Van der Kloot built the instrument with the help of his father, Albert P. Van der Kloot, who is a chemist. The two of them set out to construct an instrument of high sensitivity and the greatest possible simplicity. The first goal was achieved with impressive success: the apparatus easily detects dilutions of one part in a million. The second objective a proved to be more elusive, according to

Albert Van der Kloot, who writes:

"The experimenter who designs instruments as a pastime soon discovers that striving for maximum simplicity runs counter to human nature Part of the fun comes in devising convenient features and amusing gadgets that are not strictly essential to the operation. Bob and I yielded to this temptation more than once. We built several versions of the apparatus and shall describe two of them. One version is comparatively insensitive but useful for demonstrating the principles of gas chromatography. The second version, when it is operated properly, is capable of matching the performance of professional instruments. We shall identify the features that can be omitted as they come up in our discussion.

"Our instruments are similar in principle to the gas chromatograph described previously in these columns by Steve Langhoff and Glen Martin [see 'The Amateur Scientist,' SCIENTIFIC AMERICAN; June, 1966]. That article is an excellent introduction to the technique of gas chromatography and should be consulted by beginners who undertake the construction of our apparatus. Our design differs from the Langhoff and Alartin one principally in the provision of a heated inlet fixture for injecting small specimens into the fractionating column, the use of more effective materials in the fractionating column and the substitution of a flame-ionization detector for the incandescent filament that Langhoff and Martin used for sensing the separated materials.

"The inlet of our system is essentially a T fitting. Specimens are injected into the inlet with a hypodermic syringe, the needle of which pierces a rubber septum that closes one end of the crossarm of the T. The opposite end of the crossarm is joined to the fractionating column.

"We made the T fitting by drilling a 1/8-inch hole in the side of a tubing connector 1/8 inch in diameter and silver-soldering a 1/8-inch copper tube into the hole. This formed the leg of the T, through which nitrogen gas flows into the system and sweeps the specimen through the column. A constriction was made in the leg of the T by inserting a piece of spring-steel wire .009 inch in diameter and squeezing the tube between the jaws of a vise. The wire was then removed. The constriction thus formed helps to control the flow of nitrogen and prevents the volatilized specimen from backing up into the nitrogen inlet. A ready-made compression T would have served our purpose but none was handy.

"To encourage the smooth flow of specimen materials into the column we drilled the fitting lengthwise and slipped a length of 1/8-inch copper tubing inside. The insert extended to within 1/8 inch of the outer end. Prior to assembly we reamed the outer end of the insert to a taper resembling a small funnel.

"The septum, which consists of a disk cut from a sheet of rubber about 1/16 inch thick, is clamped between the compression nut of the fitting and the threaded end of the crossarm. We also silver-soldered a disk of brass inside the nut and drilled a hole 1/16 inch in diameter through the center of the disk. The centered hole serves as a guide for the hypodermic needle when specimens are injected into the system [see illustration below]. This arrangement is one of the nonessential elements in the design of our gas chromatograph, but it is convenient for preventing damage to the point of the needle, which might otherwise be accidentally thrust into the metal.

"The fractions constituting a specimen separate most distinctly if the specimen enters the chromatographic column as a compact 'plug' of gas or vapor. To transform liquid specimens quickly into vapor we heat the entire inlet assembly electrically. In applying the heating unit the fitting was first wrapped with glassfiber insulation. We then wound six inches of 30-gauge Nichrome heating wire around the insulation to form a single-layer coil of widely spaced turns. The turns of the coil were bound in place by a second layer of glass-fiber braid. The electrical resistance of the coil is 3.2 ohms. It operates from the five-volt tap of a transformer.

"The design of fractionating columns varies with the nature of the materials to be analyzed. Normally columns consist of copper tubing 1/4 inch in diameter filled with a loosely packed inert granular substance known as the support. The granules of the support are usually coated with a thin film of liquid called the partitioner. The partitioner largely determines the performance of the fractionating column. Demonstration columns can be packed with an uncoated material such as Tide, the detergent, but the performance of such columns is comparatively poor. If Tide is used, it should be heated to a temperature of 150 degrees centigrade for several hours to drive off perfume and other volatile components.

"Diatomaceous earth was used as the support in the first gas chromatograph. This material is produced by the Johns Manville Corporation under the brand name Celite. The material must be screened by an 80-mesh sieve to separate from the powder the granular particles that might otherwise clog the column. Crushed firebrick can be used as the support; it too must be screened.

"The effectiveness of supports depends strongly on the uniformity of the particle size. The technology of support materials is developing rapidly. A wide variety of special types is now stocked by companies that deal in supplies for gas chromatographs. Some of the new materials are surprisingly costly. We use the support known as Gas Pac W, which is available from Chemical Research Services Incorporated of Addison, Ill. The serious experimenter will find it worthwhile to use a good support. The very best and most expensive types are not essential, but good separations cannot be made with inferior supports.

"Partitioners are equally important. We used three: glycerine, squalene and polyethylene glycol 600. This last chemical is available from the Jefferson Chemical Company of Houston, Tex. Columns can be made simply by mixing the liquid with the solid. Coatings of more uniformity can be prepared, however, by dissolving the liquid in a suitable solvent, stirring the mixture thoroughly into the solid and then evaporating the solvent. We used lacquer-thinner and stirred the batch occasionally as it dried. Be sure to do this work outside and avoid inhaling the vapor. In general we use 12 to 15 grams of liquid partitioner to 100 grams of support. Usually we make up batches of 25 grams, an adequate amount for packing a tube eight feet long and 1/4 inch in diameter.

"To the eye the coated material after drying appears similar to the uncoated support. It flows freely. To pack a column we stand the tubing vertically and plug the bottom end with a wad of glass wool. The coated support is poured into the upper end of the tube through a funnel. During this operation we hold a vibrating sander against the tube to make the powder fill the voids.

"Freshly packed columns should be coiled, placed in the oven of the instrument, heated somewhat above normal operating temperature and flushed with compressed nitrogen to eliminate traces of solvent and other volatile substances. The column should not be connected to the detector during this conditioning procedure. A few minutes of flushing is adequate in the case of columns intended for demonstrations or other low-sensitivity applications, but those designed to work at the highest sensitivity may have to be flushed continuously for several days.

"The rate at which materials travel through the column is partly determined by the temperature of the packing materials; therefore the temperature must be carefully controlled. The oven we used for maintaining the temperature of our low-sensitivity instrument is identical with that described by Langhoff and Martin. It houses the inlet, fractionating column and detector. We made it from a wooden box a foot square and nine inches deep. The top of the box is closed with a removable lid. The inner surfaces are lined with sheets of cement board fastened in place with Silastic Bathtub Calk, which is available in hardware stores. Oven temperature is regulated by means of a variable autotransformer that energizes a 300-watt heating unit inside the box.

"The oven used in the high-sensitivity unit is designed for higher operating temperature and consists of a metal box lined with Marinite, a product of the Johns-Manville Corporation. It is equipped with a system of electronic temperature control. Actually a minimum system for making demonstrations does not require an oven. Interesting separations can be made at room temperature with fractionating columns ranging in length from six inches to three feet.

"The heart of our detector is a tiny flame of hydrogen, mixed with gases from the fractionating column, that burns invisibly on the tip of a hypodermic needle. The point of the 20-gauge needle was ground off to make a right angle to the axis of the tube. The tube is clamped vertically in the end of a T fitting by a Teflon bushing, which insulates the tube electrically from the fractionating column. A loop of 20-gauge Nichrome wire, about 3/16 inch in diameter, is centered horizontally about 1/4 inch directly above the flame. This wire, which functions as a negative electrode, is also insulated from the fractionating column by Teflon bushings; they are supported by a metal housing that shields the flame both electrically and from drafts.

"During operation a potential of 150 to 200 volts is applied between the flame and the loop. The heat of the tiny flame breaks some of the molecular bonds of the gases and thereby ionizes the substances more or less, depending on their nature. The electrical conductivity of the path between the flame and the loop varies in proportion to the ionization of the burning materials. Conductivity of the path is measured continuously by a meter, the movements of the pointer indicating the passage of each substance through the flame. The response of the meter does not identify the substance but merely indicates its passage. Substances must be identified by passing known materials through the fractionating column and observing the rate and sequence in which they emerge. Known materials can then be added to unknown specimens; identification is established by comparing the performances.

"The housing of our burner was made of brass tubing 3/4 inch in diameter. The ends were closed by caps made of slightly larger tubing that telescoped over the 3/4-inch stock. Disks cut from sheet brass were silver-soldered to 1/2inch lengths of the larger tubing to complete the caps.

"The base of the housing involves bit of fussy construction. Openings must be made for admitting an air supply, the burner assembly and electrical connections [see Figure 3]. A brass compression T clamps the Teflon bushing, which in turn supports the hypodermic tubing, as previously mentioned. The fractionating column is connected to the opposite end of the crossarm of the T, and the hydrogen supply enters the burner through the leg of the T. The burner is inserted into the housing through a centered hole in the bottom, and the air supply occupies an adjacent hole.

"We equipped the burner with an ignition wire-a handy but nonessential feature. The gadget consists of a two-inch length of 24-gauge Chromel wire, silver-soldered at its center to the hypodermic tubing about 1/8 inch below the tip. Gas is ignited by connecting the heater wire momentarily to a source of three volts. The flame could be ignited almost as conveniently by lifting the top of the housing and using a match, but Bob thought it would be fun to devise the ignition circuit. It was. The polarizing voltage required between the burner and the wire electrode above the flame can be connected to either lead of the ignition wire. The leads are supported by a layer of Silastic Bathtub Calk in the bottom of the housing. The calk also serves as electrical insulation. Teflon insulation was used exclusively in the construction of the burner for our high-sensitivity instrument.

"Special care must be taken to insulate the wire-loop electrode. In effect it picks up the weak signal generated by the flame.

Spurious electrical leakage between the loop and the ground can generate 'noise' that may mask the desired signals. The loop for our low-sensitivity instrument was made from a short length of Nichrome wire that we stripped from a discarded 1,000-watt heating unit. Insulators were made of three 1/4-inch lengths of 1/4-inch Teflon rod. The plastic pieces were drilled axially and slipped over the wire like widely spaced beads. The assembly was then inserted into a length of 1/4-inch brass tubing that we had previously silver-soldered into a 1/4-inch hole in the side of the housing at a point that brought the loop 1/4 inch above the flame. For our high-sensitivity instrument we substituted a loop of platinum wire for the Nichrome.

"A supply of compressed hydrogen for the burner can be bought in a cylinder, but we prefer to generate our own gas by the electrolysis of water. Gas so obtained is less expensive, more convenient and much safer than compressed gas. A small leak in a supply of compressed hydrogen can be ignited accidentally, and the fire may escape notice because the resulting flame is almost invisible. Yet the heat is so intense that an accidental fire can lead to disaster. Leaks that do not catch fire may collect in pockets of potentially explosive gas that is both odorless and colorless.

"The design of the hydrogen generator can be varied according to the materials available. The construction requires a container, preferably of glass, for holding 25 ounces or so of electrolyte, which consists of a 15 percent solution of potassium hydroxide in distilled water. (A solution of this strength is caustic and may cause a painful burn if it comes in contact with the skin.) Two electrodes of stainless steel are immersed in the caustic solution and energized from a source of 12 volts, direct current. Our generator draws about two amperes and is energized by a battery charger.

"The negative electrode, at which hydrogen is liberated, is enclosed by a glass tube about 3/4 inch in diameter. Oxygen is liberated at the positive electrode, which is located outside the glass tube, and escapes into the air. The top of the 3/4-inch glass tube is connected through a rubber stopper to a filter containing silica gel, which separates the gas from a fine mist that rises above the electrodes [see Figure 4]. The filtered gas is fed directly into the burner of our low-sensitivity instrument.

"We used a one-liter graduate for the container, but any slender bottle with a wide mouth could be substituted. Electrodes were made by winding, with stainless-steel wire, a pair of single-layer coils two inches long. One coil fits inside the 3/4-inch tube near the bottom and the other is wrapped around the outside of the tube, also near the bottom. Electrodes of similar size and shape could be made of stainless-steel sheet.

"We insert a loose tuft of glass wool inside the 3/4-inch glass tube just below the stopper to retard the mist that forms above the bubbling electrolyte. We cover the top of the graduate with a mat of the same material to protect neighboring apparatus from the mist.

"If hydrogen is admitted directly to the burner from the generator, the flame will flicker imperceptibly because of bubbles that rise to the surface of the electrolyte and burst. The electrometer of our high-sensitivity instrument responds to the flickering flame as though it were detecting noise. We suppressed the wavering flow by inserting a constriction in the supply line from the generator. The constriction was made by inserting a piano wire about .01 inch in diameter into a length of 1/8-inch tubing and squeezing the tubing in a vise. The wire was then pulled from the tubing, which was inserted into the supply line.

"The rate at which hydrogen is generated varies with the voltage applied to the stainless-steel electrodes. We control the rate by a rheostat made of a short length of Nichrome wire and a battery clip. The Nichrome wire is connected to one lead of the battery charger; the clip is connected to one of the stainless-steel electrodes. The rate of gas production is adjusted by attaching the clip to selected points along the wire. The proper voltage depends on the geometry of the electrodes. If the voltage appears to be too low to generate an adequate supply of gas, reduce the spacing between the electrodes.

"The burner must also be supplied with air. In the case of our low-sensitivity instrument we couple the output of an aquarium pump directly to the inlet of the burner housing. Any oil-free pump of similar capacity should be adequate. The rate of flow is controlled by a needle valve. For operation at high sensitivity in a laboratory where organic fumes may be present we pass the air first through a drying tube packed with silica gel and then through a similar tube packed with Linde 5A, a material that functions as a molecular sieve. It is manufactured by the Linde Division of the Union Carbide Corporation, 270 Park Avenue, New York, N.Y. 10017.

"Gas flow into the burner should be regulated to approximately 25 milliliters of hydrogen per minute and 500 milliliters of air per minute. A simple meter for measuring the rate of gas flow is described in the account by Langhoff and Martin. The optimum rate of gas flow through the fractionating column depends on the size of the column. Normally it varies from 10 milliliters per minute in columns of 1/8-inch diameter to 40 milliliters per minute in 1/4-inch columns.

"The output of the flame-ionization detector can be observed most readily by connecting a vacuum-tube voltmeter across the output. Clip one lead of the meter to the fractionating column and the other to the wire-loop electrode of the burner. We used an old meter that was calibrated for a maximum potential of three volts with an input resistance of 11 megohms. A meter designed for lower full-scale voltage indication or with a higher input resistance would in effect have increased the sensitivity of the gas chromatograph. Even so, the old meter worked well for demonstrations and simple analytical work.

"The accompanying graph [Figure 6] shows a typical analysis. The graph displays voltage peaks that were generated by the detector when a specimen consisting of 10 microliters of mixed solvents was injected into the instrument. The solvents, in order of elution, were ethyl acetate, ethyl alcohol, n-propyl alcohol, n-butyl alcohol and isoamyl alcohol.

"Plotting output voltages and time recordings simultaneously can be a busy and frustrating job. If a number of test runs are to be made, the experimenter will soon become aware of the advantages of an automatic pen recorder. Commercial pen recorders are priced beyond the reach of most amateurs, and the best expedient might be to build a recorder of the type described in this department by Thomas W. Maskell [see The Amateur Scientist, SCIENTIFIC AMERICAN; July, 1966]. The instrument is relatively inexpensive.

"Maximum advantage of the high sensitivity of our instrument can be realized only if the output of the flame-ionization detector is measured by an electrometer. The electrometer need not be elaborate, because it is not required to measure currents lower than 10 trillionths of an ampere. We use a cathode-follower electrometer with a single vacuum tube [see Figure 5].

"Care must be taken with certain details of the construction. All insulation on the input side of the electrometer should be of Teflon and all related components must be clean. Avoid touching the insulation or the vacuum tube with unprotected fingers. The tube must be shielded against light, moisture and mechanical shock. All contacts must be clean. We found it important to use switches and wire-wound resistors of the highest quality. All components except special resistors in the megohm range can be bought from large electronic suppliers. We got the special resistors from the Victoreen Instrument Company of Cleveland, Ohio.

"As explained by Langhoff and Martin, the instrument is placed in operation by first turning on the nitrogen, air and hydrogen supplies. Heating current is then applied to the inlet and the burner is lighted. Finally the oven is brought up to a predetermined temperature, say 80 degrees C. in the case of a column filled with Gas Pac W coated with polyethylene glycol 600. After all units of the system have reached normal operating temperature, 10 microliters of a selected specimen are injected into the fractionating column. The analysis will be complete within a matter of minutes.

"Interesting initial experiments usually include the analysis of 'pure' solvents. Exceptionally high sensitivity is not required for detecting impurities in most volatile reagents, although the products are not as impure today as they were a few years ago before manufacturers began monitoring their processes with gas chromatographs. Our high-sensitivity instrument can also be used to analyze volatiles created by the fermentation of yeast. Many interesting features of fermentation can be investigated because the nature of the volatiles depends on such factors as the type of yeast, the composition of the mediums and the temperature of growth.

"Gas chromatography has attracted the interest of workers in several branches of biology, notably for the rapid identification of bacterial cultures, an application that shows great promise. Again, many variables await investigation. The gas chromatograph is among the recent developments in instrumentation. Opportunity for adding to its power as an analytical tool and for using it as an instrument to probe the unknown is by no means closed to the persistent and enterprising amateur."

Bibliography

GAS CHROMATOGRAPHY: PRINCIPLES, TECHNIQUES, AND APPLICATIONS. A. B. Littlewood. Academic Press, 1962.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.