March 20, 1959
REACTIONS OF
1489
h l I N E RADICALS WITH OLEFINS
trum indicating pure material. Sorcamphane was prepared by hydrogenating a sample of bicyclo [2.2.l]heptadiene kindly supplied by the Shell Chemical Co. Chlorinations were carried out in small sealed tubes using a 5-10-fold excess of hydrocarbon. This excess is important in isomer distribution studies t o eliminate complications due t o polychlorination. Measured amounts of hydrocarbon and solvent were placed in t h e tubes, which were then attached t o a vacuum line and measured quantities of chlorine condensed in from a calibrated receiver. The tubes were finally degassed with liquid nitrogen cooling and sealed. Reactions were accomplished by placing the tubes in a thermostat and irradiating with a 100-watt incandescent light a t approximately 25 cm. until the chlorine color had disappeared. When benzene was used as solvent, some hexachlorocyclohexane sometimes was formed. While this required separation before analysis, its formation does not interfere with our relative reactivity measurements. The molar concentration of solvents employed was determined by the ratio of solvent t o substrate being chlorinated t h e balance of the system being substrate; see Table VI below for actual examples. Analyses were carried out by gas chromatography, using either a Perkin-Elmer Vapor Fractometer or a Wilkens Instrument and Research Inc. Aerograph and whichever of several column substrates gave the most satisfactory resolution. Each reaction mixture was analyzed in duplicate or triplicate, and ratios of peak heights or areas usually agreed t o within 2-3y0. Experimental errors given in Tables I-V represent the spread of independent experiments. I n competitive halogenation of different molecules relative reactivities were determined by change in peak height or area relative t o an internal standard (usually the solvent), and Table VI presents representative experimental results on pairs of cycloparaffins, chlorinated a t 68” in CCla. Cyclohexane and cyclopropane differ t o o much in reactivity for accurate direct comparison. Accordingly cyclopropane was chlorinated with neopentane, relative reactivity us. cyclohexane (determined independently) 0.555. Relative reactivities were calculated from original data via equation 1, assuming only t h a t ratios of peak heights relative t o internal standards before and after reaction were proportional t o the amounts of substrate present. Thus in the first experiment of Table VI1 react. of cyclopentane = log (1.385/1.151) 0.76 react. of cyclohexane log (1.909/0.712) In experiments involving chlorine atom attack a t different points on the same molecule t o yield isomeric products, relative reactivities were determined by the relative amounts of chlorides produced (using a large hydrocarbon/chlorine
ratio t o suppress further chlorination). The chlorination of n-butyl chloride represents the most complicated case, but clean separation was obtained using the “A” column in the Perkin-Elmer instrument a t 115”and 25 p.s.i. helium pressure. Retention times observed were 1,l-dichlorobutane, 5.4 min.; 1,2-, 6.9 min.; 1,3-, 8.8 min.; 1,4-, 16.2 min. Experiments on known mixtures showed t h a t peak areas were proportional t o concentrations of each component. The “9” column was used also for the separation of isomers in the chlorination of ethyl chloride ( a t 50”), 2,3-dimethylbutane ( a t 114’) and for the analysis of the hydrocarbon mixtures in Table I ( a t 114’). Resolution of 1- and 2chlorobutane in the presence of benzene (Table 11) required “-1”and ‘“2” columns in series a t 80”.
TABLE VI1 RESULTSIN CHLORISATIOSOF CucroREPRESESTATIVE PARAFFISS AT 68” IS CCl? Compound
CsHin CsHi? CCla
C4H8 CjHio CC1, CaH6 Seo-CsHlz
Retention time
M1.
1.1 1.1 2.0 0.2 1.9 2.0 1.1 1.1 2.0 1.1
2.5” 53“ 6.5“ 1.4 3.2 8.8
Rob
RC
1 . 3 8 5 i 0.036 1.151 rtto.031 ,909 i ,051 ,712 i ,027 ,0965 zt ,007 . i 1 5 i ,038
0 . 3 6 8 i ,002 . 2 0 8 r t ,003
0.9 ,477 zk ,009 .303 i ,012 1 . 0 1 . i 5 2 i ,012 ,0915 i .003 cc14 8.8 n-CaHio 0.9 1 . 4 2 i ,013 .886 rt ,003 CjHio 1.1 3 . 2 ,734 f ,013 ,421 i ,004 CCla 2.0 8.8 At 68’. Ratio of peak height t o CC14before reaction. Ratio of peak height t o CCl, after reaction. Experimental errors here and in R, are the spread of 3 analyses on the same reaction mixture. Using 0.3 cc. of Clz, analysis with Perkin-Elmer Fractometer, 2 meter “A” column, 25 p.s.i. helium Pressure a t 50” unless noted.
*
The major product from the chlorination of norcamphane was separated and purified on an “.4” column a t 142O, nZ5D 1.4837, lit. 1.482415 (other reported values range up t o 1.4849). I t rapidly liberated chloride ion on treatment with alcoholic ilgNOa. NEW YORK27, N. 1’.
[CONTRIBUTION NO. 499 FROM THE CENTRAL RESEARCH DEPARTMEST, EXPERIMESTAL S T A T I O S , E . 1. DU PONT DE XEMOURS AND CO.]
VIII. Reactions of Amino Radicals with
Syntheses by Free-radical Reactions.
Olefins BY C. J. ALBISETTI, D. D. COFFMAN, F. W. HOOVER, E. L.
JENNEK AND
W. E. MOCHEL
RECEIVEDJULY 28, 1958 A new synthesis uf diainincs Iins becn found ill the additive dimerization of dienes with atnillo radicals generated from hydroxylaniine by reducing ions. Thus, 1,8-diainino-2,6-octadiene was obtained from butadiene, hydroxylamine and a titanium(II1) salt. I n a similar reaction, 2-butene has given 3,4-dimethyl-2,5-hexanedianiine. Use of ethylene has given products containing from one t o several ethylene units per molecule. With titanium(II1) as reducing agent these products are monoamines and diamines, whereas with vanadium(II1) the chief products are aminoalcohols.
Amino radicals have been shown to be intermediates in the reaction of hydroxylamine with a titanous salt,‘ and radicals produced in this way NHsOH+
+ Ti3+
----f
NH2.
+ HzO + Ti4+
have been used to initiate vinyl polymerization.2
The reaction of amino radicals with benzene has been reported to yield an unstable intermediate which decomposed to ammonia, aniline and biphenyL3 Toluene gave similar results, and cyclohexene formed cyclohexylamine. It has now been found that amino radicals generated from hydroxylamine undergo additive dimer-
(1) P. Davis, M. G. Evans and W. C. E. Higginson, J . Chem. Soc., 2563 (1951). (2) E. Howard, U. S. Patents 2,693,140, July G, 19.54, and 2,587,109, (3) H. Seaman, P. J. Taylor and W. A. Waters, i b i d . , 4690 Sept. 4, 1051. (1954).
1490
C. J. ALBISETTI,D. D. COFFMAN, F. LV. HOOVER, E. L.
JENNER AND
\V.E. ~ J O C N E L V O ~81 .
ization4 with olefins to yield diamines containing two units of the unsaturated compound. The additive dimerization can be represented as
monia was the principal competitive reaction. Interestingly, 1-hexene also reacted with amino radicals forming a 12-carbon diamine, presuniably 2,3-dibutyl-l,4-butanediamine, as the main prodS I - I Z ~ ' ~ ' M +XH2-M. uct. 2NHz--M. +XHZ--R.I--M--NH2 When ethylene was treated with hydroxylamine in which M represents a unit of composition cor- and vanadium(II1) salts, the principal products responding to t h a t of the diene or nionoolefin were alkanolamines. For example, in one experiemployed. It is conveniently carried out a t 0 to ment carried out a t room temperature in water, 30' by introducing a hydroxylamine salt and a the following amines were obtained in the indicated solution of a titanium(II1) or vanadium (111) salt yields based on vanadium sulfate : ethylamine in stoichiometric amount into a well-stirred soluYO), butylamine (l5%), ethanolamine (38%), tion of the unsaturated compound in acidic aqueous &hydroxybutylamine ( 18y0),and a small amount methanol. (4yo) of unidentified higher-boiling amines. I t The mixture of additive dimers obtained from seems likely the hydroxy group is formed by oxidabutadiene and amino radicals was principally 1,8- tion of the aminoalkyl radical with vanadiuni(1V) diamin0-2,G-octadiene~as shown by the formation sulfate. Since titanium(1V) salts are weaker of octamethylenediamine in a t least 70% yield SHaOH+ V 3 + + .NH, + \ - * + HzO upon hydrogenation. The octamethylenediamine MCHL=CHZ ---+ .(CHzCHz)nXH2 *SHz formed was identified by its physical properties and by a comparison of its dibenzamide derivative .(CHzCH2)nNHz V 4 + HzO + HO(CHzCHa)nSHz H+ \T3+ with an authentic sample. Small quantities of branched-chain unsaturated 8-carbon diamines oxidizing agents than vanadium(1V) salts, they presumed to have been formed by the incorporation less readily effect oxidation of aminoalkyl radicals. of butadiene by 1,2-addition rather than by 1,4The reaction of benzene and toluene with amino addition and also higher-boiling amines were formed radicals gave unstable amines t h a t appeared to be in low yield. Similar syntheses using isoprene, identical with those reported recently. However, cyclopentadiene, cyclohexadiene and methylpenta- by using a reaction mixture of acetic acid or fordiene also gave diamines as additive dimers but in mic acid and water instead of water alone the lower yields. I n a limited study, methylamino yields of these products were markedly improved radicals were found to undergo additive dimeriza- over those previously reported. The initial tion with butadiene. amines from benzene and toluene decomposed Terminally unsaturated olefins such as ethylene, when heated, yielding ammonia, biphenyl and propylene and isobutylene reacted with amino presumably bitolyl, respectively. Aniline and p radicals generated from titanium(II1) salts and toluidine were also obtained in small amounts. hydroxylamine to form complex mixtures of monoThe ultraviolet spectrum of the unstable amine and diamines. Thus, ethylamine, butylamine, from benzene indicated the absence of conjugated hexylamine, octylamine and octamethylenediamine double bonds. Hydrogenation a t low pressure have been identified among the products obtained using Adams catalyst yielded bicyclohexylamine from ethylene in various experiments. Butyl- and a diamine whose neutral equivalent and amine was the product always formed in highest elemental analyses corresponded to a diaminoyield. The formation of monoamines involves bicyclohexyl. the reduction of an intermediate free radical, The observed stoichiometry of the Ti3+-h'HzOH probably by a Ti3+ion reaction has indicated that reactions of amino radicals with solvents and halide ions may occur. SH2CH2CH2+ . Ti3+----f ?;H2CH2CH2- Ti4+ If excess titaniuiii(II1) is treated with hydroxylN H ~ C H ~ C H S - H++K H ~ C H Y C H ~ anline in the absence of other reactants, reduction Amino radicals added preferentially to the to ammonia occurs as terminal methylene group of an olefin as shown b y S H 3 0 H + + Ti3++ SHz. + H20 + Tiar (1) the formation of propylamine and isobutylamine H + + S H 2 . + Ti3f +KHa + Ti4+ (2) from propylene and isobutylene, respectively. T h e (or S H 3 + . + Ti3+ --+NH3 + Ti&+) main product from propylene was not propylamine but was a mixture of 6-carbon monoamines, with This sequence requires two moles of Ti3+ for each 2-methylpentylamine as probably the main com- mole of hydroxylamine. In the presence of butadiene, the ainino radical ponent. The many other amines obtained from these olefins have not been completely character- may add to the diene before reduction occurs SO that step 2 is replaced by step 3. By this reaction ized. I n contrast to the reactions of terminal olefins, SHz. CaHs ----f SHLC~HP,.+dimer (3) "butene reacted with amino radicals to give as the path, one mole of Ti3f is consumed for each mole principal product 3,4-dimethyl-2,5-hexanediamine The observed stoichiometry and 1-methylpropylamine in yields of 18 and 15%, of hydroxylamine. respectively, based on the hydroxylamine con- in the presence of butadiene has been approxisumed. The reduction of hydroxylamine to am- mately 1.05 moles of Ti3+ consumed per mole of hydroxylamine. This corresponds to 95y0 effic(4) D. D. Coffman and E. L. Jenner, THISJOURNAL, 81, 2872 (1958). iency of step 3 in superseding step 2 . (5) It is believed that the actual species in the highly acidic media h number of solvents, e.g., methanol, acetone. employed in this Kork are N H r + radicals. For simplicity they are tfiiiieihylforiiiaiiii~l~~ ctliyl alcuhol a ~ isopropyl ~ l rclcrred to as NHi2. in this payer.
+
+
+
+
+
+
+
+
+
+
+
REACTIONS OF AMINE RADICALS WITH OLEFINS
March 20, 1959
alcohol, react with amino radicals by hydrogen extraction (equation 4) to yield radicals which are oxidized (step 5a or 5b) and thus furnish a chain reaction. When these processes occur, the Ti3+/
+ R H +R . + S H a R . + Ti4'R + + Ti3+ (followed by: R + + H,O +ROH + H + ) R. + XHgOH +ROH + S H ? . SHg.
(4) (5a)
(5b)
NHzOH ratio falls below two (this is in systems containing no olefin), since in step 5a titanium(II1) is regenerated and in 5b amino radicals are formed directly without the consumption of titanium(II1). Methanol is an especially reactive solvent. As shown in Table 11, the Ti3+/NH20H ratio was only 0.9 in an aqueous methanol solution as compared to 2.0 in water alone. It was found that the methanol was oxidized to formaldehyde in about 2770 yield as determined b y sulfite titration. t Butyl alcohol was less reactive than methanol, and acetic acid was inert to amino radicals. The fact that methanol can be used as solvent for reactions of amino radicals with olefins indicates that the rate of the abstraction reaction 4 is much slower than the rate of addition of amino radicals to double bonds ( 3 ) . If a halide ion is added to the reaction system containing a diene, three new reactions become possible
+ S H , . + 9 - +S H 3 + X . 9. + Ti3+ --+ T i 4 + + S X . + CaHe +XCIHe. --+dinier H-
(6)
(7) (8)
If steps G and 7 occur, this is revealed by an increase in the Ti3+/NH20H ratio. If 6 and S take place, there is no change in the stoichiometry, but the product is a halide rather than an amine. When the synthesis of diaminooctadiene was attempted in the presence of chloride ion, no interference by the latter was observed. The Ti3+/NHaOH ratio held a t 1.05, and the product was the diamine. Bromide ion, however, was extremely reactive with amino radicals, and reaction G superseded formation of diamine (reaction 3 ) . Both reactions 7 and 8 occurred. The stoichiometry changed to a Ti3+/NH20H ratio of 1.G5 (rather than 1.05), demonstrating the occurrence of step 7. This ratio indicates that the sequence of steps 6 and 7 accounts for 60-651?~ of the amino radicals. illso, step 8 was indicated by the fact that the additive dimer formed was the bromide rather than the diamine. Thus, there are two independent pieces of evidence pointing to the occurrence of step G. Experimental I.
Materials.-The aqueous titaniuin trichloride eniployed was purchased from the Fisher Scientific Co. as 1520% solutions containing a considerable quantity of zinc salts. Vanadium(II1) sulfate solutions were prepared by catalytic (Xdams catalyst) hydrogenation of vanadyl sulfate solutions a t 2 5 3 0 ' and atmospheric pressure. The precise concentration of both solutions was determined by titration with a standard ferric alum or ferric chloride solution using amrnunium thiocyanate as indicatur. The uther chemicals used were cu"ercia1 products.
1491
11. Reaction of Amino Radicals with Butadiene.6-To a well-stirred mixture of 104.3 g. (1.5 moles) of hydroxylamine hydrochloride, 81 g. (1.5 moles) of butadiene, 150 g. of sulfuric acid, 600 ml. of methanol and 300 ml. of water were added over a period of 10 minutes a t -7", 81 g . (1.5 moles) of butadiene and 500 ml. (0.75 mole) of titanium(111) chloride solution. The reaction mixture was allowed t o warm t o room temperature after the additions had been completed, and alkali was added in sufficient quantity to precipitate titanium hydroxide. The titanium hydroxide was removed by filtration, and the filtrate was concentrated. Salts that precipitated were removed by filtration, and the filtrate was made strongly alkaline with potassium hydroxide. The amine layer was separated and distilled. There was obtained 57 g . (42YGyield based on titanous(II1) chloride) of isomeric diaminooctadiene fractions distilling in the range 70-110' (4.7 m m . ) . Anal. Calcd. for CSHIZh7g:neut. equiv., 70.1. Found: neut. equiv., 70.6. The residue was a dark brown rubbery resin which contained 10.870 nitrogen. The above diamine (46 g.) was hydrogenated using dioxane as a solvent with a palladium-on-carbon catalyst. The solution was concentrated, and the product was distilled through a spinning band column a t a 2 5 : l reflux ratio t o obtain eight fractions distilling from 86 to 93" a t 3.7 mm. All of the fractions had neutral equivalents between 71.2 and 72.8 (calculated for a saturated 8-carbon diamine, 72.1). Fractions 2, 3 and 4 were partly solid at room temperature, and fractions 5 to 8 solidified completely a t room temperature. The total amount of the solid fractions was 31.9 g., m.p. 46-50". A dibenzamide derivative prepared from the A mixture of this disolid fractions melted at 174-175'. benzamide with the dibenzamide prepared from an authentic sample of octamethylenediamine melted a t 174-175'. The liquid fractions amounted to 8.3 g. and contained some octamethylenediamine as indicated by the fact that the dibenzamide derived from them melted at 170-173' after two crystallizations. Thus, a t least 709; of the diamine obtained from butadiene was the straight-chain isomer. In another experiment, vanadium(II1) sulfate was used as the reducing agent. T o a mixture of 1100 ml. of acetic acid, 200 ml. of water, ii g . of hydroxylamine hpdrochloride and 27 ml. of concentrated sulfuric acid was added a t 25' 480 ml. of 1.4 molar vanadium(II1) sulfate with vigorous stirring over a period of 49 minutes. Concurrently, butadine (gas) was added at the rate of two liters per minute. The amines obtained were isolated by concentrating under reduced pressure, adding excess 40y0 sodium hydroxide, and extracting with three 1.3-lb. portions of ether and two 0.75liter portions of methylene chloride. Distilhtion through a Vigreux column gave the following fractions: (1) 20.2 g. b.p. 86-100" (3 nim.), neut. equiv., 75.0; ( 2 ) 6.5 g., b.p. 100-140° (3 mm.), neut. equiv., 97.5; and (3) 8.6 g. of residue, neut. equiv., 82.7. Redistillation of fraction 1 through a precision still gave eight cuts with refractive indices a t 26' ranging from 1.1931 to 1.4995. The neutral equivalent of the first six cuts (8870 of total) were 71.8 i 0.9 (calculated for CaHlsS?, 70.1). The infrared spectra of the six cuts indicated that the proportion of straight-chain diamines was greatest in the high-boiling cuts. Redistillation of fraction 2 gave four cuts with refractive indices ranging from 1.4989 t o 1.5019. The neutral equivalents of these cuts were 92.0, 98.0, 97.5 and 98.3, respectively. Anal. of Cut 3 . Calcd. for CWHWS.): C . 74.17: H . .- -_ 11.40; S , 11.43; neut. equiv., 97.0. Foui;d: C, f4.22; H, 11.57; N,11.71; neut. equiv., 97.5. These data indicate that fraction 2 was principally a iiiixture of lz-carbon diamines. I n this experiment the yield of 8-carbon diamines \vas about 40%; 12-carbon diamines, 15'; ; higher-boiling amines, 18%; and ammonia, 27';%, based on the amount of vanadium(II1) sulfate consumed. 111. Reaction of Amino Radicals with Cyclohexadiene .T o a vigorously stirred mixture of 600 ml. of methanol, 300 nil. of water, 91 g . (1.3 moles) of hydroxylamine hydro.__~~.____ (6) D. D. Coffman a n d E. I,. Jenner, Canadian P a t e n t 542,1li2, June 11, 1957. Other workers h a v e also explored this reaction-Belgian P a t e n t 657,iLiO issued t u Xatiuual Distillers P r u d u c t s Curg., Alay 2 2 ,
lY57.
TABLEI FRACTIONS OBTAINED FROM ETHYLESE \YITH TiClj A N D SHzOH
-
"C.
68-76 76-98 98 OX 88-112
112-118 118-111 1 11-90
Residue
~~~
-
him.
19 19
19 19 11 I1 11-4 4-1.5
Seyt.
\Vi , :'.
equiv
1.10
118 104
1.13 0 , 50
0.25 1.2 1.5
0.5 1 .:3 7.6
75 81 91 84 98 100
-C
.
64.:3:3 64.35 65.05 66.17 68,89 67.98 69.86 71 .53
chloride and 80 g. ( 1 IlloleJ of cycloiiexadiene iilailitairiet~a t 25' was added over a period of 9 minutes 335 ml. (0.5 mole) of 1.5 A; titanium(II1) chloride solution. Isolation of the diamine in the usual manner gave 5.9 g. (12.35c) of a dialninobicyclohexene fraction, b.p. 131-133' (2 m m . ) . AnciZ. Calcd. for CI?HZ"S!: S, 14.56; C , 74.95; H , 10.48; neut. cquiv., 97.1. F w n d : S,14.45; C, 74.76; H , 10.72;neut. equiv., 96.2. Higher-boiling amines were also obtained. IV. Reaction of Amino Radicals with Isoprene.-The reduction of hydroxl-lamine hydrochloride with titaniuni(111) chloride in the presence of excess isoprene was carried out under conditions similar to those used for butadiene. There were obtained several fractions (217; yield based on TiCh), b . p . 93-110' (1 mm.) with neutral equivalents varying from 83.1 t o 86.4 (calculated for Clo-diamine, 84.0). V. Reaction of Amino Radicals with Cyclopentadiene .This reaction was carried out in a manner similar t o that described for cyclohexadiene. There was obtained d . 8 g. (21 5 0 ; ) of a diaminobicyclopenteiie fraction, b.p. 96-98' (1.1 mni.). zlnul. Calcd. for CIOHIGXz: C, 73.12; H , 9.8%; S, 17.06; neut. equiv., 82.1. Found: C , 72.57; H, 10.22; S,17.63; neut. equiv., 82.3. Higher-boiling amines were also obtained. -1fraction, b.p. 140-145" (1.1 mm.), contained 1.2.7t;i nitrogen and was apparently a product containing more than one cyclopentene unit for each amino group. VI. Reaction of Amino Radicals with Ethylene.-Reactiiins of ethylene were carried out in a 1.2-liter stainless steel autoclave. T h e products obtained depended upon the reaction conditiuns, solvent, and the nature of t h e reducing agent. Reducing Agent, TiCl:,; Solvent, Acetic Acid.-To a mixture of l l O U 1x11. of acetic acid, 200 ml. of water, 77 g . of hydroxylamine hydrochloride, 42 ml. of concentrated hydrochloric acid and 250 g. of ethylene was added 690 ml. (1.(l mole) of titanium(II1) chloride over :I period of one hour a t about 2 5 " . There were obtained trace amounts of ethylarnine, butylamine (2.(J g., 2.7"; yield), hexj-lamine (1.5 g., 1.5(,; yield) and some higher-boiling amines not completely characterized. The amines were identified by their physical properties : ~ n dby mixed melting point cieternlitiations of their pheuyl isothiocyanate derivatives with autlientic specimens-. this exReducing Agent, TiCl:, : Solvent. Methanol.-In pcritriciit t i l e k c t i o n of'ettiylciic a n c I amino radicals was c;irrictl o u t esseritiwlly :is giveii i i i tile Iireviiius experiment esccspt that inethanol was used as the solvetit instead of acetic acid. Oii,distillatioti tliere were obtained 1.4 g. ( 1 . 9 ( ~ ; , ) o f I~utylaiiiine a n d 0.5 g. ( o . S ( i ) c i f hexylamine. 1 liglier bciiliiig fractions also were obtaiiietl (see Table I ) . *. I lie elcmerital malyses anti neutral equivalents of fractions 1 arid 2 indicated t h a t these fractions contained ethanol:tinine and octylaniine. These fractions were combined and mixed with water. The water-insoluble fraction was separated and distilfed, b.p. 178-18(J0 (boiling point of octylamine, 180"). A n a l . Calcd. for CSHlgS: N,10.84. Found: S,10.74. T h e identification of this material as octylamine was confirmed by comparing its infrared spectrum with t h a t of an authentic sample. I'ractioiis :I-~t! obvioudy cotitaiiied relatively high proOct~~iiiethJ-leuctli:iiiiiie was showii til I i i i i . t i o i i i ( i f tliutiiiiic\.
-Analyses,
H
I ;
IS .kQUEOUS ~ ~ E T H A S O L --
~
-.
S
1:3.59
14.78
1X.:36
15 811 1 9 . 12 19 58 15.88 17 48 14.32 14.63
13 97 1:3.8;3 14 0 13.86 13.91 14.02
Liquid a t (J' Liquid a t ( J o Solid a t 0' ; seiriisolicl, 83" Solid a t 0'; semisolid, 25' Solid a t 25" Solid a t 25" Solid a t "3" Solid a t 25"
be a cumponent of fractirins 3-8 by paper chrr~rnatography analysis using the method of Brernner and Kenten.' Through the use of paper chromatography, i t has been shown t h a t ethylarnine, butylamine, hexylamine, octylamine, ethylenediamine, tetramethylenediamine, liexamethylenediamine and octametliylenediamine are products obtained from ethylene and amino radicals. Reducing Agent, Vanadium(II1 j Sulfate ; Solvent, Water.--A vanadium(II1) sulfate solution (1.1 liters, 1.0 mole) was added during the course (Jf 60 minutes t o a tnixture of 1.67 moles of hgdroxJ-lamine hydrochloride, 8.9 moles of ethylene, 1.0 mole (if hydrochloric acid and 1300 ml. of water. There was obtained ethylamine (5',';), butylamine (l55;), etlianolainine (38', h ) , 4-hydroxybutylamine (18(;) and about 4': ( i f higher boiling amines. Ethylamine and butylamine were identified by their boiling points and by paper chromatographic analysis. Ethanolamine n-as identified by its boiling point, neutral equivalent and infrared spectrum. -$-Hydroxybutylamine was ideiitified by its boiling puint (113" (24 m m . ) or 2U4" (atrn.)'", neutral equivalent arid cieniental analysis. I t s infrared spectrum is also consistent wit11 this structure. d n d . Calcd. for C,HIIOS: C , 53.89; H , 12.44; 0 , 17.95; S, 15.72; neut. equi , 89.1. Found: C , 53.84; H, 12.48; 0 , 1i.93(Y; N , 15.8 ; neut. equiv., 88.5. VII. Reaction of Amino Radicals with 2-Butene.-To a mixture of 1100 m l . of methanol, 200 ml. of water, 77 g. (1.1 moles) of hydroxylamine hydrochloride, 42 rnl. of concentrated hydrochloric acid and 231 g. of trans-2-butene in an autoclave was added 590 nil. (1.0 mole) of (111) chloride solution over a period of one hour a On distillation there irere obtained 6.2 g., (8.5 rnethylpropylamine, b . p . W o , neut. equlv., n p 5 D 1.3976 ias compared with literature values of 6 Y o , 73.1 and 1.3900, respectively) and 8.9 g . (12.4;s) of 1,2,:3,4tctramethyltetramethylenedianline ( b .p. 63 . a (5 IIIII~.)), ?L~;D1.1613. A d . Calcd. fur CxH2t,S1: C , 66.60; H , 13.97; S , 19.42; neut. equiv., 72.1, Foulid: C , 66.78; H, 14.03; S , 19.12; neut. equiv., 74.2. The proton magnetic resoiiuicc spectrum o f the diaiiiiiie was consistent with the H2N(CHCH:,)ISH?structure. Similar products were obtailied with cis-2-butene. VIII. Reaction of Amino Radicals with Propylene. -'Hie reaction of propylene and :imino radicals was carried I I U ~i n ;in autoclave a t room temperature in aqueous acetic acid a 5 solvent. There \vas ohtained aliprciximately %l',/Cdistillable amines and 4.X',o iion-distillable basic residue based ou the titaniutn(II1) cliloritie used. .\ sinall aniount (O.(j',i of propylamine, itleiitified 1)y its boiliiig point 4(i-48' :tl1(1 infrared spectrum, was recovered frrirn the ether extracts. On clistillatioii a fractioii (b.1~.113--121L", neut. equiv. 100) \vas obtained (yield .7' 1) hich was probably riiairlly 2rxiethylpent~-lamirle (calcd. neut. equiv. 101, reportetl'o b.p. 64" (90 tnm. 1 (extrapolated 120')J. Higher-boiliiig fractions (83' ( 2 0 inm.)-6(l0 ( 1 mIn.)) were also obtained ( 1-LC
