Fig. 1. Bottom, microelectrode tip has infinite resistance and has not been bared; top, microelectrode tip has been bared by plunging it into a hard rubber stopper, making it functional (scales, 10u)
METAL microelectrodes suitable for recording externally from single cells have been reported by various authors (1-7). None of them meets our requirements for a rugged and reliable micoelectrode which can be heat- or gas-sterilized for use in human neurosurgical procedures.
The tungsten microelectrode of Hubel(3) is the hardest and stiffest, but it, as well as all the other more fragile ones, is functionally destroyed by autoclaving or by ethylene oxide gas sterilization. Insulation such as 'Insl-X E33' clear, or vinyl lacquer such as Stoner-Mudge 'S-986s', yields a desirable microtip resistance in the low megohm range by virtue of shrinkage and a withdrawal from the point. But this kind of insulation is disrupted by heat, ehtylene oxide and mechanical abrasion by dura matter, blood clots or bone.
Electrical insulating varnishes generally present two problems in their application to microelectrode technology. They may be brittle and the coat easily ruptured during manipulation or bendin, as has been our experience with 'Formvar'. Furthermore, these varnishes completely cover the microtip when applied in a sufficient number of coats to insulate the shank well.
The insulating varnish problem has been solved by using 'Isonel 31' (Schenectady Varnish Co., Schenectady, New York). It bonds well and is cured by heat to a tough and flexible coat. Although it can be thinned with xylene, we have used it without modification. (Other insulating varnishes have been tried which are adequately flexible, but they require many more coats and sometimes more difficult curing procedures than does 'Lionel 31'. )
Quarter millimetre (10 mil) straight tungsten wire (Sylvania Electric Co., Towanda, Pennsylvania) is sharpened electrolytically by the Hubel technique(3). The wire is dipped in a saturated aqueous solution of potassium nitrite with about 6 V alternating current and sharpened with a pencil-like taper to a 1u tip. After washing in a detergent solution, it is cleaned in trichloroethylene placed in an ultrasonically agitated bath. Then the needle is slowly withdrawn from the varnish, point either up or down, at about 4 cm per min to avoid beading, as in the Wagman technique(8). We use an 'Isonel 31' bath and cure the insulation at 160° C for 0.5 h after each coat. Seven coats are required. (We now have machines which both tip and dip several dozen wires automatically.) The electrode is examined for the integrity of the shank's insulation by placing it in a 0.9 per cent sodium chloride solution with the tip just above the surface and applying a negative potential of 50V. Bubbles betray any leakage. Bending the electrode or even knotting it does not disrupt the insulation, although pinching the shank with a haemostat will.
The electrode tip immersed in the saline solution exhibits practically infinite resistance - more than a teraohm (1*10^12 ohms) as found by applying 0.10 V, yielding a current of less than 0.1 p.amp measured with a Hewlett-Packard microammeter (model 425A) Infinite resitance is also indicated by the introduction of a parallel 100 megohm resistance when the electrode is connected to the input of an electrometer cathode follower. It is necessary to break dwn the insulation at the tip and reduce the resistance to a value useful for recording single units.
One solution to this second problem is to apply a negative potential of 2-20 V for some minutes, producing bubling at the tip in saline solution. Another is to discharge a capacitor of several microfarads which has been charged with 25-250 V. It is applied between the elctrode and the saoline solution which is abrely in contact with the tip. Once the tip is open, bubbling starts at about 2 V. The test bubbling may occur as far as 30u back from the tip and the whole area of the tip insulation may be shucked. Nevertheless, a resistance of 100-1000 megaohms (a gigaohm) indicates that the electrode will often prove useful with a cathode follower despite the microscopic indications of a diffuse conduction area. (The capacitance and/or polarization of the tip results in a much lower impedance at frequencies above 5 c/s.) the obvious need of a more localized conduction area encouraged us to attempt to reduce the resistance to useful values mechanically. Mechanicaly methods have proved preferable to the electrical ones in providing better unit isolation.
Fig. 1. Bottom, microelectrode tip has infinite
resistance and has not been bared; top, microelectrode tip has
been bared by plunging it into a hard rubber stopper, making it
functional (scales, 10u)
An electrode tip is pushed several times againts an ordinary hard-rubber bottle stopper. A very sharp electrode may require only one or two penetrations to reduce its resistance to the 100 megohm range, whereas a more blunt tip may require six or more contacts before this resitance is achieved. If the resitance is much higher, no responses are obtained, and if it is much lower, single units cannot be isolated electrically in the brain. Under the microscope the insulation appears to bepolished off as much as 15u back from the tip, or the point may appear fuzzy because of the abraded insulation (which can be wiped off by drawing the electrode backwards through a cloth or soft paper tissue). Photomicrographs of the tip are shown in Fig. 1, before and after the final preparation of an electrode with a rubber stopper. The function, however, can be well predicted, not by appearance but by resistance. Fig. 2 is an example of unit responses from the lateral eniculate body of a rabbit. If the electrode is not applied perpendicularly to the hard rubber surface, the tip may bend into a small hook which can be seen under the microscope, but it does not seem to affect the isolation of single units so long as the resitance is optimum. In a few instances, the tip was broken off giving resistances in the low megohm range, but surprisingly the function was still good. An explanation could be that which has been offered for relatively large tipped electrodes, such as those of Wolbarsht et al.(5) and Gesteland et al.(6), hypothesizing 'hot spots' in the fashion of a galena crystal rectifier.
Stainless steel microelectrodes etched by the method of John D. Green(4) do not isolate units well when prepared with the new insulating varnish. In part, the steel is too soft to stand the mechanical treatment required to bring the resistance down to useful levels.
Our microelectrode can be prepared in a variety of ways besides using a rubber stopper. For example, by dropping it on the floor, point down, several times from a height of more than a metre, the resistance is reduced to several hundred megohms and the easy isolation of single units is made possible. It can also be prepared by pushing the tip against a smooth, hard surface such as glass. This flattens the tip to about 30u diameter, but the geometry does no appear to affect the performance.
Fig. 2. Single units recorded from the lateral
geniculate body of the rabbit. Time, 10 msec. Amplitude, 1/2 mV.
Perhaps the most convenient method of reducing the electrode resistance from infinity to a useful value during a physiological experiment is to push it against the skull or other bone of the animal until the shank is well bowed. Unlike other microelectrodes, generally no harm is done to its slectivity, although the development of a microscopic hook can mechanically disrupt some fibres on withdrawal. Entrance can be made through the intact dura, obviating the use of a special chamber to reduce pulsation.
In fact, we have been able to isolate singleunits after pluging the electrode through the skin and dura mater of a rabbit to reach the brain. This treatment reduced the infinite resistance to the required value. Hence, microelectrodee recordings can be made in animals which have been previously prepared only by a crainiotomy and the skin replaced and healed. Prof. Irving H. Wagman has used these electrodes to record single units in cat and in human skeletal muscle through the intact skin. he believes that these microelectrodes may prove of value for single unit electromyography in the clinic. Dr. Hubert Goodwin has used them in recording units of the lateral geniculate body of rabits during ultrasonic irradiation of the visual pathways. However, these electrodes are not very effective in the frog brain.
Just as with other microelectrodes, one sometimes hears a high-to-low frequency burst as the tip is pushed through the brain. This is generally believed to signal the disruption of a cell by the compression caused by the advance of the electrode. Also, as with other microelectrodes, the tip can approach or withdraw from a cell and the amplitude of the spike increases or decreases. There seems to be no more damage to the nerve cells than with other miroelectrodes which have finer tips.
The ruggedness and reliability of this electrode now make practicable large arrays of microelectrodes for recording from brain, muscle and nerve. They also make practicable use of microelectrodes in student laboratories where only rugged equipment survives.
In terms of our orginal objective, most important of all, has been the autoclaving or gas sterilizing of the electrode without any apparent effect on its appearance or function. This makes it of special value in human neurosurgical procedures. The impression is gained that is difficult to make a poor electrode by this method or to damage it functionally by ordinary use.
This work was performed while I was a Fellow of the John Simon Guggenheim Memorial Foundation, on leave from the University of California, San Francisco (Department of Neurosurgery) and Berkeley (Department of Optometry). It was aided by a grant from the U.S. National Science Foundation. I thank Prof. Irving H. Wagman for his advice and Mrs. Nancy Fletcher for her assitance.
ELWIN MARG
College de France,
Laboratoire de Neurophysiologie Generale,
Station de l'Institut Marey,
4 Avenue Gordon-Bennett,
Paris 16e.
(1) Grundfest, H., Sengstaken, R.W., Oettinger,
W.H., and Gurry, R.W., Rev. Sci. Instr., 21, 300 (1950).
(2) Gray, J.A.B., and Svaetachin, G., Acta Physiol. Scand., 24,
278 (1951).
(3) Hubel, D.H., Science, 125, 549 (1957).
(4) Green, J.D., Nature, 182, 962 (1958).
(5) Wolbarsht, M.L., MacNichol, jun., E.F., and Wagner, H.G.,
Science, 132, 1309 (1960).
(6) Gesteland, R.C., Howland, B., Lettvin, J.Y., and Pitts, W.H.,
Proc. Inst. Rad. Eng., 47, 1156 (1959).
(7) Pierce, D.S., and Wagman, I.H., J. App. Physiol. (in the
press).
(8) Wagman, I.H. (personal communication).