STUDIES ON PHOSPHORIC ACID. III* The Use of Phosphoric Acid in

in an open evaporating dish at 2 jo'. This acid gave the yields recorded in. Table I. The yields varied greatly at the lower temperatures so an averag...
0 downloads 0 Views 441KB Size
STUDIES ON PHOSPHORIC ACID. III* The Use of Phosphoric Acid in the Preparation of Ethylene BY ARTHUR G. WEBER WITH JAMES H. WALTON

Phosphoric acid was first used as a dehydrating agent in the preparation of ethylene by Pelouzel in 1833, although it had been demonstrated by Lassaigne2 thirteen years before, that alcohol and phosphoric acid would react with each other. iSewth3 and Prideaux4 have described laboratory methods for the preparation of ethylene using phosphoric acid while Boullayb and B a c k h a d have studied the reaction from the standpoint of its commercial possibilities. Lindinger and Moser' found that a mixture of alcohol and phosphoric acid would give off ethylene when heated to as low a temperature as 2 0 5 ' . They concluded that the formation of ethylene was due to the action of the pyro phosphoric acid into which the ortho phosphoric acid was being dehydrated] and that a more complete conversion could be obtained by using a higher temperature. By passing alcohol vapor over finely divided pumice soaked in phosphoric acid heated to a temperature of 2 5 0 ° - 3 0 0 0 a yield of 90% ethylene was obtained which was 99.5% pure. Newth demonstrated that syrupy phosphoric acid heated previously to a temperature of 200' was more efficient in the preparation of ethylene than glacial phosphoric acid or phosphorus pentoxide. The mechanism of the reaction is described by Balareff8 as being different at different temperatures] the catalysis taking place partly by the formation of intermediate compounds and partly by adsorption and resulting thermal decomposition. In order to discover the treatment which would produce the most active phosphoric acid for the production of ethylene] and to study the effect of various temperatures and accelerators on the reaction] the following experimental work was performed using a modification of Newth's apparatus. No attempt was made to devise an apparatus which would give an exceptionally high yield of ethylene] but rather one which would permit a variation of the factors influencing the reaction and make it possible to determine the effect on the yield arid purity of the gas produced.

* Contribution

from the Laboratory of General Chemistry, University of Wisconsin. Pelouze: Ann. Chirn., 52, 37 (1833). 2Lassaigne: Ann. Chim., 13, 294 (1820). 3 Newth: J. Chern. SOC.,79, 916 (1901). 4 Pndeaux: Chem. News, 113, 277 (1916). 6 Boullay: Gilbert's Ann., 44, 270 (1913). EBackhaus: U.S.P., I, 372, 736, March 29 (1921). 7 Lindinger and Moser: Monatsheft., 44, 141 (1923). 8 Balareff: Z.anorg. Chern., 158, 105 (1926). 1

ARTHUR G . WEBER WITH JAMES H. WALTOX

2694

Apparatus The apparatus consisted essentially of a reaction flask, a condenser flask, and a bottle to measure the gases generated (Fig. I ) . The reaction flask was made from a zoo cc. round-bottom pyrex flask. The top was replaced by a mercury seal and two side necks attached which extended diagonally up from the sides thus permitting the alcohol to enter thru a jet on one side (a)

FIG.I

and the gas to leave thru the jet (b) on the other side. The jet thru which the alcohol entered was a six mm. glass tube drawn out to an opening of about one mm. and extending to the bottom of the flask. To the other end of the tube carrying the jet, a burette was connected for measuring the absolute alcohol. The stirrer was made from pyrex rod flattened until it was about one centimeter wide. The reaction flask was heated in an oil thermostat a t 250'

.3'.

The reaction flask was connected to the condenser flask by means of a I O mm. glass tube (b), bent as diagrammed. Slight changes in the diameter and length of this tube caused large variations in the yields obtained. The condenser flask was a wide mouthed z j o cc. flask completely surrounded and covered with ice. The gas was measured in a five liter aspirator bottle a t 30' over a saturated solution of sodium sulfate which had previously been saturated with ethylene. hlr. Peterson of this laboratory has shown that the solubility of ethylene in a saturated sodium sulfate solution a t this temperature is less than one per cent. Method of Procedure The apparatus assembled as diagrammed was allowed to come to constant temperature. The stirrer of the reaction flask was removed and the treated phosphoric acid waa allowed to run into the flask from a weighing pipette

STUDIES ON PHOSPHORIC ACID

2695

which drained approximately 93 grams. After the acid was introduced the stirrer was replaced and the acid allowed to stir for five to ten minutes. The three-way stop-cock a t the measuring bottle was turned to allow the gas to enter the bottle, the leveling bottle was lowered so there was a slight evacuation in the apparatus, and the burette adjusted so that the alcohol dropped into the reaction flask at the rate of about a drop each second. When the run was completed the gas was analyzed, the bottle emptied, and the next run of the series started. Since there was air in the apparatus a t the start, the first run of a series always contained about 10% air, the second run about 1 . 5 % air, and the third about . 5 % ; the fourth usually contained an amount that was not detectable.

Effect of various temperatures of reaction on yield of ethylene. Newth3 used a temperature of zooo to 220' for the preparation of ethylene while Lindinger and Maser? obtained better results by keeping the reaction between 250' and 300' in an apparatus which permitted the alcohol vapor to pass thru a tube containing pumice soaked with the acid. I n order to determine the effect of temperature of reaction on the yields of ethylene, a sample of syrupy phosphoric acid was heated for 4-1/2 hours This acid gave the yields recorded in in an open evaporating dish a t 2 jo'. Table I. The yields varied greatly a t the lower temperatures so an average is The apparatus would given; uniform yields were obtained a t 250' and 260'. not permit higher temperatures. TABLE 1 Effect of Temperature on the Yields of Ethylene produced by Absolute Alcohol and Previously Heated Syrupy Phosphoric Acid Temperature of reaction 210' 220' 230' 240' 250' Yield of ethylene, (yo) 30 35 52 61 81

260" 82

T h e effect of the previous heat treatment of acid. I n order to observe the effect of the previous heat treatment of the acid on the yields of ethylene two kilos of syrupy phosphoric acid (Sp. Gr. 1.71) were heated during a period of about forty-five minutes until a temperature of 250' was reached after which a sample of approximately Ijo cc. was removed. Two and one quarter hours later another sample was removed, this was repeated a t two and one quarter hour intervals until finally a nine hour sample had been withdrawn. The different samples were preserved in glass stoppered bottles until determinations of the yields of ethylene they would produce could be made. These samples were analyzed for the total amount of phosphorus pentoxide by the method of Berthelot.9 Because hot phosphoric acid attacks glass ware, a noticeable amount of material from the evaporating dish was dissolved in the last three samples to 0

Berthelot s n d Andre: Compt. rend., 123, 773 (1896).

2696

ARTHUR C.. TVEBER WITH JAMES H . W A L T O S

be removed, the other samples appeared perfectly colorless. Runs 1,2,3,4,5 (Table 11) were made by adding 5 portions of alcohol consecutively to the same sample of acid. The addition of each portion therefore constitutes one run. The yield was always low on the first run of any series probably because of the alcohol consumed in making esters of the phosphoric acid.

TABLE I1 Yields of Ethylene produced by the Interaction of Absolute Alcohol and Phosphoric Acid Previously Heated at z i o o Temperature of Reaction 2 5 0 " No. of Hrs heated sample a t 250'

Total P20, present c c

Xiumber of run I

3

2

4

j

I

9

73 4

So ;

87

0

s2 0

2

6 3/4

i3

86

I

3 4

4

1'2

80

)C

114

68 3 68 6

78 3 SI 0

2

84 4 82 8 SI 6

77 1

80 5

80 7 77 4 80 0 81 5

5

0

66

i 2

9

73

i4 4

i4 4

1

2

1

81 04

80 46

79 79 78 8s

7 5 34

Upon t,he addition of the alcohol there appears to be an equilibrium approached between the reactants and their products which tends to give constant yields of ethylene. In a number of trials, acid treated in the same manner as sample KO. 3 seemed to permit this equilibrium to be attained more rapidly than with longer or shorter heat treatments. During the reaction the acid remained clear but became a dark amber color. Small amounts of oil, ether, alcohol, and water collected in the condenser flask.

Reproducibility of results. The possibility of reproducing the same acid and duplicating results is shown by the data in Table 111. Three series of runs were made on three independently prepared samples of acid. The first series of runs was made with the same acid one run being made each day, the apparatus being allowed to cool down after each run. In the second and third series each run was started immediately at the conclusion of the preceding run.

TABLE 111 Yields of Ethylene produced by Independently Prepared Samples of Phosphoric Acid and Absolute Alcohol. Temperature of Reaction 2 j o " Acid Heated Four and One Half Hours Previously at z i o o Sample I

2

3

Per rent gield of ethjlene 3 4 5

I

2

71 o 68 3 64 S

SI o 82

S

SI 3

So o 80 9 SI 2

o

79 o

81 o

80 o

j8

6

81

I

Sz o

2697

STUDIES ON PHOSPHORIC ACID

The variations in the first runs of each series can be ascribed to slight variations in the charge of acid used in the reaction flask and slight differences in the degree of dehydration of the acid. This acid, prepared as described above, was used in the work which follows.

Effect of adding H P & , H4P20,,and P z O ~ . The data tabulated in Table IV confirm Newth’s3 statement that neither glacial phosphoric acid nor phosphorus pentoxide are as efficient in the preparation of ethylene as a previously heated syrupy phosphoric acid. For this work Kahlbaum’s C.P. preparations were used, twenty grams being added to the usual charge of acid and stirred thoroughly before starting the run. I t appears that the pyro acid disturbs the conditions within the flask least and one may conclude ( I ) that the previously heated acid approximates the pyro acid or (2) that the addition of the pyro acid has no effect on the reaction. The former conclusion would agree with those of Lindinger and Maser' and Senderens,’O who attribute the dehydration to the pyro acid.

TABLE IV Yields of Ethylene when HP03, HaPzOl,and PZOEare added to the Csual Charge of Acid. Temperature of Reaction 250’ Yield of ethylene in per cent

Acid added I

Sone HPOs H4P207 PZOS

2

82.8 66.4 80.4 74.0

68 .o 67.6 63 . 2 74.6

3 80.9 65.1 81.9 82.6

4

81 . o 65 . o 81.2 82.1

The eflect of uarious oxides, salts, etc. on the reaction. Various catalysts have been tried in the preparation of ethylene from alcohol and sulfuric acid. Lindinger and Moser7 found that one and one-half to two per cent of anhydrous copper sulfate was the most suitable cat,alyst. hnhydrous copper sulfate also accelerates the reaction between ethylene and sulfuric acid.” If the reaction is similar in the case of phosphoric acid, copper oxide and silver sulfate should be good catalysts for the dehydration of alcohol with phosphoric acid particularly since Xuller*2has shown that these substances are accelerators for the reaction between the phgsphoric acids and ethylene. Frofn one and one-half to two per cent of the various substances were added to the reaction flask and the yields of ethylene as well as its purity determined. Most of these substances caused the acid to foam and boil out of the reaction flask on the first run. Since the foaming takes place only to a slight extent with the alcohol and the acid alone, in many cases one run was made with only the acid and the alcohol, added substances being introduced at Senderens: Ann. Chim., IS

‘I ip

1 1 7 (1922).

W.Gluud and G. Schneidir: Ber., 57B, 254 Muller: Ber., 58B, 2105 (192j).

(19241

2698

ARTHUR G. W E B E R WITH JAMES H. IVALTOX

the beginning of the second run. Before the alcohol was introduced for the second run special attention was paid to make sure that everything had come to equilibrium. Slight changes in the quantity and the purity of the various catalysts seemed to have no effect on the reaction, consequently very accurate weighing and further purification of the C.P. chemicals was unnecessary. I n each case portions of alcohol were added consecutively to the same sample of acid and catalyst, three, four, or five runs being madeineachexperiment with a given substance to make certain that the maximum yield had been reached in each experiment. The results are given in Table V, where for each catalyst the yield of ethylene and the purity of the gas are given. I n considering the effect of the added substances upon the reaction, there is a possibility that there is a change in the surface reaction at the interface as the alcohol vapor passes thru the acid, or that these compounds may act in some other role to change the equilibria wit'hin the reaction flask. Altho the side reactions produced by phosphoric acid are relatively insignificant compared to the sulfuric acid reaction, Rl~ntmollin'~ found that there is 6.2 grams of oil consisting of organic condensation products formed for each I O O grams of alcohol used in Iiewth's apparatus. A quantitative estimation of the side reactions was not made in this study because in most cases after making a run, only a large drop of oil was observed floating on the liquids of the condenser flask. KO marked change in the amount of oil was observed when the catalysts were used with the phosphoric acid. Lindinger and lloser' found that the first portion of alcohol added to the phosphoric acid mixed with difficulty owing to the high surface tension of the acid, and because of this fact, a large portion of the alcohol evaporated before entering into the reaction. They found the yield of ethylene was greater when sufficient alcohol had been added to lower the surface tension of the acid. -11though in this study the alcohol was introduced from below instead of being dropped on the surface of the acid, the lowering of the surface tension of the acid by the addition of the catalyst may be a factor here which tends to increase the yield of ethylene. The possibility that the added catalyst renders the intermediate complex of alcohol and phosphoric acid less stable or produces a slightly different "pivot" for the reaction, as referred to by Senderens,'O is shown by the following experiment: An excess of alcohol was added to about 20 cc. of the previously heated syrupy phosphoric acid. The mixture was thoroughly agitated and kept at a temperature of 250' over a period of twelve hours, during which time equilibrium was established. By means of a glass capsule suspended from the stopper inside the flask, two grams of silver oxide was added to the equilibrium mixture, Approximately 5 0 cc. of gas was liberated. I n a similar experiment with anhydrous copper sulfate slightly less gas was liberated. The existence of metallic ethyl pyro phosphates in this reaction would seem very probable in those cases where metallic oxides were added to the '3Montmollin: Bull.. i n )

19. 242

(1916).

2699

STUDIES ON PHOSPHORIC ACID

treated acid to which the alcohol had already been added. The description of several metallic alkyl ortho phosphates is given by CavalierI4 and the barium ethyl pyro phosphate has been used as a method to separate the ethyl esters of the phosphoric acids by Langheld.15

TABLEV Yields and Purity of Ethylene produced when Various Catalysts are added to the Phosphoric Acid. Temperature of Reaction 250' Exp. No. I 2 3 4 5 I. CUO 81.8 94.8 89.8 Purity CzH, 98.4 99'5 99.6 2. Fez03 76.3 86.4 91.4 89.5 Purity CzH4 99.2 99.6 99.8 3. CuSO4 93.3 89.5 89.8 87.8 Purity CzH4 99'2 99'7 99'3 99'8 4. Ag?O 92.7 92.9 90'7 Purity C2H4 99'2 99'2 99" 5. AgzSO4 78,4 89.9 93.3 90.7 Purity CzH, 98.2 98.4 99.6 99.2 63.0 76.2 78.2 82.6 80.4 6. FeCl3.6H20 Purity CzH4 96.6 99.3 98.9 99.0 64.1 75.8 78.0 7. FeC13 Purity C2H4 99.0 99.0 8.

nioo3

87.9 99.3

81.0 99.9

82.5

88.7 99.5

83.1 99.5

Turity CzH4 9.

11-03 Purity C2H4

IO.

Al,(SO,), Purity CZHI

77.4 98.8

93.5 98.9

93.2 99.3

11.

NO Purity C2H4 SasPOl Purity CzH4 Silica gel

78.4

82.5 95.8

83.9 96.3

70.8

76.8 99.2 89.7 98.0 89.5 97.3 82.8 99.6

73.9 99.6 85.3 99'3 90.9 98.9

12.

13.

82.6 96.4

14.

Animal charcoal

89.9

15.

KO

68.3 99.5

catalyst

'*Cavalier: Compt. rend., 124, 91 (189;l. 15 Langheld: Ber., 4 4 , 2076 j1911j.

80.9

99.6

87.6 99.2

84.8 99'7

85.9 99.9

81.2 99.1 81.2 99'3 81.0 99.8

Po.0

99 7

83.9 99.6

2700

ARTHUR G . W E B E R WITH JAMES H . WALTOX

An attempt was made to prepare silver ethyl pyro phosphate by adding silver oxide t o the acid which had been treated previously with a slight excess of absolute alcohol. The almost white precipitate was filtered off and washed with alcohol and dried. The impure compound thus obtained gave an uncorrected melting point between 2 j j oand 260' decomposing between 260' and 270'. I t contained 63.2% Ag and 13.2% P. Silver ethyl pyro phosphate contains 48.17~ silver and 13.8yo phosphorus. The impure compound when heated in a test tbbe fused turning gray, brown and then black, giving off a gas that burned with a yellow flame. An odor resembling that of fresh cider was detected and drops of moisture were obserred on the side of the test tube, Definite proof of the existence of a metallic ethyl phosphate in the reaction mixture would still leave a doubt as to whether it is an intermediate compound in the reaction or whether catalysis operates thru a specific surfacp phenomena created by the presence of this new compound in the reaction mixture. Summary I. The preparation of ethylene by the interaction of ethyl alcohol and phosphoric acid has been studied and the yields and purity of the gas determined. 2. Syrupy phosphoric acid previously heated in an evaporating dish at 250" for a t least 4 - i , ) 2 hours produces the highest yields of ethylene. 3. At a temperature of 2 jo" the reaction is continuous and gives a yield of 81% ethylene as long as alcohol is added to a given amount of the previously heated phosphoric acid. 4. The yield is decreased from I O to 20% at a temperature of 240' but is not increased appreciably a t a temperature of 260'. 5 . The addition of meta phosphoric acid to the regular charge of acid decreases the yield approximately 1570 while the addition of phosphorus pentoxide and pyro phosphoric acid decreases the yield only for a short period after which equilibrium is established and the usual yield is obtained. 6. The effect of various catalysts on the yields and purity of the gas was determined. Copper oxide, anhydrous copper sulfate, anhydrous aluminum sulfate, silver sulfate, silver oxide, iron oxide, and animal charcoal increase the yield as much as ten per cent, ferric chloride, molybdenum trioxide, tungsten trioxide, nickel oxide, and silica gel increase the yield by smaller amounts, while tribasic sodium phosphate tends to decrease the normal yield. In each case the yield approaches a maximum and then decreases to normal as the reaction progresses. In most cases the gas contained less than iq impurity. 7 . Evidence for the possible existence of silver ethyl pyro phosphate in the reaction mixture when silver oxide is used as a catalyst is given. 8. The data do not permit a definite conclusion as to the mechanism of the reaction, but suggests that the catalysts supply a slightly different "pivot" for the reaction or render the intermediate complex of alcohol and phosphoric acid less stable. .lfOdiS07t,

TrkCO7lSi!l.