AIDS FOR THE ANALYST

Nonbumping Digestion Heater

Andrejs Steinbergr, Division 01 Plant Industry, Commonwealth Scientific and Industrial Research Organization, Canberra, A. C. T., Australia.

HEN solid particles are boiled for a prolonged period with a w liquid on the normal type of electric heater (as in Kjeldahl digestion for determination of nitrogen in soils), bumping usually occurs, accompanied frequently by loss of some of the contents from the container. The liquid enclosed among the heavy solid particles is prevented from free movement and when heating is from beneath, this leads to local superheating and consequent bumping. n 11

A ASBESTOS CEMENT OUTER BODY

B PORCELAIN SUPPORT FOR HEATI-1

stages of the digestion has been reduced by this improved heating arrangement and creeping of the charring material into the neck of the flaak eliminated.

Salt Bridges of Porous Glass and Ion Exchange Membranes

W . N. Carson, Jr., C. E. Michelron, and Karl Koyama, Hanford Atomic Products Operation, General Electric Co., Richland, Wash.

LTHOUGH salt bridges are widely used in the analytical labora- A tory in polarography, titrimetry, and pH measurement, and their use is increasing as more electrometric methods of analysis are employed, the present forms and materials leave much to be desired in mechanical stability, small solution flow, and low elec- trical resistance. Work in this laboratory has shown that porous glass and ion exchange resin membranes can be used to make salt bridges that are superior in most respects to conventional bridges.

Porous glass bridges are made from Corning Glass Works glass 7930, which is leached but unfired Vycor glass. The glass is available in a variety of forms, such as sheet, tubing, and rod. It cannot be handled easily by glassblowing techniques (This glass burns, when touched to the lips or other moist skin areas, and will cause severe dehydration and excoriation in the manner of a burn.), as it shatters when placed in a flame. Special shapes, closed tubes, etc., must be fabricated before leaching; these can be supplied by Corning. Some properties of Corning porous glass 7930 are (Corning Glass Works technical brochure): void space (dry), 28% of volume; average pore diameter, 4 mp; composition: 96% Si02, 3% Bz08, 0.4% RnOs, traces of alkalies and arsenic; flow through 2-mm. thickness, 0.00065 ml. of water per sq. em. of area per atmosphere of pressure per hour.

C NI-CHROME HEATING ELEMENT

D PORCELAIN SUPPORT FOR R A S K

E ASBESTOS CEMENT BOTTOM PLATE

F ANGLE IRON FRAME

G BUSBARS

0 1 2 - INCHES

Figure 1 The large void space permits a large amount of electrolyte to be held in the glass, and thus give low, electrical resistance to the bridge, while the small pore size prevents more than a very small flow of solution. As the glass is electrically inert, the junction potential of a bridge is determined by the electrolyte used in the bridge. Nonaqueous solutions may be used as bridge electrolytes for nonaqueous systems. The bridges are suitable for use at elevated temperatures.

Fluoride and caustic attack the glass, as would be expected for finely divided silica. The high specific surface gives rise to ad-

A new type of digestion heater eliminates this effect. The main heating is directed toward the sides of the flask, with only slight heating from the bottom. Accordingly, the bulk of the liquid becomes heated, and readily circulates.

Two types of heater have been used: one with cylindrically shaped walls for use with Kjeldahl flasks up to 30-ml. volume or for test tubes, the other with conical shaped walls for Kjeldahl flasks up to 100-ml. volume. Details of the heaters are shown in Figure 1.

The heating element, C, is wound twice around the inside of a fireclay support, B, commencing 1.0 to 1.5 cm. from the bottom.

heater. removing i t and placing the flask directly on the concave bottom of the heater, i t is possible to adjust the heating of the flask to the desired zone.

An assembly of six heaters is mounted on an asbestos-cement The asbestos-

cement outer cover, A , for each of the heaters is fastened to the frame by means of wedge-tailed clips. Thus all three parts (outer cover, heater, and base plate) are easily firmly assembled or dis-

All Nichrome wires of the heaters are connected to bus bars, G, located under the base plate. These run along the length of the assembly under the heaters, their ends resting freely in slots in asbestos-cement sheets which are attached to the inside faces of

by a safety casing clipped to the asbestos-cement side sheets. Thus the bus bars are readily accessible.

In several hundred nitrogen determinations, with a wide range of soil types, no bumping whatsoever occurred during the diges- tions. Because of more efficient and uninterrupted heating, the nitrogen analyses gave results in many cases up to 10% higher than those obtained with the usual type of digestion stand. In nitrogen determinations on plant materials, frothing in early

A

POROUS VYCOR

1 I/;

f

I mantled for repairs. I/;

T

The flask is placed on a removable concave support, D, inside the By placing asbestos rings underneath the support or by

-

base plate, E, lying in an angle-iron frame, F. B

PLUG (SOLID)

the end panels of the heater support. The bus bars are protected c

GLASS 8 m m OD.

Figure 1. Porous glass salt bridges

472

V O L U M E 27, NO. 3, M A R C H 1 9 5 5 473

large area can be made in a similar fashion from widemouthed bottles.

The electrical resistance of bridges made from the new mate- rials was measured, employing saturated potassium chloride &B

electrolyte, by coupling the bridge in series with two mercury pool electrodes by means of potassium chloride solutions of known resistance. A 60-cycle alternating current Wheatstone bridge was employed. To obtain the resistance contributed by the bridge materials, the resistance of a column of potassium chloride solution of the same cross section and length as the bridge waa measured. The difference was assigned to the bridge material. Although the total resistance values obtained in the measure- ments were easily reproducible to within 0.1% of the total value, the differences are uncertain to 1 to 5 ohms. Hence, the values given should be considered orders of magnitude rather than abso- lute values.

1 ;

IOmm.PYREX

1

f

POROUS - VYCOR

Figure 2. Reference electrodes

sorption of materials such as dyestuffs; desorption is difficult. Mechanically, the bridges are rugged. Care must be taken to avoid drying porous glass with salt solution inside, as the ma- terial shatters when the salts crystallize.

Figure 1 gives bridge designs used in this laboratory.

Bridge A, which can be supplied by Corning, consists of a closed tube unfired at the closed end, but fired over the remaining length; the fired portion may be worked like ordinary Vycor. Bridge B is made from porous glass rod. The sleeve must fit very tightly to prevent creep of salts between the sleeve and porous glass. Tygon, or other relatively stiff tubing, is suitable for the sleeves. Bridge C is made from porous glass tubing. Ordinary glasswork- ing techniques will not work on this glass; hence seals and junc- tions must be of plastic or other material. For most work, type A is recommended. All the bridges illustrated operate satisfactor- ily with up to about 1 atmosphere pressure difference across them. Flow rates of electrolytes Tyith such a pressure difference are very low; about 0.1 ml. per week is typical.

Figure 2 gives some designs for reference half cells using porous glass bridges.

The other bridge materials used in this study which have great promise are the ion exchange membranes, which are sold under the trade name Amberplex by Rohm & Haas, Philadelphia, Pa., and under the trade name Nepton by Ionics, Inc., Cam- bridge, Mass. The membranes are available in cationic or anionic forms. The sheets are 1/18 to '/a inch thick, and handle, when wet, like medium hard gasketing materials; when dry they are fragile. I'either form can be allowed to dry under constraint, although they may be wet and dried repeatedly if no constraint is applied.

When these membranes are used in salt bridges, they do not give low junction potentials, as conduction through them is al- most exclusively either cationic or anionic. The advantage of the membranes is that they do not pass any solution per se, and can be used to block unwanted ions-e.g., mercury-from passing from the reference half cell to the other half cell. Figure 3 gives two forms of bridges using these membranes; both utilize the gasketing properties of the material. Bridge B is made from a small screw-cap vial with a hole drilled in the cap. Bridges of

ION EXCHANGE MEMBRANE

TYGON SLEEVE TO PROVIDE LENGTH a SEAL tON EXCHANGE

SCREW CAP V l A L T 2 m l . ) END SAWED OFF

Figure 3. Ion exchange membrane salt bridges

The total resistance of the A type of porous glass (of full tube length as shown in Figure 1) is about 300 ohms; of this 15 to 20 ohms are in the porous glass section. Porous glvs rods inch in diameter, such as those used in bridge B, give an average re- sistance of 9 to 10 ohms per mm. of length. The resistance of disks of porous glass 1/4 inch in diameter and ' / le inch thick (em- ployed in a bridge similar to membrane bridge A) averaged 15 to 20 ohms. Cation exchange membrane (Amberplex C-1) salt bridges of type A have resistances of 25 to 40 ohms across the membrane; similar bridges of anion exchange membrane (Amber- plex A4-1) have resistances of 40 to 100 ohms across the membrane. For comparison, the resistance of some 1 % agar-agar, saturated potassium chloride bridges was measured. The total resistance of a bridge made with tubing 7 mm. in outside diameter and 11.75 inches long was 620 ohms. For a bridge made from tubing 4 mm. in outside diameter and 11.75 inches long, the resistance was 2059 ohms. These may be compared to the 480-ohm total resistance of a bridge of porous glass Type A, l l 3 / 4 inches long (7 mm. in outside diameter). Although bridges of the same cross section show similar resistances, the mechanical properties of the porous glass bridge make it possible to use a much shorter length than would be feasible for agar-agar bridges. Bridges of 50 ohms or less are easily made with porous glass.

The limited experience with these new bridges to date has shown them to outlast conventional bridges many times, espe- cially in applications requiring evacuation of the measuring cell, and have shown little change in characteristics for several months, if the bridge was not allowed to dry. Their superior mechanical and flow resistance should make them popular, and their electrical characteristics are very satisfactory for all uses. The only ad- verse observation has been the severe leaching of salt from porous glass bridges stored in distilled water. This leaching is observed to a lesser extent with agar-agar bridges.

PRESEXTED at the Korthwest Regional 3Ieeting of the AMERICAS CHEMICAL SOCIETY, Richland, Wash., June 1 1 and 12, 1954.

V O L U M E 27, NO. 3, M A R C H 1 9 5 5

plywood base. Two wooden blocks accommodate the feet of the monochromator housing in fitted holes and raise the instru- ment to the height desired. ,4 platform with angle irons and set-screws holds the phototube housing. The light source is mounted on a metal platform and secured to the base by screws. The instrument can readily be converted from spectrophotometer to spectrofluorometer in less than 5 minutes.

The parts necessary to make the adaptation of the fluorescence accessory are Fhown in Figure 2.

Part 1 is the base for the light source, made from 1/31-inch sheet brass. This base can be modified for different light hous- ings and its height is determined by the position of the fluores- cence accessory lens. Part 2 is a brass ring collar, machined to fit on the lamp housing and into the lens opening. This collar holds in position the metal washers which determine the size of the incident beam. Part 3 is a light channel made from '/*-inch brass plate cut to size and soldered. The channel is designed at one end to fit the aperture a t the back of the mounting block and at the other to fit the opening of the phototube hou-ing. .4 projecting frame of thin metal sheet and a collar a t each end help to provide light-tight connectionp. In order to facilitate optical alignment, the face plate of the monochromator serves a s one wall of the light channel for part of its length. A layer of felt interposed between the cuvette holder and the channel helps to hold the latter tightly against the face plate. Part 4 is a gasket made of heavy blotting paper or pressed cardboard, placed between the mounting block and the fluorescence cuvette holder. Part 5 is an end plate of 1/8-inch brass made to fit the outer side of the cuvette holder. -4 laler of blotting paper and a gasket similar to part 4 should be cemented to the inside of this plate The gaskets help to provide light-tight connections and clearance for the cuvette carriage to be racked in and out. All parts are p:tinted in a dull black finish to minimize reflected radiation.

4'15

400 ' 5 0 0 ' 600 ' 7 0 0

mP Figure 3. Fluorescence spectra as determined with modified Beckman

spectrofluorometer Quinine sulfate in 0.1S sulfuric acid Fluorescein in aqueous sodium hydroxide

Rose bengal in ethyl alcohol

1. 2. 3. Pyronine in ethyl alcohol 4.

When the instrument is assembled for the first time, it is essen- tial to make certain that the ligKt source is in alignment with the mirror in the base of the fluorescence accessory and that the beam passes vertically through the cuvette. The angle of the mirror in the mounting block of the monochromator may require some adjustment, to ensure that light emerging from the lower slit passes through the center of the channel to the photocell.

RESULTS

The general principles of procedure described in detail by Huke and coworkers (9) apply equally to the above arrangement. The incident light reaches the cuvette from below rather than above and the light path through the solution is about 10 mm. rather than 2 mm., factors which, in the case of quinine in dilute sulfuric acid, reduce sensitivity by about 20y0 Other systems would be affected similarly unless they absorbed the exciting radi- ation strongly. In practice there has been no difficulty due to light scattering or fluorescence by the cell. In some applica- tions there is an advantage in being able to use covered cuvettes.

Because in many analytical situations fluorescence may be of a low order of intensity, the over-all sensitivity of the instrument is a major consideration. For comparison with other instruments the following calibration was carried out with a solution of quinine sulfate in O.l,\r sulfuiic acid. The incident light was passed through a 3-mm. Corning 9863 filter. With the photomultiplier at maximum sensitivity and a sulfuric acid blank set a t zero, a concentration of 0.15 y of quinine sulfate (V S.P.) per ml. of solution gave a scale reading of 100 at a $lit n-idth of about 1.8 mm. A straight-line relationship held for lower concentrations.

The higher sensitivity of the modification described by Huke and coworkers can be attributed mainly to the higher intensity of incident radiation supplied by the hydrogen arc and larger quartz lens LThich they used. Replacing the tungsten lamp in the present instrument with an Allen hydrogen arc operating at 0.5 ampere gave a 30% increase in fluorescence with quinine sulfate. il higher sensitivity could also be achieved with a mercury or xenon arc light source.

Figure 3 illustrates the spectral distribution curves of f l u e rescence emitted by extremely dilute solutions (less than 1 y per ml.) of quinine sulfate in sulfuric acid, fluorescein in aqueous sodium hydroxide, pyronine in ethyl alcohol, and rose bengal in ethyl alcohol. These substances were used simply to demon- qtrate the application of the instrument a t different wave lengths of fluorescent light.

4CKYOWLEDG\IENT

The authors wish to thank H. L. Kelsh and Theodor Pavlo- poulos, Department of Physics, for testing the Allen hydrogen arc in this system, and K. A Smithers, Fisher Scientific Co., Toronto. for assistance with certain technical aspects of this adaptation.

One of the authors (H. K.) is a Medical Research Fellow, Xational Research Council, Canada.

LITERATURE CITED

(1) Burdett, R. A., and Jones, L. C., J . O p t . SOC. Amer , 37, 554

(2) Huke, F. B., Heidel, R. H., and Fassel, 1'. A., Ibid. , 43, 400

(3) Lauer, J. L., and Rosenbaum, E. J., Ibid. , 41, 450 (1951).

(1947).

(1953).

Automatic Micromuffle for Determination of Ash in Carbonaceous Material

Robert Meyrowitr and C. J. Massoni, U. S. Geological Survey, Washington 25, D. C.

HE mineralogy and geochemiatry of carhonaceous rocks are T being studied as part of a program of research on the geo- chemistry of uranium that the U. S. Geological Survey is conduct- ing on behalf of the Atomic Energy Commission. To help solve certain phases of the problem, an organic microanalytical labo- ratory has been organized. This paper describes an automatic microcombustion apparatus that has been designed and used in this laboratory for the determination of the ash content of small amounts of carbonaceous materials. Often the amount of mate- rial available for analysis is small, necessitating the use of micro- methods. Norton, Royer, and Koegel (1) have shown that there is a saving of time nithout loss of precision when the microtech- nique is used for ash determinations.

These micromuffles have been made to order and require a great deal of machine-shop work in their construction. The micromuffle described by Norton, Royer, and Koegel (1) has platinum heat- ing coils, and the temperature is regulated by varying the voltage with a variable autotransformer. Either one or two samples can be burned at one time; the furnace is stationary, and the tubes are drawn through the furnace. There are two rates of travel, the slowest speed being 2.5 em. per 10 minutes. Steyer-

Two automatic micromuffles have been described ( 1 , Z ) .

476

mark's micromuffle ( 8 ) is heated by means of small Nichrome wire, and temperature regulation is by means of a variable auto- transformer. Two samples can be burned a t one time; the tubes are stationary and the furnace is moved. There is one rate of travel, 2.5 em. per 10 minutes.

The micromuffle described here can be assembled with a minimum of labor, using parts available from certain scientific laboratory apparatus companies.

A N A L Y T I C A L C H E M I S T R Y

Near the tous of A and B, cauillarv side mercuw container. C.

Figure 1. Micromuffle

The micromuffle pictured in Figure 1 consists of two parts: ( 1 ) a furnace drive, single speed (2.0 cm. per 10 minutes), elec- tric, having adjustable limit stops to arrest the advance of the furnace automatically, and (2) the moving, short-furnace section of an automatic, micro- and semimicro- combustion apparatus.

To the best of the authors' knowledge, part 1 is available only from Arthur H. Thomas Co., Philadelphia. 5, Pa. (Catalog No, 5683K) and part 2 is available only from E. H. Sargent and Co., Cdicago 30, Ill. (Cahlog No. 5-21580).

This furnace is of the radiant-heating type, with hinged shell and open-sided heating elements and operates an a 115-valt, 60-cyele, alternating current circdit. The furnace unit consists of six parts: (1) voltage-reducing transformer, designed to operate on lOO-valt, 50/60-cycle, single-phase service; the secondmy is rated 10 volts a t 28 amperes; (2) variable auto- transformer; (3) thermocouple and direct-reading pyrometer for temperature control; (4) fuse; (5) pilot light; and (6) switch. The maximum temperature for continuous operation is 900' C:

To calibrate the pyrometer, the thermocouple of a second pyrometer is inserted into the combustion tube until it reaches the position to be occupied by the combustion boat. While a stream of oxygen is flowing a t the rate to be used in the ash determination, the position of the thermocouple in the furnace is adjusted so that the reading of the pyrometer of the micro- muffle is the same as that of the second pyrometer.

LITERATURE CIT

(1) Norton, A. R., Royer, G. L., and Koegel, E., IND. ENC. CHEM.,

(2) Steyermark. AI. "Quantitative Organic Microanalysis," pp. 48-9, ANAL. ED., 12, 121-3 (1940).

Blakiston, Philadelphia, 1951.

PUBLICATION authorired by the Direotor, U. S. Geologioal Survey.

Apparatus for Automatically Changing Solvent Polarity during Chromatography

R. R. Allen and D. N. Eggenberger, Reieorch Division, Armour and Co., Chicago, 111.

N the separation of organic dibasic acids by partition chro- I matography [Higuchi, T., Hill, N. C., and Corcoran, G. B., ANAL. CHEM., 24, 491 (1952)j the mobile phase is a series of solvent mixtures progressively increasing in concentration of n-butyl alcohol. An apparatus has been designed that adds the solvents automatically in the correct proportions and, when coupled with a mechanical fraction collector, makes the whale separation procedure automatic. Also, constant pressure in the apparatus permits more even elution of the acids.

arms lead to a distributor bead D 20 which the Ohroimatograph tube is connected by a ball ioidt. Small funnels are joined throueh nressure stoncocks to the tons of A and B. The con- take< A; which con'tains chloroform; is a straight tube but B, which contains butyl alcohol, is made up of a series of progres-

ki th the first elutinr &l<ent mixture, the stopcock a t the-bot-

mercury container. C, is txen raised to a' Dosition so that the de-

mercury rises in A and B and forces both chlordark and butyl alcohol into the distributor head, The ratio of butyl alcohol to chloroform changes as the mercury rises into each new segment of B.

The diameter of each segment of B is calculated from each desired ratio of butyl alcohol to chloroform, and the length is determined by the total volume of each composition. For ex- ample, if 3% volume of solvent B in solvent A is desired, the ratio of the diameters of B a n d A would be 31'2 to 971'*. If the specific gravities of the liquids m A and B are equal, a segment of B containing 3 ml. would be the same length as a segment of A containing 97 ml. However, because the specific gravities of butyl alcohol and chloroform are different, the mercury level in B will be higher than that in A . The length of each segment of B may be corrected by the equation

(Sp. gr. A - sp. gr. B) hA (Sp. gr. Hg - sp. gr. B) Ah =

where Ah is the difference in mercury levels in A and B, and hA is the distance from the mercury level in A to the capillary-tube outlet in the distributor head.

rected for the difference'in mercury levels is the lengtb'hf each segment of B. The diameter of each segment of B is then deter- mined to obtain the required amount of butyl alcohol for each mixture.

Some space above the stationary phase in the chromatograph tube is necessary to obtain thorough mixing of the liquids as they are delivered. The change of compositions is smooth and does not give rise to false shoulders an the chromatographic peaks.