JOURNAL OF THE OPTICAL SOCIETY OF AMERICA

Accessory Equipment for Spectrochemical Analysis*

J. L. SAUNDERSON AND V. J. CALDECOURT

The Dow Chemical Company, Midland, Michigan

INTRODUCTION

THIS paper describes miscellaneous accessoryT equipment used in this laboratory for thespectrochemical analysis of magnesium alloys.Consideration of the purpose of the equipmenthere described shows that rigid control of the ex-perimental conditions during the analysis as wellas convenience and efficiency of operation dic-tates the design of equipment. In the relativelyfew cases where convenience is not compatiblewith duplication of conditions, the latter must befavored. Much of the equipment in use in thislaboratory on the routine analytical work hasbeen described by others. The spectra are photo-graphed with a Bausch and Lomb MediumQuartz spectrograph, the spark source follows thedesign of Feussner,1 the arc is similar to that ofDuffendack and Wolf,2 the microphotometer is ofthe Michigan design,3 and the calculating board4

is of the conventional type. Apparatus for con-trolling the development and drying of the plates

FIG. 1. Automatic arc starter. The starter is shown abovethe dotted line, connected to a conventional arc sourcebelow.

* Paper presented at the Twenty-Eighth Annual Meetingof the Optical Society of America, Pittsburgh, Pennsyl-vania, October 7-9, 1943.

1 0. Feussner, Arch. Eisenhfittenwes 6, 551 (1932-33);also 0. Feussner, U. S. Patent 1,971,215. See also H. B.Vincent and R. A. Sawyer, J. App. Phys. 8, 163 (1937).

2 Duffendack and Thompson, Proc. A. S. T. M. 36,Part 11 (1936).

3 H. B. Vincent and R. A. Sawyer, J. Opt. Soc. Am. 31,639 (1941).

4 J. S. Owens, Metals and Alloys (January, 1938).

is similar in principle to that now commerciallyavailable.

The equipment described below may best betermed "accessory,"' in the sense that its use isnot necessarily essential, but rather that it offersgreater convenience of operation and control ofconditions. The value of such equipment isprobably best emphasized by the fact that duringthe last two years the analytical load carried bythis laboratory has increased by a factor of 300percent, while the personnel engaged in theroutine work has increased only 50 percent. The"accessory" equipment has been an importantfactor in the increased efficiency and accuracy ofthe routine spectrochemical analysis of mag-nesium alloys.

AUTOMATIC ARC STARTER

Figure 1 shows a wiring diagram of a con-ventional high voltage a.c. arc equipped with anautomatic arc starter. Electrical ignition systemshave been described by Pfeilsticker5 and others,6

and the starter part of Fig. 1 is not essentiallydifferent from these. A luminous sign transformerof standard rating, 15,000 v at 240 v-amp., isused to power the exciter circuit. When the con-tactor in the primary of the arc power trans-former is closed, the starter circuit is excitedsimultaneously. Coupling between the startercircuit and the arc secondary circuit is obtainedby means of a Tesla type transformer. When thestarter circuit is excited, a high frequency oscil-latory discharge takes place between the elec-trodes of the analytical gap. The arc current willfollow the starter current across the gap, andwhen the electrodes become heated sufficientlythe arc will be self-maintaining. For magnesiumalloy electrodes, the arc will maintain itself aftera tenth of a second. For iron electrodes, thestarter must be run several seconds. The coil ofthe time delay relay is connected to the low

5 K. Z. Pfeilsticker, Zeits. f. Elektrochemie 43, 719(1937).

6 H. Hemmendinger, J. Opt. Soc. Am. 32, 149 (1942);F. G. Brockman and F. P. Hochgesang, Ind. Eng. Chem.Anal. Ed. 14, 796 (1942).

116

VOLUME 34, NUMBER 2 FEBRUARY. 1944

EQUIPMENT FOR SPECTROCHEMICAL ANALYSIS

voltage side of the secondary of the arc powertransformer, so that the arc current flows throughthe coil. When the arc starts, the time delay relaycloses after the delay period. The second relay isthen activated, opening the primary of theexciter circuit and introducing a five-ohm powerresistor in parallel with the time delay relay coil.The starter thus goes off automatically afterabout one-half second, and the time delay relayremains closed as long as the arc continues torun. If the arc should go out, the starter comes onimmediately until the arc is re-ignited. Whenusing electrodes for which the arc is more difficultto ignite than for magnesium alloy, a longerdelay period should be used.

A time delay relay with normally closed con-tacts could be used, and if the contacts had arating sufficient to carry the current for theluminous sign transformer, the second relaywould not be necessary. The five-ohm resistorwas inserted in parallel with the coil of the timedelay relay in order to reduce the current throughthe coil after the relay has closed. This procedurewas considered advisable in this case because theparticular time delay relay used required a muchlarger current to close the relay than was neces-sary to keep it closed. The Tesla type transformerwas constructed from two concentric solenoids.The outer coil contained 30 turns of number 24wire wound with a 4-inch diameter, and the innercoil had a diameter of 3' inches and consisted of75 turns of number 12 wire; both coils had 8/

spacing between turns. The 0.01-luf filter con-denser is sufficient to prevent any high voltagefrom damaging the power transformer. It shouldhave sufficient insulation to withstand 5000volts. The condenser in the exciter circuit wastested at 15,000 volts. Both condensers weremade by piling up alternate sheets of copper andglass in a pan containing molten wax.

STABILIZED MICROPHOTOMETERPOWER SUPPLY

The power supply here described is one whichhas been used with considerable success forseveral years. It is simple in design, gives trouble-free service, and is easily adaptable to anymicrophotometer requiring power for a lamp atlow voltage and large current. The principles in-volved in the circuit are not new, and a descrip-

FIG. 2. Stabilized microphotometer power supply. Ad-justable resistors from left to right are RI, R2, R3.

tion is justified here only by the belief that manyworkers engaged in spectrochemical analysiswould welcome publication of a complete unit.The system was first suggested to the authors byDr. J. L. Lawson at the University of Michigan.

The unit is stabilized by means of the well-known "floating" battery principle (Fig. 2). The110-volt 60-cycle current is reduced in voltage bymeans of the transformer and R1 and rectified sothat the d.c. voltage across the first battery isjust sufficient to maintain a small charging cur-rent. The voltage is reduced at the second batteryby the resistor R2 , and again from the secondbattery voltage to the voltage supplied the lampby the resistor R3. The stabilization against linevoltage fluctuations is much better when two setsof batteries are used at two voltages, as shown,than when the same number of batteries are allused at the same voltage. Ordinary 6-volt auto-mobile storage batteries have been used in thislaboratory. The first set of batteries consists of asmany batteries in series as are required to obtainthe necessary voltage, while the second set con-tains two sets of batteries arranged in parallel forincreased capacity for stabilization. A chargingcurrent of about 0.1 amp. through each batteryhas been found to be satisfactory. The secondammeter, therefore, shows about 0.2 amp.

The resistors should be of the sliding contacttype which is clamped in place rather than therheostat type which does not make sufficientlygood contact for this purpose. The additionalconvenience of the latter type is not necessarysince the unit should require little or no adjust-ment after it has been set up properly for opera-tion. The service required by the unit is limitedto checking the water in the batteries every sixweeks.

It was the original intention when this unit wasconstructed to substitute a current regulator tubefor the resistance R to minimize slow drift

117

J. L. SAUNDERSON AND V. J. CALDECOURT

* FIG. 3. Electrode clamps particularly suitable for mag-nesium alloys. The electrode spacing mechanism can beswung forward so that the glass spacer is between the twoclamps.

effects. The system has been so satisfactory,however, that this was never done. Within twentyminutes, the average time required to measurethe line intensities on a plate, the drift in trans-parency of the clear plate rarely exceeds 2 mmand usually is less than 1 mm on a 500-mm scale.

ELECTRODE CLAMPS

The electrode clamps shown in Fig. 3 aredesigned to hold rectangular or cylindrical elec-trodes very tightly and always in exactly thesame position. The clamps are heavy enough tocarry safely the currents encountered with thearcs and sparks used in spectrochemical workThe clamps require no artificial cooling whenused with a spark, but cooling may be desirablefor continuous use with an arc. Artificial coolingcan be accomplished either by soldering a smallcopper pipe on the outside of each of the jaws, orby drilling a hole inside the jaws to allow thecirculation of water or air. Several sets of clampsof this design have been constructed for use inroutine work. The advantages of the designinclude strength of construction, rigidity andlack of adjustable parts, large thermal capacity,positive pressure for gripping electrodes, andaccurate alignment of electrodes. An experiencedoperator can load the clamps with two electrodesin 10 seconds.

The clamp jaws are constructed of 4" steel.The two jaws next to the Bakelite post are bolted

to the post, electrical connections being madethrough the bolts. The front jaws are held by thehinge pins, which allow the jaws to rotate enoughto get electrodes in and out. The clamping boltspass freely through the front jaws, and arethreaded into the rear jaws. A removable key orwrench is provided to tighten the bolts. Theelectrode spacing mechanism swings parallel tothe jaws to space the electrodes, and can beswung out of the way for the discharge. TheV-groove is intended to be narrow enough incomparison to the diameter of the electrodes sothat the edges grip the electrodes firmly when theclamps are tightened.

The lack of adjustments on the clamps is con-sidered an advantage for routine work. Onceproper alignment with respect to the spectrographhas been found, the clamps require no furtheradjustment.

ELECTRODE SHAPER

A simple and straightforward design was fol-lowed in constructing the apparatus used in thislaboratory for shaping the ends of cylindricalelectrodes in preparation for the discharge. Ai-hp motor with built-in automatic brake wasfitted with a Jacobs keyless chuck mounted onthe motor shaft. The motor was mounted on aniron base together with a compound tool restfitted with a cutting bit. A three-phase motor was

FIG. 4. Intensity control sector, having a continuouslyvariable transmission from 50 percent to 100 percent.

118

EQUIPMENT FOR SPECTROCHEMICAL ANALYSIS

used because of the frequent starting required inroutine operation. The cutting operation is iden-tical to that of a lathe, but the frequent startingand stopping can be done more conveniently thanwith the usual lathe. The motor starts nearlyinstantaneously, and stops in about one second.Construction of the unit is simplified in that nobearing surfaces are required. The apparatus hasbeen in use for about one year without showingany noticeable effect of wear. Mr. W. M. Weddellwas partly responsible for the design and con-struction of this apparatus.

INTENSITY CONTROL VANE

The vane shown in Fig. 4 was constructed toprovide a continuously variable control of theintensity of the photographed spectra. The vaneis used to maintain the density of a standard lineat some constant value when a difference in thespeed of the photographic plates occurs. The vanehas a continuously variable transmission between50 percent and 100 percent, and affords a methodof intensity control which does not depend uponchanging source distance, exposure time, or theuse of filters and screens. The vane is rotated at 1revolution per second to avoid any possibility ofa stroboscopic effect when used with the inter-rupted spark. Although calibration was carriedout under the actual operating conditions usingthe vane in order to nullify any possible effectfrom the intermittancy of the exposure, noevidence for any intermittancy effect in this casehas yet been discovered.

A layout of the vane is shown in the upperright-hand corner of Fig. 4. To adjust the trans-mission factor, the vane can be rotated in thecollar mounted on the end of the shaft connectedto the 60-r.p.m. motor. The apparatus should beused such that the shaft to the motor is parallelto but about two inches lower than the opticalaxis of the spectrograph.

AUTOMATIC EXPOSURE APPARATUS

A time-saving system has been constructed tominimize the labor required to operate thespectrograph used on the routine analysis ofmagnesium alloys. The actual wiring diagram ofthe system appears to be so closely connectedwith the particular requirements of this labora-

FIG. 5. Block diagram showing operating functions of thetiming system for automatic spectrograph.

tory that a detailed description would not bejustified. The general operation of the system canbest be followed by reference to the block diagramof Fig. 5. Briefly, the timing system will make aspark exposure followed by an arc exposure, andmove the plate down the proper distance aftereach. It has been found desirable to obtain bothspark and arc exposures on magnesium alloys tocover adequately the analytical range required.

After the operator has prepared two electrodesand has loaded them into the clamps, he pushesa "start" button, and then returns to prepare twomore electrodes from the next sample to beanalyzed. The spark discharge begins immedi-ately, as well as the first timer. A magneticshutter is closed during the first part of the sparkdischarge, and is opened when the first timerreaches the time interval for which it is set. Thesecond timer starts when the shutter opens, andregulates the length of the spark exposure. Atthe end of the spark exposure, the spark goes off,the shutter closes, a high voltage relay changesthe leads from the electrode clamps over to thearc, a lens comes into position for the arc ex-posure, the arc is turned on and ignited by theautomatic starter, the clutch mechanism con-trolling the plate motion is activated, and boththe arc pre-exposure timer and the plate motiontimer are started. After several seconds, the platemotion timer releases the clutch mechanism andthe arc pre-exposure timer opens the shutter andstarts the arc exposure timer. After the arcexposure, the arc is turned off, the high voltage

119

ANNOUNCEMENT

relay goes back to the spark position, the lens isremoved, the shutter closes, the plate motionclutch is again activated, and the plate motiontimer starts. When the plate has been moveddown the correct distance, the timer again re-leases the clutch. Thus the system is entirelyreset ready to repeat the cycle on the nextsample. Provision is also made with appropriateswitches to permit running only sparks or onlyarcs. The complete cycle, under the present

operating conditions, requires one minute and 20seconds. An operator is expected to prepare thenext set of electrodes within this interval.

The timer circuit was constructed from fiveGeneral Electric Type TSA-10 time switches,several auxiliary relays and solenoids, and amagnetic clutch with a driving motor. Thespecially constructed parts included the highvoltage relay, the lens mechanism, and the mag-netic shutter.

120