Oct., 1919
I
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
961
LABORATORY AND PLANT PREPARATION AND TESTING OF HYDROGEN OF HIGH PURITY B y JUNIUS DAVID EDWARDS Received February 28, 1919
In t h e course of a study of t h e properties of pure gases, t h e preparation of pure hydrogen by several different methods was undertaken. W. A. Noyes has already shown‘ in his work on t h e atomic weight of hydrogen, t h a t hydrogen of exceptional purity could be prepared b y t h e electrolysis of a solution of barium hydroxide. His work leaves little further t o be said regarding t h e purity of hydrogen, which can be produced in t h a t manner if t h e proper precautions are adopted. Where t h e highest purity is not essential, other methods may offer certain advantages. Chief of these is t h e greater speed with which hydrogen can be generated b y such means as t h e reaction between zinc and acid or of hydrone (sodium-lead alloy) and water. For example, t h e maximum rate of generation from a small electrolytic cell taking I C amp. is about 4.4 1. per hr., whereas I O t o 2 0 times t h a t rate can be easily attained with a small-sized Kipp generator for relatively short periods cf time. T h e Kipp generator, however, has two disadvantages where purity of t h e product is t h e first requisite. The residual air must be swept out (which consumes time and material), and the acid is in contact with a;r in the upper reservoir; this air dissolving in t h e acid may slowly diffuse into and contaminate t h e hydrogen. T o obviate these difficulties, t h e generator shown in Fig. I was designed b y t h e author in cooperation with Mr. E. R. Weaver. The operation of t h e generator is apparent from t h e figure. T h e chief advantages claimed for this generator are: I -The sweeping-out period of generation is eliminated. 2-An atmosphere of pure hydrogen is maintained in all parts of the apparatus at all times. 3-Pure hydrogen can be obtained from the generator at any time--immediately after filling or after standing for long periods. 4-Fresh acid can be introduced without admitting air. T h e manipulation b y which this is accomplished is as follows: Before filling t h e apparatus i t is first evacuated through a. T h e residual pressure in t h e generator is indicated by t h e height of t h e mercury in t h e t u b e d which serves as a manometer and also as a blow-off. Any leak can t h u s be detected b y t h e indications of this manometer. If t h e generator contains any liquid, t h e pressure cannot, of course, he reduced below its vapor pressure. Continued pumping will, however, sweep out t h e generator with a stream of water vapor. The acid t o be introduced (sulfuric, I : 8) is contained in and nearly fills t h e flask f which is connected t o t h e generator by a delivery t u b e as shown. Several pieces of zinc are introduced into t h e flask, in order t o sweep all the air out of t h e acid and saturate it with hydrogen. When this has been accomplished, t h e vent e is closed until t h e pressure 1
Bureau of Standards, Bulletin 4 (1908), 345.
increases sufficiently in f, from t h e generation’of hydrogen, t o force t h e acid through t h e delivery t u b e and out t h e end of t h e Geissler stopcock; all air is t h u s removed from t h e connections. On turning t h e cock to connect with t h e generator, t h e acid will t h e n r u n into t h e evacuated generator. Before t h e acid has reached t h e level of t h e zinc, t h e hydrogen dissolved in it can be removed b y pumping in order t o further ensure complete absence of air. The proper amount of acid is then allowed t o enter and hydrogen generated until i t bubbles out through d , which shows t h a t t h e generator is under pressure. The rate of generation is regulated b y t h e cock c.
i I PIG.1-GENERATOROB EDWARDS AND WEAVERFOR PURB GASES
The generator illustrated was made from two large calcium chloride towers with glass stoppers. The stopcock at b entered through a rubber stopper which was always under liquid. All parts are rigidly mounted on a stand t o prevent breakage of t h e glass connections. The size, shape and arrangement of parts can of course, be modified as occasion demands. PURITY
OF HYDROGEN FROM
ZINC A N D ACID
Although t h e purity of t h e gas may vary with t h e purity of t h e materials employed, the present tests are of interest as showing t h e possibilities of t h e method. T h e zinc used was a very pure sample from an English source; t h e acid was J. T. Raker’s “C.P.”sulfuric acid. The possibility of t h e hydrogen containing t h e following impurities should be considered: Oxygen a n d nitrogen, sulfur dioxide, hydrogen sulfide, acetylene, methane or other hydrocarbons, carbon monoxide, carbon dioxide, arsine, phosphine, stibine, and water vapor.
962
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
Of these impurities, oxygen and nitrogen which might come from air leaks, had t o be rigidly excluded because there is no satisfactory method of removing traces of nitrogen. Sulfur dioxide, hydrogen sulfide, and carbon dioxide were removed b y passing t h e gas over solid potassium hydroxide. The gas was thoroughly dried by passing through three tubes of sublimed phosphorus pentoxide. All connections were glass t o glass sealed with de Khotinsky cement where necessqry. The following tests for impurities were applied to t h e gas. REFRACTIVITY-The refractivity of a sample of hydrogen is a delicate criterion of its purity because all gases which are likely t o be present have much higher refractive indices t h a n hydrogen; helium and neon are t h e only gases having lower refractive indices. The gas interferometer,l which measures t h e difference in refractivity of two samples of gas, may t h u s be used for detecting impurities in hydrogen. The interferometer was calibrated by t h e method of Edwards.z For t h e standard of comparison, hydrogen was generated b y electrolysis of barium hydroxide in a n all-glass electrolytic cell3 and purified b y passage over hot platinized quartz, solid potassium hydroxide, a n d sublimed phosphorus pentoxide. Every precaution was taken t o secure t h e requisite purity of t h e materials and t o exclude leaks in t h e apparatus. The hydrogen so generated was considered t o have considerably less t h a n I p a r t in 10,000of impurity present. When compared with this gas in t h e interferometer, t h e hydrogen from zinc showed a total impurity, calculated as air, of less t h a n 0.01 per cent. The interferometer was of such sensitivity t h a t 0.00j per cent air in hydrogen could be detected. If gases other t h a n air were present, the actual percentage would be still lower because their higher refractive index gives a greater sensitivity in the reading of t h e interferometer. TESTS FOR C A R B O N COMPOUNDS-TO test for carbon compounds, t h e hydrogen was burned with carbonfree air in a hard glass tube filled with copper oxide. The residual air was then passed through a solution of barium hydroxide contained in t h e spiral wash bottle4 of Weaver and Edwards. The precipitated barium carbonate was filtered off and washed free from barium hydroxide without removing i t from t h e wash bottle or allowing i t t o come into contact with air containing carbon dioxide. The small precipitate of barium carbonate which was formed yielded 4.4 mg. of barium sulfate, equivalent t o 0.83 mg. of carbon dioxide. If this came from t h e oxidation of acetylene i t would have corresponded t o 8 p. p. m. b y volume in t h e 2 5 1. of hydrogen from which i t came. Acetylene was determined directly in t h e gas by t h e colorimetric method of Weaver5 using an absorbing solution of cuprous chloride containing gelatin, hy1
For a description of the gas interferometer, see I,. I€. Adams, -7,
Am. Chem. SOG.,57 (1915), 1181. a Ibid., 59 (1917), 2382. 8 Similar in design to that of Weber, Bureau of Standards, Bulletin 9 (1913). 135. 4 THISJOURNAI,, 7 (1915), 534. 5 J . Am. Chem. Soc., 58 (1916), 352.
Vol.
11,
No.
IO
droxylamine hydrochloride, alcohol, and ammonia in t h e proportions recommended in t h e article cited. T h e limit of visibility was about 0 . 0 2 mg. of acetylene. Only t h e faintest coloration was secured which indicated not more t h a n I or 2 p. p. m. of acetylene. T h e remainder of t h e carbon found in t h e previous test may have come from traces of carbon monoxide, methane, or other hydrocarbons. Hydrogen generated from a sample of hydrone contained 4 p. p. m. of acetylene as determined in t h e same manner. TESTS F O R A R S I N E A N D STIBINE-Arsine and stibine were absent as indicated b y t h e lack of any deposit on t h e passage of t h e gas through a heated tube. Arsenic and antimony were not detected in t h e chemical analysis of t h e zinc. All of these results confirm t h e conclusions arrived a t from t h e indications of t h e interferometer, namely, t h a t hydrogen containing not more t h a n I p a r t in 10,000of impurity can be prepared from zinc with t h e generator described. It was of interest t o determine t h e purity of gas obtained from an ordinary Kipp. The gas was compared with t h e hydrogen from t h e generator shown in Fig. I , by means of t h e interferometer. T h e g a s from t h e Kipp was purified b y passage through soda lime and anhydrous, granular calcium chloride. There was a short piece of rubber tubing leading t o t h e drying tower and three rubber stoppers in t h e line. No leaks were detected in t h e system upon evacuation. T h e purity of t h e hydrogen and t h e amount generated by t h e Kipp were determined every 5 min. The results are shown in Table I. TABLEI-COMPARISONOF PURITYAND YIELDOF HYDROOBN RROX KIPP GENBRATOR
Time from Start of Generation Min. 17 22 27 32 37 42 57 77 112
Purity of Hydrogen Per cent 83.86 93.59 97.86 99.51 9.. 9.92
99.97 99.974 99.973 99.976
Calculated
Total Volume of Gas Generated Liters 3.96 6.02 8.50 11.58 15.00 18.35 28.4 41.7 74.3
volume-Of Gas Required t o Reach Indicated Purity Liters 4.00 6.03 8.45 11.70 15.69 17.84 18.17
...
18.3
The results given in Column 4 of t h e table were calculated from t h e following equation:l Total Gas Evolved Total Gas Space
-
I
loge(.
- I) IO0
I n this equation a is t h e purity of t h e hydrogen expressed as a percentage. The gas space in t h e case of t h e generator tested was 2 . 2 1. It is assumed in this equation t h a t there is perfect mixing a t all times of t h e hydrogen with t h e air in t h e generator. The agreement between t h e experimental and t h e calculated valu$s is quite satisfactory and shows t h a t t h e equation can be used in calculating t h e sweeping-out period for any similar generator when gas of any required purity is desired. 1 Weaver, “Generation of Hydrogen from Ferrosilicon,” 4th Annual Report Nat. Adv. Comm. for Agronautics, 1918.
Oct., 1919
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
Hydrogen of approximately t h e same purity (99.97 per cent hydrogen) was prepared in t h e same Kipp from J. T. Baker’s “C. P.” zinc said t o contain a trace of arsenic a n d antimony.
963
The apparatus is designed t o meet t h e requirements for these three determinations. APPARATUS
To set u p t h e a p r z r a t u s t h e following material is necessary: SUMMARY I-Double burette stam,. A generator is described with which hydrogen 2-TWO certified burettes, 30 cc. capacity, graduated t o 0.05 cc. containing not more t h a n I p a r t in 10,000 of im3-Capillary glass tubing, I t o 2 mm. inner diameter. purities can be prepared from zinc and acid. Methods 4-Four glass T tubes: Three, inner diameter 3 mm. One, of testing for contaminating gases are described. inner diameter 4 mm. T h e relation between t h e volume of gas generated and 5-Red rubber tubing, inner diameters of I to z mm. and also purity of t h e product is given b y a n equation which 4 mm. was verified experimentally. 6--Pinchcocks, three plain and one special two-way cock. 7-Adhesive tape. BUREAU O F STANDARDS WASHINGTON, D.C. Fig. I shows t h e apparatus as completed and in running order in t h e laboratory. The burette on t h e AN APPARATUS FOR RAPID GASTRIC ANALYSIS right is connected TOGETHER WITH A METHOD FOR THE t o the alkali reserPRESERVATION OF STARCH SOLUTION voir, t h a t on t h e By RAYMOND J. MILLER left t o t h e thioReceived h’ovembrr 29, 1918 sulfate. At t h e t o p Considerable of t h e success accomplished along t h e of t h e burettes is line of gastric investigation conducted a t t h e Jefferson placed a n arrangeMedical College recently is due t o a n apparatus de- ment which is signed b y one of t h e members of t h e research staff, shown in detail in t h e author of this contribution. With t h e aid of this Fig. 2 . Near t h e appas-atus t h e manipulation of different solutions used base of t h e burette i n t h e several analyses has been shortened. The stand two devices apparatus demonstrates its real value not in t h e hand- will be observed. ling of one case for analysis b u t rather when three or The one on t h e more are t o be run. right leading t o the tip of t h e alkali burette is an automatic pipette used in t h e formol titration, a n d is shown in detail in Fig. 3. The attachment on t h e left leading t o t h e tip of t h e thiosulfate burette is an ordinary FIG.2 gravity device for t h e rapid addition of starch solution. Fig. z illustrates t h e overhead arrangement whereby t h e burette can be filled b y suction and when t h e same is released any excess solution drawn over is immediately withdrawn back into t h e reservoir, leaving t h e burette reading 0.00 cc. FIG. 1 Fig. 3 shows the automatic pipette which, as soon a-Alkali Burette d-Syphon Device (Starch) as its charge is delivered, automatically fills u p t o b-Thiosulfate Burette e-Formation Reservoir approximately its former level a n d is again ready t o c-Automatic Pipette +Starch Solution be discharged. The principle upon which this deI n gastric analysis as conducted in this laboratory vice depends is a two-way pinchcock as illustrated a t t h e following determinations are considered of most A, Fig. 3. This t y p e pinchcock was designed b y importance: another author a n d reported some time since. Very little has been said in regard t o t h e assembling I-Total Acidity: Titration with N / I O Opotassium hydroxide. z--.Free Acidity (Sahli Method) : Titration with N / I O O of material for t h e several devices as i t has been taken for granted t h a t t h e figures are self-explanatory. Persodium thiosulfate, finishing with starch solution. 3--.Formol Titration or Amino Acid Nitrogen Complex : haps b u t one thing needs an explanation a n d t h a t is t h e purpose of t h e adhesive tape. It will be observed Addition of neutral formalin and titrating with N l i o o potasin Fig. I t h a t t h e automatic pipette c and t h e starch sium hydroxide.
