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 CHEMISTRY
636
EXPORTS OF CHEMICALS A N D ALLIED PRODUCTS (Continued) PAINTSPIGMENTSETC.: Copper paints ' for ships' bottoms.. ............... Paints ground in oil.. . . . . . . Zinc white and white lead. , Other paints and colors.. Varnishes. OILS. FATS AND WAXES: Coconut oil.. Linseed oil, rapeseed oil , . Train oil, gallons.. Austria-Hungary
... ................
.............. . . ......... .........
United States. . . . . . . . . . . . Turpentine oil.. . . . . . . . . . . . Vegetable oils, solid. ........ Bone f a t . .
. . . . .. .. ....................
Whale fat, solid.. . . . . . . . . . . SOAPAND CANDL Soap.
.........
Russia, . . . . . . . . Sweden, Stearin candles.
....... ............
1913 Pounds
1916 Pounds
1917 Pounds
527 068 25:850 106 75,422 46
28,984 67,228 676,334 1,453 538
23 931 354:944 1,576
(1)
25 413 7,086: 670 44,804 3,574,378 931,411 56,323 813.826 260,664
.....
53 215 535:975 53,398 175,600 100,335 18,149 23,538 212,680 14,386,000 40,785
..... 5,534
19,484 584 Metric tons
.....
..... .....
2.,119,720 10,211 5,397,964 128 232 3,230 :420 7,582 90,428 878,681 81,552
606 1,003,690
.....
585,708 238 206,323 115,763 78,038 19,758
.....
.....
690,190 660,152 8,201 24 991 23:961 53
191,780
852,849 13,823,400
..... 14,054 3,963 ..... 10,091 .....440
336,368 309,392 14,270 10,988 1,730 Metric tons
1,170 33 789 326 904 Metric tons
.....
QRES:
Bismuth ore and other ores. n.e.s Chrome ore.. Copper ore, Copper and nickel ore.. Iron ore and concentrates. Germany. Netherlands. . . . . . . . . . . United Kingdom. Briquettes. Germany. Sweden.. United Kingdom.. Lead ore.. Manganese ore.. Molybdenum ore.. ......... Nickel ore.. Pyrite (iron and copper). Belgium.. f Germany.. Netherlands. Sweden . . . . . . . . . . . . . . . United Kingdom.. Slag (rich in copper). ..... Belgium.. 1 Not shown separately.
................... .............. ............... ...... .............
...... .............. .............
.............. ..... ................ ........... ............... .... ........... ............ ..........
.....
.............
....20 74
.....
373,071 189,596 90,345 92,827 195,692 74,861 2,512 118,319 35
.. 5
..... 425,876
26,251 40,892 46,773 89,518 138,134 35,035 399
10 2,737 2,175
iij:ios
152,454
..... 35,086
216,896
.....
.....
2 16,896 10 44
..... 2,395 809 ..... 150,960 135,814 ..... 9,956 46,875 ..... ..... 46,875
.....
141
.....201
253,362
212,909
LL
.....
..... 84 511
10' 145 115:488 21,844 47,523
.....
.....
..... 4,105
8,914 101,554 50.770 41.367
Vol.
12,
No. 7
EXPORTS OF CHEMICALS AND ALLIEDPRODUCTS (Concluded) 1913 1916 1917 ARTICLES Pounds Pounds Pounds ORES (Concluded) : Pyrite slag ( ~ o n c l u d e d:) Germany. . . . . . . . . . . . . . . . 2,974 4,031 3,925 Sweden 31,662 43.492 37,307 Rutile. 51 82 1'4 Wolfram. . . . . . . . . . . . . . . . . . .. ..... L Zinc ore.. 285 1 540 PAPER: Pounds . . . . . . . . . . . . . . . . . 410 $ 8i64.2'491 976 947 453 177 497 $13 185,718,270 122 674 Value. . . . . . . . . . . . . . . . . . . $15'601'272 France .................. $7 826 $ 2:338;059 $615 '944 Germany. . . . . . . . . . . . . . . . $542 352 $9:300 United Kingdom., $ 4,9261644 $ 5 009 081 Australia. . . . . . . . . . . . . . . . 618,356 f 1:945:439 British India.. . . . . . . . . . . . 191,406 834.472 1 303 257 China.. $929 370 $106: 101 762,808 South America.. . . . . . . . . $542: 807 f 1,699,120 $ 1,408,956 PULPAND CELLULOSE: Metric tons Metric tons Metric tons Cellulose, dry. . . . . . . . . . . . . . 209.544 216,800 133,564 Argentina. . . . . . . . . . . . . 8,832 1,450 25 Belgium. . . . . . . . . . . . . . . 12,064 ..... France. . . . . . . . . . . . . . . . 22,803 32,839 6,905 Germany. 7,613 Italy.. 3,666 8,111 3,206 Japan. . . . . . . . . . . . . . . . . 5,124 8,923 ..... Mexico, . . . . . . . . . . . . . . . 3,344 51 Netherlands, . . . . . . . . . . 5,604 4,640 1,866 2,929 2 957 158 Spain. ................ Sweden . . . . . . . . . . . . . . . 2,952 3:966 1 260 United Kingdom., ..... 49,175 109,430 98 640 United States.. ........ 77,777 40,331 19,855 Wet .................... 5,540 3,169 1,634. Belgium. .............. 611 ..... 267 1,211 510 3,949 2.891)
................. ................... .................
.......
:
.................
............ ................
.....
..... .....
.....
:
..... .....
-
mlp, dry.. . . . . . . . . . . . Argentina, . . . . . . . . . . . . Spain.. France ................ ............... United Kingdom.. ..... United States.. Wet. Belgium. Denmark.. ........ France.. . . . . . . . . . . Netherlands. . . . . . . Spain, United Kingdom.. United States. VARIOUSPRODUCTS: Glass. Glue ...................... Pitch Tar: . Coal.. Wood. ................... Tar oil.. Vinegar and acetic acid. .... Leather cream. ............
........ ................ .......... ................ ..... .........
....................
..................... ................... ..................
.....
.....
30
14,661 3,585 4,133 1.505 1,608 1,672 481,080 46,992 19,773 95,730 21,163 41 295,210 184
12,962 680 5,559 2,728 1,303 207 455,522
17,319 102 2,786 695 11,195 1,489 250,099
20,416 105,470 14,316 7,707 302,418 4,825
57:035 10.862 3,498 148,861 19,881
1,366,605 453,896 549,718
3,522.987 198,872 606,644
2,415,200 51,194 251.569
6,392,898 106,854 9,057 626 1,433
1,394,732 119,975 20,948 77 109,188
1,064,877 21,605
.....
..... 9 748
.....
.....
9,085
ORIGINAL PAPERS SOME CATALYSTS WHICH PROMOTE REACTION BETWEEN ANILINE AND ETHYL ALCOHOL'
By T. B. Johnson,A. 3. Hill and J. J. Donleavy DEPARTMENT OF CHEMISYRY,YALE UNIVERSITY, NEW HAVEN, CONN. Received February 24, 1920
The facts and experimental data recorded in this publication have been obtained as the result of a preliminary research dealing with the interaction of ethyl alcohol and aniline hydrochloride. Our main object has been t o obtain information leading t o a better knowledge of the experimental conditions governing the formation of diethylaniline by direct alkylation with alcohol. The details of our investigation and the results obtained are discussed in the experimental part of the paper. It has been known for several years that aromatic amines, of which aniline is the prototype, can undergo reactions with aliphatic alcohols if these reagents are heated together a t high temperatures and under presThe previous papers of this series have 1 Researches on Amines, VII. been published in the Journal o f l h e American Chemical Society. Paper VI, J. A m . Chem. Soc., 38 (1916), 2507.
sure. They interact t o give products of two different types, namely, nitrogen substituted amines like monomethyl- and dimethylaniline, which are prepared commercially in large quantities, or true carbon derivatives resulting from substitution of the alkyl group in the aromatic nucleus of the amine. Paraethylaniline is a representative of this second type of c0rnpound.l The molecular changes productive of these different compounds can be brought about by applying the alkylation reaction to a salt of the aromatic amine like aniline hydrochloride,2 and also by heating the free amine with an alcohol in the presence of a strong dehydrating agent like zinc chloride,s or of certain catalysts which have been shown t o promote 1 Hoffmann, Bsr., 6 (1872), 720; 7 (1874), 527; 13, 1729; I S (1882), 1011, 1646, 2895; Pictet and Bund, Ibid., 2I (1889), 1847; Studer, Ann., 211 (1882), 237; Senkowski, Ber., 24 (1891). 2975; Anschutz and Beckerhoff, Ibid., 23 (1895), 407; Willgerodt and Bergdolt, Ann., 327 (1903), 286; Merz and Weith, Bey., 14 (1881), 2346; Calm, Ibid., 15 (1882), 1642; Louis, Ibid., 16 (1883), 105, 116; 17 (1884), 1221; SI (1888), 1159; Romburgh, Rec. Irav. chim., 8 (1884), 392; Sampaio, Ber., 14 (1881), 2172; Benz. Ibid., 15 (1882), 1646. Staedel, B n . , 16 (1883), 30. Willgerodt and Bergdolt, LOC.cdt. f
July,
.
I920
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
such transformations.l The most important factors which regulate the course of the reaction with a given amine a n d alcohol are, apparently, t h e catalyst, and the temperature employed t o bring about the change. While t h e operation of alkylating aniline directly with methyl alcohol t o dimethylaniline has been developed t o a high degree of efficiency, and is applied successfully on a very large scale commercially by heating t h e sulfate or hydrochloride of t h e amine with wood alcohol, t h e application of a similar reaction with amines and t h e higher aliphatic alcohols apparently has not been very successfully applied. I n t h e case of aniline and ethyl alcohol, which is of particular interest t o us a t this time, Walther2 writes as follows: "On the contrary, diethylaniline cannot be prepared with sulfuric acid." On t h e other hand, Staedels claims t o have obtained this tertiary amine in a nearly theoretical yield by heating the hydrobromide or hydriodide of aniline with ethyl alcohol a t 150' C. Schultz and Julius4 state t h a t the same change can be brought about by using aniline hydrochloride for t h e reaction, b u t t h e product is less pure and contains a considerable amount of monoethylaniline. There is apparently a definite temperature for each amine-alcohol mixture, above which it is impossible t o obtain nitrogen alkylation if t h e reactions are applied in acid solution or in the presence of zinc chloride. This temperature reaches its maximum when alkylating with t h e lowest alcohol, namely, methyl alcohol. Application of the reaction a t temperatures above this maximum leads t o the formation of nucleus-substituted amines due t o rearrangement, a n d i t seems t o be true t h a t t h e tendency t o rearrange from nitrogen t o carbon is influenced by t h e size of t h e alkyl group of t h e alcohol. I n other words, the larger t h e alkyl group of t h e alcohol t h e greater the tendency t o rearrange in acid solution, and consequently t h e lower t h e temperature a t which t h e formation of teptiary amines can be accomplished successfully.~ These rearrangements have been t h e subject of much investigation, especially in t h e case of the action of methyl alcohol on aniline and its homologs.6 T h e migrations in t h e majority of cases so far carefully examined are generally brought about by applying t h e reactions a t temperatures above 250' C. A careful review of the major researches on amine alkylation by heating with alcohols reveals the interesting fact t h a t the investigator hitherto has generally proceeded t o apply his reaction a t a high temperature, namely, in the range of 250' t o 300' C., or in some cases a t much higher temperatures. The result has been t h a t the product of his reaction has always been a nucleus-substituted compound, a n d in some cases t h e temperature has been sufficiently high t o actually alter t h e constitution of t h e aliphatic group of t h e alcohol used. It seems also t o have been quite general practice, when heating a n amine directly with a n alco1 2 8
5 6
Knoevenagel, J . grakt. Chem., 89 (1914). 31. Chem.-Zfg.,84, 641. Staedel and Reinhardt, Bey., 16 (1883), 30. "Organic Coloring Matters," p; 25. Niementowski, Bcr., 80 (1897),3071. Hoffmann, Loc. c i f .
63 7
hol, t o introduce zinc chloride t o accelerate the alkylation. This apparently promotes t h e splitting off of water and leads t o t h e same results as are obtained by heating t h e salts of t b e amines with alcohol. I n t h e case of reactions involving t h e use of methyl alcohol, which has received the most attention, t h e experimental d a t a obtained indicate t h a t temperatures below 230' C. are favorable for t h e formation of nitrogen derivatives. I n t h e case of the higher. alcohol homologs t h e optimum temperature for successful nitrogen alkylation in acid solution will undoubtedly be found t o be lower t h a n 230' C. T h e results obtained by Niementowski,l who investigated t h e behavior of aniline hydrochloride towards ethyl formate and ethyl acetate a t 2 2 5 ' C,, are in conformity with this statement. If we confine our attention t o t h e chemistry of aniline i t appears, therefore, t h a t any experimental conditions which favor the formation of alkyl halides as intermediate products of t h e reaction will lead t o the production of nucleus-substituted amines if t h e alkylation reaction is applied a t a higher temperature t h a n 2 5 0 ' C. On t h e other hand, when experimental conditions are adopted which exclude the possibility of intermediate alkyl halide formation, i t is possible t o heat aniline with alcohols a t much higher temperatures t h a n 2 5 0 ' t o 300' C. a n d obtain easily nitrogen derivatives without rearrangement of t h e alkyl group t o the benzene nucleus. For example, Mailhe and de Gordon2 have actually shown t h a t mixtures of methyl alcohol and aniline vapors, when passed over thorium oxide, aluminum oxide or zirconium oxide a t 400' t o joo' C., interact smoothly t o give mono- and dimethylaniline. They state t h a t aluminum oxide is t h e best catalyst of the three oxides. I n other words, under these conditions aniline interacts with alcohol in a manner similar t o t h a t of the corresponding reduced amine, hexaphenylamine, which was investigated by Sabatier a n d Mailhe.3 I n this connection i t is of special interest t o call attention t o t h e interesting observations made b y by Knoevenage14 who has shown t h a t the presence of traces of iodine promotes t h e reaction between aromatic amines and alcohols with formation of nitrogen alkyl derivatives. H e obtained, for example, excellent yields of dimethyl- and diethylaniline by heating aniline with methyl and ethyl alcohol, respectively, in t h e presence of traces of iodine a t 200' t o 230' C. Not only did he apply his reaction successfully with aniline and amyl alcohol under these conditions, but he also showed t h a t a- and &naphthylamines behaved in a manner similar t o aniline. I n other words, no rearrangement of alkyl from nitrogen t o carbon was observed in any case examined. These interesting observations of Mailhe and Knoevenagel suggest t h a t quaternary nitrogen combinations functionate as intermediate products in t h e molecular rearrangements of secondary and tertiary aromatic amines into their isomeric carbon-substituted compounds. I n t h e light of these results of Knoevenagel i t seemed 1 2
a
LOC. cit. Compl. rend.. 166 (1918), 467. I b i d . , 148 (1909), 898; 158 (1911). 621.
1L O C . C i f .
6 38
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
probable t o us t h a t t h e reaction between aniline hydrochloride and ethyl alcohol might be stimulated b y t h e introduction of certain reagents. So far as we are aware, no attention has been paid t o t h e question of catalytic influences in such reactions. A research in this field seemed, therefore, very desirable both from a theoretical and a practical standpoint. T h e successful development of a method of operating which would be productive of diethylaniline in good yield, without recourse t o ethyl bromide or ethyl chloride for alkylation, would be of t h e greatest practical interest. The results of our preliminary work on this problem are recorded in t h e following experimental part. T h e investigation is being continued and in our next paper we shall discuss experimental d a t a obtained by working with t h e higher homologs of aniline.
12,
No. 7
condenser tube, a n d t h e mixture heated for 30 min. t o hydrolyze t h e excess of acetic anhydride. After careful dilution in a graduated flask of convenient capacity a n aliquot p a r t is titrated with standard alkali solution in order t o determine t h e excess of acetic anhydride. A blank experiment run under similar conditions standardizes the latter. T h e difference between t h e two analytical results is a measure of t h e acetylizable impurity in t h e oil. This is t h e n calculated in terms of monoethylaniline. The aromatic bases which theoretically must be considered here together with diethylaniline are unaltered aniline, monoethylaniline, and nucleus-substituted amines such as ethylaniline and its monoethyl derivatives represented by Formulas I and 11, respectively. NHC2Hs N H2
0
EXPERIMENTAL
T h e experiments described in this paper were carried out in an iron autoclave of 1 . 7 liters capacity. This was constructed of heavy iron pipe a n d was furnished with a standard pressure gauge, thermometer well, and a glass or copper inset t o avoid corrosion. T h e charge was heated b y suspension of t h e autoclave in a b a t h of cottonseed oil. The procedure which was followed in all t h e experiments herein described was as follows: After the completion of the heating period t h e autoclave was allowed t o cool and opened. T h e contents were transferred t o a distilling flask and t h e excess of alcohol removed by heating a t 100' C. under diminished pressure. A large part of t h e excess alcohol could t h u s be profitably recovered. T h e residual oil was then made strongly alkaline with sodium hydroxide in order t o liberate t h e aromatic bases from their hydrochlorides, and steam-distilled. T h e amines were separated from t h e aqueous distillate by extraction with ether a n d t h e ether solution dried over fused sodium sulfate, After removal of t h e ether the oil was fractionally distilled a t atmospheric pressure. T h e correct boiling point of monoethylaniline is 204' C. a n d t h a t of diethylaniline is 215.5' C., a t 760 mm. During the preliminary experiments, and before t h e degree of alkylation of t h e aniline was observed t o be notably raised by catalytic influence, a tolerably good idea of the success of any given experiment was gained b y noting t h e range of distillation of t h e crude oil. We found, however, t h a t a boiling point was no criterion of purity when oils containing 2 5 per cent and less of monoethylaniline were distilled. Constant boiling mixtures were always formed and i t was absolutely necessary t o analyze such oils. T h e method which was used for the estimation of primary a n d secondary amines involved t h e use of acetic anhydride as a reagent. I n brief this method is as follows: About one gram of the oil is weighed i n t o a small flask having a ground glass neck. An equal, carefully weighed amount of acetic anhydride is added, a n d t h e flask, after being attached t o a condenser provided with a ground glass connection, is heated for 3 0 min. on t h e water bath. Fifty cc. of distilled water are then carefully added through t h e
Vol.
I
CzHs
I The proportion of these nucleus-substituted bases is increased by heating aniline with ethyl alcohol a t high temperatures. ACTION
OF
ETHYL
ALCOHOL
ON
ANILINE
IN
THE
oxIDE-Theoretically, t h e simplest procedure for t h e ethylation of aniline is t h a t involving t h e interaction of this amine with alcohol. This reaction is typified by t h e equation CsHsNHz 2C2H60H = CsH5N(C2Hs)2 2H2O and theoretically should be facilitated by t h e addition of a reagent capable of combining with water. So far as t h e writers are aware only one investigator has examined t h e influence of a basic oxide like calcium oxide on t h e activity of these two reagents. Calm1 tried t o bring about a reaction between amyl alcohol and aniline b y heating these reagents in t h e presence of calcium oxide and also calcium chloride b u t reports t h a t no change was effected b y their presence. We have repeated this experiment with ethyl alcohol a n d our first set of preliminary experiments was therefore made with this oxide present as t h e dehydrating agent. T h e autoclave charge was as follows: Aniline.. , . . . . . . . . . . . . . . . . 72 g. Alcohol2.. ... . . . . . . . . . . . . . 180 g. Calcium oxide.. . . . . . . . . . . 100 g. Time of heating.. . .. . . . , . . 8 hrs. PRESENCE O F CALCIUM
+
+
Four runs were made a t different temperatures and t h e results are recorded in Table I. Autoclave presTABLE I No. .
1 2
Aniline
G.
Alcohol G. 180 180
180
A
72 72 72 7. 2_
5
72
180
3
Lime G. 100 100 100
180 . ..
{
Temperature Time O C. Hrs. 150 8 200 175
100 220-230 No lime 10 g. K I and 150 2 g. iodine
8 8 8
Product Aniline Aniline Aniline Aniline
8
Aniline
sures were obtained as high as 500 a n d 700 lbs., but in no case was alkylation obtained under t h e conditions Ber., 16 (1882),1642. The alcohol used in all the experiments described in this paper contalned about 1 per cent or less of water. I t was prepared from ordinary 95 per cent alcohol by prolonged digestion with an excess of quicklime, until the desired gravity was attained. 1
2
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
July, 1920
described, I n Expt. 5 potassium iodide and iodine were substituted as catalytic agents in place of calcium oxide and t h e mixture heated a t a temperat u r e of 150' C., but in this case also we obtained no evidence of the formation of alkylated derivatives. ACTION
OF
ETHYL
ALCOHOL
ON
ANILINE
HYDRO-
t h e light of t h e results obtained by previous investigators, who have examined t h e behavior of ethyl alcohol toward aniline oil, we decided t o confine our work t o a study of t h e action of aniline hydrochloride on ethyl alcohol. Here we have a condition favorable for t h e production of ethyl chloride which is known t o react with aniline with formation of alkyl derivatives. Aniline hydrobromide is far more reactive t h a n t h e hydrochloride b u t is too expensive a reagent t o use in manufacturing operations, while alkylation of aniline in t h e presence of sulfuric acid is not practical, due t o t h e tendency of alcohol t o undergo dehydration in t h e presence of this reagent giving ether and ethylene.' While t h e question of alkylation of aniline by interaction of t h e halide salts of this amine with various alcohols has received much attention, no really satisfactory results have been obtained with ethyl alcohol, if one is t o draw conclusion from what has been published in t h e chemical literature, a s only generalized statements are available. The most productive results are those described in a publication by Staedel,2 and which were obtained by application of a process p a t ented by him.3 He states t h a t diethylaniline may be obtained in nearly quantitative yields by heating aniline hydrobromide or t h e corresponding iodide with a slight excess over two molecular proportions of ethyl alcohol a t a temperature of 145' t o 150' C. No description of a systematic series of alkylation experiments is recorded. Concerning t h e use of aniline hydrochloride, only rather generalized statements are available in t h e literature in so far as they pertain t o t h e synthesis of diethylaniline. For example, Schulte a n d Julius4 state t h a t diethylaniline is obtained by heating aniline hydrochloride a n d aniline with rather more t h a n two moles of ethyl alcohol. T h e product contains a considerable quantity of monoethylaniline and in order t o obtain t h e diethyl compound i t is usually treated with sufficient ethyl bromide t o complete t h e reaction. According t o them this base is obtained more readily a n d in nearly theoretical yield by heating aniline hydrobromide with 2 . 2 moles of ethyl alcohol a t 145' t o 1 5 0 ° c. I n respect t o t h e preparation of monoethylaniline these authors state t h a t i t is obtained by heating aniline hydrochloride or a mixture of aniline a n d hydrochloric acid with rather T o r e t h a n one mole of ethyl alcohol at zooo C. Cain6 states t h e conditions for t h e preparation of mono- a n d diethylaniline as follows: For monoethylaniline, 95 parts of aniline hydrochloride a n d 2 8 parts of ethyl alcohol are heated CHLoRIDE-In
Walther, Chem.-Ztg., 54, 641.
cit. D. R. P. 21.241.
LOC. 4 LOC.
6
cit.
"The Manufacture of Intermediate Products for Dyes," p 66.
639
together in an enameled autoclave a t 180' C. for some hours. T h e product contains 70 t o 73 per cent of monoethylaniline which crystallizes o u t as t h e hydrochloride. Diethylaniline may be prepared like monoethylaniline using aniline hydrochloride and ethyl alcohol. The product, however, contains a considerable amount of monoethylaniline. I n our experiments with aniline hydrochloride we first operated a t t h e temperature recommended by Staedel in his work on the corresponding hydrobromide We applied t h e reacand hydriodide, namely, 150'. tion under various conditions, using in some cases zinc chloride and calcium chloride as dehydrating agents, and also incorporating several reagents which we believed would exert a catalytic influence. It was found t h a t t h e nature of t h e product was very much influenced by t h e presence of certain reagents. T h e results of our preliminary experiments are recorded in Table 11. TABLE11-INTERACTION OF ANILINEHYDROCHLORIDE AND ETHYL ALCOHOL 100 g. of aniline tydrochloride used in each experiment Temperature. 150 Tim; of heating, 8 hrs. When ZnClz, CaClz, NaBr or K I were introduced as catalysts, 10 g. were always used Distillate
'
8 1 2 3 4 5 6 7 8 9 10 11
3 .e
ZnClz NaBr iMoles of C%lz Alco- Special 10 0. 10 G. hol Catalyst 2.2 5 .... N'aBr 5 KI 5 CaClz K I 5 ZnClz K I 5 ZnClz 5 ZnClz K I 10 ZnClz K I 5 Copper lining ZnClz K I 5 CoDDer 1Xng ZnClz K I 5 Copper
K"f
N C (
&
" P I N
8 $ 4 42
.... 87 24 24 27 9 . . . . . . . . 87 46 26 8 6 . . . . . . . . . . .... 96 7 38 26 12 2 . . . . . . . . .... 96 6 37 26 14 2 . . . . . . . . .. .. .. .. .. ... .. .. .. .. 75 72 5 . . . . . . . . . . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 70 . .. 73 .. .. ... . . . 14 . . . . . . 92 . . . . . . . . . . 6.5 . . . . . . 93 . . . . . . . . . . 70 . . . . . . .... ..............
..... . . . . ... ... ...
qtrinq
----r-
12 ZnClz K I 13 ZnClz K I
5 5
14 ZnClz K I 15 ZnClz K I
5 5
16 ZnClz K I
5
17 ZnClz K I 18 ZnClz K I 19 ZnClz K I
5
5
3 g. (CeHdzNH 5 g. CuClz 5 g . CuClz 5 g. CuClz 8 g.
KI 20 21 zriC12 22
5 5 5
5 g. CuClz 5 g CuCla 5 9: cuc12
. . . . ... ..
* ' 8 S % 2 2 2 2 Z
3 '
5
5 g. CuO 5 g. CuO 8 g. H C1 5 g. CuCl 5 g. CuCl + x 4.
+
Hci
+__
HC'I --
80
.. .. .. .. .. 42 ....
63 65
3 72
. . . . 88 ...... ...... . . . . 78 . . . . 38 .. .. .. .. .. ..
.... .19. .. .. .17. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . . . . . 10 . . . . . . . . 2 .. 9 3 . . . . . . ..4 .8. .. .. 9.1. .. .. .. .. .. .. .. 168 .. .......... 46 . . . . . . so 11 . . . . . . . . . .
Expt. 1 , 1 Table 11, shows the result of substituting aniline hydrochloride for aniline hydrobromide under the conditions used by Staedel in the case of the latter salt. A theoretical yield of monoethylaniline would have been 92 g. Our product was very largely monoethylaniline. Nine grams of material boiling a t 208' to 210' C. were obtqined. Nothing, however, distilled within the range of diethylaniline. It is rather astonishing, in the light of Staedel's work, that diethylaniline was not more predominant in the reaction mixture. The use of 5 molecules of alcohol (Expt. 2 ) did not give as good a result. This is in some degree contradictory, for in the presence of a catalyst as well as at higher temperatures an increase of alcohol up to a certain limit is decidedly favorable in its effect. 1 Except in special cases, the writers have considered it unnecessary t o record in the tables the results obtained by duplication of their experiments. It should be understood t h a t all experiments have been carefully checked, in particular where abnormalities occur. In Expts. 8, 19, a n d 20 the fractions indicated under column 206°-2080 were collected from 206°-2100, and in Expt. 22 the fraction of 80 g. was collected between 202°-2060.
640
T H E J O U R N A L O F I N D V S T R I A L A N D ENGIiZ'EERING C H E M I S T R Y
With the idea of obtaining a static condition similar to that existing when aniline hydrobromide or iodide is employed, I O g. of sodium bromides (Expt. 3) and a similar amount of potassium iodide (Expt. 4) were introduced into the reaction mixture. An increased total yield was the result. The addition of I O g. of calcium chloride as a dehydrating agent (Expt. 5) was likewise very favorable and slightly better than the former case, the total yield of monoethylaniline being of the order of 80 g. The substitution of zinc chloride (Expt. 6 ) for calcium chloride did not materially affect the result, whereas the omission of the iodide or bromide (Expt. 7) materially decreased alkylation. Increasing the alcohol factor t o I O moles (Expt. 8), together with the use of potassium iodide and zinc chloride, gave decidedly the best result yet obtained. For the first time a distillate boiling above 210' C. was obtained. T o summarize, the results of t h e first eight experiments recorded in Table I1 reveal t h e following facts: I-That aniline hydrochloride when heated with 2 . z molecules of alcohol at I 50' C. gives a product consisting largely of monoethylaniline, in contradistinction t o Staedel's results, in which t h e use of aniline hydrobromide (or iodide) gave a product composed essentially of diethylaniline. 2-That t h e alkali bromides and iodides (NaBr and KI) used in conjunction with t h e dehydrating effect produced by either calcium or zinc chloride, catalyze t o a certain degree t h e ethylation of aniline. 3-That a large excess of alcohol is favorable for ethylation when used in t h e presence of t h e metallic salts. During t h e course of our preliminary experiments it became necessary t o employ a copper inset or lining i n t h e autoclave, where a glass one had been used hitherto. The first two experiments (9 and I O ) carried out with t h e copper lining gave rather good yields of diethylaniline. Since this lining had previously been used for other operations we had good reason t o believe t h a t t h e catalytic effects produced in our reaction might possibly be attributed t o one of t h e following substances:
V O ~1. 2 , NO.7
this being evidenced by the fact that the bulk of the distillate boiled below zooo C. Somewhat more alkylation was observed in Expt. 15 when a small quantity of concentrated hydrochloric acid was added to the reaction mixture employed in Expt. 14. Expt. 16 shows the somewhat more favorable effect of diphenylamine as a catalyst. However, the yield of diethylaniline was insignificant. Expt. 17 demonstrated very clearly the nature of the catalyst in that the addition of 5 g. of cupric chloride gave a most excellent yield of diethylaniline. This experiment was repeated (Expt. 18)and a similar result obtained. It is our conclusion, therefore, from these observations that the results obtained in Expts. g and IO were due to the presence of cupric chloride in the slightly corroded containers used in these operations. The unfavorable effect of moisture, even when cupric chloride was employed, is shown in Expt. 19 where a small amount of concentrated hydrochloric acid was added to the charge. An endeavor was made to obviate the use of either zinc chloride (Expt. 20) or potassium iodide (Expt. 2 1 ) when cupric chloride was used. Neither reagent could be omitted from the charge. However, the experiments did show the greater value of the zinc chloride as a catalyst as compared with the potassium iodide. The omission of both zinc chloride and potassium iodide (Expt. z z ) gave no oil distilling within the range for diethylaniline. The experimental work thus far reviewed established t h e fact t h a t t h e autoclave charge when heated for 8 hrs. at a temperature of 150' C. would give a product consisting largely of diethylaniline. The autoclave pressures recorded in all these experiments carried o u t at 1 5 0 O C. averaged about 190 Ibs. 10 g. of zinc chloride 10 g. of potassium iodide 5 g. of cupric chloride
100 g. of aniline hydrochloride 200 g. of ethyl alcohol (99 per cent)
T o summarize, t h e following substances were found either without appreciable effect, or slightly inhibitive in their action on t h e alkylation of aniline. WITHOUT EFFECT Copper Cupric oxide
INHIBITIVE Moisture-free HC1 Cuprous chloride
POSITIVE Cupric chloride Diphenylamine
T h e next phase of our investigation which received attention was t h e proper temperature for efficient
1-Metallic copper 2-Hydrochloric acid 3-Cuprous chloride +Diphenylamine 5-Cupric chloride
Experiments were therefore designed t o establish t h e identity of t h e catalyst and were carried out in a glass inset; and a basic charge of I O g. potassium iodide, I O g. zinc chloride, I O O g. aniline hydrochloride, and 2 0 0 g. alcohol was used, together with t h e catalytic substance whose effect was sought. The result recorded in Expt. 1 1 is typical of those obtained when strips of clean copper were placed in the charge, or when a thoroughly scoured copper inset was used. No catalytic effect similar to that observed in Expts. g and I O was produced. Expt. 12 shows the effect of the addition of 5 g. of cupric oxide. The larger proportion of the reaction mixture was monoethylaniline. Expt. 13 was similar to I Z except in the fact that a little free acid was added (conc. HCI). The yield of monoethylaniline was increased, thus demonstrating the inhibitive effect of free hydrochloric acid on the final alkylation reaction. A most striking inhibitive action was observed in Expt. 14 when cuprous chloride was employed. No diethylaniline was apparently formed and much unchanged aniline was present;
0
2.5
Grams of Cuprk Chhride
5
FIQ. 1
alkylation. Several experiments were run under t h e best conditions stated above and i t was found t h a t more consistent results, attended with better yields of diethylaniline, could be obtained by working at a constant temperature of 175' t o 180' C. instead of 150' C. The results recorded in Expts. I and 2 of
T H E J O U R N A L O F I N D U S T R I A L A N D ENGINEERI,VG C H E M I S T R Y
J ~ i l y I, Q 2 0
TABLEI11
d
33
Expt.
b
d
4.8
2,
82
L
2 5 5 5 5 5 8 10 14 10 10 10 10 10 10 10 10 10 10 10
B
:-
u
p
175-180 175-180 175-180 175-180 175-180 175-180 175-180 175-1 80 175-1 80 175-180 175-180 175-180 175-180 175-180 175-1 80 175-180 175-180 175-180 175-180 190 220 175-180
8 8
95% Alc. 175-180 957a Alc. 175-180 208-212.
s 8
54
8 8 8 8 8 8 X
4 16 8
$0 'n a 5 5 5 5 5 None 2.59
5 5
Table 111 are representative of the effect of operating under such conditions. Another phase of the investigation was concerned with t h e replacement of zinc chloride and potassium iodide by sodium bromide and calcium chloride (Table 111). I n Expt. 3, t h e substitution of sodium bromide for potassium iodide was productive of results wholly comparable with those obtained when potassium iodide was used, nor was t h e calcium chloride found t o be less efficient t h a n zinc chloride (Expt. 4). I n other words, our results indicated t h a t t h e conditions for reaction and t h e relative proportions of reagents best adapted for t h e preparation of $ethylaniline were t h e following: Calcium chloride.. . . . . . . . . . . . . . . . . . . . 10 g. Cupric chloride.. .................... 5 g. Sodium bromide
...
.. ..
..
I .
83 105 96 103 100
..
..
.. .. .5
.6
5 ,
8 8
.. .. ..
..
47 ,,
1
5 5
.. ..
96 96 99 112
5 5 5 5 5 5 5 5 5 5 5
s 8
E. A,-----
---M.
c ) N
100 100
{;2
DISTILLATE--
4
0
10 KI 10 ZnClz 10 K I 10 ZnClz 2 10 ZnClz 10 NaBr 3 100 10 CaClz 10 NaBr 4 100 10 NaBr 10 CaCIz 5 100 10 NaBr 10 CaCll 6 100 10 NaBr 7 100 10 CaCL 10 CaClz 10 NaBr 8 100 10 CaClz 10 NaBr 9 100 10 CaClz 10 NaBr 10 100 10 CaClz 10 NaBr 11 100 10 CaCIz 1 NaBr 12 100 10 CaClz 25 NaBr 13 100 10 CaClz 5 NaBr 14 100 15 NaBr 10 CaClz 15 100 20 CaClz 10 NaBr 16 100 17 100 5 CaClz 10 NaBr 10 CaCh 10 NaBr 18 100 10 CaCh 10 NaBr 19 100 10 CaClz 10 NaBr 20 100 10 CaClz 10 NaBr 21 100 10 NaBr 10 CaClz Aniline 22 Aniline 10 NaBr 23 100 10 NaBr 24 100 20 CaClz 1 The fraction was collected from 1
ti
96 88
. . . . . . . . .
.. .. ..
.. .. ..
.. .. .. ..
.. .. .. .. '2
-D.
.. .. .. .. 5
3
3
..
1
'3 3 2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 . . . . . . . 24 . . . .
.. .. ..
..
..
. . . . 221 3
. .
. .
. .
. . . . . . . . . . . . . .
E. A,--
8~. X 88 92 93 89 72 95 88 90 94 83.5 77 84 87 84 92.5 87 70 91.6 85 82 11 91 83
..
.. ..
.. .. .. ..
..
..
..
.. .. .. .. ..
..
...
... ... ...
6.79 10.00 7.35 8.10 6.70 5.00 7.60 11.08 11.6
7.7 7.1 10.5 9.4 10.6 6.8 8.3 13.0 M. E. A. 12.24 11.58
because it had promoted such a high degree of alkylation. Four experiments Were carried Out, using alcohol 8s t h e variable factor. Five, eight, ten, and fourteen molecular Proportions of alcohol were employed, corresponding t o 1 6 0 , 2 8 8 , 3 6 0 , and 504 g. of this reagent, respectively. The curve for alcohol (Fig. 2 ) remains practically flat up t o 8 moles; from here it falls sharply t o I O moles, where the monoethylaniline content of t h e oil was only 5 per cent, the best p r o d u c t obtained i n all o u r e x p e r i m e n t s . Striking was t h e result, however, with t h e use of 14'moles, t h e purity of the product falling off t o t h e extent of 7.9 per cent monoethylaniline. This is shown by t h e sharp rise in t h e curve frpm I O t o 14 moles. Z6
8 hrs.
2 6.7 6.8
$
Adopting a unit charge represented above as t h e basis for further work, a series of experiments were conducted i n which t h e factors of concentration, time, and temperature were thoroughly investigated. The amount of cupric chloride productive of the best results was first ascertained, the other ingredients of t h e reaction mixture being kept constant in quantity. From inspection of Table 111, as well as t h e curve developedfrom t h e same (Fig. I ) , it may be seen t h a t t h e effect of introducing cupric chloride is quite apparent even when small quantities are used (0-1 g.), t h e effect becoming gradually less pronounced up t o j g., when a product containing b u t 6.79 per cent of monoethylaniline was obtained. This was t h e purest sample of diethylaniline t h a t we had thus far obtained in our experiments. However, at this point it should be noted t h a t t h e total yield as well as t h e 2 1 2 ' t o 2 1 8 O fraction was falling off in amount, due t o t h e initiation of some decomposition from t h e presence of t h e excess of copper. This amount of copper chloride, however, was deemed the most efficient
3
$
x
6B
Q. 5
'Moleculds df)Alcohol FIG.2
With the following factors constant, sodium bromide was made t h e variable: Aniline hydrochloride.. .................... 100 g. Alcohol .................................. 360 g. Cupric chloride.. ......................... 5 g. Sodium bromide (variable). . . . . . . . . . . . . . . . . 10 g. Calcium chloride.. ........................ 10 g. Temperature. 1 750-180° 8 hrs. Time.. ..................................
.............................
Using b u t one gram of sodium bromide, t h e curve (Fig. 3) commences a t a purity of 1 1 per cent monoethylaniline, rising slightly a t 2 . 5 g. (presumably due t o mechanical errors), then falls steadily with 5 and
4542
T H E J O U R N A L O F 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 Vol.
I O g. of sodium bromide, reaching its lowest point with the latter figure. Further introduction of sodium bromide did not better t h e result, but if anything affected t h e purity, since t h e monoethylaniline con-
12,
No. 7
I n this case, in contradistinction t o t h e 4-hr. period, t h e acetylizable material very probably is a nucleussubstituted derivative of aniline and not unchanged aniline. It was our experience t h a t temperatures below 175' C. made for results not as consistent as those obtained 10.6
c
4C J?
E3. 6.8
9P 5nO x
4 4 4
8
Time in Hours
I6
FlG. 3
FIG.5
tent rose from 5 t o 7.1 per cent. The effect of doubling the normal amount of calcium chloride is evidenced in t h e curve (Fig. 4) by an abrupt rise, denoting an increase in the acetylizable impurities, probably in t h e nature of p-ethyl substituted anilines. Using half t h e normal amount of calcium chloride was also unfavorable since incomplete alkylation was plainly manifest. I n respect t o concentration the following proportions of t h e catalysts were therefore the most efficient in their effect:
when the higher temperature was employed. T h e question as t o whether t h e optimum temperature was above 175" C. was next investigated. The temperature of 190' C. (see Fig. 6) caused a rise in t h e acetylizable impurities. This fact could be attributed only t o rearrangement of the tertiary amine through t h e influence of t h e metallic salts a t the elevated temperature. This effect was even more striking a t 220' C., when t h e impurity calculated as monoethylaniIine rose t o 13 per cent. T o summarize, t h e results obtained in our experimental work have led us t o conclude t h a t sodium
............................... ...............................
Sodium bromide.. Cupric chloride.. Calcium chloride..
..............................
Grams 10
5 10
f3.
10.5 P,
28.3 4b
x 8 9 50 x
s
d
Emperuture Cenfiqrade
FIG.4
Fro. 6
The effect of time on t h e degree of alkylation is best shown in Fig. 5 , where t h e highest point on t h e curve represents the percentage of monoethylaniline when heated for a period of 4 hrs. Incomplete reaction is evident in comparison with 8 hrs., t h e lowest point on t h e curve. Extension of t h e time factor beyond 8 hrs. was not productive of favorable results: in fact, at 16 hrs. t h e per cent of diethylaniline fell off slightly, showing t h e initiation of a reaction presumably typified by t h e expression CGHSN ( C Z H ~--t ) ~ C2Hs.CeHa.NHCzH6.
bromide, cupric chloride and calcium chloride jointly functionate as catalytic agents in t h e alkylation of aniline when t h e hydrochloride of this base is heated with ethyl alcohol. The relative proportions of t h e various reagents and t h e time and temperature factors which are productive of t h e maximum yield of t h e tertiary base-diethylaniline-are given in t h e following table representing a unit autoclave charge. A unit charge run a t a temperature of 175" t o 180' is productive of pressures in the autoclave varying between' 280 and 310 lbs.
July,
1920
T E 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
............... 100 g. 10 moles (360 9 . ) ........................... .................... 10 g.
Aniline hydrochloride.. Alcohol.. Sodium bromide.. Cupric chloride. Calcium chloride., Time ............................... Temperature..
..................... 5 g. ................... 10 g. 8 hrs. ....................... 175°-180e
C.
We have run through a great many experiments under t h e above conditions and they have all been
643
As previously stated, calcium chloride a n d sodium bromide had been utilized in place of zinc chloride and potassium iodide, respectively, purely from t h e standpoint of economical procedure. I n t h e light of this fact i t seemed of interest t o ascertain t h e relative efficiency of these salts. T h e efficiency of zinc chloride as compared with calcium chloride is shown in Fig. 7 (also Table I V ) , where Graph A represents t h e effects produced when t h e usual amounts of the catalysts are used. The second Graph (B) represents t h e effect when t h e amount of calcium chloride is reduced onehalf, a n d also the sodium bromide or potassium iodide. T h e nature of t h e two results should check. At t h e peak of each curve stands zinc chloride a n d sodium bromide with zinc chloride a n d potassium iodide just below it, while a t t h e inception of Curve A stands calcium chloride and sodium bromide. with calcium chloride and potassium iodide slightly above it. I n Graph B t h e conditions are reversed. The results may therefore be summed up in the statements, first, t h a t zinc chloride is much less efficient t h a n calcium chloride in alkylation, and, second, t h a t the iodide is b u t slightly different in its effectiveness as compared with t h e bromide. Placing more emphasis on t h e zinc chloride results, t h e conclusion may be reached t h a t the potassium iodide functions somewhat more favorably t h a n t h e sodium bromide.
48. I Fro. 7
productive of good yields of diethylaniline. The yields have always been very consistent. The total yield is more frequently above (i. e., 104-10; 9.) t h a n below I O O g. Complete alkylation, together with complete extraction of t h e alkylated material, should produce t h e theoretical yield of 1 1 5 g. Therefore, notwithstanding t h e many opportunities for mechanical losses, decomposition through autoclave heat, t o say nothing of t h e purity of the commercial aniline hydrochloride, crude yields of t h e order of 8 7 t o 91 per cent are consistently obtainable. Generally speaking, in t h e distillation of t h e crudes resulting from these best operative conditions, no oil, other t h a n a drop or two, is obtained below 212' C. T h e loss of material in distillation is largely accounted for in the residue left in the flask, no doubt resulting from some decomposition during t h e distillation. T h e product, purified by adistillation, is always a clear, highly refractive oil, possessing a slight yellow color. An a t t e m p t was made t o lower the cost factor in t h e production of diethylaniline by our method, through t h e incorporation of some aniline base in place of t h e hydrochloride, and also through t h e use of g j per cent alcohol. The substitution of aniline oil for a n equivalent amount of t h e hydrochloride in our best unit charge gave a product consisting largely of monoethylaniline (Expt. 2 2 ) . T h e use of 9 5 per cent alcohol (Expt. 2 3 ) instead of t h e product dehydrated over lime lowered t h e purity about 7 per cent; nor could this be prevented by t h e incorporation of double t h e usual amount of calcium chloride as a dehydrating agent (Expt. 24).
12.2
12.0
FIG.8
I n regard t o t h e zinc chloride, i t may be said in explanation of its action, t h a t in all probability it
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
644
Vol.
12,
No.
TABLEIV 100 g. of aniline hydrochloride were used in each experiment. The quantity of ZnCh and CaClz introduced was always 10 g., and the time of heating the autoclave was 8 hrs. In each experiment the temperature was held between 175”-180° C.
d
g
N
,-.
N
Expt.
5
9
10 11 12 13
....
.... .... ....
0 W
1R
c1
8 ZdClz ZnCle ZnCle ZnCln ZnClt ZnClz ZnClz
10 NaBr 10 K I 10 K I 5 NaBr 5 KI 5 NaBr 5 KI
.... ....
....
10 NaBr
10 10 10 10 10 10 2.2 J
10 10 10 10
5 cuc12 5 CUClZ 5 CUClZ 2.5 CuCle 2.5 CuCli 2.5 CuCle 2.5 CuClz
.... .... ....
5 PGjdine
5 Cu Powder
$
101.3 90 100.4 94 96 96.3 80 80 93 103 102 99 93
functions a t a high temperature as a de-alkylating,
as well as a dehydrating agent, and the smaller yield of diethylaniline produced by i t in comparison with calcium chloride may be due t o t h e fact t h a t this salt rearranges some of t h e diethylaniline formed t o a nucleus-substituted amine, such as the monoethylated derivative of p-ethylaniline or similar combinations. Expts. I and 2 of Table I1 are illustrative of the results obtained by heating aniline hydrochloride alone with alcohol a t 1 5 0 ’ C. I n these experiments t h e concentration of the alcohol was 2.2 and 5 moles, respectively. I n neither case was there a n y distillate within t h e range for diethylaniline. Expts. 8, 9, a n d I O of Table IV furnish very interesting comparisons, when t h e reaction temperature of 175’ t o 180’ C. is employed. Expt. 8 gave a product containing b u t 51.9 per cent diethylaniline, yet even this result was distinctly better t h a n t h e corresponding experiment carried on a t I ~ o ’ , while an increase of alcohol t o 5 (Expt. 9) a n d I O moles (Expt. IO) decreased t h e mono‘ethylaniline content t o a marked degree. These results, shown graphically in Fig. 8, are very noteworthy, in t h a t they demonstrate again the extremely favorable effect of t h e concentration of alcohol on t h e ethylation of aniline. The introduction of I O g. of sodium bromide t o t h e most favorable alcohol concentration lowered t h e monoethylaniline content some 2 per cent, while t h e addition of 5 g. of pyridine (Expt. 12) produced a rather marked negative catalysis in t h a t 15.9 per cent of t h e product was monoethylaniline. Although we have been unable t o better t h e results obtained (j per cent monoethylaniline impurity) when t h e three inorganic salts are used, mention should be made of one very interesting experiment (Expt. 13) in which the addition of 5 g. of copper powder a n d sodium bromide t o t h e usual amounts of aniline hydrochloride and alcohol gave a product containing b u t 8.6 per cent monoethylaniline. This result is rather interesting in t h e light of t h e experiment mentioned i n t h e first part of t h e paper, when clean copper strips or a thoroughly cleaned copper inset were apparently without catalytic effect. T h e difference in results may of course be attributed t o t h e state of division of t h e copper.
Distillate-
E. A,--
--M. c
L
.. ..
. . . . .. ....
c -
m
z 0
N
. . .. .. . . .. . .
s
3R
2.5 1’3 0.8 1.6 0.3 6.8
.. 26 . . .. . . .. . . .. 1 .... .. .. .. .. .. * . .. .. . 1. 0:..5 . . . . . . . .
.
I
ii
.
I
D. E. A,----
.. .. .. .. .. .. 16 7 6
.
I
8 1
.. .. .. .. .. .9.
-
2
2
0 N
N
N
N
3; 4
95.8 85 95 89.3 91.8 86 75 3 57 4 55 29 55 39 44 35 24 48
...... .. ... ... 3
2 6 4 9 11
.. .. .. .. .. .. .. .. ..
..
14.6 12.62 5.3 9.5 8.7 16,05 14.39 48.1 12.2 12.0 10.2 15.9 8.6
THE EFFECT OF AIR IN STEAM ON THE COEFFICIENT OF HEAT TRANSMISSION By C. S. Robinson MASSACHUSETTS INSTITUTE O F TECHNOLOGY, CAMBRIDGE,MASS Received March 9, 1920
P-
Recently the writer had occasion t o discuss t h e effect of the presence of norr-condensable gases in condensable vapor on t h e coefficient of heat transmission. A study of t h e d a t a in t h e literature1 showed t h a t t h e information was not in a readily usable form, and t h a t i t was necessary t o refigure it. The writer feels t h a t t h e results of these recalculations will be of interest t o engineers and t h a t a presentation of t h e methods of calculation used may be of assistance t o others who have similar problems t o be solved. It should be emphasized t h a t t h e relations hereinafter derived are approximate, and are based on very incomplete data. Until more information is available, t h e y are, however, t h e best t o be had. T h e flow of heat from one point t o another is inversely proportional t o t h e thermal resistance of t h e material between the two points. Where t h e material in t h e p a t h of flow is not homogeneous, but is made up of successive sections having different resistances, t h e method used in calculating t h e over-all resistance is t h e same as t h a t used in electricity, t h a t is, t h e overall resistance is t h e sum of t h e resistances of t h e successive sections. I n t h e special cases of t h e surface condenser a n d t h e evaporator, the heat flows from t h e condensing vapor t o t h e separating wall, through this wall, and from t h e further side of t h e wall t o t h e cooling fluid, which in t h e first case is t h e condenser water, a n d in t h e second, t h e boiling liquid. T h e coefficient of heat transfer from t h e vapor t o t h e cooling fluid may be calculated, providing t h e resistances of t h e successive sections of, t h e p a t h are known. These sections are: first, t h e film resistance on t h e vapor side; second, t h e resistance of t h e metal; a n d third, t h e film resistance on t h e liquid side. T h e conductivity of metals has been carefully studied, and i t is possible t o obtain values of t h e 1 E. W. Rerr, “Tests upon the Transmission of Heat in Vacuum Evaporators,“ Trans. A . S. M . E , 36 (1913), 731; G A. Orrok, “Air in Surface Condensation,” I b i d . , 34 (1912), 713; J. A. Smith, “kCffect of Air in Feed Water,’’ London Engineering, Oct. 7, 1904.