Fill the burette and leveling bulb p with pure distilled mercury until, when r is raised and the mercury rises to the cock g, 2 or 3 c. c. of mercury remains in r. Additional mercury is unnecessary, and only adds to the weight to be raised and lowered during the analysis.

Put 1 or 2 c. C. of distilled water, for reasons already given, into the compensation tube before it is assembled in the water-jacket.

Pour caustic solution into the storage bulb q until, with cocks e and f removed, the solution comes to the mark o" on pipette n; then replace cocks e and f. Remove the glass electrodes in combustion pipette 7 and fill them with mercury, using a capillary funnel to introduce the mercury therein and taking care that no air bubbles are present, otherwise the electrodes will not conduct the electric current. Next replace the electrodes in the pipette, and force securely into the bottom of the pipette the rubber stopper holding them. The mercury exerts a great pressure on the rubber stopper and some may leak out unless the stopper is tight. Pour mercury into the storage bulb w until when cock d is removed the level just comes to the mark o' in pipette l; cock d is then replaced.

Pour caustic pyrogallate into storage bulb æ until, with cock c removed, the level of the solution just comes to mark o. Replace cock c, slightly inflate rubber bag p, and connect it to x by a rubber stopper and a U-shaped glass tube, as shown. Caustic pyrogallate solution is viscous and disagreeable to handle and should not be exposed to the air longer than necessary. A satisfactory method of preparing the solution is to use 350-c. C. magnesium citrate bottles; from these it can easily be poured directly into the pipette with minimum contact with the air.

GREASING THE STOPCOCKS Lubricate all stopcocks carefully with a good grease. Vaseline alone is not satisfactory for the Haldane apparatus, but should be incorporated with pure gum rubber and beeswax, as described later. Wipe the cocks thoroughly with cheesecloth to remove old grease and clean the capillary bores by passing a small copper wire through them. Pipe cleaners are better than copper wire. To clean the barrel of the cock most satisfactorily wrap one or two thicknesses of cheesecloth over the end of a lead pencil and pass them through the barrel, at the same time giving the pencil a rotary motion. Then clean the capillary leading to the cock with a pipe cleaner or copper wire. Rub a small amount of the grease over the plug of the cock and distribute it evenly over the entire length. Remove the grease that gets into the bore. Replace the plug in the barrel and give it a rotary motion in both directions, exerting pressure at the same

time. If the cock is properly greased there should be an even, transparent film throughout its entire length. Should any striations or bubbles show, especially at the neck of the plug, remove the cock and regrease it.

These instructions on the greasing of stopcocks have been given because proper greasing is very important and essential to the accurate analysis of gases. The usual tendency is to use too much grease, and after the cock is turned a few times the bore and capillaries become filled with grease; endless trouble follows and the solutions often rise into the manifold.

Draw one or two drops of water into the burette by turning cock g to communicate with h, and by turning h to communicate with the outside air. Raise bulb r until mercury flows out of the left branch of cock h; then dip the capillary tube connected to h and bend at right angles into a beaker containing water, draw a few cubic centimeters into the burette, lower r until the level of the mercury in the burette is near the bottom, and elevate r until mercury again flows out of the connection at h. Enough water sticks to the sides of the burette by this procedure to saturate the gases to be analyzed, and so prevents an excess of water in the burette, which would give erroneous gas-volume readings.

The solutions in the pipettes are now brought to their respective marks by the following procedure: Take 19.5 to 20 c. c. of gas into the burette by way of cocks g and h; then turn g 180°, connecting the burette with the manifold; open cock e and raise or lower leveling bulb r, as the case may be, until the level of the solution comes to mark o''. In a similar manner bring the levels in pipettes 1 and m to the marks o' and o''.


The analyst should now test the apparatus for leaks, remembering that even a very small leak will make the results obtained in the Haldane apparatus practically valueless, especially when three combustibles are present. Large leaks are easily found and removed, but small leaks are difficult to detect.

To test for leaks proceed systematically by first drawing into the burette about 20 c. c. of air and closing cock g. Then raise leveling bulb r, and put the gas in the burette under pressure; raise r until the mercury level is just below the bulb part of the burette, and hold it in this position. Observe whether the mercury level rises in the burette. A gradual rise indicates a leak in cock g. If the mercury level remains stationary, turn cock g to connect with the manifold, which places it under pressure, and test cocks c, d, and e for leaks. A leak in any of these usually, although not always, causes the solutions to drop from the marks 0, o' and o''. If the solutions do not drop and the mercury level in the burette does not rise, the cocks c, d, and e are gas-tight. Next make tests for leaks at the rubber connections below the cocks c, d, and e. Test pipettes m and l by removing storage reservoirs x and w on the back of the apparatus and lowering them several inches with cocks c and d closed. When x and w are replaced on the supports the liquids should come back to the marks o and o'.

If there are no leaks in the manifold, test the compensator and pipette n. With cock f open to the air, raise leveling bulb q until the level of the caustic solution nearly reaches cock f, then turn f 90° to the right, making connection with the compensator tube s. Lower q as far as possible by removing it from the clamp, noting whether the liquid in the small tube below cock f falls. If the liquid falls, there is a leak in the compensatory tube, and all the rubber connections should be examined and made gas-tight. Also, if the solution in n falls when the compensator is tested there is a leak in n.

Lastly, test the tightness of the capillary line connecting the sample tube with the burette by raising leveling bulb p 'until the mercury in the burette reaches cock g, placing a finger over the open end of the gooseneck, lowering leveling bulb r, and observing the mercury level to see whether it falls. Should the mercury level continue to fall, there is a leak in the line somewhere between the gooseneck and cock g.

The above tests should detect any leaks. Each time the apparatus is used, provided it has been idle several days, all connections and cocks should be examined as outlined.

Heavy rubber bands are very satisfactory for making gas-tight connections between the rubber t'ubing and the glass. Stretch and wrap a piece of rubber band about 2 inches long around the connection several times and tuck the free end under the last winding, drawing it up tight, as in tying with string. It is difficult to wrap some parts of the apparatus with rubber bands. In such places use No. 20 copper wire, winding the wire twice around the connection, and then drawing it up tight by twisting the free ends with pliers.


CAUSTIC POTASH SOLUTION For the removal of carbon dioxide, use caustic potash or soda solution. Prepare the potash solution in the proportions of 30 grams of stick potassium hydroxide in 100 c. c. of water.

CAUSTIC SODA SOLUTION Caustic soda is equally good for removing carbon dioxide. Long use causes the precipitation of a white deposit of bicarbonate on the glass tubes, but this is easily removed by washing the pipette with dilute hydrochloric acid before it is refilled with fresh solution. Prepare the caustic soda solution in the proportion of 20 grams of stick sodium hydroxide to 100 c. c. of water.


Alkaline pyrogallate solution requires a high concentration of alkali to prevent the production of carbon monoxide when the solution reacts with the oxygen. Moreover, the proportions of pyrogallic acid must not be too great, otherwise the solution becomes too dark for easy readings. Solutions filling the above requirements are prepared as follows:


Dissolve 1,200 grams of stick potassium hydroxide in 800 c. c. of water, and in another receptacle dissolve 50 grams of pyrogallic acid in 150 C. C. of water.

When the potassium hydroxide is cool distribute it evenly in four 350-C. C. magnesium citrate bottles; add the pyrogallic acid solution, 46 c. c. to each bottle. Stopper the bottles tightly by the spring clamps provided and shake them to mix the solutions, which are then ready for use. If the reagent is prepared in the bottles as described, it may be kept indefinitely without deterioration.

2. SODIUM PYROGALLATE SOLUTION A solution that gives off a minimum amount of carbon monoxide, has a fairly high rate of absorption, and is adapted for pipettes using glass tubes 16 is prepared as follows: Dissolve stick sodium hydroxide in an equal weight of water; this constitutes the stock sodium hydroxide solution. In another receptacle make up stock pyrogallic acid solution in proportions of 1 gram of pyrogallic acid to 3 c. c. of water. If the mixture is to be kept in 350-c. c. magnesium citrate bottles, place 250 C. C. of the sodium hydroxide in each bottle and add 100 c. c. of the pyrogallic acid solution. Stopper the bottles quickly and shake to mix the solutions, which are then ready for use.


Before a sample of gas is taken into the apparatus adjust the compensator by turning cock f to the air and raising or lowering a until the level of the caustic solution comes to the mark o''' on the capillary tube below f. Then close f, making connection with the compensator tube s, and leave it closed during the entire analysis.

16 Jones, G. W., and Meighan, M. H., Sodium pyrogallate solution as an absorbent for oxygen : Jour. Ind. Eng. Chem., vol. 11, 1919, p. 311.

If the gas to be analyzed is contained in a vacuum sample bottle as shown in Figure 1 (p. 3), make a file mark part way around the neck, place the end under mercury in the jar i, and break the neck off by tapping it against the side of the jar. Then introduce gooseneck j into the sample tube as shown. If the gas is contained in a sample tube of the type having connections on either end, remove the gooseneck and connect the sample at this point, dipping the other end into the bell jar i.


To begin the analysis, turn cock g to connect with h and the outside air; raise leveling bulb p until mercury flows out of the left capillary connecting with h; then turn cock h 90° to the left, making connection with the sample of gas; lower leveling bulb r, drawing in a sample of gas. When the mercury level reaches the 20-c. C. mark, again turn cock h 90° to the right and raise r, forcing the gas out of the burette. Discard this sample because it is contaminated with air in the capillary between the gooseneck and cock h. When the mercury reaches cock h again turn h 90° to the left, making connection with the sample; lower leveling bulb r, drawing in a sample of gas. When approximately 20 c. c. has been drawn into the burette, as indicated by the mercury level, turn cock g 180°, connecting the burette with the manifold. Raise leveling bulb r about 1 inch to prevent the caustic solution from rising and getting into the manifold when cock e is opened. Open cock e, and raise or lower r until the caustic solution comes to the mark 0''. At the same time, raise or lower leveling bulb q until the caustic solution comes to the mark O''' on the compensator. When the solutions are on both marks, read the volume of gas in the burette to the nearest 0.002 c. c. This reading gives the volume of sample taken for analysis.

TEST FOR CARBON DIOXIDE Remove the carbon dioxide by raising and lowering leveling bulb r, thus forcing the gas back and forth into the caustic pipette n. When the gas has been passed six times into the caustic pipette, read its volume again by compensating in the manner described before. Check this reading by passing the gas six times more and noting whether there is any further contraction. When this volume is subtracted from the volume of sample taken, the result gives the contraction due to carbon dioxide, and when divided by the volume of the sample taken for analysis and multiplied by 100 the result is the percentage of carbon dioxide in the sample.

Next, determine the combustibles by passing the gas into pipette ? by closing cock e, opening cock d, and raising leveling bulb r. Close

« ForrigeFortsett »