L. 0lson, A ( 1970).
~ cHEL1,, ~ 42, ~ 488,
Optical Ring-Disk Electrode System James E. McClure Research Seroices Department, American Cyanamid Co., Stamford, Conn.
THEROTATED DISK electrode and rotated ring-disk electrode systems and their applications are well known ( I , 2). This paper describes a n extension of the above two systems. The new system is a n optical ring-disk electrode (ORDE) in which a n optically transparent ring surrounds the disk electrode. This arrangement allows recording of optical spectra of unstable materials generated at the disk electrode. The system has been tested by recording the spectra of a stable species (IrC16*-) and a n unstable species (monocation of triphenylamine) which were generated electrochemically.
n MOTOR
LIGHT DETECTOR WORKING ELECTRODE CONTACT BEARINGS
EXPERIMENTAL
Apparatus. The arrangement used for the O R D E studies is shown in Figure 1. The shaft of the rotating electrode assembly was machined from a length of 6i8-in. o.d., 3/16-in. i d . , stainless steel tubing (final 0.d. was 9116 in.). The bearing assembly and contacts for electrical connection to the working electrode were duplicated from a rotated disk electrode system which was designed by Stonehart (3). The shaft was driven through a pulley and belt arrangement with a n Electrocraft Model E500-M electric motor and speed control system. The Electrocraft E500-M is capable of speeds u p to 5000 rpm. A 3 to 1 pulley (PIC Design Corp.) ratio was used in the connection between the drive motor and rotating shaft. The optical ring-disk electrode was constructed by positioning several hundred light wires (American Optical Company) around a platinum rod (l/*-in. diameter) in a hole -5/32 in. drilled into a trifluorochloroethylene cylinder and then sealing with epoxy resin (see Figure lb). After the resin was cured, the face of the O R D E was sanded and then polished t o a mirror finish by conventional resinographic techniques (0.1 p alumina was used for the final polish). The thickness of the optically transparent ring, which is shown in the crosshatched area of Figure lb, is -0.5 mm. The optical ring is immediately adjacent t o the platinum disk electrode. Eccentricity of the trifluorochloroethylene holder was ~t0.0005in. A similar value was determined for the eccentricity of the platinum disk. The light wires surrounding the platinum electrode extend up through the hollow stainless steel shaft and terminate -’/* in. from a photomultiplier tube. Light from a Heath EU-700 monochromator was focused on the O R D E through a window in the bottom of the electrolysis cell with a l/s-in. 0.d. light guide. A Pacific Photometric (1) R. N. Adams, “Electrochemistry at Solid Electrodes,” Marcel Dekker, Inc., New York, 1969, p 67. (2) A. C . Riddiford in “Advances in Electrochemistry and Electro-
chemical Engineering,” P. Delahay, ed., Vol. 4, Interscience, New York, 1965, p 47. (3) P. Stonehart, Anal. Chirn. A m . , 37, 350 (1967).
WORKING ELECTRODE
\ CHROMATOR
WIRES
(ai
(b)
Figure 1. Arrangement for optical ring-disk electrode studies
Model 15 Recording Photometer with shielded 1P21 photomultiplier tube was used for light detection. A Wavetek Model 114 signal generator and Model 61RS Wenking potentiostat were used to provide a square-wave excitation signal t o the working electrode. A Princeton Applied Research Model HR-8 lock-in amplifier was used for signal detection in the light measuring circuit. The cell (3l/*-in. diameter) was blackened to exclude light. A platinum counter electrode was isolated from th6 working electrode compartment with a fine porosity frit. Reagents. Na*IrCls and NaJrCI, were obtained from A. D. Mackay and K & K Laboratories, respectively. Eastman triphenylamine and Mallinckrodt nanograde acetonitrile were used without further purification. Procedure. The O R D E system was tested with the IrC163IrC16*- e- system. IrC162- was generated from 4 m M IrC163- (in 0.2M HC104) a t the disk electrode with a square-wave excitation signal. The light attenuation due t o the periodic appearance of IrCI6*- in the light path at the optical ring was measured with a photometer and lock-in amplifier. The spectrum of the monocation of triphenylamine
+
ANALYTICAL CHEMISTRY, VOL. 42, NO. 4, APRIL 1970
551